Prolonged ingestion of ovalbumin diet by sensitized mice improves the metabolic consequences induced by experimental food allergy

Abstract

The prevalence of food allergy is rising in the western world. Allergen restriction is the chosen treatment in this condition, but continuous ingestion of the antigen has shown positive results in clinical trials. In a previous study, we have shown several allergic and metabolic alterations after 7 days of ovalbumin (OVA) ingestion by sensitized mice. The aim of this study was to investigate whether prolonged ingestion of antigen by sensitized mice would reverse the metabolic consequences caused by experimental food allergy. For this, allergic and metabolic parameters were analysed after prolonged ingestion of an OVA diet by OVA-sensitized mice. As shown previously, after 7 days of OVA consumption, sensitized mice showed increased serum levels of anti-OVA immunoglobulin (Ig)E and IgG1, aversion to the antigen ingestion, marked body and adipose tissue weight loss, followed by adipose tissue inflammation and decreased serum levels of adipokines, glucose and triglycerides. However, after 14 days of oral challenge, sensitized mice showed an anti-OVA IgE level similar to the mice that were only sensitized, but the specific IgG1 did not change. With this prolonged ingestion of OVA, sensitized mice were protected from OVA-induced anaphylaxis when the antigen was given systemically at a dose of 2 mg/animal. Moreover, various parameters analysed were significantly ameliorated, including adipose tissue inflammation, body and adipose tissue loss, as well as serum levels of adipokines and triglycerides. Therefore, our data suggest that prolonged ingestion of OVA by sensitized mice results in an improvement of the metabolic consequences caused by experimental food allergy.

Introduction

Food allergy is an adverse immune response to dietary proteins 1, affecting up to 1–3% of adults and 3–8% of infants in westernized countries, where prevalence rates are currently increasing 2,3. Immunoglobulin (Ig)E-mediated food allergy, also known as type I food allergy, is responsible for the majority of food allergic reactions. Several factors responsible for the development of IgE-mediated food allergy have been identified, such as genetic predisposition, age at which food antigen is introduced and composition of the gut microbiota 4,5. The typical immune response to dietary proteins is the induction of oral tolerance 6, while a breakdown or failure to induce this mechanism can result in allergic food reactions 7. However, even when allergic sensitization occurs, oral tolerance can be re-established 8,9. This phenomenon is the base of oral immunotherapy (OIT), a therapeutic application of oral tolerance in allergy that has gained increasing attention and is showing good results in early-phase clinical trials 10–12.

In order to propose new therapies and study the mechanisms involved in food allergy, the utilization of experimental models is an extremely important tool. Therefore, our group has developed a murine model of food allergy to ovalbumin (OVA) in which sensitized mice are challenged orally for 7 days. In this model, several signals similar to those that occur in patients with food allergy are developed, such as antigenic aversion, increased anti-OVA IgE production, intestinal oedema and intestinal eosinophil infiltration, as well as a marked weight and adipose tissue loss 13,14. In addition, we have demonstrated that this allergic process induces an adipose tissue inflammation with an increase in the number of inflammatory cells and proinflammatory cytokine levels in this tissue, resulting in systemic metabolic alterations including the decrease in the serum glucose, triglycerides and total cholesterol 14. Interestingly, we have also demonstrated that prolonged ingestion of OVA after sensitization for 14 days reduces specific serum IgE levels and the number of CD4⁺, CD8⁺ and CD4⁺CD25⁺forkhead box protein 3 (FoxP3)⁺ T blood cells when compared to sensitized mice after 7 days of OVA oral challenge; this prolonged ingestion of antigen is able to suppress non-related pathological conditions such as airway inflammation 15 and antigen-induced arthritis 16. Furthermore, this suppressive effect on airway inflammation persisted even when the oral challenge with OVA by sensitized mice was stopped for 7 days before the animals were submitted to aerosol challenge 15. However, the allergic-induced inflammatory changes in adipose tissue and consequent metabolic alterations were not studied after prolonged ingestion of antigen after sensitization.

Therefore, once prolonged ingestion of OVA by sensitized mice leads to an immunological suppression of the allergic response, the aim of this study was to evaluate if the metabolic alterations caused by food allergy would also be diminished.

