The first 2 authors contributed equally to this work.
Effects of T cell-induced colonic inflammation on epithelial barrier function†
Article first published online: 12 FEB 2010
Copyright © 2010 Crohn's & Colitis Foundation of America, Inc.
Inflammatory Bowel Diseases
Volume 16, Issue 8, pages 1322–1331, August 2010
How to Cite
Suenaert, P., Maerten, P., Van Assche, G., Van Driessche, W., Geboes, K., Bulteel, V., Simaels, J., Augustijns, P., Ceuppens, J. L., Rutgeerts, P. and Perrier, C. (2010), Effects of T cell-induced colonic inflammation on epithelial barrier function. Inflamm Bowel Dis, 16: 1322–1331. doi: 10.1002/ibd.21211
Supported by grants from the Foundation for Scientific Research Flanders (FWO-Vlaanderen, grant G.0507.06), by a Researcher Fellowship Funding of the Flemish Fund for Scientific Research (FWO-Vlaanderen, grant 1.5.054.02) (to P.S.), by a grant from the Research Council of the Catholic University of Leuven, and by a Schering-Plough IBD fellowship (to P.M., C.P.).
- Issue published online: 13 JUL 2010
- Article first published online: 12 FEB 2010
- Manuscript Accepted: 30 NOV 2009
- Manuscript Received: 18 NOV 2009
- Foundation for Scientific Research Flanders (FWO-Vlaanderen, grant. Grant Number: G.0507.06
- Flemish Fund for Scientific Research (FWO-Vlaanderen, grant. Grant Number: 1.5.054.02
- Research Council of the Catholic University of Leuven
- Schering-Plough IBD fellowship
- Crohn's disease;
- experimental colitis;
- epithelial barrier function;
Epithelial barrier disturbance is thought to contribute to the pathogenesis of inflammatory bowel diseases; however, it remains unclear whether it is a primary defect participating to the onset of inflammation or only a consequence of sustained inflammation.
A time course study of epithelial barrier functions and immune mediators was performed in the CD4+CD45RBhi T cell transfer model of colitis using Ussing chambers.
In nonreconstituted severe combined immunodeficiency (SCID) mice, no epithelial dysfunction was observed. However, after transfer of CD4+CD45RBhi T cells or total CD4+ T cells, colon of SCID mice displayed a decreased epithelial resistance, even before overt microscopic inflammation had occurred. Sustained colitis of CD4+CD45RBhi T cell reconstituted mice was also associated with enhanced subepithelial resistance, enhanced paracellular permeability, and decreased net ion transport. All these reflect a disturbance of barrier function and may contribute to diarrhea. Epithelial resistance was positively correlated with interleukin 10 (IL-10) and transforming growth factor β (TGF-β) levels and net ion transport inversely correlated with tumor necrosis factor alpha (TNF-α) levels, pointing to the protective effect of IL-10 and TGF-β and to a damaging effect of TNF-α. Indomethacin, a nonselective COX inhibitor, decreased epithelial resistance independent of T cells and inflammation, but its effect was more pronounced in inflamed colon.
Induction of colitis by transfer of CD4+CD45RBhi T cells in SCID mice leads to changes in the colonic epithelium before colitis develops. Decreased epithelium resistance might contribute to the development of colitis; however, it is not sufficient to lead to chronic inflammation. (Inflamm Bowel Dis 2010)
Crohn's disease (CD) is a segmental inflammatory disease of the intestine characterized by a massive infiltration of CD4+ T cells and macrophages and by epithelial damage. The relationship between inflammation and epithelial dysfunction is not well understood.
