Anti-tumor effects of ONC201 in combination with VEGF-inhibitors significantly impacts colorectal cancer growth and survival in vivo through complementary non-overlapping mechanisms

Background Small molecule ONC201 is an investigational anti-tumor agent that upregulates intra-tumoral TRAIL expression and the integrated stress response pathway. A Phase I clinical trial using ONC201 therapy in advanced cancer patients has been completed and the drug has progressed into Phase II trials in several cancer types. Colorectal cancer (CRC) remains one of the leading causes of cancer worldwide and metastatic disease has a poor prognosis. Clinical trials in CRC and other tumor types have demonstrated that therapeutics targeting the vascular endothelial growth factor (VEGF) pathway, such as bevacizumab, are effective in combination with certain chemotherapeutic agents. Methods We investigated the potential combination of VEGF inhibitors such as bevacizumab and its murine-counterpart; along with other anti-angiogenic agents and ONC201 in both CRC xenograft and patient-derived xenograft (PDX) models. We utilized non-invasive imaging and immunohistochemistry to determine potential mechanisms of action. Results Our results demonstrate significant tumor regression or complete tumor ablation in human xenografts with the combination of ONC201 with bevacizumab, and in syngeneic MC38 colorectal cancer xenografts using a murine VEGF-A inhibitor. Imaging demonstrated the impact of this combination on decreasing tumor growth and tumor metastasis. Our results indicate that ONC201 and anti-angiogenic agents act through distinct mechanisms while increasing tumor cell death and inhibiting proliferation. Conclusion With the use of both a murine VEGF inhibitor in syngeneic models, and bevacizumab in human cell line-derived xenografts, we demonstrate that ONC201 in combination with anti-angiogenic therapies such as bevacizumab represents a promising approach for further testing in the clinic for the treatment of CRC. Electronic supplementary material The online version of this article (10.1186/s13046-018-0671-0) contains supplementary material, which is available to authorized users.


Background
Colorectal cancer is the third most common cancer in the world and has an overall 5-year survival rate of approximately 10% for those with stage IV cancer [1,2]. The cure rate and 5-year survival rate is significantly lower for those with metastatic colorectal cancer who cannot be cured from surgery alone. Therefore, therapies that can treat metastatic colorectal cancer is needed [3].
The novel anti-cancer imipridone ONC201 induces the TRAIL and DR5 genes through dual inactivation of Akt/ERK/Foxo3a and activation of the integrated stress response (ISR). Further, in vivo, ONC201 possesses a broad spectrum of activity, wide safety margin, robust stability, aqueous solubility, and favorable pharmacokinetics [4][5][6][7][8][9][10][11][12][13]. The therapeutic activity of ONC201 in preclinical in vivo studies in solid tumors, hematological malignancies, and with targeting of cancer stem cells as well as bulk tumor cells prompted its ongoing clinical development. In Phase I clinical testing with ONC201, patients were treated with the compound once every 3 weeks and the drug showed evidence of safety and promising efficacy in multiple tumor types [14].
Tumor angiogenesis is the process by which new blood vessels are developed; a critical process in tumor progression and development [15]. Many growth factors are needed for angiogenesis including vascular-endothelial growth factor (VEGF), fibroblast growth factors, and platelet-derived endothelial growth factors, which bind to three tyrosine kinase receptors: VEGFR1/2 which promote angiogenesis, and VEGFR3 which stimulates lymphangiogenesis [16]. These corresponding receptors are located on endothelial cells of pre-existing blood vessels and promote the activation of endothelial cells [17].
High levels of VEGF has been shown to increase vascular disorganization and permeability; creating heavily leaky tumors with poor perfusion and enhancing the ability of tumor cells to spread throughout the body [18]. Further, higher VEGF expression levels has been detected in various human cancers including colorectal and non-small lung cancer and have some correlation to outcome [19][20][21].
Bevacizumab (Avastin), a humanized monoclonal antibody designed to neutralize human VEGF, inhibits VEGF-induced proliferation of endothelial cells and promotes endothelial cell apoptosis. Treatment with monoclonal antibodies such as bevacizumab have been show to inhibit growth of tumors in vivo, promote tumor cell apoptosis, and prevent the spread of metastases [22][23][24][25]. Bevacizumab functions best as a combinational agent and has shown promise in combination with several approved chemotherapies including with 5-fluorouracil or paclitaxel; causing it to be approved by FDA for metastatic CRC, non-small cell lung cancer, and metastatic breast cancer [22,[26][27][28]. Regorafenib, an oral multi-kinase inhibitor with anti-angiogenic properties is also approved for metastastic CRC but has a distinct profile of adverse advents including hepatotoxicity, fatigue, diarrhea, hypertension, and hand-foot syndrome [29,30]. Here we demonstrate that ONC201 and bevacizumab, or its murine counterpart, provide a potent combinational therapy option when compared to regorafenib that could be further pursued in the clinic.