Materials and methods

Experimental protocol design

Figure 1 shows a detailed overview of the experimental protocol used in our study. First, animals were subjected to two rounds of OVA sensitization with posterior oral challenge with the same antigen (OVA⁺ groups). Mice from control groups (OVA⁻ groups) received only adjuvant or saline in the days of sensitization and during the period of oral antigenic challenge they also received a diet containing OVA. After 7 or 14 days of oral challenge, animals were anaesthetized and prepared for intravital microscopy. After this procedure, mice were euthanized and adipose tissue was collected for histology and cytokine measurement. The blood was collected for cell counting and the serum was used for both adipokine and specific immunoglobulin evaluation. In the group that received the OVA diet for 7 days, we also analysed the adipocyte glucose uptake. Another group of animals were euthanized before the oral antigenic challenge, on day 0, to evaluate whether sensitization interferes with anti-OVA IgE and IgG1 levels. Other animals were used for anaphylaxis induction before (0 day), 7 or 14 days after OVA oral challenge and the response was analysed through the haemodynamic changes. Also, fasted mice were used after 7 or 14 days of oral challenge in order to determine the serum levels of glucose, triglyceride and insulin and also to perform the oral glucose tolerance test.

Fig. 1.

Overall experimental procedures. Mice from ovalbumin (OVA)⁺ groups were sensitized with OVA in Al (OH)₃ on day −21 and with OVA on day −7. The control groups (OVA⁻) received Al (OH)₃ on day −21 and saline on day −7. On day 0 the standard chow was replaced with a 14% OVA diet for 7 or 14 days, depending on the groups. The procedures were performed as described in this figure. Animals were euthanized for blood and adipose tissue collection as specified.

Animals

Male BALB/c 6–8-week-old mice were obtained from the animal facility of the Federal University of Minas Gerais (CEBIO/UFMG). All mice received standard mouse chow (Purina, Ribeirão, SP, Brasil) until the beginning of antigen challenge with the OVA diet. The investigations were made according to the Ethical Principles in Animal Experimentation of our institution and the experimental protocol was approved by the Ethics Committee in Animal Experimentation of the University (protocol 085/11-CETEA/UFMG).

OVA sensitization and oral challenge

Mice from control groups (OVA⁻) were sensitized with a subcutaneous (s.c.) injection of 0·2 ml saline (0·9%) with 1 mg Al(OH)₃ as adjuvant, while mice from sensitized groups (OVA⁺) received s.c. injection of 0·2 ml saline (0·9%) with adjuvant and 10 μg OVA (five times crystallized hen's egg albumin; Sigma, St Louis, MO, USA). Two weeks later, mice from the OVA⁺ groups received a booster sensitization with a s.c. injection of saline with 10 μg OVA and mice from the OVA⁻ groups received only saline. One week after this secondary sensitization, the standard chow was replaced with a 14% OVA diet for 7 or 14 days, depending on the groups. However, the day 0 group was euthanized before this oral OVA challenge. This diet was prepared using a lyophilized egg white (Salto's, Belo Horizonte, MG, Brazil) and nutrient contents were in accordance with AIN-93G 17.

Induction of anaphylaxis

Anaphylaxis was induced before (day 0), 7 or 14 days after OVA oral challenge by a single intraperitoneal (i.p.) injection of 0·2 ml with 10 μg/μl of OVA (2 mg/mice) or 0·2 ml with 100 μg/μl of OVA (20 mg/mice). Mean arterial pressure (MAP), blood volume and blood flow in the tail were evaluated after 30 min of i.p. challenge.

Haemodynamic measurements using the tail cuff method

Haemodynamic changes followed by anaphylactic shock were analysed using a very sensitive technique. Specifically, mean arterial pressure (MAP), blood volume and blood flow in the tail were evaluated by a volume pressure recording sensor and an occlusion tail cuff system, which measures mice tail blood pressures non-invasively (Kent Scientific Corporation, Torrington, CT, USA). Mice were acclimated to the restraint and to tail cuff inflation for 1 day before the beginning of the experiments. The restraint platform was maintained at 32–35°C. At each session mice were placed into an acrylic box restraint, and the tail was inserted into a compression cuff that measured the blood pressure 10 times. Following the measurement cycle, the average of these values was considered for each mouse. All these parameters were evaluated 30 min after induction of anaphylaxis.

