Long-Term Repeated Daily Use of Intragastric Gavage Hinders Induction of Oral Tolerance to Ovalbumin in Mice

Abstract

Oral tolerance is dependent on the complex architecture of the mucosal system of the gastrointestinal tract, its associated lymphoid tissue, and specialized immune cells. Changes in this architecture or the failure of any of its components may hinder the generation of oral tolerance. The larynx and esophagus are the gateway to the gastrointestinal tract, serving as the site of oral antigen introduction to the immune system and may have an important role in establishing tolerance. Intragastric gavage is a common method for precise oral dosing of rodents, particularly in studies examining oral tolerance. However, complications such as esophageal trauma can occur and induce complicating factors that affect experimental outcomes. In this study, we examined the esophageal epithelium for alterations resulting from long-term repeated daily use of intragrastric gavage and its effect on the induction of tolerance. Tolerance to ovalbumin could not be achieved after using intragastric gavage for 14 d or more consecutively to introduce ovalbumin. However, tolerance was achieved when intragastric gavage was used for shorter durations. After 14 d of gavage, disruption of the esophageal mucosal epithelium indicative of an inflammatory pathology, cellular influx into the esophageal tissue, and proinflammatory cytokines in the tissue were absent, and the CD3⁺ cell population in the esophageal epithelium decreased. These findings provide initial evidence for the important roles of esophageal integrity and cellular populations in the induction of oral tolerance and suggest possible immunologic sequelae in experiments involving the use of extended, repeated gavage.

Intragastric gavage is a commonly used method to precisely administer oral solution in rodents. Gavage techniques vary due to the variety of materials available, such as ball-tipped needles and flexible tubing, and the use (or nonuse) of lubricants. Despite the ability to customize the gavage method, problems with intragastric gavage can occur, especially when used long-term⁵ or when prolonged daily administration is required.¹³ These complications can include increased stress in the animals (as indicated by increased plasma corticosterone levels), aspiration pneumonia, unintentional tracheal administration, esophageal trauma, and gastric rupture.^3,5,12,13,16

Intragastric gavage is commonly used in studies of oral tolerance, the active suppression of an immune response to fed antigen upon subsequent challenge with that same antigen. The induction of oral tolerance involves the intestinal epithelial barrier, the gut-associated lymphoid tissue, and specialized immune cells present in intestinal mucosal tissues.¹⁸

The intestinal epithelium forms a selective barrier that absorbs necessary nutrients and allows for molecule transport yet concomitantly limits interactions with food antigens and commensal organisms. Intestinal permeability primarily results from the transport of materials through tight junctions between adjacent intestinal epithelial cells (paracellular diffusion) or through transcytosis (transcellular transport). Increased intestinal permeability, resulting from dysregulated mechanisms of paracellular or transcellular transport or other alterations to the epithelial barrier, can result in immune alterations, leading to diseases such as food allergies, celiac disease, and intestinal bowel disorders.¹¹ In addition to barrier permeability, the access of antigens within the intestinal lumen to the immune system is regulated by the components of the mucosal membranes, which line the surface of the gastrointestinal tract. These components include unique structures, such as Peyer patches, and cells, such as M cells, specialized antigen-presenting cells, and CD25⁺CD4⁺Foxp3⁺ regulatory T cells, that aid in generating tolerance responses.^10,18

The larynx is the gateway to the gastrointestinal tract and therefore serves as one of the first sites of antigen introduction to the immune system. Recent examination of the epithelial cells of the human larynx have revealed the presence of unique expression patterns of antigen-presenting molecules within the laryngeal tissue.¹⁷ A high amount of MHC class I was observed in the deep layers of tissue, but MHC class I expression decreased and CD1d levels increased on the superficial layers of laryngeal tissue. Second, the epithelial cells contain MHC class II, which is normally present only on dedicated antigen-presenting cells. The decreased expression of MHC class I molecules in the larynx as compared with other areas such as the spleen, as well as the presence of MHC class II antigens in the absence of costimulatory molecules aids to establish a tolerant atmosphere to the barrage of external antigen it receives from the outside environment. The esophagus is similarly positioned for early contact with oral antigens, yet even less is known about its immunologic architecture or its role in initiating immune responses. However, given its location and the interplay and trafficking of cells between mucosal tissues, the esophagus, too, may have a role in establishing tolerance toward fed antigens.