Cell lines and PDX tumors
All cell lines were obtained from the American Type Culture Collection. CT26 and MC38 cells were provided by Dr. Scott Waldman's lab at Thomas Jefferson University. ONC201 was provided by Oncoceutics.
The PDX tumor was provided by NexusPharma Inc., Philadelphia, PA. The PNX0229 sample was obtained from a 57-year old Caucasian male with a Stage 2A descending colon adenocarcinoma. The sample was taken from a liver metastases that formed. The patient underwent a combination of FOLFIRI and Erbitux with a partial response; and a second line therapy of FOLFOX with progressive disease before the resection.

Small molecules and dosing schedule
ONC201 was administered orally in 10:70:20 DMSO: PBS:Cremphor El as described [4] and treated weekly at the indicated doses. Bevacizumab was procured from the Fox Chase Cancer Center pharmacy and diluted in PBS. Bevacizumab was administered through retro-orbital injections every other week at a dose of 5 mg/kg. Regorafenib was procured from MedChemExpress (HY-1031) and administered orally at 10 mg/kg per day dissolved in PBS for at least 22 days. Anti-murine VEGF-A inhibitory antibody (Biolegend 512,808) was administered at 10 micrograms by i.p. twice weekly. Mouse body weight was observed every 3 days and when body weight began to drop regorafenib dosing was stopped.

In vivo animal experiments
All animal experiments were conducted in accordance with the Institutional Animal Care and Use Committee at Fox Chase Cancer Center. For subcutaneous xenografts, 6week-old female athymic nu/nu mice (Taconic Biosciences) were inoculated with 1 × 10 6 cells of the HT29-luciferase, HCT116 p53 −/− , or HCT116-GFP cell lines in each rear flank, in a 150 μl suspension of 1:1 Matrigel (BD). For subcutaneous xenografts in syngeneic models, CT26 cells were inoculated with 1.0 × 10 6 cells into six-week old female Balb/c mice (Taconic Biosciences) and MC38 cells were inoculated with 1.0 × 10 6 cells into six-week old female C57/ BL6 mice (Taconic Biosciences). All subcutaneous tumors were allowed to establish for 1 to 3 weeks after injection until they reached a volume of~200 mm 3 before treatment initiation. Mice were monitored every 3 days and tumor volumes were measured using calipers. Tumor volumes were measured according to the formula (L*W^2)/2.

In vivo pathology and toxicology
Toxicity during the course of ONC201 and combination treatment was judged by body weight decrease of greater than 10%, tumor growth of more than 10% of body weight, or a body condition scoring < 2. Serum and plasma samples were collected through orbital bleeding and cardiac puncture before sacrifice, and samples were immediately stored at 4°C and processed by Antech Diagnostics for CBC and chemistry panels. Results were analyzed by board-certified toxicologists. Tumors were measured post-mortem through caliper and water density examination. Organ and tumor samples were processed in 10% formalin and fixed in paraffin. Hematoxylin-stained samples were analyzed by a boardcertified pathologist to determine whether tumor cells existed on any organs or necrosis occurred in tumors. Board-certified veterinary pathologists also indicated whether or not signs of toxicity were present.

