Skip to main content
NCBI home page
Immunology logoLink to Immunology
. 2003 Feb;108(2):204–210. doi: 10.1046/j.1365-2567.2003.01577.x

Regulation of epithelial cell cytokine responses by the α3β1 integrin

Farah D Lubin 1, Miriam Segal 1, Dennis W McGee 1
PMCID: PMC1782878  PMID: 12562329

Abstract

Epithelial cells (EC) from various tissues can produce important cytokines and chemokines when stimulated by proinflammatory cytokines. These EC also receive signals from cell surface integrins, like the α3β1 integrin, which is important in cell migration and wound healing of epithelial monolayers. However, little is known of the effect of integrin signals on cytokine responses by EC. Colonic Caco-2 cells treated with an anti-α3 integrin antibody prior to stimulation with the proinflammatory cytokine interleukin (IL)-1 yielded suppressed levels of mRNA and secreted IL-6, IL-8 and monocyte chemoattractant protein-1 as compared to cells treated with normal mouse immunoglobulin G. Lung A549 cells also showed a similar suppression of cytokine secretion. Likewise, treatment of the Caco-2 cells with the same antibody suppressed tumour necrosis factor-α-stimulated IL-6 secretion. Fab fragments of the anti-α3 integrin antibody did not induce the suppressive effect but did block the suppressive effect of the whole antibody suggesting that the effect of the antibody required cross-linking of the integrins. Finally, culture of the Caco-2 cells on laminin type 5 (the major ligand for this integrin) yielded depressed levels of IL-1-induced IL-6 secretion as compared to cells on laminin type 1. These data are the first indication that the α3β1 integrin may cause a suppression of cytokine responses by EC which may be important in regulating the capacity of EC to respond during inflammation or wound healing.

Introduction

Epithelial cells (EC) have the capacity to secrete a wide range of cytokines which can regulate cell growth and immune or inflammatory responses. Intestinal EC have been shown to produce proinflammatory cytokines such as interleukin-6 (IL-6)1–4 and tumour necrosis factor-α (TNF-α)3 as well as several chemotactic cytokines, or chemokines, including IL-83,5 and monocyte chemoattractant protein-1 (MCP-1).3,6 Intestinal EC can be induced to produce these cytokines in response to proinflammatory cytokines, such as IL-1β or TNF-α produced by infiltrating macrophages, or in response to infection by a number of pathogenic bacteria (reviewed in 7). Likewise, EC from other tissues such as the lungs have been shown to produce similar cytokines.8,9 The capacity of EC to produce proinflammatory cytokines and chemokines makes these cells an important component of the mucosal immune and inflammatory response.

In addition to external cytokine signals, EC also respond to signals through their cell surface integrins which bind to extracellular matrix (ECM) proteins of the basement membrane. These integrins are heterodimers of α and β subunits,10 which have been shown to be important in a number of epithelial processes, such as cell differentiation11 and wound healing.12 However, the ECM protein components of the basement membrane may change due to tissue damage or proteolytic enzymes at sites of inflammation and altered matrix protein production during the process of wound healing.13–15 Such changes in the ECM content would therefore result in different integrin signals to the cell. One integrin which has been suggested as having an important role in wound healing of epithelial layers is the α3β1 integrin, which appears to be required for the migration of EC to close wounded epithelial monolayers in vitro.15,16 This α3β1 integrin along with its ligand, laminin-5 (LN-5), and the α6β4 integrin which also binds to LN-5, have been implicated as having a dominant regulatory influence on the function of other integrins and cell migration in wound healing.17

Yet little is known about the effect of integrin signals on the capacity of EC to produce cytokines important in the inflammatory response. Given the importance of the α3β1 integrin in wound healing, we have examined the effect of the α3β1 integrin on the capacity of EC to produce inflammatory cytokines in the presence or absence of an IL-1β or TNF-α stimulus. For these studies, two EC cell lines, the human colonic adenocarcinoma Caco-2 cell line and the human lung carcinoma A549 cell line, were selected as these cell lines have been commonly used in cytokine secretion experiments. To determine the role of the α3β1 integrin, the cells were treated with a monoclonal anti-α3 integrin antibody (clone P1B5) which has been shown to induce a tyrosine phosphorylation of the focal adhesion kinase (FAK)18,19 involved in integrin-associated signalling events.20 Our results suggest that cross-linking of the α3β1 integrin can down-regulate the response of EC to proinflammatory stimuli resulting in a depressed cytokine response. This novel regulatory effect of the α3β1 integrin may be a very important factor in the overall regulatory mechanisms controlling cytokine secretion by EC in inflamed, wounded and healing mucosal tissues.

