In the intestine, M cells scattered in the epithelial cells covering the lymphoid follicles of Peyer's patches perform antigen sampling as the first step in developing immune responses. M cells act as regulated gates in epithelial barriers that can be used opportunistically by pathogens to invade their host. Using a cultured model of M cell, we have studied the cellular and molecular mechanisms involved in the differentiation and uptake processes of M cells, in particular receptors involved in pathogen adhesion and specific gene regulation that controlled this process

In the Laboratory, we study the homeostasy of the epithelial barrier in response to agressions like pathogen infections and physicochemical changes, in relation with the underlying mucusal tissue that triggers immune responses, tolerance and, in patologic situations, local inflammation that can modulate the fate of epithelial cells. For these studies we have chosen M cells of intestinal Peyr's patches as a paradigm, and previously developed acultured model allowing molecular, cell biology and kinetical study. We have recently extended these studies to intestial metaplasia due to acid reflux in the Malpighian epithelium of the esophagus (Barrett syndrom). The project is conducted by four convergent approach :

1.Different pathways for transport of pathogenic BACTERIA through M cells of Peyer's patches. Leaded by S. Kerneis.

M cells of intestinal Peyer's patches (PP) sample a large variety of antigens and microorganisms present in the gut lumen, and convey them to underlying lymphoid follicles. This physiological function is essential for the development of mucosal immunity. Many pathogenic microorganisms infecting the host by the oral route use M cells as gateways to cross the epithelial barrier of the digestive tract involving possibly two distinct but not exclusive mechanisms : a passive transport via a non-specific macropinocytosis-like process and/or a specific mechanism of bacterial adhesion to apical membrane components of M cells. To identify which of these mechanisms is used by pathogens we have investigated bacteria/M cell interactions in terms of both adhesion and translocation processes, using a previously described cultured model of M-like cells. Four different bacterial genus were chosen for their tropisms and mechanisms of pathogenicity : Shigella, Yersinia, Escherichia and Listeria. Shigella flexneri M90T and Yersinia enterocolitica O8, , we have reported that four pathogenic bacteria, Shigella flexneri, Yersinia enterocolitica, enteropathogenic Escherichia coli, and Listeria monocytogenes, which infect their host by oral route, showed that their adhesion profile did not correlate with their uptake transcytosis efficiency. This suggests that there are distinct pathways in which bacteria can cross M cells.

We have evidenced a Peyer's patch lymphocyte-induced increase of apical and basolateral membrane localization of beta-1 integrin in intestinal Caco-2 cell monolayers and showed that PP lymphocyte/Caco-2 cell interactions might trigger the mobilization of pre-synthetized intracellular pool of b1 integrins towards the apical and basolateral membranes of epithelial intestinal cells. Membrane relocalization of b1 integrins induced by lympho-epithelial interactions was selective and did not concern all cell adhesion proteins, even if they are involved as receptor for bacterial entry like E-cadherin described as the receptor for L. monocytogenes. Peyer's patch lymphocytes induced shrinking of the intracellular pool of pre-b1 integrin with concomitant increase in the rate of conversion from precursor to mature glycoprotein. Immunoprecipitation and pulse-chase metabolic labeling experiments demonstrated a lymphocyte-driven acceleration of the maturation of the glycosylation of the b1 integrin pool, confirming the above described increase of membrane-anchored b1 integrins in Caco-2 cell moonolayers co-cultured with Peyer's patch lymphocytes. Peyer's patch lymphocytesalso induced the synthesis of a subunits associated with b1 integrins on epithelial cell membranes. On the apical membranes of control and co-culture cells, a2 integrin was the only a subunit that could be detected by immunoprecipitation of apically biotynylated cell lysates.There was no obvious difference between the two culture conditions in the amount of apical a2 chains expressed.When subjected to streptavidin labeling, a spot was revealed by streptavidin on biotynylated cell extracts of both Caco-2 monolayers cultured alone and co-cultured with Peyer's patch lymphocytes but a strong increase in intensity was observed on co-cultures extracts. In sharp contrast, such difference in amount of immunoprecipitated a integrin was no longer observed when the mini2D blot was labeled with anti-a 2integrin antibody . This clearly confirmed that alpha- integrin subunit expression could not account for the total lymphocyte-driven increase of alphasubunits expression at the apical membrane of Caco-2 monolayers.

