SEARCH

SEARCH BY CITATION

Keywords:

  • antioxidants;
  • calcium;
  • HLA-B27 transgenic rats;
  • colitis;
  • oxidative stress

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Background:

Oxidative stress is presumed to play an important role in inflammatory bowel disease (IBD). Accordingly, antioxidant supplementation might be protective. Dietary calcium inhibited colitis development in HLA-B27 transgenic rats, an animal model mimicking IBD. As antioxidants might act at mucosa level and calcium predominantly in the gut lumen, we hypothesize that the combination has additive protective effects on colitis development.

Methods:

HLA-B27 rats were fed a control diet or the same diet supplemented with the antioxidants glutathione, vitamin C, and vitamin E, or supplemented with both antioxidants and calcium. Oxidative stress in colonic mucosa, colonic inflammation, intestinal permeability, and diarrhea were quantified.

Results:

Intestinal permeability, diarrhea, myeloperoxidase, and interleukin-1β levels were significantly lower in rats fed both antioxidants and calcium compared to rats supplemented with antioxidants only. No beneficial effects were observed in rats fed the diet supplemented with antioxidants only. Strikingly, despite extremely low colonic mucosal glutathione levels in HLA-B27 rats, there was no oxidative stress-related damage. Subsequent analyses showed no defect in expression of glutathione synthesis genes. Additional experiments, comparing young and older HLA-B27 rats, showed that glutathione levels and also reactive oxygen species production decreased with progression of intestinal inflammation.

Conclusions:

Antioxidant supplementation was ineffective in HLA-B27 rats despite low mucosal glutathione levels, because colitis development did not coincide with oxidative stress in this model. This indicates that the neutrophilic respiratory burst, and thus innate immune defense, is compromised in HLA-B27 rats. As supplementation with both calcium and antioxidants attenuated colitis development, we speculate that this protective effect is attributed to calcium only. (Inflamm Bowel Dis 2011;)

Crohn's disease and ulcerative colitis, two forms of inflammatory bowel disease (IBD), are characterized by chronic intestinal inflammation resulting from mucosal barrier dysfunction, altered microbial factors, and dysregulated immune responses.1 Despite a genetic influence, these different determinants of the pathogenesis of IBD can be modulated by environmental factors and dietary components in particular. Therefore, nutrition might be an interesting preventive option for IBD patients as well as a supporting therapeutic strategy.2

The inflamed bowel of IBD patients is infiltrated with leukocytes, generating large amounts of reactive oxygen species (ROS), presumed to result in oxidative stress and subsequent mucosal damage.3 Antioxidants are proposed to prevent oxidative damage in general.2, 4 Numerous antioxidants exist and they all have specific roles in the mucosal antioxidant defense system. Therefore, for protection against oxidative stress, it might be more effective in nutritional interventions to supplement a mixture of antioxidants, each having different chemical properties, rather than just one antioxidant.5

In this dietary intervention study, we chose to supplement with vitamin C, vitamin E, and glutathione, being central players in the antioxidant network. Vitamin C is a water-soluble antioxidant, vitamin E a lipid soluble antioxidant, while glutathione is one of the major intracellular antioxidants.6 Moreover, the bioactivity of each of these three antioxidants is strongly dependent on the others. For example, vitamin E can be directly regenerated from vitamin E radical by vitamin C,7 and in the case of glutathione deficiency vitamin C can spare glutathione.8–10 Although the literature on oral glutathione supplementation is contradictory, orally supplied glutathione has been shown to increase intestinal mucosal glutathione levels, and also in situations of suppressed de novo glutathione synthesis.11–13

We have previously shown that dietary calcium inhibits colitis development in HLA-B27 transgenic rats.14 The HLA-B27 transgenic rat is a well-characterized animal model of chronic intestinal inflammation. Rats overexpressing the human HLA-B27/β2 microglobulin gene spontaneously develop an inflammatory disease mainly involving the gastrointestinal tract and the colon in particular.15 Also, in humans the HLA-B27 gene is associated with inflammatory disorders.16 Although the exact underlying mechanism has not been characterized yet, it has been shown that HLA-B27-expressing innate immune cells have an essential role in determining mucosal immune responses to commensal bacteria in the development of colitis in transgenic rats.17 Additionally, we have shown in several controlled rat and human studies that dietary calcium improves intestinal resistance against infections with foodborne bacterial pathogens and that it strengthens the mucosal barrier.18–20 It is hypothesized that calcium exerts this protection in part by precipitating irritating bile acids and fatty acids, thereby reducing cytotoxicity of the fecal stream. The decrease in luminal cytotoxicity reduces damage to intestinal epithelial cells and thus reinforces mucosal integrity.21–22 We speculated that dietary antioxidants, suggested to be active in the intestinal mucosa, and calcium, appreciated for its luminal effects, have an additive protective potential against colonic inflammation.

