In previous studies, we found that apical and basolateral EGF receptors (EGFR) on primary canine gastric monolayers decreased paracellular permeability, evident by increased transepithelial electrical resistance (TER) and decreased flux of [3H]mannitol (MF). After studying monolayers in Ussing chambers, we now report that treatment with apical, but not basolateral, EGF enhanced tolerance to apical H+, evident by a slower decay in TER and an attenuated rise in MF. Enhanced tolerance to apical acid was evident within 10 min of treatment with apical EGF. Immunoneutralization of endogenous transforming growth factor (TGF)-α accelerated the drop in TER and the rise in MF in response to apical acidification; apical EGF reversed these effects. Study of monolayers cultured in Transwell inserts showed that immunoblockade of basolateral, but not apical, EGFR also impaired the resistance to apical acidification and enhanced MF. We conclude that apical EGFR regulates the barrier to apical acidification via effects on paracellular resistance. Although exogenous basolateral EGF has a less apparent effect on the barrier to acid, endogenous ligand active at basolateral EGFR plays an important role in maintaining the barrier to apical acid. Our data implicate a role for an apical EGFR ligand, which may be EGF or another member of the EGF family.
- peptic ulcer
- paracellular pathway
- tight junctions
- cell polarity
multiple mechanisms account for the ability of the gastric mucosa to withstand acid and pepsin concentrations sufficient to digest most tissues. These mechanisms include three lines of defense: 1) mucus and HCO secretion (preepithelial), 2) epithelial cell mechanisms, and 3) postepithelial element due to removal of H+ and supply of substrate via mucosal blood flow. Three lines of repair (epithelial restitution, cell replication, and wound healing) (13, 14, 30, 35) are also important. HCO secretion into an unstirred layer stabilized by mucus establishes a pH gradient over the gastric surface epithelial cells that protects intracellular pH (10). However, mucosal cells lining the gastric glands withstand very high apical H+ and pepsin concentrations without the benefit of a mucus layer, implying that intrinsic cell mechanisms are critically important for glandular cells. These cellular mechanisms include the barrier to H+ diffusion provided by the apical surface, the extrusion of excess H+ entering cells via basolateral Na+/H+ and HCO /Cl− exchange, and possibly other intracellular processes, such as induction of heat shock proteins (23) or glutathione-dependent mechanisms (25).
Our previous studies (29) with monolayers formed from dispersed canine gastric mucosal cells indicated that the apical surface constituted an important component of the epithelial barrier to apical H+. We used transepithelial electrical properties and mannitol (3.5 Å diameter) and inulin (11–15 Å diameter) as probes (4, 19) to investigate the paracellular pathway. We found that this barrier function reflected two modes of interaction of H+ with epithelial cell tight junctions. First, an increase in transepithelial electrical resistance (TER) occurs as the apical pH is decreased between 6.5 and ∼3.5 (depending on the specific experimental model); this increased TER was accompanied by an decrease in [3H]mannitol flux (MF), indicating that apical H+ decreased paracellular permeability. Second, with exposure to high apical concentrations of acid, TER falls and MF, but not [14C]inulin flux, increases, indicating a paracellular leak, with preservation of the size-sieving function of tight junctions (19). Finally, after sustained exposure to apical pH of <3.5 for Transwell inserts and pH of <2.5 for Ussing chambers, transepithelial integrity is lost, marked by a decrease to fluid TER and a parallel increase in both mannitol and inulin fluxes (4), indicating loss of tight junctional size sieving as monolayer integrity is disrupted. In contrast, acidification of the basolateral surface to a pH of <5.5 leads to a rapid loss of monolayer integrity and cell viability (4). Although the exact mechanisms remain undefined, these findings confirm that the apical surface of gastric epithelial cells constitutes a major element of the barrier to the diffusion of luminal H+ and suggest that tight junctions account for the initial increase in TER and are the site of earliest injury after apical acidification. Studies with gastric glands have confirmed the very high resistance of the apical surface of gastric glandular cells to acidification (33,34).
