Article Text

PDF

Reduced mucin sulfonation and impaired intestinal barrier function in the hyposulfataemic NaS1 null mouse
  1. P A Dawson1,
  2. S Huxley1,
  3. B Gardiner2,
  4. T Tran3,
  5. J L McAuley3,
  6. S Grimmond2,
  7. M A McGuckin3,
  8. D Markovich1
  1. 1
    School of Biomedical Sciences, University of Queensland, St Lucia, Australia
  2. 2
    Institute for Molecular Bioscience, University of Queensland, St Lucia, Australia
  3. 3
    Mucosal Diseases Program, Mater Medical Research Institute, Mater Hospital, South Brisbane, Australia
  1. Professor D Markovich, University of Queensland, St Lucia, QLD 4072, Australia; d.markovich{at}uq.edu.au

Abstract

Objective: Sulfate (SO42−) is an abundant component of intestinal mucins and its content is decreased in certain gastrointestinal diseases, including inflammatory bowel disease. In this study, the hyposulfataemic NaS1 sulfate transporter null (Nas1−/−) mice were used to investigate the physiological consequences of disturbed sulfate homeostasis on (1) intestinal sulfomucin content and mRNA expression; (2) intestinal permeability and proliferation; (3) dextran sulfate sodium (DSS)-induced colitis; and (4) intestinal barrier function against the bacterial pathogen, Campylobacter jejuni.

Methods: Intestinal sulfomucins and sialomucins were detected by high iron diamine staining, permeability was assessed by fluorescein isothiocyanate (FITC)–dextran uptake, and proliferation was assessed by 5-bromodeoxyuridine (BrdU) incorporation. Nas1−/− and wild-type (Nas1+/+) mice received DSS in drinking water, and intestinal damage was assessed by histological, clinical and haematological measurements. Mice were orally inoculated with C jejuni, and intestinal and systemic infection was assessed. Ileal mRNA expression profiles of Nas1−/− and Nas1+/+ mice were determined by cDNA microarrays and validated by quantitative real-time PCR.

Results: Nas1−/− mice exhibited reduced intestinal sulfomucin content, enhanced intestinal permeability and DSS-induced colitis, and developed systemic infections when challenged orally with C jejuni. The transcriptional profile of 41 genes was altered in Nas1−/− mice, with the most upregulated gene being pancreatic lipase-related protein 2 and the most downregulated gene being carbonic anhydrase 1 (Car1).

Conclusion: Sulfate homeostasis is essential for maintaining a normal intestinal metabolic state, and hyposulfataemia leads to reduced intestinal sulfomucin content, enhanced susceptibility to toxin-induced colitis and impaired intestinal barrier to bacterial infection.

Statistics from Altmetric.com

Inorganic sulfate (SO42−) is an abundant anion in mammalian plasma and is essential for numerous metabolic and cellular processes (reviewed in Markovich1). SO42− conjugation (sulfonation) of structural components (such as glycosaminoglycans, proteins, cholesterol and carbohydrates) is essential for the maintenance of normal structure and function of tissues. 2 Sulfonated carbohydrates are an abundant component of mucins, which are the major macromolecular component of gastrointestinal mucus and the mucosal glycocalyx, that lubricates and protects the underlying epithelium.3 These complex glycoproteins consist of a central protein backbone with a dense array of O-linked carbohydrates (comprising >70% of their mass), that are assembled progressively in the Golgi apparatus by an array of glycosyltransferases and sulfotransferases.4 5 The family of human mucin genes has 17 members69 that can be divided into: secreted gel-forming mucins, cell surface mucins and secreted non-gel-forming mucins. The predominant mucins expressed in the gastrointestinal tract are the cell surface mucins (MUC1, 3A, 3B, 4, 12, 13 and 17), and the gel-forming mucins (MUC2, MUC5AC and MUC6).10 Abnormal expression of these mucins has been reported for some cases of inflammatory bowel disease.10 11 MUC1 polymorphisms have been associated with chronic gastritis and intestinal metaplasia.12 Muc1 null mice are more susceptible to intestinal bacterial infection.1315 Muc2 null mice develop spontaneous colitis16 and intestinal tumours.17 Mice containing Muc2 mutations leading to aberrant mucin complex assembly develop spontaneous colitis.18 Mice lacking the enzyme core 3 β 1,3-N-acetylglucosaminyltransferase (responsible for production of mucin core 3 O-glycans) show altered glycosylation of intestinal Muc2,19 produce less Muc2 and have increased intestinal permeability and increased susceptibility to induced colitis, suggesting that aberrant post-translational processing of mucins can lead to intestinal pathology. Decreased mucin sulfonation has been proposed to result in enhanced mucin degradation.20 Reduced sufomucin levels are found in gastrointestinal diseases such as colon cancer and inflammatory bowel disease.17 2124

