Abstract
Mucosal mast cells (MMC) or their precursors migrate through the intestinal lamina propria to reside intraepithelially, where expression of mouse mast cell protease-1 indicates the mature phenotype. Alterations in expression of integrins that govern cell adhesion to the extracellular matrix may regulate this process. As the key cytokine mediating differentiation of mouse mast cell protease-1-expressing MMC homologues in vitro, TGF-β1 was considered a likely candidate for regulation of the integrins that facilitate intraepithelial migration of MMC. Therefore, we examined adhesion of bone marrow-derived mast cells cultured with and without TGF-β1 to laminin-1, fibronectin, and vitronectin along with expression of integrins likely to regulate this adhesion. Adhesion of PMA-stimulated cultured mast cells to laminin-1 increased from 5.3 ± 3.6% (mean ± SEM) in the absence of TGF-β1 to 58.7 ± 4.0% (p < 0.05) when cultured mast cells had differentiated into MMC homologues in the presence of TGF-β1. Increased adhesion of MMC homologues to laminin-1 was also stimulated by FcεRI cross-linking and the calcium ionophore A23187. Expression of the laminin-binding integrin α7 by MMC homologues grown in the presence of TGF-β1 was demonstrated by RT-PCR and flow cytometry, and preincubation of MMC homologues with the α7-neutralizing Ab 6A11 inhibited adhesion to laminin-1 by 98% (p < 0.05), demonstrating a novel role for this molecule in adhesion of a hemopoietic cell to laminin-1.
Cell trafficking into and through tissues depends upon complex interactions, including adhesion to extracellular matrix (ECM)4 proteins (1); integrins are the major molecules that mediate this process (2). Mast cells are involved in the pathophysiology of immediate-type hypersensitivity reactions, wound healing, and chronic inflammatory diseases (3, 4). The ability to move through tissues is mandatory for their physiological function, and integrins are likely to be involved in mast cell migration.
The roles of several integrins and other adhesion molecules in mast cell adhesion to ECM proteins has been studied using murine IL-3-dependent, bone marrow-derived cultured mast cells (CMC). These cells adhere to laminin-1 and -2 (5, 6, 7, 8), fibronectin (9), and vitronectin (10), and integrins α6β1 (5, 8), α5β1 (11), and αvβ3 (10), respectively, have been shown to play a role in this adhesion. Additionally, Abs to the non-integrin 67-kDa laminin binding protein (LBP) block adhesion of IL-3-dependent CMC to laminin-1 (6), indicating a role for this molecule.
In vivo, mast cells are phenotypically classified into two major subsets, connective tissue and mucosal (12, 13), and mucosal mast cells are important in the immune response to gut-dwelling nematodes. Nematode infection of the gut results in recruitment of the precursors of mucosal mast cells (MMC) (14) and their migration intraepithelially, where expression of the MMC-specific chymase, mouse mast cell protease-1 (mMCP-1), indicates the mature mucosal phenotype (15). As IL-3-dependent CMC are thought to represent an immature mast cell phenotype (16) and lack mMCP-1 expression (17), they may be a poor model for study of the molecular interactions involved in this process, especially since it is widely recognized that the cytokine stem cell factor (SCF) is a key growth and differentiation factor for mast cells in vivo and in vitro (18). We have recently developed an in vitro model of mast cell differentiation in which addition of the cytokine TGF-β1 to bone marrow cells supplemented with IL-3, IL-9, and SCF regulates differentiation of bone marrow cells into a close homologue of the mucosal phenotype, as shown by abundant expression of mMCP-1 (19).