The above findings implicated the tight junction as the critical component of an apical membrane that otherwise had very low permeability to H+. Few important physiological processes are passive, and, therefore, it was not unexpected when we found that the paracellular pathway was regulated by EGF (5). Furthermore, both apical and basolateral EGF receptors (EGFR) increased TER and decreased MF, indicating that the change was in paracellular permeability (5). Other investigators have found that the permeability of the paracellular pathway is regulated (3, 21, 24,32), although these studies with EGF provide the best example of receptor-mediated regulation.
Transforming growth factor (TGF)-α is present in parietal cells in intact mucosa (2) and in our cultures of canine gastric glandular cells (6). Furthermore, in our cultures, endogenous EGFR ligand, presumably TGF-α, exerts control over migration (16) and thymidine incorporation (6). In the present study, we examined the effects of apical and basolateral EGF on the apical barrier to acid and sought to determine whether endogenous EGFR ligand was involved in this regulatory process. Our data indicate that these growth factor effects on paracellular permeability translate into enhanced mucosal tolerance to apical acidification, suggesting an action of physiological relevance to the maintenance of gastric mucosal integrity in vivo.
MATERIALS AND METHODS
Tissue Dispersion, Cell Separation, and Monolayer Culture
Canine oxyntic mucosal cells were dispersed using sequential treatment with collagenase and EDTA, with minor modifications from previous methods (4, 5). Cells (1.4 × 109) were loaded into a Beckman elutriator rotor (model JE-5.0) and washed at 2,000 rpm with a 100 ml/min flow rate, and fractions were collected at 1,460 rpm with a 200 ml/min flow rate. For our initial studies, cells were plated on collagen-coated Millipore cellulose filters (type SSWP; Millipore, Bedford, MA) stabilized to the bottom of culture dishes (1). Polarized monolayers formed on these filters, but the apical and basolateral solutions were not separated until the filters were mounted in the Ussing chambers. Most of our subsequent studies were performed by using collagen-coated Transwell inserts (Costar, Cambridge, MA); filters are suspended in the inserts, which are placed in a culture plate. After confluent monolayers have formed, these inserts allow separate manipulation of apical and basolateral solutions. In both systems, cells were cultured in basic growth medium (R0: DMEM/F12, 20 mM HEPES, and 100 mg/ml amikacin) with 2% calf serum. Cultures were fed with the same medium every 48 h until confluent monolayers were formed. Monolayers were then switched to R0 medium without calf serum in the presence or absence of antibodies or growth factors, as indicated.
Electrophysiological and Acid Barrier Studies with Epithelial-Volt-Ohm-Meter and Ussing Chambers
For studies with Millipore filters, monolayer confluence was achieved 60–72 h after plating and was established using phase contrast microscopy to examine the cells around the edge of the filter. In contrast with the Transwell inserts, monolayer formation was monitored using the epithelial-volt-Ohm-meter (EVOM; Millicells-ERS, Millipore) to measure TER. After monolayers formed, medium was replaced with serum-free R0; all studies with growth factors and antibodies were done in this serum-free medium. EVOM electrodes were fixed to a derrick-like stand to ensure that the TER measurement was performed when electrodes were placed in an identical position in the well. The electrodes were sterilized by soaking in 70% EtOH and equilibrated with R0 before the experiment. TER was determined as [(resistance of the monolayer) − (resistance of blank collagen-coated insert)] × cm2 (area of the insert filter) and expressed as Ω · cm2.
For initial studies using Ussing chambers, confluent monolayers after 72–120 h of cultures were mounted in a bathing solution of Hanks' balanced salt solution (10 mM HEPES, pH 7.4). Current was monitored continuously with the voltage clamped to zero. TER was calculated by Ohm's law from the voltage deflection obtained with application of a 5-mV pulse and expressed in Ω · cm2. As described earlier (4), apical pH was adjusted by adding aliquots of 2 N HCl and actual pH was confirmed with a micro electrode (Microelectrodes, Londonerry, NH) on small aliquots of bathing solution.
For studies using Transwell inserts, apical pH was adjusted by adding aliquots of 0.5 N HCl to the inside of the inserts (apical medium). TER was measured with EVOM at the indicated time point.
Monolayers were cultured in R0 on Millipore filters; just after achieving confluence, filters were treated with either control serum (1:125 dilution) or sheep polyclonal antibody to TGF-α (1:125 dilution; Biogenesis, Sandown, NH). After 24 h, monolayers were mounted in Ussing chambers and vehicle or EGF (10 nM) were added apically or basolaterally, as indicated. Electrophysiological and acid barrier studies were monitored in Ussing chambers as described above. Nonimmune goat serum was used as control.