Sufficiently high levels of sulfate and its universal sulfate donor 3′-phosphoadenosine 5′-phosphosulfate (PAPS) need to be maintained for sulfonation reactions to function effectively.25 Circulating SO42− levels are maintained by the NaS1 sulfate transporter which is primarily expressed in the ileum and kidney.26 27 Previously, we generated an NaS1 null (Nas1−/−) mouse, which has increased urinary sulfate excretion and hyposufataemia.28 One approach to better understanding the physiological consequences of altered sulfate homeostasis was to investigate the role of sulfate availability on endogenous molecules, such as mucins, which are biotransformed by sulfonation (reviewed in Nieuw Amerongen et al20). In addition to the importance of mucin sulfonation, the strong expression of NaS1 in the ileum and its role in intestinal sulfate absorption makes the ileum an appropriate tissue to investigate the effects of a disrupted NaS1 gene (Slc13a1) on the ileal transcriptional profile. Gene arrays are a valuable tool for investigating the intestinal transcriptome and have provided important insights into genes that are involved in ulcerative colitis, inflammatory bowel disease and the inflammatory response to bacterial gut pathogens, such as Campylobacter jejuni.2931 Studies of the Nas1−/− mouse can provide valuable insights into the physiological consequences of altered sulfate homeostasis and function of the NaS1 gene, which are yet to be characterised in human disease. These studies are relevant to single nucleotide polymorphisms (N174S and R12X) in the human NaS1 gene, which lead to 60% and 100% loss of function, respectively.32 The aims of this study were to determine the intestinal sulfomucin content, susceptibility to toxin-induced colitis and bacterial infection, and ileal transcriptional profile of Nas1−/− mice, and compare those with wild-type mice. Our findings show substantial diminution of intestinal sulfomucin content, enhanced colitis induced by a luminal toxin, reduced intestinal barrier function and changes in the mRNA expression of genes involved in metabolism and the immune response.

MATERIALS AND METHODS

Mice

We previously generated Nas1 knockout (Nas1−/−) mice in which the NaS1 gene was disrupted by targeted mutagenesis.28 Nas1−/− mice exhibit increased urinary sulfate excretion, hyposulfataemia (0.2 mM serum sulfate, ∼75% decrease), growth retardation, reduced fertility, as well as increased levels of hepatic and serum lipid.28 33 Studies were performed on mice with a mixed genetic background (129Sv and C57BL/6J).28 Nas1−/− mice and wild-type Nas1+/+ littermates (control) were housed at a constant temperature (23±1°C) with a 12 h/12 h light/dark cycle (lights on at 06:00 h and off at 18:00 h). Male mice were weaned at 3 weeks of age and were then fed a standard rodent chow (no. AIN93G: Glen Forrest Stockfeeders, Glen Forest, Western Australia) and water ad libitum, and were used for approved experiments in accordance with the guidelines of the University of Queensland Animal Ethics Committee.

Assessment of mucin sulfonation

The entire large intestine, ileum and jejunum was Swiss-rolled for histological analysis and stained with high iron diamine (HID) and alcian blue.34 An observer blind to the genotype of the mice scored the intensity (0–4) of HID staining (sufonation) and the proportion of goblet cells positive (0–4 in quartiles) in the jejunum, ileum, and proximal, mid and distal colon. The two distinct goblet cell lineages within the proximal and mid colons were assessed separately.