In the present study we investigated the effect of a mixture of supplemental antioxidants with or without extra calcium on colitis development in HLA-B27 transgenic rats. It is hypothesized that a mixture of vitamin C, vitamin E, and glutathione reduces the severity of colonic inflammation but that the combination of dietary antioxidants plus calcium has superior efficacy. As remarkably low glutathione levels were observed in colonic mucosa of HLA-B27 rats, we performed an additional experiment. Young and older transgenic rats were compared to determine possible causality of low glutathione levels to colitis development. Moreover, since glutathione levels were low and oxidative stress-related damage seemed to be absent in the HLA-B27 rats in the dietary intervention study, we studied the mucosal and systemic ROS production of the transgenic rats.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Experimental Design of Dietary Intervention Study: Animals and Diets

The experimental protocol was approved by the animal welfare committee of Wageningen University (The Netherlands). Female HLA-B27/β2-microglobulin transgenic rats on an inbred Fisher 344 background (n = 9 per dietary group) and their nontransgenic counterparts (n = 7; Taconic Farms, Germantown, NY), 8–10 weeks old with a mean body weight of 140 g at the start of the experiment were housed individually in metabolic cages. Animals were kept in a temperature- and humidity-controlled environment on a 12-hour light-dark cycle. Transgenic rats were fed a purified “humanized” Western-type diet ad libitum which contained in the control situation (per kg diet): 200 g acid casein, 326 g corn starch, 174 g glucose, 160 g palm oil, 40 g corn oil, 50 g cellulose, 2 g chromium EDTA (CrEDTA; see below), and 5.16 g CaHPO4.2H2O (corresponding to 30 mmol calcium/kg diet; Sigma-Aldrich, St. Louis, MO). Vitamins and minerals (other than calcium) were added to the diets according to AIN-93.23 To better mimic the composition of a Western human diet, diets were relatively low in calcium and had a high fat content in comparison with AIN-93 recommendations for rodent diets. The experimental diets were supplemented with the antioxidants glutathione (25 mmol/kg diet), vitamin C (2 g/kg diet), and vitamin E (α-tocopherol acetate; 0.68 g/kg diet extra), or with both these antioxidants and calcium (90 mmol/kg diet extra) (all supplements: Sigma-Aldrich), all at the expense of glucose. Because of the limited availability of transgenic rats, we could not include a dietary group supplemented with calcium only, but the effect of solely calcium has been reported recently.14 To prevent possible degradation of the antioxidants by light or air, feed was divided into daily portions per dietary group, vacuum-packed, and stored at 4°C until use. In addition, food was supplied daily to the rats just before dark (eating period). The nontransgenic rats were fed the control diet. Inert CrEDTA was added to the diets to quantify intestinal permeability.24 CrEDTA solution was prepared as described elsewhere and subsequently freeze-dried.25 To check complete formation of the CrEDTA complex, the prepared CrEDTA solution was passed through a cation-exchange resin column (Chelex 100 Resin; Bio-Rad, Hercules, CA) and no uncomplexed Cr3+ ions were present. Food intake was recorded daily and animal weight twice every week. Time of manifestation of colitis in this animal model is variable between different studies and therefore difficult to predict in advance. To follow up, intestinal permeability and diarrhea were monitored every 1 or 2 weeks. Nine weeks after the start of the dietary intervention, when a clear increase in intestinal permeability and diarrhea was measurable in at least one dietary group (see Results), rats were anesthetized with isoflurane. The colon was taken out and its weight and length were measured. A 1-cm piece from the middle of the colon was excised and stored in 10% neutral buffered formalin for histology. The remaining colon parts were longitudinally excised and washed in saline to remove residual luminal contents. Subsequently, the mucosa, consisting of both epithelial cells and lamina propria cells, was scraped off using a spatula and immediately snap-frozen in liquid nitrogen until further processing and analyses.

Colonic Histology

Colonic tissue specimens were embedded in paraffin. Slices were stained with hematoxylin and eosin. Before analysis, slides were recoded to guarantee blind scoring. A validated histological scoring system for inflammation was used. Scores ranged from 0 to 4 based on criteria such as inflammatory cell infiltration, depletion of goblet cells, mucosa thickening, and destruction of mucosal architecture.26

Measurement of Intestinal Permeability and Diarrhea

Total 24-hour urine samples were collected on 1 day every week. For CrEDTA measurement, urine was mixed 1:1 with 50 g/L trichloroacetic acid and centrifuged for 2 minutes at 14,000g. After dilution of the supernatants with 0.5 g/L CsCl solution, chromium was analyzed by inductively coupled plasma-atomic emission spectrophotometry (ICP-AES; Varian, Mulgrave, Australia). Total 3 × 24-hour feces was collected every 2 weeks and freeze-dried. Percentage fecal wet weight was determined to quantify diarrhea.

Myeloperoxidase (MPO) and Interleukin-1β (IL-1β) Analysis in Colonic Mucosa

Frozen colonic mucosal scrapings were manually crushed in liquid nitrogen using a precooled mortar and pestle. After homogenization, part of the powder was used for RNA isolation (see below), and the remaining portion was suspended in a solution containing 200 mM sucrose, 20 mM Tris (pH 7.4), 1 mM dithiothreitol, and protease inhibitors before protein analyses. In these samples, MPO activity was measured using a colorimetric assay as described by Grisham et al.27 MPO activity was expressed as units per mg protein, using purified MPO (Calbiochem, Darmstadt, Germany) as a standard. IL-1β was determined using an enzyme-linked immunosorbent assay (ELISA) kit (Biosource, Camarillo, CA) according to the manufacturer's instructions. Total protein content of the scrapings was measured spectrophotometrically using the BCA protein assay kit (Uptima, Montluçon, France) with bovine serum albumin (BSA; Sigma-Aldrich) as standard.