Repeated and chronic use of intragastric gavage may disrupt the cellular and mucosal tissue architecture of the larynx and esophagus, consequently potentially inducing immunologic consequences beyond known complications of intragastric gavage. The goal of this study was to investigate the hypothesis that daily gavage for 14 d or longer disrupts esophageal mucosal tissue, initiating an inflammatory reaction that supersedes the induction of tolerance to fed antigen. To investigate this hypothesis, mice were either fed ovalbumin via intragastric gavage using a ball-tipped needle or received the liquid dropwise into the mouth via a syringe daily for 14 d. Tolerance to ovalbumin was determined by assessing relative levels of ovalbumin-specific IgG in blood serum by ELISA. Changes in esophageal cell populations were assessed via immunohistochemistry, and the severity of damage from intragastric gavage was assessed by examining sections of esophageal tissue.

Materials and Methods

Mice.

BALB/cJ mice (8 to 12 wk) bred from mating pairs purchased from The Jackson Laboratory (Bar Harbor, ME) were used for each study. Mice were housed conventionally in open-top caging containing bedding (Tek-Fresh Laboratory Animal Bedding, Harlan, Indianapolis, IN), fed a commercial diet (Teklad Global Diets Rodent Chow, Harlan), given unlimited access to water, and tested regularly for pathogen status. The room environment was a 12:12-h light:dark cycle with a temperature of 20.0 to 22.2 °C (68 to 72 °F). Prior to each study, mice were housed individually in cages and separated into 4 treatment groups, with no fewer than 3 mice per group and with equal numbers of each sex between groups. All procedures involving mice were in accordance with NIH guidelines and approved by the Ball State University Animal Care and Use Committee.

Treatments, feedings, and time-point studies.

Four treatment groups were used for each tolerance experiment: nongavage, syringe-fed water only (negative control for tolerance response); nongavage, syringe-fed ovalbumin (positive control for tolerant response); gavage-fed water only, and gavage-fed ovalbumin. Gavage-fed mice were fed via intragastric gavage using a ball-tipped 18-gauge feeding needle (SouthPoint Surgical Supply, Coral Springs, FL) without lubrication for 14 d daily (total, 14 feedings) or every other day (total, 7 feedings). Nongavage, syringe-fed control mice were restrained identically to mice being fed via intragastric gavage but were fed by using a syringe without a feeding needle: the tip of the syringe was placed into the mouth of the mouse, and solution was administered dropwise. This feeding method required a holding time of 5 min, which was longer than the approximate 15-s holding to administer liquid via gavage.

These same feeding and treatment methods were used for the time-point studies, where mice in 2 groups (nongavage, syringe-fed water only, and gavage-fed water only) were treated for 1, 3, or 7 d. Untreated, control (day 0) mice were not handled or treated. After each indicated time point, mice were euthanized, and esophageal tissue was harvested and examined histologically.

Induction of oral tolerance.

Mice were fed water or 3 mg ovalbumin in a total of 200 μL daily for a total of 14 d by using either the syringe or intragastric gavage method as described previously. Half of the mice from each group were euthanized via CO₂ asphyxiation at 24 h after the final feeding, and esophageal tissue was harvested for histologic analysis.

The remaining mice were challenged via intraperitoneal immunization with ovalbumin (0.1 mg in 200 μL of 50% alum solution) at 1 and 2 wk after the final feeding. At 1 wk after the second immunization, mice were euthanized via CO₂ asphyxiation, blood was collected by cardiocentesis, serum was isolated, and levels of IgG were assessed by ELISA). This treatment plan is outlined in Figure 1 A. For comparison, Figure 1 B demonstrates the treatment plan involving nonconsecutive gavage treatments that consistently resulted in tolerance to ovalbumin.

Figure 1.

