The Two Mono-Carbonyl Curcumin Analogs, PGV-1 and CCA-1.1: The Chemopreventive Activity Against DMH-Induced Colorectal Cancer Rat and Proteins Target Candidate Involved

Pentagamavunone-1 and its newest derivatives, CCA-1.1, possess an outstanding cytotoxic activity against several cancer cell lines, especially colorectal cancer. We are continuing to explore the anti-colorectal cancer properties of PGV-1 and CCA-1.1 against DMH-induced colorectal cancer rats and expound the potential protein target in colorectal adenocarcinoma. We utilized DMH 60 mg/kg (subcutaneous injection once a week for 16 weeks) to induce colorectal cancer. PGV-1 and CCA-1.1 at 10 and 20 mg/kg (orally twice a week for 16 weeks) were co-administered with DMH. The WBC count increased in a single DMH group and was countered by co-administration of PGV-1 and CCA-1.1, but no signi�cant differences in RBC. Single DMH treatment for 16 weeks resulted in 80% adenocarcinoma. In contrast, co-administration with PGV-1 and CCA-1.1 suppressed most of the carcinogenic characteristics and symptoms of the pre-malignancy condition. Furthermore, bioinformatics analysis showed that CCA-1.1 has more speci�c targets than PGV-1, including CDK1, CDK2, MMP3, MMP14, and CYP3A4, which regulate the cell cycle arrest, cancer cell migration, and xenobiotic metabolism, respectively. Interestingly, CCA-1.1 targets CYP3A4, which possibly interrupts DMH metabolism and prevents the initiation of DMH-colorectal carcinogenesis. In conclusion, CCA-1.1 performed better chemopreventive effects against DMH colorectal cancer and might replace PGV-1 to be promoted as a more effective anti-colorectal cancer agent.


Introduction
Colorectal adenocarcinoma (COAD) displays a high mortality risk despite the grownup technology for diagnosis and treatment [1].Chemotherapy remains the furthermost preference due to several inconveniences and obstacles of surgery and radiation to treat accurately at the cancer site [2].However, the conventional chemotherapeutic drugs for colorectal cancer such as 5-FU, paclitaxel, and epirubicin possess numerous threats of side effects, recurrence, and metastasize induction, which subsequently impacts on patient's quality of life and survival [3].Hence, an alternative or new therapeutic agent that is more effective, comfortable, and less toxic to patients is urgently needed to improve colorectal cancer therapy.
Curcumin analog has been extensively studied for its anticancer activity in vitro and in vivo [4].Two kinds of monocarbonyl curcumin analogs, Pentagamavunone-1 (PGV-1) and Chemopreventive Curcumin Analog 1 (CCA-1.1)(Fig. 1A), were favorable as successor drug for colorectal cancer in several ways: (1) exhibiting irreversible cytotoxic effect in vitro against various cancer cell lines at a lower dose; (2) performing unique target on mitotic phase in the cell cycle; (3) effectively induce cellular senescence and death; (4) suppressing cancer migration in vitro culture system; and (5) inhibiting tumor growth in a xenograft mouse model of leukemia and breast cancer [5][6][7][8][9][10][11].Both compounds exhibit remarkable cytotoxic activities in various mechanisms such as apoptosis, senescence, autophagy, mitotic arrest, and ROS level generation on colorectal cancer cells.Those data give us a foundation for further exploring its mechanism in vivo system.
We currently conducted an in vivo experiment of PGV-1 compared to CCA-1.1 against the DMH-induced colorectal cancer rat model.We observed the cancer incidences, possible safety of the compound based on the body weight, blood pro les, and histopathological appearances.In addition, we also examined the cytotoxic activity of these compounds in vitro using two kinds of colorectal cancer cells, as well as exploration of the potential protein target candidate in COAD using bioinformatics analysis.Hopefully, these results will provide meaningful data on the potency of PGV-1 and CCA-1.1 to develop as colorectal cancer drug candidates, and more speci cally, we could determine the better suitable candidate of the two compounds.

Chemicals
PGV-1 and CCA-1.1 were obtained from the collection of CCRC, Faculty of Pharmacy UGM, Indonesia [9].

MTT assay
Cells (10 4 cells/well) were seeded into each well of 96-well plates and treated with various concentrations of CCA-1.1 and PGV-1 for 24 h.After incubation, the media were discharged, and cells were washed with phosphate-buffered saline, then added the MTT reagent (Sigma, St. Louis, CA, USA) and incubated for 4 h.Stopper reagent was added and incubated overnight, and then, the absorbances were measured on a microplate reader (Bio-Rad) under λ = 595 nm.

