| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
INFLAMMATION/IMMUNITY/MEDIATORS
1-dependent pathway
1INSERM U539, IMAD, University of Nantes, 44035 Nantes, France; 2INSERM U601, 44093 Nantes, France; 3Department of Anatomy, University of Luebeck, Luebeck, Germany; 4INSERM U643, 44093 Nantes, France; 5Technical University Munich, Department of Human Biology, Freising, Germany; 6Department of Gastroenterology, University of Texas Medical Branch, Galveston, Texas
Submitted 16 June 2005 ; accepted in final form 14 January 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
) is altered in colon adenocarcinoma compared with control tissue. In an in vitro coculture model of subconfluent Caco-2 cells seeded onto Transwell filters with EGCs, Caco-2 cell density and [3H]thymidine incorporation were significantly lower than in control (Caco-2 cultured alone). Flow cytometry analysis showed that EGCs had no effect on Caco-2 cell viability. EGCs induced a significant increase in Caco-2 cell surface area without any sign of cellular hypertrophy. These effects by EGCs were also seen in various transformed or nontransformed intestinal epithelial cell lines. Furthermore, TGF-1 mRNA was expressed, and TGF-1 was secreted by EGCs. Exogenously added TGF-1 reproduced partly the EGC-mediated effects on cell density and surface area. In addition, EGC effects on Caco-2 cell density were significantly reduced by a neutralizing TGF- antibody. In conclusion, EGCs have profound antiproliferative effects on intestinal epithelial cells. Functional alterations in EGCs may therefore modify intestinal barrier functions and be involved in pathologies such as cancer or inflammatory bowel diseases.
colon cancer; enteric nervous system
, and vimentin. EGCs ensheathe enteric ganglia and are considered supportive and nutritive components of enteric neurons. EGCs respond by changes in intracellular calcium levels or expression of activation markers such as c-Fos to various mediators released by enteric neurons or by immune cells (22). However, the direct role of EGCs in the control of gastrointestinal function is presently largely unknown. A major function of the gastrointestinal tract is to control the passage of nutrients and fluids while preventing passage of microorganisms and toxic or noxious agents. This function is regulated by the intestinal epithelial barrier, which is located at the interface of the body and the luminal environment. This barrier is formed by an epithelial cell monolayer that lines the lumen. The epithelium undergoes rapid and constant turnover. Interactions between intestinal epithelial cells and the subepithelial cellular components of the mucosa play a key role in the control of intestinal barrier function under physiological and pathological conditions. It is well known that pericryptal myofibroblasts, which form a sheet underlying the intestinal epithelium, can modulate intestinal epithelial cell proliferation and differentiation, increase barrier resistance, and modulate secretory responses of the epithelium to various agonists such as acetylcholine or prostaglandins (4, 12).
Recent studies have shown that the ENS, a major constituent of the mucosa, can also modulate intestinal barrier functions. The ENS is an integrative neuronal network localized along the gut. It is composed of two major ganglionated plexuses: the submucosal plexus and the myenteric plexus located within the intermuscular space. The mucosal layer also contains delicate nerve networks known as the mucosal plexus, which extends within the lamina muscularis mucosae and lamina propria mucosae beneath the epithelial lining.
Studies have shown that enteric neurons can regulate intestinal barrier functions. In particular, activation of human submucosal neurons decrease paracellular permeability (20) and intestinal epithelial cell proliferation (27). However, the role of EGCs in the control of intestinal barrier function is presently unknown.
Recent in vivo studies have also suggested that EGCs could indeed be involved in the maintenance of the intestinal barrier. Indeed, ablation of EGCs in mice using chemical or immune-mediated methods leads to intestinal inflammation, associated with an initial alteration of mucosal (6) and vascular integrity (8). However, whether the alterations of the intestinal barrier was a consequence of intestinal inflammation, an effect of EGC ablation on enteric neurons, or a direct consequence of EGC ablation on intestinal barrier integrity is presently unknown.