Experimental design. BALB/cJ mice (age, 8 to 12 wk) were fed 200 μL water or ovalbumin (OVA; 3 mg in water) (A) daily or (B) every other day for 14 d. At 24 h after the final feeding, half of the mice from each group were euthanized, and esophageal tissue was harvested for immunohistochemical analysis (only in the consecutive treatment studies denoted in Figure 1 A). Five days after the final feeding treatment, the remaining mice were challenged with intraperitoneal (100 μg) ovalbumin in a 50% alum solution, with a second challenge 1 wk later. One week after the second challenge, mice were euthanized, and blood serum was collected for analysis.

Systemic ovalbumin studies.

Treatment groups remained the same for this experiment: nongavage, syringe-fed water only; nongavage, syringe-fed ovalbumin; gavage-fed water only; and gavage-fed ovalbumin. Mice were fed water or 3 mg ovalbumin (+ ovalbumin) in a total of 200 μL daily via intragastric gavage or nongavage methods for a total of 14 d. Mice were euthanized via CO₂ asphyxiation, 14 d after the final feeding, blood was collected by using cardiocentesis, and serum was isolated. Levels of ovalbumin-specific IgG and IgM were determined by using an indirect ELISA (described following). As a control to demonstrate the generation of an immune response to ovalbumin, an additional group of mice were immunized on day 0 (intraperitoneal injection with 0.1 mg ovalbumin in PBS and 50% alum solution in a total volume of 200 μL) but were not otherwise treated. Serum isolated from these mice was pooled and used as a positive control for immunoglobulin ELISA analyses.

ELISA.

Levels of ovalbumin-specific IgG and IgM present in the serum were measured by using an indirect ELISA. Briefly, 96-well plates were coated with 100 μL ovalbumin (Sigma-Aldrich, St Louis, MO) in coating buffer (4 mg/mL; Bethyl Laboratories, Montgomery, TX) and placed at 4 °C overnight. Plates were then washed 3 times in wash buffer (Bethyl Laboratories) between all steps. Plates were blocked for 30 min with blocking solution (Bethyl Laboratories). Serum from each mouse was diluted 1:100, 1:500, 1:2500, and 1:12500; added to the plate; and incubated for 2 h at room temperature. After sample addition, plates were incubated with 100 μL rat antimouse IgG (dilution, 1:1000; conjugated with alkaline phosphatase, human adsorbed, Southern Biotech, Birmingham, AL) or 100 μL goat antimouse IgM (dilution, 1:5000; conjugated with horseradish peroxidase, Bethyl Laboratories). To catalyze the alkaline phosphatase enzymatic reaction, 100 µL of p-nitrophenylphosphate substrate solution (4 mg/mL, Sigma-Aldrich) was added to each well. For the horseradish peroxidase reaction, 100 µL tetramethylbenzidine substrate (Bethyl Laboratories) was used for detection, followed by the addition of stop solution. Samples were analyzed in duplicate by using a microplate reader (model 680, Bio-Rad, Hercules, CA). The absorbance was measured at 450 nm for samples analyzed using horseradish peroxidase with tetramethylbenzidine and absorbance was measured at 405 nm for analyzed using alkaline phosphatase with p-nitrophenylphosphate. Lines graphs were generated from dilutions of 1:100, 1:500, 1:2500, and 1:12500 to demonstrate decreasing absorbance, indicating the assay was working appropriately. Samples were considered to be positive when the absorbance value of the sample was at least 2 times the absorbance value of negative controls. Absorbance values corresponding to the 1:100 serum dilution for each sample were used for statistical analysis and graphing.

Tissue harvesting.

After the completion of described treatments, the upper thoracic segment of the esophagus, below the oropharyngeal cavity, was harvested from each mouse. Tissue was placed in a protective cassette and stored in 10% neutral-buffered formalin at room temperature for 6 to 8 h. After fixation, tissue cassettes were stored in 70% ethanol until embedding, processing, and staining could be performed.

Immunohistochemistry.

