The Influence of Mast Cells on Pathways of Transepithelial Antigen Transport in Rat Intestine¹

Local hypersensitivity reactions at mucosal surfaces play an important role in the pathophysiology of allergic diseases, including food allergy, atopic asthma, and rhinitis. Ag challenge in sensitized individuals leads to mast cell activation by cross-linking IgE bound to FcεRI on the cell surface. Released mast cell chemicals such as histamine, serotonin, proteases, and lipid mediators produce alterations in epithelial and smooth muscle physiology (1) that are responsible for many of the acute symptoms of allergic disease. Such symptoms develop very rapidly (within minutes) after an encounter with Ag (2). However, because the mucosal epithelium is believed to provide a selective barrier that restricts the influx of ingested or inhaled Ags, it is unclear how luminal Ags are transported across this barrier to reach effector cells.

Soluble luminal Ags can be taken up across the epithelium by two routes: the transcellular pathway or the paracellular pathway. The transcellular pathway involves endocytic uptake of Ag at the apical membrane and the transport of this Ag in endocytic vesicles through the cell to the basolateral membrane where it is released into the extracellular space (3). The paracellular pathway is the pathway between epithelial cells (ECs)³. However, this route is restricted by intercellular tight junctions at the apical pole of ECs that limit passage of macromolecules (4, 5).

Recently, we reported enhanced transepithelial transport of Ag in a rat model of intestinal hypersensitivity (6). Electron photomicrographs clearly demonstrated accelerated transport of protein Ag (horseradish peroxidase (HRP)) within endosomes, such that Ag was present in the lamina propria (LP) at 2 min (∼10 times faster than normal) after its addition to the mucosal buffer bathing tissues in Ussing chambers. A secretory response was evident by 3 min, as indicated by an increase in the short-circuit current (Isc) associated in time with evidence of mast cell activation (clear zones around granules). Subsequently (>30 min postchallenge), a large increase in the flux of Ag across the tissue was documented, and a larger conductance value suggested a decreased resistance of the paracellular pathway. This observation was confirmed by electron microscopy that showed Ag in the paracellular regions and tight junctional areas.

A large body of evidence indicates that mast cells regulate epithelial ion transport (7). However, there is little information on whether mast cells are involved in the regulation of epithelial permeability. Support for this possibility includes studies showing that Ag challenge of sensitized rats results in increased transport of a range of probes, from small m.w. markers (8) to proteins (9). In addition, infusion of the mast cell mediator, rat mast cell protease II, caused enhanced leakage of protein into the intestinal lumen (10). However, the exact role of mast cells in the alteration of the rate or route of transepithelial Ag transport has not been clearly defined.

Therefore, the aim of the current study was to assess directly the role of mast cells in transepithelial Ag transport across small intestine from sensitized rats by comparing results in mast cell-deficient rats and controls. White spotting (Ws)/Ws rats have a genetic mutation at the c-kit locus resulting in a lack of mast cells within the intestinal mucosa, whereas +/+ littermates have been shown to have normal numbers of mast cells (11). As before, we chose HRP as our model protein Ag, since it can be measured quantitatively by enzymatic assay and visualized by electron microscopy. The transepithelial transport of HRP across jejunal segments was assessed after its addition to the luminal side of tissues in Ussing chambers. Our analysis of electron photomicrographs demonstrated similar results in both sensitized +/+ and Ws/Ws rats in the initial phase of Ag transport; these results included enhanced uptake and transport of Ag in endocytic vesicles, indicating that mast cells are not involved in this phase. However, a hypersensitivity response occurred only in the +/+ rats, indicating an absolute requirement for mast cells in Ag-stimulated ion secretion. In addition, only +/+ rats showed an increased overall flux of Ag, increased conductance, and the presence of HRP in paracellular spaces. These findings provide evidence that mast cells regulate the permeability of the epithelial paracellular pathway.

