Polyphenols in the treatment of inflammatory bowel disease and acute pancreatitis

Polyphenols are phytochemicals that are abundant in food and beverages derived from plants. Although no deficiency state has been described for them, increased intake of polyphenols appears to protect against disease in virtue of their anti‐inflammatory and vasculoprotective properties. This article focuses on four polyphenols with established anti‐inflammatory properties: resveratrol, epigallocatechin gallate, curcumin and quercetin. In rodents, ingestion or systemic administration of these agents inhibits nuclear factor κ B‐dependent gene expression and induces phase II antioxidant and detoxifying proteins. Conditions prevented and/or ameliorated by these polyphenols include inflammatory colitis and acute pancreatitis. Polyphenols also attenuate ischaemia‐reperfusion injury and endotoxemic sepsis, which has a role in the development of multiple organ dysfunction in severe acute pancreatitis. Enteral nutrition has an important role in the management of inflammatory bowel disease (IBD)—mainly of Crohn's disease, and of acute pancreatitis. Parenteral nutrition is reserved for refractory cases and disease‐associated complications. Artificial nutrition attempts to safely administer the essential and otherwise beneficial constituents of food to patients with an impaired ability to ingest or digest food; yet, polyphenols are not included in the formulas. We suggest that the addition of polyphenols to artificial nutritional formulas would improve the outcome of patients with IBD and acute pancreatitis in need of enteral or parenteral nutrition.

Polyphenols, a group of secondary plant metabolites, are non‐essential nutrients that probably contribute to human health

Plants, like other unicellular and multicellular organisms, contain ubiquitous organic molecules (eg, amino acids, carbohydrates and fatty acids) termed primary metabolites that are essential to cell structure and basic metabolism. These compounds also serve as substrates for the synthesis of an array of chemicals called secondary plant metabolites, which are accumulated at lower concentrations and are more variably distributed among different species. Once thought to be waste products, these agents are now considered to have a role in ecological interactions with friendly and hostile microorganisms and macroorganisms, and protection from environmental stressors. After their ingestion, certain secondary metabolites from edible plants interact beneficially with the regulatory domains of functional proteins that are shared by plants and herbivores, due to evolutionary conservation. Vitamins, for instance, act as enzyme coactivators and their insufficient intake results in some disease.¹,²,³ Polyphenolic substances (PPLs), the subject of this review, belong to an ill‐defined group of secondary plant metabolites called phytochemicals or bioactive compounds in food, for which a deficiency state has not been described (denying them the title of essential micronutrients) but they have attracted attention as mediators of the disease‐preventive effect of a healthful diet.⁴,⁵,⁶,⁷,⁸,⁹,¹⁰,¹¹,¹²,¹³,¹⁴,¹⁵ PPLs' status as a non‐essential, but an established health‐promoting dietary constituent is similar to that of the marine‐life‐derived, long chain ω‐3 polyunsaturated fatty acids (LC PUFAs), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). These can be synthesised to some extent from the shorter‐chain, plant‐derived, essential ω‐3 PUFA, α‐linolenic acid. Thus, EPA and DHA are not considered essential nutrients for adults.⁹ Yet, since the initial interest in their vasculoprotective properties during the 1970s, a formidable body of knowledge from mechanistic, epidemiological and large clinical studies has evolved, defining a role for the LC ω‐3 PUFAs in the secondary prevention of cardiovascular disease.¹⁶,¹⁷ Several studies have also shown the importance of supplementing artificial nutrition formulas with these fatty acids, by showing that they reduce morbidity, and possibly mortality, in several clinical scenarios.¹⁸,¹⁹ Thus, modern nutritional science has expanded its focus from overt nutrient‐deficiency states to the relationship of diet and nutritional status to the industrialised world's common diseases, including the benefits derived from consuming certain non‐essential bioactive components of food.⁹,²⁰ Although interest in PPLs' effect on health is growing, it is more recent²¹ than the investigations dealing with the ramifications of LC ω‐3 PUFA consumption. In addition, the growing list of identified PPLs, their variable distribution in different foods, their low bioavailability and their extensive and differential intestinal and hepatic metabolism have hampered defining their significance in the prevention of disease.¹²,¹³,¹⁴,¹⁵,²¹,²² The current level of evidence precludes recommending that patients take PPL‐containing supplements to treat or prevent disease, rather than eat a variety of healthful foods with emphasis on plant sources.⁹,¹³,²¹ Nevertheless, PPLs are increasingly recognised as health‐promoting phytochemicals as: (1) they attenuate animal models of many human illnesses that have an inflammatory component;¹³,²¹ (2) their consumption by humans on a healthful plant‐based diet is approximately 1 g a day—considerably higher than that of vitamins C and E and β‐carotene;¹³ (3) they have anti‐inflammatory, antioxidative and insulin‐sensitising effects and have enhanced endothelial function in clinical trials;²³,²⁴,²⁵,²⁶,²⁷,²⁸ and (4) they have reduced cardiovascular disease in epidemiological studies.²⁹

Polyphenols inhibit pro‐inflammatory transcription factors and enhance cytoprotective ones in vivo