Anti-EGFR antibody Studies
EGFR antibody MAb-528 (Ab-1, Oncogen Research Products/Calbiochem, Cambridge, MA) was used at 4 μg/ml, a concentration that produces 91 ± 7% (n = 4) blockade of 125I-EGF binding to these cell preparations (5). Studies with MAb-528 were only done in Transwell filter inserts after confluence was established by EVOM. After monolayers were formed, MAb-528 was added either apically or basolaterally (4 μg/ml) for the indicated time period. EGF was then added, as indicated, and TER was monitored using EVOM. We (16) previously performed controls for antibody specificity in our cultures: MAb-528 blocks the action of EGF/TGF-α on thymidine incorporation and cell migration but has no effect on the actions of unrelated peptides, such as insulin-like growth factor-I or basic fibroblast growth factor.
MF studies were performed in Ussing chambers, as previously described (4); 0.5-mM cold mannitol was added to the initial apical and basolateral bathing solutions, and, after an 1-h equilibration time, [3H]mannitol (5 μCi/ml) was added to the apical side (t = 0). Aliquots (200 μl) were taken from the basolateral solution at the indicated times, and the apical-to-basolateral MF was expressed in nmol/cm2. In the Transwell inserts, mannitol flux studies were done as above, except that 5 μCi of [3H]mannitol was added to apical solution at t = 0 and aliquots of 50 μl were taken from the basolateral solution at the indicated times. Data were expressed as counts/min × 1,000.
Data are expressed as means ± SE, with nequal to the number of separate cell preparations. Statistical significance was determined by repeated-measures ANOVA methods. This analysis was computed using the MIXED procedure (SAS Institute, Cary, NC), under the expert guidance of Dr. Jeff Gornbein (Statistical/Biomathematical Consulting Clinic, Biomathematics Department, University of California at Los Angeles). As mentioned in the figure legend of a few studies, the statistical significance of differences was assessed using a paired Student's t-test.
Studies of EGF on Barrier Function in Ussing Chambers
Apical, but not basolateral, EGF enhances monolayer tolerance to apical acidification.
We studied the effect of EGF on apical resistance to acid using monolayers formed on collagen-coated filters and mounted in Ussing chambers. As previously observed (4), apical acidification to a pH of 2.5 caused a small rise and then a fall in TER (Fig.1). Also, as previously reported, apical EGF treatment increased TER (5). However, apical EGF also increased the time that monolayer integrity was maintained in the face of apical acidification (Fig. 1). In five consecutive control experiments when the apical pH was decreased to 2.5, TER fell to 50% of the initial value in 60 ± 9 min. In contrast, after apical EGF treatment, a 50% decrease in TER was not reached for 149 min (n = 5, P < 0.05, Fig. 1 B). With basolateral EGF treatment, the time decay in TER with apical acidification was not altered from control monolayers (n = 5, P > 0.2, data not shown).
A similar pattern for TER change was found in a second set of experiments in Ussing chambers in which the apical pH was lowered to pH 2.0. EGF increased the duration of time that monolayers could withstand apical acidification (Fig.2 A); however, the effect was smaller and not significant in these four experiments. In addition, despite the apical acidification, MF across these monolayers was very low in the presence or absence of EGF (Fig. 2 C). These latter experiments underscored the remarkable resilience of monolayers cultured in serum-free medium (R0) to relatively high concentrations of apical acid. Because of this high innate resistance to apical acidification, effects of EGF on the apical barrier were more easily demonstrated when endogenous TGF-α was immunoneutralized, as described below.
TGF-α antisera impairs the barrier to apical acidification.
Our previous studies with these canine monolayers demonstrated effects of endogenous TGF-α on cell replication and migration (7,16). Although these studies were done with antiserum, the specificity for TGF-α was demonstrated at the doses used (16). For our present studies, we hypothesized that immunoneutralization of TGF-α or immunoblockade of endogenous EGFR ligand would make monolayers more sensitive to apical acidification.