Ws/Ws and +/+ rats were obtained by breeding male and female Ws/+ heterozygous rats (from the original colony developed by Y.K.). A spontaneous mutation (Ws/+) was first identified in a BN/fMai rat colony, and the heterozygous rats were bred with female rats of the Donryu strain to obtain viable Ws/Ws rats (11). Ws/Ws rats have a 12-base deletion in the tyrosine kinase domain of the c-kit gene (12) that results in a lack of melanocytes, E, and mast cells. By 10 wk of age, no mast cells can be detected in skin (13) or intestine (14) from Ws/Ws rats, whereas +/+ rats have normal numbers of mast cells. Rats (>12 wk of age), were maintained on a 12-h light/dark cycle and were given food and water ad libitum. Some experiments were repeated with Sprague Dawley rats (Charles River, St. Constant, Canada) to confirm our previous findings (6). Experiments were approved by the Animal Care Committee at McMaster University.

Rats were sensitized to HRP by s.c. injection with 1 mg of HRP (type II, Sigma-Aldrich Canada, Oakville, Canada) in 1 ml aluminum hydroxide (10%) and by i.p. injection of 1 ml Bordetella pertussis vaccine (Connaught Laboratories, Willowdale, Canada) as adjuvants to stimulate IgE production (15). Naive rats that had been sham-sensitized by saline injection served as controls. Experiments were conducted at 10 to 14 days after sensitization. Rats were anesthetized, and a blood sample was obtained for the measurement of IgE. A laparotomy was performed, and a 15 to 20 cm segment of jejunum was excised, beginning at 5 cm distal to the ligament of Treitz, and immediately placed in warmed oxygenated Krebs buffer.

Intestinal segments were placed on a plastic rod, and the external muscle layers were stripped off while leaving the submucosal plexus and mucosa intact. From each rat, 4 to 8 pieces of intestine were mounted in Ussing chambers (WPI Instruments, Narco Scientific, Mississauga, Canada). Care was taken to avoid tissue containing Peyer’s patches. The chamber opening exposed 0.6 cm² of serosal surface area to 8 ml of circulating oxygenated Krebs buffer (pH 7.35) 37°C. The serosal buffer contained 10 mM of glucose that was osmotically balanced with 10 mM of mannitol in the mucosal buffer. The tissue was clamped at 0 V using a W-P Instruments automatic voltage clamp (Narco Scientific, Downsview, Canada). The Isc (in μA/cm²) was recorded continuously. At 5-min intervals, the tissue was voltage clamped at 1 mV (for a duration of 1 s), and the Isc deflection was used to determine the conductance (G, mS/cm²) according to Ohm’s law. Tissues were allowed to equilibrate until the Isc stabilized before HRP (5 × 10⁻⁵ M) was added to the luminal buffer. The Isc response to HRP was measured as the peak increase in Isc within 15 min after the addition of HRP to the luminal buffer.

To determine the mucosal to serosal flux of HRP, duplicate samples (500 μl) of serosal buffer were obtained at 0, 30, 60, and 90 min after the addition of HRP and were replaced with Krebs buffer. HRP activity was measured by assaying enzyme activity using a modified Worthington method (16). Briefly, 150 μl of sample was added to 800 μl of phosphate buffer containing 0.003% H₂O₂ and 80 μg/ml o-dianisidine (Sigma). The HRP concentration was calculated using enzyme activity (the rate of increase in OD at 460 nm over a 2-min period). Fluxes were calculated according to standard formulae and were expressed as pmol/cm²/h.

