Helicobacter pylori is a helix-shaped, microaerophilic, Gram-negative, flagellated bacterium that specifically colonizes the human gastric mucosa. It is considered as the most widespread chronic bacterial pathogen in the world, infecting more than 50% of the human population. Although over 80% of the infected individuals remain asymptomatic, H. pylori is implicated as an etiologic factor in the development of a variety of GI diseases including gastroduodenal ulcers and gastric adenocarcinoma and lymphoma. Accordingly, this pathogen is classified as a class I carcinogen (25–27). H. pylori expresses several virulence factors involved in the initial colonization of the gastric mucosa by the bacteria and in its persistence in the stomach. These factors include (1) adhesins, bacterial surface proteins allowing the adherence of this agent to the gastric epithelial cells, (2) an urease activity that releases ammonia from urea hydrolysis, allowing proton neutralization and the survival of the bacteria in the acidic gastric environment, (3) the vacuolating cytotoxin A (VacA) that alters the mitochondrial function of the epithelial cells and promotes their vacuolization and subsequent apoptosis, in addition to influencing host tolerance by suppressing T cell activation, and (4) the cag pathogenicity island encoding a type IV secretion system, which is involved in the development of inflammatory processes in the mucosa (27–29).

The administration of PACs-rich beverages decreases H. pylori colonization in humans. A prospective, randomized, double-blind, placebo-controlled trial conducted in China in 189 adult patients colonized by H. pylori provided some clinical support for the anti-H. pylori effect of cranberry juice that contains large amounts of A-type PACs (30, 31). The participants were assigned to receive 250 ml of cranberry juice or placebo twice daily for 90 days. H. pylori colonization was determined by ¹³C-urea breath test after 35 and 90 days of treatment. The eradication of the pathogen was reported in 14.4% of the treated subjects compared with 5.3% in the placebo group. Similar results were reported in a clinical trial carried out in 295 Chilean children colonized by H. pylori. The intake of 200 ml cranberry juice per day for 3 weeks induced the eradication of the pathogen in 16.9% of the treated children vs 1.5% in the placebo group. Interestingly, the coadministration of cranberry juice with the probiotic, L. johnsonii NCC533, eradicated the bacteria in 22.9% of the children (32). The mechanisms associated with the anti-H. pylori activity of cranberry PACs include their ability to inhibit urease activity and prevent bacterial adhesion as well as VacA-induced mucosal damage (Table 1). High molecular mass, non-dialyzable constituents of cranberry juice, subsequently characterized as PACs, have been shown to inhibit the adhesion of clinically isolated H. pylori strains to AGS gastric cell line. As no cross-resistance was detected between the non-dialyzable material and metronidazole, it was suggested that cranberry preparations could improve H. pylori eradication in untreated patients as well as in those under pharmacological treatment (33). In vitro studies consistently indicate that high molecular weight (HMW) components from cranberry inhibit the sialyllactose-specific (S fimbriae) adhesion of H. pylori to immobilized human mucus, erythrocytes, and cultured gastric epithelial cells (34, 35). This anti-adhesion effect was not restricted to cranberry PACs and was also reported with dietary PACs from different sources. Pycnogenol^®, a standardized concentrate elaborated from French maritime pine bark and which contains PAC from dimers (mainly type B) to polymers (up to 12 monomeric units of flavanol) (36, 37), inhibited concentration-dependently the adhesion of clinical strains of H. pylori to AGS cells (38). A root extract from Pelargonium sidoides, rich in polymeric PACs (39), has also been shown to prevent H. pylori adhesion to intact human stomach tissue (40–42).