Materials and methods

Coating of culture wells with ECM proteins

The bottom of 24-well plates were coated with 10 µg/well of mouse laminin (LN) from BD Biosciences (Bedford, MA), bovine plasma fibronectin (FN; Sigma Chemical Co., St. Louis, MO), or an affinity purified mouse LN from Chemicon International (Temecula, CA) in phosphate-buffered saline (PBS). The mouse LN from BD Biosciences was isolated from mouse Engelbreth–Holm–Swarm (EHS) tumour cell ECM and was determined by the suppliers to be greater than 90% laminin and of the type 1 form. After incubating at room temperature for 1 hr, any remaining material was aspirated and plates were air dried. In some experiments, the wells of 96-well plates were coated with 1·6 µg/well of human laminin type 5 (Chemicon) or the Chemicon affinity purified mouse LN in PBS. The plates were then treated as above.

Culturing of cells for cytokine secretion

The human colonic adenocarcinoma Caco-2 cells (ATCC HTB37; American Type Culture Collection, Rockville, MD) were maintained in RPMI-1640 (Mediatech, Washington DC) containing 10% fetal calf serum (FCS; Hyclone Laboratories, Logan UT), 10 mml-glutamine, 2 g/l sodium bicarbonate, 25 U/ml penicillin, 25 µg/ml streptomycin and 10 mm non-essential amino acids. The A549 human lung carcinoma cells (ATCC CCL 185) were maintained in Ham's F12K medium (F12K) with 2 mm l-glutamine (Mediatech) supplemented with 10% FCS (Hyclone), 1·5 g/l sodium bicarbonate, 25 U/ml penicillin and 25 µg/ml streptomycin (Sigma), designated as 10% FCS-F12K. The cells were removed from culture flasks by brief incubation with trypsin and ethylenediaminetetraacetic acid (EDTA; Sigma) and prepared in serum-free RPMI or F12K containing insulin, transferrin, and selenium (ITS; BD Biosciences) along with antibiotics, designated as ITS-RPMI or ITS-F12K. Removal of the cells using trypsin and EDTA appeared to have no effect on integrin expression as we and others21 have used this method for immunofluorescent staining of integrin subunits and trypsin treatment has been used in integrin functional studies using anti-integrin antibodies.19 The cells were then treated with 10 µg/ml of either a mouse anti-human α3-integrin antibody (clone P1B5; Life Technologies, Grand Island, NY), mouse immunoglobulin G (IgG; Sigma), rat anti-human α6-integrin antibody (clone GoH3; BD Biosciences/Pharmingen, San Diego, CA), or rat IgG (Sigma) for 30 min on ice before adding the cells at 2·5 × 105 cells/well to LN- (BD Biosciences) or FN-coated wells. In some experiments, cells were pretreated with 10 µg/ml of Fab fragments (see below) of the mouse anti-human α3-integrin antibody or Fab fragments of mouse IgG for 15 min on ice before the cells were washed and treated with the anti-α3 integrin antibody or IgG as above. Alternatively, untreated Caco-2 cells were added at 1 × 105 cells/well (96-well plates) or 2·5 × 105 cells/well (24-well plates) in ITS-RPMI to ECM protein-coated wells and incubated at 37°C. All cells were cultured at 37°C in a 90% air−10% CO2 humid atmosphere.

After 3 hr, some wells were stimulated with 1 ng/ml of recombinant human (rh) IL-1β or 50 ng/ml rhTNF-α (R & D Systems, Minneapolis, MN). These concentrations of IL-1β and TNF-α have previously been found to induce cytokine secretion by the Caco-2 cells.4 After 24 hr, the culture supernatants were collected and stored at −80°C for determination of the cytokine content. The cells were then removed with the trypsin-EDTA solution and the total cells/well were counted using a hemacytometer.

Preparation of Fab fragments

Fab fragments of the anti-α3 integrin antibody or mouse IgG were prepared by papain digestion and purified with a Protein A column using the Immunopure Fab kit (Pierce, Rockford, IL). The Fab fragments were then concentrated using a Centricon 10 000 MW Centrifugal Filter (Millipore Corporation, Bedford, MA). The purity of the Fab fragments was confirmed by non-reducing sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE).

Determination of secreted cytokine levels in the culture supernatants

The IL-6 and IL-8 content in the culture supernatants were determined using the DuoSet enzyme-linked immunosorbent assay (ELISA) development systems for human IL-6 and IL-8 (R & D Systems). MCP-1 content was determined using the Cytoscreen Immunoassay kit specific for human MCP-1 (Biosource International, Camarillo, CA). The absorbances of the samples were measured using a Bio-Tek EL312e microplate reader (Winooski, VT).

Determination of mRNA levels by reverse transcriptase–polymerase chain reaction (RT–PCR)

The Caco-2 cells (5 × 105 cells/well) were pretreated with the antibodies and cultured in 12-well plates as described above. Total RNA was then extracted using TRIZOL (Life Technologies). Samples were reverse transcribed and PCR amplified using the Gene Amp RNA PCR Kit (Perkin-Elmer, Norwalk, VT) as previously reported2 for 22–25 cycles. Primers specific for human IL-6, IL-8 and glyceraldehyde-3-phosphate dehydrogenase (G3PDH) were obtained from Clontech (Palo Alto, CA) or human MCP-1 from Biosource International. The PCR product band densities on agarose gels were quantitated from photographic negatives using a UVP Gel Documentation system (Upland, CA) and the Labworks 3·02 Analysis sofware (Media Cybernetics, Silver Spring, MD).

