In vitro model of inflamed intestinal barrier

Abstract

The present invention relates to an intestinal epithelium model accurately reproducing the pathophysiological mechanisms observed in vivo in the context an inflammatory condition of the intestinal barrier. The model according to the present invention comprises two compartments separated by a semi-permeable membrane. The first compartment, corresponding to the apical pole of the intestinal epithelium, comprises a coculture of Caco-2 cells differentiated into enterocytes and HT29-MTX cells differentiated into goblet cells. The second compartment, corresponding to the basolateral pole of the intestinal epithelium, comprises a culture of THP-1 monocytic cells differentiated into macrophages. The model according to the present invention is characterized in that the cells contained in the first compartment produce interleukin 6 (IL-6) at a concentration of more than 100 pg / ml and interleukin 8 (IL-8) at a concentration of more than 150 pg / ml, and the cells contained in the second compartment produce tumour necrosis factor α (TNF-α) at a concentration of more than 40 pg / ml and interleukin 1β (IL-1β) at a concentration of more than 90 pg / ml. The present invention also relates to the method for obtaining this model and also to a method for selecting candidate compounds for the treatment of inflammatory bowel diseases by verifying their ability to stop the inflammation as soon as the permeability is increased, or verifying the restoration of the permeability and of the mucus layer.

Description

Description Title: In vitro model of inflamed intestinal barrier Technical Field

[0001] The present invention relates to the field of cell models and particularly to an in vitro model of an inflamed intestinal barrier. The invention also relates to the method for obtaining this model and its applications. Previous technique

[0002] The intestinal epithelium forms a protective and permeable barrier that plays a major role in defending the host against the luminal microenvironment, while maintaining selective permeability and nutrient absorption. The epithelium is covered by a monolayer composed of various intestinal epithelial cells, such as enterocytes, goblet cells, M cells, and Paneth cells, and cooperates with immune cells in the lamina propria. Close cooperation between all these cells maintains intestinal homeostasis and mucosal integrity, which are challenged daily by external factors in the luminal environment. Defects in intestinal barrier function are associated with diseases of the gastrointestinal tract, such as inflammatory bowel disease (IBD).

[0003] The exact etiology of IBD is unknown, although multiple causes are involved. Studies have reported that in IBD, the role of the intestinal epithelium as a physical and immune barrier can be disrupted (Ramos and Papadakis, Mayo Clinic proceedings (2019), 94(1), 155-165). In particular, an inappropriate inflammatory response is observed, which can be explained by increased intestinal barrier permeability, impaired immune barrier function, and abnormal cooperation between the gut microbiota and the intestinal barrier (Rogler, The Lancet. Gastroenterology & Hepatology (2017), 2(7), 521-530).

[0004] Several in vitro, ex vivo, and in vivo models representing the pathophysiological manifestations of the disease have been developed for the study and development of drug treatments. While very useful for studying the physiology of an organism as a whole, in vivo models have numerous limitations, notably due to interspecies differences that make extrapolations to humans difficult, as well as ethical considerations. In vitro models using human cells have the advantage of reducing animal experimentation and offering the possibility of testing treatments in a simple and reproducible manner.

[0005] Originally used in cancer research, Caco-2 cell monoculture, an immortalized human colorectal adenocarcinoma cell line, is now recognized as a valid model of small bowel epithelium. In particular, Caco-2 cells have been shown to be capable of producing inflammatory markers in response to other cytokines such as interleukin IL1-0, tumor necrosis factor TNF-α and / or interferon IFN-γ (Rodriguez-Ramiro et al., The British Journal of Nutrition (2013), 110(2), 206-215; Hollebeeck et al., Food & Function (2012), 3(8), 875-885) and / or in some cases, in response to external stimuli such as contact with lipopolysaccharides (LPS) present on the membranes of microorganisms (Weglarz et al., Digestive Diseases and Sciences (2007), 52(1), 93-102).

[0006] Unfortunately, using a Caco-2 monoculture as the sole intestinal cell type does not accurately reproduce the complex composition of the intestinal barrier. In particular, such a monoculture fails to produce permeability similar to the intestinal barrier, nor does it allow for the formation of a mucus layer on the luminal side, which severely limits absorption and permeability studies. Therefore, several co-culture models have been developed to better represent the complex composition of the intestinal barrier (Béduneau et al., European Journal of Pharmaceutics and Biopharmaceutics (2014), 87(2), 290-298). Among these, we can notably mention the Caco-2 / HT29-MTX co-culture model, which provides a more physiologically relevant system, since it includes mucus-producing HT29-MTX cells that mimic the function of goblet cells (Lesuffleur et al., Cancer Research (1990), 50(19), 6334-6343).

[0007] In vitro co-culture models frequently use insertion systems composed of rigid porous membranes to perform permeability studies (Pan et al., International Journal of Food Sciences and Nutrition (2015), 66(6), 680-685).

[0008] Subsequently, a tri-culture model based on combining the Caco-2 / HT29-MTX co-culture model with human THP-1 monocytic cells differentiated into an adherent macrophage-like phenotype was developed. A state corresponding to an inflammatory state was obtained by stimulating THP-1 cells with a cocktail of inflammatory activators including LPS and the pro-inflammatory cytokines IL-10 and TNF-α (Kaulmann et al., Molecular Nutrition & Food Research (2016), 60(5), 992-1005). However, this type of model does not take into account the modification of the composition of the mucus and the permeability of the cell monolayer, as described in in vivo models (Michielan and D'Incà, Mediators of Inflammation (2015), 2015 628157) or found in patients with IBD (Sôderholm et al., Gastroenterology (1999), 1 17(1 ), 65-72).The inflamed intestinal barrier is characterized by several features, including altered mucus composition, modified intercellular junctions leading to increased membrane permeability associated with exposure to bacterial antigens that perpetuates the inflammatory state, altered gut microbiota composition, and recruitment of inflammatory cells that secrete pro-inflammatory chemokines, fueling the immune response. A good model should therefore be able to reproduce all these features.

[0009] Unfortunately, there are currently no in vitro models that can accurately and comprehensively simulate the state of the inflamed intestinal barrier. Therefore, the development of anti-inflammatory drugs for the treatment of IBD and intestinal tissue inflammation requires evaluation in animals to verify the efficacy of these substances.

[0010] Therefore, there is a real need for a reliable intestinal epithelium model capable of accurately reproducing an inflammatory state of the intestinal barrier. Such a model must, in particular, allow for the observation of changes in mucus composition as well as alterations in the permeability of the cell monolayer. This would reduce the need for animal experimentation in the study of inflammatory bowel diseases, especially in the search for potential therapeutic candidates applicable to these diseases. Summary

[0011] The invention is defined by the claims.

[0012] The present inventors have succeeded in developing a model of intestinal epithelium that faithfully reproduces the pathophysiological mechanisms observed in vivo in the context of an inflammatory state of the intestinal barrier.

[0013] This new model allows for consideration of changes in mucus composition, the effect of intestinal mucosal permeability on the passage of molecules through the inflamed epithelium and their ability to interact with epithelial cells, and also mimics the presence of immune cells infiltrating the intestinal epithelium during inflammation. The model according to the present invention can be used to identify and select molecules or delivery systems with therapeutic effects in inflammatory bowel diseases, thus reducing the need for animal models.

[0014] The inventors have demonstrated that the present model is stable over time and very easily reproducible. It incorporates the pathophysiological mechanisms observed in IBD, including changes in mucus composition, increased membrane permeability linked to alterations in intercellular junctions, and the secretion of pro-inflammatory cytokines.

[0015] The inflamed intestinal epithelium model according to the present invention is based on a triculture of Caco-2 / HT29-MX / THP-1 cells treated with different compounds. The inventors have demonstrated that treating a Caco-2 / HT29-MX / THP-1 cell triculture with compounds such as DSS and LPS accurately recreates the conditions observed in an inflamed intestinal barrier. Although used in some models to induce inflammation in vivo, the mechanism by which DSS affects intestinal permeability remains unclear. Studies using DSS to induce inflammation in certain rodent strains involve several days of DSS exposure. However, in the context of the present invention, the inventors have demonstrated that a short application of DSS in vitro yields an accurate model of inflamed intestinal epithelium.

[0016] Treatment of Caco-2 / HT29-MX / THP-1 cell triculture with compounds such as DSS and LPS yields a model with well-defined characteristics, particularly with regard to the expression of pro-inflammatory cytokines, mucin expression, and measured transepithelial electrical resistance.

[0017] Thus, according to a first aspect, the present invention relates to a model of intestinal epithelium comprising two compartments separated by a semipermeable membrane, said model comprising: - in the first compartment, corresponding to the apical pole of the intestinal epithelium, a co-culture of Caco-2 cells differentiated into enterocytes and HT29-MTX cells differentiated into goblet cells; and - in the second compartment, corresponding to the basolateral pole of the intestinal epithelium, a culture of THP-1 monocytic cells differentiated into macrophages, said model being characterized in that the cells contained in the first compartment produce interleukin (IL-6) at a concentration greater than 100 pg / mL, and interleukin 8 (IL-8) at a concentration greater than 150 pg / mL; and the cells contained in the second compartment produce tumor necrosis factor α (TNF-α) at ​​a concentration greater than 40 pg / mL and interleukin 1P (IL-1P) at a concentration greater than 90 pg / mL.

[0018] The model according to the present invention also exhibits a transepithelial electrical resistance lower than 85% compared to the transepithelial electrical resistance measured in a control non-inflamed intestinal epithelium model, and altered mucin expression. In particular, the expression of MUC2 mucin is reduced by at least 30% compared to a control non-inflamed intestinal epithelium model, and the expression of MUC5AC mucin is increased by at least 50% compared to a control non-inflamed intestinal epithelium model.

