Anti-colon cancer effect of caffeic acid p-nitro-phenethyl ester in vitro and in vivo and detection of its metabolites

Caffeic acid phenethyl ester (CAPE), extracted from propolis, was proven to inhibit colon cancer. Caffeic acid p-nitro-phenethyl ester (CAPE-pNO2), a derivative of CAPE, was determined to be an anti-platelet agent and a protector of myocardial ischaemia with more potent effects. In the present study, CAPE-pNO2 showed stronger cytotoxic activity than CAPE. We revealed interactions between CAPE-pNO2 and experimental cells. CAPE-pNO2 induced apoptosis in HT-29 cells by up-regulating P53, cleaved-caspase-3, Bax, P38 and CytoC; CAPE-pNO2 also up-regulated P21Cip1 and P27Kip1 and down-regulated CDK2 and c-Myc to promote cell cycle arrest in G0/G1. In xenograft studies, CAPE-pNO2 remarkably suppressed tumour growth dose dependently and decreased the expression of VEGF (vascular endothelial growth factor) in tumour tissue. Moreover, HE staining showed that no observable toxicity was found in the heart, liver, kidney and spleen. In addition, metabolites of CAPE-pNO2 in HT-29 cells and organs were detected. In conclusion, para-nitro may enhance the anticancer effect of CAPE by inhibiting colon cancer cell viability, inducing apoptosis and cell cycle arrest via the P53 pathway and inhibiting tumour growth and reducing tumour invasion by decreasing the expression of VEGF; additionally, metabolites of CAPE-pNO2 showed differences in cells and organs.

. Chemical structure of the compounds used in the present study and results of the MTT assay. (A) Chemical structure of CAPE and CAPE-pNO 2 . HT-29 cells (B) and HCT-116 cells (C) were treated with CAPE and CAPE and CAPE-pNO 2 for 48 h, and the expression of P53 after treatment by PFT-α. Values represented the means ± SD from three independent experiments, and error bars represented the STDEV (SD).
Scientific RepORTs | 7: 7599 | DOI: 10.1038/s41598-017-07953-8 CAPE and CAPE-pNO 2 induce cell cycle arrest in G0/G1 in colon cancer cells. The cell cycle distribution was detected by flow cytometry. As shown in Fig. 3, the number of HT-29 cells (Fig. 3A) in G0/G1 phase increased in a dose-dependent manner from 31.9% to 77.3% and 86.5% after treatment with CAPE and CAPE-pNO 2 , respectively. Concerning HCT-116 cells (Fig. 3B), the G0/G1 phase increased from 31.0% to 80.1% and 84.5%, but the G2/M phase was barely changed in the present study. The results proved that the progression of cells from G1 to the S phase was interrupted more obviously by CAPE-pNO 2 than by CAPE (p < 0.01) (Fig. 3C,D). CAPE and CAPE-pNO 2 regulate the expression of P53 signalling pathway related proteins in HT-29 cell and tumours. The expression of related proteins in HT-29 cells was measured by western blot assay after treatment with CAPE and CAPE-pNO 2 (Fig. 4). The results showed that these two drugs could down-regulate the expression of pro-caspase-3, which was reduced by 44.0% and 79.0% after treatment with From the microscopic vision fields, the cell number was decreased, and fluorescence was increased by treatment for 48 h in a dose-dependent manner. Values represented the means ± SD from three independent experiments, and error bars represented the STDEV (SD). *p < 0.05, **p < 0.01: CAPE and CAPE-pNO 2 compared with the control. # p < 0.05, ## p < 0.01: CAPE-pNO 2 compared with CAPE at the same concentration.