Proanthocyanidins can also interfere with H. pylori by inhibiting its urease activity and/or by inactivating VacA. For example, a methanol extract of Eucalyptus grandis (Myrtaceae) bark was shown to inhibit the urease of clinical strains of H. pylori in a concentration-dependent manner. This effect was attributed to the tannins and triterpene saponins present in the extract (43). Some studies suggest that PAC DP is a determining factor for urease inhibition or their protective effects on VacA-induced mucosal damage. LMW PAC fraction (mean DP of 3) derived from apple peel exhibited an urease inhibition fourfold lower than the high HMW fraction (mean DP of 9.5) (44). More recently, an aqueous extract of Peumus boldus Mol. (Monimiaceae) rich in catechin-derived PCs B3 and C2 was also shown to inhibit H. pylori urease (45). The separation of the extract components according to their molecular weight revealed that the higher the DP, the greater the urease inhibition. Interestingly, the catechin-derived PCs (B3 and C2) were more effective than the epicatechin-derived PCs B2 and C1 in inhibiting the urease, suggesting that not only DP is important but also the chemical nature of the monomer bound to C-4 in the PC structure (45). A structure–activity relationship between PAC DP and their ability to inactivate VacA cytotoxin has also been established with hops extracts (46). Accordingly, the administration of red wine or green tea mixture to H. pylori-infected mice significantly prevented the development of gastritis, limiting the localization of the pathogen and VacA toxin on the surface of the gastric epithelium (47).

On the other hand, exciting advances in the knowledge of the interactions between PCs and target molecules are emerging from molecular docking studies. Such studies, for example, revealed that B-type PCs (catechin dimers) inhibit urease because, according to their docking scores (−6.9 kcal/mol), they fit the binding pocket of the bacterial enzyme, being these interactions energetically favorable (48). Then, molecular docking could contribute to elucidate the interactions between PCs and other cellular or bacterial targets of interest for the protection of the gastric mucosa against H. pylori.

The gastroprotective properties of PAC extracts have been widely studied using different animal models of gastric inflammation induced by ethanol, non-steroidal antiinflammatory drugs (NSAIDs), pylorus ligature, or restrain stress (Table 2). Synthetic PAC oligomers, more particularly those bigger than tetramers, protected the gastric mucosa against ethanol-induced damage, due to their ability to scavenge free radicals and, therefore, prevented the appearance of oxidative damage (49). Such effect could also be due to the protein-binding ability of the oligomers, which would allow them to form a protective layer coating the gastric mucosa. In another study, the administration of a single oral dose of Hawthorn berries extract (a mixture of Crataegus monogyna and C. oxyacantha containing 0.44% PACs) attenuated the intensity of ethanol-induced gastric lesions in rats, an effect similar to that observed with the administration of ranitidine (50). Considering the low proportion of PACs in this extract, it is improbable that these molecules were responsible for this protective effect; however, it is important to mention that Hawthorn berries contain oligomeric PACs and more B-type than A-type (51).

Proanthocyanidins were also protective in gastric damage induced by NSAIDs. Although NSAIDs are widely used for their antipyretic, analgesic, and antiinflammatory properties, their administration is frequently associated with adverse effects mainly affecting the GI mucosa (52, 53). NSAIDs and proton pump inhibitors (like omeprazole and lansoprazole) are frequently coprescribed to minimize NSAID-related adverse effects and the enteropathy induced by such combination, though common, is often clinically silent. Thus, the lesions induced by these drugs in the GI tract could be of considerable clinical importance. Accordingly, bioactive compounds, like PACs, arise as an alternative approach for the management of the adverse effects associated with NSAID therapies (53). The oral intake of PAC extracts from Guazuma ulmifolia (Sterculiaceae) or grape seed for 2–6 days before diclofenac or indomethacin administration in rats, prevented in a dose-dependent manner the development of gastric mucosal damage and attenuated intestinal injury (54–56). These extracts decreased the area of ulceration induced by indomethacin in the stomach by decreasing lipid peroxidation in the mucosa and by increasing the superoxide dismutase and glutathione peroxidase activities as well as the glutathione levels (54). The G. ulmifolia extract was also shown to prevent neutrophil infiltration (reflected by a lower myeloperoxidase activity) in the mucosa (54).

Proanthocyanidin extracts from medicinal plants including Mangaba (Hancornia speciosa) (containing LMW PACs), Curatella americana L. (Dilleneaceae) (containing oligomeric and polymeric PACs), Cecropia glazioui Sneth (Cecropiaceae) (that contains 22% of PACs B2, B3, B5, and C1), and Byrsonima intermedia (containing oligomeric PACs, phenolic acids, and catechin derivatives and flavonoids) have also been evaluated (57–61). The oral administration of a single dose of each extract decreased the gastric lesions induced by indomethacin or piroxicam, hypothermic restraint stress, ethanol, or pylorus ligature, in the same extent that cimetidine, ranitidine, or lansoprazole (57, 58, 60, 61). Interestingly, the extraction step seems to be critical to determine the antiulcer effects of the bioactive compounds; no significant protective effects were observed when plant infusions were used (57, 58).