Animal evaluations and tissue collection

Body weight was determined weekly throughout the experiment and daily during the OVA challenge. During the oral challenge period, OVA diet consumption was assessed daily by weighting the remaining chow and comparing its weight with the previous day. The measurements were reported as a percentage of diet consumption (OVA group/control group), considering the diet consumption by the control group to be 100%. Before starting the oral challenge (day 0 group) and after 7 or 14 days of receiving the OVA diet, mice were anaesthetized by i.p. injection of 10 mg/kg xylazine and 100 mg/kg ketamine hydrochloride and blood were obtained from all groups. Later, under deep anaesthesia, mice were euthanized and the epididymal fat was collected.

Evaluation of serum OVA-specific IgE and IgG1

The enzyme-linked immunosorbent assay (ELISA) test for anti-OVA IgE was performed using plates coated with rat anti-mouse IgE, serum and biotinylated OVA, as described previously 18. The reaction was developed with streptavidin–peroxidase conjugate (ExtraAvidin; Sigma), o phenylene-diamine and H₂O₂. Anti-OVA IgG1 antibodies were determined using plates coated with OVA, 100 μl of a 1:1600 dilution of mouse sera and biotinylated goat anti-mouse IgG1 (Southern Biotechnology Associates, Birmingham, AL, USA). The reactions were developed with streptavidin–peroxidase conjugate (ExtraAvidin; Sigma), o-phenylene-diamine and H₂O₂. The plate was read at 492 nm on an automated ELISA reader. The results were reported in arbitrary units (AU) using 1000 units as a positive reference serum. For this positive reference serum, a pool of serum obtained from sensitized mice that received the OVA diet for 7 days was used; this pool was considered as a positive control once these mice showed higher levels of anti-OVA IgE and IgG1 in comparison to the other groups. Moreover, the same serum pool was used during all experimental settings. All antibodies were acquired from Southern Biotechnology Associates.

Histological analysis of epididymal adipose tissue

Samples from epididymal adipose tissue were fixed in phosphate-buffered formaldehyde for 24 h, dehydrated in absolute ethanol, cleared in xylene and then embedded in paraffin. Histological sections (5 μm) were stained with haematoxylin and eosin (H&E) and then evaluated by light microscopy (Olympus BX41; Olympus, Center Valley, PA, USA). Images of four fields from each animal were captured using a digital camera coupled to a microscope (Olympus BX41) and the area of 50 adipocytes per animal was measured using ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Intravital microscopy of the epididymal adipose tissue microvasculature

Mice were anaesthetized by i.p. injection of 10 mg/kg xylazine and 100 mg/kg ketamine hydrochloride. The right jugular vein was cannulated and rhodamine 6 G (Sigma) was injected intravenously (i.v.; 1·5 mg/kg) to label the leucocytes and endothelial cells. Rhodamine epi-illumination was achieved with a 150 W variable HBO mercury lamp in conjunction with a Zeiss filter set 15 (546/12 nm band-pass filter, 580 nm Fourier transforms, 590 nm late potentials; Zeiss, Wetzlar, Germany). The microscopic images were captured using a Nikon eclipse 50i (Nikon Instruments Inc., Tokyo, Japan) microscope (×20 objective) with a video camera (5100 HS; Panasonic, Secaucus, NJ, USA) and recorded digitally using both filter blocks consecutively. Data analysis was performed off-line. Rolling leucocytes were defined as cells passing through a given point in the venule per minute. Leucocytes were considered adherent to the venular endothelium if they remained stationary for at least 30 s or longer within a 100 μm length of venule. The two parameters analysed were measured in two or three different vessels and averaged for each animal.

Total and differential blood cell counts

Blood was collected and total white blood cells were counted with a Neubauer chamber. Blood smears were stained with May–Grünwald–Giemsa stain and the differential white blood cell count was determined using standard morphological criteria.

Cytokine measurements in adipose tissue

Epididymal adipose tissue extracts obtained during the necropsy were homogenized in extraction solution (100 mg of tissue per 1 ml) containing 0·4 M NaCl, 0·05% Tween 20, 0·5% bovine serum albumin (BSA), 0·1 mM phenylmethylsulphonyl fluoride, 0·1 mM benzethonium chloride, 10 mM ethylenediamine tetraacetic acid (EDTA) and 20 KIU aprotinin. The levels of tumour necrosis factor (TNF)-α and interleukin (IL)-6 were measured by ELISA in supernatants of epididymal adipose tissue homogenate using DuoSet ELISA kits and according to the instructions provided by the manufacturer (R&D Systems, Inc., Minneapolis, MN, USA).

