Introduction
Key functions of the cellular tumor suppressor p53 are to prevent the development of malignancy and stop the growth and expansion of abnormal cells through repair of DNA damage, arresting cell cycle, induction of apoptosis, and controlling angiogenesis [
1]. During physiological conditions, p53 levels remain low but once activated it is followed by an increase in protein levels, increase in DNA binding, increase in multiple target genes, and activation of its corresponding biological effects [
1]. Based on Pan-cancer data,
TP53 remains the most frequently mutated gene observed in human malignancy [
2]. p53 is nullified either by alterations in the
TP53 gene or through oncogenic activity of the E3 ubiquitin-protein ligase MDM2 [
3] and the related MDMX (MDM4) proteins [
4]. Two different models of biological functions of mouse double minute (MDM) proteins have previously been proposed where in the first model, Mdm2 and Mdm4 act independently by regulating specific activities of p53. In this model, it is proposed that p53 levels are mainly being regulated by MDM2, and the transcriptional activity by MDMX. In the second model, they jointly regulate p53 function [
4‐
7].
Amplification of the
MDM2 gene and/or overexpression of its protein have been found to be a main driver of malignancy in many tumor types, especially in tumors with retained wild-type
TP53 status (WT
TP53) [
8]. A previous study demonstrated estrogen receptor-positive (ER+) breast cancer comprises over 70% of breast cancers [
9] and majority of ER+ breast cancers have WT
TP53 status [
10]. Additionally, 50% of human tumors retain their WT
TP53 status but this differs across tumor histologies [
5]. Although not being the key regulator of p53, Mdm4 overexpression and amplification at the gene level has been linked with multiple human malignancies, promoting neoplastic proliferation. This was mainly observed in ER+ breast cancers [
11]. MDM2 inhibitors, also known as nutlins, were initially identified in the early 2000s [
12] and were found to interrupt the MDM2-p53 interaction. Since then, the portfolio of small molecule MDM2 inhibitors has continued to grow [
13]. They have demonstrated promising antitumor efficacy in preclinical ER+ breast cancer [
14]. Interestingly, it was found that in order to achieve full p53 activation, dual inhibition of Mdm2 and MDMX is needed [
15]. ALRN-6924 is a first of its class cell-penetrating stapled α-helical peptide that disrupts the interaction of p53 tumor suppressor protein and its endogenous inhibitors Mdm4 and Mdm2 in WT
TP53 tumors [
16‐
18]. ALRN-6924 was tolerated and has shown antitumor activity in a phase I clinical trial in patients with solid tumors (NCT02264613) [
17].
Paclitaxel is one of the most commonly used standard of care chemotherapeutic agents in breast cancer patients [
19]. It induces microtubular polymerization, which leads to stabilization of microtubules, followed by catastrophic mitotic arrest, and demise of the cancer cell [
20]. Eribulin is another safe and tolerable FDA-approved anti-microtubule inhibitor which showed an improvement in overall and progression-free survival in metastatic breast cancer patients progressing on prior chemotherapy [
21].
Resistance to chemotherapeutic agents is a major concern when treating breast cancer patients, especially with taxanes [
22]. MDM2 is a known mediator of chemoresistance [
23]; this further accentuates the pressing need to develop rational combinatorial regimens to overcome this resistance. To date, chemotherapeutic agents have not been clinically evaluated in combination with an MDM2/MDMX inhibitor. The aim of the reported studies was to determine if these drug combinations will work synergistically and increase the sensitivity of cancer cells to chemotherapeutic agents. We therefore sought to evaluate the combined antitumor efficacy of ALRN-6924 and paclitaxel or eribulin in ER+ breast cancer with WT
TP53 status by testing cell lines in vitro and in vivo. Our findings indicate that ALRN-6924 acts synergistically with both paclitaxel and eribulin in preclinical breast cancer models.