To examine the route and extent of initial Ag uptake across the intestinal epithelium, tissues were removed from Ussing chambers at 2 min after HRP challenge (∼1 min before the Isc response). To examine Ag transport after the hypersensitivity reaction, tissues were removed at 90 min. Tissues were fixed in 2% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) for 2 h at room temperature. Tissues were incubated overnight at 4°C in 0.1 M sodium cacodylate buffer and then washed three times for 5 min each in 0.05 M Tris buffer (pH 7.6). Segments were incubated for 30 min in 3,3′-diaminobenzidine tetrahydrochlorine (Sigma) (5 mg in 10 ml 0.05 M Tris buffer and 0.01% H₂O₂ (pH 7.6) at 20°C). Samples were then processed for routine electron microscopy and embedded in Epon. Tissues were oriented so that villus ECs were cut longitudinally, and photomicrographs were prepared from the mid-villus region. To assess the distribution of HRP across the epithelium, the incidence of HRP within the apical or basal regions of the cell or in the LP was recorded. The total area of HRP within endosomes was also quantified in windows of fixed area (4 × 6 μm) (see Fig. 1 for schematic representation). A total of 10 micrographs per region were used for each rat; 40 micrographs were used for each rat group. For each rat, the percentage of windows that were positive for HRP in each region were determined, and then mean values were calculated for each rat group. This analysis was performed by one investigator (M.C.B.) who was unaware of the treatment group.

IgE was determined in serum by passive cutaneous anaphylaxis (PCA) as described previously (17). Briefly, Sprague Dawley rats were injected intradermally with 100 μl of diluted serum from Ws/Ws and +/+ rats. Samples were run in duplicate dilutions from 1/8 to 1/512. In addition, sera were heat-treated (56°C, 60 min) and injected intradermally as described above. After 72 h, rats were challenged by i.v. injection of a 0.5 ml solution of 1% Evans blue (Sigma) containing 2.5 mg of HRP. Bluing of the skin was evaluated at 30 min after injection. The highest serum dilution giving a positive reaction was recorded as the PCA titer of that serum.

The statistical significance between the treatment groups was assessed using ANOVA; a Dunnett t test was used for post hoc analysis. An analysis of conductance measures was performed using a repeated measures ANOVA. The correlation between HRP flux values and conductance measurements was assessed using Pearson’s correlation test. Differences between groups were considered significant at p < 0.05.

It was confirmed that mucosal mast cells were present in intestinal tissue sections from +/+ rats; however, no mast cells were visible in the intestinal tissues of Ws/Ws rats.

In Ussing chambers, intestinal segments from HRP-sensitized +/+ rats responded to luminal HRP challenge with an increase in Isc (13.8 ± 2.0 μA/cm², mean ± SEM) beginning at 3.4 ± 0.5 min after challenge. These results were similar to but of lower magnitude than those obtained from HRP-sensitized Sprague Dawley rats (with a mean increase in Isc of 32.1 ± 2.5 μA/cm² at 3.4 ± 0.4 min after challenge). Tissues from naive rats and sensitized Ws/Ws rats showed no Isc response to HRP challenge and maintained a stable Isc baseline for the duration of the 90-min experimental period. Representative Isc tracings for tissues from sensitized +/+ and Ws/Ws rats are shown in Figure 2. To ensure that the sensitization of Ws/Ws rats was successful, specific Ab titers were measured by PCA. Sensitized +/+ and Ws/Ws rats had similar mean Ab titers of 7.4 ± 0.4 and 8.2 ± 0.2, respectively (log₂ PCA titer). Heat treatment of the serum abolished the PCA response, indicating that the reactive Abs were of the IgE isotype (18).

