Thérèse Sergent

Anti-inflammatory effects of dietary phenolic compounds in an in vitro model of inflamed human intestinal epithelium.

Phenolic compounds (PCs) are considered to possess anti-inflammatory properties and therefore were proposed as an alternative natural approach to prevent or treat chronic inflammatory diseases. However their effects are not fully understood, particularly at the intestinal level. To further understand their mode of action at the molecular level during intestinal inflammation, an in vitro model of inflamed human intestinal epithelium was established. Different representative dietary PCs, i.e. resveratrol, ellagic and ferulic acids, curcumin, quercetin, chrysin, (-)-epigallocatechin-3-gallate (EGCG) and genistein, were selected. To mimic intestinal inflammation, differentiated Caco-2 cells cultivated in bicameral inserts, in a serum-free medium, were treated with a cocktail of pro-inflammatory substances: interleukin (IL)-1β, tumor necrosis factor-α, interferon-γ and lipopolysaccharides. The inflammatory state was characterized by a leaky epithelial barrier (attenuation of the transepithelial electrical resistance) and by an over-expression at the mRNA and protein levels for pro-inflammatory markers, i.e. IL-6, IL-8 and monocyte chemoattractant protein-1 (MCP-1), quantified by ELISA and by gene expression analysis using a low-density array allowing the evaluation of expression level for 46 genes relevant of the intestinal inflammation and functional metabolism. Treatment with PCs, used at a realistic intestinal concentration, did not affect cell permeability. In inflamed cells, the incubation with genistein reduced the IL-6 and MCP-1 overproduction, to ca. 50% of the control, whereas EGCG provoked a decrease in the IL-6 and IL-8 over-secretion, by 50 and 60%, respectively. This occurred for both flavonoids without any concomitant inhibition of the corresponding mRNA expression. All the PCs generated a specific gene expression profile, with genistein the most efficient in the downregulation of the expression, or over-expression, of inflammatory genes notably those linked to the arachidonic metabolism pathway. In conclusion, this study provides evidence that genistein and EGCG downregulate the inflammatory response in inflamed intestinal epithelial cells by a pathway implicating largely a post-transcriptional regulatory mechanism.

Deoxynivalenol affects in vitro intestinal epithelial cell barrier integrity through inhibition of protein synthesis

Deoxynivalenol (DON), one of the most common mycotoxin contaminants of raw and processed cereal food, adversely affects the gastrointestinal tract. Since DON acts as a protein synthesis inhibitor, the constantly renewing intestinal epithelium could be particularly sensitive to DON. We analyzed the toxicological effects of DON on intestinal epithelial protein synthesis and barrier integrity. Differentiated Caco-2 cells, as a widely used model of the human intestinal barrier, were exposed to realistic intestinal concentrations of DON (50, 500 and 5000 ng/ml) during 24h. DON caused a concentration-dependent decrease in total protein content associated with a reduction in the incorporation of [(3)H]-leucine, demonstrating its inhibitory effect on protein synthesis. DON simultaneously increased the paracellular permeability of the monolayer as reflected through a decreased transepithelial electrical resistance associated with an increased paracellular flux of the tracer [(3)H]-mannitol. A concentration-dependent reduction in the expression level of the tight junction constituent claudin-4 was demonstrated by Western blot, which was not due to diminished transcription, increased degradation, or NF-kappaB, ERK or JNK activation, and was also observed for a tight junction independent protein, i.e. intestinal alkaline phosphatase. These results demonstrate a dual toxicological effect of DON on differentiated Caco-2 cells consisting in an inhibition of protein synthesis as well as an increase in monolayer permeability, and moreover suggest a possible link between them through diminished synthesis of the tight junction constituent claudin-4.

Interaction of ochratoxin A with human intestinal Caco-2 cells: possible implication of a multidrug resistance-associated protein (MRP2).