Discussion
CAPE is a bioactive natural ingredient exacted from propolis and exhibits inhibition activity against various cancers, such as prostate cancer, breast cancer and colon cancer. In a previous study, the inhibitory effect of CAPE on colon cancer cells was determined. Wang D. et al. 26 reported that CAPE induced HCT-116 cell cycle arrest in G0/ G1 and apoptosis by decreasing the expression of β-catenin. Xiang D. et al. 27 suggested that CAPE inhibited the proliferation of HCT-116 cells and SW480 cells via the β-catenin/T-cell signalling pathway by down-regulating cyclin D1 and c-myc, and En-Pei Isabel Chiang et al. 28 discovered that caffeic acid phenylpropyl ester (CAPPE), a derivative of CAPE, could inhibit the growth of the HCT-116 and SW480 cells more significantly than CAPE through the PI3K/AKT and AMPK signalling pathways in vivo and in vitro. He, Y. J. et al. 3 also proved that CAPE induced SW480 cell apoptosis by down-regulating PSMA1 and PSAT1 while up-regulating GNPDA1 and GPX-1. These studies proved that the signalling pathway that mediated the apoptosis of colon cancer cells and tumour The cell cycle-related protein expression levels of P21 Cip1 , P27 Kip1 , P53, CDK2 and c-Myc were determined by western blotting. The blots were representative of three independent experiments. Data represented the means ± SD from three independent experiments, and error bars represented the STDEV (SD). *p < 0.05, **p < 0.01: CAPE and CAPE-pNO 2 compared with the control. # p < 0.05, ## p < 0.01: CAPE-pNO 2 compared with CAPE at the same concentration.
after treatment with CAPE was not unique. P53 is one of the tumour suppressor proteins that has an intimate connection with the occurrence and progression of many tumours in humans, and it can induce apoptosis and cell cycle arrest of carcinoma 29 . The result in the above literature was not involved in the P53 signalling pathway after treatment with CAPE in colon cancer. By contrast, the P53 signalling pathway was selected to explore the anti-cancer mechanism of CAPE and CAPE-pNO 2 in HT-29 cells.
CAPE-pNO 2 , a derivative of CAPE like CAPPE but with quite a different chemical structure, was synthesized by adding a nitro moiety to the para position in our laboratory. It was proven to be a platelet anticoagulant for collagen-induced platelet aggregation and a protector of acute myocardial ischaemia-reperfusion injury in previous studies 24,25 , and these effects of CAPE-pNO 2 were more potent than those of CAPE, but there is no report concerning the anti-cancer effect of CAPE-pNO 2 .
In this study, the MSI-type cell line HCT-116 and the MSS-type cell lines HT-29 and SW480 were selected to verify the anti-cancer effect of CAPE and explore the anti-cancer effect of CAPE-pNO 2 in colon cancer. In the MTT assay, CAPE and CAPE-pNO 2 inhibited cell proliferation in a dose-dependent manner, and the IC 50 of CAPE-pNO 2 was lower than that of CAPE. Moreover, the inhibitory effect of CAPE and CAPE-pNO 2 was weaken when the expression of P53 was decreased by PFT-α, this result proved that P53 signalling pathway was one of the approaches which inhibited the proliferation of HT-29 and HCT-116 cells. Hoechst 33342 staining showed that the number of cancer cell was decreased, and the fluorescence intensity was increased with increasing concentrations of drugs. Compared with the CAPE treated group, after treated with CAPE-pNO 2 , the apoptosis rate of HT-29 cell and HCT-116 cells increased 15 and 8 percentage points (Fig. 2), the number of HT-29 cells and HCT-116 cells in G0/G1 phase were increased 10 and 5 percentage points (Fig. 3), these data showed that CAPE-pNO 2 exhibited stronger inhibitory effect on HT-29 cells than that on HCT-116 cells. Based on the above results, we selected HT-29 cells of MSS type cell lines for the further study on anti-colon cancer mechanism of CAPE-pNO 2 in vivo and in vitro.
The results of western blotting showed that both CAPE and CAPE-pNO 2 could up-regulate P53, P38, Bax, CytoC and cleaved-caspase-3 and down-regulate pro-caspase-3. These proteins associated with the P53 signalling pathway were closely connected to the occurrence and growth of tumours 30 , and up-regulated CytoC proved that CAPE and CAPE-pNO 2 could induce the mitochondrial apoptotic and active pro-caspase-3 to become cleaved-caspase-3, our results were consistent with those of Liu X. et al. 31 .