Another mechanism by which PACs may protect the gastric mucosa is through the modulation of HCl secretion. Plant extracts decreased the secretion of HCl by gastric parietal cells when administered intraduodenally to pylorus-ligated mice (57, 58, 60, 61). The C. glazioui extract reversed the histamine or bethanechol-induced acid secretion to basal values, indicating it inhibits the proton pump. This antisecretory effect was comparable to that observed with the histamine H₂ receptor-antagonist, ranitidine (60). It is possible that this effect relies on the type-B2 PACs since these molecules displayed the highest inhibitory activity against the gastric H⁺, K⁺-ATPase in vitro, compared with the other type of PACs isolated from C. glazioui (like B3, B5, and C1) (60).

The PAC antiulcer properties are also related to their ability to stimulate mucus synthesis and secretion (62), and to their mucosal repair activity. Indeed, they were shown to accelerate the healing of gastric ulcer induced by acetic acid administration (57, 58, 61). It is probable that nitric oxide synthase (NOS), sulfhydryl compounds (SH), and TRPV-vanilloid receptors were involved in this phenomenon, as pretreatment with NOS inhibitor (L-NAME), SH-blocker (NEM), or TRPV receptor inhibitor (ruthenium red) blocked the PACs protection against ethanol-induced gastric damage (57, 58, 61). The increased expression of key molecules implicated in the restitution of the gastric epithelium during chronic gastroduodenal ulcers arises as another mechanism of protection exerted by PACs. For example, a sea buckthorn extract containing 96.5% PACs was shown to decrease the ulcer index in acetic acid-induced gastric lesions, in association with increased plasma concentrations of epidermal growth factor (EGF) and higher expression of EGF receptor and proliferating cell nuclear antigen in the gastric mucosa (63). These molecules are considered as crucial for the ulcer healing process and are implicated in epithelial restitution and gland reconstruction. The effect of PACs on trefoil peptides, also involved in the epithelial restitution process, has not been studied.

On the other hand, PACs could also exert their protective effects through endocrine and neural mechanisms. A C. americana extract, for example, exhibited ulcer healing properties by increasing the mucosal levels of prostaglandin E₂ and somatostatin and decreasing those of gastrin (58). Gravinol S containing 89.3% PAC (25.2% dimers–pentamers, 74.8% oligomers) from grape seeds, when administered ad libitum for 2 weeks, prevented gastric mucosal damage induced by water immersion restraint stress in rats. The authors proposed that PACs inhibit gastrin secretion by G cells and subsequently that of histamine and somatostatin, in addition to increase prostaglandin levels and superoxide dismutase activity in the gastric mucosa (64). Using the same animal model, a single dose of Viburnum opulus (Caprifoliaceae) PAC extract attenuated the gastroduodenal lesions. This effect was abrogated by capsaicin pretreatment, suggesting the implication of the nervous system. Moreover, the administration of this extract was also associated with the stimulation of the nitric oxide system, the increased resistance of the mucus layer, and the stimulation of mucosal superoxide dismutase and catalase activities (65).

A few studies have evaluated the impact of PAC intake on the release of digestive hormones, some of them in relation to GI motility. González-Abuín et al. showed that rats fed a cafeteria diet exhibited a lower density of enteroendocrine cells in the intestinal epithelium and plasma concentrations of active glucagon-like peptide-1 (GLP-1), and that these alterations were prevented with a grape seed PC extract (66). Since GLP-1 is an incretin hormone participating in the regulation of food intake and insulin secretion, it is possible that dietary PACs are beneficial for the control of energetic metabolism in humans. There is no doubt that it is a promising field of research. Serrano et al. showed that the administration of a grape seed proanthocyanidin extract (423 mg phenolics/kg body weight) to rats increased the portal concentrations of active GLP-1 and ghrelin and decreased those of cholecystokinin. These findings were accompanied by a delayed gastric emptying and lower food intake in the treated animals (67). In another study, Ko et al. evaluated the effect of a whole grape juice (with skin and seeds) in rats treated with cisplatin, a chemotherapeutic drug known to provoke acute GI disorders. Cisplatin treatment decreased significantly the rate of gastric emptying compared with the control group, and pretreatment with grape juice (10 ml/kg body weight) prevented this disturbance (68).

