SCFAs (short-chain fatty acids), fermentation products of bacteria, influence epithelial-specific gene expression. We hypothesize that SCFAs affect goblet-cell-specific mucin MUC2 expression and thereby alter epithelial protection. In the present study, our aim was to investigate the mechanisms that regulate butyrate-mediated effects on MUC2 synthesis. Human goblet cell-like LS174T cells were treated with SCFAs, after which MUC2 mRNA levels and stability, and MUC2 protein expression were analysed. SCFA-responsive regions and cis-elements within the MUC2 promoter were identified by transfection and gel-shift assays. The effects of butyrate on histone H3/H4 status at the MUC2 promoter were established by chromatin immunoprecipitation. Butyrate (at 1 mM), as well as propionate, induced an increase in MUC2 mRNA levels. MUC2 mRNA levels returned to basal levels after incubation with 5–15 mM butyrate. Interestingly, this decrease was not due to loss of RNA stability. In contrast, at concentrations of 5–15 mM propionate, MUC2 mRNA levels remained increased. Promoter-regulation studies revealed an active butyrate-responsive region at −947/−371 within the MUC2 promoter. In this region we identified an active AP1 (c-Fos/c-Jun) cis-element at −818/−808 that mediates butyrate-induced activation of the promoter. Finally, MUC2 regulation by butyrate at 10–15 mM was associated with increased acetylation of histone H3 and H4 and methylation of H3 at the MUC2 promoter. In conclusion, 1 mM butyrate and 1–15 mM propionate increase MUC2 expression. The effects of butyrate on MUC2 mRNA are mediated via AP-1 and acetylation/methylation of histones at the MUC2 promoter.
- histone acetylation
- histone methylation
- human milk feeding
- necrotizing enterocolitis
- short-chain fatty acid
SCFAs (short chain fatty acids) are produced by fermentation of undigested carbohydrates. SCFAs, and more specifically acetate, propionate and butyrate, are the major anions in the lumen of the large intestine. Several functions of SCFAs have been described, i.e. lowering intestinal pH, an energy-source for colonocytes, stimulation of colonic blood flow, smooth muscle contraction, transepithelial chloride secretion and exertion of proliferative stimuli of colonic epithelial cells . It is known that dietary fibres and SCFAs have beneficial effects in IBD (inflammatory bowel disease), e.g. by inhibition of pro-inflammatory cytokine-induced NF-κB (nuclear factor κB) activation and absorption of sodium and water [2–4]. In addition, SCFAs, and especially butyrate, are known to influence intestinal-specific gene expression, thereby influencing immune responses and oxidative and metabolic stress [5–9].
The composition of SCFAs in the intestine is determined by the composition of the microbiota, which in turn is influenced by the diet. For example, prebiotics selectively stimulate the growth and/or activity of bifidobacteria and thereby influence the SCFA composition .
Moreover, in human milk-fed infants, the large bowel is generally dominated by bifidobacteria and lactic acid bacteria. The flora of formula-fed infants on the other hand, is more diverse, less stable and often contains more Bacteroides, Clostridium and Enterobactericeae [11–14]. This difference in the composition of the microbiota results in a different SCFA composition between human milk-fed and formula-fed infants. It is well known that the ratio between the SCFAs butyrate, propionate and acetate differ in breast-fed infants compared with formula-fed infants (i.e. 2:6:90 in human milk-fed infants compared with 5:20:70 in formula-fed infants) . Based on the fact that more than 90% of the infants who develop NEC (necrotizing enterocolitis), which is the most common gastro-intestinal emergency in premature infants, have received formula feeding, as opposed to human milk solely, one could suggest that the production of SCFAs by bacteria and the composition of SCFAs in the intestine might play a role in the pathogenesis of NEC. Altered faecal concentrations of butyrate have also been reported in patients with UC (ulcerative colitis). In addition, a diminished capacity of the intestinal mucosa to oxidize butyrate has been reported in patients with active UC [16–18].