[0019] According to a second aspect, the present invention also relates to a method for obtaining a model of intestinal epithelium comprising the following steps: - the culture, in a medium comprising two compartments separated by a semi-permeable membrane, of Caco-2 cells differentiated into enterocytes and HT29-MTX cells differentiated into goblet cells in the first compartment, and of THP-1 monocytic cells differentiated into macrophages in the second compartment; and - bringing the differentiated Caco-2 and HT29-MTX cells contained in the first compartment into contact with at least one compound chosen from the group consisting of sodium dextran sulfate (DSS), trinitrobenzenesulfonic acid (TNBS) and dinitrobenzenesulfonic acid (DNBS); Then - bringing differentiated THP-1 monocyte cells contained in the second compartment into contact with a compound chosen from the group consisting of lipopolysaccharides (LPS), TNF-α, interferon-γ (IFN-γ), and IL-1P; or - the replacement of said at least one compound chosen from DSS, TNBS and DNBS with LPS, TNF-α, IFN-γ and / or IL-1 p.

[0020] The inventors have demonstrated that the model according to the present invention, by faithfully reproducing the state of the inflamed epithelium, allows for the reliable selection of compounds with therapeutic activity in the context of inflammatory bowel diseases. The inventors have indeed demonstrated that the model according to the present invention makes it possible, for example, to observe the effects of betamethasone, a synthetic corticosteroid used primarily for its anti-inflammatory effect, in the treatment of IBD.

[0021] Also, according to a third aspect, the present invention relates to a method for selecting candidate compounds for the treatment of inflammatory bowel diseases, said method comprising the steps of: A) bringing a candidate compound into contact with the intestinal epithelium model according to the present invention; B) selection of the compound if it allows obtaining, in said model, at least one of the following effects a) to e): a) A decrease of at least 10% in the level of IL-6 and / or IL-8 in the first compartment; b) A decrease of at least 10% in the level of TNF-a and / or IL-10 in the second compartment; c) An increase in the expression of MUC2 of at least 5%; d) A decrease in the expression of MUC5AC of at least 5%; e) An increase in transepithelial electrical resistance of at least 10%. Brief description of the figures: Fig. 1

[0022] [Fig. 1]: Effect of apical stimulation with DSS (5% w / v) and basolateral exposure with LPS (100 ng / mL) on the synthesis of the pro-inflammatory cytokines TNF-α / IL-10 by macrophages and IL-6 / IL-8 by intestinal cells. This figure shows the synthesis of the pro-inflammatory cytokines TNF-α / IL-10 by macrophages and IL-6 / IL-8 by intestinal cells after exposure of Caco-2 / HT29-MTX cells to DSS and THP-1 cells to LPS. Fig. 2

[0023] [Fig. 2]: Effect of apical DSS stimulation (5% m / V) and basolateral LPS exposure (100 ng / mL) on the expression of MUC2, MUC5AC, ZO-1, and Claudin-1 by Caco-2 / HT29-MTX cells. This figure shows the expression levels of mucins MUC2 and MUC5AC and tight junctions ZO-1 and Claudin-1, after contacting Caco-2 / HT29-MTX cells with DSS and THP-1 cells with LPS. Fig. 3

[0024] [Fig. 3]: Effect of apical betamethasone (100 pM) on the synthesis of the pro-inflammatory cytokines TNF-α / IL-10 by macrophages and IL-6 / IL-8 by intestinal cells, after induction of inflammation. This figure shows the synthesis of the pro-inflammatory cytokines TNF-α / IL-10 by macrophages and IL-6 / IL-8 by intestinal cells, after induction of inflammation and apical treatment with betamethasone. Fig. 4

[0025] [Fig. 4]: Effect of apical betamethasone (100 pM) on the expression of MUC2, MUC5AC, ZO-1, and Claudin-1 by Caco-2 / HT29-MTX cells after inflammation induction. This figure shows the expression levels of the mucins MUC2 and MUC5AC and the tight junctions ZO-1 and Claudin-1 after inflammation induction and apical treatment with betamethasone. Fig. 5

[0026] [Fig. 5]: Effect of DSS / LPS exposure on epithelial layer permeability in a Caco-2 / HT29-MTX / THP-1 tri-co-culture model. Epithelial layer permeability was characterized after exposure to DSS (5% mA) and stimulation with LPS (0.1 pg / mL) for 24 ha by measuring JL transport (5A) and rhodamine 123 efflux (5B). Values ​​are expressed as mean ± standard error of the mean, with three wells per condition in three independent experiments (n = 9 wells / conditions; *P < 0.05 compared to the unstimulated control). Fig. 6

[0027] [Fig. 6]: Images of bioadhesion to the intestinal membrane of a fluorescent marker conjugated to anti-MUC5AC, obtained by confocal microscopy. Different experimental conditions were studied: non-inflamed membrane + marker / anti-MUC5AC (6A); inflamed membrane without marker / anti-MUC5AC (6B); inflamed membrane + marker (6C); inflamed membrane + marker / anti-MUC5AC (6D). Detailed description of the invention

[0028] A first aspect of the present invention relates to a model of intestinal epithelium comprising two compartments separated by a semipermeable membrane, said model comprising: - in the first compartment, corresponding to the apical pole of the intestinal epithelium, a co-culture of Caco-2 cells differentiated into enterocytes and HT29-MTX cells differentiated into goblet cells; and - in the second compartment, corresponding to the basolateral pole of the intestinal epithelium, a culture of THP-1 monocytic cells differentiated into macrophages, said model being characterized in that the cells contained in the first compartment produce interleukin (IL-6) at a concentration greater than 100 pg / mL, and interleukin 8 (IL-8) at a concentration greater than 150 pg / mL; and the cells contained in the second compartment produce tumor necrosis factor α (TNF-α) at ​​a concentration greater than 40 pg / mL and interleukin 1P (IL-1P) at a concentration greater than 90 pg / mL.

[0029] The present invention aims to provide a model of intestinal epithelium. The model according to the invention is advantageously a model of vertebrate intestinal epithelium, particularly human.

[0030] The model according to the invention therefore comprises two compartments. Each compartment represents a specific pole of the intestinal epithelium. As those skilled in the art know, the wall of the vertebrate gastrointestinal tract has several characteristic layers and is thus oriented. The innermost layer is the intestinal epithelium (and therefore the part reproduced in the model according to the present invention), which lines the interior of the tract, the lumen. The underlying tissue layer, the submucosa, is made of connective tissue and is backed by layers of smooth muscle. Another layer, the serosa, composed of epithelium and connective tissue, covers the external surface of the tract. The so-called "apical" pole is the pole that is in contact with the lumen, as opposed to the "basolateral" pole, which corresponds to the pole on the side of the external surface of the digestive tract.

[0031] The model according to the present invention aims to reproduce the intestinal epithelium. It comprises two compartments: one corresponding to the apical pole of the intestinal epithelium, i.e. the part in contact with the intestinal lumen, and the other corresponding to the basolateral pole, i.e. the part in contact with the submucosa.

[0032] Each compartment is a clean culture medium. In the context of the present invention, the two compartments are separated by a "semi-permeable membrane," that is, a membrane with pores ranging in diameter from 0.2 to 10 µm, preferably from 0.4 to 8 µm. Examples of membrane materials for the semi-permeable membrane include, but are not limited to, polycarbonate (PC), poly(diallyldimethylammonium chloride) (PDDA), polyamide (PA), or polyethylene terephthalate (PET).

[0033] Those skilled in the art are aware of numerous supports that can be used in the context of the present invention. In a particularly advantageous manner, the compartments used are cell culture inserts. Transwell® technology is particularly well-suited for this purpose. Such inserts are commonly used as supports in models of intestinal epithelia (see, for example, Rahman et al., International Journal of Molecular Sciences, 22(24), 13472).

[0034] In the context of the present invention, the compartment corresponding to the apical pole of the intestinal epithelium (upper compartment of the Transwell® when this technology is used) comprises a co-culture of Caco-2 and HT29-MTX cells. "Co-culture" here refers to the joint culture, in the same medium, of these two cell lines.

[0035] Caco-2 cells are an immortalized cell line of human colorectal adenocarcinoma cells. Caco-2 cells are capable of spontaneous differentiation into enterocytes and, after differentiation, express the main enzyme systems found in enterocytes of the intestinal epithelium. This is the most widely used cell line as an intestinal model for studies of absorption, transport, and bioavailability. These cells are readily available to those skilled in the art. For example, they are available under the reference HTB-37™ from the American Type Culture Collection (ATCC).

[0036] In the context of the present invention, the Caco-2 cells used in the intestinal epithelium model are differentiated into enterocytes. Those skilled in the art are well aware of the conditions required to achieve complete differentiation of a population of Caco-2 cells into enterocytes. A method for this purpose is disclosed, for example, in the experimental portion of this application.

[0037] HT29-MTX cells are a stable subpopulation of mucus-secreting cells obtained after treating human HT29 colorectal cancer cells with methotrexate (Lesuffleur et al., Cancer Research 50.19 (1990): 6334-6343). HT29-MTX cells spontaneously differentiate into goblet cells. They have the capacity to produce MUC2-type mucus, which is predominantly produced in the small intestine and colon. These cells are widely used in co-culture systems with the Caco-2 cell line to mimic the human intestinal barrier in absorption and permeability studies. These cells are readily available to those skilled in the art from specialized suppliers.

[0038] In the context of the present invention, HT29-MTX cells used in the intestinal epithelium model are differentiated into goblet cells. Those skilled in the art are well aware of the conditions required to achieve complete differentiation of a population of HT29-MTX cells into goblet cells. A method for this purpose is disclosed, for example, in the experimental portion of this application.

[0039] According to a preferred embodiment, the co-culture of Caco-2 and HT29-MTX cells is in the form of a monolayer of polarized cells, with a heterogeneous distribution of the two cell types at the membrane level.

[0040] In the context of the present invention, Caco-2 cells were typically seeded at a concentration half that of HT29-MTX cells. For example, Caco-2 cells were seeded at a concentration of 5,000 cells per compartment and HT29-MTX cells were seeded at a concentration of 10,000 cells per compartment.