Measurement of Oxidative Damage in Colonic Mucosa

Lipid peroxidation generates malondialdehyde (MDA), which reacts with thiobarbituric acid. By measuring the amount of thiobarbiturate reactive substances (TBARS) in colonic mucosa, lipid peroxidation can be determined, as previously described by Ohkawa et al.28 The amount of mucosal TBARS was expressed as μmol MDA/mg protein. Protein carbonyls were analyzed as a marker of oxidative protein damage in colonic mucosa by ELISA as described.29 The concentration of carbonyls was expressed as nmol/mg protein.

Analysis of Antioxidants in Colonic Mucosa

The nonthiol antioxidant capacity was measured in colonic mucosa using the ferric reducing ability assay (FRAP) as described by Benzie and Strain.30 In the presence of antioxidants, ferric ions are reduced to ferrous ions, leading to formation of a colored ferrous-2,4,6-tripyridyl-s-triazine complex which can be measured spectrophotometrically. Values are expressed as nmol Fe2+/mg protein. Total glutathione and cysteine were measured using a slight modification of the assay described by Mansoor et al.31 This method determines total (reduced, oxidized, and protein-bound forms) glutathione and cysteine by high-performance liquid chromatography (HPLC). Compared to the original method, twice the sample volumes and reagents were used. Briefly, samples of 25 μL were injected into a 150 × 4.6 mm, 3 μm PLRP-S column, equipped with a PLRP-S guard column (Polymer Laboratories, Amherst, MA). Flow rate was 1 mL/min at 30°C with elution solvent A (0.1% trifluoroacetic acid [TFA], 5% acetonitrile) and solvent B (0.1% TFA, 80% acetonitrile), both diluted with distilled water. The elution profile was as follows: 0–20 minutes, 0% B; 20–25 minutes, 16% B; 25–30 minutes, 50% B, with retention time of bimane derivatives of glutathione of 23 minutes and cysteine of 13 minutes. By using a Spectra Systems FL2000 fluorometer (Spectra Physics, Mountain View, CA) excitation and emission were at 394 and 480 nm, respectively. Plotting and integration of peaks were performed by Chromeleon software 6.6 (Dionex, Sunnyvale, CA). Total glutathione and cysteine results are expressed as μmol/g protein.

Quantitative Real-time Polymerase Chain Reaction (Q-PCR) Analysis of Mucosal Samples

Total RNA was isolated from frozen homogenized colonic scrapings using TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. RNA was purified with RNeasy columns (Qiagen, Venlo, The Netherlands), and purity and concentration were quantified using the Nanodrop ND-1000 (Isogen Life Science, Maarssen, The Netherlands). Subsequently, 1 μg of RNA was reverse-transcribed to obtain complementary DNA (cDNA) using reverse transcriptase and oligo d(T)16 primer (Applied Biosystems, Foster City, CA) at 48°C for 30 minutes, followed by 5 minutes at 95°C. Differential expression of individual genes was assessed by Q-PCR using primers presented in Table 1. For each reaction, 2 μL of 10-fold diluted cDNA was added to 23 μL premix, containing 1 μL forward primer (10 μmol/L), 1 μL reverse primer (10 μmol/L), 8.5 μL RNase-free water, and 12.5 μL Power SYBR green PCR Master Mix (Applied Biosystems). The samples were incubated for 10 minutes at 95°C, followed by 40 amplification cycles. Each cycle consisted of 15 seconds at 95°C and 1 minute at 60°C, using the 7500 Fast Real-Time PCR System (Applied Biosystems). The run was completed with melting curve analysis for every PCR product to check for single product formation. A standard curve for each particular gene was run with every assay using serial dilutions of a reference sample (cDNA synthesized from a pooled RNA sample). mRNA levels were calculated from the appropriate standard curve, and these data were normalized against the housekeeping genes aldolase and ADP-ribosylation factor 1 as indicated.

Table 1. Primer Information
GeneAcc No.Forward Primer (5′ [RIGHTWARDS ARROW] 3′)Reverse Primer (5′ [RIGHTWARDS ARROW] 3′)
  1. Arf1, ADP-ribosylation factor 1; γGCSc, γ-glutamylcysteine synthetase catalytic subunit; γGCSm, γ-glutamylcysteine synthetase modifier subunit; GSS, glutathione synthetase; xCT, cystine transporter of system xmath image.

AldolaseNM_012495ATGCCCCACCCATACCCAGCACTAGCAGCAGTTGGCGGTAGAAGCG
Arf1NM_022518CTCCGAGATGCCGTTCTCTTGGGTAGCCTGAATGTACCAGTTCC
γGCScNM_012815TTACCGAGGCTACGTGTCAGACTGTCGATGGTCAGGTCGATGTC
γGCSmNM_017305ACCGGGAACCTGCTCAACTGGATTTGGGATCTCCATTCATTCAAG
GSSNM_012962CGGAGCTGGGTATTTTTGGAGTGCATGTTCAATGGCTTTGGT
xCTNM_001107673CCTGGCATTTGGACGCTACATTCAGAATTGCTGTGAGCTTGCA