Both UC and NEC share the feature of an impaired intestinal barrier function. Mucins are required for the maintenance of an adequate mucus layer that covers the intestinal epithelium and thereby forms a physical barrier that protects the intestinal epithelium against toxic agents. The mucin MUC2 is the predominant mucin in the colon and MUC2 synthesis is diminished in UC [19,20] and presumably also NEC.
It has been shown in cell-line studies, experimental animal models and fresh human intestinal tissue specimens, that butyrate alters MUC2 expression in a dose-dependent manner [21–25]; however, the mechanisms that are responsible for these alterations have not so far been studied in detail.
In the present study, we investigated the role of increasing concentrations of butyrate, as well as acetate and propionate, in MUC2 expression in LS174T cells, a human goblet-cell-like cell line. Furthermore, the effects of butyrate on MUC2 expression were studied at the promoter, mRNA and protein levels respectively. We identified butyrate-responsive regions and cis-elements within the MUC2 promoter and determined the effects of butyrate treatment on histone H3 and H4 status at the MUC2 promoter.
The LS174T colonic cancer cell line was cultured in a 37 °C incubator with 10% CO2 in Dulbecco's modified Eagle's minimal essential medium supplemented with non-essential amino acids and 10% (v/v) foetal calf serum (Boehringer Mannheim) as described previously .
Cell-proliferation assay and morphological alterations
LS174T cells (2×105 cells) were pre-cultured in 24-well plates overnight to allow them to adhere. Subsequently cells were stimulated with physiological concentrations of butyrate (0, 1, 5 or 10 mM; sodium butyrate, Sigma–Aldrich) diluted in cell culture medium for 24 h. After removal of the culture medium, cells were treated with trypsin/EDTA solution and counted. All experiments were performed in triplicate in at least three separate experiments. In addition, cell proliferation and cell death were determined using the WST-I (WST-I proliferation agent; Roche Molecular Biochemicals) and Trypan Blue exclusion assays respectively. In addition, butyrate-induced morphological changes were studied microscopically.
qRT-PCR (quantitative real-time PCR)
LS174T cells were seeded in 6-well plates at 0.5×106 cells/well. Cells were incubated 16 h after seeding with either a low (1 mM), moderate (5 mM) or high (10 and 15 mM) concentration of butyrate, acetate or propionate (Sigma–Aldrich). After 24 h of stimulation, cells were lysed and harvested. Total RNA was prepared using the Nucleospin RNA II-kit from Macherey–Nagel. Total RNA (1.5 μg) was used to prepare cDNA. The mRNA expression levels of MUC2, as well as the housekeeping gene GAPDH (glyceraldehyde-3-phosphate dehydrogenase) were quantified using qRT-PCR analysis (TAQman chemistry) based upon the intercalation of SYBR Green on an ABI prism 7900 HT Fast Real-Time PCR system (PE Applied Biosystems) as described previously . The primer combinations for MUC2 (5′-CTCCGCATGAGTGTGAGT-3′ and 5′-TAGCAGCCACACTTGTCTG-3′) and GAPDH (5′-GTCGGAGTCAACGGATT-3′ and 5′-AAGCTTCCCGTTCTCAG-3′) were designed using OLIGO 6.22 software (Molecular Biology Insights) and purchased from Invitrogen. The effect of butyrate on MUC2 transcription and de novo protein synthesis was studied by co-incubating cells with butyrate and either actinomycin D (0.5 μg/ml) or cycloheximide (10 μg/ml) (Sigma–Aldrich) respectively. qRT-PCR was performed as described above.
RNA stability assay
LS174T cells were seeded at 1×106 cells/well in a 6-well cluster dish 24 h prior to the experiment. At t=0, the cells were treated with either 0 mM or 5 mM butyrate in combination with 4 μg/ml actinomycin D (Sigma–Aldrich). Cells were harvested after 0, 4, 6, 8, 10 and 24 h of butyrate/actinomycin D treatment. RNA isolation, cDNA synthesis and qRT-PCR were performed as mentioned above. To verify the amplification efficiency, as well as the amount of mRNA present in the treated cells, a serial dilution of cDNA derived from non-treated LS174T cells was amplified in duplicate on each plate.