Ochratoxin A (OTA), a nephrotoxic mycotoxin, is absorbed from small intestine and, in plasma, binds to serum albumin. Prolonged half-live results from reabsorption by proximal tubules and enterohepatic circulation. The mechanism whereby OTA crosses intestine was investigated by means of a cell culture system consisting of Caco-2 cells, as in vitro model of human intestinal epithelium. Cytotoxicity assays on proliferating Caco-2 cells showed that 0.4 microM OTA inhibits MTT reduction by 50%. Transepithelial transport and intracellular accumulation of OTA were studied in Caco-2 cells, differentiated in bicameral inserts. At pH 7.4, OTA is transported preferentially in basolateral (BL) to apical (AP) direction, suggesting a net secretion. Conditions closer to in vivo situation in duodenum (AP pH 6.0, BL pH 7.4) increase intracellular accumulation and transepithelial transport. AP to BL transport becomes higher than BL to AP transport, suggesting OTA absorption. Addition of serum albumin in BL compartment further increases OTA absorption across Caco-2 cells and suggests that in vivo OTA transport from serosal to luminal side of enterocytes is prevented, due to its binding to plasma proteins. Competition experiments showed that carrier systems for large neutral amino acids, H(+)/dipeptides cotransporter, organic anion (p-aminohippurate) carrier and organic anion transporter (oatp) are not implicated in OTA transport across Caco-2 cells, in contrast to what was reported in kidney and liver. AP and BL transport and intracellular accumulation of OTA are increased in the presence of non specific inhibitors of MRPs (indomethacin, genistein and probenecid) and of 1-chloro-2,4-dinitrobenzene (biotransformed into 2,4-dinitrophenyl-gluthatione, a specific inhibitor of MRPs), but are affected by verapamil, an inhibitor of P-gp. This suggests that the multidrug resistance-associated protein (MRP2) could be implicated in transepithelial transport. Therefore, absorption of OTA across the intestinal mucosa would be limited thanks to its excretion through MRP2 at the apical pole of enterocytes.

Molecular and cellular effects of food contaminants and secondary plant components and their plausible interactions at the intestinal level

The intestinal mucosa is not simply a barrier allowing entry of compounds such as nutrients or chemicals, and restricting that of others. Intestinal cells and activities perform selective absorption, biotransformations and efflux back to the lumen. Furthermore, food substances affect both bioavailability and intestinal function. Some are able to act as transcriptional regulators and enzyme modulators. This review points out plausible interactions between food contaminants and/or natural constituents at molecular and cellular levels and focuses on the effects of classical (pesticides and veterinary drugs), environmental (heavy metals, PCBs, dioxins, etc.) and food processing generated (PAHs, heterocyclic amines, etc.) contaminants on absorption, metabolism and efflux. Special attention is given to secondary metabolites of molds (mycotoxins) and plants (polyphenols). Molecular targets are briefly described as well as regulation mechanisms. Where possible, data referred to deal with human intestinal functions in vivo, and with in vitro studies on human intestinal Caco-2 cells; however, since data related to the intestine are rather scarce, effects on molecular targets in liver are also considered. This review also points out the urgent need for fully validated high throughput in vitro tools to screen combinations of substances, at realistic intestinal concentrations. A higher priority could then be given to combinations of nutrients, xenobiotics and food contaminants, with hazardous or beneficial impacts on human health.

Food flavonoid aryl hydrocarbon receptor-mediated agonistic/antagonistic/synergic activities in human and rat reporter gene assays.

Aryl hydrocarbon receptor (AhR) is a ligand-activated transcription factor mediating the adverse effects of dioxins and polycyclic aromatic hydrocarbons (PAHs). In this study, we investigated the genetic-, time-, dose-, species- and tissue-dependent AhR-mediated agonistic/antagonistic activities of three food flavonoids: quercetin, chrysin and genistein. To that end, four stably transfected cell lines were used in cell-based luciferase reporter gene assays: three lines were transformed with the ptKLuc vector harbouring four dioxin-responsive elements (DREs) upstream of the thymidine kinase promoter and the luciferase gene (HepG2-Luc, T-47D-Luc and H4IIE-ULg). The fourth is a patented cell line transformed with a different construct: H4IIE DR-CALUX((R)). Both H4IIE cells were compared for their genetic construction. Human hepatoma (HepG2-Luc) and human breast tumour (T-47D-Luc) cells were compared for tissue-dependent effects. Rat hepatoma (H4IIE-ULg) and human hepatoma (HepG2-Luc) cells were compared for species-dependent activities. We concluded that quercetin, chrysin and genistein act in a time-, dose-, species- and tissue-specific way. For example, genistein displayed agonistic activities when exposed to rat hepatoma cells during 6h but not after 24h. Flavonoids displayed agonistic/antagonistic activities in human breast tumour cells, depending on the exposure time, while in human hepatoma cells, only antagonistic activities of flavonoids were measured. In addition, we report, in all the cells, a synergy between an isoflavone and two food contaminants; the 2,3,7,8-tetrachlorodibenzo-p-dioxin and 3-methylcholanthrene, a PAH. In rat cells, this synergy occurred when cells were exposed to flavonoids and contaminant for 6h, while it was observed in human cells only after 24h.