Experimental Design of Study with Young and Older Rats: Animals and Diets

The experimental protocol was approved by the animal welfare committee of Wageningen University (The Netherlands). As in the nutritional study, HLA-B27 transgenic rats (n = 17) and their nontransgenic counterparts (n = 17) were used. The transgenic rats were 9–13 weeks old with a mean body weight of 170 g at the start of the experiment. The animals were housed and fed the control diet as described above. The eight youngest transgenic and eight nontransgenic rats were killed directly at the beginning of the experiment (t = 0), while the rest of the rats were monitored until they developed colitis, as explained above for the nutritional study. After 10 weeks (t = 10) the older rats were killed. Orbital blood was collected under anesthesia for measurement of ROS production (see below). During the dissections the colon was taken out, longitudinally excised, and washed in saline. Then the mucosa was scraped off. Part of this mucosal scraping was immediately snap-frozen for analysis of glutathione, cysteine, and MPO as described above. The other part was placed in ice-cold preoxygenated phosphate-buffered saline (PBS) for ex vivo analysis of ROS production (see below). In addition, part of the liver was excised and frozen in liquid nitrogen for measurement of glutathione and cysteine after being homogenized in the same sucrose solution as used for the colonic scrapings (see above).

Measurement of ROS Production

Oxygen radical production was determined by luminol-enhanced chemiluminescence, measured in an automated LB96V Microlumat Plus Luminometer (EG&G Berthold, Bad Wildberg, Germany) as described.32 Briefly, luminol reacts with ROS, which causes light emission (chemiluminescence). For the ROS production in blood, 200 μL of diluted blood (1:100 diluted in Hank's Balanced Salt Solution [HBSS]) was added per well in a microplate together with 20 μL luminol (0.5 mM luminol HBSS / 0.5% BSA) and 20 μL serum treated zymosan (human serum incubated with zymosan, 10 mg/mL) to activate phagocytes. For determination of the ROS production in colonic mucosa, the scrapings were cut into smaller pieces which were added to 200 μL PBS/tri-ethylamine per well, after which 20 μL luminol was added. Chemiluminescence was then monitored every minute for 60 minutes at 37°C. Data were analyzed using Winglow software (EG&G Berthold). One part of every blood sample was used for neutrophil counting. Briefly, leukocytes were counted by a hematology analyzer (Cell-Dyn; Abbott Diagnostics, Santa Clara, CA), after which they were stained with May–Grünwald to quantify neutrophils. Mean chemiluminescence (area under the curve) was expressed in relative light units (RLU) per neutrophil for the blood measurements, and in RLU/mg protein for the mucosal measurements.

Statistical Analysis

All results are expressed as mean ± SEM. The two predefined main comparisons of interest for the nutritional study were HLA-B27 rats on the control diet versus HLA-B27 rats on the antioxidants diet, and HLA-B27 rats on the antioxidants diet versus HLA-B27 rats on the diet supplemented with both antioxidants and calcium. Statistics on the dietary effects in transgenic rats were done by using one-way analysis of variance (ANOVA) or Kruskal–Wallis, depending on normality of the data. If significant, this was followed by Student's t-test (for normally distributed data) or Mann–Whitney U-test (for nonnormally distributed data) to identify the significant dietary effects. Nontransgenic rats were included in this study as a noncolitic reference to provide some general comparative information on baseline values of the measured parameters. Accordingly, statistics were done on nontransgenic rats (fed the control diet) versus the transgenic rats on the control diet to determine the inflammatory status as such. Again, Student's t-test was applied for normally distributed data and Mann–Whitney U-test for nonnormally distributed data. The predefined comparisons of interest for the study with young and older rats were young HLA-B27 rats versus older HLA-B27 rats, young nontransgenic rats versus young transgenic rats, and older nontransgenic rats versus older HLA-B27 rats. The latter two comparisons were to check the inflammatory status, while the first comparison was to test the potential aging effect. Statistics were done as described above for the nutritional study. Differences were considered statistically significant when P < 0.05 (all two-sided).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Nutritional Intervention: Animal Growth, Food Intake, and Colon Length and Weight

Animal growth and food intake of the transgenic rats were not affected by the different diets (Table 2). Nontransgenic rats had a higher body weight gain (Table 2; P < 0.01) than transgenic rats despite a similar food intake. Colons of HLA-B27 rats of the control group were the same length as the colons of the antioxidants group, which were shorter than colons of rats fed both extra antioxidants and calcium (Table 2; P < 0.05). Colon weight and the weight/length ratio of the colon of transgenic rats were much higher than that of nontransgenic rats (Table 2; P < 0.001 and P < 0.0001, respectively), which is one of the characteristics of intestinal inflammation. Colon weight and the colon weight/length ratio were not influenced by the different diets.

Table 2. Effect of Diet (Control, Supplemented with an Antioxidant Mixture (AOX) or Supplemented with Both Antioxidants and Calcium (AOX + Ca)) on Animal Growth, Food Intake, Colon Length, and Weight in HLA-B27 Transgenic Rats
 Nontransgenic ControlHLA-B27
ControlAOXAOX + Ca
  • Results are expressed as mean ± SEM.

  • a

    Significant difference between nontransgenic rats and control HLA-B27 rats.

  • b

    Significant difference between the AOX-group and the AOX+Ca-group.