Statistics

The results are shown as representative experiments from two to three experiments and the values are reported as the means of triplicate wells for each condition ±SD. Representative experiments are shown as some experiments were repeated less than three times. Statistical significance was determined by Student's t-test.

Results

Role of the α3β1 integrin in cytokine secretion by Caco-2 and A549 cells

Previous studies have shown that the Caco-2 cells express the α3 and α6 integrin subunits21 and using immunofluorescence, we have found them to express the α3, β1 and β4 integrin subunits (unpublished results). Therefore, these cells should be capable of expressing the α3β1, α6β1 and α6β4 integrins. As noted before, the α3β1 and α6β4 integrins appear to be important in the regulation of cell migration and wound healing.17

As shown in Table 1, treatment of the Caco-2 cells with the P1B5 anti-α3 integrin antibody resulted in some suppression of basal, unstimulated levels of IL-6 and IL-8 secretion. More importantly, the antibody treatment yielded a suppression of IL-1β stimulated IL-6 (76%) and IL-8 (60%) secretion by the cells as compared to cells treated with normal mouse IgG. In a single experiment, both unstimulated and IL-1β stimulated MCP-1 levels were also significantly suppressed with anti-α3 integrin antibody-treated cells (3 ± 1 pg/105 cells for unstimulated and 67 ± 8 pg/105 cells for IL-1β stimulated) as compared to that of the IgG-treated cultures (25 ± 6 pg/105 cells for unstimulated and 164 ± 24 pg/105 cells for IL-1β stimulated; P < 0·01). Treatment of the cells with the anti-α3 antibody resulted in no noticeable effect on their adherence to the LN as compared to the IgG-treated cells and the viability of the cells also remained high. Still, all values were adjusted for the total cell numbers recovered to account for any differences.

Table 1.

Effect of anti-α3 integrin antibody binding on cytokine secretion by IL-1β-stimulated Caco-2 cells cultured on LN

Cytokine secreted (pg/105 cells)
Treatment* IL-6 IL-8
Mouse IgG 6 ± 1 50 ± 8
Mouse IgG + IL-1β 121 ± 22 675 ± 111
Anti-α3 antibody 2 ± 1† 10 ± 2†
Anti-α3 antibody + IL-1β 29 ± 5† 273 ± 43†
Open in a new tab
*

Caco-2 cells were treated with the anti-α3 integrin antibody (P1B5) or normal mouse IgG (10 µg/ml) in ITS-RPMI for 30 min and added to LN-coated wells at 2·5 × 105 cells/well. After 3 hr, rhIL-1β (1 ng/ml) was added. Supernatants were harvested at 24 hr for determination of cytokine levels and counting cells. Values are the means of triplicate wells ± SD for each condition. This experiment is representative of three separate experiments.

†

Significant difference from IgG-treated cultures (P < 0·01).

We next determined the effect of the anti-α3 integrin antibody treatment on Caco-2 cells cultured in FN-coated wells. Treatment of the cells with the anti-α3 integrin antibody and subsequent culture on FN also resulted in a significant suppression of both unstimulated (74%) and IL-1β stimulated (53%) IL-6 secretion (Fig. 1). These results suggest that the α3β1 integrin may play an important role in cytokine responses by EC.

Figure 1.

Figure 1

Open in a new tab

Effect of the anti-α3 antibody on IL-6 secretion by Caco-2 cells cultured on FN. Caco-2 cells were treated with the P1B5 anti-α3 integrin antibody or mouse IgG as described in Table 1 and cultured in FN coated plates. After 3 hr, the cells were stimulated with IL-1β (1 ng/ml) and at 24 hr the culture supernatants were collected and cells counted as before. Values are the means ± SD for triplicate wells from a representative experiment of three separate experiments. *Significant difference from cells treated with the normal mouse IgG (P < 0·05).

The A549 cells are also known to express the α3, α6, β1, and β4 integrin subunits,22 and therefore, should be capable of expressing the α3β1 and α6β4 integrins (and presumably the α6β1 integrin) as well. When the effect of the anti-α3 integrin antibody on cytokine responses by the A549 cell line was also tested, similar results as seen with the Caco-2 cells were obtained. Treatment of the A549 cells with the anti-α3 integrin antibody resulted in a suppression of unstimulated MCP-1 (41%) and IL-8 (43%) secretion levels and IL-1β stimulated IL-6 (48%), MCP-1 (40%), and IL-8 (43%) secretion levels as compared to cells treated with a normal mouse IgG (Table 2). These results taken together with those of the Caco-2 cells indicate that the α3β1 integrin may play an important regulatory role in proinflammatory cytokine responses by EC.