Determination of serum parameters

Serum fasting glucose and triglyceride levels were assayed using enzymatic kits (Katal, Belo Horizonte, MG, Brazil). The levels of serum adiponectin, resistin, leptin and fasting serum insulin were determined by ELISA (R&D Systems). The homeostasis model assessment of insulin resistance (HOMA-IR) was calculated as follows: HOMA-IR = fasting glucose level (mmol/l) × fasting insulin level (IU/ml)/22·5.

Oral glucose tolerance test

Mice that were fasted overnight received d-glucose orally by gavage (2 mg/g body weight). Glucose levels were monitored from tail blood samples at 0, 15, 30, 60 and 90 min after glucose overload using an Accu-Check glucometer (Roche Diagnostics, Indianapolis, IN, USA).

Adipocyte isolation and glucose uptake by adipocytes

Isolated adipocytes were obtained from epididymal fat pads, as described by Rodbell [16]. Briefly, collagenase digestion (1 mg/ml) was performed at 37°C with constant shaking (140 cycles/min) for 40 min. Cells were filtered through nylon mesh and washed three times with HEPES buffer plus 1% bovine fatty acid free-serum albumin. Glucose uptake was measured as described previously 19. Isolated adipocytes were incubated in the presence or not of insulin (25 ng/ml; Sigma-Aldrich, Carlsbad, CA, USA) for 45 min at 37°C. Then, 2-deoxy-d-[³H]-glucose (2-DG, 50 μl-0,2 μCi) was added for 3 min. Cell suspensions were centrifuged in microtubes containing silicon to separate adipocytes from the buffer and count the intracellular [³H]2-DG.

Statistical analysis

Results were expressed as mean ± standard error of the mean (s.e.m.) and analysed using GraphPad Prism version 4·0 (GraphPad Software, San Diego, CA, USA). Parametric data were evaluated using one-way analysis of variance (anova), followed by the Tukey test. Differences were considered statistically significant at P < 0·05.

Results

Prolonged ingestion of OVA by sensitized mice decreases specific serum IgE resulting in a breakdown of antigenic aversion

IgE has a substantial role in the allergic response and in the resulting aversion to the allergen 20. Indeed, the sensitization by itself induced the production of OVA-specific IgE, as shown on day 0 (Fig. 2a). Also, continuous ingestion of OVA for 7 days by sensitized mice resulted in a further significant increase in this production. As shown previously by our group 16, after 14 days of OVA ingestion by sensitized mice the serum anti-OVA IgE levels decreased to titres shown by animals that were only sensitized (Fig. 2a). Moreover, the immunoglobulin IgG1 production, also related to T helper type 2 (Th2) response, was induced by the sensitization process, and with the ingestion of OVA for 7 days by previously sensitized mice there was a significant increase in its production, which was maintained even with 14 days of oral challenge by those mice (Fig. 2b).

Fig. 2.

Markers of food allergy after ovalbumin (OVA) consumption by sensitized mice. Kinetics of serum anti-OVA immunoglobulin (Ig)E (a) and IgG1 (b) in non-sensitized or sensitized mice after OVA challenge. Food intake was assessed every day during the oral antigenic challenge and the data were reported as a percentage of diet consumption: sensitized group/control group (c). Body weight was assessed weekly before the antigenic challenge and daily after this time (d). The epididymal fat was collected, weighted and correlated to body weight (e) after it was used for histological analysis. The area of 50 adipocytes from each animal was measured in haematoxylin and eosin-stained sections (f). Representative photomicrographs of haematoxylin and eosin-stained epididymal adipose tissue of mice sensitized or not after 7 and 14 days of OVA challenge (f). Bars indicate 100 μm. All data, except food consumption, are reported as means ± standard error of the mean for six mice in each group. *P < 0·05 compared to OVA⁻ groups and #P < 0·05 compared to the OVA⁺ group after 7 days of OVA consumption.

In order to follow the development of antigen aversion, analysis of diet consumption was performed daily for 14 days of continuous and restricted diet containing OVA to sensitized or non-sensitized mice, in order to follow the development of antigen aversion. Sensitized mice showed a continuous decrease in OVA diet consumption, slightly noticeable after 1 day of this diet and more marked after 4 days in comparison to the control group. This consumption was persistently decreased until the seventh day of antigen exposure. However, after this time the OVA aversion was abrogated and sensitized mice showed higher food consumption until day 10 and comparable amounts after this point in comparison to the control group (Fig. 2c).