[0041] Typically, Caco-2 cells are seeded at a concentration of 10,000 to 20,000 cells per cm³ 2 , preferably at a concentration of 14,000 and 16,000 cells per cm 2 , even more preferably at a concentration of approximately 15,000 cells per cm 2 .

[0042] Typically, HT29-MTX cells are seeded at a concentration of 20,000 to 40,000 cells per cm³ 2 , preferably at a concentration of 28,000 and 32,000 cells per cm 2, even more preferably at a concentration of approximately 30,000 cells per cm 2 .

[0043] Within the framework of the present invention, the compartment corresponding to the basolateral pole of the intestinal epithelium (lower compartment of the Transwell® when this technology is used) comprises a culture of THP-1 monocytic cells differentiated into macrophages.

[0044] "THP-1 monocyte cells" are a human monocyte cell line derived from acute monocytic leukemia. This cell line can differentiate into macrophages upon treatment with phorbol-12-myristate-13-acetate (PMA). It is a cell line widely used in inflammatory models. These cells are readily available to those skilled in the art. They are, for example, available from ATCC under the reference TIB-202™. Those skilled in the art will know the conditions required to obtain a differentiated macrophage population from THP-1 monocyte cells. For example, they can treat THP-1 monocyte cells with PMA at a concentration of 200 ng / mL for 48 to 72 hours to obtain a fully differentiated macrophage population.

[0045] Typically, THP-1 monocytic cells are seeded at a concentration of 10,000 to 20,000 cells per cm³ 2, preferably at a concentration of 14,000 and 16,000 cells per cm 2 , even more preferably at a concentration of approximately 15,000 cells per cm 2 .

[0046] A person skilled in the art knows what conditions and medium to use for culturing these different cell lines. The culture medium can be any physiologically acceptable medium that allows the growth of the cell lines used herein. Physiologically acceptable media are media containing electrolytes such as sodium, potassium, magnesium, and / or calcium salts, including anions such as chloride, carbonate, hydroxide, or caprylate. A suitable culture medium according to the present invention may include D-glucose, non-essential amino acids, vitamins, inorganic salts, and albumin. A suitable culture medium that can be used to culture the co-culture of Caco-2 and HT29-MTX cells is, for example, DMEM (Dulbecco's Modified Eagle Medium) GlutaMAX™, marketed by TermoFisher Scientific.In parallel, a suitable culture medium that can be used to culture differentiated THP-1 monocytic cells is, for example, RPMI (Roswell Park Memorial Institute) / RPMI-1640 medium, marketed by TermoFisher Scientific. These media can be supplemented with albumin, for example, by adding fetal bovine serum, and / or with antibiotics such as penicillin and / or streptomycin.

[0047] The model according to the present invention is characterized in that the cells contained in the first compartment produce interleukin (IL-6) at a concentration greater than 100 pg / mL, preferably at a concentration greater than 300 pg / mL, even more preferred at a concentration above 600 pg / mL; and of interleukin 8 (IL-8) at a concentration above 150 pg / mL, preferably at a concentration above 500 pg / mL, even more preferably at a concentration above 1000 pg / mL.

[0048] The model according to the present invention is characterized in that the cells contained in the second compartment produce tumor necrosis factor α (TNF-α) at ​​a concentration greater than 40 pg / mL, preferably at a concentration greater than 400 pg / mL, and even more preferably at a concentration greater than 700 pg / mL; and interleukin 10 (IL-10) at a concentration greater than 90 pg / mL, preferably at a concentration greater than 200 pg / mL, and even more preferably at a concentration greater than 250 pg / mL.

[0049] IL-6, IL-8, TNF-α, and IL-10 are well-known inflammatory cytokines. Their human sequences are available in the UniProt database under references P05231 for interleukin 6, P10145 for interleukin 8, P01375 for TNF-α, and P01584 for interleukin 10.

[0050] A person skilled in the art is also perfectly aware of how to quantify these proteins and assess the concentration of these cytokines produced in a cell culture medium. The experimental part of this application provides an example of a method for such quantification within the scope of the present invention. Typically, the concentration of the different cytokines can be assessed in the cell culture supernatant (in the first or second compartment, as appropriate) by enzyme-linked immunosorbent assay, for example, by ELISA.

[0051] The model according to the present invention can also be characterized by the transepithelial electrical resistance measured therein. Transepithelial electrical resistance, or TEER, is a measurement used to assess the degree of impermeability of a tissue. It consists of measuring the electrical resistance across a cell layer to evaluate its integrity and permeability. TEER measurement is a well-known technique commonly used by those skilled in the art, particularly in the context of in vitro models (Srinivasan, Balaji, et al. Journal of Laboratory Automation 20.2 (2015): 107-126). A method for measuring TEER is presented in the experimental part of the present invention.

[0052] In the model according to the present invention, the transepithelial electrical resistance is less than 85%, preferably less than 50%, even more preferably less than 35% compared to the transepithelial electrical resistance measured in a control non-inflamed intestinal epithelium model.

[0053] The model according to the present invention can also be characterized by the expression of mucins by the cell cultures constituting it, and in particular by the HT29-MTX cells contained in the first compartment. Mucins are macromolecules that are components of many mucus substances. They coat epithelia in contact with the external environment and protect them against numerous aggressions of endogenous or exogenous origin. At least 21 genes encoding mucins have been identified to date in humans.

[0054] In the model according to the present invention, the expression of the MUC2 mucin (whose amino acid sequence is accessible via reference Q02817 in the Uniprot database) is reduced by at least 30%, preferably by at least 40%, even more preferably by at least 50% compared to a control non-inflamed intestinal epithelium model.

[0055] Similarly, in the model according to the present invention, the expression of the MUC5AC mucin (whose amino acid sequence is accessible via reference P98088 in the Uniprot database) is increased by at least 50%, preferably by at least 75%, even more preferably by at least 100% compared to a control non-inflamed intestinal epithelium model.

[0056] The term "control non-inflamed intestinal epithelium model" refers to any intestinal barrier model that replicates a healthy intestinal barrier. In the context of the present invention, said control model corresponds to the triculture described above, but in which the concentrations of inflammatory cytokines are below the thresholds specified in the model according to the invention. Thus, in the control non-inflamed intestinal epithelium model, the cells in the first compartment can produce IL-6, but at a concentration below 100 pg / mL, and / or IL-8, but at a concentration below 150 pg / mL. The cells in the second compartment can produce TNF-α, but at a concentration below 40 pg / mL, and / or IL-1P, but at a concentration below 90 pg / mL.Such a model is typically obtained by the same process as described below, but in which the cell cultures have not been brought into contact with DSS, TNBS, DNBS, LPS, TNF-α, IFN-γ and / or IL-1p.

[0057] In another aspect, the present invention also relates to the method for obtaining the inflamed epithelium model as described above. Such a method typically comprises the following steps: - the culture, in a medium comprising two compartments separated by a semi-permeable membrane, of Caco-2 cells differentiated into enterocytes and HT29-MTX cells differentiated into goblet cells in the first compartment, and of THP-1 monocytic cells differentiated into macrophages in the second compartment; and - bringing the differentiated Caco-2 and HT29-MTX cells contained in the first compartment into contact with at least one compound chosen from the group consisting of sodium dextran sulfate (DSS), trinitrobenzenesulfonic acid (TNBS) and dinitrobenzenesulfonic acid (DNBS); Then - bringing differentiated THP-1 monocyte cells contained in the second compartment into contact with a compound chosen from the group consisting of lipopolysaccharides (LPS), TNF-α, interferon-γ (IFN-γ), and IL-1P; or - the replacement of said at least one compound chosen from DSS, TNBS and DNBS with LPS, TNF-α, IFN-γ and / or IL-1 p.

[0058] The definitions and conditions mentioned above for the model itself also apply to the process for obtaining it.

[0059] Therefore, the culture media used in step a) are the same as those mentioned above. Similarly, the support used advantageously consists of cell culture inserts such as the Transwell® system. The first compartment is the upper compartment and corresponds to the apical pole of the intestinal epithelium. The second compartment is the lower compartment and corresponds to the basolateral pole of the intestinal epithelium. The semi-permeable membrane is also as defined above.

[0060] DSS, or dextran sulfate sodium, is a sulfated polysaccharide used in vivo to induce ulcerative colitis in mouse models. It can be readily obtained from specialist suppliers such as Sigma-Aldrich.

[0061] TNBS, or trinitrobenzenesulfonic acid, and DNBS, or dinitrobenzenesulfonic acid, are both acids used to induce models of ulcerative colitis and intestinal inflammation in mice. They can be readily obtained from specialist suppliers such as Sigma-Aldrich.

[0062] Lipopolysaccharides (LPS), also known as lipoglycans or endotoxins, are the major component of the outer membrane of Gram-negative bacteria. LPS is a molecule comprised of three parts: a lipid A, an oligosaccharide core, and an O antigen. LPS are commonly used to induce inflammation in cultured cells or organs. They can be readily obtained from specialized suppliers such as Sigma-Aldrich.

[0063] Interferon-γ, or IFN-γ, is a well-known cytokine mediator of innate immunity. The protein sequence of interferon-γ is available under reference P01579 in the UniProt database.

[0064] Therefore, there are two different ways to obtain the model according to the present invention.

[0065] According to a first embodiment, the process includes bringing differentiated Caco-2 and HT29-MTX cells contained in the first compartment into contact with at least one compound selected from the group consisting of DSS, TNBS and DNBS; and bringing differentiated THP-1 monocyte cells contained in the second compartment into contact with a compound selected from the group consisting of LPS, TNF-α, IFN-γ and IL-10.