We provide evidence that PP lymphocytes can themselves modulate gene expression in PP in vivo and in an in vitro model of FAE. Transgenic mice carrying a reporter gene under the control of a modified L-pyruvate kinase promoter (SVPK) exhibit strong transgene expression in PP and FAE, but not in the adjacent villous cells. We used the mouse intestinal epithelial cell line m-ICcl2 transfected with the SVPK promoter fused to ß-galactosidase to investigate the direct effect of PP lymphocytes on SVPK promoter activity. ß-galactosidase expression was 4.4-fold higher in transfected m-ICcl2 cells when they were cultured with PP lymphocytes. Conversely, GFP expression was 1.8-fold lower in stably transfected differentiated intestinal Caco-2cl1 cells with the sucrase-isomaltase promoter (SI) fused to GFP cDNA when they were cultured with PP lymphocytes, indicating that the in vivo FAE down-regulation of SI is transcriptionally regulated.

4. Development of an in vitro model system to study the cellular conversion in Barrett's syndrome. Leaded by M. Marchetti.

Barrett's oesophagus (BE) consists in a metaplasia in which the pluristratified epithelium of the oesophagus is replaced by a monostratified intestinal epithelium. This condition is mainly associated to the development of oesophageal cancer. The aim of the project consisted in setting up an in vitro model system for the understanding and identification of factors responsible of this cellular differentiation (oesophageal versus intestinal type) by assessing the transcription of intestinal homeobox genes. We have established a long-standing cultures of normal mouse oesophageal keratynocites that is able to differentiate and form multilayer an in vitro model system. Cells were cultured on filter supports and cell culture was monitored by expression of cyokeratine 14 (CK14) and cytokeratine 4 (CK4). It resulted that cell culture progressed in the formation of a primary basal layer of undifferentiated basal cells (CK 14 positive) and a secondary differentiated suprabasal cell layer (CK 4 positive) thus, mimicking the proliferation and stratification patterns as found in the normal oesophagus. Cells were then successfully transfected with a plasmid containing the GFP reporter gene under the control of the cdx2 promoter region and exposed to acidic pH. When cultured at pH 5 and 3.5, acid exposure did not altered cell proliferation and differentiation patterns compared to neutral pH. However, long-term exposure to pH 5, induced activation of the promoter region of the intestinal homeobox gene, cdx2 as visualized by GFP expression. Thus, exposure to low pH in the apical chamber of transwell devices modified at least in part the keratynocite genetic program and activate intestinal homeogene cdx2 promoter. We conclude that this model could be an useful in vitro model to study the physiology of oesophageal cells and the pathological changes that occur in the oesophagus as seen in the intestinal metaplasia of Barrett oesophagus.

Figure 1 : Transmission electron microscopy view of an M cell of Peyer's patch. In blue, a parasite., in yellow a lymphocyte settled in the M cell pocket.

Figure 2: Conversion of mouse intestinal crypt m-ICcl2 cells into M-like cells by PP lymphocytes.

A: Double indirect immunostaining with FITC-conjugated phalloidin and CY3-conjugated anti-Ig Ab of confluent m-ICcl2 cells cocultured with mouse PP's lymphocytes. The F-actin network (green) outlines the cells, and 20-30% of the cells display the intense CY3 staining (red) labeling of PP lymphocytes. The inset shows the labeling of the intraepithelial lymphocytes at higher magnification. B: Temperature-dependent kinetics of the translocation of fluorescent latex beads. The basal-to-apical transport of fluorescent latex beads was detected by FACS analysis and expressed in arbitrary units (A.U.), in confluent m-ICcl2 cells cultured on porous filters with or without PP lymphocytes at 4°C and 37°C, as described in Materials and methods. , m-ICcl2 alone; , m-ICcl2 cells cocultured with PP's

Figure 3: In vitro activation of the pyruvate-kinase promoter by PP lymphocytes

A: Western blot analysis of b-gal (121 Kd) and Tag (66 Kd, internal control) levels in monolayers of m-ICcl2 cells cultured alone (lane 1) or cocultured with PP lymphocytes (lane 2). B: Quantification of the chemiluminescent signals from Western blots on m-ICcl2 cells (1) and m-ICcl2 cells cocultured with PP lymphocytes (2) using a STORM laser scanner and Image Quant software. Values (in arbitrary units) correspond to the peak area intensity of b-gal and Tag signals and are the means of two independent experiments. C: In-situ detection of b-gal activity (in blue) in confluent monolayers of m-ICcl2 cells cultured alone on filters (1) or cocultured with PP lymphocytes (2). The fine black dots are the pores of the 3 m m porous filter. D: merged image of light transmission and fluorescence microscopy of the same field showing nuclear b-gal staining and FITC-conjugated particles inside WGA-labelled epithelial cells. In blue, b-gal+ nucleus; in green, FITC-conjugated particles; in red, WGA labelling. Note that b-gal^- surrounding epithelial cells did not contain any intracellular fluorescent particles.

Hamzaoui Nadim, MD, Institut Pasteur, physician-researsher. No e-mail

El Bahi Sophia, PhD student, CANAM fellow. elbahis@pasqteur.fr

Caliot Elise, Institut Pasteur, Ingeneer . ecaliot@pasteur.fr