As shown above, luminal HRP challenge elicited a very rapid Isc response that was dependent upon the presence of mast cells. We have previously shown that the sensitization of Sprague Dawley rats enhances Ag transcytosis across the intestinal epithelium to the extent that HRP reaches the LP within 2 min (6). To confirm this finding and to examine the role of mast cells in enhanced Ag transcytosis, results were compared in tissues from naive and sensitized Sprague Dawley and +/+ rats and sensitized Ws/Ws rats. In electron photomicrographs of tissues from all sensitized rats, the distribution of HRP-containing endocytic vesicles was similar: HRP-containing vesicles were identified in the apical and basal regions of enterocytes and in the LP as soon as 2 min after HRP challenge. HRP was also found in goblet cells below the level of the theca. Examples are shown in Figure 3, a–c. HRP was not visualized within any tight junctions or paracellular regions in the photomicrographs from any of the rat groups at this timepoint. Compared with results from unsensitized naive rats, the sensitization of rats increased the incidence of HRP-containing vesicles in both the apical and basal region of enterocytes. Sensitized Sprague Dawley rats had HRP within endosomes in 65 ± 11% of apical windows and 45 ± 11% of basal windows compared with 30 ± 11% and 0% in the apical and basal regions of the epithelium from naive Sprague Dawley rats. Naive +/+ rats had endosomal HRP in the apical region of enterocytes in only 18 ± 3% of the photomicrographs examined compared with 41 ± 9% and 48 ± 6% in sensitized +/+ and Ws/Ws rats, respectively. The basal regions of enterocytes contained endosomal HRP in 13 ± 3% of photomicrographs from naive +/+ rats compared with 29 ± 6% and 36 ± 9% of photomicrographs from sensitized +/+ and Ws/Ws rats, respectively (Fig. 4). The area of endosomal HRP was also not statistically different between sensitized Ws/Ws and +/+ rats in all regions (apical: 180 ± 42 vs 120 ± 53; basal: 125 ± 90 vs 107 ± 58 nm² for +/+ and Ws/Ws rats, respectively). The lack of significant differences between sensitized +/+ and Ws/Ws rats indicated that mast cells do not influence the initial uptake and transport of Ag.

We have previously shown that the mucosal to serosal flux of HRP across jejunal segments (measured over a 90-min period after challenge) was significantly increased in Sprague Dawley rats that had been sensitized to HRP compared with naive controls or rats that had been sensitized to an irrelevant Ag. To determine the role of mast cells in Ag transport at this stage, the mucosal to serosal flux of intact HRP and tissue conductance were compared in tissues from naive and HRP-sensitized +/+ and Ws/Ws rats. The HRP flux was significantly higher across intestine from sensitized +/+ rats compared with unsensitized +/+ rats beginning in the second 30-min flux period (53.9 ± 10.6 vs 25.4 ± 3.6 pmol/cm²/h) and was even more pronounced in the third flux period (114.0 ± 19.5 vs 45.2 ± 5.7 pmol/cm²/h). Intestine from sensitized Ws/Ws rats had an HRP flux that was comparable with unsensitized controls (21.5 ± 4.2 and 21.7 ± 3.5 pmol/cm²/h in the second and third flux periods, respectively) and was significantly less than sensitized +/+ rats (Fig. 5). Conductance measures at 90-min postchallenge correlated closely with the HRP flux (r = 0.83). Conductance gradually increased throughout the experimental period in sensitized +/+ rats but not in naive or Ws/Ws rats. A statistically significant increase was observed at 60 min after HRP challenge; this increase was maximal at 90 min (Δ conductance was 12.0 ± 2.0 mS/cm² for sensitized +/+ rats vs −0.3 ± 0.4 for sensitized Ws/Ws rats) (Fig. 6).

Electron microscopy of tissues that were fixed at 90 min postchallenge demonstrated HRP in the paracellular regions of sensitized +/+ rats but not the same regions in Ws/Ws or naive +/+ rats (Fig. 7). This observation is consistent with previous findings in Sprague Dawley rats that demonstrated that increased HRP flux and conductance were associated with the presence of HRP within the tight junctions and paracellular regions between ECs.

In this study, we demonstrated that intestinal transepithelial Ag transport occurred in two phases, an initial phase of Ag uptake and translocation and a second phase that followed after the hypersensitivity reaction. The initial phase was increased by sensitization and was mast cell-independent. The second phase was dependent upon both sensitization and the presence of mast cells. Initially, HRP was transported by transcytosis in endocytic vesicles. In this phase, the uptake of HRP at the apical membrane and its transport through the cell was significantly greater in rats that had been sensitized to HRP. Similar data in +/+ and Ws/Ws rats indicated that mast cells did not affect this phase. After the hypersensitivity reaction, both the HRP flux and the tissue conductance values were significantly elevated only in mast cell-containing intestine from rats that had been sensitized to HRP, indicating that mast cells were required for this second phase of Ag transport.