Adhesion molecule expression in mast cells may be modulated during differentiation. The integrin α4 is expressed by IL-3-dependent CMC, is down-regulated in older cultures (11, 20), and is not present in mature tissue-derived connective tissue mast cells (20), but is required for the intestinal recruitment of MMC (14). In our in vitro model and in IL-3-dependent CMC (21), TGF-β1 up-regulates the expression of integrin αE (22), which may facilitate retention of MMC intraepithelially by binding E-cadherin (23). Therefore, as the key cytokine mediating differentiation of the mucosal phenotype, TGF-β1 may also change the spectrum of integrins expressed to reduce adhesion to ECM proteins and increase adhesion to the basement membrane (BM) protein laminin, thus facilitating intraepithelial migration. In support of this, one study showed that supplementing IL-3-dependent CMC with TGF-β1 for 48 h increased adhesion to laminin-1 (7); the mechanism was thought to be increased adhesion receptor expression.
To test the proposed hypothesis that the adhesion properties of CMC are regulated by TGF-β1, we have compared adhesion of mast cells cultured with and without TGF-β1 to the ECM proteins fibronectin and vitronectin and the BM protein laminin-1. We have also examined the expression of several integrins using RT-PCR and flow cytometry. In addition to those integrins previously implicated in mast cell adhesion, we have investigated the role of the laminin-binding integrin α7 in adhesion of TGF-β1-dependent CMC. This integrin was originally thought to be skeletal muscle specific (24), but expression of the α7B isoform has since been found in other tissues, including intestinal epithelium, where it correlates with intestinal cell differentiation (25). In this study we show the expression of α7B integrin in cultured MMC homologues, which is regulated by TGF-β1. We also demonstrate, by use of the blocking Ab 6A11, a novel role for α7 in promoting adhesion of mucosal mast cell homologues to laminin-1.
Materials and Methods
Mast cell culture
Mouse bone marrow cells were isolated from the femurs of male 12-wk-old BALB/c mice and suspended at 5 × 105/ml in DMEM (Life Technologies, Paisley, U.K.)/10% FCS (Sigma, Poole, U.K.), 2 mmol of l T+ and CMCT−, respectively. CMC were fed every 2–3 days by centrifuging and resuspending in half-volume of original culture medium and half-volume of fresh medium. Cells used for studies were mature cells, generally 14–30 days old, and consisted of >99% mast cells, as shown by toluidine blue staining. Immunohistochemistry was used to identify mMCP-1-positive cells (15). Typically, <2% of CMCT−, but >98% of CMCT+, were positive for mMCP-1 expression.
MC/9 culture
A mast cell line, MC/9, known to express the integrin α6β1 (5) was cultured for use as a positive control for flow cytometry and for blocking studies using a neutralizing rat anti-α6 mAb (GoH3). MC/9 cells were obtained from American Type Culture Collection (Manassas, VA) and cultured in DMEM (American Type Culture Collection) with 10% FCS (Sigma) and 10% T-stim (BD Biosciences, Oxford, U.K.), as instructed by American Type Culture Collection.
Antibodies
Mouse IgG1 mAbs to α7β1 integrins were obtained by immunization of α7-deficient mice with wild-type primary myoblasts, as described previously (26). Clone 6A11 was used for inhibition of CMCT+ adhesion to laminin, since it has the strongest adhesion inhibition activity of all the anti-α7 clones obtained. Clone 3C12 did not inhibit cell adhesion to laminin and was used for flow cytometry.
Murine IgG1 and rat IgG1 and IgG2a isotype controls, rat anti-mouse αv IgG1 (clone RMV-7), rat anti-mouse α5 IgG2a (clone 5H10-27), rat anti-mouse α6 IgG2a (clone GoH3), and rat anti-mouse FcγRIII/II (Fc block, clone 2.4G2) were obtained from BD PharMingen (Oxford, U.K.). Biotinylated-anti rat IgG2a or IgG1 secondary Abs were also obtained from BD PharMingen, and anti-mouse IgG1-Alexa Fluor 488 was obtained from Molecular Probes (Leiden, The Netherlands).