Table 2.

Effect of anti-α3 integrin antibody binding on cytokine secretion by IL-1β-stimulated A549 cells cultured on LN

Cytokine secreted (pg/105 cells)
Treatment* IL-6 MCP-1 IL-8
Mouse IgG 18 ± 1 213 ± 12 616 ± 67
Mouse IgG + IL-1β 434 ± 16 644 ± 21 4635 ± 98
Anti-α3 antibody 9 ± 1 126 ± 6† 353 ± 49†
Anti-α3 antibody + IL-1β 277 ± 15† 384 ± 30† 2653 ± 63†
Open in a new tab
*

A549 cells were treated with the anti-α3 integrin antibody (P1B5) or normal mouse IgG for 30 min and added to LN coated wells as in Table 1. After 3 hr, the cells were then stimulated with rhIL-1β as appropriate and incubated for 24 hr. The culture supernatants and cells were then collected as before. Values are the means of triplicate wells for each condition ± SD. This experiment is representative of three experiments.

†

Significant difference from IgG-treated cultures (P < 0·005).

The effect of treating the Caco-2 cells with a monoclonal rat antibody to the α6 integrin subunit (clone GoH3; Pharmingen/BD) was also tested to determine if integrins containing this subunit (α6β1 and α6β4 which bind LN) could affect EC cytokine responses. Pretreatment of the Caco-2 cells with the anti-α6 integrin antibody and subsequent culture in FN-coated wells had no effect on unstimulated or IL-1β-stimulated IL-6 and IL-8 secretion as compared to cells treated with normal rat IgG (data not shown). A culture well coating of FN was used as the α6-containing integrins are not known to bind to FN and the antibody treated cells readily attached to the FN, whereas anti-α6 integrin antibody treatment was found to prevent attachment of the cells to LN coated wells.

The anti-α3 integrin antibody effect requires antibody cross-linking

The P1B5 anti-α3 integrin antibody is known to block the binding of the α3β1 integrin to ECM proteins, but other reports have also used this antibody to examine intracellular signalling pathways and morphological changes induced by binding of the antibody to the α3β1 integrin.18,19,23 In order to determine if the suppressive effect of the anti-α3 integrin antibody treatment was a result of simply blocking the binding of the α3β1 integrin to its ligand, Fab fragments of the antibody were used. Treatment of the Caco-2 cells with Fab fragments of the anti-α3 integrin antibody resulted in no suppression of IL-1β-stimulated IL-6 secretion as compared to cells treated with Fab fragments of normal mouse IgG (Fig. 2). However, pretreatment of the Caco-2 cells with Fab fragments of the anti-α3 integrin antibody before treatment of the cells with the whole anti-α3 integrin antibody blocked the suppressive action of the anti-α3 integrin antibody. For comparison, pretreatment of the Caco-2 cells with Fab fragments of normal mouse IgG, which should not bind to the α3β1 integrin, before treatment of the cells with the whole anti-α3 integrin antibody resulted in suppressed levels of IL-1 stimulated IL-6 secretion. This suggests that the Fab fragments of the anti-α3 integrin antibody were able to bind to the Caco-2 cells and block binding of the whole anti-α3 integrin antibody, thus preventing the suppressive effect of the anti-α3 integrin antibody treatment. Therefore, the suppressive effect was probably not caused simply by blocking the binding of the α3β1 integrin to the ECM protein, but was an active effect because of antibody binding, which required cross-linking of the α3 integrin.

Figure 2.

Figure 2

Open in a new tab

Effect of anti-α3 integrin antibody Fab fragments on IL-6 secretion by Caco-2 cells cultured on FN. Caco-2 cells were treated with or without 10 µg/ml of Fab fragments of the P1B5 anti-α3 integrin antibody or Fab fragments of normal mouse IgG for 30 min in ITS-RPMI. The cells were then washed and treated with or without 10 µg/ml of the P1B5 anti-α3 integrin antibody or normal mouse IgG for 15 min as appropriate. The cells were then added at 2·5 × 105 cells/well to FN-coated wells. After 3 hr, the cells were stimulated with 1 ng/ml rhIL-1β as indicated and then cultured for 24 hr. The culture supernatants were then collected and the cells counted as before. *Similar values which show a significant difference from all other values (P < 0·001).