Intestinal cell proliferation in Nas1+/+ and Nas1−/− mice

Proliferation was determined by immunohistochemical detection of 5-bromodeoxyuridine (BrdU) incorporation 3 h after intraperitoneal injection of 100 μg BrdU/g body weight in eight mice of each genotype. The numbers of BrdU-positive and -negative nuclei were counted in 10 crypt–villus units in the ileum, and 10 crypts in the proximal and distal colon from each mouse.

DSS administration

Nas1+/+ and Nas1−/− mice (n = 8–10 per group) received 5.0% dextran sulfate sodium (DSS, molecular weight of 30–50 kDa, MP Biomedicals, Solon, Ohio, USA) in their drinking water to model acute colitis.35 Daily assessment of DSS-treated mice was performed blinded to genotype, and included measurements of body weight, evaluation of stool consistency and the visual detection of blood in the stool, using a scale from 0 to 4, as previously described.36

C jejuni challenges

C jejuni strain 81116 was grown and suspended in warmed Brucella broth as previously described.14 Nas1+/+ and Nas1−/− mice (n = 10 per group) were orally inoculated with 107 colony-forming units (CFU) of C jejuni and sacrificed at 3 days postchallenge. Large and small intestines and liver were dissected and either placed in 10% formalin for histological analyses or homogenised in broth then plated onto selective agar to determine CFU/g tissue, as previously described.14

Histopathological analysis

Tissues were dissected into approximately 50 volumes of 10% buffered formalin and fixed for 3 days prior to paraffin embedding. Embedded tissue was sectioned, stained with H&E and examined by light microscopy. Intestinal sections were scored blinded to genotype, using the quantitative histological scale37 from 0 to 2, 3 or 5, as previously described.18

Haematological measurements

Blood samples were collected from Nas1−/− and Nas1+/+ mice (n = 8–10 per group), and immediately mixed with K2EDTA. Haematological analysis was completed using an automated haemotological analyser (Sysmex, Kobe, Japan).

Assessment of intestinal permeability

Intestinal macromolecular permeability was determined in Nas1+/+ and Nas1−/− mice (n = 8 per group) by administering 400 μg of fluorescein isothiocyanate (FITC)–dextran (4 kDa, Sigma-Aldrich, Sydney, NSW, Australia)/g body weight by gavage and measuring the fluorescence of FITC in plasma samples (FLA-5100 fluorimeter, Fujifilm, Toyko, Japan) 2 and 5 h after gavage, as previously described.19 In order to assess the translocation of resident bacterial flora across the mucosa, we used quantitative real-time PCR (qRT-PCR) to quantitate bacterial 16S rDNA from mesenteric lymph nodes of Nas1+/+ and Nas1−/− mice (n = 8 per group) as previously performed.38 39

Gene expression profiling

We isolated total RNA from the ileum of male Nas1+/+ and Nas1−/− mice at 09:00 h, using previously described methods.28 Pooled total RNA (n = 14 for each genotype) was amplified, labelled, and hybridised overnight at 42°C using previously described methods40 and 22K mouse Compugen long oligonucleotide arrays from the SRC Microarray Facility, University of Queensland (ARC Centre for Functional and Applied Genomics). Ileal RNA samples from Nas1+/+ and Nas1−/− mice were directly compared in quadruplicate with dye swapping incorporated into two of the hybridisations to account for bias. Arrays were scanned on an Agilent scanner 600B (Agilent Technologies, Forest Hill, VIC, Australia) and foreground/background signal intensities quantified using Imagene 5.0 (BioDiscovery, El Segundo, California, USA). All data were normalised using a combination of the print tip-dependent lowess and between-array scaling algorithms from Bioconductor (LIMMA) and SMA.41 All primary data (including images), data transformations and methods are available via the comprehensive microarray relational database, BASE (http://kidney.scgap.org/BASE). The entire data set is also available from GEO (accession number: GSE9958). Genes were functionally annotated via the web-enabled tools: Clonefinder (http://microarray.imb.uq.edu.au/clonefind.html) and DAVID (http://david.niaid.nih.gov/david/version2/index.htm).