Diarrhea, which is frequently associated with intestinal inflammatory diseases including enteric infections, ulcerative colitis, CD, and celiac disease, reflects an imbalance between intestinal absorption and secretion of water and electrolytes. Inflammation-induced changes in epithelial cell function may induce a decrease in active colonic sodium absorption and/or increased chloride secretion. Moreover, inflammation can be associated with altered tight junction structure, resulting in an impaired epithelial barrier function with increased loss of electrolytes and water into the lumen.1–3
The gut epithelial barrier is an essential component of the innate immune function preventing or limiting the entry of potentially noxious substances, such as microbes and antigens, into the body. A variety of defense mechanisms are located in close proximity to the epithelium, and it is well established that immune cells and inflammatory mediators affect epithelial cell function.4
In vivo murine models of colitis offer a potential approach to characterize the intestinal epithelial function in inflamed mucosa in comparison to healthy circumstances. In the T cell transfer model, severe combined immunodeficiency (SCID) mice,5 which lack T and B lymphocytes, are reconstituted with syngeneic CD4+CD45RBhi T cells and develop a wasting disease 6–10 weeks after T cell reconstitution.6 The disease is characterized by diarrhea, weight loss, and transmural inflammation in the colon. The colitis results from a Th1 immune response toward enteric bacterial antigens, since CD4+CD45RBhi T cell transfer into SCID mice raised in sterile circumstances fails to induce disease.7 The histopathological and immunological features, as well as gene expression profile, resemble those observed in CD.6, 8, 9 Concomitant reconstitution with regulatory T cells (contained within the CD4CD45RBlo T cell subset) protects against colitis through regulatory T cell secreted interleukin 10 (IL-10) and transforming growth factor β (TGF-β).10, 11 The advantage of the T cell transfer model of colitis for the study of the effects of inflammation on the epithelial barrier is the absence of any intrinsic barrier defect or direct intestinal epithelial damage, as is the case in other models in which exogenous agents like dextran sulfate sodium, ethanol trinitrobenzene sulphonic acid, iodoacetamide, or indomethacin are used.12
Diarrhea in several colitis models is due to tumor necrosis factor (TNF)-mediated enteral fluid loss and decreased sodium and water absorption rather than by enhanced ion secretion.13 On the other hand, TGF-β treatment downregulates CFTR (cystic fibrosis transmembrane conductance regulator) and diminishes the chloride secretion in monolayer experiments.14, 15 This suggests an important role of TGF-β in the maintenance of water balance. Furthermore, IL-10 is able to inhibit epithelial cell dysfunctions that are induced by interferon-γ (IFN-γ) or superantigen-activated immune cells in an in vitro cell culture model.16, 17 However, a great deal of uncertainty remains about the changes in epithelial barrier and secretory function in vivo in the early phase of intestinal inflammation.
In the present study we investigated the temporal relationship between the T cell-induced mucosal colonic inflammation and the change in epithelial barrier function in the transfer model of colitis.
MATERIAL AND METHODS
Reagents and Mice
Specific pathogen-free female Balb/c mice were obtained from Harlan (Horst, The Netherlands). Female C.B-17 SCID mice were obtained from the REGA Institute (Leuven, Belgium). All mice were maintained in the animal care facility of the faculty of medicine, Catholic University of Leuven, in micro-isolator cages with filtered air and free access to autoclaved food and water. All studies were approved by the local ethical committee on animal experimentation of the Catholic University of Leuven.
FITC-conjugated antimouse CD45RB monoclonal antibody (mAb) (clone 16A, rat IgG2a,κ) and PE-conjugated antimouse CD4 mAb (clone GK1.5, rat IgG2b) were purchased from PharMingen (San Diego, CA).
Isolation of CD4+ T cell Subpopulations and Induction of Colitis
Spleen cells from Balb/c mice were used as a source of CD4+ cells and further separated into a CD4+CD45RBhi and CD4CD45RBlo fraction under sterile conditions by 2-color sorting on a FACS Vantage (Becton Dickinson, CA), as described previously.18 All populations were >98% pure on reanalysis. To induce colitis, 8- and 12-week-old SCID mice were intraperitoneally injected with CD4+CD45RBhi T cells (4–5 × 105 cells/mouse). Some mice were injected i.p. with total CD4+ T cells (5–6 × 105 cells/mouse) or alternatively with CD4+CD45RBhi T cells (4 × 105 cells/mouse) plus CD4CD45RBlo T cells (4 × 105 cells/mouse). Disease activity was monitored weekly on the basis of body weight and soft stool or diarrhea. The mice were sacrificed 1, 2, 3, 4, and 6 weeks after T cell reconstitution.