Therefore, we used a noncontact coculture system to explore the role of EGCs in the control of a major function of intestinal barrier, i.e., intestinal epithelial cell proliferation. We show that EGCs inhibit intestinal cell proliferation and concomitantly increase the cell surface of intestinal epithelial cell. Furthermore, we identified TGF-1 as a mediator secreted by EGC to be involved in the inhibition of cell proliferation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Fragments of human adenocarcinoma and "healthy" control tissue taken at 10 cm from the tumor were obtained from patients undergoing surgery for colon carcinoma (n = 5 patients). According to the guidelines of the French Ethics Committee for Research on Human Tissues, these were considered as "residual tissues" and were not relevant to pathological diagnosis.
The tunica muscularis with adhering serosal fat was removed, with the submucosal and mucosal layers remaining in place. After fixation in PBS-buffered 4% paraformaldehyde (pH 7.4) at 4°C overnight and dehydration in graded alcohols, paraffin-embedded specimens were cut (4 µm) perpendicular to the gut axis. Sections were preincubated with 10% normal goat serum (Dakopatts, X-970) diluted in PBS-0.5% Triton X-100 (Sigma) for 30 min followed by incubation with polyclonal rabbit anti-S-100 (1:400, Dako) diluted in PBS-10% goat serum-0.5% Triton X-100 for 24 h at room temperature. After sections were rinsed in PBS, sections were incubated for 2 h in a buffer solution containing an anti-rabbit antibody conjugated to fluorescein isothiocyanate (1:200, Jackson ImmunoResearch). The slides were rinsed in PBS, mounted, and viewed with an Axiophot (Zeiss) equipped with an adequate filter system.
For whole-mount immunohistochemistry, specimens were stretched and pinned flat on wax-based Petri dishes before fixation to achieve optimal spacing of the remaining intestinal layers. After treatment with PBS-0.05% thimerosal (Sigma) overnight and 0.1% NaCNBH3 for 30 min, whole mounts of both the mucosal and submucosal layers were prepared by use of watchmaker's forceps under stereomicroscopic control. Whole mounts were treated with 10% normal goat serum diluted in PBS-0.5% Triton X-100 for 1 h followed by incubation with polyclonal rabbit anti-S-100 (1:400) diluted in PBS-10% goat serum-0.5% Triton X-100 for 48 h at room temperature. After extensive rinsings in PBS, whole mounts were incubated with biotinylated goat anti-rabbit IgG (1:200; Jackson ImmunoResearch) for 2 h, rinsed in PBS, and incubated with the avidin-biotin complex (Vectastain ABC Elite kit; Vector, Burlingame, CA) conjugated with horseradish peroxidase according to the instructions of the supplier. Peroxidase activity was detected with the chromogen 4-chloro-1-naphtol, resulting in a dark-blue reaction product. Whole mounts were rinsed in PBS, placed on slides, covered with Aquatex (Merck), and examined with an Axiophot microscope.
For transmission electron microscopy, small samples (
5 mm border length) of the mucosal-submucosal tissue layer were immediately fixed by immersion in 0.1 M cacodylate buffer containing 2.5% glutaraldehyde and 2% paraformaldehyde at pH 7.4 for 24 h. The specimens were postfixed in 1% OsO4 and stained en bloc with 2% uranylacetate. After dehydration in graded alcohols, the specimens were embedded in Araldite. Semi-thin sections were stained with methylene blue and azure II to visualize the regions of interest, in particular the enterocyte lining and the lamina propria mucosae. Ultrathin sections were cut and stained with lead citrate and examined with a transmission electron microscope (Phillips, EM 109). The findings were recorded both by conventional films (Agfa) and a digital image system (analySIS; Soft Imaging System, Münster, Germany).
Enteric Glial Cells
Nontransformed or transformed EGCs were generated as previously described (23). EGC cultures were isolated and purified from enzymatically dissociated preparations of rat longitudinal muscle-myenteric plexus. EGCs were cultured in DMEM (4.5 g/l glucose; GIBCO, Cergy-Pontoise, France) supplemented with 10% heat-inactivated FBS (GIBCO), 50 IU/ml penicillin, and 50 µg/ml streptomycin (GIBCO). EGCs were seeded at a concentration of 50,000 cells/cm2 in 12-wells plates (Corning, Avon, France). Cells were grown to confluence (3 days).
To assess the paracrine effect of EGCs on proliferation of intestinal epithelial cells, EGCs were cultured for 48 h after they reached confluence. Cell culture supernatants were collected and centrifuged at 1,000 g for 10 min to remove accidentally transferred cells and stored at –20°C until use as a conditioned medium.