Materials and methods
Reagents
ALRN-6924 was kindly provided by Aileron Therapeutics Inc. (Watertown, MA, USA). RG7112 and RG7388 (idasanutlin) and paclitaxel were purchased for in vitro studies from MedChemExpress (Monmouth Junction, NJ, USA). For in vivo paclitaxel efficacy studies, the agent was obtained from MD Anderson Cancer Center pharmacy. This source of paclitaxel was later changed to the National Cancer Institute for pharmacodynamic experiments. Paclitaxel was prepared in 15% ethanol, 15% kolliphor (Sigma-Aldrich Co., LLC, St Louis, MO, USA), and water for in vivo studies. Eribulin was obtained from the MD Anderson Cancer Center pharmacy. Dimethyl sulfoxide (DMSO) was obtained from Sigma-Aldrich (Sigma-Aldrich Co., LLC, St Louis, MO, USA). All drugs used for in vitro studies were initially prepared in DMSO prior to dilution in a tissue culture medium.
Cell line panel screening
Two commercial cancer cell line panels, the Eurofins OncoPanel (St. Charles, MO;
n = 233 cell lines) and a focused Horizons Discovery OncoSignature panel (Cambridge, UK;
n = 89), were employed. Total 302 tested cell lines representing multiple solid and hematologic tumor types were composed of 100 cell lines with WT
TP53 and 202 cell lines with mutant or null
TP53 (Supplementary Table
1). Cell lines from solid and hematologic tumors were seeded into 384-well plates and incubated in a humidified atmosphere of 5% CO
2 at 37 °C overnight. Cells were treated with test compounds over 10 concentrations, up to 30 μM, for 72 h. Cells were then fixed and stained in the same wells with nuclear dye and fluorescently labeled antibodies (Eurofins) or ATP levels were measured by adding Perkin Elmer ATPLite (Horizons). Relative cell proliferation was assessed by changes in nuclear dye uptake or ATP levels compared to the untreated cell plate and presented as the effective concentration to inhibit 50% of cell proliferation (IC
50).
In an independent screen of ER+ breast cancer cell lines, 6 cell lines were used, including MCF-7, ZR-75-1 (WT TP53), and HCC-1954, SK-BR-3, BT-474, CAMA1 (mutant
TP53). These ER+ breast cancer cell lines were obtained from American Tissue Culture Collection (Manassas, VA, USA). Except for ZR-75-1 cells that were cultured in RPMI-1640 medium, other cell lines were cultured in Dulbecco’s modified Eagle’s medium/F12 supplemented with 10% fetal bovine serum (FBS) at 37 °C and humidified in 5% CO
2. Cells were seeded in 96-well plates overnight at a density of 5000–10,000 cells per well based on cell line growth characteristics. Cells were treated with ALRN-6924 single agent in a dose range of 0.64–50,000 nM for 3 days. Sulforhodamine B staining (SRB) assay was performed to determine cell survival rate [
24]. IC
50 was determined based on the sigmoid drug-inhibition curve using GraphPad Prism v7.03 software.
Drug combination assay
Two ER+ breast cancer cell lines with WT TP53, MCF-7 and ZR-75-1, were used to determine ALRN-6924 combination efficacy. Cells were treated with ALRN-6924 in combination with chemotherapy agents at individual dose ranges, based on the dose-response curves for ALRN-6924 single drug treatment as described above and for the chemotherapy drugs that we previously established. Each combination treatment group has 6 doses with a fixed combination ratio and a fixed serial dilution. For example, in the combination of ALRN-6924 + paclitaxel, we chose 5000 nM and 10 nM as the highest doses for ALRN-6924 and paclitaxel respectively (combination ratio, 500). Then we had 5 serial dilutions at 5-fold from the highest combo doses. In the combination of ALRN-6924 with eribulin, 5000 nM and 80 nM were chosen as the highest doses respectively (combination ratio, 62.5). To evaluate combination efficacy, a combination index (CI) was determined using the Chou-Talalay drug combination model (CI < 1.0 (curve left-shift): synergistic; CI = 1.0, additive; CI > 1.0 (curve right-shift): antagonistic). Experiments were performed in triplicate and repeated three times.
Cells were seeded in 60-mm plates at a density of 1 × 103 for MCF-7 and 5 × 103 for ZR-75-1 per well. Drug treatment started following 24 h of incubation, with changing of culture media and drugs at least twice a week. Each treatment group was tested in duplicate. Culturing and treatment of cells continued for 2–3 weeks. At the end of the experiment, the colonies were fixed in 10% buffered formalin and stained with 0.05% crystal violet in 25% methanol. Total colony area was measured using NIH ImageJ v1.52a software (NIH, Bethesda, MD, USA).