Effect of LN isoforms on IL-6 secretion by Caco-2 cells

Up to this point, the experiments have focused on the α3β1 integrin. Because the major ligand for the α3β1 integrin is LN-5,24,25 we next investigated the effect of LN isoforms on IL-6 secretion. First, culture wells were coated with an affinity purified LN isolated from the EHS mouse tumour ECM (Chemicon). The major isoform of LN produced by the EHS mouse tumour cells is LN-1. For comparison, wells were coated with FN as the α3 and α6 containing integrins do not bind (or bind only weakly as with the α3β1 integrin) to this ECM protein. Uncoated plastic was not used as a control as this may have allowed cell binding interactions that were non-integrin mediated, although we have found that the levels of IL-1β-stimulated IL-6 secreted by Caco-2 cells cultured on FN were not significantly different from that of cells cultured on plastic (data not shown). Culture of the Caco-2 cells on the LN-1 containing LN preparation yielded IL-1β-stimulated IL-6 levels that were not significantly different from cells cultured on FN (Fig. 3a). However, culture of the Caco-2 cells on a purified human LN-5 preparation resulted in IL-1β stimulated IL-6 levels that were only 51% of the levels of IL-6 secreted by the cells cultured on the LN-1-containing LN preparation (Fig. 3b). Because LN-5 is known to be the major ligand for the α3β1 integrin,24,25 whereas both the α6β1 and α6β4 integrins bind well to both LN-5 and LN-1, these results provide additional support for our finding that antibody cross-linking of the α3β1 integrin can provide a signal to the cell, which results in a suppression of EC cytokine responses.

Figure 3.

Figure 3

Open in a new tab

Secretion of IL-6 by Caco-2 cells grown on FN or LN preparations. (a) Caco-2 cells at 2·5 × 105 cells/well in ITS-RPMI were cultured in wells coated with FN or an affinity purified LN extracted from EHS cell cultures. After 3 hr, rhIL-1β (1 ng/ml) was added. (b) Caco-2 cells (1 × 105 cells/well) in ITS-RPMI were cultured in wells coated with an affinity purified LN extracted from EHS cell cultures or human LN-5 and after 3 hr, IL-1β (1 ng/ml) was added. In both experiments, the culture supernatants were harvested at 24 hr for determination of IL-6 levels and cell counts as before. Values shown are the means ± SD for triplicate wells and are representative of (a) three or (b) two separate experiments. *Significant difference in IL-6 secretion levels (P < 0·05).

Effect of the anti-α3 integrin antibody on cytokine secretion by TNF-α-stimulated cells

The IL-1 and TNF receptor signalling pathways that result in the activation of cytokine genes are known to share several components.26 Therefore, we next determined the effect of treating the Caco-2 cells with the anti-α3 integrin antibody on TNF-α-stimulated cytokine responses. Caco-2 cells were treated with the P1B5 anti-α3 integrin antibody or normal mouse IgG and added to LN-coated wells as in Table 1. After 3 hr, rhTNF-α (50 ng/ml) was added to the appropriate cells and the culture supernatants were collected at 24 hr and cells counted as before. Treating the cells with the anti-α3 integrin antibody also resulted in a suppression of IL-6 secretion by unstimulated cells (7 ± 1 pg/105 cells for mouse IgG-treated cells compared to 2 ± 1 pg/105 cells for anti-α3 integrin antibody-treated cells; P ≤ 0·01) and IL-1β stimulated cells (123 ± 8 pg/105 cells compared to 82 ± 4 pg/105 cells; P ≤ 0·01; representative of three separate experiments), the latter of which was decreased by 33%. This suggests that the binding of the anti-α3 integrin antibody may induce an effect common to other cytokine signalling pathways.

Effect of the anti-α3 integrin antibody treatment on cytokine mRNA levels

Using semiquantitative RT–PCR, the effect of the anti-α3 integrin antibody on cytokine mRNA expression by the Caco-2 cells was determined. As shown in Fig. 4, treatment of the cells with the anti-α3 integrin antibody suggested a suppression in the levels of both unstimulated and IL-1β-stimulated cytokine mRNA levels as compared to cells treated with normal mouse IgG. Densitometric analysis of the 0·125 µg RNA bands suggested a decrease in MCP-1 (82%), IL-8 (47%), and IL-6 (88%) mRNA levels for the unstimulated anti-α3 integrin antibody treated samples. Densitometric analysis of the RNA bands from the IL-1 stimulated cells suggested a decrease in MCP-1 (70%) and IL-6 (55%) (both for the 0·25 µg samples) and IL-8 (63%) (for the 0·125 µg sample) mRNA levels for the IL-1 stimulated anti-α3 integrin antibody-treated samples. These results support our finding that the anti-α3 integrin antibody treatment suppressed IL-6, IL-8 and MCP-1 responses.

Figure 4.

Figure 4

Open in a new tab

RT–PCR analysis of cytokine mRNA levels from Caco-2 cells treated with the anti-α3 integrin antibody and cultured on LN. Caco-2 cells were pretreated for 30 min with the anti-α3 integrin antibody and then cultured as in Table 1. At 3 hr, rhIL-1β was added to the appropriate cultures. Four hours after IL-1 stimulation, the cells were harvested for total RNA isolation and reverse transcription of (a) 0·25 µg or (b) 0·125 µg RNA samples. The samples were then PCR amplified from the same RT samples. The unstimulated and IL-1-stimulated samples were PCR amplified for different numbers of cycles and should not be directly compared. This figure is representative of three separate experiments.

