Glutamine, Glutamate, and Aspartate Improves Morphology and Energy Production of Small Intestine in Piglets with Different Energy Levels Diets

Background: Weaning-stress may cause reduced energy intake for maintenance of mucosal structure. Glutamine (Gln), glutamate (Glu), and aspartate (Asp) are major energy sources for small intestine. This study investigated whether Gln, Glu, and Asp improve the intestinal morphology via regulating the energy metabolism in weaning-piglet. A total of 198 weaned-piglets were assigned to the following treatments: Control (Basal diet + 1.59% L-Alanine); T1 (Basal diet + 1% L-Gln + 0.5% L-Glu + 0.1% L-Asp); T2 (Low energy diet + 1% L-Gln + 0.5% L-Glu + 0.1% L-Asp). Jejunum and ileum were obtained on d-5 or d-21 postweaning. Results: Improved growth performance on d-21 post-weaning were observed in T1 treatment. Both T1 and T2 treatments improved small intestinal morphology by increasing villus height, goblet cell number and decreasing crypt depth on d 5 or d-21 post-weaning. Gln, Glu, and Asp could restore the small intestinal energy homeostasis via replenishing the Krebs’ cycle and down-regulate the AMPK pathway. However, when piglets fed by a low energy diet, Gln, Glu, and Asp are not sucient to maintain the intestinal energy balance on d-5 post-weaning so that the AMPK, beta-oxidation, and mitochondrial biogenesis are activated to meet the high energy demand of enterocytes. Conclusion: These data indicated that Gln, Glu, and Asp could restore the energy homeostasis of intestinal mucosa of piglets. And the mucosal energy metabolism showed different responses to the intervention of Gln, Glu, and Asp in piglets with a low energy diet between d5 and d21 post-weaning. These ndings provide new information on the nutritional intervention for the insucient energy intake in weaning-piglet. After dehydration, embedding, sectioning, and staining, images were acquired at various magnications with computer-assisted microscopy (Micro-metrics TM; Nikon ECLIPSE Tokyo, Japan). Villous height, crypt depth, goblet cell and lymphocyte counts were measured by Image-Pro Plus software, Version 6.0 on images at 200- or 400-fold magnication in ve randomly selected respectively [19]. supplementation with Gln, Glu, and Asp in either a normal energy diet or a low energy diet replenish the Krebs’ cycle in enterocyte converting to the intermediate metabolites. energy supplementation with Glu, Gln, Asp to promote glycolytic inux on d 5 instead of d 21 post-weaning. Unlike the low diet treatment, a normal diet addition with Gln, Glu, Asp did not signicantly affect the glycolysis during the weaning-period. challenge, Gln, Glu, and Asp are not sucient to maintain the intestinal energy balance so that the AMPK, fatty acids beta-oxidation, and mitochondrial biogenesis pathways are activated to meet the high energy demand of enterocytes of piglets. Our results aid in providing new information about the nutritional intervention of Gln, Glu and Asp for the insucient energy intake in weaning-piglet.

One piglet closed to the average body weight from each pen was slaughtered on day 5 (w5d) and 21 (w21d) post-weaning. After being stunned electrically, piglets were sacri ced and the jejunum and ileum was dissected and rinsed thoroughly with ice-cold physiological saline. The middle segments of the jejunum (2 cm) and ileum (2 cm) were cut and xed in 2.5% glutaraldehyde or 4% formaldehyde for morphological and immunohistochemical analysis. Samples of the jejunal and ileal mucosa were scraped, immediately snap frozen in liquid nitrogen and stored at -80 °C for further analysis.

Intestinal morphology
The segments of the jejunum and ileum xed in 4 % formaldehyde were used to determine morphology using hematoxylin-eosin staining. After dehydration, embedding, sectioning, and staining, images were acquired at various magni cations with computer-assisted microscopy (Micro-metrics TM; Nikon ECLIPSE E200, Tokyo, Japan). Villous height, crypt depth, goblet cell and lymphocyte counts were measured by Image-Pro Plus software, Version 6.0 on images at 200-or 400-fold magni cation in ve randomly selected elds, respectively [19].
Segments of the jejunum and ileum at 150-fold magni cation were also designated for analysis by scanning electron microscopy as described by German [20] and Liu et al. [21]. Brie y, tissue segments were xed with 2.5 % glutaraldehyde for 2 h at 4 °C, and rinsed 3×10 min in PBS at 4 °C. The tissues were then post xed in 1 % osmium tetroxide for 12 h at 4 °C, and rinsed 3×10 min in PBS at 4 °C. After samples were mounted onto stubs by means of quick-drying silver paint, the tissues were coated with gold-palladium and examined by a JEOL JSM-6360LV scanning electron microscope at 25 KV. The apparent characteristics of microvillus were observed and described.
After samples were mounted onto stubs by means of quick-drying silver paint, the tissues were coated with gold-palladium and examined by a JEOL SU8010 scanning electron microscope at 1.0 KV. The apparent characteristics of microvillus were observed and described.