Miller et al. evaluated the effect of an extract of Croton palanostigma (100 mg/g of PACs) in the treatment of emesis induced by the administration of morphine-6-glucuronide in adult ferrets. The extract reduced by 77% the morphine-induced vomiting and retching, suggesting that it could suppress the activation of sensory afferent nerve implicated in the emetic reflex (69). According to these observations, Li et al. reported that a grape seed proanthocyanidin extract inhibits in a non-competitive manner the 5-hydroxytryptamine-3 receptors involved in the initiation and coordination of the vomiting reflex, in NCB-20 neuroblastoma cells (70). Taking together, these results suggest that PAC-containing extracts or foodstuffs could be used as a complementary medicine for the management of nausea/emesis, without the side effect usually associated with cannabinoid-based antiemetic agents.

Starch constitutes the main source of carbohydrates and energy in the occidental diet, although the disaccharides sucrose and lactose and the monosaccharides glucose and fructose are also present (104). Starch is digested in the intestinal lumen by pancreatic α-amylase and the resulting disaccharides, trisaccharides, and limit dextrin are subsequently digested by the brush-border disaccharidases, maltase-glucoamylase and saccharase-isomaltase, while lactose is hydrolyzed by the lactase-phloridzin hydrolase. The resulting monosaccharides (glucose, galactose, and fructose) are absorbed into the enterocytes through active [sodium–glucose cotransporter 1 (SGLT1)] and facilitated [glucose transporter (GLUT5 and GLUT2)] apical transporters. Dietary polyphenols, including PACs, may delay carbohydrate digestion and reduce postprandial glucose absorption, which represents an alternative approach for diabetes prevention and management (15, 102).

Triglycerides (TG) constitute the majority of the dietary lipids, while the contribution of cholesterol (CS) and phospholipids is much lower. Due to their hydrophobicity, lipids must be solubilized to be digested and posteriorly absorbed (130). Dietary fats are first emulsified in the stomach, a phenomenon that increases the rate of TG hydrolysis by the lipase and the release of diacylglycerol and free fatty acids (FFAs). Once in the small intestine, the fat emulsion is stabilized by bile salts, enabling the action of the colipase/pancreatic lipase (PL), CS ester hydrolase, and phospholipase A2 (PLA2) (130). PL hydrolyzes the TGs, releasing the fatty acids esterified in the carbon 1 and 3 of the molecule, and monoacylglycerol. Regarding CS ester hydrolase and PLA2, these enzymes hydrolyze CS esters and phospholipids, releasing free CS, FFAs, and lysolecithin (130). As these final products of digestion (monoglycerides, FFAs, CS, and lysolecithin) are hydrophobic, they must be incorporated into biliary mixed micelles as they are released in the lumen. This process of solubilization allows them to diffuse across the unstirred water layer until the enterocyte surface, where they are released from the micelles. Posteriorly, they enter into the absorptive cells by passive diffusion according their concentration gradient or by using specific transporters such as FAT/CD36, FATP4, and FABPpm for FFAs and Niemann–Pick C1-like protein 1 (NPC1L1) for CS (131, 132). In the enterocytes, FFAs are reesterified with glycerol or CS, and the resulting TG and CS ester are subsequently incorporated to chylomicrons and exported into lymphatic circulation, to finally end up in the bloodstream. Bile salts are reabsorbed in the terminal ileum through the apical sodium bile acid transporter, they reach the circulation and are taken up by the liver and resecreted by the biliary system. This enterohepatic circulation takes place 10 times per day so that less than 5% of bile acids enter the large intestine during this period for fecal elimination (133).

Similar to carbohydrates, the enzymes and transporters involved in lipid digestion and absorption are subjected to the action of dietary polyphenols. Most of the studies have focused on the inhibitory activity on of these compounds on PL, PLA2, or bile salts, due to their implication in fat absorption and their potential to prevent obesity and its complications. Fewer studies have focused on their interaction with CS esterase or lipid transporters (Table 4).

The digestion of dietary protein occurs first in the gastric and intestinal lumen through the action of HCl/pepsin and pancreatic proteases, including pepsin, trypsin, chymotrypsin, elastase, carboxypeptidase A, B, and aminopeptidase, and concludes with peptidases located in the brush-border membrane and cytoplasm of the enterocytes (130, 145). Pepsin and pancreatic proteases hydrolyze proteins in amino acids, di-, tri-, and oligopeptides. Posteriorly, amino acids are absorbed into the enterocytes through specific transporters according to their chemical structure (146), while di- and tripeptides are transported by the proton-dependent cotransporter, Pept-1 (130, 147). Oligopeptides need further hydrolysis by brush-border proteases to yield absorbable molecules (130). In the absorptive cells, the di- and tripeptides are hydrolyzed to amino acids by cytosolic peptidases, and these are exported to the bloodstream by facilitated diffusion (130).

As described above, a great proportion of the dietary polyphenols, including PACs, remain undigested in the intestine and reach the colon, where they are used as substrates by specific bacterial populations, stimulating their growth much like prebiotics do. In addition, PACs with high DP exert bacteriostatic and eventually bactericide effects that might also contribute to the modulation of IM composition and bacterial adhesion to colonocytes (164, 171).

Several studies have investigated the effect of PAC intake on the composition of the IM. In an in vitro model of colonic fermentation, the addition of (+) catechin (150 mg/l) for 48 h increased significantly the growth of C. coccoides–E. rectale group, Bifidobacterium spp., and E. coli and inhibited that of C. histolyticum. In the same conditions, the incubation with epicatechin only increased the growth of C. coccoides–E. rectale group (172). In a similar model, a grape seed extract with 28% PACs (600 mg/l) promoted the growth of Lactobacillus/Enterococcus group at 5 and 10 h, while the extract with 78% PAC at the same concentration only decreased C. histolyticum (173). In a recent interventional study, healthy volunteers had to ingest a dairy-based cocoa beverage with high (494 mg) or low (29 mg) flavanol content once a day for 4 weeks (174). The high flavanol beverage was composed of 110 mg monomers (catechin and epicatechin), 99 mg dimers, and 285 mg polymers (trimers to decamers), while the low flavanol beverage only contained 6, 11, and 12 mg of these compounds, respectively. The high flavanol drink significantly increased the counts of Bididobacterium, Lactobacillus, and Enterococcus and decreased those of C. histolyticum, while the low flavanol drink increased C. histolyticum. Both beverages stimulated the growth of the E. rectale/C. coccoides group. Accordingly, cocoa flavanols display a prebiotic potential by promoting a healthy IM in humans. In another human study, the effect of 2-week administration of 0.19 g/day of a PAC-rich grape seed extract (DP 2–15) was evaluated on the IM composition and fecal odor in healthy adults (175). An increase of Bifidobacterium and a non-significant decrease of Enterobacteriaceae were reported. Interestingly, the concentrations of potentially harmful bacterial metabolites such as ammonia, phenol, p-cresol, 4-ethylphenol, indol, and skatol (eventually implicated in the development of colorectal cancer) tended to decrease. Methyl mercaptan gas concentrations and fecal odor decreased significantly. Therefore, in this study, the modification of the IM by PACs was associated with a healthier environment in the colonic ecosystem.

The study of PAC effects on the IM is not restricted to humans. In ruminant nutrition, PACs are recognized to modulate fermentation processes in the rumen, inhibiting methane-producing archaea, decreasing bloating and protein degradation, and favoring the formation of conjugated linoleic acid. Some studies suggest that body weight gain, milk yields, and reproductive performance could be also improved by incorporating tannins in the animal diet, even if there is not yet a clear explanation for these beneficial effects (155).

The enzymatic degradation of flavonoids by the colonic microbiota results in a huge array of new metabolites. Bacterial enzymes may catalyze many reactions including hydrolysis, dehydroxylation, demethylation, decarboxylation, and deconjugation (176). Many flavonoids undergo ring-fission in which their C-ring is degraded, A-ring forms hydroxylated aromatic compounds, and B-ring phenolic acids derivatives (177). Clostridium and Eubacterium have been proposed as the main bacterial genera involved in the metabolism of phenolic compounds including flavan-3-ols (178). It is considered that about 40% of the flavan-3-ols ingested with green tea are converted to phenolic acid metabolites in the colon, which are excreted in urine. About 8% of them are methyl, glucuronide, and sulfate derivatives of flavanols, which reflect the fact that these metabolites were previously absorbed by the colonic epithelium (179).