Discussion

Epithelial cells are now known to play an important role in mucosal immune responses by producing regulatory cytokines in response to proinflammatory cytokines and infectious agents. Because of its importance in wound healing, we have investigated the role of the α3β1 integrin on cytokine responses by EC as these cells may also be responding to inflammatory signals in addition to wound healing in vivo. Our studies have shown that treatment of the Caco-2 and A549 cell lines with an antibody to the α3 integrin subunit resulted in a suppression of IL-1 stimulated IL-6, IL-8 and MCP-1 secretion. Furthermore, the suppressive effect of the anti-α3 integrin antibody required cross-linking of the integrin receptor, suggesting that the effect of the antibody was to provide a signal to the cell. Indeed, others have shown that treatment of fibroblasts with the same P1B5 anti-α3 integrin antibody could induce cells to form lammellipodia which were similar to those obtained by culturing the cells on LN-5.23 Also, treatment of melanoma cells18 and even Caco-2 cells19 with the P1B5 anti-α3 integrin antibody has been shown to induce tyrosine phosphorylation of the focal adhesion kinase (FAK) associated with integrin signalling. Finally, Caco-2 cells cultured on purified LN-5 produced lower levels of IL-1 stimulated IL-6 as compared to cells cultured on LN-1 containing ECM proteins, to which the other LN-5 binding integrin (the α6β4 integrin) binds as well. Taken together, we now suggest that binding of the α3β1 integrin to its ligand LN-5 or cross-linking this integrin with the anti-α3 integrin antibody provides a signal to the cell that results in the suppression of IL-1-stimulated cytokine responses.

The results from RT–PCR analysis of cytokine mRNA levels suggested that binding of the anti-α3 integrin antibody also led to depressed IL-1-stimulated MCP-1, IL-6 and IL-8 mRNA levels, which corroborates the effect on the secretion of these cytokines. It is interesting that the basal unstimulated levels of cytokine mRNA and secreted cytokine levels were also suppressed in the anti-α3 integrin antibody-treated cells. These results indicate that one effect mediated the α3β1 integrin signal may have been to down-regulate basal and IL-1-stimulated cytokine mRNA expression.

Treatment of the Caco-2 cells with the anti-α3 integrin antibody also had a similar effect on TNF-α-stimulated IL-6 secretion. The intracellular signalling pathway associated with TNF and IL-1 stimulation share some common components such as activation of the transcription factor nuclear factor (NF)-κB via the activation of the IκB kinase (IKK) and phosphorylation of the inhibitor IκBα.26 Indeed, NF-κB is a major transcription factor involved in the activation of many proinflammatory response genes, including cytokine genes for IL-6, IL-8 and MCP-1.27 Perhaps binding of the α3β1 integrin may ultimately affect the activation of this pathway common to IL-1 and TNF stimulation; however, further study is needed to address this hypothesis.

In the intestine, expression of the α3β1 integrin receptor and LN-5 coincide in the villus region of the crypt–villus axis.28,29 However, the deposition of FN and LN isoforms may change during inflammation and wound healing due to injury or enzymatic destruction of the ECM.13–15 In a recent study, the expression of LN-5 was found to be altered in the small intestine of Crohn's disease (CD) patients.30 LN-5 was found to be expressed in both the villus and crypt regions of inflamed CD specimens whereas LN-2, normally found only in the crypt region, was shown to be absent. This confirms that the distribution of LN isoforms can change in diseased mucosal tissues and suggests that LN-5 may be an important factor in the disease process of CD. Based on our findings, such a change in the types of ECM proteins resulting from inflammation or wound healing could affect the capacity of EC to produce proinflammatory cytokines.

Recently, Hodivala-Dilke and coworkers31 using mice deficient in the α3 integrin subunit have proposed a skin wound healing model where wounding would result in the activation of proteases which would allow a degradation of LN-5 in the normal basement membrane. This would lead to a decreased binding of the α3β1 integrin, which would then allow for activation of FN and collagen binding integrins resulting in cell migration and wound healing. Repair of the basement membrane and LN-5 replacement would then initiate α3β1 integrin binding and inactivation of the other integrin receptors to inhibit cell migration. The results from our study would fit well with a mucosal model like this in which the destruction of LN-5 during wounding and inflammation would allow for the production of elevated levels of proinflammatory cytokines by EC which would be needed for a localized inflammatory response. Then, as the wound is healed and LN-5 is replaced, binding of the α3β1 integrin on EC to LN-5 would down-regulate the capacity of the cells to produce these proinflammatory cytokines. However, it remains to be determined whether EC attached to the different combinations of ECM proteins associated with normal, diseased and healing tissues actually do produce altered levels of cytokines in vivo.

The results described in this report provide the first indication that EC proinflammatory cytokine responses may be regulated by signals from the α3β1 integrin. This could indicate that EC may be able to detect changes in the make-up of the ECM through integrin signals and may represent a mechanism to allow a greater control of the levels of inflammatory cytokines secreted by EC in inflamed or healing tissues. A further examination of the signalling pathways involved in this effect may provide important insight into a novel mechanism for regulating EC proinflammatory cytokine responses.