Metabolomic pro ling
The ileal mucosal metabolomic pro le was analyzed by liquid chromatography (LC)-mass spectrometry (

Western blotting
Jejunal and ileal mucosa samples were homogenized, and protein concentrations were measured using the bicinchoninic acid assay method with BSA as standard (Beyotime Institute of Biotechnology, Shanghai, China). All samples were adjusted to an equal concentration (70 µg protein). The supernatant uid (containing tissue proteins) was diluted with 5 × sodium dodecyl sulfate (SDS) sample buffer and heated in boiling water for 5 min and cooled down for Western blot analysis [23]. The following rst antibodies were used for protein quanti cation: AMPKα

Statistical analysis
The growth performance was performed by one-way ANOVA using SPSS software 19.0 (SPSS Inc., Chicago, IL). The differences among treatments were evaluated using Tukey's test.
The intestinal morphology, metabolomic pro les, gene and protein expression were analyzed using the general linear models' procedure, with diet treatments, days after weaning, and the interaction between treatment and day as the main effects. The results in the tables were presented as main effects. When no signi cant interaction between treatment and day was found, the differences among treatments or days were evaluated using Tukey HSD test (showed in Table 3, 4, 5, 6). When signi cant interaction between treatment and day was found, a simple main effect analysis would be performed and the differences among 3 treatments across 2 levels of days post-weaning were evaluated using Bonferroni test (showed in Figure 1, 2, 3). The developmental changes of the intestine were not the main outcome, so the differences between d 5 and d 21 post-weaning across 3 levels of treatments were not showed. Differences were declared as signi cant at P < 0.05.

Growth performance
Initial body weight (BW), nal BW, average daily gain (ADG), average daily feed intake (ADFI), and the ratio of average daily feed intake to average daily gain (F / G) were showed in Table 2. The initial BW was similar among treatments (P = 0.959). On d 5 post-weaning, there were no differences on BW (P = 0.952), ADG (P = 0.782), ADFI (P = 0.127) and F/G (P = 0.240) among different treatments. On d 21 post-weaning, basal dietary supplementation with Gln, Glu, Asp (T1) improved the nal BW and ADG of piglets in comparison to those in control group (P < 0.05). However, the digestible energy intakes of weanling piglets neither on d 5 (P = 0.190) nor d 21 (P = 0.709) post-weaning were not signi cantly different.