Prolonged ingestion of OVA for 14 days by sensitized mice results in a partial recovery of body and adipose tissue weight loss

Weight loss is one feature shown by allergic mice in our experimental model 14, so we followed this parameter during all the experiments. Before the antigenic challenge there was no significant difference in the body weight variation between sensitized (OVA⁺) and non-sensitized mice (OVA⁻). However, after the oral challenge sensitized mice showed significant weight loss that started on the first day and peaked on the seventh day. After this point it was possible to observe the beginning of recovery in the body weight of these mice, but with 14 days of challenge this parameter still did not reach the same level as the control group (Fig. 2d). Akin to what was observed in body weight, we also observed a marked reduction of epididymal adipose tissue mass and adipocyte area in sensitized mice after 7 days of oral challenge. Thereafter, there was a significant increase in these parameters after 14 days of OVA diet by sensitized mice (Fig. 2e–g).

Prolonged ingestion of OVA by sensitized mice decreases systemic and local adipose tissue inflammation induced by food allergy

OVA diet ingestion for 7 days by sensitized mice increased the total number of blood leucocytes (Fig. 3a), with an increase in the number of both mononuclear cells and neutrophils (data not shown). It is interesting to note that leucocytes in blood reached normal levels after 14 days of antigenic challenge. In order to analyse the cellular recruitment to adipose tissue, intravital microscopy was performed and there was an increased rolling and adhesion of leucocytes to the microcirculation of this tissue after 7 days of OVA challenge by sensitized mice. Akin to what was observed in the number of blood leucocytes, rolling and adhesive events fell on day 14 (Fig. 3b,c). Levels of IL-6 and TNF-α were determined to evaluate whether prolonged OVA diet ingestion by sensitized mice could influence the production of inflammatory cytokines in adipose tissue. Concentrations of these cytokines increased significantly in adipose tissue of sensitized mice after 7 days of antigenic challenge in comparison to the control group, and after 14 days they showed a significant decrease in these levels (Fig. 3d,e).

Fig. 3.

Markers of systemic and local adipose tissue inflammation. Whole blood was collected for white blood total cell count (a). Visualization was performed of leucocyte–endothelium interaction during different times of ovalbumin (OVA) diet challenge. Intravital microscopy was used to assess rolling (b) and adherent (c) leucocytes in microvasculature of epididymal adipose tissue in vivo. Kinetics of interleukin (IL)-6 (d) and tumour necrosis factor (TNF)-α (e) production in epididymal adipose tissue in non-sensitized or sensitized after OVA challenge. All data are reported as means ± standard error of the mean for six mice in each group. *P < 0·05 compared to OVA⁻ groups and #P < 0·05 compared to the OVA⁺ group after 7 days of OVA consumption.

Kinetics of changes in adipokine profile and glucose metabolism in sensitized mice after oral challenge

Serum levels of adipokines (adiponectin, leptin and resistin) were significantly lower in sensitized mice after 7 days of OVA challenge. After 14 days, adiponectin and leptin, but not resistin, levels showed a significant increase in comparison to mice that received the OVA diet for 7 days (Fig. 4a–c).This altered adipokine profile was associated with significant changes in serum levels of glucose and triglycerides. Glucose levels in sensitized mice were decreased at 7 days after initiation of the OVA diet and were persistently decreased throughout the observation period (Fig. 5a). However, levels of triglycerides, which were also decreased in sensitized mice after 7 days of oral challenge, returned to basal levels after 14 days (Fig. 5b). The fasting serum level of insulin was measured and there was no significant difference between the groups (Fig. 5c). Despite this, HOMA-IR index was lower in sensitized mice after 7 and 14 days of challenge (Fig. 5d). Glucose tolerance tests were performed in mice after 7 and 14 days of OVA intake and it was seen that after 7 days of the OVA diet sensitized mice had a significantly higher tolerance to glucose challenge, once there was a significant decrease in the area under the curve in those animals when compared to non-sensitized mice (Fig. 5e). However, this feature was lost after 14 days of oral challenge. In order to understand more clearly why, after 7 days of antigenic challenge, sensitized mice had a higher tolerance to glucose challenge without changes in insulin levels, an evaluation was performed of glucose uptake by adipocytes. Through this analysis, we have found that both basal and insulin-stimulated glucose uptake were increased significantly in sensitized mice in comparison to non-sensitized animals (data not shown).