PPLs are powerful free radical scavengers in vitro,¹³ but present opinion holds that their in vivo anti‐inflammatory effect stems from their interaction with proteins involved in signal transduction and gene expression.¹²,¹³,²¹,²²,³⁰ One important target of PPLs' action in vivo is the pivotal inflammatory transcription factor, nuclear factor κ B (NFκB; fig 1). Usually bound to and inactivated by the cytoplasmic inhibitor of κ B (IκB), NFκB is released from its association with the latter by activated IκB kinases (IκKs). Pathogen‐associated molecular patterns (eg, endotoxin), inflammatory cytokines (eg, tumour necrosis factor‐α (TNFα)), T‐cell‐activating signals and reactive oxygen species trigger signal transduction cascades that converge on, phosphorylate and rapidly activate IκKs. Released NFκB is then translocated into the nucleus where it promotes the expression of inflammatory gene products such as TNFα, inducible nitric oxide synthase (iNOS), cyclo‐oxygenase (COX)‐2, intercellular adhesion molecule and induces inflammatory activation of lymphocytes, monocyte/macrophages and endothelial cells.³¹ Although it may protect parenchymal cells from apoptosis, NFκB is considered to be a target for the treatment of diverse pathological conditions with an inflammatory component.³¹,³² By inhibiting IκK phosphorylation and/or preventing proteasomal degradation of IκB, numerous PPLs attenuate the in vivo NFκB activation induced in inflammatory disease states.¹²,¹³ These include resveratrol, which is found in grapes and wine and probably contributes to their cardioprotective effects;³³,³⁴ epigallocatechin gallate (EGCG), an anti‐inflammatory component of green tea;³⁵ curcumin, the major anti‐inflammatory PPL in turmeric;³⁶,³⁷ and quercetin, which is found in apples, onions, leafy green vegetables and tea.³⁸ Interestingly, the anti‐inflammatory properties of NSAIDs and salicylates are also partially attributable to inhibition of IκK.³⁹ As PPLs do not appear to prevent COX‐1‐mediated synthesis of constitutive/protective prostaglandins, they may be safer, if less powerful, anti‐inflammatory agents.

Figure 1 Polyphenols attenuate injury to parenchymal cells by down‐regulating inflammatory genes and up‐regulating cytoprotective ones. COX‐2, cyclo‐oxygenase‐2; HO‐1, haem‐oxygenase‐1; IκB, inhibitor of κ B; ICAM, intercellular adhesion molecule; iNOS, inducible nitric oxide synthase; NFκB, nuclear factor κ B; Nrf2, nuclear factor E2‐related factor; TNFα, tumour necrosis factor‐α.

In addition to down‐regulating the expression of inflammatory mediators, preneoplastic lesions also induce in vivo the transcription of cytoprotective phase II antioxidant and detoxifying molecules, such as haem‐oxygenase‐1 (HO‐1) and glutathione‐related enzymes⁴⁰,⁴¹,⁴²,⁴³ (fig 1). These are part of the endogenous defence system against xenobiotic and chemical toxicity and protect from carcinogenesis and inflammatory and autoimmune disease.⁴⁴ Electrophilic PPLs release nuclear factor E2‐related factor (Nrf2) from its complex with the cytoskeleton‐associated protein Keap1, through an interaction with thiols in Keap1. Nrf2 can then bind to the antioxidant responsive element, a regulatory element of phase II genes.⁴³ Interestingly, the Nrf2 pathway is physiologically activated by the very mediators of inflammatory pathways, such as increased oxidative stress, certain protein kinase(s) C and mitogen‐activated protein kinases, perhaps as a counter‐regulatory mechanism to reduce collateral tissue injury during inflammation.⁴⁴ Thus, through differential modulation of gene expression, PPLs inhibit the injurious consequences of cell injury, oxidative stress and inflammation while enhancing cytoprotective responses. Other genomic and non‐genomic actions of different PPLs are reviewed elsewhere.¹²,¹³,²⁹,³³,³⁴,³⁵,³⁶,⁴²

Let artificial nutrition be thy medicine

Artificial nutrition formulas are an attempt to partially reconstruct food from its essential and healthful components, in a form that can be safely administered by the enteral or parenteral route.⁴⁵ Over the years, formulas have been modified in an attempt to replicate healthful food intake, correct malnutrition, attenuate catabolism, inflammation and immunosuppression and to enhance disease resolution and healing.⁴⁵,⁴⁶ To this end, formulas may include pharmacological doses of nutrients (eg, glutamine and ω‐3 PUFAs), non‐essential constituents of food with healthful and disease‐modifying properties (eg, dietary fibre, prebiotics and probiotics), and even molecules not related to food (eg, transforming growth factor‐β).⁴⁶,⁴⁷,⁴⁸,⁴⁹ Yet, a patient who temporarily or permanently becomes dependent on enteral or parenteral nutrition is deprived of PPLs that she/he may have previously obtained from fruit, vegetables, tea, wine, chocolate or spices.