Tissue Preparation and Mounting
After intraperitoneal anesthesia with 300 μL Nembutal per 100 g body weight, the colon and cecum (with the ileocecal junction) were removed, rinsed twice with ice-cold buffer, and immediately put into ice-cold modified Krebs-Ringer buffer solution (KRB, pH 7.4, NaCl 111.9, KCl 5.0, CaCl2 1.2, MgCl2 1.2, NaH2PO4 0.4, Na2HPO4 1.6, NaHCO3 25.0, Na glutamate 4.9, Na pyruvate 4.9, Na2 fumarate 5.4, glucose 11.5, β-hydroxy-butyrate 0.5 mM) that was continuously bubbled with carbogen O2/CO2 (95%/5%). The specimens were cut open along the mesenteric border with blunt-end scissors and the intestine was divided into 3 segments, representing ascending, transverse, and distal colon, respectively. From each segment a tissue sample was taken for histological analysis of inflammation and for mRNA analysis. The serosa and outer muscle layer of the remaining sample were stripped off in ice-cold oxygenated buffer and the mucosa was mounted in the Ussing diffusion chamber with a 0.07 cm2 exposed tissue area.
Ussing Chamber Experiments and AC Impedance Analysis
Each half-cell was continuously perfused with 37°C Krebs-Ringer Buffer, bathing the bowel segment on both the mucosal and serosal side. The temperature of Krebs-Ringer Buffer was kept constant at 37°C in a warm water bath. The Krebs-Ringer Buffer was continuously oxygenated with carbogen. In order to assess the pure epithelial function in the presence of several subepithelial layers between the electrodes in the Ussing chamber, we used the impedance analysis technique. This is essential as only the epithelial layer is involved in barrier function and as subepithelial resistance alters during inflammation.2, 19
Moreover, the reported transepithelial resistances of colonic tissue are relatively low and may even be lower than the resistance of the bathing solutions. This phenomenon leads to a markedly less accurate measurement of the resistance by means of direct current (DC) analysis and is avoided by the alternating current (AC) technique. The subepithelium analyzed in this study consisted of lamina propria, muscularis mucosae, and submucosa.
Electrophysiological measurements were done under transepithelial voltage clamp. The mean transepithelial potential was clamped at zero. Transepithelial impedance analysis was performed in the frequency range of 1 Hz to 22 kHz. Ninety voltage sine waves of 5 mV amplitude were simultaneously applied transepithelially in the frequency range 1–22 kHz and the corresponding current response was recorded. Impedance spectra were calculated by Fourier analysis. The impedance spectrum was plotted in a Nyquist diagram and revealed a single semicircle in all experiments. Therefore, the electrical properties of the stripped mucosa can be modeled as an electrical circuit that consists of a resistor–capacitor (RC) network composed of a resistor and a capacitor in parallel that represents the epithelium, and a resistor in series with this RC network representing the subepithelium and bathing solution. The capacitance (Cep) is formed by the lipid membranes of the apical and basolateral borders of the epithelial cells, while the epithelial resistance (Re) results form conductive pathways in the cellular and paracellular pathway. With such an electrical model, the AC impedance analysis technique permits to determine accurately the epithelial (Re) and the sum of the subepithelial (Rsub) resistance of the mucosa and the solution resistance (Rsol) that resides between the voltage sensing electrodes: Rs = Rsol + Rsub. The epithelial capacitance and resistance as well as Rs were determined by a fit of the Cole–Cole equation to the impedance data. On the basis of these data, the electrophysiological parameters were calculated. Rt = Rs + Re is obtained at low frequencies, whereas Rs is obtained at high frequencies. Therefore, determination of Rsub relies on an estimation of the solution resistance (Rsol) that was determined by impedance analysis of the empty chamber performed at the end of the experiment: Rsub = Rs − Rsol. Transepithelial currents (Isc) were corrected for clamping errors caused by Rs and offset potentials of the electrodes (Vofs). However, if Isc were not corrected for Rsub, true active transport would be underestimated.20 Therefore a correction factor Rt/Re is applied to calculate true Isc.