Western blot assay
Pierce BCA protein assay kit (ThermoFisher, Waltham, MA, USA) was utilized to obtain protein concentration of cell lysates. Membranes were blocked with 0.1% casein blocking buffer following separation using sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transfer of proteins onto a 0.2-μm nitrocellulose membrane (Bio-Rad Laboratories, Hercules, CA, USA). Next, immunoblotting was performed by probing membranes with antibodies (anti-Mdm2 (#86934), anti-p53 (#9282), anti-p21 (#2947), anti-p-Histone-H3 (Ser10) (#29237), anti-c-Myc (#9402), anti-LC3A/B (#12741), anti-p-Rb (Ser807/811) (#8516), anti-Rb (#9309), anti-p-S6 (Ser235/236) (#4858), and anti-S6 (#2217)). All described antibodies were purchased from Cell Signaling Technology (Boston, MA, USA). Anti-β-actin antibody (#A5441) was purchased from Sigma-Aldrich (Sigma-Aldrich Co., LLC, St Louis, MO, USA). Immunoblotting signals were visualized by Odyssey IR imaging system (Li-Cor Biosciences, Lincoln, NE, USA). Image Studio software v4.0 (Li-Cor Biosciences, Lincoln, NE, USA) was used for the analysis of the bands.
Annexin V assay
To evaluate for apoptosis, we utilized the Annexin V Apoptosis kit (#11858777001 Roche, Indianapolis, IN, USA). Cells were treated for 72 h with either vehicle, ALRN-6924 10 μM, paclitaxel 1 μM, or combination of both followed by collection of both floating and attached cells. These concentrations were chosen because of findings of p53 reactivation and apoptosis on immunoblotting, respectively. Cells were then probed with both Annexin V and propidium iodide (PI). Labeled cells were analyzed by flow cytometry in the MDACC Flow Cytometry and Cellular Imaging Core. Annexin V-positive cells were counted.
Cell cycle analysis
To determine the effects of the combination of ALRN-6924 and paclitaxel on the cell cycle, we used MCF-7 and ZR-75-1 cell lines. Cells were plated and treated the following day in duplicates for 24 h with either vehicle, ALRN-6924 10 μM, paclitaxel 1 μM, or combination of both. Samples were collected, fixed, stained with PI (Abcam, #ab139418, Cambridge, MA, USA) according to manufacturer protocol, and analyzed by flow cytometry in the MDACC Flow Cytometry and Cellular Imaging Core Facility. We evaluated the percentages of cells in each cell cycle phase: subG1, G0-G1, S, and G2-M.
In vivo studies
All animal experiments were approved by the Animal Care and Use Committee of MD Anderson. Cells were inoculated into the 4th mammary fat pad of 6- to 8-week-old female athymic BALB/c nu/nu mice (Department of Experimental Radiation Oncology, MD Anderson) at a density of 5 × 106 in both MCF-7 and ZR-75-1 xenograft models. All mice were implanted with 60-day 17β-estradiol pellets (Innovative Research of America, Sarasota, FL, USA) subcutaneously prior to inoculation of the tumor. Treatments began when tumor diameter reached between 150 and 200 mm3. The mice were euthanized at the end point of the study or when tumor diameter reached 1500 mm3.
In order to determine in vivo antitumor efficacy of ALRN-6924 in combination with paclitaxel, mice were randomized into 6 groups (vehicle, ALRN-6924 5 mg/kg, ALRN-6924 10 mg/kg, paclitaxel 10 mg/kg, ALRN-6924 5 mg/kg + paclitaxel 10 mg/kg, and ALRN-6924 10 mg/kg + paclitaxel 10 mg/kg, n = 6–8 in MCF-7, and n = 5 in ZR-75-1 xenograft combination therapy experiment). ALRN-6924 (5, 10 mg/kg) was dosed twice weekly and paclitaxel (10 mg/kg) was dosed weekly, with paclitaxel administered 6 h prior to ALRN-6924. All drugs were administered intravenously via the tail vein over a 26–28-day treatment window.