Although consumption of phytochemicals as a supplement does not necessarily confer the same benefit as ingesting foods rich in these compounds, addition of PPLs to otherwise complete nutritional formulas would bring them a step closer to what has been termed the “entire biological package” of food.⁵¹

By inhibiting NFκB, inclusion of the PPLs discussed here in artificial nutritional formulas may boost their therapeutic effect in acute and chronic diseases that necessitate enteral nutrition or parenteral nutrition and in which NFκB activation is implicated, such as sepsis,⁵² acute respiratory distress syndrome,⁵³ postoperative organ dysfunction,⁵⁴ cachexia,⁵⁵ IBD⁵⁶ and acute pancreatitis.⁵⁷

It has previously been proposed that phytochemicals other than vitamins may be important for patients receiving artificial nutrition.⁴⁵ The rest of this article reviews the literature suggesting that administration of the four above‐mentioned PPLs is beneficial in IBD and acute pancreatitis, thus offering a rationale for their inclusion in artificial nutrition formulas for patients with these conditions.

Polyphenols for IBD

Crohn's disease and ulcerative colitis, the two forms of IBD, are multifactorial disorders resulting from a dysfunctional epithelial, innate and adaptive immune response to intestinal microorganisms. Pharmacological treatment typically targets the ensuing robust autoimmune and inflammatory response that damages the gastrointestinal mucosa, impairing its absorptive and protective barrier function.⁵⁸ Most patients with IBD will suffer at some stage a degree of nutritional deficiency, owing to any combination of anorexia, malabsorption, enteropathic protein and blood loss, and a systemic inflammatory‐catabolic response. These have deleterious intestinal and extraintestinal consequences.⁵⁸,⁵⁹,⁶⁰

Paediatric and adult patients with Crohn's disease and malnutrition may need enteral nutrition and rarely parenteral nutrition to replenish macronutrients and micronutrients and enhance anabolism and growth. In patients with active Crohn's disease, enteral nutrition is effective as a remission‐inducing and a glucocorticoid‐sparing treatment to maintain remission. Parenteral nutrition may benefit malnourished patients before major surgery, those with spontaneous or postsurgical enterocutaneous fistulas and possibly also glucocorticoid‐resistant patients. Patients with Crohn's disease having short bowel syndrome after extensive resection of the intestines often depend on artificial nutrition as a source of nutrients. The benefit of nutritional support in ulcerative colitis has received less interest and is presently less supported.⁵⁸,⁵⁹,⁶⁰

Studies in rodent models of IBD⁶¹,⁶²,⁶³,⁶⁴,⁶⁵,⁶⁶,⁶⁷,⁶⁸,⁶⁹,⁷⁰,⁷¹,⁷²,⁷³,⁷⁴,⁷⁵,⁷⁶,⁷⁷ (table 1) indicate that administration of PPLs is effective in preventing and treating intestinal inflammation and injury. Acute or chronic colitis was induced in these studies by intrarectal administration of dinitrobenzene sulphate,⁶³,⁷¹ or trinitrobenzene sulphate,⁶¹,⁶²,⁶⁶,⁶⁷,⁶⁸,⁶⁹,⁷¹,⁷²,⁷³ addition of dextran sulphate sodium (DSS) to the drinking water,⁶⁴,⁷⁵,⁷⁶,⁷⁷ or by knock‐out of the interleukin‐2 gene⁶⁵ (IL‐2^‐/‐ mice). Rodents were treated with PPLs before, during and/or after induction of colitis, by oral (added to food or drinking water or through an orogastric tube),⁶¹,⁶²,⁶⁴,⁶⁵,⁶⁶,⁶⁷,⁶⁸,⁷⁰,⁷¹,⁷²,⁷³,⁷⁴,⁷⁵,⁷⁶,⁷⁷ rectal⁷² or intraperitoneal⁶³,⁶⁹,⁷⁰ administration. They were killed, and indices of disease were assessed between 48 h and 6 weeks after induction of colitis. Resveratrol,⁶¹,⁶² EGCG/green tea PPL extract,⁶³,⁶⁴,⁶⁵ curcumin⁶⁶,⁶⁷,⁶⁸,⁶⁹,⁷⁰,⁷¹ and quercetin and its naturally occurring glycones⁷²,⁷³,⁷⁴,⁷⁵,⁷⁶,⁷⁷ (quercetrin and rutin) reduced mortality rates, attenuated colonic (eg, diarrhoea, bloody stools) and extracolonic (eg, weight loss) signs of disease, colon macropathology and micropathology (eg, hyperaemia, ulcerations, inflammatory infiltrate, serosal adhesions) and/or indices of inflammation and autoimmunity (eg, colonic myeloperoxidase and NFκB activity, increased TNFα, IL‐1β, IL‐12, iNOS and reduced IL‐10, Crohn's disease 4⁺ T cell and neutrophil infiltration). In studies that assessed several dosages, the ratio between the highest and lowest effective dose ranged between 2 and 6, suggesting a relatively wide therapeutic window.

Table 1 Prophylactic and therapeutic effects of polyphenol administration in rodent models of inflammatory colitis.