A potential difference of 1 mV or more was a prerequisite for inclusion of colon specimens. At 30–60 minutes after equilibration of the tissue in the chambers, the experiment was started by adding sodium fluorescein to the luminal side of the colon. Indomethacin was administered at a dose of 500 μM to the luminal side 1 hour after the experimental startpoint and all parameters were recorded for 1 hour.
To validate the setup for fluxes, flux measurements with sodium fluorescein were performed by luminal administration of different doses. A dose of 500 μM yielded reproducible and consistent values as measured by spectrofluorometry (λ excitation 485 nm, λ emission 515 nm). The 750-μL samples were taken at 0, 20, 40, 60, 80, 100, and 120 minutes. Transport of sodium fluorescein from the mucosal to serosal compartment was calculated as apparent permeability coefficient (Papp cm/s):
where ΔQ/Δt is the rate (nmol/s) of total transport of the compound in a stable flux period (between 20 and 60, and between 60 and 120 minutes), A is the surface area (0.07 cm2), and C0 is the initial donor concentration (nmol/mL).
Colon tissue samples were collected and fixed in phosphate-buffered saline (PBS) containing 6% formalin. Paraffin-embedded sections (5 μm) were stained with hematoxylin and eosin. The sections were analyzed without prior knowledge of the type of T cell reconstitution. Microscopic sections were graded by the number and severity of lesions. The degree of transmural inflammation (epithelium, submucosa, and muscularis) was calculated using a previously described scoring system.18
RNA Extraction and Quantitative Polymerase Chain Reaction (PCR)
Colonic tissue samples were obtained from SCID mice, either 1, 2, 4, or 6 week after T cell reconstitution as indicated as well as from nonreconstituted SCID mice. All samples were immediately frozen in liquid nitrogen after dissection and stored at −80°C until use. Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Leusden, The Netherlands) according to the manufacturer's instructions. A constant amount of 2 μg target RNA was used for cDNA synthesis (Ready-to-go-kit; Pharmacia, Uppsala, Sweden). After 90 minutes at 37°C, the reverse transcriptase was inactivated by incubating the cDNA samples for 5 minutes at 95°C. The cDNA samples were then subjected to real-time quantitative PCR, performed in the ABI prism 7700 sequence detector (Applied Biosystems, Foster City, CA) as described previously.21 The sequences of the primers and probes for IFN-γ, TNF-α, TGF-β, IL-10, and β-actin have been previously reported.21 All primers and probes were designed with the assistance of the computer program Primer Express (AB) and purchased from Eurogentec (Seraing, Belgium). The 5′-nuclease activity of the Taq polymerase was used to cleave a nonextendable dual-labeled fluorogenic probe. Fluorescent emission was measured continuously during the PCR reaction. PCR amplifications were performed in a total volume of 25 μL containing 5 μL cDNA, 12.5 μL Universal PCR Master Mix, no AmpErase UNG (AB), 100–300 nM concentrations of each primer, and 200 nM concentrations of the corresponding detection probe. Each PCR amplification was performed in triplicate wells using the following conditions: 94°C for 10 minutes, followed by 40 or 45 cycles at 94°C for 15 seconds and 60°C for 1 minute. cDNA plasmid standards, consisting of purified plasmid DNA specific for each individual target, were used to quantify the target gene in the unknown samples, as described.21 All results were normalized to β-actin to compensate for differences in the amount of cDNA in all samples.
Data are given as median and interquartile range (IQR). Standard error of the mean (SEM) was calculated based on the mean of the individual mice. Differences between groups were tested by the 2-tailed Student's t-test for unpaired and paired data or the nonparametric Mann–Whitney and Wilcoxon tests where appropriate. Correlations between epithelial function, histology score, and cytokines were calculated with the Spearman correlation test. P < 0.05 was considered significant. All tests were performed with SPSS 11.5 (Chicago, IL).