pGL3-MUC2 promoter constructs covering the −371/+27, −947/−1, −2096/+27 and −2627/−1 regions of the MUC2 promoter have been previously described . The AP-1-Luc (where Luc is luciferase) reporter construct was a gift from Dr August Avery (Center for Molecular Immunology & Infectious Disease, Department of Vetinary & Biomedical Sciences, Pennsylvania State University, PA, U.S.A.). LS174T cells were seeded at 2.0×105 cells/well in 24-well plates. Transfections and co-transfections were performed the next day by adding 0.25 μg of the pGL3 construct of interest and 0.15 μg of phRG-B as an internal control. Transfection and co-transfection experiments were performed using Effectene® reagent (Qiagen) as described previously . Cells were incubated with the transfection mixture for 24 h at 37 °C. Stimulation with various doses of butyrate was performed over 24 h. Total cell extracts were prepared using 1×passive lysis buffer (Promega), as described in the manufacturer's instructions. Cell extract (10 μl) was used to determine luciferase activity in a Glomax luminescence counter (Promega) using the dual-luciferase assay system (Promega). The luciferase activity is expressed as the fold-induction of the non-stimulated sample compared with that of the SCFA-stimulated samples, after correction for transfection efficiency as measured by the Renilla luciferase activity. All experiments were performed in triplicate in at least three separate experiments.
The consensus AP-1 site (ATGAGTCAGA) found in the MUC2 promoter at −817/−808 was mutated using the QuikChange® site-directed mutagenesis kit (Stratagene) according to the manufacturer's instructions. The sequence of the oligonucleotides used to mutate the AP-1 site are shown in Table 1. The mutation was confirmed by DNA sequencing.
EMSA (electrophoretic mobility-shift assay)
Nuclear extracts from LS174T cells, untreated or treated with butyrate, were prepared as described previously , quantified using the bicinchoninic acid assay (Pierce assay, Perbio Science) and stored at −80 °C. Oligonucleotides were synthesized by MWG-Biotech and the sequences are shown in Table 1. Annealed oligonucleotides were radiolabelled using T4 polynucleotide kinase (Promega) and [γ-32P]dATP (GE Healthcare) and purified by chromatography on a Bio-Gel P-6 column (Bio-Rad). Nuclear protein incubation with radiolabelled probes and competitions with unlabelled probes were performed as described in Perrais et al. . For supershift analyses, 2 μl of the antibody of interest [anti-c-Fos (K-25, SC-253X) and anti-c-Jun (SC-44X); Santa-Cruz Laboratories and Tebu-Bio respectively], were added to the proteins and left for 1 h at room temperature (22 °C) before adding the radiolabelled probe. Electrophoresis conditions and gel processing were as described previously .
Western blot analysis
Nuclear proteins (10 μg) were separated by SDS/PAGE (10% gel), followed by electrotransfer on to a 0.45 μm PVDF membrane (Millipore). The membranes were incubated either with specific antibodies against c-Fos (sc-253, 1:10000) or c-Jun (sc-44, 1:5000) (Santa Cruz Laboratories) for 1 h at room temperature, or with specific antibodies against histone H3 (anti-acetylated lysine, mono-/di-/tri-methylated lysine 4) and histone H4 (anti-acetylated lysine) [Upstate Biotechnology #06-599 (1:10000 dilution), #05-791 (1:10000 dilution) and #06-598 (1:1000 dilution) respectively] for 2 h at room temperature. Secondary antibodies used were HRP (horseradish peroxidase)-conjugated anti-rabbit IgGs (Pierce). For detection, blots were processed with West® Pico chemiluminescent substrate (Pierce) and the signal was detected by exposing the processed blots to HyperfilmTM ECL® (enhanced chemiluminescence; Amersham Biosciences). For Sp1 detection, the membranes were incubated for 1 h at room temperature with anti-Sp1 antibody (sc-59, 1:10000, Santa Cruz Laboratories) and alkaline phosphatase-conjugated anti-goat IgG (Promega) was used as a secondary antibody. For detection, the membrane was incubated with Nitro Blue Tetrazolium chloride and 5-bromo-4-chloro-3-indolyl phosphate substrate (Life Technologies) .