Quantitative RT-PCR

In order to validate our microarray data, we examined the mRNA levels of seven genes that were found to be increased (Bok, Car4, Avil and Pnliprp2) or decreased (Car1, Casp11 and Orm1). Total RNA (2 μg) was reverse transcribed by using random hexamers and an Omniscript RT kit (Qiagen, Doncaster, VIC, Australia) as recommended by the manufacturer. PCR was performed in quadruplicate with 5 μl of cDNA (from 25 ng of RNA) and 10 μl of mastermix containing Platinum SYBR Green qPCR SuperMix UDG (Invitrogen, Mount Waverley, VIC, Australia) and 200 nM forward and reverse primers (table 1) in a Rotor-Gene 3000 thermal cycler (Corbett Research, Sydney, Australia). The thermal cycling protocol was: 50°C for 2 min; 94°C for 2 min; 45 cycles of 94°C for 1 s, 60°C for 10 s, and 72°C for 15 s. RNA expression levels and absolute threshold cycle values (Ct values) of each gene were normalised to those of rRNA with the Rotor-Gene 6 software (Corbett Research). Amplification specificity was confirmed by melting curve analysis and agarose gel electrophoresis.

Table 1 Primers used for PCR amplification

Statistical analyses

Where data were normally distributed, the statistical significance of differences between Nas1+/+ and Nas1−/− groups was assessed by an unpaired Student t test. Non-continuous data (disease activity index and histology scores) were analysed by the Mann–Whitney test. Differences in body weight were analysed by the Dunnett multiple comparison test. For the gene expression profiling analyses, differential expression was defined using the B statistic method, where both fold change and variance of signals in replicates is used to determine the likelihood that genes are truly differentially expressed.41

RESULTS

Sulfomucin content in Nas1−/− mice

Small and large intestinal sections stained with HID and alcian blue showed predominant sialomucin staining in Nas1−/− mice, whereas high sulfomucin content was observed in the goblet cells of Nas1+/+ mice (fig 1). Significantly reduced mucin sulfonation was observed in the jejunum, ileum, gland lineage of the proximal and mid colon, as well as the surface lineage of the mid colon of Nas1−/− mice when compared with Nas1+/+ mice (fig 1), demonstrating that decreased mucin sulfonation occurs in both the small and large intestine of Nas1−/− mice. Furthermore, there was a slight reduction in the number of goblet cells in the ileum of Nas1−/− mice when compared with Nas1+/+ mice (fig 1). Increased intestinal proliferation is a hallmark of murine colitis and was found to be increased in Muc2/17 and Muc2 mutant mice.18 In order to ascertain whether intestinal proliferation was altered in Nas1−/− mice, we used BrdU metabolic labelling to examine ileum, and proximal and distal colon epithelial proliferation. No significant differences were found for the number of BrdU-positive epithelial cells in Nas1−/− and Nas1+/+ mice (Supplementary table 1), suggesting that an intestinal inflammatory and proliferative response does not ensue spontaneously when mucin sulfonation decreases.

Figure 1

Assessment of intestinal mucin sulfonation in wild-type (Nas1+/+) and NaS1 sulfate transporter null (Nas1−/−) mice. Representative photomicrographs (n = 8 per group) from the ileum, as well as proximal and distal colon, showing a high sulfomucin content in the goblet cells of Nas1+/+ mice (dark brown-stained cell thecae), whereas more predominant sialomucin staining (blue-stained thecae) was observed in Nas1−/− mice. Bar, 50 μm. The individual mucin sulfonation scores (product of the intensity and proportion scores), median, range and p values from the Mann–Whitney U test are shown. Goblet cell numbers were counted in 10 crypt–villus units from the ileum of each mouse and the mean (SEM) determined and plotted (bottom right), with the p value from a unpaired t test shown, NS, non-significant.