Jejunal and ileal morphology
In Figure 1A and B, the scanning electron microscopy showed that the villi in the jejunum and ileum on d 5 post-weaning were thin and sparse, and then grew shorter, stouter and denser on d 21 post-weaning. In the jejunum, the villi heights among control, T1, and T2 groups were not showed signi cantly different on w 5 d, but there was an increase in villi density and villi-fold in T1 and T2 group on w 21 d. In the ileum, the increased villi densities were observed in control and T1 groups in comparison to T2 group on w 5 d, whereas villi in T1 and T2 group were denser and stouter than villi in control group on w 21 d.
The main effects of diet treatments and days after weaning on parameters from the intestinal hematoxylin-eosin staining were showed in Table 3. When signi cant interactions between treatments and days after weaning were obtained, differences among control, T1 and T2 on d 5 or d 21 post-weaning were showed in Figure 1C. In Table 3, the number of goblet cells in the jejunum of T2 group was higher than these in T1 group (P < 0.05). There were interactions (P < 0.05) between treatments and days on the jejunal villus height, ileal villus height and crypt depth, which means days after weaning in uenced the treatments' e cacy.
In Figure 1C, a low energy diet supplemented with Glu, Gln, and Asp increased (P < 0.05) the villus height of jejunum on d 21 post-weaning and decreased (P < 0.05) the crypt depth of ileum on d 5 post-weaning as compared to the control group, whereas it had lower (P < 0.05) ileal villus height than the control and T1 groups on d 5 post-weaning. A basal diet supplemented with Gln, Glu, and Asp increased the villus height of ileum on d 21 post-weaning compared to the control group.
Metabolomic pro ling of jejunal mucosa of weanling piglets The jejunal mucosa energy metabolite ngerprints were presented in Figure 2A and Figure  The main effects of diet treatments and days after weaning on metabolites' levels of jejunal mucosa were showed in Table S2.
When signi cant interactions between treatments and days after weaning were found, differences among control, T1 and T2 on d 5 or d 21 post-weaning were showed in Figure 2C. In Table S2, supplementation with Gln, Glu, and Asp in both basal and low energy diets increased (P < 0.05) the succinate level but decreased (P < 0.05) the phosphoenolpyruvate level in jejunal mucosa of weanling piglets. Days post-weaning had in uences on the diet treatments' e cacy (P interaction < 0.05). As shown in Figure   2C, a basal diet supplemented with Gln, Glu, and Asp (T1) increased (P < 0.05) the pyruvate, α-ketoglutarate (AKG), isocitrate, and fructose 1,6 phosphate (F-1,6-P) levels but decreased (P < 0.05) the NADPH level on d 5 post-weaning compared to control group. A low energy diet supplemented with Gln, Glu, and Asp (T2) increased (P < 0.05) the pyruvate, AKG, isocitrate and GMP levels compared to the control group and increased (P < 0.05) DHAP, NADPH, and AMP levels as compared to control and T1 groups but reduced (P < 0.05) 3-phosphoglycerate, F-6-P, and G-6-P levels on d 5 post-weaning. On d 21 post-weaning, T1 and T2 treatments showed higher (P < 0.05) oxaloacetate level than the control group.
Expression of gene involved in energy metabolism of jejunal and ileal mucosa of piglets The gene expressions of several key enzymes of energy metabolism were presented in Table S3 and Figure 2D. The main effect in Table S3 showed T1 treatment increased (P < 0.05) the ileal Ucp2 mRNA level compared to the control group but decreased (P < 0.05) the ileal Cpt1a mRNA level compared to control and T2 groups. T2 treatment increased (P < 0.05) the jejunal Acox1 and ileal Cpt1a mRNA levels in comparison with the control group. Days post-weaning had in uences on the diet treatments' e cacy (P interaction < 0.05; Figure 2D). On d 5 post-weaning, a low energy diet supplemented with Gln, Glu, and Asp had lower (P < 0.05) Pdk4 and Pck1mRNA abundances than control or T1 groups. On d 21 post-weaning, a basal diet supplemented with Gln, Glu, and Asp enhanced (P < 0.05) the Sdh1 mRNA abundance in comparison to these in the control group. A low energy diet supplemented with Gln, Glu, and Asp increased (P < 0.05) the Pck1 and Sdh1 mRNA abundances as compared to T1 or control groups, respectively.
mRNA and protein abundances of AMP-activated protein kinase signaling pathway in jejunal and ileal mucosa of piglets The relative mRNA and protein abundances of AMPK pathway in the jejunal and ileal mucosa were presented in Table S4, S5 and Figure 3, respectively. The main effects on AMPK pathway gene expressions in Table S4 showed a basal diet supplemented with Gln, Glu and Asp (T1) lowered (P < 0.05) jejunal Pgc1α, Tfam, and Nrf1 and ileal Sirt1, Pgc1α mRNA levels compared to control and T2 group, as well as ileal Tfam mRNA level compared to the T2 group. The main effects on AMPK pathway protein abundances in Table S5 showed T1 treatment enhanced (P < 0.05) ileal ACC protein abundance as compared to that in the control group. Days post-weaning did not affect the diet treatments' e cacy on the AMPK pathway gene expression (P interaction > 0.05) but had in uenced the treatments' e cacy on AMPK pathway protein abundances (P interaction < 0.05). As shown in Figure 3, a basal diet supplemented with Gln, Glu and Asp (T1) enhanced (P < 0.05) the jejunal AMPKα protein abundance compared to the control group but reduced (P < 0.05) the ileal p-AMPKα and PGC1α protein abundances as compared to control and T2 groups, respectively, on d 5 post-weaning. A low energy diet supplemented with Gln, Glu and Asp (T2) reduced (P < 0.05) the ileal AMPKα protein abundance compared to control and T1 groups, while increased (P < 0.05) jejunal and ileal SirT1 and PGC1α, as well as jejunal ACC and p-ACC protein abundances compared to control or T1 groups on d 5 post-weaning. On d 21 post-weaning, T1 and T2 treatments lowered (P < 0.05) the jejunal ACC, ileal AMPKα and PGC1α, as well as jejunal and ileal p-AMPKα, p-ACC, and SirT1 protein abundances as compared to the control groups. Meanwhile, T1 treatment showed lower (P < 0.05) jejunal and ileal PGC1α and ileal SirT1 protein abundances than these in the T2 group.