Fig. 4.

Profile of circulating adipokines. The serum was collected for 7 and 14 days after oral antigenic challenge and the concentrations of adiponectin (a), leptin (b) and resistin (c) were measured in non-fasted animals. All data are reported as means ± standard error of the mean for six mice in each group. *P < 0·05 compared to ovalbumin (OVA)⁻ groups and #P < 0·05 compared to the OVA⁺ group after 7 days of OVA consumption.

Fig. 5.

Metabolism of glucose and triglyceride. The serum was collected for 7 and 14 days after oral antigenic challenge and the concentrations of glucose (a), triglyceride (b) and insulin (c) were measured in overnight fasted mice. Insulin resistance index (HOMA-IR) was calculated (d). Oral glucose tolerance tests (OGTT) in fasted non-sensitized and sensitized mice after 7 and 14 days of oral antigenic challenge were determined (e). All data are reported as means ± standard error of the mean for six mice in each group. *P < 0·05 compared to ovalbumin (OVA)⁻ groups and #P < 0·05 compared to the OVA⁺ group after 7 days of OVA consumption.

Prolonged ingestion of OVA by sensitized mice protects against systemic anaphylaxis induced by 2 mg, but not 20 mg, of antigen

OVA was administered systemically in order to test if the decreasing anti-OVA IgE level caused by prolonged ingestion of OVA by sensitized mice also leads to an improvement in the metabolic consequences caused by food allergy could also result in a protection against systemic anaphylaxis. Anaphylactic shock is an allergic reaction associated with haemodynamic changes such as hypotension, caused mainly by vasodilatation. Mice only sensitized, that did not receive an oral challenge with OVA (day 0) after receiving 2 mg of the same antigen systemically, showed a significant decrease in MAP (Fig. 6a), blood volume (Fig. 6b) and flow (Fig. 6c) in the tail when compared to non-sensitized mice. After 7 days of OVA ingestion, sensitized mice had a strong response to this dose of antigen and in this case, most of the animals died fewer than 30 min after the injection, which did not allow us to assess the haemodynamic changes in this group. Interestingly, with this same dose of i.p.-administered OVA, prolonged ingestion of OVA for 14 days was able to prevent anaphylactic shock, and these animals showed MAP (Fig. 6a), blood volume (Fig. 6b) and flow (Fig. 6c) in the tail similar to non-sensitized mice. In order to determine if this protection would be maintained if the antigen was given at a higher dose, mice received 20 mg of OVA i.p. Sensitized mice before oral challenge (day 0) showed the same profile of haemodynamic changes presented when they received the 2-mg i.p. injection. However, the prolonged ingestion of OVA by sensitized mice after 14 days was not enough to avoid the haemodynamic changes caused by 20 mg of OVA i.p. in comparison to the mice that were only sensitized and did not receive the oral challenge. Moreover, with this dose the anaphylactic shock on sensitized mice also resulted in death after 7 days of OVA diet.

Fig. 6.

Haemodynamic analysis following anaphylaxis induced by two different doses of ovalbumin (OVA). Mean arterial pressure (MAP) (a), blood volume (b) and flow (c) in the tail were evaluated by a volume pressure recording sensor and an occlusion tail-cuff system. All data are reported as means ± standard error of the mean for six mice in each group. *P < 0·05 compared to OVA⁻ groups and #P < 0·05 compared to the OVA⁺ group before OVA consumption (day 0).

Discussion

Our data demonstrate that after a prolonged oral challenge with OVA diet for 14 days, OVA-sensitized mice showed a decrease in the anti-OVA IgE level and protection against systemic anaphylaxis. Along with this prolonged ingestion of OVA, most metabolic alterations shown by allergic mice were improved. We have shown a reduction in the adipose tissue inflammation, weight and adipose tissue mass loss as well as the return to basal serum levels of adiponectin and triglycerides. These results mark a decrease in the inflammatory process in adipose tissue and in the systemic metabolic consequences induced by experimental food allergy, as a result of prolonged ingestion of OVA by previously sensitized mice.

In the present study, we confirm the importance of investigating the metabolic changes that occur in food allergy, as they can be extremely deleterious, especially in children, who are the most affected by food allergic disorders. Indeed, food allergies in children are related to nutritional deficiencies that can interfere with growth, neurological and cardiovascular development 21. Moreover, there is no curative treatment for food allergy and the existing standard of care is the triggering of food avoidance, which is worrying from a nutritional perspective, mainly in children who must avoid multiple foods 1. However, some therapeutic strategies, such as OIT, are under investigation, which involves the regular administration of increasing doses of allergen by the oral route in order to induce desensitization or even oral tolerance to the allergen 10–12.

Although within the allergy research field there is a great deal of discussion concerning the effectiveness of OIT, it has been demonstrated that the OIT protocols could show good results for allergies to cow's milk and peanut, for example 9,22–24. Although OIT seems to be a promising therapeutic strategy, there is a high risk of systemic reactions during this procedure, and because of remaining unanswered questions, this type of therapy is not ready to be implemented fully. As little is known regarding the systemic consequences of this acquired desensitization, or in some cases oral tolerance, the study of their effects in metabolism, for example, is necessary. Although our protocol does not reflect oral immunotherapy, as the antigen was not given in increasing controlled doses, in this study we attempted to show the consequences of continuous and prolonged ingestion of the antigen by previously sensitized mice. Moreover, through this prolonged ingestion of antigen it was possible to reach a significant decrease in the allergic response in mice and even protection against systemic anaphylaxis, so to some extent this protocol can mirror OIT. Our study also showed that prolonged ingestion of OVA for 14 days by sensitized mice resulted in a significant decrease in the serum levels of anti-OVA IgE, while the serum IgG1 titres remained the same in comparison to sensitized mice that received the OVA diet for 7 days. The specific IgE level in those animals reached the same level of sensitized mice that were not orally challenged with OVA, but the difference between these two groups regarding the humoral response is in the concentration of specific IgG1. It has been shown that beyond neutralization, IgG (IgG1 in mice and IgG4 in humans) blocks mast cell activation 25 and can even protect IgE-mediated anaphylaxis 26. Indeed, this is one of the explanations of why the IgE levels do not necessarily decrease during oral immunotherapy despite showing efficiency, both in humans and in murine experimental models 22,23,27. In our model, prolonged ingestion of OVA for 14 days by sensitized mice protected them against anaphylactic shock when the antigen was given systemically in a dose of 2 mg, and the balance of those immunoglobulins is one of the factors that might be contributing to this response, together with other tolerogenic mechanisms that will be investigated in a future work. However, when the OVA diet was given in a dose 10 times higher (20 mg), prolonged ingestion of OVA was not enough to prevent anaphylaxis.

As well as being a key event during anaphylactic shock, mast cell activation has an important role in allergen aversion by sensitized mice. It was shown that this process happens through IgE-mediated mast cell degranulation with consequent 5-HT₃ signalling 28. It has also been shown that this aversion is immune-based 29 and is dependent upon both IgE 20 and IL-4 18. Therefore, these data support the breakdown of antigen aversion that happens in sensitized mice after 14 days of prolonged OVA ingestion when the specific IgE levels are decreased significantly, while IgG1 is kept at the same level of sensitized mice that were only challenged orally for 7 days.

This phenomenon, the breakdown of antigen aversion, can partly explain the weight loss that happened until the seventh day of oral challenge by sensitized mice. Indeed, a previous study by our group showed that this loss also involves an inflammatory process in the adipose tissue, and this inflammation results in systemic metabolic consequences 14. This study also demonstrated that weight loss in allergic mice occurs mainly because of the adipose tissue loss followed by adipocyte hypotrophy, with a higher lipolysis rate in fat tissue 14. It has been demonstrated in both animal and human studies that IL-6 30 and TNF-α induce adipocyte lipolysis, and administration of this last cytokine in human subjects can promote a 40% increase in whole body lipolysis 31,32. In our work, we have shown that after 7 days of oral challenge sensitized mice showed an increase of IL-6 and TNF-α levels in the adipose tissue coinciding with the peak of body and adipose tissue loss, as well as the adipocyte area. However, after 14 days the levels of such cytokines were decreased significantly, which could be responsible for the partial recovery of these parameters.