DSS-induced colitis in Nas1−/− mice

The reduced sulfomucin levels in Nas1−/− mice provide a model of reduced mucin sulfonation, as occurs in some human inflammatory bowel diseases, including ulcerative colitis.23 4244 Intestinal mucus is an important component of the barrier to luminal toxins, and altered mucin glycosylation has been shown to enhance susceptibility to luminal toxins,19 which led us to compare the sensitivity to DSS-induced colitis between Nas1−/− and Nas1+/+ mice. After acute administration of 5% DSS in drinking water for 6 days (to model acute colitis), Nas1−/− mice showed significantly enhanced colitis, including increased rectal bleeding and diarrhoea when compared with Nas1+/+ mice (fig 2A). Body weights were significantly decreased in Nas1−/− mice after 4 (11% decrease), 5 (18% decrease) and 6 (26% decrease) days of DSS treatment, whereas Nas1+/+ body weights were not significantly decreased until 5 (12% decrease) and 6 (16% decrease) days of DSS treatment (fig 2B), indicating an earlier onset of colitis in Nas1−/− mice. Whole blood profiles revealed no change in white blood cell and platelet counts, whereas significant decreases (by 25%) in red blood cell count, haematocrit and haemoglobin levels were found in Nas1−/− mice when compared with Nas1+/+ mice (table 2), implying that Nas1−/− mice have enhanced DSS-induced intestinal bleeding. This was supported by histological findings of increased ileal villus damage (fig 2C,F), colon crypt loss and increased lamina propria neutrophil cell counts in Nas1−/− mice (fig 2D,H), when compared with Nas1+/+ mice (fig 2C,D,E,G). These data show that Nas1−/− mice are more sensitive to an acute 5% DSS treatment than wild-type mice.

Figure 2

Assessment of dextran sulfate sodium (DSS)-induced colitis in Nas1+/+ and Nas1−/− mice. (A) Individual disease index scores, median, range and p values after administration of 5% DSS in drinking water for 6 days, showing an increased disease activity in Nas1−/− mice. (B) Body weights (mean (SD)) of Nas1−/− mice (n = 10) were reduced after 4, 5 and 6 days of 5% DSS treatment, whereas body weights of Nas1+/+ mice (n = 10) were reduced after 5 and 6 days. #p<0.01 compared with baseline, *p<0.05 and **p<0.01 when compared with Nas1+/+ mice. (C and D) Individual histopathology scores, median, range and p values showing increased (C) ileal and (D) proximal colon damage in Nas1−/− mice after 7 days of 5% DSS treatment. (E–H) H&E-stained ileum (E and F) and proximal colon (G and H) sections after 7 days of 5% DSS in drinking water, showing villi shortening, leucocytic infiltration and epithelial cell loss in the ileum (F, arrowhead), and crypt loss and ulceration in the proximal colon (H, arrow), which were not observed in Nas1+/+ mice (E and G). Bar, 50 μm. Nas1−/−, NaS1 sulfate transporter null mice; Nas1+/+, wild-type mice.

Table 2 Blood profiles for Nas1+/+ and Nas1−/− mice after administration of 5% DSS in drinking water (t = 7 days)

C jejuni infection in Nas1−/− mice

To compare colonisation of the gastrointestinal tract and penetrance through the mucosal barrier by bacterial pathogens, mice were challenged orally with C jejuni and then sacrificed 3 days after infection, and the gastrointestinal tract and liver removed for bacterial quantitation. C jejuni equally colonised the small and large intestine of both Nas1+/+ and Nas1−/− mice. However, C jejuni were present in the liver of 6/10 Nas1−/− mice vs 0/10 Nas1+/+ mice (fig 3), suggesting that sulfate depletion leads to an impaired intestinal barrier to infection in Nas1−/− mice. This was supported by histological findings of increased villus damage in the ileum of Nas1−/− mice (fig 4B,D,F,H,I), when compared with Nas1+/+ mice (fig 4A,C,E,G,I).

Figure 3

Colonisation of the large (LI) and small (SI) intestine, and liver of Nas1−/− and Nas1+/+ mice, orally inoculated with C jejuni strain 81116. Tissues were sampled at 3 days postinoculation. Each bar is the mean (SEM) and is representative of two experiments with similar data. *p = 0.011 compared with Nas1+/+ mice. CFU, colony-forming unit; Nas1−/−, NaS1 sulfate transporter null mice; Nas1+/+, wild-type mice.