Immunohistochemistry was performed on the esophageal cross-sections to assess the infiltration of immune cells within the tissue and to identify any inflammation that might result from intragastric gavage. Hematoxylin and eosin staining was performed to assess total immune cell infiltration in the submucosa and identify any alterations of the esophageal mucosal tissue after examination by a pathologist (Dr Tom Kocoshis, Indiana University Health). The chloroacetate esterase or Leder stain was used to determine the presence of granulocytes in the submucosa and epithelial layer. T cells in the submucosa and epithelial layer were identified by using and antiCD3 antibody and diaminobenzidine stain. All stains were provided and performed by the Indiana University School of Medicine Immunohistochemistry Laboratory (Indianapolis, IN). In addition, an immunostain for major basic protein was performed (laboratory of Dr Marc Rothenberg) to assess eosinophils within the esophageal tissue; the antibody to myelin basic protein was a kind gift of Drs Jamie and Nancy Lee (Mayo Clinic, Scottsdale, AZ). Cells were counted by using a 10 μm × 10 μm grid. Within this grid, 5 different quadrants along the mucosal–epithelial layer of the esophageal tissue were assessed and cells were counted, to get the best representation of the tissue condition. The average of these cells counts were taken to determine the number of cells per μm². All cell counts were performed by a blinded investigator.

Reverse transcription and quantitative PCR.

Total RNA was prepared from esophageal tissue with the use of Trizol reagent (Life Technologies, Carlsbad, CA). cDNA was synthesized by using iScript cDNA Synthesis Kit (Bio-Rad), and the samples were diluted in water (1:10). A total volume of 25 μL containing 5 μL cDNA template, 0.3 μM of each primer, and SYBR Green PCRMaster Mix (Applied Biosystems, Carlsbad, CA) was analyzed in triplicate. Gene expression was analyzed by using an ABI PRISM 7300 Sequence Detector and ABI Prism Sequence Detection Software version 1.9.1 (Applied Biosystems). Expression of TNFα, IL1β, and IL6 was analyzed. Results were normalized by dividing the value for the tested gene by that obtained for β-actin. The primers used were: β-actin–F, 5′ TGG AAT CCT GTG GCA TCC ATG AAA G 3′; β-actin–R, 5′ TAA AAC GCA GCC TCA GTA ACA GTC CG 3′; IL1bF, 5′ GGA CAG AAT ATC AAC CAA CAA GTG AT 3′; IL1bR, 5′ ATT ACA CAG GAC AGG TAT AGA TTC T 3′; IL6F, 5′ CCG GAG AGG AGA CTT CAC AG 3′; IL6R, 5′ TTC TGC AAG TGC ATC ATC GT 3′; TNFαF, 5′ CCC AGA CCC TCA CAC TCA GAT C 3′; and TNFαR, 5′ CAG AGT AAA GGG GTC AGA GTG GGG G 3′.

Statistical analysis.

Results from immunohistochemistry cell counts were analyzed by using a Mann–Whitney rank sum test to determine statistical significance. All ELISA results were analyzed by using one-way ANOVA with Holm–Sidak posthoc analysis. A P value less than 0.05 was considered significant for all experiments. All statistical analyses were performed by using SigmaStat 11 software (Systat Software, San Jose, CA).

Results

Extended treatment with intragastric gavage abrogates the induction of oral tolerance.

In our model, mice with low levels of serum ovalbumin IgG are considered to have developed tolerance, whereas serum ovalbumin IgG levels in nontolerant mice are higher and similar to those of water-fed mice. In the current study, the nongavage-fed ovalbumin mice had low levels of ovalbumin IgG, whereas mice that were fed 3 mg ovalbumin in water via intragastric gavage daily for 14 d consecutively and then challenged with ovalbumin had serum levels of ovalbumin IgG that were similar to those of the water-fed nontolerized control mice. (Figure 2). The normal levels of ovalbumin IgG in mice gavaged daily for 14 d therefore indicates that these mice had not developed tolerance. Conversely, using intragastric gavage to administer 3 mg ovalbumin every other day over 14 d (total, 7 feedings) in water followed by ovalbumin challenge resulted in decreased production of serum ovalbumin-specific IgG in the gavaged mice, indicative of the induction of tolerance (Figure 2 B). The total amount of ovalbumin differed between mice gavaged consecutively (42 mg) or every other day (21 mg); however, this should not affect the ability of the mice to be tolerized to ovalbumin because oral tolerance can be achieved after the administration of various concentrations of oral antigen.⁴ Similarly, our preliminary data demonstrated that mice receiving either 1 mg or 3 mg of ovalbumin via gavage for 7 consecutive days or 20 mg of ovalbumin once weekly for 2 wk followed by ovalbumin challenge similarly had lower levels of serum ovalbumin IgG compared with control mice gavage-fed water only. Because tolerance to ovalbumin was induced by a variety of antigen concentrations but only after relatively few intragastric administrations, the data suggest that repeated consecutive daily use of an 18 gauge ball-tipped feeding needle without the use of lubrication may hinder the induction of tolerance to the fed antigen.

