Inhibition of MDM2 via Nutlin-3A: A Potential Therapeutic Approach for Pleural Mesotheliomas with MDM2-Induced Inactivation of Wild-Type P53

Previously, our group demonstrated that nuclear expression of E3 ubiquitin ligase (MDM2) in malignant pleural mesothelioma (MPM) is significantly associated with decreased overall survival. A possible explanation may be that overexpression of MDM2 leads to a proteasomal degradation of TP53 that eventually results in a loss of TP53-induced apoptosis and senescence. It is well known from other tumor entities that restoration of TP53 activity, e.g., by MDM2 inhibition, results in an instant TP53-induced stress and/or DNA damage response of cancer cells. Nutlin-3A (a cis-imidazoline analogue) has been described as a potent and selective MDM2 inhibitor preventing MDM2-TP53-interaction by specific binding to the hydrophobic TP53-binding pocket of MDM2. In the present study, the effects of MDM2 inhibition in MPM via Nutlin-3A and standard platinum based chemotherapeutic agents were comparatively tested in three MPM cell lines (NCI-H2052, MSTO-211H, and NCI-H2452) showing different expression profiles of TP53, MDM2, and its physiological inhibitor of MDM2—P14/ARF. Our in vitro experiments on MPM cell lines revealed that Nutlin-3A in combination with cisplatin resulted in up to 9.75 times higher induction of senescence (p=0.0050) and up to 5 times higher apoptosis rate (p=0.0067) compared to the commonly applied cisplatin and pemetrexed regimens. Thus Nutlin-3A, a potent inhibitor of MDM2, is associated with a significant induction of senescence and apoptosis in MPM cell lines, making Nutlin-3A a promising substance for a targeted therapy in the subgroup of MPM showing MDM2 overexpression.

Several studies have shown the efficacy of the evaluation of intratumoral expression of members of the folic acid metabolism for prediction of multitargeted antifolate therapy response in patients with different cancer entities but are discussed controversially [10,[13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28]. As platin-analoga are genotoxic compounds that induce DNA damage [29] leading to TP53 induced cell cycle arrest and apoptosis [30], it is basically conceivable that the DNA repair mechanism might be one of the keys associated with an impaired therapy response. As the identification of molecular properties shared by MPMs may help to overcome the poor treatment response observed, several studies addressed this question [11,12,27,[31][32][33][34]. However, the reasons for the rather poor efficacy of platinum compounds remain largely unknown.
Summing up, neither reliable predictive biomarkers nor individualized therapeutic concepts for MPM exist until now. Therefore, current guidelines emphasize the need of innovative and novel therapies [35].
A restoration of TP53 activity, e.g., by MDM2 inhibition, might result in an instant TP53 induced stress and/or DNA damage response of cancer cells. Nutlin-3A (a cis-imidazoline analogue) is a potent and selective MDM2 inhibitor with an IC 50 value of 90nM [54] and prevents MDM2-TP53interaction by binding to the hydrophobic TP53-binding pocket of MDM2 [55].
Thus, the aim of this study was to test the effect of MDM2 inhibition in MPM via Nutlin-3A in comparison to the contemporary common chemotherapeutic strategies using three cell lines showing different marker profiles concerning TP53-status, P14/ARF-and MDM2 expression level.
Human MPM cell lines were obtained from the American Type Culture Collection in 2012-08 (Manassas, VA, USA). The cell lines were authenticated and tested for contaminations by using a commercial service (Multiplexion, Heidelberg, Germany) and were last retested directly after the experiments were finished.
NCI-H2052, NCI-H2452, and MSTO-211H were cultured in Roswell Park Memorial Institute (RPMI) medium (Invitrogen, CA, USA) containing 10% fetal bovine serum (Invitrogen) at 37 ∘ C in a 5% CO 2 -humidified atmosphere. Cells were grown until 85% to 95% confluency, then washed with phosphate-buffered saline (Invitrogen), and trypsinized with 1 ml of 0.05% trypsin-0.53 mM ethylenediaminetetraacetic acid, phenol red (Invitrogen). Trypsinization was stopped by adding fresh medium to the reaction. Approximately 10 l was transferred to a hemocytometer (BRAND, Wertheim, Germany) for cell counting purposes. 1,000 cells per well (100 l) were seeded into microplates 96/U (Eppendorf, Hamburg, Germany) suitable for luminescence and fluorescence detection. The cells were allowed to attach overnight at 37 ∘ C and 5% CO 2 . At the next day, the medium was removed and fresh medium containing either one of the cytostatics or without additive was applied to each well.  Table 1. Cell cultures containing cytostatics and blank medium were incubated for three days at 37 ∘ C and 5% CO 2 . Within 72 hours, necrosis, "+" "+" "+/-" MSTO-211H "+/-" "+" "-" NCI-H2452 "-" "-" "+" -: minimal to no expression +: expression measurable +/-: little expression measurable apoptosis, and cell viability were assessed by using the following luminescence assays: CytoTox-Glo6 Cytotoxicity Assay (Promega), Caspase-Glo5 3/7 Assay (Promega), and CellTiter-Glo5 Luminescent Cell Viability Assay (Promega). The assays were performed as recommended by the supplier. Per cytostatic drug and luminescence assay at least four data points were measured. Luminescence was assessed using a SpectraMax L Luminescence Microplate Reader (Molecular Devices, CA, USA). Luminescence (relative luminescent units; RLU) was measured at 570nm and integration time was adjusted to 1 second. Temperature of the SpectraMax L was kept between 21.5 ∘ C and 24.5 ∘ C during measurements.
Additionally, from each cell line a FFPE block was prepared for immunohistochemical and qPCR analysis.