CYP1A1 and CYP3A4 modulation by dietary flavonoids in human intestinal Caco-2 cells

Flavonoids have been proposed to exert beneficial effects in a multitude of disease states. However, evidence of potential toxic actions has also emerged. Since large doses of flavonoids can be encountered in the intestine simultaneously with ingested drugs and pollutants, this study aimed at investigating nine individual flavonoid compounds and their interactions with the major intestinal isoforms of cytochrome P450, i.e. CYPs 1A1 and 3A4, using human intestinal Caco-2 cells cultivated in a serum-free medium. Genistein, quercetin and chrysin provoked a dose-dependent inducing effect on the CYP1A1 activity, measured with the EROD assay. However, they did not affect the CYP1A1 mRNA expression, suggesting they are not aryl hydrocarbon receptor-ligands in intestinal cells and act at a post-transcriptional level. Chrysin, at 50 microM, was detected as a potent inhibitor of the TCDD-induced CYP1A1 activity, leading the activity to ca. 10% of the TCDD-control value (n=3), this effect involving, at least partly, direct interactions at the enzyme level. Quercetin was also shown to significantly inhibit the constitutive CYP3A4 activity, measured by the 6beta-(OH)-testosterone assay, and to impair its induction by 1,25-vitamin D(3). Chrysin, quercetin and genistein, were detected as significant inhibitors of the 1,25-vitamin D(3)-induced CYP3A4 activity. In vivo, these effects could result in reduced activation of procarcinogens and/or in drug bioavailability limitation. They underline the importance of intestinal studies to assess food safety and health risks linked to the ingestion of flavonoid-enriched supplements or functional foods.

Anti-inflammatory effect and modulation of cytochrome P450 activities by Artemisia annua tea infusions in human intestinal Caco-2 cells

In an attempt to understand the beneficial health effects of Artemisia annua other than its anti-malaria properties, extracts from different cultivars prepared as tea infusions were investigated using Caco-2 cells on the intestinal inflammation and cytochrome P450 (CYP) activities. The characterisation of their phenolic compound (PC) profile revealed rosmarinic and chlorogenic acids as the main PCs. The extracts, assayed on Caco-2 cells at a plausible intestinal concentration, significantly decreased the secretion of pro-inflammatory cytokines, IL-8 and IL-6. This effect could be attributable at least to their content in rosmarinic acid, detected as a potent anti-inflammatory compound. The extracts also inhibited the activity of CYP3A4, whose expression was induced by 1,25-dihydroxyvitamin D(3), and of CYP1A1, induced by benzo(a)pyrene. Our results highlight the advantage of drinking A. annua infusions for their potent anti-inflammatory effect, linked to PC content, which could synergise their antimalarial activity.

CYP1A1 induction and CYP3A4 inhibition by the fungicide imazalil in the human intestinal Caco-2 cells - comparison with other conazole pesticides.

Imazalil (IMA) is a widely used imidazole-antifungal pesticide and, therefore, a food contaminant. This compound is also used as a drug (enilconazole). As intestine is the first site of exposure to ingested drugs and pollutants, we have investigated the effects of IMA, at realistic intestinal concentrations, on xenobiotic-metabolizing enzymes and efflux pumps by using Caco-2 cells, as a validated in vitro model of the human intestinal absorptive epithelium. For comparison, other conazole fungicides, i.e. ketoconazole, propiconazole and tebuconazole, were also studied. IMA induced cytochrome P450 (CYP) 1A1 activity to the same extent as benzo(a)pyrene (B(a)P) or 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), in a dose- and time-dependent manner. Cell-free aryl hydrocarbon receptor (AhR) binding assay and reporter gene assay suggested that IMA is not an AhR-ligand, implying that IMA-mediated induction should involve an AhR-independent pathway. Moreover, IMA strongly inhibited the CYP3A4 activity in 1,25-vitamin D(3)-induced Caco-2 cells. The other fungicides had weak or nil effects on CYP activities. Study of the apical efflux pump activities revealed that ketoconazole inhibited both P-glycoprotein (Pgp) and multidrug resistance-associated protein 2 (MRP-2) or breast cancer resistance protein (BCRP), whereas IMA and other fungicides did not. Our results imply that coingestion of IMA-contaminated food and CYP3A4- or CYP1A1-metabolizable drugs or chemicals could lead to drug bioavailability modulation or toxicological interactions, with possible adverse effects for human health.