In a study of the cell cycle, CAPE and CAPE-pNO 2 induced colon cancer cell cycle arrest in the G0/G1 phase; up-regulated P53, P21 Cip1 and P27 Kip1 ; and down-regulated CDK2 and c-Myc. P53 mainly promoted tumour cell apoptosis and induced cell cycle arrest. When the cells of the body are damaged, P21 Cip1 mRNA and protein expression levels are elevated following activation by P53 protein, and the cell cycle is blocked in G1, G2 or S phase 32,33 . Down-regulated CDK2 by CAPE and CAPE-pNO 2 treatment might induce pRb dephosphorylation to promote cancer cell ageing and prevent cell cycle progression from G1 to S phase 34,35 . Up-regulated c-Myc induces cell cycle arrest in the G1 phase and inhibits the repair effects on telomeres, preventing the cells from being immortalized 36  induces up-regulation of Bax, cleaved-caspase-3, CytoC, P53, P38, P21 Cip1 and P27 Kip1 and down-regulation of pro-caspase-3, CDK2 and c-Myc. In one word, CAPE and CAPE-pNO 2 inhibited proliferation of cells and suppressed tumours growth by regulating the P53 signalling pathway, and CAPE-pNO 2 is more effective than CAPE in inhibiting cell growth, inducing apoptosis and cell cycle arrest in G0/G1 and suppressed tumours growth.
To explore the anticancer effect of CAPE-pNO 2 in vivo, HT-29 cells were xenografted into nude mice. The turning point of the tumour growth curve appeared on the 37th, 35th and 33rd days after treating with CAPE-pNO 2 at doses of 5, 10 and 20 mg/kg/day, respectively. However, the tumour growth curve in the CAPE (10 mg/kg/day) group showed a relatively steady trend on the 41st day. Based on the results in vitro, we used HT-29 cells to establish xenograft models. According to the report by En-Pei Isabel Chiang 28 , after injecting HCT-116 cells into nude mice, CAPE treatment lasted for six weeks. Although CAPE markedly inhibited the tumour growth, the growth trend of the treatment group showed a downward trend, a finding that was different from ours. The cause of this discrepancy might be the different cell lines used. Additionally, Wu J 20 reported that the growth curves of xenograft tumours using MDA-231 cells and MCF-7 cells after treatment with CAPE were also different. At the end of experiment, all nude mice were euthanized, and the tumours were removed. The results of HE and TUNEL staining showed that CAPE and CAPE-pNO 2 inhibited tumour growth through inducing tumour tissue necrosis and apoptosis. More importantly, CAPE-pNO 2 exhibited more potent effects than CAPE on tumours. Interestingly, no morphological changes were found in the heart, liver, spleen and kidney after treatment with CAPE and CAPE-pNO 2 for a long time. In clinical settings, many drugs used to cure colorectal cancer, such as 5-fluorouracil, have serious toxic and side effects and can even lead to patient death 39,40 . Thus, CAPE-pNO 2 might have great clinical application value. Meanwhile, the results of immunohistochemistry indicated that CAPE and CAPE-pNO 2 decreased the expression of VEGF to disturb the pervasion and growth of tumours, while there was almost no expression of VEGF in normal colon tissues, and many reports have shown that the combination of VEGF with tyrosine kinases and neuropilins on the tumour cell surface promoted the progress of tumour invasion and cancer stem cell formation [41][42][43] , and VEGF could be related to the survival of patients with colorectal carcinoma and should be considered a predictor of the prognosis clinically 44 . Thus, CAPE-pNO 2 may be regarded as a better inhibitor of VEGF in colon tumours (p < 0.01).
The nitro group at the para position was the only difference between CAPE and CAPE-pNO 2 . Consequently, our results imply that the anticancer effects of CAPE were enhanced by the para-nitro moiety. Similarly, it was confirmed that para-nitric oxide-donating acetylsalicylic acid was more purposeful in chronic lymphocytic leukaemia cells and more applicable to clinical treatment than NO-ASA 45 . For further study on the para nitro, LC-MS/MS was applied to investigate the difference in metabolites between CAPE and CAPE-pNO 2 in HT-29 cells and in organs (tumour, heart, liver, spleen and kidney). In our results, the main difference is that CAPE can combine with the glucose acid, while para nitro-benzene alcohol combined with glucose acid after the hydrolysis of CAPE-pNO 2 . In CAPE, caffeic acid from CAPE hydrolysis was methylated; however, in CAPE-pNO 2 , caffeic acid from CAPE-pNO 2 hydrolysis combined with L (+)-cysteine (maybe from the HT-29 cell line). C 17 H 15 NO 5 only occurred in the metabolite of CAPE-pNO 2 .