Intestinal Epithelial and Fibroblast Cell Lines
Experiments were performed with four different human intestinal epithelial cell lines with distinct phenotypes. Cells were seeded onto porous filters (12-well Transwell Clear, 0.40-µm porosity; Corning). The Caco-2 cell line (EATCC, Port Down, UK) was cultured in DMEM (4.5 g/l glucose) supplemented with 10% heat-inactivated FBS, 2 mM glutamine (GIBCO), 50 IU/ml penicillin, and 50 µg/ml streptomycin. The T84 cell line (EATCC) was cultured in DMEM-F12 (1:1; GIBCO) supplemented with 10% heat-inactivated FBS and 50 IU/ml penicillin and 50 µg/ml streptomycin. The HT-29 and the mucus-secreting HT-29-Cl.16E cell lines were cultured in DMEM (4.5 g/l glucose) supplemented with 10% heat-inactivated FBS, 50 IU/ml penicillin, and 50 µg/ml streptomycin. The nontransformed rat small intestinal epithelial cell line IEC-6 (EATCC) was maintained in DMEM (4.5 g/l glucose) supplemented with 10% heat-inactivated FBS, 2 mM glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomycin. CCD 18Co cells (normal human colon fibroblasts; ATCC) were cultured in MEM (GIBCO) supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine, 0.1 mM nonessential amino acid (GIBCO), 50 IU/ml penicillin, and 50 µg/ml streptomycin and seeded at a density of 50,000 cells/cm2. All epithelial cell lines were seeded at a density of 65,000 cells/cm2, except T84 cells, which were seeded at a density of 180,000 cells/cm2.
Coculture Model
One day after epithelial cells were seeded onto Transwell filters, filters were cultured in the presence of EGCs seeded in the bottom of the 12-well plates. Components of the coculture model were cultured with the culture medium for epithelial cells. Half of the culture medium was changed daily.
The resistance of the epithelial monolayer was measured with an EVOM resistance meter (World Precision Instruments) and calculated by subtracting the electrical resistance of a blank insert from the measured value.
Animal Model
Transgenic mice expressing the thymidine kinase gene of the herpes simplex virus (HSV-TK) from the mouse glial fibrillary acidic protein promoter were used in this study (6). The in vivo ganciclovir (GCV) procedure was approved by an independent IACUC committee. Subcutaneous injection of GCV was performed for 7 days in transgenic and nontransgenic mice as previously described (6). Exposure to GCV in transgenic but not in nontransgenic mice caused a significant disruption of the enteric glial network (6). After 14 days of initiating GCV treatment, mice were injected intraperitoneally with 37 kBq of [methyl-3H]thymidine (Amersham) per gram body weight 90 min before the ileum was collected (at 11:00 AM so as to avoid circadian variation). Tissues were fixed in 10% phosphate-buffered formal saline (pH 7.2) for 24 h and embedded in paraffin using conventional methods. Paraffin sections (5 µm thick) were rehydrated through graded levels of alcohol and were processed for autoradiography by Ilford K2 photographic emulsion (Amersham). Slides were developed after a 3 wk-exposure in the dark at 4°C. Finally, all slides were counterstained with hematoxylin (Sigma) and mounted in Ralmount (Merck). Scoring of intestinal proliferation (>3 silver grains recorded over a blue nucleus) was performed with CRYPTS software as described previously (25) on a minimum of 20 crypt columns per sample using a Nikon x100 oil-immersion objective.
Cell Growth Studies
Cell counting. Epithelial cells were cocultured for various periods of time, harvested with 1% trypsin-EDTA (GIBCO), and homogenized in their respective epithelial cell culture medium. The resulting cell suspensions were counted in a blind fashion with the use of Malassez slides (VWR International, Strasbourg, France).