RNA Isolation and Real-Time qPCR.
Expression levels of ACTB (reference gene), MDM2 and P14/ARF, were investigated by TaqMan real-time qPCR in the three MPM cell lines. Therefore, RNA was isolated by cutting three to five sections of 4 m from the FFPE block using a microtome (Leica, SM 2000 R, Wetzlar, Germany). Total RNA was isolated using the miRNeasy FFPE kit (Qiagen, Hilden, Germany) and manufacturer's protocol, except for two modifications (proteinase K digestion overnight; elution in 25 l). RNA concentrations were measured using UV/VIS spectrometry (NanoDrop ND-1000, PEQLAB Biotechnologie GmbH, Erlangen, Germany). RNA was stored at -80 ∘ C. For cDNA synthesis, the iScript Select cDNA Synthesis Kit and protocol (Bio-Rad Laboratories, Inc., CA, USA) was used with an input of 1 g total RNA per reaction. For real-time qPCR, the TaqMan Gene Expression Assays on Demand (AoD) for ACTB (Hs03023943 g1), MDM2 (Hs01066942 m1), and P14/ARF (Hs99999189 m1) were used (Applied Biosystems5; CA, USA). The reaction volumes were modified by using 50% of the recommended total reaction volumes with 50 ng cDNA input. Each target was measured in triplicate. Ct-values of P14/ARF and MDM2 were normalized to the mean values of ACTB. Real-time qPCR and data analysis were performed on a Roche LightCycler 480 II (Roche, Basel, Switzerland) and corresponding software. All real-time qPCR experiments were performed in accordance with the MIQE-guidelines [58].

Statistical Analysis. Statistical and graphical analyses
were performed with the R statistical programming environment (v3. 4.2).
For analysis between single groups, either the Wilcoxon Mann-Whitney rank sum test (non-parametric) or twosided students t-test (parametric) was applied. For ordinal variables with more than two groups (luminescence signal differences between all treatment groups), either the Kruskal-Wallis test (non-parametric) or ANOVA (parametric) was used to detect group differences.
The level of statistical significance was defined as p<0.05.