Figure 2.

Oral tolerance induction to ovalbumin is successful with syringe feeding but not intragastric gavage. Mice were divided into 4 treatment groups (n = 3 to 6 per group): nongavage,-fed H₂O (negative control for tolerant response); nongavagesyringe-fed ovalbumin (positive control for tolerant response), gavage-fed H₂O, and gavage-fed ovalbumin. Mice received water (filled cirlces) or 3 mg ovalbumin in water (open circles) (A) consecutively for 14 d or (B) every other day for 14 d. At 5 and 12 d after feeding, all mice were challenged with an intraperitoneal injection of 0.1 mg ovalbumin in alum solution. At 7 d after the second ovalbumin challenge, blood serum was isolated, and serum levels of ovalbumin-specific IgG were assessed by indirect ELISA. Data (mean ± 1 SD) are representative of 4 experiments. Different symbols represent significant (P < 0.05, one-way ANOVA with Holm–Sidak posthoc analysis) differences in values. *, Value for nongavage-fed + ovalbumin significantly less than that for nongavage-fed water control; #, Value for gavage-fed + ovalbumin significantly less than that for gavage-fed water control.

Repeated daily use of intragastric gavage does not induce an influx of cells into the esophageal epithelium.

To assess the presence of inflammation in the esophagus (as demonstrated by an influx of cells into the tissue), mice were euthanized 24 h after the 14th consecutive day of feeding, and esophageal tissue was harvested for staining. Tissues were stained with hematoxylin and eosin, and cells within the submucosa were enumerated. No significant difference in the total number of cells (mean ± SEM) in the esophageal submucosa was observed between day 0 control mice (2.04 ± 0.11 cells/µm²); nongavage, syringe-fed mice (2.15 ± 0.08 cells/µm²); and mice fed via intragastric gavage (2.05 ± 0.06 cells/µm²). Furthermore, no gross physical alterations indicative of inflammatory pathology occurred in any of the treatment groups.

Extended use of intragastric gavage results in a small amount of systemic antigen exposure.

Introduction of antigen into the circulatory system could result in the activation of a systemic immune response, potentially hindering the generation of a tolerant response. To assess the immunologic consequences of this possibility, mice were fed water or 3 mg ovalbumin by using either intragastric gavage or nongavage, syringe-fed methods daily for 14 d. However, ovalbumin immunizations were not given after the completion of treatment regimens. At 2 and 14 d after the completion of the feedings, serum was isolated, and levels of ovalbumin IgG and IgM were determined. Whereas ovalbumin-specific IgM and IgG were present in the immunized control group, they were not detected in any of the treatment groups at 2 d after the completion of feeding treatments (data not shown). At 14 days after the feeding treatments, serum ovalbumin IgG levels did not exceed background in any of the treatment groups (Figure 3 B), and the water-fed control groups and the nongavage ovalbumin-fed group all lacked noteworthy levels of ovalbumin IgM. However, slightly elevated levels of ovalbumin IgM were detected in mice fed ovalbumin by gavage (Figure 3 A). Due to the large variation in ovalbumin IgM levels among mice gavage-fed ovalbumin, this elevation of ovalbumin IgM was not significantly higher than that of water-only controls, nor was it significantly lower than values from the immunized control mice. Taken together, these data suggest that fed antigen is not being introduced directly into the bloodstream in sufficient quantity to elict a strong antibody response.

Figure 3.

Extended use of intragastric gavage results in a small amount of systemic antigen exposure. Mice were divided into 4 treatment groups (n = 3 per group): nongavage, syringe-fed water; nongavage, syringe-fed ovalbumin; gavage-fed water; and gavage-fed ovalbumin. Mice were received water or 3 mg ovalbumin in water every day for 14 d. At 2 wk after the final feeding, blood serum was isolated, and serum levels of (A) ovalbumin-specific IgM and (B) ovalbumin-specific IgG were determined by indirect ELISA. A separate group of naïve mice was immunized intraperitoneally with 0.1 mg ovalbumin in alum solution. Blood serum was collected 7 d after immunization, pooled, and assessed for serum levels of ovalbumin-specific IgM and IgG levels (immunized control). *, Value significantly (P < 0.5, Kruskal–Wallis one-way ANOVA with Holm–Sidak post hoc analysis) lower than that for the immunized control.

Total granulocyte and T cell populations are altered after feeding treatments.

To investigate a mechanism that explains the diminished induction of tolerance to fed antigen in mice fed via intragastric gavage, we analyzed the infiltration of specific cell populations into esophageal tissues. The presence of increased numbers of CD3⁺ cells and granulocytes can be an indicator of inflammation. To determine whether alterations in specific populations of CD3⁺ T cells and granulocytes correlated with intragastric gavage treatments and an inability to induce tolerance, the embedded esophageal tissues that were used previously for histopathology were analyzed. To identify CD3⁺ T cells, the embedded tissues were stained by using an antiCD3 antibody and diaminobenzidine, whereas granulocytes were identified by using the Leder stain.

In contrast to our expectation that esophageal tissue would be inflamed in mice fed via intragastric gavage and thus have an influx of cells, the stained tissues from animals after 14 d of consecutive treatments had fewer CD3⁺ T cells when mice were fed via intragastric gavage as compared with nongavage, syringe-fed mice and untreated (day 0) controls (Figure 4 A). The esophageal tissues of nongavage, syringe-fed mice showed significantly more granulocytes than did tissues of mice fed via intragastric gavage and untreated (day 0) controls (Figure 4 B). Furthermore, eosinophils are not normal residents of the esophagus, and their presence can indicate disease.¹⁵ To identify whether eosinophils contributed to the slight increase in granulocytes seen in the syringe-fed mice, tissues were examined by morphometric analysis using an antibody against myelin basic protein (Figure 4 C). However the total number of eosinophils was similar between mice fed via intragastric gavage and nongavage, syringe-fed mice, indicating that the change in granulocyte count was not driven by eosinophil migration to the esophagus.

Figure 4.

Levels of T cells and total granulocytes but not eosinophils are altered in the submucosa and epithelium of the esophagus after feeding treatments. Mice were divided into 2 groups (n = 14 per group): nongavage-fed and gavage-fed. Mice received feedings via intragastric gavage or nongavage, syringe-fed methods every day for 14 d. At 24 h after the final feeding, esophageal tissue was harvested for (A) immunohistochemical analysis of T cell infiltration (antiCD3 antibody and diaminobenzidine stains), (B) granulocyte infiltration (Leder stain), and (C) eosinophilic infiltration (stained by using an antibody against myelin basic protein). On bar graphs, horizontal lines indicate average cell counts from day 0 mice (untreated mice). Data from 2 experiments were pooled. *, Value for nongavage, syringe-fed mice significantly (P < 0.05, Mann–Whitney rank-sum test) different from that for gavage-fed mice.

To investigate the occurrence of any alterations in esophageal cell populations throughout the 14-d study, esophageal tissues were harvested at days 1, 3, and 7 and examined for the presence of granulocytes and CD3⁺ cells. Numbers of CD3⁺ cells did not differ significantly between gavage-fed and nongavage, syringe-fed mice from day 0 to day 7 (Figure 5 A), nor did the granulocyte population in the esophagus of gavage-fed mice increase over time (Figure 5 B), suggesting that the gavage treatment did not induce inflammation of the esophageal tissue. Furthermore, examination of esophageal tissue at days 0, 1, 3, and 7 for the presence of the proinflammatory cytokines IL1β, IL6, and TNFα by qRT-PCR revealed no differences in cytokine levels between gavage- and nongavage-treated mice (data not shown). However, on day 3, there was a significant increase in granulocyte count in the esophageal tissue of nongavage, syringe-fed mice (Figure 5 B).