[3H]thymidine incorporation. After various times in coculture in the presence or absence of EGC, Caco-2 monolayers were incubated alone with DMEM containing [3H]thymidine (0.5 µCi/well) for 12 h. Cells were then removed from the filter with trypsin and harvested by a Titertek cell harvester (Flow Laboratories, Rickmansworth, UK) on a glass fiber filter (Wallac-Perkin Elmer, Courtaboeuf, France). The resulting filters were dried and incubated with Betaplate scintillation liquid (Wallac-Perkin Elmer), and beta radioactivity was counted by use of a scintillation spectrometer (Wallac-Perkin Elmer). [3H]thymidine incorporation for each culture duration in the presence or absence of EGC was analyzed with the Microbeta Windows Workstation software (Wallac-Perkin Elmer) and was normalized to the level of [3H]thymidine incorporation measured in Caco-2 cultured alone during the first day.
Protein, DNA, and RNA quantitation. Cell protein and nucleic acids were precipitated in 10% cold TCA for 1 h. DNA and RNA concentrations were determined after hydrolysis with 5% TCA (90°C for 20 min) using the orcinol colorimetric assay for RNA (16) and the diphenylamine reaction for DNA determination (5). Protein content in the insoluble fraction was quantified after acid hydrolysis. Proteins were dissolved in 1 M NaOH and quantified with the Bradford protein assay (Bio-Rad, Marnes La Coquette, France) and BSA (Sigma-Aldrich, St. Quentin Fallavier, France) as standard. Protein, DNA, and RNA concentrations were expressed as micrograms per well, and protein-to-DNA and RNA-to-DNA ratios were calculated for each coculture condition.
Two-Color Flow Cytometry Analysis of Cell Viability Using Annexin V and Propidium Iodide Staining
After 2–3 days in culture in the presence or absence of EGCs, Caco-2 Transwell filters were washed with 0.1 M PBS before harvesting of Caco-2 cells with trypsin. Cell suspension was then pooled with the supernatant and centrifuged for 10 min at 1,000 g. After centrifugation, the pellet was washed twice with PBS and resuspended in 100 µl of the staining solution of the annexin V-FITC and propidium iodide (PI) staining kit (BD Biosciences, Le Pont de Claix, France). After a 30-min incubation at room temperature, the percentage of cells undergoing apoptosis was determined by two-color flow cytometry using a FACScalibur (BD Biosciences). Results (annexin V- or PI-positive or -negative cells) were expressed as percent of total cells.
Morphological and Immunohistochemical Analysis
After coculture, filter-grown cells were fixed for 1 h in PBS containing 4% paraformaldehyde at room temperature. ZO-1, a tight-junction-associated protein, was used to determine the apical cell surface area and detected by immunofluorescence. Filters were preincubated for 30 min in PBS-4% horse serum (Sigma)-0.5% Triton X-100 (Sigma). Monolayers were then exposed to a monoclonal mouse anti-ZO-1 antibody, diluted in PBS-horse serum-Triton X-100 (1:500; Zymed, San Francisco, CA), for 1 h at room temperature and washed with PBS. Monolayers were then incubated for 30 min in a buffer solution containing an anti-mouse antibody conjugated to carboxymethylindocyanine (1:500; Beckman Coulter, Roissy, France). Monolayers were mounted and viewed with an Olympus IX 50 (Olympus, Rungis, France) connected to a black and white video camera (model 4910, Cohu). Apical cell surface area was measured with DP-Soft software (Olympus). An average of 456 ± 13 epithelial cells were analyzed for each experimental condition.
RT-PCR Analysis of TGF-1 mRNA Levels in EGC
Extraction of total RNA from EGC cells was performed with TriReagent (Euromedex, Mundolsheim, France) according to the manufacturer's instructions. For reverse transcription, RNA (1 µg) was mixed with 0.5 µg of random hexamer primers (Amersham, Orsay, France), transcription buffer (50 mM Tris·HCl, pH 8.3, 75 mM KCl, 3 mM MgCl2, 10 mM DTT; GIBCO), dNTPs (10 mM each, GIBCO), and RNasin (50 units; Promega, Charbonnieres, France) to synthesize single-stranded cDNA using the Superscript II reverse transcriptase kit (GIBCO) according to manufacturer's instructions in a total volume of 20 µl. Incubation was performed at 42°C for 60 min. PCR amplifications were performed with Goldstar red Taq DNA polymerase (Eurogentec, Angers, France). The cycling conditions were as follows: denaturation for 5 min at 95°C, amplification for 30 cycles, with denaturation for 5 s at 95°C, annealing for 30 s at 55°C, and extension for 45 s at 72°C. The following primers were used as previously described (15): TGF- all isoforms (forward: 5'-TACATTGACTTTAGGAAGGA-3'; reverse: 5'-ATCATGTTGGACAACTGCTCC-3'), PCR product size of 252 bp; TGF-1 (forward: 5'-CAAAGACATCACACACAGTA-3'; reverse: 5'-GGTGTTGAGCCCTTTCCAGG-3'), PCR product size of 448 bp; -actin (forward: 5'-CCTTCCTGGGCATGGAGTCCTG-3'; reverse: 5'-GGAGCAATGATCTTGATCTTC-3'), PCR product size of 201 bp.