Total CD4+ Recipient SCID Mice Develop Transient Inflammation, Whereas CD4+CD45RBhi Recipients Develop Chronic Colitis
SCID mice were reconstituted with either total CD4+ T cells or CD4+CD45RBhi T cells and weighed every week. Upon reconstitution of CD4+CD45RBhi T cells, mice lost weight and developed colitis in contrast to total CD4+ reconstituted mice (Fig. 1A). We then performed a time course analysis of histologic scores and observed that 1 week after reconstitution both groups displayed no microscopic signs of colitis (Fig. 1B). At week 2 after reconstitution with CD4+CD45RBhi T cells, mice had colonic inflammation characterized by infiltration of large numbers of lymphocytes within the lamina propria and submucosa, prominent epithelial hyperplasia with loss of goblet cells, crypt abscesses, and ulcerations. At this same timepoint the colon of SCID mice reconstituted with total CD4+ T cell was infiltrated by lymphocytes within the lamina propria, although to a lesser extent, compared to CD4+CD45RBhi T cell recipients (P = 0.048) (Fig. 1B). At later timepoints, histology returned to normal in total CD4+ T cell reconstituted recipients, while inflammation and epithelial changes persisted in CD4+CD45RBhi T cell reconstituted SCID mice. Thus, total CD4+ T cell reconstituted SCID mice had transient inflammation, while chronic inflammation developed in CD4+CD45RBhi T cell reconstituted SCID mice.
Colonic Epithelial Function During Colitis Development
Epithelial resistance (Re), measured by AC impedance analysis, is a measure for passive ion movement across the epithelium (transcellular and paracellular pathways). Re was measured in colonic samples of Balb/c mice, nonreconstituted SCID mice, total CD4+ T cell, and CD4+CD45RBhi T cell reconstituted SCID mice throughout the development of the disease. Nonreconstituted SCID mice displayed a mean Re of 19.58 ± 1.00 Ω.cm2 which was not different from Re in Balb/c mice at any timepoint (21.22 ± 0.35 Ω.cm2) (Fig. 2, dotted line).
Before the development of histological inflammation, i.e., 1 week after reconstitution, total CD4+ T cell SCID recipients had a lower resistance than the Balb/c mice or CD4+CD45RBhi T cell SCID recipients (Fig. 2). From week 2 on, i.e., the timepoint of onset of inflammation, Re in the CD4+CD45RBhi T cell recipients also decreased compared to the values in Balb/c (P < 0.05). Interestingly, in both reconstituted groups Re was restored to the normal level at week 6, independently of resolution of inflammation.
Paracellular permeability for the water soluble small molecule sodium fluorescein was measured in all groups at week 1 and 6 after reconstitution. As shown in Figure 3, paracellular permeability was increased at week 6 in the CD4+CD45RBhi T cell reconstituted mice that had developed colitis, compared to CD4+ T cell reconstituted mice.
Rsub represents the resistance of the ssubepithelial layers consisting of lamina propria, muscularis mucosae, and submucosa. At baseline, Rsub did not differ between CD4 T cell reconstituted mice, nonreconstituted SCID mice, or Balb/c ones. In contrast, Rsub paralleled histological scores in CD4+CD45RBhi T cell or total CD4 T cell reconstituted SCID mice (Fig. 4). Rsub was increased in total CD4+ T cell reconstituted mice at week 2, the moment these mice had developed inflammation, and was normalized at later timepoints, when inflammation was no longer present. In contrast, CD4+CD45RBhi T cell recipients had a persistent increase in Rsub compared to Balb/c mice (P = 0.011) and to nonreconstituted SCID mice (P < 0.05) (Fig. 4).
Net ion transport measured by short circuit current (Isc) represents active ion secretion and absorption, which is the driving force for water movement in the colon. Isc started to decrease in the CD4+CD45RBhi T cell reconstituted mice from week 2 after reconstitution on, in contrast to all the other groups (Fig. 5). Active sodium reabsorption constitutes a significant portion of the measured short circuit current. Applying this to Figure 5, this means that there was clearly less Na+-absorption, which allows more intraluminal water accumulation and loss in the inflamed colonic epithelium. This net ion transport remained significantly lower in the colitic mice 4 and 6 weeks after reconstitution compared to all the other groups. In contrast, Isc did not decrease in the total CD4+ T cell recipients (Fig. 5).