LS174T cells were grown on poly-L-lysine-coated microscope glass slides 24 h prior to butyrate treatment. Cells were treated with 0, 1, 2, 5 and 10 mM butyrate for 24 h. Cells were fixed in ice-cold methanol at −20 °C for 10 min and rinsed in PBS. The MUC2 mucin expression was determined by immunocytochemistry. For this purpose, cells were incubated for 60 min at room temperature with the monoclonal MUC2 antibody (WE9)  diluted in PBS (1:200), rinsed four times with PBS, and incubated at room temperature for 60 min with the biotinylated horse-anti-mouse antibody (Vector) diluted in PBS (1:1000), followed by 1 h incubation with ABC-PO complex (Vectastain Elite Kit; Vector Laboratories), each component diluted 1:400 in PBS. After incubation, binding was visualized using 0.5 mg/ml DAB (diaminobenzidine), 0.02% H2O2 in 30 mM imidazole and 1 mM EDTA (pH 7.0). The slides were counterstained with Haematoxylin.
ChIP (chromatin immunoprecipitation)
Cells untreated or treated with butyrate (10×106 cells) were fixed in 1% (v/v) formaldehyde and chromatin was sonicated and immunoprecipitated as described in Piessen et al.  with either 5 μg of specific antibodies against histone H3 (anti-acetylated lysine, mono-/di-/tri-methylated Lys4, methylated Lys9 and trimethylated Lys27) and histone H4 (anti-acetylated lysine; Upstate Biotechnologies) or with normal rabbit IgGs (Upstate Biotechnologies) at 4 °C. Immunoprecipitated chromatin (50 ng) was then used as a template for PCR using the following pairs of primers: forward primer1, 5′-TTGGCATTCAGGCTACAGGG-3′ and reverse primer1, 5′-GGCTGGCAGGGGCGGTG-3′, covering the −236/+24 region of MUC2 promoter. PCR was performed using AmpliTaq Gold polymerase as described by Piessen et al. . PCR products (15 μl) were separated on a 2% (w/v) agarose gel containing ethidium bromide run in 1×TBE [Tris/borate/EDTA (1×TBE=45 mM Tris/borate and 1 mM EDTA)] buffer.
All values are means±S.D.
Butyrate affects cell morphology and proliferation
SCFAs are known to affect epithelial proliferation, differentiation, apoptosis and gene expression. Butyrate is, compared with acetate and propionate, the most effective SCFA in inducing alterations in these processes. Therefore we first analysed the effects of various concentrations of butyrate on morphology, proliferation and apoptosis of LS174T cells. Butyrate induced marked changes in morphology, which are characterized by elongation/stretching of the cells (Figure 1A). Furthermore, butyrate treatment inhibited the proliferation of the LS174T cells in a dose-dependent manner, as reflected by a decrease in cell number upon butyrate treatment (Figure 1B). These results were confirmed by WST-1 cell-proliferation assays (results not shown). Finally, none of the butyrate concentrations (1–10 mM) used in the present study induced cell death of LS174T cells, as determined by Trypan Blue exclusion assays and analysis of cell morphology.