Adhesion assays
Murine Engelbreth-Holm Swarm sarcoma BM laminin-1 and bovine fibronectin and vitronectin were obtained from Sigma, and optimum coating concentrations for these proteins were determined from dose-response studies (data not shown). Ninety-six-well ELISA plates were coated overnight with 100 μl of proteins at 20 μg/ml in PBS, with the exception of vitronectin, which was used at 10 μg/ml; 3% BSA in PBS was used as a control. Excess coating proteins were then removed, and nonspecific binding was blocked by incubation for 2 h at 37°C with 3% BSA in PBS. CMC (and MC/9 cells for blocking studies) were washed twice in PBS/0.1% BSA before resuspending at 5 × 105 cells/ml in their respective culture medium (without cytokines). CMC (100 μl) were then loaded in quadruplicate into wells. PMA (50 ng/ml; Sigma) was added to some wells immediately after loading to investigate the effect of cell activation on adhesion. Cells were then incubated for 1 h at 37°C in 5% CO2, after which nonadherent cells were aspirated, and the wells washed three times with PBS. The number of cells adherent to wells was estimated using the β-hexose aminidase assay (27).
Regulation of CMCT+ adhesion to laminin-1
The effect of sensitization with IgE followed by addition of specific Ag and the effect of calcium ionophore A23187 on adhesion of CMCT+ to laminin-1 were investigated. For stimulation by FcεRI cross-linking, CMCT+ were first sensitized by incubation overnight in complete culture medium with 100 ng/ml of IgE anti-DNP (Sigma). They were then washed twice in PBS/0.1% BSA and resuspended in culture medium (without cytokines) before loading into wells and immediate addition of 10 ng/ml of DNP-human serum albumin (Sigma). Controls included IgE-sensitized cells to which no Ag was added and unsensitized cells to which Ag was added. For stimulation with calcium ionophore, CMCT+ cells, washed and resuspended in culture medium without cytokines, were loaded into wells, and 1 μM calcium ionophore A23187 (Sigma) was immediately added. PMA-stimulated cells were also included as a positive control for stimulation of adhesion, and as a negative control, wells were included where no activating agent was added. The adhesion assay was then performed as described above.
Analysis of integrin expression by RT-PCR
Expression of integrins by cells from four separate cultures of CMCT− and CMCT+ was investigated by RT-PCR. Total RNA was recovered from 5 × 106 cells/culture in Tri-Reagent (Sigma) by phenol-chloroform extraction. RNA was then DNase treated using a DNA-free kit (Ambion, Houston, TX) and 1 μg of RNA reverse-transcribed in a 20-μl volume using a Promega RT kit (Promega, Southampton, U.K.). One microliter of each RT reaction was used for semiquantitative PCR, employing gene-specific primers (Table I⇓); cycle numbers were optimized so that increases in expression would result in corresponding increases in signal intensity. Negative controls were set up for each sample containing RNA only (no cDNA), and each PCR experiment included a negative control omitting cDNA. PCR products were separated on a 1.4% agarose gel containing 0.5 μg/ml of ethidium bromide and were visualized and recorded under UV light using a Kodak Image Station 440cf imaging system (Eastman Kodak, Rochester, NY). PCR product identities were confirmed by Southern hybridization using gene-specific oligonucleotide probes as described previously (28).
Primer sequences and conditions used for RT-PCRa
Analysis of integrin expression by flow cytometry
Surface expression of integrins was analyzed by flow cytometry (FACScan; BD Biosciences). To detect the expression of α5, α6, and αv, 1 × 106 cells were incubated for 5 min with 0.5 μg of murine Fc block, followed by incubation for 60 min with 1 μg of rat mAb against α5 (5H10-27), α6 (GoH3), or αv (RMV-7), respectively, or isotype control in PBS/10% mouse serum. The cells were then washed twice in wash buffer (PBS/0.1% BSA) and incubated for 30 min with 1 μg of biotin-anti rat IgG2a or IgG1 in PBS/10% mouse serum before washing twice and incubating for 30 min with 100 μg of streptavidin-PE (Vector, Peterborough, U.K.). To detect α7 expression, 2 × 105 cells were incubated with 0.1 μg of murine Fc block before incubation for 60 min with 2 μg of mouse anti-α7 (3C12) or mouse IgG1 isotype control in PBS/5% FCS. They were then washed twice in wash buffer and incubated for 30 min with anti-mouse IgG1-Alexa Fluor 488 diluted 1/2000 in PBS/5% FCS. Following both labeling protocols, cells were washed twice and fixed for 10 min in 2% paraformaldehyde before analysis by flow cytometry; all procedures were performed on ice.