Figure 4

H&E-stained ileum sections from Nas1+/+ and Nas1−/− mice, orally inoculated with C jejuni strain 81116. Representative histological demonstration of increased damage in the ileum of two Nas1−/− mice (B,D and F,H) when compared with Nas1+/+ mice (A,C and E,G). Note that in Nas1+/+ mice even when inflammation is present (E and G), the villi remain intact, whereas in Nas1−/− mice there is more damage and shedding of villus epithelium (B and D) leading to loss of villi and elongation of crypts (F and H). Bar, 50 μm. (I) Individual histopathology scores, median, range and p values showing increased ileal villus damage in Nas1−/− mice after an oral administration of C jejuni. Nas1−/−, NaS1 sulfate transporter null mice; Nas1+/+, wild-type mice.

Intestinal barrier function in Nas1−/− mice

Since C jejuni colonised the liver of Nas1−/− mice (fig 3), we aimed to determine if there were any differences in the intestinal permeability of Nas1−/− and Nas1+/+ mice by measuring plasma fluorescence after a single oral administration of FITC–dextran. Mean plasma FITC–dextran concentrations were increased by 30% (p = 0.14) and 24% (p = 0.059) at 2 and 5 h, respectively, in Nas1−/− mice (fig 5A), demonstrating mildly increased intestinal permeability to macromolecules. To assess the translocation of resident bacterial flora across the submucosa, we used qRT-PCR to amplify bacterial 16S rDNA from mesenteric lymph nodes of Nas1−/− and Nas1+/+ mice. No significant differences were found for 16S rDNA levels (fig 5B), indicating a similar mucosal integrity against the resident bacterial flora of Nas1−/− and Nas1+/+ mice, which is consistent with the lack of spontaneous colitis in Nas1−/− mice.

Figure 5

Assessment of intestinal permeability and barrier function against resident bacterial flora in Nas1+/+ and Nas1−/− mice. (A) Plasma concentration of fluorescein isothiocyanate (FITC)–dextran 2 and 5 h after oral adminstration. Means (SEM), n = 8 per group. (B) Relative levels of bacterial 16S rDNA in mesenteric lymph nodes (MLN) of Nas1+/+ and Nas1−/− mice, as assessed by using quantitiative real-time PCR. Individual data points, medians, ranges and p values are shown. GADPH, glyceraldehyde phosphate dehydrogenase; Nas1−/−, NaS1 sulfate transporter null mice; Nas1+/+, wild-type mice.

Ileal transcriptional profile of Nas1−/− mice

Since NaS1 is highly expressed in the ileum26 and altered mucin sulfonation was observed in the ileum of Nas1−/− mice, we determined the ileal gene expression profile of Nas1−/− and Nas1+/+ mice using cDNA microarrays, which revealed transcriptional differences in Nas1−/− mice. We have listed all genes assessed as having a B-statistic score >0 (tables 3 and 4). Of these, 23 transcripts were upregulated and 18 downregulated in Nas1−/− mice when compared with Nas1+/+ mice. No significant differences (B<0) were found for mRNA levels of mucin genes, including Muc1 (Nas1−/−:Nas1+/+ ratio = 1.04; B = −6.02), Muc2 (ratio 0.98, B = −6.24), Muc3 (ratio 1.02, B = −6.23), and Muc6 (ratio 1.08, B = −5.68). Expressed sequence tag (EST) sequences, found to be upregulated (n = 6) and downregulated (n = 21) in Nas1−/− mice, were excluded in tables 3 and 4, as their molecular functions are unknown and are therefore not discussed hereafter. qRT-PCR was used to verify changes in transcript levels for a selected subset of eight genes, representing different functional categories with increased (Bok, Car4, Avil and Pnliprp2) or decreased (Car1, Casp11 and Orm1) transcriptional levels. The expression levels of these eight genes determined by qRT-PCR were similar (R2 = 0.98) to those obtained with gene arrays (Supplementary Fig. 1). Notably, there were no differences in expression of major inflammatory cytokines or chemokines in the absence of NaS1, consistent with the lack of histological evidence of colitis and the normal epithelial proliferation.