SCFAs alter MUC2 mRNA expression
LS174T cells were stimulated with increasing concentrations, from 1 mM to 15 mM, of butyrate, propionate or acetate to determine the effects of the different SCFAs on MUC2 mRNA expression (Figure 2A). Butyrate at a concentration of 1 mM induced a 2.5-fold increase in MUC2 mRNA levels compared with untreated cells. In contrast, stimulation with higher concentrations, i.e. 5–15 mM butyrate, did not induce an increase in MUC2 mRNA levels, as at these concentrations MUC2 mRNA levels were comparable with control levels. Similar to butyrate, 1 mM propionate induced a 2.5-fold increase in MUC2 mRNA levels. Furthermore, at 5 mM propionate, MUC2 mRNA levels increased 4.2-fold, whereas at higher concentrations, MUC2 mRNA levels decreased again. Acetate treatment resulted in a dose-dependent increase in MUC2 mRNA levels at 5 mM, reaching a 2.2-fold induction at 15 mM. To determine whether the activating effect of SCFAs on MUC2 expression occurred at the transcriptional level, cells were pretreated with actinomycin D, which inhibits transcription. The results indicate that activation of MUC2 expression by 1 mM butyrate occurred at the transcriptional level, as MUC2 mRNA levels returned to basal levels when cells were pretreated with actinomycin D (Figure 2B). This process also requires de novo protein synthesis as pretreatment of LS174T cells with cycloheximide, an inhibitor of mRNA translation, decreased MUC2 mRNA levels to basal levels as well (Figure 2B). Similar results were obtained when cells were stimulated with propionate or acetate instead of butyrate (results not shown). Since butyrate increased MUC2 mRNA levels at low concentrations (1 mM), in contrast with no effect at moderate (5 mM) and high (10 and 15 mM) concentrations, we studied whether this decrease was due to a decrease in MUC2 RNA stability. For this we pre-treated cells with actinomycin D over 24 h, and then incubated cells with 5 mM butyrate before measuring the amounts of MUC2 by qRT-PCR. The results show no differences in MUC2 mRNA stability between butyrate-stimulated and non-stimulated cells (Figure 2C).
Effect of butyrate on MUC2 protein expression
To determine whether butyrate also induced an increase in MUC2 protein expression in LS174T cells, immunocytochemistry was performed with an antibody specific for MUC2. In non-stimulated cells, MUC2 staining was hardly visible (Figure 3). Stimulation with 1 mM butyrate clearly showed an increase in MUC2 staining. This effect was even more pronounced in cells stimulated with 2 mM butyrate (Figure 3).
Identification of butyrate-responsive regions in the MUC2 promoter
Transfections with MUC2 promoter constructs were performed to identify butyrate-responsive regions. The MUC2 promoter constructs used are indicated in Figure 4(A). Stimulation of LS174T cells with low (0.5–2 mM) concentrations of butyrate demonstrated a dose-dependent increase in luciferase activity after transfection with each of the promoter constructs used (Figure 4B). The highest transactivation was seen using the MUC2 promoter construct −947/−1, indicating a possible butyrate-responsive element within the −947/−372 region. Analysis of the MUC2 promoter sequence indeed revealed the presence of a consensus putative binding site (ATGAGTCAGA) for the transcription factor AP-1 at −817/−808, a transcription factor known to mediate butyrate-induced transcriptional effects. To determine whether this putative AP-1-binding site was responsible for the butyrate-induced MUC2 promoter transactivation, specific nucleotides within the sequence were mutated (Table 1). The mutation resulted in a 50% reduction of the butyrate-induced MUC2 transactivation (Figure 5A). Activation of AP-1 by butyrate in LS174T cells was confirmed by treating AP-1-Luc-transfected cells with butyrate. The stimulation was dose-dependent with a maximal 13.3-fold induction at 2 mM butyrate (Figure 5B).
c-Fos and c-Jun bind to the AP-1 element in the MUC2 promoter
As the transcription factors c-Fos and c-Jun are known to bind as a complex to AP-1-binding elements within promoters, EMSAs were carried out to show the binding of these two transcription factors to the AP-1 element found at −817/−808. When incubated with nuclear extract from untreated and butyrate-stimulated (1 mM and 10 mM) LS174T cells, the radiolabelled probe T282 (containing the putative AP-1 binding site; see Table 1) produced one retarded band (Figure 6, lane 2). Specificity of the protein–DNA complex was confirmed by a strong decrease of the shifted band when unlabelled competition was performed with a 50-fold excess of unlabelled T282 probe (lane 3), whereas competition with a 50-fold excess of unlabelled mutated T282 probe (lane 4) did not affect the shifted band. Involvement of c-Jun and c-Fos in the complex formation was then proven in supershift experiments carried out with antibodies specific for c-Jun (lane 5) and c-Fos (lane 6) respectively. Addition of the two antibodies indeed resulted in a supershift that was observed both in untreated cells and butyrate-stimulated cells. This was well-correlated with the amount of c-Fos and c-Jun found in the cells (Figure 7). Taken together, this suggests that the decreased MUC2 mRNA levels after stimulation of cells with 5 mM or 10 mM butyrate compared with 1 mM butyrate stimulation (see Figure 2), are not caused by a decreased binding capacity of AP-1 (i.e. the c-Fos–c-Jun complex) to its cis-element within the MUC2 promoter.