Animal growth (g/day)1.0 ± 0.1a0.70 ± 0.080.76 ± 0.060.74 ± 0.07
Food intake (g/day)10.2 ± 0.410.4 ± 0.310.6 ± 0.210.3 ± 0.2
Colon length (cm)12.1 ± 0.3a13.1 ± 0.312.7 ± 0.313.6 ± 0.2b
Colon weight (g)0.9 ± 0.1a2.1 ± 0.12.1 ± 0.12.0 ± 0.1
Colon weight/length ratio0.07 ± 0.01a0.16 ± 0.010.17 ± 0.010.15 ± 0.01

Effect of Antioxidants and Calcium on Intestinal Permeability and Diarrhea

Intestinal inflammation is characterized by an increased intestinal permeability and diarrhea. In week 1 of the experiment, intestinal permeability was similar in all groups as measured by urinary CrEDTA excretion. In week 9, a clear increase in intestinal permeability was observed for the HLA-B27 rats on the control diet and the HLA-B27 rats fed the antioxidants diet. In contrast, transgenic rats on the diet with both extra antioxidants and calcium did not develop this colitis-related increase in intestinal permeability when compared to transgenic rats supplemented with antioxidants only (Fig. 1; P < 0.05). A similar protective effect was observed on diarrhea development. Diarrhea was quantified by the measurement of percentage fecal wet weight. Transgenic rats fed both antioxidants and calcium suffered less from diarrhea than transgenic rats fed antioxidants only (Fig. 2; P < 0.05). Intestinal permeability and diarrhea development were similar in HLA-B27 rats fed the diet supplemented with antioxidants only and the transgenic rats fed the control diet.

thumbnail image

Figure 1. Effect of dietary antioxidants (AOX) and calcium (Ca) on urinary chromium EDTA (CrEDTA) excretion, a marker for intestinal permeability, in HLA-B27 transgenic rats. CrEDTA was added to the diets and analyzed in urine by inductively coupled plasma-atomic emission spectrophotometry. Results are expressed as mean ± SEM. In week 1, intestinal permeability was similar for all groups. *Significant difference between the AOX-group and the AOX+Ca-group. #Significant difference between nontransgenic rats and control HLA-B27 rats.

Download figure to PowerPoint

thumbnail image

Figure 2. Effect of dietary antioxidants (AOX) and calcium (Ca) on diarrhea development in HLA-B27 transgenic rats. Diarrhea was quantified as the increase of percentage fecal wet weight compared to week 1. Results are expressed as mean ± SEM. *Significant difference between the AOX-group and the AOX+Ca-group. #Significant difference between nontransgenic rats and control HLA-B27 rats.

Download figure to PowerPoint

Antioxidants, Calcium, and Inflammatory Parameters

To quantify inflammation, MPO activity and IL-1β were measured in colonic mucosa. In addition, mucosal inflammation was graded histologically using a semiquantitative scoring system.26 As expected, the histological inflammation score, mucosal MPO activity, and IL-1β concentration were higher for HLA-B27 rats compared to nontransgenic rats. No significant dietary effects were observed for the histological inflammation score (Fig. 3A). As observed for the intestinal permeability and diarrhea outcomes, transgenic rats supplemented with both antioxidants and calcium had lower MPO activity and IL-1β levels compared to transgenic rats fed antioxidants only (Fig. 3B,C; P < 0.05). Dietary antioxidants alone did not influence the inflammatory parameters compared to the transgenic rats on the control diet.

thumbnail image

Figure 3. Effect of dietary antioxidants (AOX) and calcium (Ca) on inflammatory parameters in HLA-B27 transgenic rats. Rats were fed the different diets for 9 weeks, after which they were sacrificed to isolate the colonic mucosa for subsequent analyses. (A) Representative colonic sections were assigned a histological score in a blinded fashion using a 0–4 scale. (B) Colonic mucosal MPO activity was quantified using a colorimetric assay. (C) Colonic mucosal IL-1β levels were measured using an ELISA kit. Results are expressed as mean ± SEM. *Significant difference between the AOX-group and the AOX+Ca-group. #Significant difference between nontransgenic rats and control HLA-B27 rats.

Download figure to PowerPoint

Oxidative Stress-related Damage to Colonic Mucosa

As dietary antioxidants are proposed to protect against oxidative stress involved in intestinal inflammation, we measured oxidative damage in colonic mucosa. TBARS were assessed as a marker for lipid peroxidation. However, no differences were observed in colonic mucosal TBARS between the dietary groups. Moreover, TBARS levels were similar in nontransgenic and HLA-B27 rats (Table 3). Additionally, protein carbonyls were determined in colonic mucosa to evaluate oxidative damage to proteins. Similar to TBARS, diet did not influence the protein carbonyl content of colonic mucosa of transgenic rats. Moreover, protein carbonyls were even lower in transgenic rats compared to nontransgenic rats (Table 3; P < 0.005).

Table 3. Effect of Diet (Control, Supplemented with an Antioxidant Mixture (AOX) or Supplemented with Both Antioxidants and Calcium (AOX + Ca)) on Oxidative Stress Markers in Colonic Mucosa of HLA-B27 Transgenic Rats
 Nontransgenic ControlHLA-B27
ControlAOXAOX + Ca
  • Rats were fed the different diets for 9 weeks, after which they were sacrificed to isolate the colonic mucosa for subsequent analyses. TBARS are a marker for lipid peroxidation, whereas carbonyls were determined as a marker of oxidative protein damage. Total non-thiol antioxidant capacity was measured by the FRAP assay. Results are expressed as mean ± SEM.

  • FRAP, ferric reducing ability assay; GSH, glutathione; MDA, malondialdehyde; TBARS, thiobarbiturate reactive substances.

  • a

    Significant difference between nontransgenic rats and control HLA-B27 rats. No significant dietary effects were observed in the HLA-B27 rats.