After addition of 5 µl of loading buffer (6x; Sigma) and 1 µl of Syber green (1:1,000; Roche Diagnostics, Meylan, France), PCR products were loaded onto a 1.5% agarose gel (Eurobio, Ulis, France) and separated by electrophoresis. DNA size markers were run in parallel to validate the predicted sizes of the amplified bands (100-bp DNA ladder; GIBCO). PCR products were visualized with use of the Gel Doc 2000 system (Bio-Rad, Paris, France) and analyzed with Quantity One software (Bio-Rad).
Pharmacological Tools
Rat TGF-1 levels were determined by ELISA kit (Diaclone, Besançon, France). Immunoneutralization of TGF-1 was achieved with a pan-specific TGF- antibody (R&D Systems, Lille, France).
Statistics
Data are expressed as means ± SE when normally distributed or as the median (25–75%) when nonnormally distributed. A paired or unpaired t-test, a Mann-Whitney test, and one-way ANOVA followed by a Bonferroni t-test or two-way ANOVA on repeated measures were performed to compare different populations. Differences were considered as significant for P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
in Healthy Areas and in Human Colonic Adenocarcinoma
Immunohistochemical analysis of human colonic mucosa revealed a prominent honeycomb-like latticework of S-100-immunoreactive EGCs surrounding the colonic epithelial crypts (Fig. 1A) and present all along the crypt axis (Fig. 1C). These glial cell networks extended throughout the lamina propria mucosae, forming part of the subglandular and periglandular portions of the mucosal plexus directly located beneath the epithelial lining. In addition, strong S-100 immunoreactivity was observed in elongated cells accompanying mucosal blood vessels (Fig. 1B), indicating the presence of glial cells ensheathing nerve fibers of the perivascular nerve network. Transmission electron microscopy studies revealed a close topographic proximity between EGC associated with mucosal nerve fibers and the basal lamina of intestinal epithelial cells (Fig. 1D). Distance between EGC and enterocytes ranged between 0.5 and 2 µm.
|
|
To characterize the role of glial cells on intestinal epithelial cell proliferation, glial cells were specifically ablated in a transgenic mice model previously developed by Bush et al. (6). After glial cell ablation by GCV in transgenic mice, the mean [3H]thymidine labeling index in the intestinal crypts was significantly increased by 51% (n = 4) compared with control mice (Fig. 3). In addition, analysis of the distribution of [3H]thymidine-labeled cells along the crypt-axis revealed 1) that at a given crypt cell position the proliferation index was increased and 2) that labeled cells were distributed further up in the crypt in transgenic animals compared with controls.
|
To study the direct effects of EGCs on intestinal epithelial cells, a coculture model was used. After seeding procedures, subconfluent Caco-2 cells cultured alone (control) initially formed clusters of cells. For several days, we observed spaced islands of growing cells (Fig. 4, A and C). In contrast, rapid and drastic changes were observed when Caco-2 cells were cocultured with EGC. Indeed, in the presence of EGC, filter-grown Caco-2 cells did not form large multicellular clusters as observed when cultured alone; instead, the entire filter covered rapidly and uniformly within 2–3 days after seeding (Fig. 4, B and D). After 6 days in culture, the filter was uniformly covered under both experimental conditions (Fig. 4, E and F). These observations were correlated with changes in filter resistance. In the presence of EGCs, a significant and time-dependent increase in the resistance of Caco-2 monolayer was measured compared with that observed in control samples (Fig. 4G). Furthermore, incubation of Caco-2 cells with the conditioned medium from EGC cultures reproduced the effect of coculture with EGCs (Fig. 4G).