[0066] According to this embodiment, the process therefore comprises the following steps: i. the culture, in a support comprising two compartments separated by a semi-permeable membrane, of Caco-2 cells differentiated into enterocytes and of HT29-MTX cells differentiated into goblet cells in the first compartment, and of THP-1 monocytic cells differentiated into macrophages in the second compartment; ii. contacting differentiated Caco-2 and HT29-MTX cells contained in the first compartment with at least one compound selected from the group consisting of DSS, TNBS and DNBS; iii. contacting differentiated THP-1 monocyte cells contained in the second compartment with a compound selected from the group consisting of LPS, TNF-α, IFN-γ and IL-1p.

[0067] This embodiment makes it possible to recreate the acute phase and therefore allows, for example, testing the effect of compounds on a state that is already inflamed.

[0068] According to one embodiment, the co-culture of Caco-2 and HT29-MTX cells contained in the first compartment can advantageously be contacted with DSS, TNBS, or DNBS before the culture of THP-1 monocytic cells differentiated into macrophages contained in the second compartment has been contacted with LPS, TNF-α, IFN-γ, and / or IL-1β. In other words, step ii above is performed before step iii. According to this embodiment, the co-culture of Caco-2 and HT29-MTX cells has therefore been pretreated with DSS.

[0069] Alternatively, the co-culture of Caco-2 and HT29-MTX cells contained in the first compartment can be contacted with DSS, TNBS or DNBS at the same time as the culture of THP-1 monocytic cells differentiated into macrophages contained in the second compartment is contacted with LPS, TNF-α, IFN-γ and / or IL-1β. In this context, step ii. above is carried out at the same time as step iii.

[0070] According to another embodiment, the process according to the present invention comprises bringing the differentiated Caco-2 and HT29-MTX cells contained in the first compartment into contact with at least one compound selected from the group consisting of DSS, TNBS and DNBS and then replacing this compound with LPS, TNF-a, IFN-y and / or IL-1 p.

[0071] According to this embodiment, the process therefore comprises the following steps: i'. the culture, in a medium comprising two compartments separated by a semi-permeable membrane, of Caco-2 cells differentiated into enterocytes and HT29-MTX cells differentiated into goblet cells in the first compartment, and of THP-1 monocytic cells differentiated into macrophages in the second compartment; ii'. the contacting of the differentiated Caco-2 and HT29-MTX cells contained in the first compartment with at least one compound selected from the group consisting of DSS, TNBS and DNBS; iii'. the replacement of said at least one compound with LPS, TNF-α, IFN-γ and / or IL-1p.

[0072] This embodiment allows the stages of ignition to be recreated and therefore allows, for example, testing the preventive effects of compounds in the establishment of an inflamed state.

[0073] By replacement we mean here that the culture medium of the cells in which they had been in contact with DSS, TNBS or DNBS is changed by a new culture medium into which LPS, TNF-a, IFN-y and / or IL-10 are introduced. Preferably, DSS, TNBS or DNBS is therefore not eliminated by rinsing, but simply by replacement of the medium.

[0074] In the context of the present invention, regardless of the method used, DSS, TNBS and / or DNBS are brought into contact with the co-culture of differentiated Caco-2 and HT29-MTX cells at concentrations ranging from 1 to 10% by weight per unit volume (w / v), preferably at a concentration of 5% w / v. The co-culture was typically incubated with DSS, TNBS and / or DNBS for a period of 6 to 72 hours, preferably for a period of approximately 12 hours.

[0075] In the context of the present invention, LPS, TNF-α, IFN-γ, and / or IL-10 are brought into contact with a culture of differentiated THP-1 monocyte cells or with a co-culture of differentiated Caco-2 and HT29-MTX cells at concentrations ranging from 0.1 to 1000 ng / mL, preferably at a concentration of 100 ng / mL. The culture of differentiated THP-1 monocyte cells, or, depending on the embodiment, the co-culture of differentiated Caco-2 and HT29-MTX cells, is typically incubated with LPS, TNF-α, IFN-γ, and / or IL-10 for 12 to 72 hours, preferably for approximately 24 hours. Once this incubation is complete, the epithelial model is ready for use.

[0076] A person skilled in the art knows how to obtain the various cell cultures used in the process according to the present invention. Typically, these cell cultures are obtained as follows: - On day 1 (J1), Caco-2 cells are seeded at a concentration ranging from 5000 to 100,000 cells / cm³ 2 , preferably of 15000 cells / cm 2 , in the first compartment (apical compartment); - Between days 1 and 7, preferably on day 5, HT29-MTX cells are inoculated at a concentration ranging from 5000 to 100,000 cells / cm² 2 , preferably of 30,000 cells / cm 2 , in the first compartment; - Between days 18 and 30, preferably on day 18, seed THP1 monocytic cells at a concentration ranging from 1000 to 1,000,000 cells / cm³ 2 , preferably 800,000 cells / cm² 2 , in the second compartment (basolateral compartment) in the presence of phorbol-12-myristate-13-acetate (PMA) at a concentration ranging from 100 to 300, preferably 200 ng / mL, to obtain a population completely differentiated into macrophages; - incubation of the different cultures until complete differentiation of Caco-2 into enterocytes is obtained (21 days of culture are generally required); - bringing the two compartments into contact through the semi-permeable membrane to carry out the process according to the invention.

[0077] Finally, the present invention also relates to the use of the model according to the invention for screening candidate compounds for the treatment of chronic inflammatory bowel diseases (IBD) or acute inflammatory colitis. IBD includes Crohn's disease (CD) and ulcerative colitis (UC), which are characterized by inflammation of the lining of a portion of the digestive tract due to dysregulation of the immune system. The inflamed intestinal epithelium model according to the present invention accurately represents the state of the intestinal epithelium observed in IBD. It also allows observation of the effects obtained by applying a test compound, or "candidate" compound, to an IBD intestinal epithelium.It is therefore possible to see if the said compound has beneficial, anti-inflammatory effects in the context of IBD and intestinal inflammation in general, and thus to identify compounds with potential therapeutic action in the context of intestinal inflammation.

[0078] Also, according to another aspect, the present invention relates to a method for selecting candidate compounds for the treatment of inflammatory bowel diseases, said method comprising the steps of: i) bringing a candidate compound into contact with the inflamed intestinal epithelium model according to the present invention; and ii) selecting the compound if it makes it possible to obtain, in said model, at least one of the following effects a) to e): a) A decrease of at least 10%, preferably at least 20%, even more preferably at least 50% of the level of IL-6 and / or IL-8 in the first compartment; b) A decrease of at least 10%, preferably at least 20%, even more preferably at least 50% of the level of TNF-α and / or IL-10 in the second compartment; c) An increase in MUC2 expression of at least 5%, preferably of at least 20%, even more preferably of at least 40%;d) A decrease in MUC5AC expression of at least 5%, preferably of at least 20%, even more preferably of at least 40%; e) An increase in transepithelial electrical resistance of at least 10%, preferably of at least 20%.

[0079] The compound being tested, or "candidate" compound, can be of any type. It can be of natural origin or produced by chemical synthesis. It can be a library of structurally defined chemical compounds, uncharacterized compounds or substances, or a mixture of compounds.

[0080] The level of expression and synthesis of pro-inflammatory cytokines, as well as the study of the permeability of the intestinal barrier by measuring transepithelial resistance and mucin expression, allows us to assess the level of inflammation of the epithelium.

[0081] The present invention is presented in more detail in the examples below. Examples Example 1:

[0082] Materials and methods

[0083] Cells and reagents

[0084] Enterocyte-like Caco-2 cells and human monocyte-like THP-1 cells were obtained from the American Type Culture Collection (ATCC). HT29-MTX cells differentiated into goblet cells were kindly provided by Dr. Thécla Lesuffleur of INSERM UMR S 938 (Paris, France).

[0085] All cell culture reagents and betamethasone were purchased from Fisher Scientific (Illkirch, France). Sodium dextran sulfate (DSS) and lipopolysaccharides (LPS) were obtained from Sigma-Aldrich (Saint Quentin-Fallavier, France). The SuperMApo was kindly provided by Dr. Sylvain Perruche of Med'Inn'Pharma in Besançon, France.

[0086] Cell culture methods

[0087] Co-culture Caco-2 HT29-MTX

[0088] The culture medium for Caco-2 and HT29-MTX cells consisted of Dulbecco's Modified Eagle Medium Gluta-Max™ containing 15% inactivated fetal bovine serum (FBS), 1% non-essential amino acids, and 1% antibiotics (100 pg / ml streptomycin and 100 IU / ml penicillin). The cells were maintained in an incubator (Binder, Tuttlingen, Germany) at 37°C and 5% CO2 in a water-saturated atmosphere.

[0089] After thawing, the Caco-2 and HT29-MTX cells were seeded separately in 25 cm³ flasks. 2 After three passages using 0.25% Trypsin-EDTA, Caco-2 and HT29-MTX cells were seeded at densities of 5,000 and 10,000 cells per insert, respectively, in 0.33 cm² Transwell® 24-well permeable support plates 2with a polycarbonate membrane (0.4 pm pores, Corning Costar, Cambridge). After 21 days of incubation, the co-culture is fully differentiated.

[0090] THP-1 cell culture

[0091] The culture medium for the THP-1 cell line consisted of Roswell Park Memorial Institute medium containing 10% inactivated FBS and 1% antibiotics (100 pg / mL streptomycin and 100 IU / mL penicillin). The cells were maintained in an incubator (Binder, Tuttlingen, Germany) at 37°C and 5% CO2 in a water-saturated atmosphere.

[0092] After 6 passes through 75 cm bottles 2 The cells were then seeded into 24-well culture plates at a density of 1.5 x 10 6 cells per well. Monocytes were differentiated into macrophage-like cells after 3 days of incubation with 100 ng / mL of phorbol 12-myristate 13-acetate (PMA).

[0093] Tri-culture Caco-2, HT29-MTX, THP-1

[0094] The insert containing the Caco-2 / HT29-MTX co-culture was transferred, 21 days after inoculation, into a 24-well plate containing differentiated THP-1 cells (after 3 days of incubation in PMA). The tri-culture model combining differentiated Caco-2 / HT29-MTX cells (located on the insert membrane) and THP-1 cells (adhering to the well surfaces) was ready for use.

[0095] Inflamed tri-culture model

[0096] Stimulation of differentiated THP-1 and Caco-2 / HT29-MTX cells

[0097] For tri-culture characterization, THP-1 and Caco-2 / HT29-MTX cells were stimulated with 0, 1, 1, 10, 100, and 1000 ng / mL of LPS solution (24, 48, and 72 h) and 1, 3, 5, 7, and 10% w / v of DSS (24, 48, and 72 h), respectively. For kinetic studies, Caco-2 / HT29-MTX cells were first stimulated with 5% w / v of DSS (6, 12, and 24 h) and then, after removal of the medium, with 100 ng / mL of LPS solution (24 h).

[0098] After stimulation, all supernatants were collected and stored at -80 °C for cytokine assay. THP-1 cells and co-cultured cells (Caco-2 and HT29-MTX cells) were harvested separately and stored in RLT lysis buffer at -80 °C for total RNA isolation, reverse transcription (RT), and quantitative polymerase chain reaction (qPCR) analysis.

[0099] Anti-inflammatory treatments for inflamed tri-culture

[0100] 100 pM of betamethasone (BTM) in culture medium or pro-resolving factors produced by macrophages after efferocytosis (SuperMApo) were added to the apical compartment for 24h after induction of inflammation, as described above.

[0101] After stimulation and treatment, all supernatants were collected and stored at -80°C for cytokine assays. THP-1 cells and co-cultured cells (Caco-2 and HT29-MTX cells) were harvested separately and stored in RLT lysis buffer at -80°C for total RNA isolation, RT, and qPCR analysis.

[0102] Cell viability test

[0103] The metabolic activity of intestinal epithelial cells (IECs) was tested using the MTS assay. After removal of the supernatant, the cells were incubated with 200 pL of MTS solution. The formazan content was quantified by spectrophotometry at 450 nm using the FLUOstar Omega spectrophotometer (BMG Labtech, Champigny-sur-Marne, France). All reactions were performed in triplicate.

[0104] Permeability study

[0105] The integrity of the co-culture monolayer was controlled by measuring transepithelial electrical resistance (TEER) and the penetration of lucifer yellow (LJ) through the IECs.

[0106] The TEER was evaluated using the Millicell electrical resistance system (Millipore, USA). Measurements were taken in triplicate, and the values ​​were normalized by subtracting the resistance value of an empty insert (membrane) and expressed as follows:

[0107] [Math. 1]: [ 1 0108] J %TEER = 100 x ( TEERf ~ TEERe ) (TEERi— TEERe)

[0109] TEERe is the resistance value of a cellless insert, TEERf is the final resistance value and TEERi is the initial resistance value.

[0110] Paracellular permeability was assessed by measuring the transport of lucifer yellow (LJ) from the apical to basolateral compartments. LJ concentrations were determined by fluorescence (e.g., 405 nm and em., 535 nm) using the FLUOstar Omega spectrophotometer (BMG Labtech, Champigny-sur-Marne, France). Triplicate measurements were performed, and the permeability coefficient Pa PP (cm / s) was estimated using the following formula:

[0111] [Math. 2]: dQ

[0112] Papp =

[0113] dQ / dt is the flux, where Q is the quantity of molecules transported during the incubation time t, and A is the surface area of ​​the polycarbonate membrane (0.33 cm²). 2 ), and Co is the initial concentration in the donor compartment.

[0114] Cytokine levels measured by enzyme-linked immunosorbent assay

[0115] The concentrations of various pro-inflammatory cytokines (interleukin (IL)-ip, IL-6, IL-8, and tumor necrosis factor (TNF)-α) were assessed in apical and basolateral supernatants. The IL-10 Human ELISA Kit (Life Technologies-ThermoFisher), the IL-6 Human ELISA Kit (Diaclone), the IL-8 Human ELISA Kit (Life Technologies-ThermoFisher), and the TNF-α Human ELISA Kit (Diaclone) were used according to the manufacturer's instructions. All assays were performed in triplicate.

[0116] Quantification of protein expression by real-time PCR

[0117] Total RNA was isolated from cultured cells using the RNeasy Mini Kit with DNase I column treatment (Qiagen) according to the manufacturer's instructions, and quantified using the Nanodrop 2000 (Thermo Fisher Scientific, Waltham, USA).

[0118] cDNA was synthesized by reverse transcription of equal amounts of total RNA (2 pg) from each sample using a High-Capacity RNA-to-cDNA kit (Applied Biosystems) and measured by qRT-PCR using the TaqMan Universal Master Mis II kit (Applied Biosystems), according to the manufacturer's protocol.

[0119] The gene expression levels of each sample were quantified using the comparative method 2 -AACt and normalized with respect to endogenous P-actin gene expression. The reference mRNA level was designated as 1 and the relative levels of the transcripts of Genes in the samples were expressed as fold changes. All quantifications were performed in triplicate.

[0120] Statistical analysis

[0121] Statistical analysis was performed using Graphpad Prism® 8 software. Data are expressed as mean ± standard deviation (SD). For statistical significance analysis, rank ANOVA was applied, followed by Dunn's test for all pairwise comparisons. In all cases, P < 0.05 was considered significant.

[0122] Results

[0123] Effects of DSS and LPS on CEI and THP-1

[0124] The metabolic activity of CEIs and THP-1 exposed to different concentrations of DSS (1 to 10% mA / L) or LPS (0.1 to 1000 ng / mL) after 24 hours of incubation was studied. Cell viability of CEIs remained stable at 100% and decreased at concentrations above 5% mA / V (81.95% and 68.21% for 7% and 10% mA / V DSS, respectively). Below 7% mA / V DSS, the metabolic activity of THP-1 was not affected. Below 1000 ng / mL, LPS did not alter the viability of CEIs or THP-1.

[0125] The integrity of the IECs was monitored by the TEER and was not influenced by the tested LPS concentration. Furthermore, the TEER was reduced to 68% at a DSS concentration of 1% w / v and decreased to a minimum of 19% at a DSS concentration of 10% w / v. These results were confirmed by evaluating the apparent permeability of Lucifer Yellow (Papp) under the same conditions to determine the paracellular transport of IECs. After 24 h of exposure, the Papp increased fourfold (2.22 x 10⁻⁶). -6 cm / s) with 5% m / V of DSS relative to the control (0.51 x 10 -6 cm / s) or LPS at 100 ng / mL (0.57 x 10 -6 cm / s).

[0126] The ability of DSS 5% w / v to induce the secretion of pro-inflammatory cytokines by IECs and THP-1 cells was studied separately. Slight secretion of IL-6 and IL-8 by IECs and of TNF-α and IL-1P by THP-1 cells was induced after 24 hours of direct exposure of these cells to DSS. The secretion of these inflammatory cytokines was not altered when IECs were combined with THP-1 cells in the tri-culture condition and DSS was applied apically (TNF-α: 41.09 pg / mL; IL-1P: 101.02 pg / mL; IL-6: 84.91 pg / mL; and IL-8: 284.26 pg / mL). However, if the DSS is deposited on the basolateral side, the secretion of IL-6 and IL-8 is slightly but significantly reduced (IL-6: 32.24 pg / mL; IL-8: 197.62 pg / mL).

[0127] After 24 hours of incubation with LPS, the secretions of inflammatory cytokines were completely different. When LPS was in direct contact with THP-1, all cytokines reached their highest concentrations, 5 to 7 times greater than those obtained with apical / basolateral DSS and apical LPS (TNF-α: 699.12 pg / mL; IL-10: 199.29 pg / mL; IL-6: 459.00 pg / mL; and IL-8: 861.86 pg / mL). When LPS was added apically, the cytokine concentration was similar to that obtained with DSS apically.

[0128] Permeability and secretion of inflammatory cytokines in a tri-culture model after exposure to a combination of DSS and LPS.

[0129] To achieve increased epithelial barrier permeability and significant release of pro-inflammatory cytokines, co-stimulation with DSS and LPS was investigated. DSS was fixed at 5% w / v to induce increased inflammatory cell permeability, and LPS at 100 ng / mL to induce THP-1-mediated secretion of inflammatory cytokines without any change in cell viability.

[0130] The TEER value obtained with DSS was not modified by LPS deposition (37.84% and 31.54% for DSS(A) + LPS(A) and DSS(A) + LPS(B), respectively, versus 33.84% for DSS(A); 97.20% and 99.91% for DSS(B) + LPS(A) and DSS(B) + LPS(B), respectively, versus 96.99% for DSS(B)). While apical DSS is necessary to increase IEC permeability, the TEER is independent of the LPS deposition location in this condition. These results are consistent with those obtained with the apparent JL permeability. Indeed, this parameter is also independent of LPS location but is significantly increased by apical DSS (2.21 and 2.00 × 10, respectively). -6 cm / s for induction by DSS (A) + LPS (A) and DSS (A) + LPS (B) vs 0.51*10 -6 cm / s compared to the unstimulated control, p<0.05).

[0131] Concentrations of inflammatory cytokines secreted by THP-1 cells (TNF-α and IL-1P) and by IECs (IL-6 and IL-8) increased under all conditions. When LPS was deposited on the apical surface with or without DSS, the expression and secretion of these cytokines were slightly increased. On the other hand, LPS on the basolateral side induced high secretions, and a maximum was reached for both apical DSS and basolateral LPS [Fig. 1] (for induction by DSS (A) + LPS (B) and DSS (B) + LPS (B): respectively 802.12 and 789.25 pg / mL vs 10.17 pg / mL for TNF-a, p<0.05; respectively 301.82 and 291.99 pg / mL vs 26.31 pg / mL for IL-1 P, p<0.05; respectively 684.91 and 500.49 pg / mL vs 31.02 pg / mL for IL-6, p<0.05; respectively 1084.26 and 878.09 pg / mL vs 90.23 pg / mL for IL-8, p<0.05).