Chemistry around imidazopyrazine and ibuprofen : discovery of novel fatty acid amide hydrolase (FAAH) inhibitors

Based on the imidazo-[1,2-a]-pyrazin-3-(7H)-one scaffold, a dual action prodrug has been designed for combining antioxidant and anti-inflammatory activities, possibly unmasked upon oxidation. The construction of the target-molecule requires two building blocks, namely a 2-amino-1,4-pyrazine and a 2-ketoaldehyde. Attempts to synthesize the 2-ketoaldehyde (5a) derived from ibuprofen failed, but led to the corresponding 2-ketoaldoxime (7a) which could not be condensed with the pyrazine synthons. However, a model compound, i.e. phenylglyoxal aldoxime, reacted well under microwave activation to furnish novel imidazo[1,2-a]-pyrazine-3-(7H)-imine derivatives (18a,b). These heterobicycles behave as antioxidants by inhibiting the lipid peroxidation, and one compound (18b) is endowed with a significant anti-inflammatory effect in a cellular test. Unexpectedly, all the synthetic intermediates derived from ibuprofen are good inhibitors of FAAH, the most active compound (4a) featuring the 1,3-dithian-2-yl motif.

Study of the gastrointestinal biotransformation of zearalenone in a Caco-2 cell culture system with liquid chromatographic methods

A high‐performance liquid chromatography method with fluorescence detection (HPLC‐FLD) was developed and validated for the detection of zearalenone (ZON), α‐zearalenol (α‐ZOL) and β‐zearalenol (β‐ZOL) in in vitro biological samples. Furthermore, a liquid chromatography–tandem mass spectrometry (LC‐MS/MS) method was developed and validated for the detection of ZON, α‐ZOL, β‐ZOL, α‐zearalanol (α‐ZAL) and β‐zearalanol (β‐ZAL) in in vitro biological samples. Zearalanone (ZAN) was used as internal standard in both methods. The limit of detection/limit of quantitation (LOD/LOQ) values for ZON, α‐ZOL and β‐ZOL were 2/7, 2/7 and 4/13 µg l−1, respectively, for the HPLC‐FLD method. For the LC‐MS/MS method LOD/LOQ values for ZON, α‐ZOL, β‐ZOL, α‐ZAL and β‐ZAL were 6/20, 5/17, 4/14, 9/30 and 6/19 µg l−1, respectively. Within‐day and between‐day precision were less then 11 and 14%, respectively for the HPLC‐FLD method, and both less then 20% for the LC‐MS/MS method. The recovery of ZON and its metabolites ranged between 73 and 89% for the HPLC‐FLD method and between 69 and 112% for the LC‐MS/MS method. The methods were used for the detection of the compounds in in vitro biological samples, obtained with human intestinal Caco‐2 cells culture experiments. The 8‐days post‐confluent Caco‐2 cells were treated with ZON or a mixture of ZON and imazalil (IMA). After an incubation time of 24 h the samples were analysed with the HPLC‐FLD method. Neither ZON nor its derivatives were detected in the samples. The disappearance of ZON could possibly point out the formation of phase II metabolites like glucuronide conjugates. Therefore, samples were pretreated with β‐glucuronidase before LC‐MS/MS analysis. The LC‐MS/MS results showed that ZON, α‐ZOL and β‐ZOL could only be detected in the β‐glucuronidase pretreated samples. This confirmed the formation of glucuronide conjugates and the hydroxylation of ZON during the incubation with Caco‐2 cells. Copyright