Figure 5.

Granulocyte but not CD3⁺ cell populations fluctuate in the submucosa and epithelium of the esophagus between 1 and 7 d of feeding treatments. Mice were divided into 3 groups (n = 8 to 13 per group): untreated (day 0), nongavage, and gavage-fed. Mice were exposed to the intragastric gavage (open circles) or nongavage, syringe-fed (filled circles) method consecutively for a maximum of 7 d. At days 1, 3, and 7, esophageal tissue was harvested (A) immunohistochemical analysis of T cell infiltration (antiCD3 antibody and diaminobenzidine stains) or (B) granulocyte infiltration (Leder stain) at each timepoint. Data from 2 experiments were pooled. *, Value for nongavage, syringe-fed mice significantly (P < 0.05, Mann–Whitney rank-sum test) different from that for gavage-fed mice.

These data fail to demonstrate an influx of cells indicative of inflammation in the esophageal tissue of mice fed via intragastric gavage. However, other cellular alterations, such as the observed reduction in CD3⁺ T cells within the esophageal tissue after 14 d of gavage may correlate with an inability to induce tolerance. Taken together, these results reveal a potentially important role for the cellular composition of the esophageal mucosa in inducing tolerance to fed antigen.

Discussion

This study examined how prolonged use of intragastric gavage might alter esophageal mucosal tissues and subsequent immune mechanisms necessary for inducing tolerance to fed antigens. The data from this study demonstrate that oral introduction of ovalbumin was sufficient to induce oral tolerance in nongavaged mice. However, the use of intragastric gavage either daily or every other day over 14 d hindered the induction of oral tolerance to the fed ovalbumin, such that levels of ovalbumin IgG in mice fed ovalbumin via intragastric gavage were comparable to those of mice gavaged with water only (Figure 2 A). Furthermore, findings from this study demonstrate fluctuations of cellular populations within the esophageal mucosal, which may provide insight into the tissue perturbations that may contribute to an inability to induce tolerance.

Intragastric gavage is a commonly used method due to the ability to introduce a solution both quickly and accurately. Problems with intragastric gavage include increased stress, as indicated by increased plasma corticosterone levels,^3,12,16 and even the development of granulomas after long-term use.⁵ One aspect of this study was to identify whether any trauma or physical alterations to the esophageal epithelium indicative of inflammation resulted from extended use of intragastric gavage and whether such breakdown in the integrity of the esophageal epithelium resulted in systemic exposure of ovalbumin that prevented the induction of tolerance to ovalbumin. Although no physical damage of the esophageal tissue after gavage was observed, serum ovalbumin IgM levels were increased among mice gavaged with ovalbumin but not challenged with ovalbumin (Figure 3 A). This small increase in ovalbumin IgM resulted from variable results in the 3 mice assessed for serum levels of ovalbumin IgM, among which only 1 of the 3 mice had levels of ovalbumin IgM that were within the range of ovalbumin IgM in the immunized controls. This result supports previous findings indicating that approximately 30% of mice fed radiolabeled antigen via gavage demonstrated the antigen in the circulatory system within 4 h of gavage.² The authors of the cited study² concluded that antigen accessed the circulatory system more likely via aspiration and exchange with blood in the lungs than as a result of disruption in the distal esophageal mucosa. Our findings support that conclusion, given the lack of observed pathology demonstrating a break in the integrity of the esophageal epithelium. Our study adds to these findings by showing that the amount of antigen introduced systemically was sufficient to induce some IgM production but that it was insufficient to induce a strong immune response that involved antibody class switching (Figure 3 B), which typically occurs after challenge with ovalbumin.