Blocking studies
The effects of integrin-specific mAbs on CMCT+ adhesion to laminin-1 were examined. Adhesion assays were as previously described, except CMCT+ were preincubated with respective mAbs or isotype controls before loading into wells. Preincubation conditions and optimum Ab concentrations were determined in pilot studies (data not shown). To determine the effects of α6 and α7 integrins on adhesion of CMCT+ to laminin-1, CMCT+ were preincubated for 10 min at room temperature with the α6 mAb GoH3 (1 μg/ml) or for 30 min on ice with the α7 mAb 6A11 (10 μg/ml). MC/9 cells were included as a positive control in experiments investigating the effect of GoH3 on adhesion to laminin-1(5).
Statistical analysis
PRISM (version 3.0 for Windows; GraphPad, San Diego, CA) statistical software was used to compare data using the nonparametric Mann-Whitney U test, with a statistical significance level of p < 0.05.
Results
MMC homologues exposed to TGF-β1 adhere to laminin-1
We investigated the effect on their adhesion properties of culturing mast cells with and without TGF-β1 (1 ng/ml). Results are from four separate experiments, each performed in quadruplicate, using CMC from different cultures and are expressed as the mean ± SEM of each experiment. The range of means ± SEMs of separate experiments is also shown. Experiments established that 27.7 ± 10.7% (range, 5.2 ± 0.3 to 56.4 ± 2.9%) of CMCT+ (mMCP-1+ MMC homologues) adhered spontaneously to laminin-1. CMCT+ adhesion to laminin-1 increased to 58.7 ± 4.0% (range, 51.4 ± 4.6 to 69.1 ± 2.9%) following PMA stimulation (Fig. 1⇓a), and adhesion of both control (unstimulated) and PMA-stimulated populations of CMCT+ was significantly greater (p < 0.05) than that of equivalently stimulated CMCT− (controls, 0.0 ± 0.1 to 3.5 ± 2.8%; PMA-stimulated, 0.0 ± 0.8 to 15.6 ± 2.6%) that were grown in the absence of TGF-β1. In all experiments CMC adhesion to BSA was <5% (data not shown). Adherent CMCT+ flattened and took on a more polarized morphology on laminin-1 (data not shown). PMA-stimulated CMCT+ adhered poorly to fibronectin (7.9 ± 4.3%) and did not adhere to vitronectin (Fig. 1⇓b)
CMC adhesion to ECM and BM proteins. Adhesion of unstimulated (control) and PMA-stimulated CMCT− and CMCT+ to laminin-1 (a), fibronectin (b), and vitronectin (c). CMCT+ grown in the presence of TGF-β1 (MMC homologues) adhered to laminin-1, but not to fibronectin or vitronectin, whereas CMCT− grown in the absence of TGF-β1 adhered poorly to laminin-1, but adhered to fibronectin and vitronectin following PMA stimulation. The adhesion of both control and PMA-stimulated CMCT+ to laminin was significantly greater (p < 0.05) than the adhesion of equivalently stimulated CMCT−. Also, the adhesion of CMCT− to fibronectin and vitronectin was significantly greater (p < 0.05) than the adhesion of PMA-stimulated CMCT+. Results are the mean ± SEM of four separate experiments, each performed in quadruplicate and using cells from four different cultures.