Butyrate alters histone status at the MUC2 promoter in a dose-dependent manner
Since butyrate is known to affect HDAC (histone deacetylase) activity, and the MUC2 promoter is known to be regulated by HDAC , we hypothesized that histone status at the MUC2 promoter may be involved in MUC2 regulation by butyrate. To determine whether alterations in MUC2 expression correlated with changes in histone acetylation and/or methylation we first examined the effect of butyrate on the levels of acetylated H3 and H4 histones as well as mono-/di-/tri-methylated H3-Lys4 in LS174T cells by Western blotting (Figure 7). Acetylated histone H3 and H4 and mono-/di-/tri-methylated H3 on Lys4, which correlate with activation of transcription, were strongly increased after stimulation with both moderate (5 mM) and high (10 mM) concentrations of butyrate.
To establish the effects of butyrate on histone H3 and H4 status at the MUC2 promoter, ChIP assays were performed with chromatin from non-stimulated and butyrate-stimulated LS174T cells (Figure 8). In untreated and 1 mM butyrate-treated cells, the MUC2 promoter covering the −236/+24 region was mainly associated with mono-/di-/tri-methylated H3-Lys4, as well as to a lower extent with acetylated histone H3 and H4, which correlate with activation of transcription. At 5 mM butyrate this status of chromatin activation was confirmed with a stronger association with acetylated H3. At these two concentrations we also observed an increase in trimethylated H3-Lys27, which is usually indicative of transcription inhibition. At 10 mM butyrate, histone modifications at the MUC2 promoter were characterized by modifications inducing active chromatin (acetylated H3, acetylated H4 and mono-/di-/tri-methylated H3-Lys4) and usually indicative of inactive chromatin (methylated H3-Lys9).
In the present study we analysed the effect of SCFAs on epithelial cell morphology, proliferation and, as a marker for epithelial protection, MUC2 expression. Moreover, we identified the mechanisms responsible for the butyrate-induced changes in MUC2 expression. By studying these parameters in conjunction we aimed to gain more insight into the effects of SCFAs on epithelial protection.
The present study revealed that butyrate altered the morphology of LS174T cells by inducing cell elongation/stretching. This suggests that butyrate affects LS174T cell differentiation. Additionally, butyrate caused a dose-dependent decrease in cell number. As this SCFA did not induce apoptosis at the concentrations used in the present study, we conclude that butyrate inhibits epithelial proliferation. Several in vitro studies support our finding that butyrate inhibits proliferation [33–37]. For example, butyrate inhibited the proliferation of non-confluent and subconfluent HT-29 cells in a dose-dependent manner (1–8 mM) . Furthermore, Siavoshian et al.  demonstrated that the mechanism through which butyrate inhibits proliferation in HT-29 cells is exerted via the induction of cyclin D3, an inhibitor of cell-cycle progression, and p21, a stimulator of cell differentiation.
Next, we studied the effect of SCFAs on MUC2 synthesis in the LS174T cell line. Butyrate and propionate, as well as acetate, were able to increased MUC2 mRNA synthesis. Specifically, butyrate increased MUC2 mRNA levels at low concentrations and had no effect at moderate and high concentrations. Both low and moderate propionate concentrations increased MUC2 mRNA levels, whereas at higher concentrations, MUC2 mRNA levels were still increased, but to a lesser extent. Finally, a dose-dependent increase in MUC2 mRNA levels was seen after stimulation with acetate, with the smallest increase at low concentrations and the highest increase at high concentrations.
Of the SCFAs used in the present study, only the effects of butyrate on mucin expression have been described extensively. Hatayama et al.  also showed that concentrations of 1–2 mM butyrate increased MUC2 expression in LS174T cells. Barcelo et al.  demonstrated a significant discharge of mucins at concentrations of 5 mM butyrate, whereas increasing the concentration to 100 mM decreased this mucus response in mice. Importantly, ex vivo stimulation of macroscopically normal fresh colon tissue with 0.05–1 mM butyrate stimulates MUC2 synthesis, whereas on stimulation with 10 mM butyrate MUC2 synthesis levels returned to basal levels . These studies correlate with the results of the present study with respect to the effects of butyrate on MUC2 expression (i.e. an increase in MUC2 at low concentrations and no effect at high concentrations).