Animals and treatment schedules
Thirty-six male Wistar rats (150 -200 g) were obtained from Gadjah Mada University animal house, divided into 6 treatment groups (Fig. 2A): (I) untreated; (II) DMH and Na-CMC 0.5% (colorectal cancer group); and treatment group of co-administration of DMH with: (III) PGV-1 10 mg/kg BW, (IV) PGV-1 20 mg/kg BW, (V) CCA-1.1 10 mg/kg BW and (VI) CCA-1.1 20 mg/kg BW.The body weight was recorded every week.Animals were sacri ced two weeks after all treatments were completed and the colon was collected.Red blood count (RBC) and white blood count (WBC) [6] were measured using the Automatic Hematology Analyzer Sysmex KX-21.The nodules were recorded for analysis using equation 1 [12].where VN = volume of nodules, NL = nodule length, NW = nodule width.

Histopathological examination
The colon organs were xed in 10% Neutral Buffered Formalin Fixatives (Surgipath, Leica).The para nembedded tissues were cut 5 µm in serial sections using a microtome (Leica Microsystems Inc., Buffalo Grove, IL) and mounted to slides.Slides were then de-waxed by standard depara nization with xylol and followed by step-by-step rehydration (100, 95, and 70 % ethanol).Hematoxylin & Eosin (HE) staining was performed by core facilities at the Pathological Anatomy laboratories, Faculty of Medicine, Universitas Gadjah Mada.

Bioinformatic exploration of PGV-1 and CCA-1.1 in COAD
We utilized the SwissTargetPrediction (http://www.swisstargetprediction.ch/) [13] to obtain the target genes of PGV-1 and CCA-1.1.Potential gene targets in COAD were obtained from GeneCards [14].Venn diagram was used to overlap the genes and classi ed its protein class through PANTHER v.16 (http:// www.pantherdb.org/)[15].The UALCAN database [16] was used to assess the expression pro le of candidate target genes.We used the GEPIA tool [17] to predict the overall survival, using a Kaplan-Meier survival curve, applying the median cut-off and COAD dataset.The gene expression of colon cancer samples was categorized into high expression (with transcripts per million [TPM] values above the median) and low/median expression (with TPM values below the median).We re-create the graph using GraphPad Prism.

Statistical analysis
Data were presented as mean ± SEM (n = 3).We used GraphPad Prism software v6.0 (GraphPad Software, La Jolla, CA) for statistical analysis and visualization.The parametric data were analyzed using Student's t-test or one-way analysis of variance (ANOVA) with Tukey's multiple comparison post hoc test (*p<0.01;*p<0.001;ns: not signi cant).Then, the Kolmogorov-Smirnov chi-square test was used for non-parametric data.

Cytotoxic effect of PGV-1 and CCA-1.1
CCA-1.1 was more potent on colorectal cancer cells than PGV-1 but less toxic in non-cancerous cells (Fig. 1B, Table 1).In Caco-2 cells, CCA-1.1 exhibited a signi cantly stronger cytotoxic effect than PGV-1 with the IC 50 values of 2.8 and 11.2 µM, respectively.CCA-1.1 also displayed a cytotoxic effect in lower IC 50 value (2 µM) in CT-26 cells.In addition, both compounds were less toxic against NIH-3T3 and RAW 264.7 cells with IC 50 values more than 50 µM (Fig. 1C, Table 1).CCA-1.1 exhibited selectivity index (SI) values of more than 17 and 25 in Caco-2 and CT-26 cells, respectively.In comparison, PGV-1 displayed the CI value of more than 4 and 15 in Caco-2 and CT-26 cells, respectively.These ndings suggest that both compounds have excellent selectivity and are considered safe chemotherapeutic agents, but CCA-1.1 is preferred.Subcutaneous injection of 60 mg/kg DMH signi cantly decreased the body weight compared to untreated and other treatment groups (Fig. 2B).Body mass loss was prevented signi cantly (*p<0.001) on co-administration of DMH with PGV-1 and CCA-1.1, especially at the dose of 20 mg/kg.There is no signi cant difference in WBC level between DMH + PGV-1/ CCA-1.1 treatment groups and untreated groups (Fig. 2B).CCA-1.1 most effectively countered the WBC elevation caused by DMH at the dose of 20 mg/kg (**p<0.001).For the RBC level, we found no changes at all.Macroscopic examination of single DMH and co-administration of DMH with PGV-1 or CCA-1.1 displayed uctuating nodule formation in various sizes (Fig. 2C).Colorectal wall thickening was also observed, accompanying nodule formation.In more detail, single administration of DMH induced signi cantly higher (**p < 0.001) number and volume nodules than DMH + PGV-1/ CCA-1.1 (Fig. 2C).Here, CCA-1.1 at the 20 mg/kg dose was also performed as the most effective in inhibiting tumor development.These ndings give us more detailed information that PGV-1 and CCA-1.1 could inhibit tumor progression in DMH-induced colorectal cancer rats with chemoprotective properties.
3.3 Effect of PGV-1 and CCA-1.1 on adenocarcinoma, aberrant crypt, and colitis incidence Since we used DMH as a colon carcinogen that exhibited multistep carcinogenesis involving a series of pathological alterations [18], we then analyzed the incidence of adenocarcinoma, aberrant crypt, and colitis-associated colorectal cancer of all treatment groups.We could characterize between normal (blue arrow) and multistep of colorectal carcinogenesis as adenocarcinoma (green arrow), aberrant crypts (red arrow), and colitis (yellow square) (Fig. 3A) from DMH-induction and how the treatment of PGV-1 and CCA-1.1 could prevent.Single DMH administration induced 80% adenocarcinoma development throughout the colon.Co-administration with PGV-1 at the doses of 10 and 20 mg/kg BW reduced the percentage of adenocarcinoma (incidence/ group) with the value of 20 and 40%, respectively (Fig. 3B).
There is no adenocarcinoma formation on rats administered with DMH + CCA-1.1 at both doses.However, an early characteristic of colorectal carcinogenesis as the aberrant crypt and colitis-associated colorectal cancer in DMH + PGV-1/ CCA-1.1 groups was higher than the single DMH group.At the 20 mg/kg BW dose, CCA-1.1 is more e cient in inhibiting colorectal carcinogenesis than PGV-1 at the same doses.We believe that adenocarcinoma formation is always accompanied by a multi-combination of aberrant crypts, colitis, and in ammation [19].Still, we have not counted it as an individual case of aberrant crypts or colitis and only focused on the representative pathological incidence of each colon/ rat to evaluate the carcinogenic step of DMH induction.Overall, both PGV-1 and CCA-1.1 could suppress the carcinogenesis of DMH-induction in rats, particularly CCA-1.1.