Discussion
The mammalian intestine plays a key role in nutrients intake and has a high rate of energy expenditure because the digestion and absorption processes are highly dependent on energy [24,25].Weaning results in a growth check caused by a period of low feed intake, which also reduces the energy intake of intestine of piglets [26,27]. Successful adaptation to these changes requires profound morphological and energy metabolic adjustments of the gastrointestinal tract [28]. Gln, Glu and Asp were used to help the faster restitution of mucosal barrier after stress injury, and to improve the intestinal energy metabolism of piglets with different energy levels diets in this study.
The most serious damages in piglets during weaning-stress occurs on d 2 to d 5 post-weaning [25,29], and is usually associated with a dramatic reduction of feed intake resulting in marked changes in the growth pattern on body weight [28]. In the present study, supplementation with 1 % Gln, 0.5 % Glu, and 0.1 % Asp in a normal energy level diet enhanced the BW and ADG on d 21 post-weaning. The bene cial effect of Gln, Glu and Asp on improving the growth-suppression induced by weaning stress in our study was consistent with previous researches, which have shown that Glu and Asp exert positive effects on growth performance of animals in healthy and oxidative-stress conditions [30,31] [32,33]. Although low energy diet did not signi cantly increase the ADFI of weanling piglets, the digestive energy intake in T2 group was comparable with these in control and T1 groups. Piglets in low energy diet treatment increased their satiety through increasing food intake.
In addition to poor growth performance, the growth-check induced by weaning-stress is accompanied by impaired mucosal integrity, villous atrophy, an increased rate of cell loss, or a decreased rate of crypt-cell production [34,35]. Supplementation with Gln, Glu, and Asp in normal energy diet increased ileal villus height on d 21 post-weaning. The major reasons for villus atrophy are de cient enteral nutrition and reduced energy intake independent of diet type [36,37]. We found that a low energy diet supplemented with Gln, Glu, and Asp could alleviate impaired intestinal structure by raising jejunal villus height and goblet cells number or reducing ileal crypt depth on either d 5 or d 21 post-weaning. The critical roles of these three amino acids in intestine have been reported in many well-designed studies, for example, Gln could increase the villus length of jejunum [38], and Asp could alleviate the intestinal barrier damage [32,39,40]. The results in our current study indicated that Gln, Glu, and Asp could improve the intestinal barrier structure of weanling piglets fed either a low energy diet or a normal energy diet, which led us to further explore the mechanism of these three amino acids regulating energy homeostasis in the small intestine.
ATP production and utilization are very active in pig enterocytes. The relative abundances of intermediate metabolites involved in the energy metabolism of the small intestine of piglets on d 5 and d 21 post-weaning were not comparable, and the regulation of Gln, Glu, and Asp on energy metabolites varied between d 5 and d 21 post-weaning. It may indicate that the energy metabolism in the small intestine changes during the weaning period [41], and the pattern of metabolism has different responses to amino acids. It is well established that dietary Gln, Glu, and Asp plus arterial Gln provide up to 80 % of ATP to the small intestine mucosa in mammals [11,42,43]. After these three amino acids are absorbed into enterocytes, they could be metabolized and oxidized via the Krebs' cycle to water and CO2 yielding ATP [44]. The oxidative steps of Gln and Glu in pig enterocytes are similar. Gln rstly enter the mitochondrial in order to be degraded to Glu and ammonia by the mitochondrial glutaminase [45]. Glu could be transaminated to produce AKG, which is the key molecular of the Krebs' cycle [46]. Asp replenished the Krebs cycle by being converted to oxaloacetate [44]. In the present study, supplementation with Gln, Glu, and Asp either in a regular energy diet or a low energy diet could replenish the jejunal Krebs' cycle, manifested as increased pyruvate, oxaloacetate, isocitrate, AKG, and succinate levels on d 5 or d 21 post-weaning. Meanwhile, supplementation with Gln, Glu and Asp in either regular energy diet or low energy diet enhanced Sdh1 mRNA abundance on d 21 post-weaning, whereas low energy diet supplemented with Gln, Glu, and Asp declined Pdk4 mRNA abundance on d 5 post-weaning. SDH1 is part of Krebs' cycle and oxidizes succinate to fumarate [47]. The higher Sdh1 expression was consistent with the higher succinate level in the jejunal mucosa. Sdh1 is responsible to promote the oxidation of succinate and prevents the accumulation of succinate in the mitochondria [48].PDK4 is a regulator of pyruvate dehydrogenase (PDH), as it could inactivate PDH by phosphorylation [49]. PDH represents a cornerstone in cellular energy metabolism, linking glycolysis and Krebs' cycle and lipid metabolism [50]. In this study, PDH activity was failed to detected, whereas the decreased Pdk4 gene expression might suggest the in ux of acetyl-CoA from glycolysis into the Krebs' cycle increased in response to T2 treatment on d 5 post-weaning. Indeed, a low energy diet addition with Gln, Glu, and Asp (T2) showed lower G-6-P, F-6-P, F-1,6-P, and 3-phosphoglycerate, but higher DHAP in jejunal mucosa on d 5 post-weaning, whereas the T2 did not affect these intermediate metabolites on d 21 post-weaning. Nevertheless, a normal energy diet addition with Gln, Glu, and Asp had no effect on the levels of glycolytic metabolites on neither d 5 nor d 21 post-weaning, except for the increased F-1,6-P levels on d 5 post-weaning. Our results suggested that dietary supplementation with Gln, Glu, and Asp in either a normal energy diet or a low energy diet can replenish the Krebs' cycle in enterocyte by converting to the intermediate metabolites. However, a low energy diet supplementation with Glu, Gln, and Asp seemed to promote glycolytic in ux on d 5 instead of d 21 post-weaning. Unlike the low diet treatment, a normal diet addition with Gln, Glu, and Asp did not signi cantly affect the glycolysis during the weaning-period.