An important metabolic feature to be analysed regarding these alterations in adipose tissue is the circulating levels of leptin, because it has an essential role in energetic homeostasis and is synthesized and secreted mainly by adipocytes. Leptin is responsible for sending information to the hypothalamus concerning the amount of energetic resources in the organism 33, inhibiting the appetite and increasing energetic waste through the induction of an increased lipid oxidation rate 34. Moreover, there is a positive correlation between the circulating levels of this adipokine and the amount of adipose tissue in the body. In fact, when we verified the circulating concentrations of leptin in our animals, we could confirm this correlation. However, it was seen that the leptin levels were decreased significantly on the seventh day of oral challenge by sensitized mice, and there was an increase in the production of this adipokine by sensitized mice on the fourteenth day in comparison to mice that received the OVA diet for only 7 days, without reaching the basal level. This diminished concentration on the seventh day could also be responsible for the breakdown of the aversive behaviour shown by sensitized mice, once this adipokine is related intimately to satiety. Moreover, both behaviours, satiety and immunological aversion, are associated with the brain–gut axis 29,35,36.

Other important adipokines are adiponectin and resistin. Adiponectin is synthesized almost exclusively by adipocytes and is present at high levels in the blood. Its production by adipocytes is inhibited by proinflammatory factors, such as the cytokines TNF-α and IL-6 37. In agreement with this study, after oral challenge a negative correlation was observed in sensitized mice between the levels of adiponectin and TNF-α and IL-6. Resistin is a cysteine-rich protein secreted by adipocytes, which has been proposed to link obesity with insulin resistance and diabetes 38. Previous in-vitro studies have shown that 3T3-L1 adipocytes treated with TNF-α have decreased resistin mRNA expression and protein secretion 39. According to these data, low levels of resistin in sensitized mice after 7 days of OVA diet could be explained by the high levels of TNF-α produced by adipose tissue. Another important function of resistin is the maintenance of glucose homeostasis and, indeed, knock-out mice to resistin show a significantly lower glucose level in comparison to wild-type animals, due to the decreased hepatic production of glucose 40 and an improvement in insulin sensitivity 41. We have demonstrated that OVA-sensitized mice had a lower glucose concentration along with lower levels of resistin after 7 days of OVA diet. To understand more clearly the low levels of glucose shown by sensitized mice after 7 and 14 days of oral challenge, quantification of serum insulin was performed in fasting mice. The levels of insulin were not different between the groups, showing that the low glucose level is not due to a hyperinsulinaemic state. However, low HOMA-IR, indicating more insulin sensitivity shown by sensitized mice after 7 days of oral challenge, was maintained after 14 days of OVA diet. Indeed, it was demonstrated previously that small adipocytes are more sensitive to insulin 42, which could be one factor responsible for the low levels of glucose in sensitized mice after oral challenge. Another factor to be taken into account is that the allergic suppression achieved is an active process 43 that consumes energy, or ultimately glucose.

Although the allergic response in sensitized mice after 14 days of oral challenge is suppressed, low levels of circulating anti-OVA IgE are maintained. These remaining circulating IgEs can bind the high-affinity IgE receptor present in mast cells, resulting in degranulation of these cells with the release of inflammatory molecules; even with the protective role of IgG1 against activation of these cells, this event might be happening to some extent. Adipose tissue is one depot rich in this cellular type, especially during obesity and food allergy 14,44.Therefore, this low persistent level of anti-OVA IgE shown by sensitized mice after 14 days of oral challenge could be responsible for maintaining the residual concentration of inflammatory cytokines in adipose tissue. Also, the remaining levels of this immunoglobulin could be responsible for the anaphylactic shock that happened when the antigen was administered at a higher dose (20 mg).

Recently, food allergy research has received special attention mainly because of the promising development of OIT with the possibility of cure 10,11. However, little is known about the systemic consequences of the desensitization or even oral tolerance developed within this treatment in animals and patients. Our work contributes towards clarifying some metabolic consequences in mice submitted to the food allergy protocol with posterior suppression of the allergic response through prolonged ingestion of antigen. Although the study protocols and processes cannot be transferred directly from mice to humans, these data can be used for future studies regarding the metabolic alterations of food allergy and offer a new viewpoint concerning the consequences of therapies using the continuous ingestion of the allergen.

Disclosure

The authors declare that there are no conflicts of interest.

Author contributions

N. V. B. designed, performed, analysed the experiments and wrote the paper. R. V. S. P. performed the experiments. L. P. A. D and M. L. M. N. performed the experiments and wrote the paper. D. A. P and G. F. performed the anaphylaxis experiment. A. V. M. F., A. J. F and D. C. C. designed the experiments, supervised and edited the paper.