Polyphenol	Model, reference	Route, dose, duration and timing of administration	Outcome	Comments
Resveratrol	TNBS enema in Wistar rats61	i.g. 5 or 10 mg/kg resveratrol 48, 24 and 1 h before and 24 h after TNBS	Reversal of weight loss; ↑ stool consistency, ↓colonic macropathology (presence and degree/extent of hyperaemia, ulceration, inflammation, adhesions), ↓ colonic histopathology (presence and degree/extent of necrosis, inflammatory infiltrate and mucus depletion); ↓ mucosal IL‐1β, COX‐2, PGD2 levels; ↓ colonic MPO activity	PGE2 levels were not significantly reduced; 5 mg/kg resveratrol improved fewer parameters than 10 mg/kg
	TNBS enema in Wistar rats62	i.g. 10 mg/kg/day resveratrol for 14 days starting 24 h after TNBS	↑ Stool consistency; colonic macropathology (weight, presence and degree/extent of hyperaemia, ulceration, inflammation, adhesions), ↓ colonic histopathology (presence and degree/extent of necrosis, inflammatory infiltrate, fibrosis and mucus depletion); ↓ mucosal TNFα and p65 NFκB; ↓ colonic MPO activity; ↑ mucosal PGE2; ↑ epithelial apoptosis	PGE2 levels were reduced during the chronic stage of colitis in untreated animals; resveratrol normalised PGE2 levels
EGCG/green tea polyphenol extract	DNBS enema, in Sprague–Dawley rats63	i.p. 50 mg/kg/day green tea polyphenol extract 1 day before and 4 days after DNBS	Reversal of weight loss; ↓ diarrhoea; ↓ colon weight and macropathology; ↓ histopathology (presence and degree/extent of oedema, necrosis, neutrophil infiltrate, haemorrhage); ↓ colonic TNFα, ICAM‐1 and nitrotyrosine levels; ↓ colonic MPO activity;
			↑ Colonic HO‐1
	DSS in drinking water, in BALB/C mice64	p.o. green tea polyphenol extract (in food) for 3 days before and 7 days after DSS	↓ Weight loss and diarrhoea; ↑ colon length; ↓histopathology (presence and degree/extent of inflammatory infiltrate, mucosal expansion, crypt epithelium disruption, ulceration); ↓ epithelial cytoskeleton distortion/fragmentation (as per laser scanning confocal microscopy); ↓ serum rGSH, SAA, TNFα, ↑ serum Hct and GSSG	Colonic glutathione status not significantly improved
	IL‐2‐/‐ C57BL/6J mice65	p.o. green tea polyphenol extract (5 g/l in drinking water) for 6 weeks after disease establishment	↑ Weight gain, ↓ colon weight; ↓ histopathology (presence and degree of inflammatory infiltrate, epithelial hyperplasia, goblet cell depletion, erosions, ulcerations and crypt abscesses); ↓ ex vivo colonic TNFα and IFN‐γ, ↓plasma SAA levels, ↑ Hct	IL‐2‐/‐ mice spontaneously develop autoimmune disease characterised by colitis, haemolytic anaemia and cachexia; all components responded to green tea intake
Curcumin	TNBS enema in Wistar rats66	p.o. curcumin (2% in food) for (a) 3 days before or (b) 14 days after TNBS	↑ Survival; ↓ weight loss; ↓ colon histopathology (presence and degree/extent of inflammatory infiltrate, thickening of colon wall, goblet cell depletion); ↑ colonic IκB; ↓ colonic NFκB activation and IL‐1β mRNA (both groups)	Prophylactic (group a) and therapeutic (group b) curcumin showed similar efficacy to prophylactic sulphasalisine
	TNBS enema in BALB/c mice67	i.g. curcumin at 25, 50, 100 or 300 mg/kg/day for 10 days before and 8 days after TNBS administration	↓ Weight loss; ↓ diarrhoea; ↓ colon weight and macropathology (presence and degree/extent of inflammation and ulceration); ↓ histopathology (presence and degree/extent of inflammatory infiltrate, thickening of colon wall, goblet cell depletion); ↓ colon MPO activity; ↓ colon NO, O2− (groups b–d). ↓ colon serine protease and NFκB activity; ↓ colon IFN‐γ, IL‐12 mRNA, ↑ IL‐4 mRNA (assessed in group b only).	Doses of 50–300 mg/kg/day were similarly effective, but 25 mg/kg/day did not significantly improve colitis. Curcumin attenuated the expression of T‐helper 1 profile of cytokines
	TNBS enema in C57BL/6 and BALB/c mice68	p.o. curcumin in food at (a) 2% for 3 days before; (b) 0.5% for 7 days after; (c) 2% for 7 days after; (d) 5% for 7 days after TNBS administration or (e) 2% for 5 days starting 2 days after TNBS	↑ Survival (group a); ↓ weight loss (groups a–d); ↓ histopathology (presence and degree of inflammatory infiltrate, vascularity, thickening of colon wall) (groups a,c,d); ↓ mucosal CD4+ T cell infiltration, NFκB activation and levels of IL‐6, IL‐12,TNFα and IFN‐γ mRNA (assessed in group a only)