These experiments were performed in Ussing chambers using monolayers cultured on collagen-coated filters and treated with sheep antisera to immunoneutralize TGF-α. After achieving confluence, monolayers were incubated overnight with either control sera (nonimmune sheep sera) or anti-TGF-α sera and then mounted in Ussing chambers. TGF-α antiserum was not added to the bathing solutions of the Ussing chambers. After TGF-α immunoneutralization, monolayers more rapidly deteriorated in the face of apical exposure to pH 2.0; a dramatic fall in TER (Fig. 2, B vs. A) and rise in MF (Fig. 2,D vs. C) were observed.
In monolayers exposed to TGF-α antisera, apical EGF induced a rapid enhancement of monolayer integrity in the face of an apical pH of 2.0 (Fig. 2 B), as evidenced by a delayed fall in TER and rise in MF (Fig. 2 D). In contrast, the acute addition of EGF to the basolateral solution did not produce significant restoration of the barrier function to apical acidification (n = 3,P > 0.2, data not shown).
Studies of EGF on Barrier Function Studying Monolayers in Transwell Inserts
Apical EGF and apical acidification produce a sustained, synergistic effect on TER.
To examine the effects of apical acid and EGF treatment over sustained periods, monolayers grown on Transwell inserts were studied using EVOM chopsticks under sterile conditions. In previous studies with Ussing chambers, we found that apical acidification over a pH range from 6 to 2.5 increased TER and decreased MF (4). A similar phenomenon was consistently observed when studying monolayers grown on Transwell inserts over a pH range of 6.5 to 3.75 (data not illustrated,n = 7). However, monolayers in Transwell inserts were more sensitive to acid; pH of <3.0 produced a rapid and irreversible fall in TER (n = 10, data not illustrated). With the Transwell inserts at a pH range from 3.0 to 3.6, the TER response to apical acid was somewhat variable, especially in the first 60 min of exposure. The degree of apical mixing in a Transwell insert is significantly reduced due to the small volume inside, and the lack of quick mix when acid is added could change the sensitivity. Despite these limitations, Transwell inserts allowed long-term studies under sterile conditions and separate manipulation of apical and basolateral solutions, experiments that were not possible in Ussing chambers.
When monolayers were simultaneously exposed to both apical acid (pH 4.0) and apical EGF treatment, we found a synergistic effect on TER (Fig. 3, A and B). In this illustrated set of experiments examining the increment in TER over basal, TER in response to the combination of apical acidification and apical EGF treatment was greater than the sum of the individual TER values with apical EGF alone and apical acidification alone (n = 4, P < 0.05). In other experiments with lower apical pH (3.7 to 3.2), the early TER response to apical acid was more variable, as noted above, and this early synergy was not consistently observed (n = 3,P > 0.05, data not shown).
We (5) previously observed that treatment with basolateral EGF produced a transient increase in TER and a transient decrease in MF. In our present studies, basolateral EGF also produced a transient effect on TER (Fig. 3 C). Furthermore, basolateral EGF did not produce enhancement of the TER response to apical acid (Figs.3 D and 4 A).
The ability of apical EGF to enhance the TER response to apical acid was evident over prolonged exposure periods. Over exposures from 2 to 36 h, the combination of apical EGF and apical acidification produced a dramatic enhancement of TER compared with acid alone or EGF alone (Figs. 3, C and D and 4 A). For example, 3 h after treatment at pH 3.5, TER in response to apical EGF was increased by 79.8 ± 37.1% over baseline, compared with an untreated control of 5.7 ± 20%, and basolateral EGF was 4.2 ± 24.2% over basal (mean ± SE, n = 5,P < 0.05, Fig.4 A). This same magnitude of difference was evident for periods ≤36 h (Fig. 4 A). Although studies with an apical pH of 4.0 allowed consistent demonstration of an early enhancement of the TER response to apical acidification by apical EGF, the ability of apical EGF to enhance TER in the face of apical acidification was also evident at pH 3.5 over more prolonged periods.