Cytokine Expression and Correlations with Inflammation-induced Alterations of Epithelial Function
One week after reconstitution, increased mRNA levels of both regulatory/antiinflammatory (IL-10 and TGF-β) and proinflammatory cytokines (IFN-γ and TNF-α) were found in the mucosa of total CD4+ T cell and of CD4+CD45RBhi T cell reconstituted SCID mice (Fig. 6A–D) as compared with nonreconstituted mice (not shown). No histologic signs of colitis were yet present at that timepoint. At week 2, CD4+CD45RBhi T cell as well as the total CD4+ T cell reconstituted SCID mice displayed inflammatory changes in the colonic mucosa and a parallel increase in expression of the proinflammatory cytokines TNF-α. Four and 6 weeks after reconstitution, expression of the inflammatory cytokines was further enhanced in the CD4+CD45RBhi T cell reconstituted mice which had severe colitis, whereas whole CD4+ T cell reconstituted SCID mice displayed a relative increase in regulatory cytokines compared to the proinflammatory ones (Fig. 6). The histological scores correlated significantly with TNF-α (Fig. 7A) and IFN-γ (data not shown), supporting their role in the development of colitis.
In order to assess the relationship between the cytokines on the one hand and epithelial function on the other, we calculated the correlations between the cytokine mRNA levels and epithelial resistance and net ion transport. The levels of IL-10 and TGF-β mRNA correlated with epithelial resistance (Fig. 7B,C), suggesting that epithelial barrier properties were primarily maintained or restored by regulatory cytokines IL-10 and TGF-β. TNF-α mRNA was positively associated with decreased net ion transport, confirming the increase in water secretion under the influence of this proinflammatory cytokine (Fig. 7D).
Effect of Nonspecific Inhibition of COX-1 and -2 on Epithelial Function Is Enhanced in Inflamed Colonic Mucosa
Indomethacin (a nonspecific COX inhibitor) was added in the Ussing chamber and its effect on Re was measured in vitro. Indomethacin significantly decreased Re in colonic tissue from all groups: nonreconstituted, total CD4+ T cell, and CD4+CD45RBhi T cell reconstituted SCID mice. The effect of indomethacin was expressed as the ratio of indomethacin-induced Re/baseline Re (Table 1). The ratio indomethacin-induced Re/baseline Re correlated significantly with histological score in the colitic mice group (Spearman correlation coefficient r = −0.598, P = 0.0001). This is also apparent when comparing the effect of indomethacin on Re in CD4+CD45RBhi T cell reconstituted SCID mice at the moment they develop colitis (i.e., weeks 2, 4, and 6 postreconstitution) with the total CD4 T cell reconstituted mice without colitis (i.e., weeks 1, 4, and 6) and the non-T cell reconstituted SCID mice. Altogether these results show that exposure to indomethacin induces a significantly greater decrease of Re in inflamed tissues.
|Variable||Mice||Week 1||Week 2||Week 4||Week 6|
|Epithelial resistance (Re): ratio of indomethacin-induced Re/baseline Re||Nonreconstituted SCID mice||0.81 ± 0.02||0.90 ± 0.02||0.86 ± 0.02||0.80 ± 0.02|
|Total CD4+ reconstituted SCID mice||0.87 ± 0.01||0.87 ± 0.09||0.79 ± 0.05||0.72 ± 0.04|
|CD4+CD45RBhi reconstituted SCID mice||0.91 ± 0.01||0.77 ± 0.01||0.65 ± 0.03||0.69 ± 0.03|
The main objective of this study was to better understand the relationship between inflammation and the epithelial function in the gut. The T cell transfer model of colitis was chosen since no primary epithelial dysfunction is induced, in contrast to what occurs in several other models of colitis. Indeed, dextran sodium sulfate, trinitrobenzene sulphonic acid, oxazolone, and indomethacin all induce direct intestinal epithelial damage.