	TNBS enema in Sprague–Dawley rats69	i.p. 30 mg/kg/day curcumin for 14 days after TNBS	↑Survival; ↓weight loss; macropathology (hyperaemia, ulceration, inflammation, adhesions); ↓ micropathology (epithelial necrosis, destruction of glands, inflammatory infiltrate, ↑granulation tissue); ↓colonic IL‐1, TNFα, IFN‐γ mRNA, ↑colonic IL‐4 mRNA;	Curcumin tended to be more effective than both dexamethasone 2 mg/kg/day and combined dexamethasone and curcumin. Only curcumin increased formation of PGJ2 and granulation tissue
			↑ Colonic PPAR‐γ and 15d‐PGJ2
	TNBS enema in Sprague–Dawley rats70	i.p. 30 or 60 mg/kg/day curcumin for 14 days after TNBS	↑Survival; ↓macropathology (hyperaemia, ulceration, inflammation, adhesions); ↓ micropathology (epithelial necrosis, destruction of glands, inflammatory infiltrate); ↓ colon MPO activity; ↓colonic iNOS, COX‐2, TNFα, IFN‐γ mRNA, ↑PGE2 (30 and 60 mg/kg/day)
	DNBS enema in C3H mice71	p.o. curcumin (0.25% in food) for 5 days before and 5 days after DNBS	↓ Colon macropathology (presence and degree of hyperaemia, bowel wall thickening, ulceration, inflammation; ↓ colon histopathology (presence and degree of inflammatory infiltrate, ulceration, necrosis); ↓colonic IL‐1β; ↓colonic MPO, p38 MAPK and NFκB activity
Quercetin and its naturally occurring glycones, quercetrin and rutin	TNBS enema in Sprague–Dawley rats72	(a) i.g. rutin 10 mg/kg/day from day 1 to 6 after TNBS;	↓ Colon macropathology (presence and degree/extent of hyperaemia, inflammation ulceration, scabbing, stricture, serosal adhesion); ↓colon MPO activity (rutin and quercetin 25‐100 μM/day)	p.r. quercetin 10 μM/day was ineffective. PO rutin and PR quercetin were as effective as PO sulphasalazine and PR 5‐ASA; rutin was metabolised to quercetin by colonic microbial glycosidases
		(b) p.r. quercetin (enema) 10, 25, 50 or 100 μM/day from day 1 to 6 after TNBS
	TNBS enema in Wistar rats73	i.g. quercetrin 1 or 5 mg/kg (single dose) 2 h before TNBS	↓ Colonic alkaline phosphatase and NOS activity; ↑ colonic water and electrolyte absorption; ↓ colonic MDA (1 and 5 mk/kg)	No reduction in colonic MPO activity, perhaps due to the single dose and/or insufficient time for pharmacological effect
	TNBS enema in Wistar rats74	Acute colitis: i.g. quercitrin 0.125, 0.25, 0.5, 1, 5, 10, 25 or 50 mg/kg 2 h before to and 24 h after TNBS;	Acute colitis: ↓ diahroea, ↓ colon macropathology, ↓ colon MPO and AP activity, ↑ colon glutathione content, ↑ colon fluid absorption. (1 and 5 mg/kg only)	0.125–0.5 mg/kg and 10–50 mg/kg were not effective in acute colitis; quercetrin 1 and 5 mg/kg did not reduce markers of inflammation and oxidative stress in chronic colitis despite reduced macropathology
		Chronic colitis: i.g. quercitrin 1 or 5 mg/kg, 2 h before to and daily for 2–4 weeks after TNBS	Chronic colitis: ↓ diahroea, ↓ colon macropathology, ↑ colon fluid absorption (1 and 5 mg/kg/day)
	DSS in drinking water, in Wistar rats75	i.g. quercetin or quercetrin, 1 mg/kg/day for 10 days with DSS.	↓ Disease activity index (weight loss, diarrhoea, blood in faeces), ↓ colonic MPO and NFκB activity, ↓ colonic IL‐1β, TNFα, iNOS (quercetrin)	i.g. quercetin was ineffective. Quercetrin is metabolised by colonic microbial rhamnosidases to the active quercetin
	DSS in drinking water, in ICR mice76	p.o quercetin or rutin (0.1% in food) for 7 days before and 7 days concomitantly with DSS	↓ Weight loss; ↓ colonic shortening; ↓ histopathology (presence and degree/extent of oedema, inflammation, regenerative changes, fibrosis); ↓ colon and pΦ IL‐1β, ↓ colon IL‐1β, IL‐6, GM‐CSF and COX‐2 mRNA (rutin)	i.g. quercetin was ineffective
	DSS in drinking water, in Wistar rats77	p.o quercitrin in drinking water: (a) 1 mg/kg/day together with DSS for 8 days; (b) 5 mg/kg/day together with DSS for 8 days; (c) 1 mg/kg/day after 10 days of DSS, for 5 days	↓ Disease activity index (weight loss, stool consistency, rectal bleeding) (all groups); ↓histopathology (presence and degree/extent of inflammatory infiltrate, ulceration, mucus depletion, mitotic activity in crypts, oedema, vascularity, fibrin deposition) (a,c); ↓ colon MPO and iNOS activity (a,c); ↑ colon glutathione content (all groups); ↓ colon NFκB activation	Early administration of quercitrin at 1 mg/kg/day improved more parameters than did 5 mg/kg/day; colon LTB4 levels were not reduced by quercitrin
			(assessed in group c only)