MF studies were essential to confirm that the effects of apical EGF on the TER response to apical acidification reflected effects on the paracellular pathway (Fig. 4 B). The addition of apical EGF produced a marked decrease in the MF induced by apical acidification. These effects of apical EGF were sustained over 20 h. In contrast, although basolateral EGF suppressed MF, the effects were only evident in the first few hours of treatment, consistent with a transient effect on TER (Fig. 4 B). These studies confirm that the effects of EGF on epithelial resistance to apical acidification reflect effects on paracellular permeability.
Basolateral EGFR antibodies impair barrier function in Transwell inserts.
We studied monolayers on Transwell inserts to further define the role and sidedness of endogenous EGFR ligands in supporting barrier function to apical acidification. We tested the effects on barrier function of EGFR immunoblockade with MAb-528, a monoclonal antibody to human EGFR. The specificity of the MAb-528 in our system was established in prior studies (16). Treatment with basolateral MAb-528 alone for 7 h lowered initial TER by ∼40% and attenuated the TER rise in response to an apical pH of 3.5 (Fig.5 A). In MAb-528-treated monolayers, TER in the face of apical acidification was consistently lower than found in untreated monolayers (n = 4,P < 0.05, Fig. 5 A). In parallel experiments, MF under basal conditions was slightly increased by MAb-528 alone (P < 0.05, n = 4). However, pretreatment with MAb-528 markedly increased MF in response to apical acidification to pH 3.5 (Fig. 5 B).
In contrast to these findings with basolateral MAb-528, treatment with apical MAb-528 did not alter baseline TER or the response to apical acidification (data not shown, n = 3, P> 0.2). These data indicated that blockade of the basolateral, but not apical, EGFR leads to a deterioration of barrier function. The data suggest that an EGFR ligand, presumably TGF-α, acts at basolateral receptors to support monolayer resistance to apical acidification.
Coincubation of apical EGF (10 nM) with basolateral MAb-528 effectively reversed the deleterious effects of basolateral EGFR immunoblockade (Fig. 6 A). At the end of this 7-h preincubation period, TER was numerically, but not statistically, higher in the group treated with apical EGF (Fig. 6 A). Subsequent apical acidification to pH 3.5 produced an initial decrease in TER in all groups, reflecting the early variability observed in this system. However, the group treated with apical EGF showed a significant increase in TER 1 h after the acidification, and this increase was sustained for a prolonged period (30 h; Fig. 6 A). This protective effect in TER by apical EGF was mirrored by decrease in MF that would otherwise be observed in response to apical acidification (Fig. 6 B). Basolateral EGF also reversed the effects of basolateral MAb-528, producing an intermediate level of MF and TER in response to apical acid (Fig. 6).
Apical EGFR Regulates the Apical Barrier to Acid.
Our data demonstrate that apical EGF increases monolayer TER and decreases MF, indicating regulation of paracellular permeability. Apical EGF also increases epithelial monolayer integrity and barrier function, as evidenced by an attenuated fall in TER and rise in MF in response to apical acidification. Although basolateral EGF induced a transient increase in TER and a decrease in MF, effects on barrier function to apical acidification were much less pronounced than those produced by apical EGFR activation. Because the effects of apical EGF were reflected in both increased TER and decreased MF, we conclude that the paracellular pathway mediates these effects of EGFR. Furthermore, these actions of EGF occurred at nanomolar concentrations, which are in the range of that observed in gastric juice (17, 26).
Apical EGFR Has Both Rapid and Sustained Effects on Barrier to Acid
Effects of apical EGFR on TER were observed within minutes of treatment. We speculate that enhanced barrier function in response to apical EGF is also a rapid action of EGF.
Cell activation mechanisms mediating these acute regulatory effects of apical EGF have not been defined. In our studies with regulation of paracellular permeability by apical acid exposure and treatment with basolateral secretin and apical EGF, we have found consistent effects of these treatments on tyrosine phosphorylation of β-catenin and γ-catenin (5, 8) (M. C. Chen, E. Rozengurt, and A. H. Soll, unpublished observations). Because the catenins bridge the cytoplasmic tail of E-cadherin to α-catenin and the apical cytoskeleton, they are reasonable candidates for modulating junctional permeability. This possibility requires direct testing. Our data also indicate that EGF enhances the barrier to acid for a period ≤36 h. In this more prolonged period, different cellular mechanisms are likely to come into play, such as protein synthesis or assembly of junctional components.