On the other hand, the metabolites of CAPE and CAPE-pNO 2 in HT-29 cells were different in the heart, liver, spleen and kidney. CAPE was transformed into C 13 NO 7 . The differences in the metabolites in cells and organs may be due to the discrepancy of the species of metabolic enzyme and their types. However, there were no metabolites detected in tumours, implying that CAPE and CAPE-pNO 2 were transformed to other compounds in tumours and need to be further studied in the future. Additionally, according to the pharmacokinetic analysis of CAPE and CAPE-pNO 2 in rats, the half-lives (t 1/2 ) of these compounds were 4.2 h and 20.9 h, respectively, and the area under concentration-time curve (AUC all ) values were 1659 and 3239, respectively, indicating that the bioavailability of CAPE-pNO 2 was higher than that of CAPE, and the metabolic processes of CAPE and CAPE-pNO 2 would be different 46 . The most of above metabolites are inactive excrement of CAPE or CAPE-pNO 2 , and there are no literatures reported the anti-cancer effect of these metabolites detected in this study, the results of LC-MS/MS provided a basis for further study to explore the compounds to against colon cancer with stronger effect only.

Conclusion
In this study, para-nitro enhanced the anti-colon cancer activity of CAPE, and the present data showed that CAPE-pNO 2 is more effective than CAPE in inducing colon cancer cell death, apoptosis and cell cycle arrest in the G0/ G1 phase by regulating the relative proteins in the P53 pathway and inhibiting tumour growth. Moreover, CAPE-pNO 2 significantly decreased the expression of VEGF. The metabolites of CAPE-pNO 2 were different from those of CAPE in HT-29 cells and organs. This study contributes to further development of the pharmacological activity of CAPE-pNO 2 .
Cell Viability Assay. HT-29 and HCT-116 cells were seeded at a density of 3 × 10 3 cells per well in 96-well plates and were incubated for 24 h. Next, different concentrations CAPE and CAPE-pNO 2 were administered. For pifithrin-α (PFTα, P53 inhibitor) group, cells were treated with the above concentration of CAPE and CAPE-pNO 2 after pre-treatment with 5 μM PFTα, and then the expression of P53 was detected by Western Bolt. After treatment with CAPE and CAPE-pNO 2 for 48 h, MTT (5 μg/mL) was added to each well for 4 h at 37 °C. The amount of MTT crystals (formazan) dissolved by DMSO was measured using a Plate Reader (Bio Tek) at 570 nm. All experiments were repeated three times, and IC 50 values were calculated.
Hoechst 33342 staining. Hoechst 33342 was dissolved in PBS, and the final concentration was 10 μg/mL.
HT-29 and HCT-116 cells were treated with different concentrations of CAPE and CAPE-pNO 2 after seeding in 6-well plates (8 × 10 4 per well). After 48 h, the cells were rinsed three times with PBS, and Hoechst 333342 was added. The cells were then incubated at 37 °C in the dark for 30 min. Thereafter, the dye liquor was removed, and the cells were washed three times with PBS. A fluorescence microscope (Olympus U-RFLT50, Tokyo, Japan) was used for the examination of each well at 6 different fields of view.
Cell apoptosis and cell cycle analysis by flow cytometry. HT-29 and HCT-116 cells were seeded in 6-well plates at a density of 8 × 10 4 per well. After 24 h, different concentrations CAPE and CAPE-pNO 2 were added. Next, floating cells (digested by trypsin) were collected after treatment with CAPE and CAPE-pNO 2 for 48 h. Cells were re-suspended in 500 μL of binding buffer after washing twice with PBS. Next, 5 μL of Annexin V-FITC and 5 μL of PI were added, and the cells were incubated for 1 h in the dark. Apoptosis was measured with a FACSCalibur flow cytometer (Keygen Biotech, Co Ltd, Nanjing, China). For cell cycle analysis, after treatment for 48 h, the cells were collected and then fixed with 70% ice-cold ethanol at 4 °C for 24 h. The stationary liquid was removed through centrifugation, and then 100 μL of RNase A and 400 μL of PI were added, followed by incubation at room temperature for 30 min without light. Next, a FACSCalibur flow cytometer (Becton-Dickinson) equipped with a 488-nm argon laser was used, and cell cycle analysis was performed using Cell Quest software and ModFit.