Stimulation with Ag/IgE and with the calcium ionophore A23187 up-regulates adhesion of MMC homologues to laminin-1
Adhesion of IL-3-dependent CMC to ECM proteins can be stimulated by treatment with Ag/IgE or calcium ionophore A23187 (7), and similar mechanisms may control adhesion in vivo. We therefore investigated whether the TGF-β1-induced adhesion of CMCT+ (mMCP-1+ MMC homologues) to laminin-1 was also regulated by these mechanisms. Both Ag/IgE and ionophore treatments significantly increased (p < 0.05) adhesion by CMCT+ to laminin-1 (Fig. 2⇓). Ag/IgE treatment resulted in flattening and polarization of cells, but after calcium ionophore treatment adherent cells retained a rounded shape (data not shown). This experiment was repeated twice using cells from different cultures, with similar results.
Regulation of adhesion of MMC homologues to laminin-1. Adhesion was significantly increased (p < 0.05) by stimulation using PMA, Ag/IgE, and the calcium ionophore A23187. Overnight sensitization of MMC homologues with IgE also significantly increased adhesion to laminin, but addition of specific Ag to sensitized cells significantly increased adhesion above that seen in cells sensitized only. Results are the mean ± SEM (n = 4) of one experiment that was repeated twice more using cells from different cultures, with similar results.
MMC homologues exposed to TGF-β1 express the laminin-binding integrin α7B at both mRNA and protein levels
To determine the mechanism of TGF-β1-mediated binding of CMC to laminin, we compared the expression of integrin transcripts by CMCT− and CMCT+ populations. Previous studies (5, 8) have shown involvement of α6 in adhesion of CMC to laminin; therefore, we examined the expression of this integrin and of the laminin-binding integrins α3 and α7 in unstimulated CMC cultured with and without TGF-β1, using semiquantitative RT-PCR. The α7 primers could amplify both α7A and α7B mRNA, resulting in PCR products of 480 and 366 bp, respectively, but only a single band of 366 bp was detected, indicating the expression of α7B transcripts only. The expression of α7 integrin was low to undetectable after 35 cycles of PCR in CMCT−, whereas inclusion of TGF-β1 in cultures resulted in a substantial up-regulation of α7B mRNA expression by CMCT+ (Fig. 3⇓). The expression of α7 integrin was confirmed by flow cytometry using the mouse mAb 3C12. Surface expression of α7 was consistently detected in TGF-β1-supplemented CMCT+ from three different cultures during separate experiments, whereas no expression was detected in CMCT− (Fig. 4⇓).
Analysis of adhesion molecule expression in CMCT− and CMCT+ by semiquantitative RT-PCR. Expression of α7B integrin mRNA was highly up-regulated in CMC cultured with TGF-β1, whereas TGF-β1 moderately down-regulated the expression of integrin β3. Transcripts for other adhesion molecules were expressed at comparable levels by both CMCT− and CMCT+.
Analysis of integrin expression in CMCT− and CMCT+ by flow cytometry. MC/9 cells were included as positive controls for the expression of α5 and α6 integrins. CMCT+ expressed α7 integrin, whereas there was no expression in CMCT− cultured without TGF-β1. α6 expression was virtually absent in both CMCT− and CMCT+; high expression was detected in the positive control MC/9 cells. α5 integrin was expressed by CMCT−, but there was decreased expression in CMCT+ and high expression in MC/9 cells. The expression of integrin αv was virtually absent in both CMCT− and CMCT+. These experiments were repeated twice more using cells from separate cultures, with similar results.
Transcripts for other laminin-binding integrin subunits, 3A and 6A, and of the very late Ag integrin β-subunit β1 were expressed at similar levels in both CMCT+ and CMCT− (Fig. 3⇑). However, because others have reported the expression of α6 integrin in murine CMC and the mast cell line MC/9 (5, 8, 10), we compared α6 expression in MC/9 cells and CMC with or without TGF-β1 by flow cytometry. The expression of α6 was virtually absent in CMCT+ and CMCT−, but, as expected, α6 expression was high in MC/9 cells (Fig. 4⇑). LBP transcripts were also detected at similar levels in both cell types (data not shown).