TBARS (μmol MDA/mg protein)0.70 ± 0.0090.72 ± 0.040.72 ± 0.090.67 ± 0.07
Carbonyls (nmol/mg protein)0.67 ± 0.08a0.40 ± 0.040.42 ± 0.050.43 ± 0.07
FRAP (nmol Fe2+/mg protein)30.1 ± 6.740.9 ± 5.948.8 ± 8.637.3 ± 7.9
Total GSH (μmol/g protein)13.81 ± 1.84a0.02 ± 0.0020.15 ± 0.130.59 ± 0.38
Cysteine (μmol/g protein)0.76 ± 0.51a3.60 ± 1.406.53 ± 1.303.52 ±1.26

Effects on Mucosal Antioxidant Capacity

To determine whether the dietary antioxidant mixture indeed increased the antioxidant capacity of colonic mucosa, we applied the FRAP assay. Results showed that nonthiol antioxidants were not affected by the different diets. In addition, no substantial differences were observed in the FRAP assay when comparing nontransgenic rats to HLA-B27 rats (Table 3). Colonic mucosal glutathione concentrations were measured to determine whether oral glutathione supplementation increases mucosal antioxidant capacity by raising intestinal glutathione levels. Remarkably, glutathione levels were extremely low in colonic mucosa of HLA-B27 rats compared to nontransgenic rats (Table 3; P < 0.0001) and they were not affected by dietary treatment. As we were surprised about the huge difference in glutathione levels, we performed a post-hoc analysis in colonic mucosal samples from our previous study with HLA-B27 rats14 and found a similar difference between nontransgenic rats (18.29 ± 1.15 μmol/g protein) and transgenic rats (1.74 ± 0.41 μmol/g protein). To further investigate these exceptionally low mucosal glutathione levels, mucosal cysteine levels were analyzed as well, because cysteine is often the rate-limiting precursor for glutathione synthesis.33 The cysteine concentration in colonic mucosa of transgenic rats was notably high compared to nontransgenic rats and again no dietary effects were observed (Table 3; P < 0.01). Apparently, more than sufficient cysteine was present in colonic mucosa for the synthesis of glutathione. This may indicate that glutathione synthesis itself is compromised in some way in HLA-B27 transgenic rats. Therefore, we quantified mRNA expression of genes involved in glutathione synthesis. Gene expression of these enzymes (corrected for expression levels of reference genes) in transgenic rats fed the control diet relative to those of their nontransgenic counterparts was 0.8 for γ-glutamylcysteine synthetase (γ-GCS) catalytic subunit, 1.7 for γ-GCS modifier subunit (P = 0.046), and 1.3 for glutathione synthetase (GSS). Another potential explanation for the low mucosal glutathione levels is a defect in cellular cysteine uptake, resulting in impaired intracellular glutathione synthesis. Interestingly, the cystine transporter in system xmath image (xCT), responsible for transporting cystine into the cell in exchange for glutamate, was 6.2-fold upregulated in HLA-B27 rats compared to nontransgenic rats (P < 0.0001).

MPO, Glutathione, and Cysteine Levels in Young and Older HLA-B27 Rats

As the glutathione content of the colonic mucosa of the transgenic rats was remarkably low in comparison to the nontransgenic rats, we were interested to see whether this phenomenon was already present before colitis developed and thus could be a causal factor in this IBD model. Therefore, an additional experiment was performed in which young and older HLA-B27 rats were compared. MPO was measured to quantify colonic mucosal inflammation. At the start of the experiment, MPO levels were slightly higher in transgenic rats (0.27 ± 0.11 U/mg protein) compared to nontransgenic rats (0.07 ± 0.01 U/mg protein; P < 0.005). After 10 weeks, levels were further increased in the older HLA-B27 rats (1.1 ± 0.14 U/mg protein; P < 0.005). Interestingly, the colonic mucosal glutathione levels in young HLA-B27 rats were similar to the levels measured in nontransgenic rats (Fig. 4). However, after 10 weeks glutathione concentrations decreased in colonic mucosa of transgenic rats compared to those observed in HLA-B27 rats at the start of the experiment (Fig. 4; P < 0.0001). Although the decrease in glutathione was still quite substantial, it was not yet as large as observed in the nutritional intervention study (Table 3). Mucosal cysteine levels were slightly higher in transgenic rats compared to nontransgenic rats at t = 0 (Fig. 4; P < 0.001) and were further increased with time in older HLA-B27 rats compared to young HLA-B27 rats (Fig. 4; P < 0.0001). Since glutathione is predominantly present in the liver,34 we also studied its levels in this organ. The glutathione concentration was similar in the liver of young transgenic rats (29.6 ± 2.7 μmol/g protein) and young nontransgenic rats (29.0 ± 2.1 μmol/g protein). Comparable levels were observed in nontransgenic rats at t = 10 (29.1 ± 2.4 μmol/g protein). After 10 weeks, liver glutathione levels were decreased in the transgenic rats (24.3 ± 1.7 μmol/g protein), but this was not significant. In summary, the low glutathione levels in colonic mucosa of HLA-B27 rats are only present when intestinal inflammation has developed, and not when the rats were young and still relatively healthy.

thumbnail image

Figure 4. (A) Glutathione and (B) cysteine levels in colonic mucosa of HLA-B27 transgenic rats and their nontransgenic counterparts, determined at a young (t = 0) and older (t = 10) age. *Significant difference between HLA-B27 rats (t = 0) and HLA-B27 rats at (t = 10). #Significant difference between HLA-B27 rats (t = 0) and nontransgenic rats (t = 0).