Immunohistochemistry
All antibodies were purchased from Cell Signaling. After fixation, the tumor samples were embedded in paraffin and 8 μm sections were cut and mounted on slides. The sections were then processed and analyzed using immunohistochemistry with VEGF, Ki67, and CD31 [4]. CD31 was only scored on those tumors +/− 600 mm 3 from the average tumor size in order to best control for tumor size variation. CD31, and Ki67 levels where calculated by independent blind-scoring and the use of VECTRA 3.0 Automated Quantitative Pathology Imaging system and Inform 2.0 software cursory of the Fox Chase Cancer Center Biosample Repository.

HUVEC assays
Human normal primary umbilical vein endothelial cells where purchased from ATCC and HUVEC assay was used following the protocol from Millipore (ECM625). Imaging occurred at 10 h unless indicated. IC50 doses where used as previously published. Sprouting was counted and calculated as described in Millipore methods.

Fluorescent probe imaging
Mice where anesthetized with isoflurane and injected through retro-orbital i.v. injections at recommended amounts for either Superhance 680 (NEV10116) or Angiosense 750 (NEV10011EX). Mice were allowed to equilibrate for the recommended times for each probe; and then they were imaged as directed by Perkin Elmer on a Maestro in-vivo imaging system (Cri, Woburn, MA). Mice with Superhance were also imaged for GFP-expressing tumors. Relative luminescence units (RLU) where calculated by measuring tumor fluorescence subtracted from the background of mouse auto-fluorescence divided by the area of tumor, measured by ImageJ.

ELISA assays
A total of 500 μL blood was collected through orbital blood draw while mice where anesthetized and plasma was collected in EDTA tubes and serum collected in Heparin separating tubes. Tubes were spun at 1000 rpm for 15 min. Samples were analyzed using Human TRAIL/TNSFSF10 Quantikine ELISA kit (R&D Systems, Minneapolis MN) or Human Apoptosense M30 ELISA kit (MyBiosource MBS9300524). All analyses were performed according to the manufacturer's directions.

Statistical analysis
Data are presented as means + SD. To assess the statistical significance of the differences, unpaired Student's t test with Holm-Sidak correction for multiple comparisons (maximum of three comparisons were made) was performed, with P < 0.05 deemed statistically significant. Measurements from three biological replicates per treatment group were compared. Unless otherwise noted in the figure legend, comparisons were made against the vehicle control.

ONC201 shows combinatorial efficacy with bevacizumab in vivo
ONC201 has been shown to be efficacious in colorectal xenografts [4]. However, some aggressive xenografts are more resistant to ONC201's treatment. We screened for potential synergy between ONC201 and approved CRCbased therapies (Additional file 1: Figure S1). Our in vitro results suggested no major synergies with 5fluorouracil, oxaliplatin or irinotecan although there may be some additive effects. We further sought to investigate the combination of ONC201 and bevacizumab (Avastin) since bevacizumab is an FDA-approved and relatively well-tolerated drug used to treat patients with advanced colorectal cancer. Assessing for a potential synergy between ONC201 and bevacizumab in vivo, we observed that in the HT29 xenograft model, the combination of ONC201 and bevacizumab led to a significant reduction in tumor growth and development ( Fig. 1a-b). Both bevacizumab alone and the combination led to no metastases forming during tumor development (Additional file 2: Figure S2). This combination was also synergistic against colorectal PDX tumors, and more potent in suppressing tumor growth as compared to ONC201 or bevacizumab alone. (Fig. 1c). The combination of ONC201 plus bevacizumab was also shown to be nontoxic (Additional file 3: Figure S3).