Bioinformatic exploration of PGV-1 and CCA-1.1 on COAD
For further comprehension of PGV-1 and CCA-1.1 targets in COAD, bioinformatic analysis elucidate that CCA-1.1 targets more genes than PGV-1 (Fig. 4A).The sixteen CCA-1.1 target genes were classi ed into a variety and rich protein classes than the ve target genes of PGV-1 (Fig. 4B).We then emphasize the ve genes which are speci cally involved in cancer metabolism (CYP3A4), cell proliferation (cell cycle regulator) (CDK1 and CDK2), and cell migration (MMP3 and MMP14) as the potential target of CCA-1.1.We found that CDK1, CDK2, MMP3, and MMP14 were signi cantly overexpressed in tumors than in normal colon tissues (*p<0.01;**p<0.0001)(Fig. 4C).CYP3A4 was higher expressed in normal tissue than in the tumor, con rming that this gene plays a role as a xenobiotic metabolizer to counter the pathological cascade [20,21]; in this case, DMH-carcinogenesis.On the other hand, the overall survival was not signi cantly different between high and low expression of those respective genes for 150 months span (Fig. 4C).Although not statistically different, COAD patients who had increased MMP14 and CYP3A4 intended to have a higher risk of relapse than patients with lower expression.By looking at this information, we believed that CCA-1.1 target genes in these results were unique and interesting points to explore and strongly correlated with in vitro and in vivo experiments.
In the in vitro study, we con rm that CCA-1.1 was more potent than PGV-1 and underlined with a superior safety level.Caco-2 and CT-26 colorectal cancer cells represent the COX-2 level in colorectal cancer since there was uctuate expression of COX-2 in COAD patients [22,23].The different characteristics of the two cells that we used provide different responses for each cell.We found compelling evidence that PGV-1 cellular response, in this case, cytotoxic activity, might be altered by COX-2 expression.Further con rmation on the selectivity index of CCA-1.1 and PGV-1 indicate that both compounds are promising candidates of selective anticancer agents.It was also con rmed by the WBC and RBC level in the in vivo experiment that PGV-1 and CCA-1.1 might defeat tumor formation with fewer or no observable opposing effects on the normal lineage of cells.The RBC and WBC pro les could explain a slight pathological condition of colorectal cancer, such as anemia, cruor, and in ammation [1,24].The safety properties of CCA-1.1 and PGV-1 are important to develop new anticancer agents.However, this study is still limited in the normal proliferating immortal cells that should be evaluated in the normal primary cells as well as in the in vivo model.
We performed the chemopreventive examination of PGV-1 and CCA-1.1 that simultaneously administered with DMH for inducing colorectal cancer, whether they were able to reverse, suppress, or prevent the initial phases of DMH carcinogenesis.We believed that the study of in vivo colorectal carcinogenesis had a remarkably long journal accompanying the development of diagnostic, pathway mechanism, and therapy of cancer [25].For the inducing agent, we used DMH as a procarcinogen that metabolizes to a methyl free radical and generates hydroxyl radicals of metal ions resulting in lipid peroxidation, molecular genetic alteration, and cancer initiation [26].We found that co-administration of PGV-1 and CCA-1.1 in both doses (10; 20 mg/kg) with DMH were adequate to suppress tumor growth with noticeable toxic effects to the normal lineage cells caused by DMH, particularly CCA-1.1 at 20 mg/kg.The mechanism may, in part concern with the in vitro anticancer activities of PGV-1 and CCA-1.1, including induction of apoptosis, targeted cell cycle arrest, related cancer marker protein inhibition, and selective on cancer cells [7,8].Then, PGV-1 and CCA-1.1 may play a role as free radical eliminators through enzymatic antioxidants since curcumin analogs are known as dual-oxidant on their anticancer features [27].This phenomenon should be an exciting point to further exploration related to their antioxidant activities.