Results
The expression profiles of MDM2, TP53, and P14/ARF differ between the investigated cell lines and are summarized in Table 2. Scans of immunohistochemical staining's are shown in Figure 1; qPCR results are visualized in Figure 2. NCI-H2052 showed pronounced MDM2-immunoexpression, but little P14/ARF and TP53-expression.   contrast, treatment with pemetrexed alone showed significantly elevated cell viability. Treatment with cisplatin alone showed higher cell viability than cisplatin and pemetrexed in combination.

MSTO-211H.
Pemetrexed combined with cisplatin was associated with the highest cell viability, followed by cisplatin alone and the lowest Nutlin-3A concentration (p=0.0952).
Pemetrexed combined with cisplatin reduced cell viability significantly, but Nutlin-3A (10 M) exhibited a slightly stronger reduction. The highest Nutlin-3A concentration reduced cell viability to a minimum.

Apoptosis
NCI-H2052. In the NCI-H2052 cell line, the highest apoptosis rate was found for 20 M Nutlin-3A, whereas the other treatment approaches showed similar apoptosis induction (p=0.14).

MSTO-211H.
In MSTO-211H, highest apoptosis rates were found for pemetrexed followed by pemetrexed in combination with cisplatin and different Nutlin-3A concentrations (p=0.0219). Almost no apoptosis was observed for cisplatin alone and Nutlin-3A.
NCI-H2452. NCI-H2452 revealed the highest apoptosis rate in response to Nutlin-3A in the highest concentration (20 M) followed by cisplatin (p=0.0359). Significantly lower apoptosis rates were found for the remaining cytostatics.
The results for apoptosis are summarized in Figures 4(a)-4(c).

Necrosis. Necrosis of cells was not influenced by any
of the chemotherapeutics compared to the control (data not shown).

Apoptosis
NCI-H2052. In the NCI-H2052 cell line, higher Nutlin-3A concentrations combined with cisplatin applied induced significantly increased apoptosis compared to pemetrexed alone or combined with cisplatin (p=0.0069). The highest apoptosis rates were found for 10 M Nutlin-3A in combination with cisplatin.

MSTO-211H.
Cell line MSTO-211H exhibited the highest apoptosis when treated with pemetrexed alone (p=0.0035). The second highest apoptosis rate was found for 10 M Nutlin-3A combined with cisplatin. Pemetrexed in combination with cisplatin resulted in the third highest apoptosis rate. Cisplatin in combination with 20 M Nutlin-3A was more potent than cisplatin alone, Nutlin-3A alone, and the lowest concentration of Nutlin-3A (5 M) in combination with cisplatin.
NCI-H2452. The highest apoptosis rates were found for the 20 M Nutlin-3A single agent as well as 10 M and 20 M Nutlin-3A concentrations combined with cisplatin, followed by cisplatin (p=0.1).
The results regarding apoptosis are summarized in Figures 4(a)-4(c).