ONC201 shows combinatorial efficacy with syngeneic colorectal tumors and mouse anti-angiogenic anti-VEGF or regorafenib in vivo
One limitation that has been noted over the years is that in preclinical models of human xenografts, tumor growth is rapid and not supported by vascular structures; particularly from a different species such as mouse host. Using humanized monoclonal antibodies also only attacks the human vasculature that was able to grow within the xenograft, and not within the host. Thus, tumor vessels within murine tumors implanted in syngeneic mouse models tend to be more prevalent and are able to grow within the tumor [31,32]. To better demonstrate the relevance of VEGF inhibitors in a complete system, we chose to use a murine VEGF-A inhibitory antibody, a murine counterpart of bevacizumab, in syngeneic models injected with murine CRC.
To determine whether ONC201 can also synergize with other anti-angiogenic agents; we treated mice with regorafenib and a murine VEGF-A inhibitor in combination with ONC201. ONC201 significantly synergized with either bevacizumab or the murine anti-VEGF-A (Fig. 2). Although the combination of ONC201 with the less specific compound, regorafenib, was not as significant as compared to VEGF-inhibitors, regorafenib/ ONC201 therapy still tended to be more potent than mono-agent therapy ( Fig. 2a-b, Additional file 2: Figure S2). Similar to what was observed with bevacizumab, regorafenib in combination with low doses of ONC201, reduced the incidence of metastases as compared to low doses of ONC201 alone (Fig. 2c). These drug combinations were also found to be non-toxic in mice as seen by no notable significant side effects during treatment, no significant impact on serum blood  Figure S3).

ONC201 does not significantly impact VEGF levels or human vascular endothelial cell (HUVEC) growth in vitro and does not impact vascular growth in vivo
To determine potential mechanism(s) of synergy between ONC201 and anti-VEGF therapy, we screened for effects of the combinational therapies with focus on mono-agent key mechanisms of actions. In our colorectal xenograft, VEGF expression does not appear to be significantly impacted by weekly ONC201 treatment (Fig. 3). Not surprisingly, in the presence of bevacizumab, as an inhibitory VEGF monoclonal antibody, very little VEGF expression was observed (as bevacizumab masks VEGF detection by anti-VEGF antibodies). The combinational therapy also showed significantly less VEGF expression than the vehicle and ONC201 treatment (Fig. 3a, Additional file 4: Figure S4). We tested the effect of ONC201 on HUVEC sprouting and growth. While ONC201 did not impact the amount of HUVEC sprouting on Matrigel, bevacizumab and the ONC201-bevacizumab combination did inhibit HUVEC sprouting. (Fig. 3b-c). Interestingly, regorafenib alone or in combination did not impact HUVEC sprouting (Additional file 5: Figure S5a-b).
We assessed the impact of the ONC201-bevacizumab combination therapy on tumor vascularization in vivo by performing histopathological analyses along with in vivo imaging. There have been several advancements in non-invasive imaging for monitoring vascularization within tumors. These methods can provide a means of detecting angiogenesis within the entire perimeter of a tumor and include PET studies [33]. For our purposes, we chose to use fluorescence imaging through the use of angiogenic probes. Bevacizumab and murine VEGF inhibitors decreased vascularization significantly, whereas ONC201 treatment did not impact blood flow in either mouse models. Further, ONC201 did not significantly increase the inhibition of vascularization caused by anti-angiogenic compounds. (Fig. 4a-c, Additional file 6: Figure S6). The most commonly used method of assessing the success of anti-angiogenic therapies is measuring micro-vessel density (MVD) from biopsies taken before and after treatment, evaluating certain biomarkers including CD31 [34]. However, measurement of MVD can be problematic since blocking angiogenesis can lead to reduction of tumor growth and can affect the change in MVD. As an alternative, CD31 levels within tumors of similar size can be scored [35]. Therefore, we measured CD31 in tumors of comparable size after long-term treatment or measured CD31 after short treatment before tumor reduction could occur. ONC201 did not appear to impact CD31 levels within two human xenografts and two murine colorectal tumors (Fig.  4d-e, Additional files 7: Figure S7 and 8: Figure S8).