Intestine from sensitized Ws/Ws mast cell-deficient rats did not respond to luminal Ag challenge with an increase in Isc. This observation is consistent with findings from previous studies regarding the critical role for mast cells in the intestinal hypersensitivity reaction. Doxantrazole, a mast cell stabilizer, blocked the Isc response to Ag challenge (19). Experiments in W/W^v mast cell-deficient mice and +/+ littermate controls showed that mast cells were responsible for the majority of the Ag-induced increase in Isc; however, a small component remained in intestine from W/W^v mice (20). We obtained similar findings in our current experiments using Ws/Ws and +/+ rats, but the entire Isc response to Ag was eliminated in the absence of mast cells, suggesting that this is a better model to study mast cell-mediated changes in physiology. The absent hypersensitivity reaction in Ws/Ws rats was not due to an inability of these rats to mount an IgE Ab response, since both Ws/Ws and +/+ rats had similar PCA titers that were abolished by heat treatment.

An analysis of electron photomicrographs of tissues that were fixed at 2 min after luminal HRP challenge demonstrated that the route of initial Ag uptake was transcellular. The incidence of HRP-containing endocytic vesicles throughout enterocytes was greater in sensitized vs naive rats. However, the absence of mast cells in Ws/Ws rat intestine did not reduce either the amount (as measured by area) or the incidence of HRP in various regions of the cell. These findings imply that the total amount of HRP transported across the epithelium was enhanced by sensitization but was not influenced by the presence of mast cells.

Bockman and Winborn (21) reported that the sensitization of hamsters to ferritin up-regulated the intestinal absorption of ferritin after luminal exposure. We previously documented that the endocytic transport of protein was enhanced only for the Ag to which the rats had been sensitized and not for an irrelevant Ag. Taken together, the results of Bockman and Winborn and those from our studies provide support for the recognition of Ag by ECs, possibly by surface-bound Ig in sensitized animals (for further discussion see Ref. 6 and a related editorial, 22 . This study was conducted to examine whether mast cells, which are known to bind Ag via IgE and high affinity FcεRIRs and are occasionally described within the epithelium (23, 24), are involved in the early phase of Ag transport that occurs before the hypersensitivity reaction. Our findings of a similar amount of HRP uptake and distribution in sensitized Ws/Ws and +/+ rats rule out a contribution of mast cells to this initial phase of transepithelial Ag transport.

In mast cell-containing intestine from HRP-sensitized rats, a large increase in the flux of intact HRP was observed beginning at 30 min after the hypersensitivity reaction. The increased HRP flux was associated in time and correlated with increased tissue conductance, which is a measure of the integrity of the tight junctions. This observation was supported by electron photomicrographs showing HRP in the paracellular regions. A number of studies in both sensitized animal models (9, 25) and allergic patients (26, 27) have shown that intestinal permeability to small m.w. probes and “bystander” Ags increases following Ag challenge. In contrast to the specificity observed with the initial transcytosis of Ag, this phase of Ag transport appears to be a nonspecific permeability defect. We have observed that OVA-sensitized Sprague Dawley rats that have been challenged with luminal OVA also develop an increased luminal to serosal flux of HRP (M.C.B., unpublished observations), confirming the lack of specificity of this Ag transport pathway. The increased HRP flux and conductance we observed in HRP-sensitized +/+ and Sprague Dawley rats along with the presence of HRP in the paracellular regions and within tight junctions suggests a recruitment of the paracellular pathway. Heyman et al. also showed an increased intestinal flux of HRP after Ag challenge across biopsies from children with cow’s milk allergy (28) and across intestinal segments from sensitized guinea pigs (29), but in an apparent absence of an alteration in tissue conductance. No electron microscopy analysis was conducted in those studies to examine the route of HRP transport.