The first study carried out in humans by Das in 1971 identified 11 metabolites in urine after (+)-catechin intake; among the most important were m-hydroxyphenylpropionic acid, δ-(3,4-dihydroxyphenyl)-γ-valerolactone, and δ-(3-hydroxyphenyl)-γ-valerolactone (180). These results were confirmed more recently by Li et al. who described phenyl-valerolactones as the main tea catechin metabolites produced by gut microorganisms and detected in human urine and blood (181). Consistently, Tzounis et al. reported that the incubation of (−)-epicatechin or (+)-catechin with fecal bacteria led to the generation of 5-(3′,4′-dihydroxyphenyl)-gamma-valerolactone, 5-phenyl-gamma-valerolactone, and phenylpropionic acid (172). Valerolactones were also detected in human urine by Ottaviani et al. after the consumption of flavanols and PCs (182). Aura et al. observed that 3-hydroxyphenyl propionic acid and 3-phenylpropionic acid were the main metabolites originated from (+)-catechin and (−)-epicatechin by human microbiota (183). These results are consistent with those described by Gonthier et al. who also identified urinary and plasma 3-hydroxybenzoic acid and 3-hydroxyhippuric acid in rats fed a diet supplemented with red wine polyphenols (184). Flavanol degradation by the microbiota seems to be a rapid process completed in 4–8 h, as reported in a “pig cecum in vitro model” (185). In another study with pig microbiota, it was shown that about 80% of PC A2 and 40% of cinnamtannin B1 were degraded after 8 h of incubation (186). 3-O-Gallate derivatives of epicatechin and epigallocatechin were extensively metabolized by a human fecal microbiota after 24-h-incubation but remained unaffected in presence of rat fecal microbiota, even after 48-h-incubation. These results suggest differences in the microbiota and their associated metabolic ability between both species (187). Phenylvaleric, phenylpropionic, cinnamic, phenylacetic, and benzoic acids as well as conjugated derivatives of benzoic acid have been identified in urine samples of rats fed with PC dimer B3, trimer C2, and catechin polymer (74). In another study, the main metabolites identified after the in vitro fermentation of purified PAC dimers with a human fecal microbiota were 2-(3,4-dihydroxyphenyl) acetic acid and 5-(3,4-dihydroxyphenyl)-γ-valerolactone (167). m-hydroxyphenylpropionic acid, ferulic acid, 3,4-dihydroxyphenylacetic acid, m-hydroxyphenylacetic acid, vanillic acid, and m-hydroxybenzoic acid were identified in urine samples of human volunteers after cocoa intake (188). These results were confirmed by Urpi-Sarda et al. who reported increased urinary concentrations of caffeic acid, ferulic acid, 3-hydroxyphenylacetic acid, vanillic acid, 3-hydroxybenzoic acid, 4-hydroxyhippuric acid, and hippuric acid in humans, while in rats were identified 3,4-dihydroxyphenylpropionic acid, m-coumaric acid, 3-hydroxyphenylacetic acid, protocatechuic acid and vanillic acid (189). These slight differences in the metabolite profile probably reflect interspecies differences in their microbiota composition. Ward et al. detected urinary 3-hydroxyphenylpropionic acid and 4-O-methylgallic acid after regular consumption of PAC-containing grape seed extract in humans (190). Interestingly, Jenner et al. measured the concentrations of dietary polyphenols and their bacterial metabolites in fecal waters from healthy omnivorous subjects under normal diet (i.e., without additional supplementation with fruits and vegetables). The major components detected were phenylacetic acid (479µM), 3-phenylpropionic acid (166µM), 3-(4-hydroxy)-phenylpropionic acid (68µM), 3,4-dihydroxycinnamic acid (52µM), benzoic acid (51µM), 3-hydroxyphenylacetic acid (46µM), and 4–hydroxyphenylacetic acid (19µM). Other phenolic acids ranged between 0.04 and 7µM in fecal waters (19). The bacterial degradation of black tea polyphenol-rich and red wine/grape juice extracts by the colonic microbiota were compared in an in vitro five-stage GI model (SHIME^®). The levels of gallic acid and 4-hydroxyphenylpropionic acid remained elevated throughout the colon with red wine/grape juice feeding, while these compounds were consumed in the distal colon and 3-phenylpropionic acid was strongly produced during a polyphenol-rich black tea extract feeding. The gut microbial production of phenolics was dependent on their location in the colon and the extract source (191).