[0132] The effect of the apical DSS / basolateral LPS combination on inflammatory bowel disease (IBD) was investigated. First, apical and basolateral LPS did not induce any change in mucin expression (for induction by LPS (A) and LPS (B): 0.89 and 0.91-fold for MUC2, respectively; 0.94 and 1.12-fold for MUC5AC, respectively) compared to the control. The DSS-A / LPS-B combination induced MUC2 underexpression and MUC5AC overexpression (MUC2: 0.44-fold, p<0.05; MUC5AC: 2.01-fold, p<0.05) [Fig. 2]. Tight junction protein expression was also assessed, and ZO-1 and Claudin-1 were underexpressed in the DSS-A / LPS-B condition compared to the other conditions (ZO-1: 0.10-fold, p<0.05; Claudin-1: 0.21-fold, p<0.05) [Fig. 2]. Occludin was also underexpressed, but not significantly so compared to the control (occludin: 0.96-fold).

[0133] Permeability and secretion of inflammatory cytokines in a tri-culture model after exposure to a DSS pretreatment

[0134] To reproduce the kinetics of acute inflammation characterized by an increase in the permeability of the epithelial layer allowing the passage of bacterial endotoxins, DSS and LPS were applied to the apical side of the tri-culture, and different exposure sequences were explored.

[0135] DSS at 5% w / v led to a significant reduction in TEER after 6 h (79.21% TEER) and to a minimum after 12 h of incubation (46.15% at 12 h and 38.64% at 24 h TEER). Removal of DSS and replacement with a culture medium containing LPS for 24 h did not alter these TEER values ​​(for DSS-6, 12, or 24 h (A) + LPS-24h (A): 72.81%, 42.84%, and 33.93% TEER, respectively). The change in TEER values ​​appears to depend on the duration of DSS exposure and is not influenced by LPS.

[0136] After 6 h of DSS incubation or 24 h of DSS / LPS co-incubation, a very slight increase in inflammatory cytokines secreted by macrophages was observed compared to the reference (without DSS / LPS). The secretion of IL-1P and TNF-α was increased by the duration of DSS exposure (for 12 h and 24 h: 549.01 and 773.92 pg / mL respectively for TNF-α; 261.69 and 291.21 pg / mL respectively for IL-10) to reach the same level of expression as the positive control DSS (A) + LPS (B) - 24 h (802.12 pg / mL for TNF-α; 301.82 pg / mL for IL-1P).

[0137] A 12-hour pre-incubation with DSS induced the secretion of inflammatory mediators (IL-6 and IL-8) by the CEIs (669.99 pg / mL for IL-6; 1100.01 pg / mL for IL-8) at the same level as the positive control. For a shorter exposure time, cytokine secretion by the CEIs was consistent with that of DSS(A) + LPS(A) - 24 h (289.12 and 111.03 pg / mL for IL-6, respectively; 420.02 and 163.03 pg / mL for IL-8, respectively).

[0138] Effect of anti-inflammatory treatments on a tri-culture model after exposure to DSS and LPS

[0139] The glucocorticoid BTM was chosen as the reference anti-inflammatory drug for its well-known inhibitory effects on cytokine synthesis. BTM 100 pM significantly reduced cytokine release after 24 h from both activated macrophages (TNF-α: 314.52 pg / mL vs 802.12 pg / mL; IL-10: 160.90 pg / mL vs 301.82 pg / mL, p<0.05 vs without anti-inflammatory treatment) and EICs (IL-6: 200.20 pg / mL vs 684.91 pg / mL; IL-8: 657.91 pg / mL vs 1084.26 pg / mL, p<0.05 vs without anti-inflammatory treatment) [Fig. 3]. This anti-inflammatory effect was also observed after exposure to DSS and LPS for 48 hours or after pre-stimulation with DSS for 12 hours and exposure to LPS for 24 hours.

[0140] The TEER values ​​(after DSS / LPS 24 or 36 h exposure: 73.79% and 69.98% TEER vs 36.72% TEER, respectively, p<0.05) and JL Papp values ​​(after DSS / LPS 24 or 36 h exposure: 1.01*10, respectively) -6and 0.92*10 -6 vs 2.23*10 -6 cm / s, p<0.05) were significantly altered with BTM compared to the untreated reference. The effect of 100 mM BTM on IEC permeability is consistent with mucin expression and tight junction protein expression (Claudin-1; ZO-1) by IECs [Fig. 4]. BTM treatment induces a barrier stabilization effect after stimulation with DSS and LPS for 24 h, 36 h, or 48 h.

[0141] The secretome, called SuperMApo, was composed of high molecular weight molecules called pro-resolvent factors produced by macrophages after the induction of efferocytosis (Bonnefoy F, Gauthier T, Vallion R, et al. Front. Immunol. 2018;9:2586). These pro-resolvent factors exert therapeutic effects in experimental colitis and were chosen here to assess drug permeability in this model. Treatment with SuperMApo after DSS / LPS stimulation for 24 ha decreased the secretion of IL-6 and IL-8 (IL-6: 222.17 pg / mL vs 684.91 pg / mL; IL-8: 581.16 pg / mL vs 1084.26 pg / mL, p<0.05) and of TNF-a and IL-1p (TNF-a: 302.99 pg / mL vs 802.12 pg / mL; IL-1: 149.57 pg / mL vs 301.82 pg / mL, p<0.05) compared to the reference. As with BTM, these effects were also observed after exposure to DSS and LPS for 48 hours or after pre-stimulation with DSS for 12 hours and exposure to LPS for 24 hours.Furthermore, this biological treatment restored the normal percentage of TEER (after DSS / LPS 24 or 36 h of exposure: respectively 77.33% and 63.23% TEER) and the transport of JL (after DSS / LPS 24 or 36 h of exposure: respectively 0.77*10. -6 and 0.96*10 -6 cm / s). These results are consistent with the change in the relative expressions of mucins and tight junction proteins.

[0142] Discussion

[0143] The intestinal barrier of patients with IBD, a physical barrier formed by the epithelium, mucus, and underlying host immune system, is impaired. Altered intestinal barrier function has been implicated in the pathogenesis of IBD, as demonstrated by the DSS colitis model (Ramos GP, Papadakis KA. Mayo Clin. Proc. 2019;94:155-165). The development of an in vitro model mimicking the permeability of inflamed epithelium and cytokine secretion observed in humans and animals will enhance the relevance of in vitro IBD models. Altered intestinal permeability can be induced directly on intestinal endothelial cells (IECs) by a chemical agent (Wirtz S, Neufert C, Weigmann B, et al. Nat. Protoc. 2007;2:541-546) or indirectly via an inflammatory secretome stimulating TNF receptor 2 signaling on IECs (Su L, Nalle SC, Shen L, et al. Gastroenterology 2013;145:407-415). LPS did not show any direct or indirect effects on permeability, probably due to the low level of TNF receptor 2 expression in Caco-2 cells (Vamadevan AS, Fukata M, Arnold ET, et al. Innate Immun. 2010;16:93-103). Unlike LPS, DSS in the 1–5% range induced an increase in IEC permeability by reducing TEER and increasing Papp without any alteration to barrier integrity. This effect could be a direct physical action on the membrane.Chassaing et al. hypothesize that the interaction of DSS with CEIs could be explained by the negative charge of DSS, conferred by sulfate groups, which damages negatively charged cell membranes through repulsive forces and increases the permeability of the colonic epithelium (Chassaing B, Aitken JD, Malleshappa M, et al. Curr. Protoc. Immunol. 2014; 104: 15.25.1–15.25.14). This physical interaction might be suitable for in vitro conditions, but remains uncertain in vivo when DSS is diluted in stool and at concentrations often lower than those in our study. The effect of DSS on permeability could also be attributed to an indirect effect via a complex formed between DSS and fatty acids. medium chain fatty acids (MCFAs). Laroui et al. suggest that this complex, which is metabolized by intestinal inflammatory response cells (IICs), triggers intestinal inflammatory signaling cascades. In this in vitro model, medium-chain fatty acids contained in the cellular medium's SVF could generate this complex and contribute to reducing the functions of the Caco-2 / HT29-MTX cell monolayer (Laroui H, Ingersoll SA, Liu HC, et al. PLoS One 2012;7:e32084). These results correlate with the low expression obtained after DSS exposure on IICs of genes encoding tight junction (TJ) proteins (ZO-7, Claudin-1, Occludin), which play an important role in maintaining intestinal permeability and regulating paracellular permeability (Suzuki T. Cell. Mol. Life Sci. CMLS 2013;70:631-659). Deregulation of genes encoding components of JS (Bischoff SC, Barbara G, Buurman W, et al. BMC Gastroenterol.2014; 14:189) could lead to long-term alteration of JS structure, as has been observed in murine models and human IBD (Poritz LS, Garver Kl, Green C, et al. J. Surg. Res. 2007;140:12-19; ​​Capaldo CT, Powell DN, Kalman DJ Mol. Med. Berl. Ger. 2017;95:927-934; Shaoul R, Okada Y, Cutz E, et al. J. Pediatr. Gastroenterol. Nutr. 2004;38:488-493), and could therefore be responsible for the concentration- and time-dependent increase in CEI monolayer permeability in this tri-culture model. The effect of DSS on intestinal cavities (ICs) induces a modification of intestinal barrier permeability, which may also be regulated by mucins (Bischoff SC, Barbara G, Buurman W, et al. BMC Gastroenterol. 2014;14:189), as demonstrated in IBD (Capaldo CT, Powell DN, Kalman DJ. Mol. Med. Berl. Ger. 2017;95:927-934). The expression of genes encoding mucins (MUC2 and MUC5AC) was dysregulated in the triculture, as observed in IBD patients or in the in vivo colitis model, following direct DSS exposure of ICDs. An increase in MUC5AC expression (Shaoul R, Okada Y, Cutz E, et al. J. Pediatr. Gastroenterol. Nutr. 2004;38:488-493; Olli KE, Rapp C, O'Connell L, et al. Inflamm. Bowel Dis. 2020;26:1353-1367) and a decrease in MUC2 (Van der Sluis M, De Koning BAE, De Bruijn ACJM, et al. Gastroenterology 2006;131:117-129;Gut Microbes 2010;1:51-54) were observed, and these changes in expression levels were reversible after treatment with BTM. In the long term, these changes could induce alterations in mucus composition, similar to what is observed in IBD (McGuckin MA, Eri R, Simms LA, et al. Inflamm. Bowel Dis. 2009;15:100-113).

[0144] Unlike DSS, LPS did not alter permeability but induced the secretion of inflammatory cytokines at higher levels than DSS. The mechanism by which LPS induces the release of inflammatory molecules, such as TNF-α, IL-1P, IL-6, and IL-8, is well known (Bryant CE, Spring DR, Gangloff M, et al. Nat. Rev. Microbiol. 2010;8:8-14). TLR4, expressed on the cell membrane, is associated with the MD-2 coreceptor via its extracellular fragment. This combination is essential for LPS recognition and binding. The interaction between LPS and its recognition site causes dimerization of the receptor complex, leading to the recruitment of adaptor proteins to the intracellular domain of TLR4 and the activation of an intracellular cascade that results in the translocation of the factor. transcription of nuclear factor (NF)-KB in the nucleus and to the production of inflammatory cytokines (Ciesielska A, Matyjek M, Kwiatkowska K. Cell. Mol. Life Sci. CMLS 2021;78:1233-1261).

[0145] Incubation of LPS with IECs alone or in triculture induced inflammatory cytokines at relatively low levels, in the same range as in vitro models with Caco-2 cells or the Caco-2 and HT29-MTX coculture described in the literature. This hyporesponsiveness of IECs to LPS stimulation is likely due to the limited expression of the MD-2 coreceptor (Vamadevan AS, Fukata M, Arnold ET, et al. Innate Immun. 2010;16:93-103), which is involved in LPS recognition and essential for NF-κB activation and the subsequent secretion of inflammatory cytokines (Shimazu R, Akashi S, Ogata H, et al. J. Exp. Med. 1999;189:1777-1782). To increase these secretions, immune cells can be associated with IECs to enhance cell-to-cell crosstalk. Indeed, the effect of LPS in contact with macrophages generated a significantly greater secretion of cytokines (IL-6 and IL-8) by the CEI.The sensitizing effect on intestinal cell function was probably mediated by soluble mediators. TNF-α is one of the mediators involved in the initiation of the immune response by intestinal endothelial cells and is responsible for increasing levels of pro-inflammatory cytokines (Haller D, Bode C, Hammes WP, et al. Gut 2000;47:79-87).

[0146] Direct exposure to DSS slightly increased the expression and synthesis of the inflammatory cytokines TNF-α and IL-10 by THP-1, as well as IL-6 and IL-8 by CIE or in triculture. This slight increase in gene expression of these cytokines could suggest an intracellular action of DSS and not just disruption of the intestinal barrier. The intracellular mechanism of action of DSS remains unclear, but the production of inflammatory cytokines by CIE and THP-1 could be explained by the activation of the Nlrp3 inflammasome by endocytosed DSS, which induces the cleavage of cytokines into their biologically active forms and their subsequent secretion, thereby triggering and sustaining an inflammatory cascade (Bauer C, Duewell P, Lehr HA, et al. Dig. Dis. Basel Switz. 2012;30 Suppl 1:82-90). However, unlike LPS, DSS did not produce a sufficient amount of cytokines to induce sufficient inflammation on IECs, particularly TNF-α.Its concentration is 10 times lower than that of LPS. Thus, DSS at a non-toxic dose had a weak effect on cytokine synthesis by IEs or macrophages and was insufficient to generate inflammation, whereas LPS did not alter the permeability of the co-culture. To reproduce the characteristics of the acute phase of an in vivo inflammatory model in patients, a DSS / LPS combination would be necessary. Combining basolateral LPS incubation with apical DSS incubation proved useful in achieving both altered epithelial barrier permeability and the release of pro-inflammatory cytokines into the cell supernatants. Similarly, under these conditions, altered mucin and JS protein expression was observed.This tri-culture model could mimic an acute inflammatory phase of IBD which can be used to test treatments for the purpose of evaluating curative action and is called a curative model (CU).

[0147] DSS was applied, under different conditions, to the apical side to study the increase in CEI layer permeability during inflammation induction. LPS was deposited for 24 h to induce maximum inflammatory cytokine production as described (Shimba A, Ikuta K. 2020;42:669-680). DSS induced membrane permeability and allowed JS opening only after 6 h. Under these conditions, the secretion of inflammatory cytokines by CEIs and THP-1 cells was not of the same order as that obtained in the CU model. The amount of LPS diffused during 24 h was much lower than that which stimulates THP-1 cells in the CU model. This low cytokine synthesis could be explained by a low concentration of LPS in contact with THP-1 cells and therefore probably by insufficient monolayer permeability to allow a sufficient rate of LPS diffusion.A longer incubation period (6–24 h) with DSS was necessary to sufficiently increase the permeability of the co-culture and allow LPS to diffuse from the apical to the basolateral compartment in sufficient quantities to induce the secretion of inflammatory cytokines by THP-1 cells after 24 hours. Pretreatment with DSS confirmed that maximum permeability was reached between 6 and 12 h and that the amount of LPS traversed and the duration of contact with macrophages were sufficient to induce an inflammatory secretome, enabling the secretion of inflammatory mediators (IL-6 and IL-8) by intestinal endothelial cells (IECs). This tri-culture intestinal model, called the preventive model (PR), could also be used to evaluate the ability of active molecules to diffuse through IECs, restore intestinal permeability, and / or reduce the inflammatory secretome.

[0148] Thus, we have developed two in vitro models, the CU model to mimic an acute inflammatory model, and the PR model to break down the different phases of inflammation, with an increase in permeability followed by cytokine secretion. Since barrier dysfunction is a significant factor in the pathogenesis of IBD (Turner JR. Nat. Rev. Immunol. 2009;9:799-809), drug development and the search for treatments to restore the intestinal barrier (ICB) represent a major challenge. Glucocorticoids are the cornerstone of IBD remission induction and the treatment of collagenous colitis (Baumgart DC, Sandborn WJ. Lancet Lond. Engl. 2007;369:1641-1657) and of restoring increased intestinal permeability in patients with Crohn's disease (Wild GE, Waschke KA, Bitton A, et al. Aliment. Pharmacol. Ther. 2003;18:309-317). The capacity of the active glucocorticoid BTM has been used to evaluate the reversibility of the in vitro model. BTM significantly reduced cytokine release by activated macrophages (TNF-α and IL-10) or CEIs (IL-6 and IL-8), similarly after 24 h incubation (in the CU model) or 36 h (in the PR model).Indeed, glucocorticoids are known to suppress several inflammatory cytokines originating from inflammatory endothelial cells (IECs) and macrophages, including IL-1, IL-6, IL-8, and TNF-α (Shimba A, Ikuta K, Semin. Immunopathol. 2020;42:669-680). These effects are mediated by the binding of glucocorticoids to the cytoplasmic glucocorticoid receptor (Auphan N, DiDonato JA, Rosette C, et al. Science 1995;270:286-290; Ramamoorthy S, Cidlowski JA. Rheum. Dis. Clin. North Am. 2016;42:15-31, vii; Jonat C, Rahmsdorf HJ, Park KK, et al. Cell 1990;62:1189-1204). In addition to its anti-inflammatory activity, BTM exerts a stabilizing effect on the IE barrier and restores intestinal barrier permeability in our in vitro model, in correlation with glucocorticoid therapy which restores increased intestinal permeability in CD, by inducing the expression of the inhibitor. cytosolic IKB to hinder intestinal translocation of NFKB to the nucleus and inhibit its activation in vivo and in vitro (Wild GE, Waschke KA, Bitton A, et al. Aliment. Pharmacol. Ther. 2003;18:309-317).

[0149] The anti-inflammatory effect of the biological complex, called SuperMApo, was also studied in this tri-culture. This drug contains all the pro-resolving factors released by efferocyte macrophages, involved in the resolution of inflammation and tissue healing (Bonnefoy F, Gauthier T, Vallion R, et al. Front. Immunol. 2018;9:2586). SuperMApo has been shown to exert a therapeutic effect by stimulating the healing, proliferative, and migratory properties of the main cell types orchestrating the tissue repair process, particularly intestinal endothelial cells (IECs), and by promoting intestinal mucosal healing in a DSS-induced in vivo colitis model (Martin-Rodriguez O, Gauthier T, Bonnefoy F, et al. Front. Immunol. 2021;12:754475).Given the very high molecular weight of the SuperMApo components (including Transforming Growth Factor (GF)-beta ~28 kDa; Vascular Endothelial GF ~40 kDa; and Insulin-Like GF-1 ~8 kDa), they cannot cross an intact intestinal barrier. Without sufficient barrier permeability, the effect of SuperMApo on THP-1 cells was not observed. After apical incubation, the therapeutic effect of SuperMApo was confirmed in PR and CU models after 24 h (CU model) or 36 h (PR model), suggesting that the permeability of the CEI monolayer by DSS incubation was sufficient to allow this biological complex to cross the barrier and exert its effect. These results are consistent with the in vivo properties of SuperMApo (Martin-Rodriguez O, et al (2021)).

[0150] This study suggests the relevance of these two models for the pre-selection of therapeutic agents for the treatment of IBD, by evaluating their anti-inflammatory and pro-healing effects before their use in in vivo models.

[0151] Conclusion

[0152] A novel tri-culture model, characterized by altered mucus layer composition and altered cell junction expression and inflammatory crosstalk between immune and epithelial cells, has been developed to enhance the relevance of in vitro models for IBD treatment development. The anti-inflammatory effect of two well-known anti-inflammatory agents, betamethasone and a cocktail of pro-resolvent factors, was confirmed in this tri-culture model, reducing inflammatory cytokine secretion and partially restoring intestinal monolayer permeability, suggesting its potential for selecting therapeutic agents for IBD treatment. Example 2:

[0153] Materials and methods:

[0154] Permeability study:

[0155] Paracellular permeability was assessed by measuring the transport of lucifer yellow (LJ) at a concentration of 50 pg / mL from the apical (donor) compartment to the basolateral compartment (receptor). The amount of JL transported paracellularly to the recipient side was determined by spectrofluorescence (excitation / emission wavelengths of 405 / 535 nm) using the FLUOstar Omega spectrofluorimeter (BMG Labtech, Champigny-sur-Marne, France). P-glycoprotein (P-gp) efflux was assessed by measuring the transport of 5 pM rhodamine 123 (alone or in the presence of 100 pM verapamil) from the basolateral (donor) compartment to the apical (recipient) compartment. Rhodamine 123 levels in the transport study were also measured by spectrofluorescence (excitation / emission wavelengths of 485 / 538 nm).

[0156] For each transport study (JL permeation through intestinal epithelial cells and P-gp-mediated efflux), cell models were incubated at 37°C with 5% CO2 for 2 h.

[0157] The permeability coefficient Papp (cm / s) of JL and rhodamine 123 was estimated from triplicate using the formula [Math 2]: Papp = (dQ / dt) / (A*C0), where dQ / dt represents the flux, with Q being the quantity of molecules transported over time t of the incubation, and A the surface area of ​​the polycarbonate membrane (0.33 cm²). 2 ) and C0 the initial concentration in the donor compartment.

[0158] Targeting the mucus:

[0159] Intestinal epithelial cells were washed three times with HBSS (Hank's Balanced Salt Solution) and then incubated for 1 h with anti-MUC5AC antibodies coupled to a green fluorescent marker (100 pL at 7.5 pg / mL). After removal of the supernatants, the cells were washed three times with HBSS. The membrane of the Transwell® inserts was removed and sealed with DAKO fluorescent mounting medium. The slides were analyzed using a Zeiss LSM 800 Airy-Scan confocal microscope (DlmaCell platform, Besançon) with excitation at 460 nm and emission at 500 nm to study mucin 5AC secretion.

[0160] Results :

[0161] The study of intestinal epithelial cell layer permeability included paracellular JL transport (Figure 5A) and P-gp efflux protein activity using rhodamine 123 as a substrate (Figure 5B). Application of DSS to the apical side of the membrane resulted in a fourfold increase in paracellular permeability, independent of the presence of LPS (2.00 ± 0.12 × 10⁻⁶). -6 cm / s for induction with DSS-A / LPS-B, compared to 0.51 ± 0.11*1O 6 cm / s in the unstimulated control). Under DSS / LPS conditions, a halving of rhodamine efflux was observed compared to the unstimulated control (5.22 ± 0.74 × 10⁻¹⁰). 6 cm / s vs 10.51 ± 0.90*10' 6 cm / s). The Papp values ​​obtained with rhodamine were similar to those obtained with verapamil at 100 pM, a P-gp inhibitor.

[0162] Impaired intestinal barrier integrity has been shown to be closely linked to increased paracellular permeability and decreased P-gp-mediated efflux (Michielan and D'Incà, Mediators of inflammation (2015), 2015:628-157; Cario, World journal of gastroenterology (2017), 23(9), 1513-1520). Furthermore, these in vitro data are consistent with those obtained in in vivo and clinical models. Indeed, decreased P-gp-mediated efflux has been observed in patients with ulcerative colitis (Ma et al., International Journal of Clinical and Experimental Pathology (2017), 10(4):4552-4558) as in experimental models of DSS-induced murine colitis where P-gp activity and expression are decreased (lizasa et al., Journal of pharmaceutical sciences (2003) 92(3):569-76).

[0163] After incubation and washing, observations made by confocal microscopy revealed that the anti-MUC5AC conjugated markers had a higher rate of adhesion to the inflamed intestinal membrane than to the healthy membrane, visible by an intense white coloration (Figure 6).

[0164] This result shows that the increased expression of mucin 5AC mRNA was indeed associated with the secretion of this glycoprotein. These results correlate with clinical data where it was observed that MUC5AC was overexpressed by goblet cells in patients with ulcerative colitis (Shaoul et al., Journal of Pediatric Gastroenterology and Nutrition (2004), 38:488-493; Forgue-Lafitte et al., Int J Cancer (2007), 121:1543-1549) but also with preclinical data in an in vivo model of murine colitis (Riemann et al., Nanoscale (2024), 16(11):5715-5728).

[0165] Conclusion :

[0166] The model, characterized by increased permeability of the epithelial layer and altered composition of inflammatory mucus, followed by the secretion of inflammatory cytokines, reproduces the different phases of inflammation observed clinically and in vivo, demonstrating its suitability as a replacement for in vivo models of ulcerative colitis. Data obtained with treatments demonstrate its usefulness for evaluating preventive therapies. Furthermore, the model simulating acute inflammation offers the possibility of evaluating curative treatments. These models are valuable for selecting therapeutic agents because they allow for the assessment of the anti-inflammatory effect of treatments as well as their ability to penetrate the inflamed epithelial layer to interact with immune cells.

Claims

Claims

1. A model of intestinal epithelium comprising two compartments separated by a semipermeable membrane, said model comprising: - in the first compartment, corresponding to the apical pole of the intestinal epithelium, a coculture of Caco-2 cells differentiated into enterocytes and HT29-MTX cells differentiated into goblet cells; and - in the second compartment, corresponding to the basolateral pole of the intestinal epithelium, a culture of THP-1 monocytic cells differentiated into macrophages, said model being characterized in that the cells contained in the first compartment produce interleukin (IL-6) at a concentration greater than 100 pg / mL, and interleukin 8 (IL-8) at a concentration greater than 150 pg / mL; and the cells contained in the second compartment produce tumor necrosis factor a (TNF-a) at a concentration greater than 40 pg / mL and interleukin 1 P (IL-1 P) at a concentration greater than 90 pg / mL.

2. The model of claim 1, wherein the transepithelial electrical resistance is less than 85% relative to the transepithelial electrical resistance measured in a control non-inflamed intestinal epithelium model.

3. Model according to claim 1 or 2, characterized in that the expression of the mucin MUC2 is reduced by at least 30% compared to a control non-inflamed intestinal epithelium model, and / or the expression of the mucin MUC5AC is increased by at least 50% compared to a control non-inflamed intestinal epithelium model.

4. A method for obtaining a model of inflamed intestinal epithelium, said method comprising the following steps: iv. culturing, in a support comprising two compartments separated by a semi-permeable membrane, Caco-2 cells differentiated into enterocytes and HT29-MTX cells differentiated into goblet cells in the first compartment, and THP-1 monocytic cells differentiated into macrophages in the second compartment; v. contacting the differentiated Caco-2 and HT29-MTX cells contained in the first compartment with at least one compound selected from the group consisting of dextran sodium sulfate (DSS), trinitrobenzenesulfonic acid (TNBS) and dinitrobenzenesulfonic acid (DNBS); vi.bringing differentiated THP-1 monocytic cells contained in the second compartment into contact with a compound selected from the group consisting of lipopolysaccharides (LPS), TNF-α, interferon-γ (IFN-γ) and IL-10.

5. A method for obtaining a model of inflamed intestinal epithelium, said method comprising the following steps: iv'. culturing, in a support comprising two compartments separated by a semi-permeable membrane, Caco-2 cells differentiated into enterocytes and HT29-MTX cells differentiated into goblet cells in the first compartment, and THP-1 monocytic cells differentiated into macrophages in the second compartment; v'. contacting the differentiated Caco-2 and HT29-MTX cells contained in the first compartment with at least one compound selected from the group consisting of DSS, TNBS and DNBS; vi'. replacing said at least one compound with LPS, TNF-α, IFN-γ and / or IL- 1β.

6. The method of claim 4 or 5, wherein the Caco-2 cells were seeded at a concentration half that of the HT29-MTX cells.

7. A method according to any one of claims 4 to 6, wherein the Caco-2 cells were seeded at a concentration of 5,000 cells per compartment and the HT29-MTX cells were seeded at a concentration of 10,000 cells per compartment.

8. A method according to any one of claims 4 to 7, wherein the DSS, TNBS and / or DNBS has been brought into contact with the coculture of differentiated Caco-2 and HT29-MTX cells at concentrations ranging from 1 to 10% by weight per unit volume (w / V), preferably at a concentration of 5% w / V.

9. Method according to any one of claims 4 to 8, wherein the LPS, TNF-a, IFN-y and / or IL-1 has been brought into contact with the culture of THP-1 monocytic cells or into contact with the coculture of Caco-2 and HT29-MTX cells at concentrations ranging from 0.1 to 1000 ng / mL, preferably at a concentration of 100 ng / mL.

10. A method for the selection of candidate compounds for the treatment of inflammatory bowel diseases, said method comprising the steps of A) contacting a candidate compound with the intestinal epithelium model according to any one of claims 1 to 3; B) selection of the compound if it makes it possible to obtain, in said model, at least one of the following effects a) to e) a) A decrease of at least 10% in the level of IL-6 and / or IL-8 in the first compartment; b) A decrease of at least 10% in the level of TNF-a and / or IL-1 P in the second compartment; c) An increase in the expression of MUC2 of at least 5%; d) A decrease in the expression of MUC5AC of at least 5%; e) An increase in transepithelial electrical resistance of at least 10%.

Images