Table 3 Genes upregulated in Nas1−/− ileum
Table 4 Genes downregulated in Nas1−/− ileum

DISCUSSION

The hyposulfataemic Nas1−/− mouse shows a markedly reduced sulfomucin content, which is clinically relevant since a decreased sulfomucin content has been observed in some gastrointestinal diseases, including colon cancer and inflammatory bowel disease.17 2124 Nas1−/− mice show mildly increased permeability to macromolecules, enhanced susceptibility to acute DSS-induced colitis and reduced intestinal barrier function to the C jejuni bacterial pathogen, as well as an altered ileal transcriptional profile. These abnormal features in the Nas1−/− mouse are likely to be a consequence of reduced sulfate availability in the intestine, secondary to the sulfate depletion due to increased urinary excretion that is characteristic of these mice.28 However, a non-epithelial contribution to the phenotype we describe cannot be completely excluded. The altered intestinal physiology in Nas1−/− mice is of high relevance to human gastrointestinal diseases associated with altered sulfate homeostasis2224 and suggest that intestinal sulfate deficiency is likely to enhance susceptibility to luminal toxins and pathogens.

Several sulfotransferase-deficient mice have been characterised, but none is deficient in the major sulfotransferases responsible for sulfonation of O-linked mucin oligosaccharides.4549 Previous studies have shown that sufficient levels of sulfate need to be maintained for sulfotransferase reactions to function effectively.25 These findings led us to investigate the sulfomucin content in hyposulfataemic (∼0.2 mM serum SO42−) Nas1−/− mice and compare them with normosulfataemic (∼1 mM serum SO42−) Nas1+/+ mice. Our data demonstrate that the Nas1−/− mouse provides a model of substantial diminution of mucin sulfonation, which is of particular relevance to gastrointestinal physiology because the sulfonated oligosaccharides of mucins are proposed to have evolved as a mechanism for protecting mucins from degradation.20 This is consistent with our histological data, showing reduced sulfomucin content in goblet cells of the Nas1−/− mouse jejunum, ileum and colon. This is an important finding because mucins are a major component of the mucosal glycocalyx which lubricates and protects the underlying epithelium, and plays an important role in preventing colitis and bacterial infection.14 15 50 51 It is not possible to distinguish between effects of NaS1 deficiency on the cell surface versus secreted mucins in this mouse model and this issue warrants further investigation. Nonetheless, our findings show that NaS1 is critical to maintaining the sulfate content of gastrointestinal mucins.

Reduced sulfomucin content in the small and large intestines of Nas1−/− mice led us to compare the susceptibility to colitis induced by DSS treatment between Nas1−/− and Nas1+/+ mice. DSS-induced colitis is considered a model for the intestinal pathology of human ulcerative colitis.35 52 Nas1+/+ mice were mildly affected by 5% DSS in drinking water for 7 days, whereas Nas1−/− mice showed significant intestinal pathology, which led to increased rectal bleeding and reduced red blood cell counts, when compared with Nas1+/+ mice. Whilst changes in mucin sulfonation have been well documented in ulcerative colitis,23 4244 it is not clear whether these changes are primary defects or secondary responses to colitis. Regardless of whether changes in mucin sulfonation in human disease are primary or secondary, our murine model shows that altered mucin sulfonation is likely to contribute to pathology. Whilst the precise mechanism underlying this contribution is at present not clear, our data demonstrate that greater epithelial damage accompanies reduced mucin sulfonation, consistent with a reduced barrier function of mucus containing undersulfonated mucins.

Mucosal epithelial cells (MECs) form a contiguous lining between the external milieu and the host’s internal environment, protecting the host from microbial attack. Consequently, MECs have evolved multiple defence mechanisms, including the production of mucins which play an important role in preventing bacterial infection and inflammation (reviewed in Linden et al50 and McGuckin et al51). Numerous interactions between microorganisms and mucins and/or mucin-type carbohydrates have been demonstrated, and pathogenic microbes have evolved a wide array of adhesins for binding to carbohydrates found in the glycocalyx of MECs (reviewed in Linden et al50). We propose that mucins represent polymeric complex “libraries” of glycocalycal carbohydrates that are likely to be encountered first by invading pathogens and therefore act as decoys for microbial carbohydrate adhesins. An important example of interaction with sulfonated mucin carbohydrates is the interaction of Helicobacter pylori with sulfonated oligosaccharides on salivary and gastric mucins.53 Furthermore, mucin barrier subversion strategies used by microbes include the production of degradative enzymes, such as glycosulfatases, sialidases, sialate O-acetylesterases, N-acetyl neuraminate lyases and mucinases, to destabilise the mucus gel and remove decoy carbohydrates for adhesins from the mucins.5458 There is evidence that these degradative enzymes are critical for bacterial pathogenesis, including the H pylori glycosulfatase, which has mucinolytic activity59 which is required for translocation through mucin-containing gels.60 Decreased mucin sulfonation has been proposed to lead to enhanced mucin degradation and penetration of the secreted mucus barrier by microbes, thereby giving increased access to the cell surface.20 This is consistent with our study showing enhanced systemic infection by C jejuni in Nas1−/− mice, which have reduced sulfomucin content. These findings suggest that individuals with impaired mucin sulfonation could be more susceptible to gastrointestinal pathogens and luminal toxins, which in turn could initiate chronic colitis and lead to an increased risk of carcinogenesis. Taken together, our findings indicate the importance of sulfate in maintaining intestinal barrier function.