To evaluate the drug effects and their mechanism of action, we performed pharmacodynamics studies consisting of two additional in vivo experiments. Groups of 4–5 mice with stable tumor sizes ranging between 150 and 200 mm3 were randomized into vehicle, ALRN-6924 10 mg/kg, paclitaxel 15 mg/kg, and ALRN-6924 10 mg/kg + paclitaxel 15 mg/kg. Mice were treated for a total of 14 days and euthanized 16 h after the last treatment. Tumors were cut into halves, one half of the tumor was fixed in formalin followed by paraffin embedding, while the other half was snap-frozen.
Reverse-phase protein arrays
Reverse-phase protein array (RPPA) was conducted by the MD Anderson Functional Proteomics core facility as previously described [
25]. Differentially expressed protein association with the canonical pathways was analyzed using the Ingenuity Pathway Analysis (IPA) software (Qiagen, Hilden, Germany).
Immunohistochemistry
Xenograft tumor samples were formalin-fixed and paraffin-embedded (FFPE). Tumor tissue on hematoxylin and eosin (H&E)-stained slides was confirmed by a board-certified pathologist. Additionally, the presence of tumor necrosis was estimated and given as percentage of the entire tumor.
Immunohistochemistry (IHC) was performed on 4-μm sections using the following antibodies: anti-Ki67 (clone: MIB-1, #M7240, Dako/Agilent, Glostrup, Denmark), anti-cleaved PARP (Asp214) (clone: D64E10, #5625, Cell Signaling Technology, Danvers, USA), anti-p21 (clone: F5, #sc-6246, Santa Cruz, Dallas, USA), anti-p53 (clone: DO-7; Leica Biosystems, Wetzlar, Germany), and anti-TUNEL (ApopTag Peroxidase In Situ Apoptosis Detection Kit, Merck Millipore Sigma, # S7100). For Ki67, p-Histone-H3, and cleaved PARP, the percentage of all positive tumor cells were estimated regardless of staining intensity. The percentage of positive tumor cells and their staining intensity were evaluated for p21 and p53 resulting in an H-score (range 0–300) [
26]. TUNEL staining was performed at the Houston Methodist Research Institute (HMRI, Houston, TX, USA). Percentage (%) and density (n/mm
2) of positive tumor cells were evaluated for TUNEL staining. For digital image analysis, IHC slides were scanned at 20× using Aperio AT2 scanner (Leica Biosystems), and the images were analyzed in HALO software (Indica Labs), using a cytonuclear algorithm. Distinguished necrotic areas and artifacts were excluded. All specimens but one (p21: insufficient staining) could be evaluated for IHC. The IHC evaluation was performed by a board-certified pathologist.
Unsupervised hierarchical cluster analysis was performed in order to generate MCF-7 heat maps of differentially expressed proteins (DEPs) with false discovery rate (FDR) 0.05 (ALRN-6924 vs vehicle and ALRN-6924 plus paclitaxel vs paclitaxel).
Statistical analysis
For in vitro studies, Student’s t-test was carried out for comparison between two groups.
Treatment/control ratio (T/C %) was performed day 27 or day 29, depending on the study by using the formula: (median tumor volume of treatment group/median tumor volume of control) × 100. Antitumor activity was defined as ≤ 40% [
27].
Tumor volume (TV) was obtained by using the formula: TV (mm3) = ((width)2 × length)/2. Change in tumor volume was obtained by using the formula: (TVDayX − TVDay0)/TVDay0. One-way ANOVA was carried out for in vivo studies in order to determine statistical significance.