Download figure to PowerPoint

ROS Production in Blood and Colonic Mucosa

Another important finding in the nutritional intervention study was the absence of oxidative stress-related damage in the HLA-B27 rats having colitis, despite the low colonic mucosal glutathione levels (Table 3). To examine whether this was due to a defect in ROS production capacity, we measured these ROS in both blood and colonic mucosa in the additional experiment. The ROS production in whole blood was similar in the young nontransgenic and young transgenic rats (Table 4). However, in the transgenic rats at t = 10 the ability to produce ROS was drastically lower than in nontransgenic rats at t = 10 (Table 4; P < 0.0001). This suggests that the respiratory burst of neutrophils becomes compromised in older HLA-B27 rats. As expected, the amount of neutrophils increased in the older transgenic rats (5.7 ± 0.8 × 109 neutrophils/L; P < 0.0001) compared to the young transgenic rats (1.3 ± 0.3 × 109 neutrophils/L), which reflects the development of colitis. Notwithstanding that, the production of ROS in colonic mucosa did not increase in older HLA-B27 rats compared to young transgenic rats, and was also not different from nontransgenic rats (Table 4).

Table 4. Reactive Oxygen Species (ROS) Production in Blood and Colonic Mucosa of Young and Older HLA-B27 Transgenic Rats Compared to Nontransgenic Rats
ROS Production NontransgenicHLA-B27
  • ROS production was quantified by luminol-enhanced chemiluminescence in blood and colonic mucosa of HLA-B27 transgenic rats and their nontransgenic counterparts, both at a young (t = 0) and older (t = 10) age. Results are expressed as mean ± SEM. RLU, relative light unit.

  • a

    Significant difference between HLA-B27 rats (t = 10) and nontransgenic rats (t = 10).

Blood (RLU/neutrophil)Young105 ± 1786 ± 27
 Older134 ± 1427 ± 3a
Colonic mucosa (*106 RLU/mg protein)Young1.1 ± 0.21.9 ± 0.5
 Older1.5 ± 0.32.2 ± 0.6

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

In this study we showed that supplementation of an antioxidant mixture had no protective effect on colitis development in HLA-B27 transgenic rats despite very low glutathione levels in colonic mucosa. In addition, we clearly demonstrated that a dietary intervention with both antioxidants and calcium resulted in less severe colitis in HLA-B27 transgenic rats compared to supplementation with the antioxidant mixture only. Supplementation with antioxidants plus calcium significantly attenuated the colitis-related increase in intestinal permeability, diarrhea, and inflammatory markers. Overall, these results imply that only calcium contributed to the protective effects. Furthermore, we showed that colonic glutathione levels decreased when colonic inflammation progresses, suggesting that glutathione deficiency is not a primary cause of colitis development in this rat model. The present study also suggests that the absence of oxidative stress-related mucosal damage in the HLA-B27 rats resulted from compromised ROS production, which might explain the ineffectiveness of antioxidant supplementation.

During active inflammation in IBD, leukocytes massively infiltrate the intestinal mucosa, releasing large amounts of ROS. ROS are thought to be partly responsible for the tissue damage in IBD. To counteract the possible harmful effects of ROS, the intestinal mucosa is equipped with an antioxidant defense system.3 As the balance between the amount of ROS produced and the available antioxidative defense is suggested to be impaired in IBD,35 we investigated the hypothesized relief of inflammatory symptoms in the HLA-B27 rats due to antioxidant supplementation. This study showed that the applied antioxidant mixture neither offered protection to any of the determined characteristics of intestinal inflammation nor affected oxidative stress-related outcomes. One possible explanation for the absence of an antioxidant effect in the present study might be that the transgenic animals did not suffer from mucosal oxidative tissue damage. Indeed, we could not detect signs of oxidative damage in the HLA-B27 rats at the time of tissue collection despite the presence of significant inflammation. Frequently used biomarkers of cellular lipid peroxidation (TBARS) and protein oxidation (carbonyls) were not increased at all in HLA-B27 rats with evident colitis. This shows that intestinal inflammation does not necessarily coincide with evident oxidative stress, as is often presumed. To gain further insight into the observed absence of oxidative damage in HLA-B27 rats, we performed a second animal experiment to investigate whether the leukocyte influx indeed caused elevated ROS levels in colonic mucosa. The results showed that there was no increase of ROS production in the colon of transgenic rats suffering from intestinal inflammation. Moreover, in blood the capacity to release ROS was even lower per neutrophilic granulocyte in HLA-B27 rats. So the absence of increased ROS production in colonic mucosa might explain the lack of oxidative stress-related damage in HLA-B27 rats. Apparently, the observed histological tissue damage in this model is mainly caused by other processes, for instance, by activation of matrix metalloproteinases.14, 36 Thus, the compromised respiratory burst might explain why antioxidant supplementation is not effective in this model. Interestingly, the observation that antioxidant supplementation did not ameliorate clinical endpoints can be interpreted as additional proof that oxidative stress is not a pathogenic factor in colitis development in the HLA-B27 transgenic rats. We like to stress that antioxidant supplementation can still be protective in IBD, but our results show that they must be tested in another animal model mimicking IBD.