Of the total 22 029 genes analysed, 41 genes with known functional roles in metabolism, cell signalling, immune response, cell structure, transcription or protein synthesis were differentially expressed in the ileum of Nas1−/− mice. Importantly, we found no change in mRNA levels for mucin genes that are normally expressed in the mouse ileum,10 suggesting that the abnormal intestinal features of Nas1−/− mice are likely to be due to reduced mucin sulfonation rather than altered mucin core protein expression, and that reduced mucin sulfonation does not induce changes in mucin mRNA expression. Similarly, no changes were found for mucin sulfotransferase61 mRNA levels in Nas1−/− mice, indicating that sulfotransferase gene expression is not altered when substrate levels are low in an attempt to overcome the reduced sulfomucin content in Nas1−/− mice. Of great interest is the markedly reduced (by ∼90%) mRNA level of carbonic anhydrase 1 (Car1) in Nas1−/− ileum. Car1 is the most abundant isoform of the 16 mammalian Car members (reviewed in Pastorekova et al62), which play important physiological roles in gastrointestinal processes, including acid and bicarbonate secretion.63 Interestingly, SO42− appears to have an inhibitory effect on carbonic anhydrase enzyme activity.64 These findings, together with reduced Car1 mRNA levels and impaired intestinal function in Nas1−/− mice, may be relevant to the reduced intestinal bicarbonate secretion in individuals with ulcerative colitis.65 Unlike the downregulation of Car1, we observed increased mRNA levels of Car4, suggesting that members of the Car family may be differentially regulated at the transcriptional level in Nas1−/− mice. This finding is consistent with the previously demonstrated differential mRNA expression of individual Car family members in the mouse gastrointestinal tract.66 The most upregulated genes in Nas1−/− ileum include pancreatic lipase-related protein 2 and colipase, which play important roles in dietary fat digestion.67 These findings may be relevant when considering the altered lipid metabolism in Nas1−/− mice and humans with inflammatory bowel disease.68 Taken together, these data show an altered transcriptional profile in Nas1−/− ileum, and the loss of NaS1 leads to changes in the expression of genes linked to bicarbonate and fatty acid homeostasis in the gastrointestinal tract.

This study shows for the first time the importance of sulfate in mucosal infection and toxin-induced colitis. The hyposulfataemic Nas1−/− mice characterised here provide a unique model of mucin sulfate deficiency, sharing physiological features of human gastrointestinal diseases with reduced sulfomucin content, including inflammatory bowel disease and colon cancer. The implications of this study are relevant to humans, suggesting that altered sulfate homeostasis could play a major role in human intestinal diseases, prompting future assessment of intestinal health to be included in human population studies with altered sulfate homeostasis.

Acknowledgments

We thank Drs T Walker and H Cooper for advice on qRT-PCR. SG is a recipient of an NHMRC career development award. MM is a recipient of a Senior Research Fellowship from The Cancer Council of Queensland. The array reagents were provided by the Australian Cancer Research Foundation DNA microarray initiative. This work was supported in part by the Australian Research Council and the National Health and Medical Research Council.

REFERENCES

View Abstract

Supplementary materials

Footnotes

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • Digest
    Robin Spiller Severine Vermeire