Endogenous EGFR Ligands Support Monolayer Integrity
Our data indicate that endogenous EGFR ligands in our cultures influence barrier function. Immunoneutralization of TGF-α markedly impaired the barrier to acid, and treatment with apical EGF rapidly reversed these effects; these studies were done using cells cultured on filters lightly fixed to the bottom of the petri dishes, so that polarized treatment was not possible in culture. Additional studies were done using cells cultured on Transwell inserts. In this model, basolateral, but not apical, treatment with the monoclonal antibody to EGFR lowered baseline TER, increased MF, and also impaired the barrier to apical acidification. TGF-α has been found in parietal cells (2), and we have found secretion of TGF-α into the culture media of comparable cultures of canine gastric mucosal cells (6). Taken together, our findings are consistent with TGF-α being released at the basolateral surface of our cultures and exerting important effects on monolayer integrity. Dempsey and Coffey (11) reported secretion of TGF-α at the basolateral surface of MDCK epithelial cells.
The finding that basolateral TGF-α plays an important role in maintaining epithelial integrity may appear inconsistent with our findings that basolateral EGF treatment does not acutely enhance apical barrier function. However, these experiments were done with endogenous EGF/TGF-α present in the basolateral solution. We surmise that removal or immunoblockade of this endogenous ligand interferes with essential housekeeping functions mediated by EGFR. It is of interest that, in our model, treatment with apical EGF can replace the loss of the basolateral EGFR ligand effect.
Studies of immunoneutralizing TGF-α were done using antisera at a dilution ∼1:125. The specificity of this concentration of anti-TGF-α sera was established in previous studies based on binding to monolayers (6) and mitogenic and migrational effects on these cultures (16). As noted previously (16), several attempts to purify this antibody or develop other immunoneutralizing antibodies by our group and by others have been unsuccessful. We propose that our present experiments are interpretable because of the numerous controls indicating that this concentration of antisera was specific for TGF-α (16). Our present findings demonstrating reversal by exogenous EGF are also consistent with actions mediated by immunoneutralization of TGF-α.
In these experiments in Ussing chambers, TGF-α-antiserum was not added to the bathing solutions of the Ussing chambers because of the large volumes (6 ml for the apical and basolateral solutions, respectively) and because of the short-term nature of the studies. We reasoned that in this short time period, it would be unlikely for the concentrations of endogenous EGFR ligand to build up to sufficient concentrations to obviate differences. We surmise that the persisting effects of TGF-α antiserum reflected immunoneutralization of TGF-α during the culture period and the dilution of endogenous TGF-α secreted in the large bathing solutions during the short period of these experiments in the Ussing chambers. However, it is also possible that some TGF-α antibodies remained bound to membrane-associated TGF-α precursor.
Although immunoneutralizing TGF-α would remove a mitogenic influence on these monolayers, other mechanisms seem more plausible as an explanation for the effects. Previous studies indicated that immunoneutralizing TGF-α inhibited [3H]thymidine incorporation into DNA, with this effect evident in studying cultures plated at low cell densities before confluence was achieved (6). In addition, we (16) previously demonstrated that hydroxyurea (20 mM) inhibited EGF/TGF-α-stimulated thymidine incorporation in the subconfluent cultures. However, it did not impair EGF-stimulated migration in wounded confluent monolayers nor change the basal TER of the confluent monolayers. In the present study, high cell plating densities were used so that confluence was achieved 48–60 h after plating. The addition of TGF-α antiserum during the 24-h period after confluence was reached did not significantly reduce cell number compared with control monolayers (cell number in TGF-α antiserum-treated monolayers was 93 ± 3% of control, n = 3, P > 0.1). Therefore, the trophic effects of endogenous TGF-α are minimal in confluent monolayers. Furthermore, TER was increased and barrier function was restored toward control levels within minutes after apical but not basolateral treatment with EGF. Therefore, we conclude that a component of the EGF effects on paracellular permeability reflect rapid regulatory events that are independent of the mitogenic actions. We speculate that mitogenic effects primarily reflect occupation of the more dominant basolateral receptors. Our present studies do not evaluate the possibility that apical EGFRs also induce the synthesis and assembly of tight junctional components.