The α7-neutralizing mAb 6A11 blocks adhesion of MMC homologues to laminin-1
To demonstrate the role of integrins in adhesion of MMC homologues to laminin, CMCT+ were preincubated with neutralizing Abs before use in adhesion assays. The α7-neutralizing mAb 6A11 at 10 μg/ml resulted in a 98% reduction in adhesion of PMA-stimulated CMCT+ (MMC homologues) to laminin-1 (Fig. 5⇓a), which was statistically significant (p < 0.05). This experiment was repeated twice more in quadruplicate and triplicate, with reductions in adhesion of 98 and 100%, thus clearly indicating a role for α7 in adhesion to laminin-1. 6A11 also blocked spontaneous adhesion to laminin-1 (not shown) and adhesion following Ag/IgE and A23187 stimulation (Fig. 5⇓, b and c). The use of neutralizing Abs has shown a role for α6 in adhesion of IL-3-dependent CMC to laminin-1 (5, 8, 10), but the anti-α6 mAb GoH3 (1 μg/ml) had no effect on adhesion of CMCT+ to laminin-1 (Fig. 5⇓d), whereas adhesion of MC/9 cells, which are reported to express the integrin α6β1 (5) and were used as a positive control for this Ab, was reduced by 98% in the presence of GoH3.
Role of integrins in the adhesion of CMCT+ (MMC homologues) to laminin-1 (20 μg/ml). The α7 integrin-neutralizing Ab 6A11 (10 μg/ml) reduced the adhesion of PMA-stimulated CMCT+ to laminin-1 by 98% (p < 0.05; a) and similarly reduced the adhesion of CMCT+ stimulated by Ag/IgE (b) and that of the calcium ionophore A23187 (c) to laminin-1. In contrast, the α6-neutralizing Ab GoH3 (1 μg/ml) had no effect on the adhesion of CMCT+ to laminin-1, but significantly (p < 0.05) reduced the adhesion of MC/9 cells to laminin-1 by 98% (d). Results are the mean ± SEM (n = 4) for single experiments. Experiments a and d were repeated twice more using cells from different cultures, with similar results.
Mast cells cultured in the absence of TGF-β1 adhere to fibronectin and vitronectin, but not to laminin-1
Previous studies of the binding of murine CMC to matrix proteins have predominantly used cells cultured in the presence of IL-3 alone (5, 7, 8, 10, 11). Adhesion of CMC grown in a combination of IL-3, SCF, and IL-9, all of which are expressed in the gut (29, 30) and are known mast cell growth factors (31, 32, 33), has not previously been studied. As shown above, CMC grown in the absence of TGF-β1 did not adhere to laminin-1. Furthermore, these cells did not adhere spontaneously to vitronectin, but had low (7.1 ± 4.9%; range, 0.5 ± 1.0 to 21.8 ± 4.3%) spontaneous adhesion to fibronectin (Fig. 1⇑b). However, after PMA stimulation, CMCT− adhered to vitronectin (38.4 ± 12.2%; range, 8.3 ± 3.6 to 65.6 ± 6.0%) and fibronectin (66.8 ± 8.6%; range, 47.9 ± 6.3 to 82.6 ± 1.4%) at levels which were significantly greater (p < 0.05) than those observed for CMCT+ (Fig. 1⇑, b and c). Adhesion of CMCT− to both ECM proteins was associated with cell flattening (data not shown).