AP, alkaline phosphatase; 5‐ASA, 5‐aminosalicylic acid; CCK‐8, cholecystokinin octapeptide; COX, cyclo‐oxygenase; CYP, cytochrome P450; 15d‐PGJ₂, 15‐deoxy‐D 12, 14‐prostaglandin J₂; DNBS, dinitrobenzene sulphate; DSS, dextran sulphate sodium; EGCG, epigallocatechin gallate; ICAM, intercellular adhesion molecule; IFN, interferon; i.g., intragastric; i.p., intraperitoneal; p.o., oral; i.v., intravenous; s.c., subcutaneous; iNOS, inducible nitric oxide synthase; MAPK, mitogen‐activated protein kinase; MDA, malondialdehyde; MPO, myeloperoxidase; NFκB, nuclear factor κ B; PPAR, peroxisome proliferator‐activated receptor; SAA, serum amyloid A; TNBS, trinitrobenzene sulphate; TNF, tumour necrosis factor.

Quercetrin (3‐rhamnosylquercetin) and rutin (3‐O‐rhamnosyl‐glucosyl‐quercetin), the two commonly occurring glycones of quercetin, lack anti‐inflammatory properties in vitro but act as pro‐drugs when ingested for the treatment of colitis. They are not well absorbed in the small intestine and are metabolised in the colon to the locally active aglycone form, quercetin, by microbial rhamnosidases.⁷²,⁷⁵ Quercetin was found to be effective when administered intrarectally,⁷² but its ingestion failed to ameliorate colitis despite the efficacy of its glycosides,⁷²,⁷³,⁷⁴,⁷⁵,⁷⁶,⁷⁷ probably due to its avid absorption in the small bowel. Protection against small intestinal disease was not discussed in these studies, but their findings suggest that ingestion of PPL glycones may not be beneficial for inflammation proximal to the colon.

Inhibition of NFκB and of leucocyte and T‐cell infiltration and activation probably contributes to PPLs' therapeutic effect in colitis.⁵⁶ PPL induction of HO‐1,⁴¹,⁴² which reduces oxidative stress and increases carbon monoxide formation, may also blunt injury in IBD.⁷⁷,⁷⁸,⁷⁹ A 2‐week treatment with curcumin, itself a peroxisome proliferator‐activated receptor‐γ agonist,⁸⁰,⁸¹,⁸² increased the intestinal level of this transcription factor and its endogenous agonist 15‐deoxy‐D 12, 14‐prostaglandin J₂.⁶⁹ Thus, intestinal peroxisome proliferator‐activated receptor‐γ activation, which inhibits NFκB and attenuates colitis,⁸³ may also underlie curcumin's therapeutic action. PGE₂ levels are reduced in some models of chronic colitis, and may be increased by PPLs, despite reduced transcription of COX‐2.⁶²,⁶⁹,⁷⁰ In one study, curcumin, but not dexamethasone, increased the formation of granulation tissue 2 weeks after induction of trinitrobenzene sulphate colitis.⁶⁹ Thus, PPLs may somehow modulate the eicosanoid response such as to promote the resolution of inflammation and enhance wound healing.

Animal studies suggest that PPLs improve graft function and survival after organ transplantation, acting alone or synergistically with cyclosporine or mycophenolate mofetil, suggesting that they have immunosuppressive properties as well.⁸⁴,⁸⁵,⁸⁶ Oral curcumin and quercetin reduced acute rejection and immunosuppressive‐drug side effects in a double‐blind study on cadaveric renal transplant recipients.⁸⁷ Also of pertinence to the nutritional support of patients with IBD in a catabolic state is the finding that intraperitoneal administration of resveratrol for 10 days attenuated skeletal muscle cachexia in tumour‐bearing mice.⁸⁸ A small pilot study showed that oral curcumin therapy improved clinical symptoms, histopathology and laboratory indices in all five patients with ulcerative colitis and four out of five patients with Crohn's disease who had an insufficient response to conventional treatments.⁸⁹ The patients with ulcerative colitis received 550 mg curcumin twice a day for 1 month followed by the same dose once a day for another month. Patients with Crohn's disease were treated with 360 mg curcumin three times a day for 1 month followed by 360 mg four times a day for two additional months.⁸⁹ Finally, the preliminary results from a 6‐month placebo‐controlled trial of curcumin therapy in 89 patients with ulcerative colitis in remission were recently presented.⁹⁰ All patients were on 5‐aminosalicylic acid therapy. Relapse was seen in 5% of curcumin‐treated and 21% of placebo‐treated patients, and no serious adverse effects were reported.⁹⁰

Polyphenols for acute pancreatitis

Acute pancreatitis is an inflammatory process triggered by a number of instigating factors, with variable involvement of peri‐pancreatic tissues and remote organs. After initial acinar cell injury by a specific pathogenic agent or process, the release and activation of proteases and leucocyte infiltration exacerbate a destructive autodigestive and inflammatory cascade. Most cases are mild and resolve spontaneously within 48 h with supportive therapy. But in patients with severe acute pancreatitis (SAP), spillover of inflammatory molecules can initiate early or late multiple organ dysfunction syndrome.⁹⁰ Early enteral nutrition, rather than nothing per os or parenteral nutrition, is the recommended approach for patients with severe, and especially necrotising acute pancreatitis, despite the rationale behind avoiding stimulation of the pancreas in such a scenario. Parenteral nutrition is reserved for patients with SAP that develop critical illness and in whom attempts at administering enteral nutrition have failed.⁹¹,⁹²