|
We aimed to characterize whether EGCs altered Caco-2 cell density. Under control conditions, Caco-2 cell density increased significantly over time and reached (1.21 ± 0.29) x 106 cells/filter after 6 days in culture. In contrast, in the presence of EGCs, after an initial increase in Caco-2 cell density during the first 48 h, cell density remained constant for up to 6 days in coculture (Fig. 5A). Under control conditions, Caco-2 cell density was significantly higher after 3 days compared with Caco-2 cells cultured in the presence of EGC [(3.8 ± 0.5) x 105 vs. (2.2 ± 0.3) x 105 cells/filter; P < 0.05; n = 4]. In addition, incubation of Caco-2 cells with EGC-conditioned medium significantly decreased the Caco-2 cell density compared with control and with a similar time course as with EGCs (data not shown). After 6 days of culture in the presence of EGC-conditioned medium, Caco-2 cell density was significantly decreased compared with that in control condition [(7.2 ± 1.6) x 105 vs. (12.1 ± 3.0) x 105 cells/filter; P < 0.05; n = 4].
|
To determine whether changes in cell density were associated with intestinal epithelial cell death, we assessed the effect of EGCs on the viability of Caco-2 cells by two-color flow cytometry analysis after annexin V and PI staining. The vast majority of Caco-2 cells (>90%) cultured or not with EGCs were PI and annexin V negative (Fig. 6), suggesting that Caco-2 cell viability was not affected by EGC. In addition, there was no difference in the percentage of dead cells observed under the different experimental conditions (Fig. 6).
|
To explain the simultaneous increase in the sealing of the monolayer and decrease in density of Caco-2 cells in coculture with EGCs, the effect of EGCs on the surface area of Caco-2 cells was investigated. The apical surface area of individual epithelial cells was assessed by measuring the area of cells labeled with an antibody directed against ZO-1 (Fig. 7, A and B). The surface area of Caco-2 cells was 2.2 times greater when Caco-2 cells were cocultured with EGC than under control conditions (Fig. 7, A–C).
|
Effect of EGCs on Cell Density in Other Intestinal Epithelial Cell Lines
To determine whether EGCs are able to decrease cell density in other types of intestinal epithelial cells, we characterized their effect in other transformed and nontransformed intestinal epithelial cell lines. After 6 days in coculture, a significant decrease in cell density (ranging between 72 and 66%) was observed in all transformed cell lines tested (T84, HT-29, and HT-29-Cl.16E) when cultured with EGCs compared with control cultures without EGCs (Fig. 8, A–C). Similarly, coculture with EGCs decreased cell density in cultures of nontransformed rat epithelium-derived IEC-6 cells by 76% compared with control (Fig. 8D).
|
To test whether the anti-proliferative effect of EGCs was specific for EGCs, we investigated the effects of intestinal pericryptal myofibroblasts, another major constituent of the intestinal mucosa. In contrast to the effects observed with EGCs, after 6 days in culture, Caco-2 cell density slightly increased, although not significantly, when cultured in the presence of human intestinal fibroblasts (CCD 18Co) as in control conditions (Fig. 9A). In addition, after coculture with intestinal fibroblasts, [3H]thymidine incorporation in Caco-2 cells was increased compared with control values after 3 days in culture (Fig. 9B).
|
on EGC-Mediated Effects
TGF- is known to significantly inhibit intestinal cell proliferation by inducing cell cycle arrest. We wanted to verify whether EGCs were a source of TGF-, which may partly account for the effect of EGC on cell proliferation.
TGF-1 mRNA was found to be expressed in EGCs cultured alone or in the presence of Caco-2 cells (Fig. 10A). Furthermore, as assessed by ELISA, TGF-1 was secreted by EGCs in the culture medium of EGCs when cultured alone or cocultured with Caco-2 cells (6.7 ± 3.2 and 8.8 ± 1.3 ng/ml, respectively; P = 0.2; n = 4) (Fig. 10B). TGF-1 levels in Caco-2 cell culture were below the detection limit of the kit.