Fatty acids oxidation also plays critical role in energy homeostasis [51]. UCP2 is involved in the regulation of mitochondrial substrate oxidation, which could be stimulated by fatty acid and glutamine [52,53]. Several studies suggested that the role of UCP2 would be to promote oxidation of glutamine and fatty acids rather than that of the pyruvate derived from glucose [54,55]. Indeed, a normal energy diet supplementation with Gln, Glu, and Asp raised the Ucp2 mRNA abundance but decreased Cpt1α in ileal mucosa. Although a low energy diet supplementation with Gln, Glu, and Asp did not affect the Ucp2 gene expression, it upregulated the jejunal mucosal Acox1 and ileal mucosal Cpt1α mRNA abundances. CPT1 and Acox1 play key roles in the fatty acid beta-oxidation in either mitochondria or peroxisomes, respectively. CPT1 is the key enzyme in the carnitine-dependent transport across the mitochondrial inner membrane [56], while Acox1 is the rst enzyme of the fatty acid beta-oxidation in peroxisomes, which catalyzes the desaturation of acyl-CoAs to 2-trans-enoyl-CoAs [57]. Based on these results, it may be suggested that a low energy diet addition with Gln, Glu, and Asp activated the fatty acids beta-oxidation. Besides, supplementation with Gln, Glu, and Asp in a low energy diet increased jejunal mucosal NADPH, GMP and AMP levels, as well as the Pck1 gene expression on d 5 post-weaning. NADPH is mainly produced through pentose cycle, which is an alternative way of glucose use [58]. Because the current study did not detect other enzymes or intermediate metabolites involved in the pentose cycle, it is not clear whether a low energy diet supplementation with Gln, Glu, and Asp regulates the NADPH levels via the pentose cycle. Moreover, the decreased Pck1 gene expression, which is a main control point of the regulation of gluconeogenesis [59], may implicate the reduced gluconeogenesis in the T2 group on d 5 post-weaning. The higher AMP and GMP levels activates the AMPK signaling pathway, which is a key sensor of energy balance responding to low energy status and upregulating catabolic pathways and downregulating anabolic pathways [60]. Thus, it is reasonable to speculate that a low diet addition with Gln, Glu, and Asp activates the AMPK signaling pathway.
A low energy diet supplementation with Gln, Glu, and Asp (T2) up-regulated the AMPK and ACC phosphorylation, as well as the SirT1 and PGC1α protein abundance in the small intestine on d 5 post-weaning, whereas it down-regulated the AMPK pathway phosphorylation levels on d 21 post-weaning. Under conditions of low energy, activated AMPK phosphorylates downstream substrates, increasing ATP levels by promoting glucose and lipids breakdown and inhibiting their synthesis and storage [7,61]. This is consistent with the observation from energy metabolites and these enzymes' expression mentioned above. AMPK could acutely regulate glycolysis and reduce gluconeogenesis, and controls overall lipids metabolism through direct phosphorylation of ACC and simultaneously promoting fatty acid oxidation by relieving the suppression of CPT1 [7]. Another interesting nding is that the PGC1α gene and protein abundances were enhanced by a low energy diet supplementation with Gln, Glu, and Asp (T2). PGC1α provides a direct link between external physiological stimuli and the regulation of mitochondrial biofunction [62]. PGC1α regulates mitochondrial biogenesis via regulation of Nrf1 [63], which in turn activates TFAM that involved in the mDNA replication [6]. In the current study, the increased Nrf1 and Tfam gene expression in T2 treatment suggested that a low energy diet addition with Gln, Glu, and Asp stimulated the mitochondrial biogenesis in response to the acute energy crises. Unlike the low energy diet treatment, supplementation with the amino acids in a normal diet did not affect the AMPK system on d 5 postweaning while down-regulated the phosphorylated AMPK pathway levels on d 21-post weaning. Previous studies have demonstrated that Gln could regulate enterocyte growth under normal energy conditions through activating mTOR pathway but inhibiting AMPK pathway [64,65]. Days 2 to 5 after weaning is the most severe stage during the weaning-period [22]. Piglets on d 2 to 5 post-weaning requires more energy to rebuild the impaired intestinal mucosal barrier [66]. It is possible that Gln, Glu, and Asp could restore the small intestinal energy homeostasis via replenishing Krebs' cycle under a normal energy diet condition, whereas it cannot provide su cient energy sources to piglets fed with a low energy diet so that the AMPK pathway and mitochondrial biogenesis are activated by low energy conditions. While on d 21 post-weaning, the remodeling of the intestinal mucosa has been achieved, and the digestive system has adapted to utilize solid feed components to absorb and transport more energy sources [67, 68], Gln, Glu, and Asp could balance the intestinal energy homeostasis in piglets fed with a low energy diet.