To analyze the effects of inflammation on the epithelial barrier function, we did a kinetic study of various parameters related to intestinal permeability in the colon of SCID mice that were nonreconstituted or reconstituted with total CD4+ T cells or CD4+CD45RBhi T cells. Segments of colon were harvested at different timepoints and mounted in the Ussing chambers. Mucosal barrier functions were assessed by 4 different characteristics: first, Re, which is a measure of passive transcellular and paracellular ion transport; second, Rsub, which is a measure of resistance of lamina propria, muscularis mucosae, and submucosa; third, paracellular permeability for small molecules (sodium fluorescein); and fourth, net ion transport, which is a measure of active sodium absorption in this tissue. We first observed that epithelial barrier functions in nonreconstituted SCID mice were not disturbed, as they were similar to that in Balb/c mice. This suggests that, in steady state conditions, T and B lymphocytes do not play a major role in the maintenance of a normal gut barrier function and supports the hypothesis that the epithelial gut barrier function rather depends on the innate immune system, which is intact in SCID mice.
Upon transfer of CD4+ CD45RBhi T cells, colonic inflammation develops and mice display colitis, which is progressive over time and correlated with enhanced levels of the proinflammatory cytokines TNF-α and IFN-γ. In contrast, upon total CD4+ T cells, a mild and transient colonic inflammation develops shortly after the cell transfer (week 2), which is subsequently resolved at later timepoints (weeks 4 and 6). We thus analyzed epithelial barrier functions before inflammation (week 1), at the onset of inflammation (week 2), and during sustained inflammation/after resolution (weeks 4 and 6).
We first observed a decreased Re in both total CD4+ and CD4+CD45RBhi T cell reconstituted animals, suggesting that the intestinal barrier is affected by inflammation. However, the decrease of Re was transient, especially in CD4+ T cells reconstituted mice, which display a normal Re at week 6. Interestingly, the decrease of Re appears before the onset of inflammation and suggests that it is not only a consequence of inflammation but also an early defect that might actually strengthen inflammation by increasing the amount of foreign antigens in the mucosa. Similar observations were reported in IL-10 KO mice, which display an increased ileal and colonic permeability before inflammation.22 However, since CD4+ T cell reconstituted mice can overcome this defect, it suggests that altered Re is not a critical defect that leads inevitably to overt inflammation and colitis. We found a significant correlation between Re and TGF-β and IL-10, suggesting that these cytokines are involved in the maintenance/recovery of intestinal barrier function. Indeed, IL-10 has been shown to play an important role in the regulation of intestinal epithelial integrity.16, 17, 23, 24 In addition, TGF-β possesses protective properties for epithelial barrier25 and suppresses inflammatory-dependent loss of barrier functions.26, 27 It remains to be determined whether IL-10 and TGF-β act directly on epithelial cells, or whether it is an indirect effect mediated by intermediate cells, such as regulatory T cells that globally downregulate inflammation and, among it, expression of TNF-α. Indeed, TNF-α has a proinflammatory effect on apoptosis of epithelial cells28 and modifies expression of claudin 2 in HT29 cells,29 both affecting epithelial barrier function. Indeed, we observe a correlation between TNF-α expression and net ion transport (Isc), which is in accordance with previous reports.30
In contrast to Re, Rsub was found to be significantly increased in CD4+CD45RBhi T cell reconstituted mice 6 weeks posttransfer. Moreover, Rsub of CD4 T cell reconstituted mice is transiently increased at week 2 (i.e., during inflammation). This apparent discrepancy might be explained by the increased cellularity present in the mucosa and submucosa upon overt inflammation. If Re reflects movement of ions through the epithelium, Rsub is related to the influx of cells and fluid in the mucosa and submucosa.