|
1 was also shown to induce a dose-dependent decrease in Caco-2 cell density, leading to a maximum 20% decrease in cell density (Fig. 10C). When Caco-2 cells and EGCs were cocultured, incubation of the basolateral medium with an anti-TGF--neutralizing antibody (10 µg/ml) significantly increased the number of Caco-2 cells. Cell density increased by 38 ± 15% (n = 4; P < 0.05) compared with that in the absence of blocking antibody (Fig. 10D). Addition of the neutralizing antibody to control cultures did not affect the total cell density [(7.5 ± 0.4) x 105 to 106 vs. (7.6.105 ± 0.4) x 105 cells/filter; P = 0,89; n = 5]. Finally, TGF- (10 ng/ml) significantly increased the surface area of Caco-2 cells compared with that of Caco-2 cells cultured alone (177 ± 2 vs. 119 ± 1 µm2; n = 4; P < 0.001), reproducing in part the effect of EGC.
Additional experiments also revealed that TGF-1 significantly reduced cell proliferation in other intestinal epithelial cell lines tested. Indeed, incubation of IEC-6 and T84 with 10 ng/ml of TGF-1 significantly reduced cell number by 14% (n = 9; P < 0.05) and 11% (n = 5; P < 0.05), respectively compared with control.
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
1 was identified as a factor synthesized by EGCs and shown to mediate at least partly the effects of EGCs. Finally, the decreased EGC network observed in colorectal tumors could favor cell proliferation or invasion.
Previous studies have reported in various species (guinea pigs, mice, rats, or humans) that EGCs are abundantly distributed within enteric ganglia and also form dense networks within the intestinal mucosa (7, 11, 14, 18). Our study is in agreement with these previous observations and reveals that, in human healthy colonic mucosa, a dense network of S-100-positive EGCs is present along the crypt axis (Fig. 1). The close contact revealed by the electron microscopic study between EGCs (which ensheathe enteric nerves) and epithelial cells strongly suggests that these cell types could interact through a diffusible factor.
A major finding of this study is that EGCs significantly inhibit intestinal epithelial cell proliferation in vivo and in vitro. In vivo, ablation of glial cells leads to an increase in thymidine incorporation in epithelial cells, indicating an increase in epithelial cell proliferation. However, one cannot exclude that this increase in cell proliferation was associated in response to the intestinal inflammation that occurs in this model (6). Direct evidence of EGC control of cell proliferation is further provided by the in vitro study. Indeed, to characterize direct interactions between EGCs and intestinal epithelial cells, we used a coculture model. Rat EGCs were used because no methods for the isolation and long-term culture of human EGCs are currently available. Although our results need further confirmation in humans, interspecies differences are unlikely to be involved because rats EGC had similar effects on human and rat intestinal epithelial cell lines. We also showed that EGCs exert their action probably through a paracrine mechanism. Indeed, no cell-to-cell contact exists in our model, and the effects of EGCs were reproduced with conditioned EGC culture medium. Furthermore, coculture of EGCs with transformed (HT-29, Caco-2, T84) or nontransformed (IEC-6) intestinal epithelial cell lines suggests that EGCs play a role in the inhibition of primary intestinal epithelial cell proliferation. The profound inhibitory role of EGCs on intestinal epithelial cell proliferation is to oppose the role of another major constituent of intestinal mucosa, i.e., intestinal fibroblasts. Indeed, intestinal fibroblasts have been shown to promote intestinal cell proliferation predominantly through the release hepatocyte growth factor (12).
An important finding of our study is that we identified TGF-1 as being synthesized and released by EGCs and mediating in part the effects of EGCs on epithelial cell proliferation and cell surface. The phenotypical characteristics of EGCs have been poorly studied, although it is presently admitted that they synthesize factors similar to those produced by astrocytes of the CNS such as glial-derived neurotrophic factors (2), L-arginine (19), and IL-6 (21). Our study shows that, similarly to CNS astrocytes, EGCs can also synthesize TGF-1. In the CNS, TGF-1 secreted by astrocytes plays a key role in neuronal homeostasis and, in particular, has a neuroprotective role (9, 26). However, the role of TGF-1 in the control of the homeostasis of the ENS remains unknown, although it is tempting to speculate that TGF-1 released by EGC acts as a regulator of enteric neuronal survival, as ablation of EGCs leads to enteric neuronal degeneration (6).