Mast cells cultured in the absence of TGF-β1 do not express α7 integrin, but show increased expression of fibronectin- and vitronectin-binding integrins
As previously described, analysis of integrin expression by RT-PCR established that CMCT− do not express α7 integrin, but express β1A, α3A and α6A at comparable levels to those seen in CMCT+ (Fig. 3⇑). However, as CMCT− adhered to fibronectin and vitronectin, we wondered whether this was also due to TGF-β1-mediated alterations in integrin expression. We therefore investigated the expression of integrins α5 and αvβ3 in CMCT− and CMCT+ by RT-PCR, as these integrins play a role in the adhesion of IL-3-dependent CMC to fibronectin and vitronectin, respectively. The expression of α5 and αv transcripts was similar in both cell types, but the expression of the vitronectin-binding integrin β3 was up-regulated in CMCT− (Fig. 3⇑) compared with that in CMCT+. In the absence of differences at the mRNA level, flow cytometry was used to investigate surface expression of α5 and αv (Fig. 4⇑). Expression was measured on cells from three different cultures of CMCT− and CMCT+ during separate experiments, and the mast cell line MC/9, which has been shown previously to express integrin α5 (11), was used as a positive control for the expression of this integrin. Surface expression of α5 was increased in CMCT− compared with CMCT+, but the expression in both CMC was lower that seen in the positive control MC/9 cells. The expression of αv was low to absent in both CMCT− and CMCT+.
Discussion
CMC have previously been shown to adhere to laminin (6, 7), fibronectin (9), and vitronectin (10). Here we show TGF-β1 modulation of adhesion of CMC to these proteins and a novel role for α7 integrin in adhesion of MMC homologues to laminin-1. We propose that TGF-β1-mediated up-regulation of α7 expression in conjunction with differentiation of the mucosal mast cell phenotype and expression of mMCP-1 and αEβ7 (22) may have a role to play in the intraepithelial location of MMC in vivo.
α7β1 has been described as being essentially muscle specific (34), and although several non-muscle locations have been described (35, 36), expression has not previously been shown in leukocytes of any subset. Additionally, TGF-β1-mediated regulation of α7 expression has not been shown in any cell type, and although developmental regulation of α7 has been described in muscle and has been proposed for intestinal epithelium (25), the molecular signals controlling expression in these cells are unknown.
The α7B isoform expressed by MMC homologues is also expressed in skeletal myoblasts; the α7A isoform is restricted exclusively to mature skeletal muscle (26, 37, 38). However, as both isoforms promote adhesion on laminin-1 and laminin-2/4 (39, 40), which may be rich in epithelial BM (41), α7B expression in vivo could either promote intraepithelial migration or limit egress of MMC from the epithelium into lamina propria via adhesion to BM laminin. Additionally, restricted expression of α7B by epithelial cells of the crypt-villous junction (25), the primary location of intraepithelial MMC in vivo (42), suggests that α7 may be important for the adhesion of both epithelial cells and mast cells to the BM at this site.
The phorbol ester PMA stimulated maximal adhesion of MMC homologues to laminin-1, presumably due to alterations in receptor affinity or cytoskeletal rearrangements (43) and the anti-α7 mAb 6A11 blocked this adhesion. FcεRI cross-linking and use of the calcium ionophore A23187 also stimulated adhesion to laminin-1, as previously shown in IL-3-dependent CMC (7, 8), and these may represent in vivo mechanisms by which adhesion could be stimulated. Spontaneous adhesion of CMCT+ to laminin-1 was observed in some cultures, although this was inconsistent and may have been due to low levels of endotoxin contamination, as LPS has been shown to cause mast cell activation (44). Preincubation with 6A11, however, almost completely blocked adhesion via all the above pathways, showing a universal role for integrin α7 in the adhesion of MMC homologues to laminin-1.
Previous studies (5, 8) have implicated integrin α6 in the promotion of murine mast cell adhesion to laminin-1 and –2; however, these studies used several mast cell lines and IL-3-dependent CMC, which may be more representative of immature mast cells. The mMCP-1+ CMCT+ closely resemble intraepithelial MMC (15, 19), and while these cells expressed α6 transcripts at similar levels to CMCT−, flow cytometry showed low α6 expression in both cell types. This made α6 an unlikely candidate for regulation of CMCT+ adhesion to laminin, and inclusion of the anti-α6 mAb GoH3 in adhesion assays had no effect on adhesion of CMCT+ to laminin-1. Comparison with positive α6 expression by flow cytometry and GoH3-mediated inhibition of adhesion to laminin-1 in MC/9 cells further supports our findings that α6 integrin plays no role in the adhesion of MMC homologues to laminin-1.