Bioinformatic exploration was con rmed that CCA-1.1 was more potent than PGV-1.The proteins target candidate of CCA-1.1 in COAD was richer than PGV-1, validating previous in vitro observation of CCA-1.1 that report several activities such as apoptosis, anti-migration/ metastasis, and speci c G2/M cell cycle arrest [11,[28][29][30][31][32].We underlined the two proteins CDK1 and CDK2, which play a speci c role in the G1/S and G2/M cell cycle stage, respectively [33].Those two proteins might become an essential protein in the anticancer effect of CCA-1.1 in the eld of cell proliferation and cell cycle regulation on colorectal cancer.On the other hand, targeting MMP3 and MMP14 [34] might also underline the better effect on antimigration of CCA-1.1 than PGV-1.CCA-1.1 targets CYP3A4 should be highlighted as superior to PGV-1.By targeting this protein, CCA-1.1 possibly suppressed or inhibited DMH-carcinogenesis since DMH is mainly metabolized into toxic metabolites by CYPs in the liver, including CYP3A4 [20,21].We believed that this nding could emphasize the superiority of CCA-1.1 compared to PGV-1 to develop as a colorectal cancer chemotherapeutic candidate.An additional laboratory experiment is required for a better and detailed comprehension of the mechanism.
Since we used the DMH-induced colorectal cancer model, we could also observe an early pre-neoplastic hyperproliferative lesion that formed aberrant crypt foci (ACF) as intermediary indicators of carcinogenesis [35].Here, we focused on the late and early stages of colorectal carcinogenesis to de ne the effect of suppression or inhibition of PGV-1 and CCA-1.1.We found that single DMH induced late stage of colorectal cancer, which is adenocarcinoma with severe alteration of goblet cell's structure, in ammation, and invasive growth of cancer cells.However, co-administration of DMH with PGV-1/ CCA-1.1 induced an early stage of colorectal carcinogenesis, such as aberrant crypt [36] and colitis-associated colorectal cancer (severe in ammation and hemorrhagic) [37], without adenocarcinoma formation, especially CCA-1.1 at both doses.These ndings indicated that PGV-1 and CCA-1.1 could suppress or inhibit the carcinogenesis of DMH in rats.It should also be superior to a new anti-colorectal cancer due to its solubility, stability, and effectiveness, especially in an acidic environment since both compounds were administered orally to rats.However, several conventional chemotherapeutic drugs are still limited for oral use due to their stability in the alimentary tract [38][39][40].In this study, we used a chemopreventive setting on the treatment schedule, in which DMH was co-administered with PGV-1/ CCA-1.1 at the same period.Further exploration using another treatment design would be great to establish the detailed mechanism of those compounds and the possibility to pharmaceutically developed as orally administered drugs for colorectal cancer.
To our knowledge, it is the rst time that in vivo exploration of PGV-1 and CCA-1.1 against colorectal cancer has been conducted along with in vitro and bioinformatic exploration.These ndings revealed a great potential of PGV-1 and CCA-1.1 to develop as anti-colorectal cancer agents.However, the optimum dose to give as well as the treatment setting may not be answered yet.A deeper investigation is needed to settle down the potency of PGV-1 and CCA-1.1 as anti-colorectal cancer agents.

Conclusion
co-administration of PGV-1 and CCA-1.1 (10 and 20 mg/kg) for 16 weeks signi cantly inhibited colorectal carcinogenesis, compared to single DMH and untreated groups without toxicity in normal lineage cells.The suppressing effect of PGV-1 on COAD development is suggested to target signal transduction and cell cycle, whereas CCA-1.1 targets cell cycle, signal transduction, cell migration, and carcinogen metabolism.CCA-1.1 performed a better chemopreventive effect as a new anti-colorectal cancer candidate against DMH colorectal cancer than PGV-1, possibly supported by the additional target on CYP3A4.In general, both compounds exhibited strong chemopreventive potential against colorectal cancer.

Figure 2 Co
Figure 2