The PAC DP is an important factor affecting the microbial metabolism of these compounds. The presence of PACs with high DP prevents the in vitro microbial metabolism of PACs (192), and A-type PACs are also more resistant to microbial degradation than B-type PACs (193). The main flavanol bacterial metabolites described in these studies are summarized in Table 6.

Whether these PAC metabolites can be used by the cells as a source of energy is unclear. A recent study showed that 3,4-dihydroxyphenylacetic acid, a microbial metabolite of quercetin and PACs, can protect against the mitochondrial dysfunction induced by CS in Min6 pancreatic β-cells (194). CS (320µM) decreased the mitochondrial membrane potential, the intracellular concentrations of ATP and the rate of oxygen consumption, while 3,4-dihydroxyphenylacetic acid (100–250µM) prevented, in a concentration-dependent manner, these mitochondrial function alterations. At 250µM, this metabolite was capable of preventing the drop in oxygen consumption and complex I activity, suggesting that 3,4-dihydroxyphenylacetic acid improves the energetic metabolism of these cells.

Inflammatory bowel diseases, mainly Crohn’s disease and ulcerative colitis, are considered as a growing problem of public health in the world. The etiology of these diseases remains poorly understood. Although the mechanisms underlying the occurrence of Crohn’s disease and ulcerative colitis differ, both diseases are characterized by a chronic inflammation and increased oxidative stress in the mucosa, with an inappropriate activation of the immune system (195). Considering that dietary PACs exhibit antiinflammatory, immunomodulatory, and antioxidant properties, a number of studies have tested their impact in animal models of IBD (Table 7).

The oral administration of apple (mainly containing B1, B2, and C1 PACs) or grape seed extract (with 75% PACs: 40% polymers, 14% dimers, 12% trimers, 8% tetramers) has been shown to prevent colonic damage in an experimental colitis model induced by dextran sodium sulfate or oxazolone in rodents (196–199). Accordingly, the mortality rate of the treated animals as well as their weight loss was lower, compared with those receiving the placebo. The grape extract also contributes to the normalization of the ileal villus morphology and mucosal thickness, reducing the histological severity score in the distal ileum and proximal colon (197). It has been postulated that one of the mechanisms implicated in these effects should be the modulation of the proportions of the TCR_γδ/TCR_αβ intraepithelial lymphocytes. Indeed, these cells play a key role in the maintenance of mucosal homeostasis, contributing to the modulation of the activated release of pro-inflammatory cytokines by epithelial cells and, therefore to the prevention of inflammatory states in the mucosa (196, 200).

Grape seed extract (containing PACs >95%, dimeric >1.8%, oligomers >60%) also exerted a protective effect in recurrent colitis induced by the intracolonic injections of 2,4,6-trinitrobenzene sulfonic acid, reducing the colonic weight/length ratio and the microscopic and macroscopic damage scores (201–204). Such effect relied on the antiinflammatory and antioxidant properties of the extract, as it was shown to inhibit the NF-κB signaling pathway, reducing the expression levels of TNF-α, p-IKKα/β, p-IκBα, and the translocation of NF-κB to the nucleus of colonic epithelial cells (203, 204). As a consequence of these events, the animal treated with the extract exhibited a lower neutrophil infiltration in their colonic mucosa, a decrease of IL-1β concentration, lipid peroxidation, and colonic inducible nitric oxide synthase activity, while the concentrations of IL-2 and IL-4, the antioxidants enzymes activities, and the levels of glutathione increased. These effects were comparable to those obtained with sulfasalazine, the standard drug used for IBD treatment (201–204).