Human skin mast cells also adhere to laminin, but do not significantly express α6 integrin, and adhesion is inhibited by Abs to α3 integrin (45). Abs to α3 integrin were not included in our studies, but in view of conclusive evidence for the role of α7 integrin in the adhesion of MMC homologues and because the expression of α3 mRNA expression in CMCT+ was similar to that in CMCT−, which do not adhere to laminin, a major role for α3 integrin in laminin binding seems unlikely in MMC. The importance of different integrins in adhesion to laminin may vary between mast cell phenotypes, with α6 possibly playing a role only in he adhesion of immature mast cells, while α3 and α7 may mediate the adhesion of mature connective tissue mast cells and MMC, respectively, to laminin.
LBP has been implicated in the adhesion of IL-3-dependent CMC to laminin-1 (6, 7), but the expression of LBP mRNA was similar in both CMCT− and CMCT+ (data not shown). However, it is reported that LBP expression is post-transcriptionally regulated (46); therefore, a cofactor role in the adhesion of MMC homologues to laminin-1 is possible, as suggested in other cell types (47).
CMC cultured in the absence of TGF-β1 (CMCT−) do not have a recognizable in vivo counterpart since they lack mMCP-1 and do not show any morphological resemblance to the very well-characterized serosal mast cell population. Their role in this study was simply as a comparator for MMC homologues cultured in TGF-β1, and they bound preferentially to fibronectin and vitronectin, but not to laminin. Post-transcriptional down-regulation of the fibronectin-binding integrin α5 and transcriptional down-regulation of the vitronectin-binding integrin β3 in MMC homologues compared with CMCT− suggest that TGF-β1 may also regulate mast cell adhesion to these ECM proteins by alteration in expression of specific integrins.
In vivo, secreted TGF-β1 must be activated to form a functional molecule, and epithelially expressed integrin αvβ6 has been implicated in this process (48). Our most recent studies have shown coexpression of TGF-β1 and integrin αvβ6 in murine jejunal epithelium, and that β6−/− mice have significantly reduced numbers of intraepithelial MMC following Nippostrongylus brasiliensis infection (49). This result is highly suggestive of a role for TGF-β1 in the intraepithelial MMC response to nematode parasites. It is possible that mechanisms include regulation of the expression of integrins, including α7β1 and αEβ7, that could be critical for intraepithelial migration and retention of MMC.
Acknowledgments
We thank Dr. J. Brown and A. Sanderson for assistance with flow cytometry, and Dr. A. Pemberton for help with the preparation of the manuscript. Thanks also to Eileen Duncan and Liz Moore for maintenance of BALB/c mice.
Footnotes
-
↵1 This work was supported by grants from the Wellcome Trust (Grant 060312) and the Biotechnology and Biological Sciences Research Council (Grant 99/V1/S/5158).
-
↵2 Current address: Pathology Department, Safety Assessment, GlaxoSmithKline, Park Road, Ware, U.K. SG12 0DP.
-
↵3 Address correspondence and reprint requests to Dr. Hugh R. P. Miller, Department of Veterinary Clinical Studies, University of Edinburgh, Easter Bush Veterinary Center, Roslin, Midlothian, U.K. EH25 9RG. E-mail address: hugh.miller{at}ed.ac.uk
-
↵4 Abbreviations used in this paper: ECM, extracellular matrix; BM, basement membrane; CMC, bone marrow-cultured mast cell; LBP, 67-kDa laminin binding protein; MMC, mucosal mast cell; mMCP-1, mouse mast cell protease-1; SCF, stem cell factor.
- Received July 15, 2002.
- Accepted September 18, 2002.
- Copyright © 2002 by The American Association of Immunologists