Discussion
The majority of human tumors have a mutated
TP53 gene. We have previously shown that almost a third of patients with metastatic hormone receptor-positive breast cancer have
TP53 mutations, and
TP53 mutations were associated with a shorter recurrence-free survival and overall survival [
36], demonstrating the importance of p53 pathway. However, most patients with hormone receptor-positive breast cancers still have WT
TP53 status [
2]. Many breast cancers exhibit
Mdm2/Mdm4 amplifications; however, these alterations often do not co-occur with
TP53 mutations [
37]. Interestingly, MDM2 overexpression was found to induce chemoresistance and additionally enhances invasiveness and motility of breast cancer cells [
5]. Based on this knowledge, it provided us a therapeutic rationale to analyze a possible cancer therapeutic target by itself and in combination with standard chemotherapeutic agents in breast cancer models. Our findings revealed the targeted agent ALRN-6924 was able to enhance the antitumor efficacy of both paclitaxel and eribulin in vitro and in vivo. We were also able to demonstrate that both ALRN-6924 alone and in combination with paclitaxel restore p53 activity and inhibit progression of cell cycle mainly through the activity of p21.
We demonstrated enhanced induction of apoptosis in vivo but not in vitro. This could potentially be explained by the effect of ALRN-6924 combined with paclitaxel on the stromal elements present in xenograft models. p-HH3 marker has emerged as a proliferative or mitotic marker compared to other proliferative markers such as Ki67, Mitotic Activity Index (MAI), and demonstrated greater accuracy [
38]. p-HH3 levels were lower in both single agent ALRN-6924 and combination cohorts compared to vehicle, suggesting that ALRN-6924 possibly is not acting through mitosis, but rather through DNA damage [
39]. The increased p-HH3 levels observed with single agent paclitaxel has previously been described when cells are exposed to noxious stimuli and subsequent impending apoptotic event [
40,
41]. There was a response to ALRN-6924 single agent on cell proliferation and apoptosis in ZR-75-1 cells. Cell proliferation and apoptosis involve different mechanisms. Thus, the apoptosis pathway in ZR-75-1 cells could be more vulnerable than the proliferation pathway, compared to MCF-7 cells. However, this warrants future investigation. Several molecular markers of autophagy have been studied to date, but the conversion of LC3-I to LC3-II has been accepted as the gold standard for autophagosome formation. Upon autophagic signal, the cytosolic LC3-I is conjugated to phosphatidylethanolamine (PE) to form LC3-PE conjugate (LC3-II), which is recruited to the autophagosomal membranes. Thus, the reduced ratio of LC3-I/LC3-II in this paper represents activation of autophagy by MDM2/X inhibition. MDM2 interacts with pRb which forms a triplex complex with p53. Our RPPA data showed a co-regulation of p53 and pRb by the MDM2 inhibitor (Fig.
5a). MDM2 is also known to have p53-independent actions. MDM2 promotes cell proliferation by activating NFκB pathway but inducing degradation of tumor suppressor FoxO3A which acts via ERK. MDM2 also downregulates E-Cadherin, leading to invasion. However, we did identify the changes in these markers in our RPPA study (Fig.
5a). Therefore, the p53-independent roles may not be primary roles of MDM2 in these cell lines.
In a separate study, we examined the target specificity of MDM2/X inhibitors. We created MDMX knockout cell lines in OCI-AML3 cells (Fig. S
4A, B). The data showed that MDMX KO cells were more sensitive to MDM2/X inhibitor ALRN-6924 and MDM2 inhibitor RG7388 than WT cells (Fig. S
4C, D). However, further work is needed to dissect the role of MDMX inhibition in ALRN-6924’s antitumor efficacy, and how this efficacy profile differs from those of MDM2 inhibitors. Interestingly, clinical trials appear to also demonstrate differences in toxicity profile, as ALRN-6924 appears to have less bone marrow toxicity noted in phase I trials compared to what has been reported with other MDM2 inhibitors [
42,
43].
Another question we have not specifically addressed is the effect of ALRN-6924 on the tumor microenvironment. Cell line-derived xenografts have mouse stroma. They are not the optimal model to assess the effect of ALRN-6924 on the tumor microenvironment. Further, our RPPA assay has been optimized to detect human markers and is focused on tumor markers, and thus, our current studies are not conducive to address this question. Therefore, we plan to test the effect of ALRN-6924 on the tumor microenvironment in ongoing clinical trials.
mTOR pathway is one of the classical canonical pathways involved in cancer cell proliferation and metastasis [
44]. Emerging evidence has demonstrated that the mTOR pathway has been associated with endocrine therapy resistance [
45], especially in metastatic breast cancers [
46]. By inhibiting this pathway, it has been shown to prevent resistance and restore sensitivity in hormone-positive breast cancers [
47]. Clinically, mTOR inhibitor everolimus is known to increase progression-free survival when given in combination with exemestane compared to exemestane alone in post-menopausal women with advanced hormone receptor-positive breast cancer [
48]; thus, mTOR is a proven target in breast cancer. Our study revealed an off-target effect of ALRN-6924 combined with paclitaxel, downregulating p-mTOR and p-S6, possibly explaining one of the potential mechanisms of decreased proliferation when combining these two drugs in vivo.