Western Blotting Analysis. The proteins of HT-29 cells were extracted after treatment with RIPA buffer and 1 mM PMSF (phenylmethylsulfonyl fluoride) on the ice for 30 min, the proteins of tumours were extracted after homogenate with RIPA buffer and 1 mM PMSF at 4 °C for 30 min. Proteins with different molecular weights were separated on 6.0%, 8.0% and 12% SDS-polyacrylamide gels. The proteins on the gel were transferred to a PVDF membrane, which was soaked for 10 seconds with methanol. Next, 5.0% skim milk prepared with PBST (0.1% Tween-20 and 99.9% PBS) was used to block the membrane for 1.5 h. Thereafter, the blots were incubated with primary antibody overnight at 4 °C. The blots were then incubated with the corresponding HRP-linked secondary antibody for 1.5 h after washing with PBST three times. The PVDF membranes were developed by ECL (enhanced chemiluminescence). The protein expression levels were determined by the grey values from the PVDF membrane that were calculated using software Quantity One.
Xenografts in athymic mice. Four-week-old Male BALB/c nude mice were purchased from the Beijing HFK Bioscience Co Ltd. (Beijing China) and were placed in specific pathogen-free (SPF) conditions. Trypsinized HT-29 cells were injected into the right flanks of athymic mice at a density of 2 × 10 7 /0.1 mL. Water and food for mice were sterilized. When the size of the tumour was approximately 100 mm 3 , the mice were divided into five groups of 10 mice each. These five groups were the control group, CAPE group (10 mg/kg/day), and CAPE-pNO 2 groups (5 mg, 10 mg and 20 mg/kg/day). The tumour volumes were measured every 3 days using a Vernier calliper and were quantitated according to the formula (length × width 2 × π)/6 48 ; the inhibition rate was calculate by the formula inhibition rate (%) = [1 − (the weight of tumours in the experimental/the weight of tumours in control)] × 100%. The tumours were removed and used for detecting other biochemical indexes after CAPE and CAPE-pNO 2 were given for 42 days.
Scientific RepORTs | 7: 7599 | DOI:10.1038/s41598-017-07953-8 Haematoxylin and Eosin staining. After treatment, tumours and organs were removed from nude mice and were fixed with 4.0% paraformaldehyde. Paraffin sections were dewaxed to water for different experiments. For HE (haematoxylin and eosin) staining, sections were put into haematoxylin for 8 min and then washed twice before incubating with eosin for 3 min. Finally, the sections were sealed with neutral gum. Nuclei appeared bluish, while the cytoplasm appeared red through the microscope. TUNEL staining. Paraffin sections were covered with broken membrane fluid for 20 min. The TUNEL kit reagents were added to the sections for 2 h at 37 °C while maintaining the humidity. Next, endogenous peroxidase was blocked by 3.0% hydrogen peroxide solution prepared in methanol for 15 min. DAB chromogen was used for staining; when positive cells became brown, the staining was stopped by washing with distilled water. Finally, the nuclei were stained with haematoxylin before sealing.
Immunohistochemistry. Antigen repair buffer (pH 9.0) was added before blocking endogenous peroxidase. The sections were sealed with 3.0% BSA for 30 min. The sections were incubated with the primary antibody (VEGF) overnight at 4 °C. After incubating with PBS three times, the sections were incubated with the corresponding secondary antibody for 50 min. Next, DAB and haematoxylin were used for staining as previously described. The IOD value measured the expression of VEGF.
Statistical analysis. All data were analysed by SPASS 16.0, the data normality was verified by K-S test, and if the P value was more than 0.05, the data obeyed a normal distribution. The data were presented as the means ± SD for at least three independent experiments. One-way ANOVA and Student's t-test were performed for statistical analysis. A P value less than 0.05 was considered statistically significant.

Ethics Statement.
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals from the National Institutes of Health. All animal procedures were approved by the Ethical Committee for Animal Experiments of Southwest University (Permit Number: SYXK 2015-0002). All efforts were made to minimize suffering.