The anti-proliferative effect of TGF-1 released by EGCs is consistent with the key role played by TGF-1 in the control of intestinal epithelial cell proliferation. It has been widely demonstrated that TGF-1 inhibits epithelial cell proliferation (3, 16) while stimulating epithelial cell migration (10). In accordance with these findings, our data show that TGF-1 inhibits proliferation of intestinal epithelial cells in a dose-dependent manner. Interestingly, TGF-1 levels present in EGC-conditioned medium induces a decrease in the Caco-2 cell density equivalent to the increase in Caco-2 cell density induced by EGCs in the presence of the neutralizing TGF-1 antibody. The molecular mechanisms responsible for the inhibition of epithelial cell proliferation were not investigated. However, it is well known that TGF-1 induces a significant inhibition of intestinal cell proliferation by promoting cell cycle arrest (24). This effect is either due to a downregulation of the levels and/or activities of G1/S cyclins and cyclin-dependent kinases or to an upregulation of the level and/or activities of cyclin-dependent kinase inhibitors (24). Various intracellular pathways, such as MAPK pathways, the phosphatidylinositol 3-kinase cascade, or small G proteins such as RhoA or Rac, can be directly activated by TGF-. Activation of the latter pathways, which have been shown to be involved in the control of cytoskeletal filament assembly, may be partly involved in the increase in cell surface induced by TGF-1.
Because TGF-1 accounts for 12–30% of the effects of EGC on intestinal epithelial cell proliferation (depending on the cell line studied) and cell surface, other major factors secreted by EGCs that may mediate these effects remain to be identified. In particular, factors secreted by EGCs that act on the intestinal epithelial cytoskeleton may contribute to increased cell surface area of epithelial cells. This increase in cell surface may be sufficient to induce cell contact-mediated cell cycle arrest. Such a mechanism may be compatible with our model because, before changes in cell number, i.e., during the first 48 h of culture, changes in the cell surface already occurred (unpublished data).
Our study supports the role of EGCs as a major cellular component of the gut involved in the control of the intestinal barrier homeostasis. Our results support in vivo findings that showed that controlled immune-mediated glial cell alteration increases intestinal paracellular permeability (1) and that full ablation of EGCs leads to the disruption of the intestinal barrier (5).
Another potentially important finding of this study is the alterations in the enteric glial network (identified with S-100) observed in colon carcinoma. Indeed, although one cannot presently determine whether altered glial cells are involved in tumorigenesis, the absence of enteric glia in tumors could favor cellular proliferation and/or invasion. The cellular nature of the tumor microenvironment has been shown to have a profound impact on tumor development. Therefore, besides fibroblasts or endothelial or immune cells, EGCs could be regarded in the future as an important component regulating the tumoral environment.
In conclusion, our results reinforce the emerging concept that the ENS and, in particular, EGCs are major regulators of intestinal barrier functions and homeostasis. In particular, neuronal and glial components of the ENS could be seen as acting in concert with other components of the mucosa, such as fibroblasts, to finely tune intestinal epithelial cell proliferation and to contribute to the maintenance of the intestinal barrier integrity. In addition, functional alterations in EGCs could be involved in various intestinal pathologies affecting intestinal barrier functions such as cancer or intestinal inflammatory bowel diseases. Identification of other glial cell factors involved in the control of intestinal cell proliferation could therefore be of future great therapeutic interest.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
* M. Neunlist and P. Aubert contributed equally to this study. 
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
. Gastroenterology 105: 1323–1332, 1993.[Web of Science][Medline]
and in rat intestinal epithelial cells. J Clin Invest 83: 1768–1773, 1989.[Web of Science][Medline]-induced SMAD signaling and gene regulation: consequences for extracellular matrix remodeling and wound healing. J Dermatol Sci 35: 83–92, 2004.[CrossRef][Web of Science][Medline]-amyloid-induced neuronal death. Endocrinology 145: 5080–5086, 2004.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
0.05). B: in transgenic mice (
), the proportion of [3H]thymidine-labeled cells was increased along the crypt axis compared with control (
). In addition, besides increased thymidine incorporation, crypt cell hyperplasia was also observed (n = 4 mice).
) than under control conditions (
). The effect of the EGC on filter resistance was also observed with EGC-conditioned medium (