Acknowledgments

The authors would like to thank Dr Tabbi Miller, Mate Stulic, Christian Goess, Jessica Wong, and Shari Hanifin for their technical assistance. This work was supported by US PHS Grant DK 54049. The work of M.S. was also supported by an undergraduate education grant from the Howard Hughes Medical Institute to Binghamton University.

References

  • 1.McGee DW, Beagley DW, Aicher WK, McGhee JR. Transforming growth factor-β enhances interleukin-6 secretion by intestinal epithelial cells. Immunology. 1992;77:7–12. [PMC free article] [PubMed] [Google Scholar]
  • 2.McGee DW, Beagley DW, Aicher WK, McGhee JR. Transforming growth factor-β and IL-1β act in synergy to enhance IL-6 secretion by the intestinal epithelial cell line, IEC-6. J Immunol. 1993;151:970–8. [PubMed] [Google Scholar]
  • 3.Jung HC, Eckmann L, Yang SK, Panja A, Fierer J, Morzycka-Wroblewska E, Kagnoff MF. A distinct array of proinflammatory cytokines is expressed in human colon epithelial cells in response to bacterial invasion. J Clin Invest. 1995;95:55–65. doi: 10.1172/JCI117676. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Vitkus SJD, Hanifin SA, McGee DW. Factors affecting Caco-2 intestinal epithelial cell interleukin-6 secretion. In Vitro Cell Dev Biol. 1998;34:660–4. doi: 10.1007/s11626-996-0017-7. [DOI] [PubMed] [Google Scholar]
  • 5.Eckmann L, Kagnoff MF, Fierer J. Epithelial cells secrete the chemokine interleukin-8 in response to bacterial entry. Infect Immun. 1993;61:4569–74. doi: 10.1128/iai.61.11.4569-4574.1993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Reinecker HC, Loh EY, Ringler DJ, Mehta A, Rombeau JL, MacDermott RP. Monocyte-chemoattractant protein-1 gene expression in intestinal epithelial cells and inflammatory bowel disease. Gastroenterology. 1995;108:40–5. doi: 10.1016/0016-5085(95)90006-3. [DOI] [PubMed] [Google Scholar]
  • 7.McGee D. Inflammation and mucosal cytokine production. In: Ogra PL, Lamm ME, Bienenstock J, Mestecky J, Strober W, McGhee JR, editors. Mucosal Immunology. 2. San Diego: Academic Press; 1999. pp. 559–73. [Google Scholar]
  • 8.Lin Y, Zhang M, Barnes PF. Chemokine production by a human alveolar epithelial cell line in response to Mycobacterium tuberculosis. Infect Immun. 1998;66:1121–6. doi: 10.1128/iai.66.3.1121-1126.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Borger P, Koeter GH, Timmerman JA, Vellenga E, Tomee JF, Kauffman HF. Proteases from Aspergillus fumugatus induce interleukin (IL)-6 and IL-8 production in airway epithelial cell lines by transcriptional mechanisms. J Infect Dis. 1999;180:1267–74. doi: 10.1086/315027. [DOI] [PubMed] [Google Scholar]
  • 10.Beaulieu J-F. Integrins and human intestinal cell functions. Front Biosci. 1999;4:310–21. doi: 10.2741/beaulieu. [DOI] [PubMed] [Google Scholar]
  • 11.Streuli C. Extracellular matrix remodeling and cellular differentiation. Curr Opin Cell Biol. 1999;11:634–40. doi: 10.1016/s0955-0674(99)00026-5. [DOI] [PubMed] [Google Scholar]
  • 12.Raghow R. The role of extracellular matrix in postinflammatory wound healing and fibrosis. FASEB J. 1994;8:823–31. doi: 10.1096/fasebj.8.11.8070631. [DOI] [PubMed] [Google Scholar]
  • 13.Goke M, Zuk A, Podolsky DK. Regulation and function of extracellular matrix in intestinal epithelial restitution in vitro. Am J Physiol. 1996;271:G729–40. doi: 10.1152/ajpgi.1996.271.5.G729. [DOI] [PubMed] [Google Scholar]
  • 14.Saarialho-Kere UK, Vaalamo M, Puolakkainen P, Airola K, Parks WC, Karjalainen-Lindsberg M-L. Enhanced expression of matrilysin, collagenase, and stromelysin-1 in gastrointestinal ulcers. Am J Pathol. 1996;148:519–26. [PMC free article] [PubMed] [Google Scholar]
  • 15.Lotz MM, Nusrat A, Madara JL, Ezzell R, Wewer UM, Mercurio AM. Intestinal epithelial restitution. Involvement of specific laminin isoforms and integrin laminin receptors in wound closure of a transformed model epithelium. Am J Pathol. 1997;150:747–60. [PMC free article] [PubMed] [Google Scholar]