Additional mechanistic studies are required in order to explore and determine the interplay between ALRN-6924 and paclitaxel. The question also remains on how the combination of ALRN-6924 and paclitaxel or eribulin affect other additional p53-dependent and independent pathways, potentially revealing additional affected pathways.
Changes in a single residue by hot-spot mutations can impair p53 activity which transforms p53 into a gain-of-function oncoprotein [
49]. Two scenarios have been demonstrated where both
TP53 WT and mutant p53 (
TP53 MT) can affect each other, one being “dominant-positive” effect, where
TP53 WT activity suppresses
TP53 MT activity, or “dominant-negative effect” where
TP53 MT allele inactivates
TP53 WT allele [
50]. We did not test the combination of ALRN-6924 and paclitaxel or eribulin in a setting with both
TP53 MT and
TP53 WT; further studies are needed in order to determine if this combination can rescue
TP53 WT and suppress mutant p53 activity in order to enhance paclitaxel or eribulin sensitivity.
Notably, our studies have been conducted in immunocompromised mice, not allowing us to assess the immune effects of Mdm4/Mdm2 inhibition. Recently, there have been anectodal reports of hyperprogression on immunotherapy in patients with
Mdm2 amplification [
51]. The role of p53 in signal transduction pathways involved in the regulation of the immune response has been extensively discussed [
52,
53]. In murine syngeneic models, ALRN-6924 and selective Mdm2 antagonists were shown to have co-stimulatory activity in T cells and increase PD-L1 expression, and in combination with anti-PD1 and anti-PDL1, demonstrated enhanced antitumor activity not only in
TP53 WT but also in
TP53 MT tumors [
54]. Thus, further exploration is needed to determine whether the antitumor efficacy of ALRN-6924 combination with chemotherapy would be further enhanced in immunocompetent models.
Competing interests
S. Pairawan, K. Evans, A. Akcakanat, X. Zheng, Y. Rizvi, E. Yuca, F. Yang, P. Subash Chandra Bose, M. Zhao, and E. IIeana Dumbrava declare no potential conflict of interest.
C. Tapia is an employee and shareholder of Epizyme Inc. and performed contract work for Armo Bioscience.
A. Annis, L. Carvajal, JG. Ren, S. Santiago, V. Guerlavais, and M. Aivado are employees of Aileron Therapeutics Inc., and David Sutton is a paid consultant for Aileron Therapeutics Inc.
F. Meric-Bernstam reports receiving commercial research grants from Aileron Therapeutics Inc., AstraZeneca, Bayer Healthcare Pharmaceutical, Calithera Biosciences Inc., Curis Inc., CytomX Therapeutics Inc., Daiichi Sankyo Co. Ltd., Debiopharm International, eFFECTOR Therapeutics, Genentech Inc., Guardant Health Inc., Millennium Pharmaceuticals Inc., Novartis, Puma Biotechnology Inc., and Taiho Pharmaceutical Co. She also served as a consultant for Aduro BioTech Inc., DebioPharm, eFFECTOR Therapeutics, F. Hoffman-La Roche Ltd., Genentech Inc., IBM Watson, Jackson Laboratory, Kolon Life Science, OrigiMed, PACT Pharma, Parexel International, Pfizer Inc., Samsung Bioepis, Seattle Genetics Inc., Tyra Biosciences, Xencor, and Zymeworks. Additionally, she serves on the advisory committee for Immunomedics, Inflection Biosciences, Mersana Therapeutics, Puma Biotechnology Inc., Seattle Genetics, Silverback Therapeutics, and Spectrum Pharmaceuticals.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.