Bevacizumab and regorafenib do not impede ONC201's TRAIL serum induction levels but do augment apoptotic fragment M30 levels
In order to determine the mechanism of synergy between ONC201 and bevacizumab, we explored whether bevacizumab might promote ONC201-induced TRAIL expression (4) or ISR activation (5). We found that bevacizumab and regorafenib do not impact ONC201's ability to induce serum TRAIL (Fig. 5a, Additional file 9: Figure S9) or CHOP/DR5 (Fig. 5b-c, Additional file 5: Figure S5 c) expression.
We assessed whether the ONC201-bevacizumab combination increases apoptosis and/or inhibits proliferation by analyzing the apoptotic caspase cleavage product M30 and Ki67 expression as a proliferation marker, respectively. A slight increase in M30 serum levels was observed in tumor tissues from combination-treated mice. In addition, a significant decrease in Ki67 was observed in mice treated with either bevacizumab-ONC2091 or regorafenib-ONC201 combination as compared to single agents (Fig. 6a-c, Additional files 10: Figure S10 and 11: Figure S11). Furthermore, ONC201 and bevacizumab in combination significantly inhibited HUVEC migration at low doses that did not induce toxicity compared to monotherapy treatment (Fig. 6d). This data supports the idea that ONC201 and bevacizumab mechanistically act independently of one another, but do increase each other's ability to promote tumor apoptosis and decrease tumor cell growth (Fig. 7).

Discussion
While as a single agent, ONC201 can be efficacious at a dose of 50 mg/kg in vivo, we have demonstrated here that in combination with the clinically approved agents bevacizumab or regorafenib, ONC201 can significantly reduce tumor growth in colorectal xenografts. Based on significant effects on tumor growth over a 4-week period, the combination of ONC201 with anti-VEGF therapies seems promising as compared to single agent therapies. Given the common use of VEGF inhibitors for combinational therapies in the clinic for CRC patients, this may be advantageous. Further, while bevacizumab and regorafenib are potent at preventing metastases from forming in mice, one might expect additive effects with the ONC201 therapies in the clinic, which should be addressed in clinical trials.
The success of ONC201 combination with murine VEGF inhibitors in syngeneic models was an important finding to demonstrate that VEGF therapies act both systemically on the host and locally within the tumor. In a direct comparison of efficacy, bevacizumab was far superior in combination with ONC201 than that or regorafenib. Further, the lack of toxicity observed in the ONC201-bevacizumab combination-treated mice, compared to the effects seen in the regorafenib-ONC201 combination mice, supports the conclusion that ONC201-bevacizumab combination would be favored for initial pursuit in the clinic.
As shown here, ONC201 has been demonstrated to not significantly impact VEGF, CD31+ cells, or angiogenesis. Further, VEGF-based compounds appear to not affect ONC201's ability to induce TRAIL, CHOP, or DR5 levels; suggesting that ONC201 and anti-angiogenic agents mechanistically act independently of one another. Importantly, combinations of ONC201 with either clinically approved agent led to a greater increase in M30 and decrease in Ki67, along with a greater inhibition of HUVEC migration. This demonstrates that while acting through two separate mechanisms, both ONC201 and combination with either bevacizumab or regorafenib enhance tumor cell death and inhibit proliferation when used in combination. Based on our results, such a trial of ONC201 and bevacizumab in combination for the treatment of CRC would include biomarkers for each drug individually, and a biomarker selected for the effect of the drugs in combination, such as caspasecleavage product M30. This is commonly used in dual-therapeutic treatments. One limitation is that we pursued in preclinical models only dosing at the same time, and other schedules should be pursued. Further, chemo-resistant tumors were not analyzed and need to be taken into consideration. Biomarker panels as discussed would aid in determining the optimal patient populations for this treatment.

Conclusion
Our data here indicates that ONC201 and bevacizumab act through independent mechanisms; with ONC201 impacting TRAIL and DR5 pathways and bevacizumab acting as an anti-angiogenic agent. In combination, they increase tumor cell death markers and decrease proliferation at a higher rate than single agent alone without inducing any signs of drug toxicity. Our proposal of ONC201 and bevacizumab in combination provides a reasonable combination therapy with non-overlapping mechanisms of action, a strategy that historically has been associated with improved efficacy in the clinic. Moreover, this strategy could reduce the emergence of resistance cells in patients in the clinic and not subject patients to toxicities common with combination chemotherapies. Thus, ONC201 and bevacizumab could be a potent combination in the clinic for metastatic CRC patients and should be pursued further in clinical trials for patients with advanced or refractory metastatic disease.

Additional file
Additional file 1: Figure S1.