PPLs are protective in experimental acute pancreatitis⁹³,⁹⁴,⁹⁵,⁹⁶,⁹⁷,⁹⁸,⁹⁹,¹⁰⁰,¹⁰¹,¹⁰²,¹⁰³,¹⁰⁴ induced either by injection of tert‐butyl hydroperoxide⁷⁹ or sodium taurocholate into the pancreatic duct,⁸¹,⁸²,⁸⁵ intraperitoneal⁸³ or intravenous⁸⁶ injection of cerulein, intraperitoneal injection of DL‐ethionine⁸⁴ or intravenous cholecystokinin octapeptide (CCK‐8) together with oral ethanol (table 2).⁸⁶ Resveratrol,⁷⁹,⁸⁰,⁸¹,⁸² EGCG/green tea PPL extract⁸³,⁸⁴ and curcumin⁸⁵,⁸⁶ attenuate acute pancreatitis in rodents when administered before, concomitantly to or after the instigating agent. Pancreatic damage (pancreatic macropathology and micropathology, neutrophil infiltrate, trypsin activity, lipoperoxides and inflammatory cytokines) was reduced in all studies except for two that were performed by the same group,¹⁰²,¹⁰³ and in which curcumin did reduce serum amylase, TNFα and IL‐6 as well as bacterial translocation.

Table 2 Prophylactic and therapeutic effects of polyphenol administration in rodent models of acute pancreatitis.

Polyphenol	Model and reference	Route, dose, duration and timing of administration	Outcomes	Comments
Resveratrol	ButOOH injection into pancreatic duct of Wistar rats94	i.p. resveratrol 2 mg/day for 8 days before ButOOH injection (rats weighed ∼300 g)	↓Pancreatic weight, ↓pancreatic histopathology (acinar vacuolisation, focal oedema, necrosis, haemorrhage), ↓ pancreatic carbonyl and SH groups, ↓acinar RES cistern dilation and mitochondrial swelling (per electron microscopy), ↓serum amylase activity	Administration of diethylstilbesterol, a synthetic analogue of resveratrol, was equally effective
	Taurocholate injection into pancreatic duct of Sprague–Dawley rats95,96	i.p. resveratrol 30 mg/kg (single dose) after taurocholate injection	↓ Pancreatic histopathology (haemorrhage, microthrombi, exudates, inflammatory infiltrate), ↓ pancreatic NFκB activity, ↓pancreatic TNFα and IL‐8, ↓lung histopathology (alveolar septum thickening, interstitial oedema, inflammatory infiltrate), ↓lung water content and capillary permeability, ↓ lung MPO activity, ↓ lung ICAM‐1, ↓ blood viscosity
	Taurocholate injection into pancreatic duct of Sprague–Dawley rats97	i.v. 10 mg/kg resveratrol 5 min after sodium taurocholate	↓Micropathology (intralobular and interlobular oedema, inflammatory infiltrate, haemorrhage, necrosis); ↓PΦ NFκB activation and iNOS activity; ↓serum TNFα, IL‐1β and NO
	Taurocholate injection into pancreatic duct of Sprague–Dawley rats98	i.v. 20 mg/kg resveratrol after sodium taurocholate	↓Macropathology (oedema, necrosis, haemorrhage, saponification); ↓amount and turbidity of ascitic fluid; ↓micropathology (intralobular and interlobular oedema, inflammatory infiltrate, haemorrhage, thrombosis, necrosis); ↓pancreatic TBARS, ↓pancreatic MPO activity, ↑pancreatic SOD
	s.c. CCK‐8 in Wistar rats99	i.p. 10 mg/kg resveratrol 30 min before CCK‐8	↓ Pancreatic wet weight, ↓pancreatic histopathology (acinar cell vacuolisation, intralobule and interlobule oedema, inflammatory infiltrate)
EGCG/green tea polyphenol extract	i.p. cerulein in Wistar rats100	p.o. green tea polyphenol extract (in drinking water) for 10 days before i.p. cerulein	↓ Pancreatic wet weight, ↓pancreatic histopathology (acinar cell vacuolisation, intralobule and interlobule oedema), ↓ pancreatic MDA, ↓ serum amylase
	i.p. DL‐ethionine in Wistar rats101	p.o. green tea polyphenol extract (in drinking water) for 10 days before i.p. DL‐ethionine)	↓ Pancreatic wet weight, ↓ pancreatic histopathology (acinar cell necrosis, intralobule and interslobule oedema), ↓ pancreatic MDA, ↓ serum amylase
Curcumin	Taurocholate injection into pancreatic duct of Wistar rats102	(a) i.g. curcumin 100 mg/kg/day for 20 days before and 6 days after taurocholate administration; (b) as in group (a) and i.p. ciprofloxacillin and metronidazole for 6 days after taurocholate administration	↓ Bacterial translocation; ↓ serum amylase, MDA and NO	Combination of curcumin and antibiotics produced better results than curcumin alone, but neither groups reduced pancreatic histopathologic scores
	Taurocholate injection into pancreatic duct of Wistar rats103	i.g. curcumin 100 mg/kg/day for 20 days before and 6 days after taurocholate administration	↓ Serum TNFα and Il‐6	Curcumin did not reduce tissue injury
	(1) i.v. cerulein or (2) p.o. ethanol + i.v. CCK‐8 in Sprague–Dawley rats104	i.v. curcumin 200 mg/kg (single dose) concomitantly to cerulein/CCK‐8	↓ Pancreatic histopathology, ↓pancreatic trypsin activity, ↓neutrophil infiltration, ↓pancreatic NFκB and AP‐1 activation, ↑pancreatic IκB, ↓pancreatic IL‐6, TNFα, iNOS mRNA, ↓serum amylase and lipase	Curcumin did not reduce CCK‐mediated amylase secretion, suggesting that the reduced pancreatic trypsin activity is due to its antineutrophil effect

CCK‐8, cholecystokinin octapeptide; EGCG, epigallocatechin gallate; ICAM, intercellular adhesion molecule; iNOS, inducible nitric oxide synthase; i.g., intragastric; i.p., intraperitoneal; p.o., oral; i.v., intravenous; s.c., subcutaneous; MDA, malondialdehyde; MPO, myeloperoxidase; NFκB, nuclear factor κ B; SOD, sphincter of Oddi dysfunction; TNF, tumour necrosis factor.