As indicated above, HRP challenge to mast cell-containing intestine from sensitized rats resulted in an elevated Isc that was followed by increased conductance and a flux of HRP. However, transport parameters in intestine from sensitized Ws/Ws mast cell-deficient rats were indistinguishable from those in intestine from unsensitized +/+ controls. This result indicates that both mast cells and sensitization are required for the later phase of nonspecific Ag transport. The role of mast cells in the regulation of epithelial ion secretion has been well-studied, and it has been clearly established that mast cell mediators such as histamine and prostaglandins can act via specific receptors on the intestinal epithelium to initiate chloride ion secretion (7). Although a number of studies have reported that Ag challenge in sensitized animals produces an increase in intestinal epithelial permeability, this study is the first to show directly that the development of the Ag-induced barrier defect is mast cell-dependent. It should be noted that c-kit deficiency can have effects on cell populations other than mast cells. Intraepithelial lymphocytes (IELs) also express the c-kit receptor, and interactions with stem cell factor-producing intestinal ECs may be important for the normal development of these IELs. Although IEL populations have not been examined in Ws/Ws rats, W/W^v mast cell-deficient mice demonstrate age-dependent changes in IEL subsets. As mice age, there is a decrease in the percentage of TCRγδ IELs and an increase in TCRαβ IELs in W/W^v mice compared with +/+ mice (30). As there is not a dramatic depletion of IELs in c-kit-deficient animals, it is unlikely that the IELs are responsible for the recruitment of the paracellular Ag transport pathway that we have observed in +/+ but not Ws/Ws rats. However, it cannot be completely ruled out that an alteration in IEL function in Ws/Ws rats may play a role in transepithelial Ag transport. Reconstitution experiments with a pure population of mast cells obtained from +/+ bone marrow would most likely confirm that mast cells are responsible for the increased Ag transport observed after the hypersensitivity reaction. However, due to the heterogeneous genetic background of the Ws/Ws and +/+ animals (F₂ generation of two inbred rat strains), reconstitution is not feasible.

Although the mechanism of a mast cell-induced increase in Ag transport has not been explored in the current study, it has been shown that a number of mast cell products such as cytokines, prostaglandins, nitric oxide, and proteases can alter epithelial permeability (31). Alternately, mast cells could potentially regulate tight junction permeability indirectly through nerves. It has been shown that nerves are activated by mast cells or specific mast cell mediators and may act to amplify the hypersensitivity response (32). Stimulating intestinal tissue with a cholinergic agonist, carbachol, has been shown to increase the paracellular permeability to HRP (5, 33); it has also been demonstrated that stress-induced barrier disruption is mediated by cholinergic nerves (34). We have previously shown that pretreating intestinal segments with tetrodotoxin prevents an Ag-induced increase in ⁵¹Cr-EDTA flux in OVA-sensitized rats (25). Therefore, it is possible that mast cells recruit the paracellular transport pathway for Ag via the activation of enteric nerves.

In conclusion, our study suggests that transepithelial Ag transport occurs in two distinct phases in sensitized rat jejunum. We have provided evidence that in phase I, Ag is initially taken up rapidly and transported across the epithelium by an endocytic mechanism that is enhanced by sensitization but is independent of the presence of mast cells. In phase II, Ag transport is sensitization- and mast cell-dependent and leads to a large flux of Ag across the epithelial barrier; this flux most likely occurs through the recruitment of the paracellular pathway. These findings suggest that in an allergic individual, even small amounts of Ag within the lumen can be preferentially transported to the LP, where the subsequent activation of mast cells further induces a nonspecific barrier defect. This sequence of events may be extremely important in initiating and sustaining allergic inflammation not just in the gastrointestinal tract but at all mucosal sites in the body.

We thank Michelle Benjamin, Greet Scholten, and the staff of the McMaster University Faculty of Health Sciences Electron Microscopy Unit for excellent technical assistance. We are grateful to the staff of the Health Sciences Central Animal Facility for their help in establishing and maintaining the Ws/Ws rat breeding colony.