  • 16.Goldfinger LE, Hopkinson SB, deHart CW, Gollawn S, Couchman JR, Jones JCR. The α3 laminin subunit, α6β4 and α3β1 integrin coordinately regulate wound healing in cultured epithelial cells and in the skin. J Cell Sci. 1999;112:2615–29. doi: 10.1242/jcs.112.16.2615. [DOI] [PubMed] [Google Scholar]
  • 17.Nguyen BP, Ryan MC, Gil SG, Carter WG. Deposition of laminin 5 in epidermal wounds regulates integrin signaling and adhesion. Curr Opin Cell Biol. 2000;12:554–62. doi: 10.1016/s0955-0674(00)00131-9. [DOI] [PubMed] [Google Scholar]
  • 18.Lauer JL, Gendron CM, Fields GB. Effect of ligand conformation on melanoma cell α3β1 integrin-mediated signal transduction events: implications for a collagen structural modulation mechanism of tumor cell invasion. Biochemistry. 1998;37:5279–87. doi: 10.1021/bi972958l. [DOI] [PubMed] [Google Scholar]
  • 19.Basson MD, Emenaker NJ, Sanders MA. Alpha integrin subunits regulate human (Caco-2) intestinal epithelial proliferation and phenotype. Cell Physiol Biochem. 2000;10:27–36. doi: 10.1159/000016332. [DOI] [PubMed] [Google Scholar]
  • 20.Giancotti FG, Ruoslahti E. Integrin signaling. Science. 1999;285:1028–32. doi: 10.1126/science.285.5430.1028. [DOI] [PubMed] [Google Scholar]
  • 21.Ebert EC. Mechanisms of colon cancer binding to sustratum and cells. Dig Dis Sci. 1996;41:1551–6. doi: 10.1007/BF02087899. [DOI] [PubMed] [Google Scholar]
  • 22.Falcioni R, Cimino L, Gentileschi MP, D'Agnano I, Zupi G, Kennel SJ, Sacchi A. Expression of β1, β3, β4, and β5 integrins by human lung carcinoma cells of different histotypes. Exp Cell Res. 1994;210:113–22. doi: 10.1006/excr.1994.1017. [DOI] [PubMed] [Google Scholar]
  • 23.Dogic D, Rousselle P, Aumailley M. Cell adhesion to laminin 1 or 5 induces isoform-specific clustering of integrins and other focal adhesion components. J Cell Sci. 1998;111:793–802. doi: 10.1242/jcs.111.6.793. [DOI] [PubMed] [Google Scholar]
  • 24.Carter WG, Ryan MC, Gahr PJ. Epiligrin, a new cell adhesion ligand for integrin α3β1 in epithelial basement membranes. Cell. 1991;65:599–610. doi: 10.1016/0092-8674(91)90092-d. [DOI] [PubMed] [Google Scholar]
  • 25.Delwel GO, de Melker AA, Hogervorst F, et al. Distinct and overlapping ligand specificities of the α3Aβ1 and α6Aβ1 integrins: recognition of laminin isoforms. Mol Biol Cell. 1994;5:203–15. doi: 10.1091/mbc.5.2.203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Baud V, Karin M. Signal transduction by tumor necrosis factor and its relatives. Trends Cell Biol. 2001;11:372–7. doi: 10.1016/s0962-8924(01)02064-5. [DOI] [PubMed] [Google Scholar]
  • 27.Baeuerle P, Henkel T. Function and activation of NF-κB in the immune system. Annu Rev Immunol. 1994;12:141–79. doi: 10.1146/annurev.iy.12.040194.001041. [DOI] [PubMed] [Google Scholar]
  • 28.Beaulieu J-F. Differential expression of the VLA family of integrins along the crypt–villus axis in the human small intestine. J Cell Sci. 1992;10:427–36. doi: 10.1242/jcs.102.3.427. [DOI] [PubMed] [Google Scholar]
  • 29.Leivo I, Tani T, Laitinen L, Bruns R, Kivilaakso E, Lehto VP, Burgeson RE, Virtanen I. Anchoring complex components laminin-5 and type VII collagen in the intestine: association with migration and differentiating enterocytes. J Histochem Cytochem. 1996;44:1267–77. doi: 10.1177/44.11.8918902. [DOI] [PubMed] [Google Scholar]
  • 30.Bouatrouss Y, Herring-Gillam FE, Gosselin J, Poisson J, Beaulieu J-F. Altered expression of laminins in Crohn's disease small intestinal mucosa. Am J Pathol. 2000;156:45–50. doi: 10.1016/S0002-9440(10)64704-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Hodivala-Dilke KM, DiPersio CM, Kreidberg JA, Hynes RO. Novel roles for α3β1 integrin as a regulator of cytoskeletal assembly and as a trans-dominant inhibitor of integrin receptor function in mouse keratinocytes. J Cell Biol. 1998;142:1357–69. doi: 10.1083/jcb.142.5.1357. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Immunology are provided here courtesy of British Society for Immunology

RESOURCES