The antiinflammatory properties of polymeric PACs from Pistacia vera L. nuts and apple has been studied in cell models (Caco-2 and T84 cell lines) simulating some conditions of IBD through their activation by pro-inflammatory cytokines such as IL-1β or IFN-γ/IL-1β/TNF-α (205, 206). Apple and pistachio PACs was shown to prevent the cytokine-induced translocation of NF-κB to the nucleus and the subsequent secretion of pro-inflammatory cytokines by these epithelial cells. More specifically, B1 PAC inhibited in a dose-dependent manner the expression of pro-inflammatory genes and repressed NF-κB-, IP-10-, and IL-8-promoters and STAT1-dependent signal transduction (206). The PAC extract from pistachio also contributes to conserve the integrity of IL-1β-stimulated Caco-2 cell monolayers by attenuating the disruption of tight-junctions, preventing the drop in transepithelial electrical resistance and the alterations of the gut barrier function (205).

In a recent in vitro study, we investigated the protective effect of PAC-containing polyphenol extracts from apple, avocado, cranberry, or grape and PACs microbial metabolites on the deleterious effect induced by p-cresol in human colonic epithelial cells (HT-29 and Caco-2). In HT-29 cells, the cranberry and avocado extracts prevented the loss in cell viability (measured as lactate dehydrogenase leakage) and the diminution in ATP contents, while bacterial metabolites only prevented the loss in cell viability. In Caco-2 cells, all extracts and bacterial metabolites prevented the p-cresol-induced alterations of barrier function (measured as transepithelial electrical resistance and fluorescein-dextran transport). These results suggest that PAC-containing polyphenol extracts and PAC metabolites likely contribute to the protection of the colonic mucosa against the deleterious effects of p-cresol (207).

Colorectal cancer is strongly related to dietary habits and is one of the most common cancers worldwide (208). Many studies have assessed the impact of flavonoids including PACs on the risk of this cancer (11, 209–212). In vitro studies using human cells lines derived from colonic adenocarcinoma have reported an inhibitory effect of PACs on cell proliferation and growth (213–215), an increase of apoptosis associated to caspase 3 (213, 215, 216) or caspase 8 activation (217), an arrest of the cell cycle in G1 phase (215), or the suppression of the angiogenic factors vascular endothelial growth factor and angiopoietin 1 (in a model of tumor growth in a xenografted chick chorio-allantoic membrane) (218). In animals, grape seed PACs (0.1–1.0% of diet) inhibited by 72–88% the formation of aberrant colonic crypt foci induced by azoxymethane and by 20–56% the activity of ornithine decarboxylase, an enzyme involved in cell growth and differentiation and related to tumor promotion, in the distal colon (219). In another study using the same model, the administration PACs (0.002%) and PC B2 (0.05%) decreased aberrant crypt foci formation and cell proliferation in the colonic epithelium and increased apoptosis compared with control rats (220).

In humans, a case–control study carried out in Italy reveals a decreased risk of colorectal cancer in the subjects with higher intakes of PACs [odds ratio (OR) = 0.74], being this effect more effective for rectal than for colon cancer (221). Another case–control study in Scotland showed a reduction in the risk of colorectal cancer with the consumption of catechin (OR = 0.68 for the highest vs lowest quartile), epicatechin (OR = 0.74), and PCs (OR = 0.78) (222), while a study in Spain showed a decreased risk in the quartile of highest intake of PAC compared with that of lowest intake [OR = 0.58 CI_95% (0.35–0.96), p = 0.02]. In this study, PACs were mainly brought by fruits, wine, and legumes and similar to the Italian study, the protective effect of PACs was more pronounced for rectal than for colon cancer (223). In opposition with these studies, another Scottish case–control study and the prospective Iowa Women’s Health Study did not detect any association between PAC intake and colorectal cancer (224, 225) and Bobe et al., in 1,859 participants of the Polyp Prevention Trial, reported an association between PAC intake and the risk of colorectal adenoma recurrence in men (226).

In conclusion, although in vitro, in vivo, and epidemiologic studies suggest that PACs could act as a chemopreventive agent, further studies are necessary to confirm these results and to clearly establish the subjacent mechanisms implicated and the type and concentration of PACs involved in the protection, prior to use them for the prevention or management of colorectal cancer.