We observed that CD4+CD45RBhi T cell reconstituted mice displayed a significant decrease in net ion transport whenever inflammation was present. As active Na+ absorption constitutes a significant portion of Isc in colonocytes, this reveals a decrease in Na+ absorption during colitis. These changes contribute to the appearance of diarrhea, which results from the imbalance between intestinal absorption and secretion of water and electrolytes. Absorption requires the active transport of Na+ out of the epithelial cells mediated by the Na+/K+-ATPase located on the basolateral cell membranes. Interestingly, although both total CD4+ T cell and CD4+CD45RBhi T cell reconstituted mice had similar colonic inflammation and similar expression of proinflammatory cytokines at week 2, net ion transport was not altered (even tended to increase) in total CD4+ T cell reconstituted mice in contrast to the decreased net ion transport present in CD4+CD45RBhi T cell reconstituted mice. This suggests that the effects of inflammation on epithelial barrier function are complex and regulated by many different mediators. Nonetheless, we found a positive correlation between expression of TNF-α and the decrease in active cellular transport (i.e., sodium absorption). These observations are in agreement with data from Musch et al,13 who observed that T cell activation and TNF-treatment in a murine model significantly reduced basolateral Na+/K+-ATPase activity in jejunal mucosa, leading to less absorption.
Finally, we assessed the role of the cyclooxygenase pathway in this animal model by studying the effect of indomethacin, a nonspecific COX-1 and -2 inhibitor, on epithelial barrier function. We found that in vitro addition of indomethacin acutely induced a significant epithelial barrier decrease in all murine groups, including the nonreconstituted SCID mice. Thus, T cells were not required for this effect of COX-inhibition. Also, the rat model of indomethacin-induced bowel inflammation suggests that these proinflammatory T cells are not necessary for the initial mucosal damage induced by indomethacin.31 Several mechanisms may contribute to the T cell-independent mucosal effect of indomethacin, which depends partially on the presence of neutrophils but not of T or B lymphocytes in knockout mouse models.32 Apart from the immunological involvement of the COX family,33 prostaglandins play a key role in maintenance and repair of intestinal epithelial integrity and secretory and barrier function.34 We further found that the mucosal response to indomethacin was clearly enhanced in the presence of inflammation in the colonic mucosa, as was apparent in CD4+CD45RBhi T cell reconstituted mice, indicating that inflammation renders the epithelial cells more susceptible to dysfunction induced by nonselective COX-inhibitors. However, a direct correlation with proinflammatory cytokines could not be shown. Other factors present in the inflammatory tissue thus seem necessary for this effect, which is in accordance with data from Schmitz et al35 and Jacob et al.36
In conclusion, we showed that T cell-induced colitis in SCID mice results in epithelial dysfunction. We observed a decreased epithelial resistance before histological signs of inflammation that persisted during ongoing inflammation, together with an enhanced Rsub, enhanced paracellular permeability, and a decreased net ion transport. These changes may all contribute to diarrhea and further increase the development of inflammation. We found that epithelial resistance correlated with IL-10 and TGF-β levels, and net ion transport inversely correlated with TNF-α expression, pointing to the protective effect of IL-10 and TGF-β and the damaging effect of TNF. Finally, we observed that indomethacin, a nonspecific COX inhibitor, decreased Re independently of T cells, but that the effect of indomethacin was more pronounced in inflamed colon.
We conclude that induction of inflammation by transfer of CD4+CD45RBhi T cells in SCID mice leads to changes in the barrier functions of the colon, which probably participate to the exacerbation of the disease. Changes in epithelial resistance appear even before the development of inflammation, and thus suggest that barrier function disturbances are early events in the development of colitis. However, changes in mucosal barrier functions are not mandatory mechanisms for pathology in this model, since inflammation can resolve, despite these changes in barrier function. These data demonstrate that immune system and gut barrier function are closely interrelated.
We thank S. Decat for expert technical assistance in preparing histology slides.
- 30Regulation of mucosal structure and barrier function in rat colon exposed to tumor necrosis factor alpha and interferon gamma in vitro: A novel model for studying the pathomechanisms of inflammatory bowel disease cytokines. Scand J Gastroenterol. 2009: 1–10., , , et al.