A second aspect of this study was to investigate the possibility that use of a feeding needle to perform intragastric gavage for an extended period of time induces inflammation in the esophageal tissue, stimulating an immune response that is not tolerant to the fed antigen. In contrast to our anticipated results, the esophageal submucosa of gavage-fed mice did not demonstrate an influx of cells; however CD3⁺ T-cell populations were reduced after intragastric gavage for 14 d (Figure 4 A). One potential explanation for the significant decrease in CD3⁺ cells is that it is a result of a reduction in γδ T cells. γδ T cells have an important regulatory role in oral tolerance through their maintenance of the barrier function of the epithelium and control of intestinal epithelial cell turnover; they also have a protective role in the colitis induced experimentally in mice by using dextran sodium sulfate.¹ This crosstalk between intestinal epithelial cells and γδ T cells has recently been shown to be crucial for routine maintenance and repair of the intestinal epithelium, which is important in establishing tolerance to fed antigen.⁸ Perhaps the damage caused by the intragastric gavage disrupts the balance between intestinal epithelial cells and the γδ T cell population, in turn disrupting their normal crosstalk and homeostasis. This explanation would be multifaceted, however, given that γδ T cells have inflammatory properties as well.¹ Our data provide preliminary information that warrants continued investigation in an appropriate esophageal inflammatory model and additional immunohistochemistry to investigate CD3⁺ T cell subtype present in the esophageal tissue.

Natural killer cells are another subpopulation of CD3-expressing T cells and require CD1d for antigen presentation. Recently natural killer cells have been shown to be involved in the induction of oral tolerance to ovalbumin, by inducing IL10-producing T regulatory cells and aiding in the clonal deletion of ovalbumin-specific CD4⁺ T cells.⁷ Given the presence of natural killer cells and CD1d within the esophagus¹⁴ and their role in establishing oral tolerance, the reduction in CD3⁺ T cells that we observed may be due to decreased numbers of natural killer T cells.

Alternatively, completely removing surface epithelial cells from both the small and large intestine of humans allows the migration of large amounts of T cells from their normal residence in the lamina propria.⁹ Perhaps prolonged use of a feeding needle results in similar damage to the esophagus, allowing for the egress of T cells from the tissue and causing the disparity in T cell counts between syringe- and needle-fed mice. However, it should be noted that our gavaged mice showed no significant or obvious esophageal damage. Given these possibilities, the underlying mechanism responsible for the observed fluctuations in our results likely is complex.

Studies examining tolerance often involve the assessment of T-cell function. The experimental model in which we did this investigation limited our assessment of tolerance to the presence or absence of antigen-specific IgG antibodies. Although the measurement of IgG levels is an indirect assessment of T cell function, given the requirement of T and B cell interactions during activation for antibody class switching, our findings provide the basis for investigation into alterations in tolerance as demonstrated by alterations in T cell function both within lymphoid and epithelial tissues. Using ovalbumin-specific T-cell receptor transgenic mice, which are commonly used in tolerance experiments, will allow the investigation of ovalbumin-specific T cell functions, such as cytokine production and proliferation. Findings from such studies would yield additional insight into the specific tolerance mechanisms that are affected by extended, repeated gavage treatment.

Interestingly, our experiments revealed a slight increase in the esophageal granulocyte population on days 3 and 14 in nongavage, syringe-fed mice compared with the gavage-fed group and day 0 control mice (Figures 5 B). Physical stress caused by strenuous exercise in humans results in an increase in granulocytes as well as fluctuations in lymphocyte makeup within the blood, and these changes correspond to an increase in plasma catecholamine levels.⁶ The prolonged restraint of the mice during nongavage feeding treatments may have caused a similar response and accounts for the significantly increased granulocytes within the esophagus of those mice. Although we were unable to examine levels of catecholamine in this study, this is an important line of investigation. Importantly, the ability to induce tolerance to fed ovalbumin in these nongavage-fed mice demonstrates that the granulocyte increase is not detrimental to the process of tolerance induction.

Taken together, these data demonstrate an inability to tolerize mice to ovalbumin fed via intragastric gavage for 14 d or more and highlight alterations in the esophageal cellular composition that occur as a result of intragastric gavage treatments. Although this study did not reveal a specific subset of cells responsible for the loss of tolerance induction, intragastric gavage clearly causes pronounced fluctuations in the CD3⁺ cell population within the esophageal mucosal tissue. This study provides initial evidence of the importance of esophageal integrity and cellular populations in oral tolerance induction and highlights the possibility of immunologic complications for experiments involving the use of extended, repeated gavage.