Conclusions
The present result shows that supplementation of Gln, Glu, and Asp in a normal energy diet can promote the growth performance of weaning-piglet on d 21 post-weaning. Supplementation with Gln, Glu, and Asp in either a normal energy diet or a low energy diet improved the small intestinal mucosal barrier structure by raising jejunal villus height and goblet cells number or reducing ileal crypt depth on d 5 or 21 post-weaning. In general, Gln, Glu, and Asp could restore the small intestinal energy homeostasis via replenishing the Krebs' cycle and down-regulate the AMPK pathway. However, when piglets under speci c conditions, such as energy crisis and severe weaning-stress challenge, Gln, Glu, and Asp are not su cient to maintain the intestinal energy balance so that the AMPK, fatty acids beta-oxidation, and mitochondrial biogenesis pathways are activated to meet the high energy demand of enterocytes of piglets. Our results aid in providing new information about the nutritional intervention of Gln, Glu and Asp for the insu cient energy intake in weaning-piglet.

Declarations
Availability of data and materials The datasets generated or analyzed during this study are presented in this manuscript and available to readers.

Ethics approval
All animals used in this study were humanely managed according to the Chinese Guidelines for Animal Welfare. The experimental protocol was approved by the Animal Care and Use Committee of the Chinese Academy of Sciences (Beijing, China).

Con icts of interest
There are no con icts of interest to declare.

Consent for publication
Not applicable.    2 The unit of goblet cells and lymphocyte is cell number per area. 3 The results were presented as main effects, and the differences among treatments or days were evaluated using Tukey HSD test. a-b Values with different letters within the same row are different (P < 0.05); n = 6. Table S1. Primers used for real-time quantitative reverse transcription-PCR 1 .