In contrast to the effects of basolateral EGFR blockade, blockade of apical EGFR did not alter monolayer barrier properties. These findings indicate that, in these cultures, endogenous EGF ligand is released into the basolateral, but not apical, solutions. However, in vivo, EGF family members are present in the gastric lumen, including EGF itself (18, 20, 26). Although EGF is digested into smaller forms in gastric juice, activity persists (28), and EGF is one candidate for exerting physiological effects on apical EGFR. Several other factors that induce mitogenesis are also present in gastric juice (15). There is no shortage of candidates. However, the identity and physiology of the putative apical EGFR ligand remains unknown.
An Enhancement Between the Actions of Apical EGF and Apical Acidification
Gastric monolayers respond to apical acidification at pH values above ∼3.5 with increased TER (4). Apical EGF treatment enhances rise in resistance and monolayer integrity in response to apical acid. In contrast, when monolayers are treated with antisera to TGF-α or with basolateral MAb-528, the rise in resistance seen with apical acidification was attenuated. Therefore, apical EGF induces rapid effects that enhance the rise in resistance in response to apical acidification. However, in addition, a tonic level of endogenous TGF-α at basolateral receptors appears to support the ability of the apical membrane to respond to acidification with a rise in resistance.
Synergistic interactions between EGF and apical acid exposure suggest cell activation by different signaling pathways. Recent data support this notion. Apical acid exposure in the absence of added EGF induces tyrosine phosphorylation of Src and Pyk2 (8). Furthermore, both the decrease in paracellular permeability and phosphorylation effects of apical acid are Src dependent (sensitive to Src kinase inhibitors) (8). This mechanism appears to be independent of EGFR activation, which is not Src dependent (9). These data are consistent with our observations that EGF enhances the effects of apical acid (present study) and of secretin (Chen and Soll, unpublished observations) on paracellular permeability. In contrast, secretin and apical acid do not exhibit such synergistic interactions.
Are Apical EGFR Real and of Physiological Importance?
Luminal EGF does have effects on mucosal repair (27). Furthermore, EGFRs have been found on apical membrane preparations (31), on the apical surface of intestinal cells (12), and on the apical surface of trophoblastic syncytium (22). Other studies (27) have not detected apical EGFR on enterocytes. We have several lines of evidence supporting the existence of functional EGFR in our gastric epithelial monolayers. Binding studies (5) indicate that specific apical receptors are present at a density of ∼5% of basolateral receptors. Immunoblockade of EGFR by the EGFR antibody MAb-528 displaced binding from both apical and basolateral sites and blocked EGF-induced increases in TER at both apical and basolateral receptors (5). These studies were performed in polarized systems where the apical and basolateral solutions were separate; the addition of basolateral anti-EGFR blocked basolateral, but not apical, binding. Despite the considerable differences in receptor density and in EGFR autophosphorylation, apical and basolateral EGFR activation produced roughly comparably phosphorylation of β-catenin (5). Furthermore, of the several receptors we have studied, we have only detected EGFR on the apical surface; receptors for secretin, insulin-like growth factor-I, fibroblast growth factor, and muscarinic agonists are localized to only the basolateral surface (M. C. Chen and A. H. Soll, unpublished observations). Therefore, we are confident that the presence of apical EGFR is not simply an artifact of the system or nondiscriminate missorting of basolateral receptors.
Our present data also provide considerable additional evidence for the existence of apical EGFR. Apical EGF enhanced the barrier to acid in a fashion not mimicked by basolateral EGF. Furthermore, apical EGF reversed the deleterious effects of basolateral MAb-528 on TER in a more effective fashion than did basolateral EGF treatment.
We thank the late professor John Harley Walsh for his belief in our perseverance. We also are indebted to Tom Gonzales and the Gonzales Family Foundation; without his timely support this work would not have been completed.
This study was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-30444 and DK-19984, by the Medical and Research Services of the Department of Veterans Affairs, and by a generous gift of The Gonzales Family Foundation.
Address for reprint requests and other correspondence: A. H. Soll, Veterans Affairs Wadsworth Hospital Center, Bldg. 115, Rm. 215, 11301 Wilshire Blvd., Los Angeles, CA 90073 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
July 31, 2002;10.1152/ajpgi.00507.2001
- Copyright © 2002 the American Physiological Society