Discussion
In previous studies we identified MDM2 as a prognostic biomarker in patients with MPM and that expression is regulated through specific miRNA [44,52,59]. Nutlin-3A inhibits MDM2-TP53 interaction and thereby induces cell cycle arrest, senescence, and apoptosis depending on the cell type [61,62]. Additionally, it is a nongenotoxic drug that exhibits little toxicity in animal models and is associated with a lower risk of resistance than conventional drugs [61][62][63].
Against this background we hypothesized that MDM2 overexpression, maybe in combination with partial or complete loss of P14/ARF, can be targeted by a Nutlin-3A based therapy regimen to restore TP53 activity in a subgroup of MPM.
In this in vitro approach, the effects of the nowadays state-of-the-art chemotherapeutics cisplatin and pemetrexed, alone and in combination, compared to Nutlin-3A were investigated in three cell lines covering the pattern found in patients [44,59]. Nutlin-3A induced senescence efficiently in all three MPM cell lines and was superior compared to cisplatin and/or pemetrexed, whereas apoptosis could only be induced at high concentrations. It is known from the literature, that the effects of Nutlin-3A are cell type specific [61,62,64], rather inducing cell cycle arrest and senescence than apoptosis [64]. Accordingly, we investigated cisplatin and Nutlin-3A in combination to increase cellular stress by inducing platin-based DNA damage. The combination of Nutlin-3A with cisplatin results in increased apoptosis and senescence rates compared to Nutlin-3A alone, as a major function of TP53 is DNA damage and stress response [46].
The same mechanism seems to be true when combining Nutlin-3A and radiotherapy to provide additional cellular damage and shift the cellular TP53-response towards apoptosis, already shown in TP53 wild-type esophageal squamous cell carcinoma in vitro and in vivo [65]. Interestingly, Shimazu et al. [66] found an additional growth inhibitory effect in MPM when combining Nutlin-3A with metformin, an mTOR inhibitor, suggesting a possible cross-talk between the mTOR-and TP53-pathway. Of note, the authors confirmed our findings of the cell lines NCI-H2052 and MSTO-211H as best responders to Nutlin-3A therapy, postulating an IC50 value of 0.37 M (MSTO-211H) and 0.50 M (NCI-H2052), respectively [66].
Interestingly, even the low MDM2 expressing cell line MSTO-211H as well as the MDM2 and TP53 negative cell line NCI-H2452 shows reduced but clearly detectable, induction of apoptosis via Nutlin-3A combined with cisplatin. Also, immunohistochemically negative cells have, as reported previously [59], detectable gene expression pattern of MDM2, resulting in MDM2 protein concentrations below the detection limit of IHC. We hypothesize, as MDM2 driven regulation of TP53 is an essential mediator of apoptosis and cell state in a physiological situation, also inhibition of the TP53-MDM2 interaction at this low MDM2 levels will have a beneficial effect on cytotoxicity of platinum compounds, explaining the occurring side effects of Nutlin-3A therapy [68]. For NCI-H2452, a cell line with absent expression of TP53, the observed effect must be TP53 independently and is most likely based on RB1 inhibitory effects.
Currently, Nutlin-3A is administered per os as substance R05045337 in a multicentre phase I clinical trial for therapy of hematologic neoplasia [69]. Additionally, RG7112, a derivative of Nutlin-3A has entered phase I clinical trials in patients with liposarcomas that are TP53 wild-type tumors with amplified MDM2 [70]. In this clinical trial, RG7112 was administered per os in 20 patients in a neoadjuvant setting [68]. One patient showed partial remission and 14 showed stable disease, but all patients suffered from side effects as neutropenia [68]. A possible explanation might be the high doses of medication of 1440 mg m −2 day −1 per os [68]. In previous in vivo studies, oral administration of Nutlin-3A showed several limitations as high input amounts of Nutlin-3A (200-400 mg/Kg) and difficulties in administering these high dosages [69]. It is noteworthy that efficient delivery systems were developed using polymers as poly(lactide-coglycolide) (PLGA) and monoclonal antibodies [69].

Conclusion
In this in vitro study, our hypothesis that MDM2-overexpressing MPM can be targeted by a Nutlin-3A based chemotherapy was proven. Particularly, for an optimal biomarker setting of MDM2-overexpression and low/absent P14/ARF expression, superior apoptosis and senescence rates were seen compared to the conventional chemotherapeutics. Even for a less optimal biomarker setting with minimal MDM2 expression, a favorable induction of apoptosis and senescence was obvious for Nutlin-3A in combination with cisplatin compared to the conventional drug regimen. Therefore, Nutlin-3A based therapy approach could be of great value for a subgroup of patients with MPM.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Disclosure
The results of the present study have been presented at the German Cancer Consortium (DKTK), 1st Essen Translational Oncology Symposium (ETOS) (Essen, 2018), and the 33rd German Cancer Congress (Berlin 2018).

Conflicts of Interest
The authors declare no conflicts of interest.