Inhibition of NFκB⁵⁷ and enhanced expression of HO‐1¹⁰⁵ may mediate PPLs' protective effects in acute pancreatitis.

Some evidence suggests that administration of PPLs early in the course of acute pancreatitis may prevent the development of multiple organ dysfunction and septic shock: resveratrol attenuated SAP‐associated acute respiratory distress syndrome.⁹⁵,⁹⁶ Prophylactic administration of PPLs attenuates ischaemia‐reperfusion injury to the bowel,¹⁰⁶,¹⁰⁷ which has been implicated in bacterial translocation, pancreatic infection and development of sepsis in SAP.¹⁰⁸,¹⁰⁹ PPLs also protect against ischaemia‐reperfusion injury to the kidneys, liver and heart, ¹¹⁰,¹¹¹,¹¹²,¹¹³,¹¹⁴,¹¹⁵,¹¹⁶ and may thus reduce dysfunction of these organs in the context of the severe inflammatory response syndrome. Finally, pre‐emptive administration of PPLs significantly improves survival rates in rodent models of endotoxinemia,¹¹⁷,¹¹⁸,¹¹⁹ whereas intravenous curcumin treatment attenuates the liver injury, systemic inflammation and mortality associated with caecal ligation and puncture, even when initiated 5 h after the puncture.⁸⁰

Some considerations for future research and development

Beyond confirming their efficacy, the animal studies reviewed here contribute little to the formulation of a PPL extract (ie, which PPL(s), what dosage(s) would be effective when administered with enteral nutrition/parenteral nutrition). Effective dosage(s) varied among different models, rodent species and studies. Controlled clinical studies of PPL supplementation⁹⁰ should provide a closer approximation, but the slow coadministration of PPLs together with nutrient‐rich enteral nutrition/parenteral nutrition formulas may influence PPLs' pharmacodynamics. For instance, quercetrin supplementation potentiates ω‐3 PUFAs' anti‐inflammatory effect in DSS‐induced colitis¹²⁰ but quercetin inhibits the induction of heat shock protein 70 (hsp70),¹²¹ which partially mediates the beneficial response to glutamine.¹²²,¹²³ What then would be the sum effect of combining quercetin (or its glycones) with an “immunonutritional” package containing glutamine and ω‐3 PUFAs? Other PPLs actually induce hsp70 or at least do not interfere with its cytoprotective properties.¹²⁴,¹²⁵

Uncertainties notwithstanding, the animal studies do suggest that PPLs have a wide therapeutic window, and that potentially, numerous combinations/dosages would be beneficial. PPLs generally remain non‐toxic, even at relatively high doses¹²⁶,¹²⁷,¹²⁸,¹²⁹,¹³⁰ and safety studies in healthy volunteers, followed by dose‐finding/safety studies in patients receiving enteral nutrition/parenteral nutrition should help identify therapeutic regimens.

Food–drug interactions are another issue which needs to be discussed. Quercetin inhibits cytochrome P450 (CYP) 3A4¹³¹ and increases blood levels of cyclosporine (which is occasionally used to treat IBD)¹³² in healthy volunteers.¹³³ Consumption of a green tea extract did not interfere with CYP3A4‐mediated or 2D6‐mediated drug metabolism.¹³⁴ To the best of our knowledge, the effect of curcumin and resveratrol on clinically relevant pharmacokinetics in humans has not been assessed. Despite their potential to increase cyclosporine levels, the PPLs reviewed here actually protect against cyclosporine‐induced nephrotoxicity.¹³⁵,¹³⁶,¹³⁷,¹³⁸,¹³⁹,¹⁴⁰ while enhancing its therapeutic immunosuppressive effect (in organ transplantation).⁸⁵,⁸⁶

In conclusion, PPLs are anti‐inflammatory and cytoprotective constituents of plant‐derived food that reduce the severity of experimental IBD and acute pancreatitis and may safely enhance the therapeutic effect of enteral and parenteral nutrition in patients with these conditions. Further preclinical and clinical studies are indicated.

The body responds to cell injury by initiating an inflammatory response aimed at immobilising, destroying and removing the injurious agent, and by cytoprotective adaptations that reduce direct and co‐lateral damage to parenchymal cells. Numerous injurious agents initiate intracellular cascades that converge on transcription factors, such as NFκB and Nrf2, allowing their translocation to the nucleus where they enhance, respectively, the transcription of pro‐inflammatory and cytoprotective genes. Through their divergent effect on NFκB and Nrf2, PPLs inhibit the synthesis of potentially injurious mediators while enhancing that of antioxidative and anti‐inflammatory ones.