One of the remarkable findings of the present study is the extremely low colonic mucosal glutathione level in the HLA-B27 transgenic rats. Even this very low glutathione concentration did not lead to measurable oxidative damage. This raises doubts about the importance of colonic mucosal glutathione in preventing oxidative stress. Our finding corroborates a study with rats having intestinal glutathione totally depleted with buthionine sulfoximine, in which also no signs of oxidative damage could be detected after intestinal infection.37 Importantly, the absence of an effect of glutathione on oxidative stress in the present study might also result from impaired ROS production. The low intestinal mucosal glutathione levels did not positively respond to oral administration of this thiol tripeptide. To further investigate the glutathione deficit, colonic mucosal mRNA expression of the glutathione synthesis enzymes GSS and γ-GCS was determined. Levels were similar in transgenic and nontransgenic rats, except for γ-GCS modifier subunit, which was slightly increased in transgenic rats. Also, availability of precursor amino acids seems sufficient, since cysteine is the rate-limiting substance for glutathione synthesis38 and colonic mucosal cysteine levels in HLA-B27 rats were even significantly higher compared to nontransgenic rats. Hence, cystine supplementation might have been ineffective to increase mucosal glutathione levels. Furthermore, xCT mRNA expression was notably increased in the transgenic rats. This suggests an increased demand for intracellular cysteine to be used for cellular glutathione synthesis. To follow up on the low glutathione levels, an additional study was performed in which young and older HLA-B27 rats were compared to determine the potential causality of low glutathione to colitis development. Since glutathione levels were not decreased in the young transgenic rats, the intestinal glutathione deficit developed with progression of colitis in these rats. Moreover, the deficit was only present in the colon, as liver levels were not significantly decreased. Interestingly, this corresponds with the lower expression of HLA-B27 mRNA in liver compared to colon tissue.16 The cause of the decrease of glutathione in the colonic mucosa is unknown. A defect in the production of NADPH, which is necessary for the reduction of oxidized glutathione, might play a role.33 This defect could as well account for the compromised ROS production, since an NADPH-dependent oxidase catalyzes the ROS production by neutrophils,3 but this requires further investigation. Another explanation for the low glutathione levels in the transgenic rats might be the presence of inhibitors of glutathione synthesis, like cysteamine.39, 40 In the nutritional intervention study and our previous HLA-B27 study,14 elevated cysteamine levels in colonic mucosa were observed (data not shown). However, in the additional experiment with young and older rats, no cysteamine was detected in the colonic mucosa, indicating that its role, if any, is inconsistent. Possibly, other inhibitors are implicated in the low glutathione levels of the transgenic rats.40 Remarkably, El Yousfi et al41 report increased glutathione levels in the colon of HLA-B27 rats compared to nontransgenic rats but the expression units and colonic part (mucosa or whole colon) are unclear. Therefore, it is difficult to compare their results with ours, which were consistent in three independent HLA-B27 studies.

For better understanding of the HLA-B27 model and its application to study mechanisms relevant to human IBD, it is important to investigate whether the compromised capacity of neutrophils to produce ROS together with the low colonic glutathione levels plays a role in the development of colitis in this model. As HLA-B27-expressing innate immune cells have an essential role in this model,17 we speculate that these cells do not adequately respond to invading bacteria due to their compromised ROS production, which might lead to an overload of antigen presentation to T cells, resulting in an overreactive immune system. Further research is necessary to support this hypothesis.

The beneficial effect of dietary calcium on colitis development in HLA-B27 transgenic rats was shown before by our group.14 Additionally, we have shown in several controlled rat and human infection studies that dietary calcium improves intestinal resistance and strengthens the mucosal barrier.18–20 The present study shows that supplemental calcium is also protective in colitis development on another background diet, i.e., a diet with additional antioxidants. Nevertheless, it is inappropriate to draw strong conclusions about the effect of calcium in this study, as no dietary group supplemented with calcium only was included. However, since studies in the field of colon cancer research have shown that calcium affects intestinal epithelial cell homeostasis,42, 43 we were curious whether cell cycle function was also affected in mucosal samples from the present study. The dietary treatments did not affect either cell proliferation (Ki-67 immunohistochemistry) or apoptosis (caspase-3 activity) (data not shown). These results suggest that the protective influence of calcium on colitis symptoms, e.g., intestinal permeability, diarrhea, and inflammation markers, cannot be explained by effects on mucosal cell proliferation and apoptosis. We hypothesize that the beneficial effect of calcium involves precipitation of cytotoxic surfactants by luminal calcium and thus modulation of the intestinal microbiota.

In conclusion, supplementation of a mixture of various antioxidants had no effect on colitis severity in HLA-B27 transgenic rats, whereas the combination of this mixture plus calcium showed significantly protective effects. Therefore, we speculate that this protective effect is contributed by calcium only. Very low glutathione levels were observed in colonic mucosa of HLA-B27 rats, which developed during progression of intestinal inflammation. In addition, compromised ROS production might explain why there is no oxidative stress-related mucosal damage in the transgenic rats and why antioxidant supplementation was ineffective. The presumed absence of adverse effects of nutritional interventions, in contrast to commonly prescribed IBD medication, makes it worthwhile to further explore the potential benefits of nutritional intervention in intestinal inflammation in humans.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The authors thank the biotechnicians of the Small Animal Centre (Wageningen University and Research Centre, The Netherlands) for expert assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES