Polyphenols in Colorectal Cancer: Current State of Knowledge including Clinical Trials and Molecular Mechanism of Action

Polyphenols have been reported to have wide spectrum of biological activities including major impact on initiation, promotion, and progression of cancer by modulating different signalling pathways. Colorectal cancer is the second most major cause of mortality and morbidity among females and the third among males. The objective of this review is to describe the activity of a variety of polyphenols in colorectal cancer in clinical trials, preclinical studies, and primary research. The molecular mechanisms of major polyphenols related to their beneficial effects on colorectal cancer are also addressed. Synthetic modifications and other future directions towards exploiting of natural polyphenols against colorectal cancer are discussed in the last section.


Introduction
Epidemiological studies exhibiting protective effect of diets rich in fruits and vegetables against different types of cancer have drawn increased attention to the possibility of exploiting biologically active secondary metabolites of plants to fight against cancer. Among the vast array of phytochemicals, compounds called "polyphenols" constitute one of the most numerous and widely distributed groups, covering more than 10,000 different chemical structures [1]. Polyphenols (PP) are reported to have antioxidant, anticarcinogenic, antiatherosclerotic, anti-inflammatory, spasmolytic, hepatoprotective, antiviral, antiallergic, antidiarrheal, antimicrobial, and oestrogenic activity [2].
Colorectal cancer (CRC) is the third most common diagnosed cancer in men after lung and prostate cancer throughout the world. While in women CRC occupies the second position after breast cancer worldwide. Prevalence of CRC is 18% higher in developed countries than developing and undeveloped nations. People of more than 50 years old are more prone to be affected by CRC, and incidence in males is greater than in females. Although diet and Western lifestyle are still considered as being the main factors responsible for CRC, no specific food or other environmental agent has been identified as an exact causative factor [3]. Thus far, clearly identified types or causes of CRC are hereditary nonpolyposis colorectal cancer, familial adenomatous polyposis, inflammatory bowel diseases, human papillomavirus, and acquired immunodeficiency syndrome [4]. Although surgical resection remains the only curative treatment for CRC, an alternative approach to reduce the mortality rate is chemoprevention, use of synthetic or natural compounds in pharmacologic doses [5].
Colon cancers result from a series of pathologic changes that transform normal colonic epithelium into invasive carcinoma. Dietary PP affect these different cellular processes by acting as chemopreventive blockers. So far, only one review article that has been published concentrated on the effect of polyphenols on colorectal cell lines [6], and only a limited number of polyphenols have been considered. This review focuses on the updated research on a wider variety of polyphenols as applied to colorectal cancer.   PP has been given in Figure 1 on the basis of the chemical structures of the aglycone portions and Figure 2 gives the basic structures of major groups [7].

Chemistry of PP and Their Dietary Sources
A list of the 100 richest dietary sources of PP has been produced using comprehensive Phenol-Explorer data [8]. The richest sources are various spices and dried herbs, cocoa products, some dark coloured berries, some seeds (flaxseed) and nuts (chestnut, hazelnut), and some vegetables, including olive and globe artichoke heads. Top ten of the list containing the highest amount of PP is in the following order: cloves > peppermint (dried) > star anise > cocoa powder > Mexican oregano (dried) > celery seed > black chokeberry > dark chocolate > flaxseed meal > black elderberry.

Pathogenesis of CRC and Its Signalling Pathways
Acquired functional capabilities of cancer cells that would allow them to survive, proliferate, and disseminate are known as the hallmarks of cancer, that is, sustaining proliferative signalling, evading growth suppressors, resisting cell death, enabling replicative immortality, inducing angiogenesis, activating invasion and metastasis, reprogramming of energy metabolism, and evading immune destruction [9]. Underpinning these hallmarks are genomic instability and inflammation. While genomic instability confers random mutations including chromosomal rearrangements, causing genetic diversity that expedites the acquisition of hallmarks of cancer, the inflammatory state of premalignant and frankly malignant lesions that is driven by cells of the immune system also fosters multiple hallmark functions.
Based on investigation of different stages of tumour initiation and progression, Fearon and Vogelstein proposed a model of colorectal carcinogenesis that correlated specific genetic events with evolving tissue morphology [10]. The Wnt/ -catenin pathway plays a dominant role in an initial stage of CRC development. Inactivation of the adenomatous polyposis coli gene is a key starting event in carcinogenesis of more than 60% of colorectal adenomas and carcinomas leading to stimulation of the Wnt pathway via free -catenin [10].
Stimulation of the epidermal growth factor receptor (EGFR) leads to the activation of KRAS or phosphatidylinositol-3-kinase pathways, which is important in CRC development from early adenoma to intermediate adenoma.
Subsequently, numerous signal transduction molecules initiate a cascade of downstream effectors that trigger tumour growth, angiogenesis, and metastasis [11].
Transforming growth factor-(TGF-) is a multifunctional polypeptide that binds to specific TGF-receptors for paracrine and autocrine signalling. This ligand and receptor complex triggers intracellular signalling cascades that include the canonical Smad2 signalling pathway, which complexes with Smad4 and accumulates and translocates into the nucleus. In the nucleus, activated Smad complexes regulate the transcription of specific genes and ultimately regulate cell cycle and tissue repair [12]. TGF pathway contributes to a favourable microenvironment for tumour growth and metastasis throughout all the steps of carcinogenesis [13]. TGF-also induces apoptosis, from the association of death-associated protein 6 (DAXX) with the death receptor Fas. After binding, DAXX is then phosphorylated by   BioMed Research International homeodomain-interacting protein kinase 2 (HIPK2), which then activates apoptosis signal-inducing kinase 1 (ASK1). ASK1 activates the Jun amino-terminal kinase (JNK) pathway that causes apoptosis [14][15][16]. Inactivation of TGF-beta pathway components is first detected in advanced adenomas and affects 40-50% of all CRCs [17].
Almost 50% of all CRCs show p53 gene mutations, with higher frequencies observed in distal colon and rectal tumours and lower frequencies in proximal tumours and those with the microsatellite instability or methylator phenotypes [18]. The mutations in p53 or the loss of its functionality occurs mainly at the transition from adenoma to cancer, and the frequency of alterations in the gene increases with the corresponding progression of the lesion [19].
CRC cells share many properties in common with stem cells which are conserved in both dormant and actively proliferating cancer cells [20]. On top of maintaining "stemness" characteristics, CRC cells with metastatic potential dissociate from the tumour mass and spread to other organs in the body [21]. This is achieved through a dedifferentiation program called epithelial-mesenchymal transition (EMT). This key developmental program allows stationary and polarized epithelial cells to undergo multiple biochemical changes that enable them to disrupt cell-cell adherence, lose apical-basal polarity, dramatically remodel the cytoskeleton, and acquire mesenchymal characteristics such as enhanced migratory capacity, invasiveness, and elevated resistance to apoptosis [22]. Adhesion molecules that maintain cell-cell contact in the differentiated tumour cells, such as E-cadherin, are downregulated in the undifferentiated cells, while molecules that impart invasive and migratory behaviour would be upregulated. To accommodate both the "stemness" and mesenchymal properties of invasive CRC cells, it has been proposed that CRC cells with metastatic potential are like "migratory stem cells" [23]. The EMT process is initially driven by three core groups of transcriptional regulators described as follows. The first is a group of transcription factors (TFs) of the Snail zinc-finger family, including SNAI1 and SNAI2 (SLUG) [24]. The second group is the distantly related zinc-finger E-box-binding homeobox family of proteins ZEB1 and ZEB2 (SIP1) [25]. The third group is the basic helixloop-helix (bHLH) family of transcription factors, including TWIST1, TWIST2, and E12/E47 [26]. In CRC, 85% of resected specimens have moderate to strong TWIST1 expression [27]. The earlier steps of the metastatic cascade EMT program include local invasion, intravasation, survival while transiting through the circulation, and extravasation. EMT programs are dynamically regulated, and during the last step of the metastatic cascade, colonization, carcinoma cells are thought to switch back to an epithelial state through the reverse process, mesenchymal-epithelial transition (MET) [28]. The final stage of the invasion-metastasis cascade, colonization, is likely to require adaptation of propagated CRC cells to the microenvironment of a distant tissue [29].
Increased matrix metallopeptidases (MMPs) expression and their activation generally promote hallmarks of CRC progression including angiogenesis, invasion, and metastasis and correlate with shortened survival. MMPs comprise a large family of at least 25 zinc-dependent endopeptidases capable of degrading all components of the extracellular matrix (ECM) and are categorized primarily by their structural features as gelatinases, collagenases, membrane-type, stromelysins, and matrilysins [30]. Intercellular adhesion molecule-1 (ICAM-1) is a 90-kDa cell surface glycoprotein that is known to be a member of the immunoglobulin gene superfamily of adhesion molecules. ICAM-1 expression is closely associated with metastasis and may be a useful indicator of prognosis in patients with colorectal cancer [31].
It is evident from the above discussion that the pathogenesis of CRC is characterized by regulatory pathways that are complex involving several layers of communication, cascades, crosstalk, and extensive networking. CRC usually develops through interaction of cytokines, the chemical mediators of inflammation; cytokine receptors, present on the surface of a variety of cell types; secondary messengers which convey signals from cell surface to the interior; transcription factors, which regulate the expression of several genes that affect CRC. Figure 3 depicts the signalling pathways involved in CRC.

Roles of PP in CRC Related to Chemoprevention and Apoptosis
Consumption of PP rich food proved to be beneficial in occurrence of CRC in a national prospective cohort study [32]. Numerous studies have evaluated the efficacy of dietary polyphenols against CRC in vivo, in vitro model and in clinical trials [33][34][35]. Polyphenols can affect the overall process of carcinogenesis by several mechanisms and cause tumour cell death through apoptotic pathway. PP have been shown to be highly effective in scavenging singlet oxygen and various free radicals, which leads to DNA damage and tumour promotion [36]. PP also displayed chemopreventive effect through their impact on the bioactivation of carcinogens. Most carcinogens of chemical origin undergo biotransformation by Phase I metabolizing enzymes to be converted into more reactive form suitable for binding with DNA and proceed towards carcinogenesis process. PP were found to inhibit cytochrome P450 enzymes of the CYP1A family and thus act as chemopreventive agents [37]. On the other hand, by increasing the activity of Phase II metabolizing enzymes (glutathione reductase, glutathione peroxidase, glutathione S-reductase, catalase, and quinone reductase), PP are able to provide beneficial effects against CRC [38,39]. For example, PP obtained from apple inhibited growth of HT-29 human colon cancer cells by modulating expression of genes (GSTP1, GSSTT2, MGST2, CYCP4F3, CHST5, CHST6, and CHST7) involved in the biotransformation of xenobiotics [40].
Apoptosis is a vital physiological process in the normal development, and induction of apoptosis is highly anticipated mode as a therapeutic strategy for cancer control [44][45][46]. Bcl family of protein, caspase signalling proteins, and p53 genes are the key factors that regulate apoptosis [47]. PP are effective general inhibitor of cancer cell growth and inducers of apoptosis in different cancer cell lines, including leukaemia, skin, lung, stomach, colorectal, and prostate cancer cells [34,[48][49][50][51][52]. Anthocyanin, ellagic acid, curcumin, flavone induced apoptosis in various colon cancer cell lines by different mechanisms in miscellaneous observations [34,[53][54][55].
PP can prevent the DNA damage caused by free radicals or carcinogenic agents through diverse mechanisms: (a) direct radical scavenging [56,57], (b) chelating divalent cations involved in Fenton reaction [58], and (c) modulation of enzymes related to oxidative stress (glutathione peroxidase, glutathione reductase, superoxide dismutase, nitric oxide synthase, lipooxygenase, xanthine oxidase, etc.) [59]. Dietary PP can also act as prooxidants depending on the cell type, dose, and/or time of treatment, as they can enhance reactive oxygen species production and therefore induce apoptosis [58,60,61]. In colon cancer HT-29 cells, flavone enriched the mitochondrial pyruvate or lactate uptake, which augmented the superoxide radical production and led to apoptosis [62].

Recent Update of Key PP as Applied to CRC
Reported antitumour activity of PP against CRC is largely based on in vitro studies, rodent model studies, and even human clinical trials. During in vitro studies on antitumour activity of PP, different colorectal cancer cells (HT-29, SW480, Caco-2, Colo-205, Colo-115, HCT-115, HCT-116, DLD-1, LoVo etc.) were cultured, and cell viability was determined via MTT [3-(4,5-Dimethylthiazol-2-yl)-2,5-Diphenyltetrazolium Bromide] reduction assay [154], SRB (Sulforhodamine B) colorimetric assay [155], and crystal violet method [156]. In vivo animal model models were produced by inducing tumour chemically, resulting in APC Min/+ mouse and rodent xenograft models. Carcinogen [azoxymethane (AOM), dimethylhydrazine (DMH), dextran sodium sulphate (DSS)] induced colon cancer in rodents can recapitulate in a highly reliable way and frequently used to assess activity of PP. Mutations in the adenomatous polyposis coli (APC) gene are required to initiate familial adenomatous polyposis (FAP) and are also important in CRC tumorigenesis. Several studies have been conducted with PP in APC Min/+ mouse that contains (multiple intestinal neoplasia Min) a point mutation in the APC gene and develops numerous adenomas. The role of PP has also been investigated in xenograft model where human tumours are injected and established in immunodeficient mouse strains (nude or SCID mice). This section contains the outcome from the studies conducted with PP against CRC.

Benzoic Acid Derivatives.
From literature, only gallic acid among benzoic acid derivatives showed anticancer activity against CRC in vitro and in vivo model [157,158], but no study has been conducted to identify the anticancer mechanism of gallic acid in CRC. However, gallic acid is believed to exhibit its anticancer effect by upregulating Bax and downregulating Bcl-2 in other tumour models [157]. Vanillic acid showed significant activity with IC 50 values less than 30 M in three different CRC cell lines but mechanism has not been studied [159] although vanillic acid and protocatechuic acid did not show significant anticancer activity against CRC [160,161].

Cinnamic Acid Derivatives.
Caffeic acid showed apoptotic cell death against HCT 15 cell lines although IC 50 value was very high (800 M). Similar findings were made by other researchers [162,163]. In a recent study caffeic acid did not show any significant activity against HT-29 cell lines up to 200 M concentration nor did chlorogenic acid [164] that did not show any significant activity against different human colorectal carcinomas [161]. IC 50 values of p-coumaric acid against some other CRC cell lines were around 1 mM and apoptosis was the mechanism of cell death [163,165,166]. Ferulic acid inhibited CRC progression at adhesion and migration steps but no IC 50 value was greater than 1 mM concentration [163].
Carnosic acid showed IC 50 values in the range of 24-96 M against Caco-2, HT29, and LoVo cell lines. It inhibited cell adhesion and migration, possibly by reducing the activity of secreted proteases such as urokinase plasminogen activator and metalloproteinases. These effects may be mediated through a mechanism involving the inhibition of the COX-2 pathway [167]. Sinapic acid showed IC 50 values of less than 25 M in three different CRC cell lines but mechanism has not been studied [159].

Isoflavones, Neoflavonoids, and Chalcones.
Among isoflavones, biochanin A showed ID 50 values below 15 g/mL against two CRC cell lines and was found to enhance radiotoxicity in vitro [170,171]. Formononetin that showed dose dependent cell killing, both in vitro and in vivo in RKO cell line, induces apoptosis by modulating Bax/Bcl-2 activities, inactivating ERK pathway and TNF-/NF-B pathway [172]. Formononetin also showed anticarcinogenic activity in HCT-116 cells via promotion of caspase-dependent apoptosis and inhibition of cell growth, with contribution by downregulation of the antiapoptotic proteins Bcl-2 and Bcl-xL [173]. Daidzein killed 50% of HCT cells, LoVo cells, and DLD-1 cells at concentration of 40 M, 68.8 M, and 46.3 M, respectively, but against LoVo cells it exhibited biphasic effects by killing cells in dose dependent manner at higher concentrations (≥5 M) and vice versa at lower concentrations (≤1 M) [75,174,175]. Most commonly studied isoflavone, genistein, showed cytotoxicity against HCT, LoVo, and DLD-1 cell lines with IC 50 values of 15 M, 57.3 M, and 56.1 M, respectively, whereas in HCC-44B2 cells and HCC50-D3 the value was 11.5 g/mL and 9.5 g/mL [75,170,174]. Genistein reduces the density of cell surface charge and increases the order in membrane protein conformation which might be one of the mechanisms of its anticancer effect [174].
No literature reporting neoflavonoids activity against CRC was found. Among chalcones, phloretin caused apoptotic cell death to HT-29 cells with an IC 50 value close to 100 M. The mechanism involved changes in mitochondrial membrane permeability and activation of the caspase pathways [176]. Phloretin also has the potential to increase adoptive cellular immunotherapy against SW-1116 CRC cells [177]. Xanthohumol, another important chalcone, is found to show cytotoxicity in different CRC cells in vivo and in vitro with IC 50 values less than 5 M [178][179][180]. The apoptosis involved downregulation of Bcl-2, activation of the caspase cascade, and inhibition of topo I activity. In combination with chemotherapy, it is recommended for use in HCT-15 cell lines, being aimed to reduce drug resistance by inhibition of efflux transporters [180]. Xanthohumol inhibits metastasis by inhibiting expression of CXCR4 chemokine receptor [181].

Flavones, Flavonols, Flavanones, and Flavanonols.
Among all different types of flavones, apigenin and luteolin were most commonly investigated phytochemicals for their anticancer activity against CRC. Important flavones that have been studied against CRC are given in Table 1.
Among all different types of flavonols quercetin, chrysin and rutin were studied most for their anticancer activity against colorectal cancer models. Important flavonols that have been investigated against CRC are given in Table 2.
Naringenin appears to be the most commonly studied phytochemicals among the flavanones that can act against colorectal cancer. It suppressed colon carcinogenesis through the aberrant crypt stage in azoxymethane-treated rats [74]. Another study showed that antiproliferative activity of naringenin was estrogen receptor dependent [182], while other in vitro studies gave mixed results in different CRC cell lines [65,80,152,183]. Another flavanone, hesperetin, significantly reduced the formation of preneoplastic lesions and effectively [70] Apigenin HCT116 Induced cell death due to apoptosis and autophagy where apoptosis is via decreased expression of cyclin B1, Cdc2, and Cdc25c; increased expression of p53 and p21CIP1/WAF1; decreased levels of procaspase-8, -9, and -3. [71] HT-29 and HCT-15 Oxidative stress resulted in senescence and chemotherapeutic effect. [72] SW480 and HCT-15 Suppressed cell proliferation, migration, and invasion via inhibition of the Wnt/ -catenin signalling pathway. [  Downregulated the activation of the PI3K/Akt and ERK1/2 pathways via reduction in IGF-IR signalling which may be one of the mechanisms responsible for the observed apoptosis and cell cycle arrest. [78] Luteolin HT-29, SW480 In HT-29 cells, IC 50 value was greater than 200 M but in SW480 cells it is 90 M. [79,80] Male Balb/c mice Inhibited azoxymethane-induced colorectal cancer growth through activation of Nrf2 signalling; altered carbohydrate metabolizing enzymes; decreased expressions of iNOS and COX-2; restored reduced glutathione and protein thiols; decreased lysosomal enzymes, induced apoptosis by modulating Bcl2, Bax, and caspase-3; decreased mucin depleted foci, levels of glycoconjugates; controlled cell proliferation by inhibiting wnt/ -catenin/GSK-3 pathway. Luteolin also acts as antimetastatic agent by decreasing MMP-9 and MMP-2.
[100] HT-29 Resveratrol and quercetin in combination showed anticancer activity in colon cancer cells and repressed oncogenic microRNA-27a. [108] HT-29 xenografts in female nude mice Quercetin and trans-pterostilbene in combination facilitated elimination of colorectal cancer by chemoradiotherapy through a Bcl-2-and superoxide dismutase 2-dependent mechanism. [109] CF1 mice, F344 rats, Wistar rats Azoxymethane and dimethylhydrazine induced colon cancer study showed reduction of aberrant crypt foci and focal areas of dysplasia.
[113] HCT-116 Chrysin sensitized tumour necrosis factor--induced apoptosis in human tumor cells via suppression of nuclear factor-kappaB. [122] HCT-116 Promoted tumour necrosis factor-(TNF-) related apoptosis-inducing ligand (TRAIL) induced apoptosis. [123] SW480 Chrysin caused cell-cycle arrest at the G2/M phase in a dose-dependent manner. [80] HCT116, DLD1 and SW837 Aryl hydrocarbon receptor was required for the chrysin induced apoptosis and the upregulation of TNF-andgene expression and consequent activation of the TNF-mediated transcriptional pathway. [124] Caco-2 Blocked topotecan-induced apoptosis in spite of inhibition of ABC-transporters. [125] Kaempferol SW480 Sensitized TRAIL-induced apoptosis. [126] HCT-116 The IC 50 of kaempferol was 53.6 M in HCT116 (p53+/+) cells and 112.7 M in HCT116 (p53−/−) cells. It induced via ataxia-telangiectasia mutated-p53 pathway with the involvement of p53 up-regulated modulator of apoptosis. [127] HT-29 Kaempferol increased chromatin condensation, DNA fragmentation, and the number of early apoptotic cells in a dose-dependent manner. Kaempferol increased the levels of cleaved caspase-9, caspase-3, and caspase-7 as well as those of cleaved poly (ADP-ribose) polymerase. Moreover, it increased mitochondrial membrane permeability and cytosolic cytochrome c concentrations. [128] Isorhamnetin HT-29, FVB/N mice Chemoprotective effects of isorhamnetin were linked to its inhibition of oncogenic Src activity and consequential loss of nuclear -catenin, activities that were dependent on CSK expression. Had IC 50 value less than 350 g/mL after 48 h and induced apoptosis by modulation of Bcl-2 family members and Fas receptor. [136] modulated the xenobiotic-metabolizing enzymes in rats during DMH-induced colon cancer study [184,185]. Among flavanonols, only taxifolin acts as an effective chemopreventive agent against colon carcinogenesis due to its antioxidant mediated apoptosis and antiproliferative activities [186,187]. Taxifolin is found to control NF-kB-mediated Wnt/ -catenin signalling via upregulating Nrf2 pathway [188].   HCT116, HT29, SW480, and SW837 cell lines, respectively. Mechanism of action has been linked to the inhibition of growth and activation of the epidermal growth factor receptor and human epidermal growth factor receptor-2 signalling pathways [189]. Among eleven different types of flavanol, EGCG showed the highest antiproliferative activity against HCT-116 cells at 50 M [190]. In APC Min/+ mice, EGCG significantly inhibited intestinal tumorigenesis. Oral administration of EGCG showed increased levels of E-cadherin and decreased levels of nuclear -catenin, c-Myc, phospho-AKT, and phosphoextracellular signal-regulated kinase 1/2 (ERK1/2) in small intestinal tumours [191]. In another mice model, EGCG reduced inflammation-related colon carcinogenesis induced by azoxymethane and dextran sodium sulphate. Antitumour activity has been ascribed to decrease in mRNA expression levels of COX-2 and inflammatory cytokines (TNF-, IFN-, IL-6, IL-12, and IL-18) in the colonic mucosa [192]. Other studies mentioned the proposed anticancer mechanisms of EGCG including cell cycle arrest and apoptosis through inhibition of cyclooxygenase-2 expression, activation of AMPactivated protein kinase, cyclin D1 degradation and p21 transcriptional activation, and inhibition of HES1 and Notch2 signalling in different colorectal cancer cell lines [43,[193][194][195][196]. EGCG also inhibited invasion and MMP expression, angiogenesis, through blocking the induction of VEGF [197,198]. Molecular mechanism for antitumour activity of EGCG and quercetin is shown in Figure 4.

Flavanols and
Chemopreventive effects of theaflavin have been reported in azoxymethane induced colon cancer study in male Sprague Dawley rats [199]. The mechanism behind the beneficial effects of proanthocyanidins against CRC has been related to inhibition of angiogenesis through suppressing the expression of VEGF and Ang1 [200], induction of apoptosis [201], and antioxidant activity [202].

Polyphenolic Amides.
Capsaicin is the most studied polyphenolic amide against CRC. In vitro and in vivo studies in mice bearing Colo-205 tumour xenografts showed that capsaicin significantly reduced tumour progression by activating caspase-3, caspase-8, caspase-9, Bax, Fas and reducing Bcl-2 [218]. Capsaicin and 3,3 -Diindolylmethane worked synergistically against CRC via modulating transcriptional activity of NF-B, p53, and target genes linked with apoptosis [219]. Other studies showed that anticancer action of capsaicin was related to nitric oxide production, reactive oxygen species generation, and suppression of transcriptional activity of -catenin/TCF pathway [220][221][222]. In our study, we have found synergism in binary sequenced combination of capsaicin with oxaliplatin in all sequences (0,0 h; 0,4 h; 4,0 h) and more so in higher concentration in Lim 2405 cell line (unpublished data). Dihydrocapsaicin, a saturated structural analogue of capsaicin, was found to possess greater activity than capsaicin against HCT-116 cells and induced autophagy in a catalase-regulated manner [223]. Avenanthramides significantly inhibited proliferation of HT29, Caco-2, LS174T, and HCT116 human colon cancer cells [224].

Other Polyphenols
5.4.1. Resveratrol. 3,5,4-Trihydroxystilbene, known as resveratrol, is one of the most studied polyphenols against CRC. It entered into clinical trial after a number of preclinical studies for its encouraging activity and nontoxicity. All studies conducted in rodent model and clinical trial regarding the activity of resveratrol up to 2009 have been described by Bishayee in his review [225]. The authors mentioned 9 in vivo studies related to CRC, among which three were conducted in APC Min/+ mice model and others related to chemically induced tumour [226][227][228][229][230][231][232][233][234]. All the studies discussed in the review (except one on APC Min/+ mice study) provided support for therapeutic potential of resveratrol against CRC. Later on, in 2012, Juan et al. published another review article that focused on the effects of resveratrol on CRC from conducted in vivo studies and clinical trials [235]. The authors provided details on the molecular mechanism of action of resveratrol against CRC. Two more research articles recently described the promising effect of resveratrol in mouse model against CRC, which were not mentioned in earlier reviews [236,237]. The results from the investigations of the anticancer activity of resveratrol against CRC have not been detailed here because they have been considered well in the previously mentioned reviews. However, a pictorial representation of the molecular mechanism of action of resveratrol against CRC is given in Figure 5. The effect of 5-fluorouracil increased in combination with resveratrol due to chemosensitizing property [238]. Antimetastatic activity of resveratrol in CRC has also been reported [237,239]. Since resveratrol is found to downregulate multidrug resistant protein 1 by preventing activation of NF-B signalling and suppressing cAMP-responsive element transcriptional activity, it can be used to overcome drug resistance by combining with other chemotherapeutic drugs [240]. We have found synergism at higher concentration in binary sequenced combination (at 0,0 h; 0,4 h; 4,0 h) of resveratrol with cisplatin but additive to antagonism at lower concentration in HT-29, Caco-2, and Lim 2405 cell lines (unpublished data).

Curcumin.
Curcumin is the main active compound in turmeric (dried rhizome of Curcuma longa). We have found 82 research articles (in vitro and in vivo preclinical studies, clinical trial) describing effect of curcumin including mechanism of action against CRC. Here, we have not considered the anticancer activity of curcumin in detail because three review articles discussed well the therapeutic potential of curcumin for CRC along with its mechanism of action [241][242][243]. Curcumin can inhibit the initiation of carcinogenesis by increasing glutathione S-transferase, induce cell cycle arrest in S and G2/M phase, induce apoptosis, and inhibit metastasis by decreasing CD31, VEGF, IL-8, and mir-21 expression [244][245][246][247]. Mechanism of curcumin in CRC as applied to apoptosis is shown in Figure 6. Curcumin has proved to be beneficial in combination with chemotherapy and radiotherapy as well [248,249]. Synergistic activity of curcumin has also been observed in our study in different sequences and doses with oxaliplatin and cisplatin in four colorectal cell lines (unpublished data).

Rosmarinic Acid and Gingerol.
Rosmarinic acid is the main component of Rosemary which at high dose showed antitumour activity applying to in vitro and DMH-induced in vivo study against CRC [250,251]. It is thought that MAPK/ERK pathway is linked with the apoptotic mechanism of rosmarinic acid in CRC [252]. Rosmarinic acid has also been reported to possess antimetastatic activity [253].
6-Gingerol is the most important ingredient of ginger showing antiproliferative activity in a dose dependent manner against LoVo cell lines via G2/M cell cycle arrest. Exposure to 6-gingerol induced intracellular ROS and upregulate p53, p27Kip1, and p21Cip1 levels leading to consequent decrease of CDK1, cyclin A, and cyclin B1 [254]. In HCT-115 cell line, its anticancer action was found to be mediated by inhibition of Leukotriene A4 hydrolase [255]. Another  study showed that 6-gingerol stimulated apoptosis through upregulation of NAG-1 and G1 cell cycle arrest through downregulation of cyclin D1 that involved protein degradation as well as -catenin, PKC , and GSK-3 pathways [256]. However, 6-gingerol did not show anticancer activity in Colo-205 cell line [257].

Ellagic Acid, Secoisolariciresinol, and Matairesinol.
Ellagic acid is a dilactone of hexahydroxydiphenic acid that occurs naturally in berries and nuts such as the raspberry, strawberry, walnut, and pecan. It inhibited growth of Caco-2 cell line, possibly mediated by regulation of matrix metalloproteinases, vascular endothelial growth factor expression, and induction of apoptosis [258]. According to others, the anticancer action is mediated through downregulation of cyclins A and B1 and upregulation of cyclin E, cell cycle arrest in S phase, induction of apoptosis via intrinsic pathway (Fas-independent, caspase 8-independent) via Bcl-XL downregulation with mitochondrial release of cytochrome c into the cytosol, and activation of initiator caspase-9 and effector caspase-3 [54]. Studies on DMH-induced colon carcinogenesis also proved the beneficial effect of ellagic acid and showed the mechanism to be linked with reduced expressions of NF-, COX-2, iNOS, TNF-a, and IL-6 as well as inhibition of AKT-phosphoinositide-3 kinase pathway [259,260]. However, metabolic products urolithins were found to be more potent against CRC compared to ellagic acid itself [261]. Mixed urolithins and ellagic acid inhibited phenotypic and molecular colon cancer stem cell features as well [262]. Secoisolariciresinol and matairesinol are lignans. They constitute one of the major groups of phytoestrogens that have been investigated against CRC but the activity remains controversial. Earlier study with flax seed diet containing secoisolariciresinol and matairesinol showed significant antitumour activity [263]. However, in vivo and in vitro studies done later did not provide insights into the anticancer potential of secoisolariciresinol and matairesinol [264,265].

Chemical Modifications of PP in Nature and Synthetic Analogues
The most important impediment to the successful development of natural PP as clinical therapy against CRC is their low bioavailability. To overcome the problem towards reaching therapeutic concentrations of PP and increasing efficacy, many researchers tried to produce a number of synthetic analogues through structural modifications. Increase in potency in vitro and bioavailability in vivo has been observed in many studies. Some selected reports concerning chemical modifications of PP applied against CRC are represented in Table 3.

Combination of Polyphenol with Chemotherapy/Radiotherapy and Other PP
Emerging evidence suggests that a single-agent approach is probably less likely to be very effective in the management cancer. The rationale for recommending a multidrug regimen is to attack cancer cells through multiple targets and diverse mechanisms of actions with reduced toxicity, ultimately leading towards improved clinical outcomes. With that aim, PP have been investigated with chemotherapy and radiotherapy in various cancer models. For example, curcumin given in combination potentiated the cytotoxic effects of doxorubicin, 5-FU, and paclitaxel in prostate cancer cells [266]; enhanced the antitumour activities of cisplatin, doxorubicin, and Taxol in HA22T/VGH hepatic cancer cells, HeLa cells, or CAOV3 and SKOV-3 ovarian cancer cells [267,268]; sensitized multiple myeloma cells to vincristine and melphalan [269]. Similar evidence is available in literature for resveratrol, EGCG, quercetin, genistein, proanthocyanidin, and daidzein in various types of cancer with different classes of chemotherapeutic drugs [169,[270][271][272][273]. A few studies also have been conducted against CRC where it has been observed that PP in combination with other chemotherapeutic drugs produced synergism, for example, curcumin with 5-fluorouracil against HCT-116 and HT-29 cell line, resveratrol metabolites and oxaliplatin in SW-480 and SW-620 cell lines, genistein with 5-fluorouracil in HT-29 cell line [274][275][276][277]. Outcomes of some other combination studies already have been mentioned in describing the effect of individual PP. Like chemosensitization, PP also have shown the potential of radiosensitization in various cancers, but investigations of the same effects against CRC is very scarce [278][279][280][281]. In one study quercetin has been shown to increase chemoradiosensitivity against colorectal cancer in xenograft mouse model [109]. Plenty of evidence exists in literature regarding the benefits of the effect of one polyphenol combined with another polyphenol against different types of cancer. Combination of resveratrol with black tea polyphenols resulted in a synergistic tumour suppressive response in mouse skin tumour [282]. Resveratrol showed better chemopreventive response when combined with curcumin by maintaining adequate zinc and modulating Cox-2 and p21 level in mouse model of lung cancer [283]. Combination of genistein with resveratrol reduced the most severe grade of prostate cancer in SV40 Tag-targeted probasin promoter rat model [284]. EGCG in combination with curcumin synergistically inhibited oral premalignant epithelial cells [285]. Quercetin and resveratrol in combination with ellagic acid showed synergism against leukaemia [286]. However, studies on the effect of combination of pure polyphenols against CRC are not numerous. Curcumin showed synergistic antitumour effects in combination with resveratrol in one report of colon cancer model in SCMID mice [287]. Combination of epicatechin and EGCG also exhibited synergistic outcome against HT-29 cancer cells. There are few reports in literature related to the beneficial effects of plant extracts or juices that possess mixture of polyphenols against CRC [288,289].
Bioprospecting and molecular pharmacology studies have shown that PP can modulate the survival pathways induced by cancer cells, carcinogens, and chemotherapeutics. The possible mechanisms of chemoresistance are shown in Figure 7.
In Section 5, we have considered the molecular mechanisms by which PP would produce anticancer action in CRC. It is thought that PP have the ability to effectively modulate the various mechanisms of chemoresistance. For example, EGCG and quercetin can directly inhibit PI3/AKT pathway, NF pathway, EGFR family pathway, and IAP family pathway and increase p53 (Figure 4). Similarly curcumin can inhibit Bcl-2 family pathway, EGFR family pathway, and NF pathway ( Figure 6). In terms of IC 50 values, synthetic analogue found to be more sensitive against 3 CRC cell lines. In vivo study also proved its greater activity. [137] Resveratrol 3, 5, 4 -Trimethoxystilbene, 3, 3 , 4, 5 -tetramethoxystilbene Inhibited HT-29 cell growth. [138] Digalloyl resveratrol Inhibited HT-29 cell growth more effectively than gallic acid and resveratrol. [139] Flavone 3 , 4 , 5 , 5, 7-Pentamethoxyflavone More active compared to tricin and apigenin in APC Min/+ mice model. [140] Curcumin Dimethoxycurcumin In HCT-116 cell lines, dimethoxycurcumin is more potent in terms of ability to kill cancer cells by apoptosis, less extensively metabolized in microsomal systems, and more stable in vivo compared to curcumin. [141] Curcumin difluorinated At higher concentration synthetic analogue showed greater potency than curcumin in HCT-116 cells. [142] EF31 and UBS109 Both analogues showed significant antitumour activity in colorectal xenograft model possibly via inhibition of NF-B and cell cycle progression at G0/G1 phase. [143] GO-Y030, FLLL-11, and FLLL-12 All of the analogues exhibited 4 to 20 times greater activity than curcumin against SW480, HT-29, and HCT116 cell lines but with minimal toxicity against normal cell line. [144] EGCG Peracetylated EGCG Administration of the synthetic analogue was more effective than EGCG in preventing the shortening of colon length and the formation of aberrant crypt foci and lymphoid nodules in mouse. [145] Procyanidin dimer [150]

Current Status of PP in Clinical Trials Related to CRC
Following the discovery of significant anticancer potential of curcumin, resveratrol, EGCG, and genistein seen in studies related to in vitro and in vivo rodent model, the compounds entered into clinical trials for efficacy and toxicity study in human model. Many of the reported Phase I, Phase II, and Phase III studies further validated the potential of using PP against CRC. Benamouzig and Uzzan provided a summary of 21 clinical trials conducted by 2016 related to the use of curcumin against CRC in their review [243]. Similarly 17 clinical trials on resveratrol have been reviewed elsewhere [290]. Few other chemotherapeutics and chemoprevention clinical trials conducted on PP have been summarized by Vinod et al. [169]. Table 4 represents the recently completed or ongoing clinical trials on PP against CRC, which have not been covered by others.

Future Perspective and Directions
From our study we have found that curcumin, resveratrol, quercetin, luteolin, apigenin, EGCG are the most investigated polyphenols against CRC. In terms of in vitro cytotoxicity these polyphenols gave average IC 50 value around 15-60 M against different colorectal cancer cell lines, which is comparatively larger than clinically used anticancer drugs. Moreover some results from single administration of curcumin or resveratrol produced contradictory evidence against in vitro and in vivo model data. It would be unwise to be overoptimistic and battle against CRC with a single polyphenol only. Rather the strength of polyphenols can be exploited by combining them with clinically used chemotherapeutic drugs to reduce dose related side effects of chemotherapy and overcoming drug resistance. Therefore, we can search among polyphenols for the ones that give synergistic effect with chemotherapeutic drugs or with other phytochemicals. In our laboratory we have been working with curcumin, resveratrol, EGCG, quercetin, capsaicin, 6-gingerol, genistein in combination with cisplatin and oxaliplatin against different CRC cell lines. In many cases we found synergism (unpublished data) like others [291]. Chemical modifications of natural PP can be continued to improve their activity and bioavailability.
Another area of research that could be explored is related to exploiting ability of PP to interact with stem cells in  CRC. Cancer stem cells are multipotent cells that possess self-renewal capacity and high proliferative capacity and lead to metastasis through migration. Although cancer stem cells represent less than 2.5% of the tumour mass, they may be responsible for the resistance to cancer therapies and relapse in CRC [292]. Wnt/ -catenin, Hedgehog, and Notch have been identified to play pivotal roles in cancer stem cells self-renewal. Presently researchers are targeting the hallmark stem cell-like properties of tumour cells to overcome cancer. A number of phytochemicals have also been investigated against cancer stem cells in several studies. Curcumin suppressed mammosphere formation along serial passage in breast cancer, and the effect of curcumin on breast cancer stem/progenitor cells was seen to be mediated through its potent inhibitory effect on Wnt/ -catenin signalling [293]. Genistein also reduced breast cancer stem cells by inhibiting AKT and increasing PTEN [294]. Likewise, resveratrol inhibited pancreatic cancer stem cell characteristics in human and mouse model by inhibiting EMT [295]. In pancreatic cancer model quercetin decreased ALDH1 activity, induced apoptosis, and reduced the expression of proteins implicated in EMT in vitro, while it inhibited stem cell-derived xenografts in vivo, reducing the expression of proliferation, stemness, and angiogenesis related genes [296]. However, very little has been studied to modulate the stem cells by PP in CRC.

Conclusion
As oxidative stress is an inescapable part of aerobic life, it can be said that cancer with its origin in mutations is a disease of living even though it evokes death sentence in many minds. However, as nature creates problems, it also provides solutions. As tumour active polyphenols have been a part of human diet for thousands of years, but without any adverse side effect, it is thought that selected tumour active polyphenols or their derivatives in combination with targeted therapy may provide an affordable means of overcoming drug resistance and reducing side effects in colorectal cancer and indeed in many other cancers.

Conflicts of Interest
Md Nur Alam, Muhammad Almoyad, and Fazlul Huq declare that they have no conflicts of interest.

Authors' Contributions
Muhammad Almoyad conducted the literature review, Md Nur Alam has drafted the entire review, and Fazlul Huq supervised the whole process and critically reviewed the manuscript.