Despite previous studies showing induction of MUC2 by butyrate, no precise analysis of the molecular mechanisms has been performed [21–25]. Since butyrate is known to mediate its effects via the AP-1 transcription factor, and because we found a putative consensus binding site (ATGAGTCAGA) for AP-1 at −817/−808 in the MUC2 promoter, we studied its regulation by AP-1. AP-1 is a multiprotein complex, composed of the products of c-Jun and c-Fos proto-oncogenes. Growth factors, neurotransmitters, polypeptide hormones, bacterial and viral infections, as well as a variety of physical and chemical stresses, employ AP-1 to translate external stimuli, both into short-term and long-term changes of gene expression. Interestingly, we found that butyrate was able to activate an AP-1 reporter construct and to induce c-Fos and c-Jun protein expression in the LS174T cell line, indicating that butyrate-induced MUC2 transcription might occur via AP-1 binding to the MUC2 promoter. That is what we indeed demonstrated by mutating the consensus AP-1-binding site, which abolished both binding of AP-1 and inhibited butyrate-induced MUC2 activation. As butyrate only increased MUC2 mRNA and protein levels at low concentrations (1–2 mM), but not at high concentrations (5–10 mM), this suggested that activation of the MUC2 promoter, and up-regulation of MUC2 RNA and protein levels, at low concentrations of butyrate was, at least partly, regulated by AP-1.
Since butyrate is a well-known HDAC inhibitor, butyrate-induced alterations in gene expression can also reflect changes in histone modification status and chromatin marks. To assess whether the increase of MUC2 expression following butyrate treatment was associated with chromatin status, we performed ChIP experiments. As expected, our results indicate that butyrate treatment is associated with a dose-dependent increase in both global histone acetylation levels and histone H3 and H4 acetylation at the MUC2 promoter region. Since cross-talk between histone post-translational modifications are important in establishing the histone code, increased mono-/di-/tri-methylation of H3-Lys4 observed at the MUC2 promoter after butyrate treatment may be directly linked to increased histone H3 acetylation, as previously shown by Nightingale et al. . For a long time methylation of H3-Lys9 has been considered as a chromatin mark associated with heterochromatin and gene silencing. However, a recent study has shown that higher H3-Lys9 monomethylation levels are detected in active promoters surrounding gene transcription start sites, suggesting that this modification may be associated with transcription activation . These results are concordant with the present study, showing that in LS174T cells, the proximal region of the MUC2 promoter is associated with monomethylation of H3-Lys9. However, to our knowledge, this is the first time that a positive effect of high concentrations of butyrate on H3-Lys9 methylation has been shown. Strikingly, we found that treatment with a low concentration of butyrate induced an increase in H3-Lys27 trimethylation at the MUC2 promoter, which therefore adopted a bivalent chromatin pattern. This bivalent profile has already been described for embryonic stem cell genes as well as DNA-hypermethylated genes which were re-expressed by demethylation . We have previously shown that MUC2 is regulated by a complex combination of DNA (de)methylation and establishment of a (de)repressive histone code . The changes of global epigenetic profile stated at the MUC2 promoter may thus be partly responsible for the increase in the MUC2 expression level induced by a low concentration of butyrate.
Surprisingly, treatment with moderate and high concentrations of butyrate, although associated with active chromatin marks at the MUC2 promoter, did not induce increased MUC2 expression. Dual effects of HDAC inhibitors on gene expression have already been shown for numerous genes , including mucin genes. In particular, Augenlicht et al.  demonstrated that cell treatment with butyrate induced an inhibition of MUC2 expression, correlated with repression of secretory functions of colonic cells. This repression may be due to changes in histone modification patterns, since trichostatin A, another HDAC inhibitor, has the same inhibiting effect on MUC2 expression in LS174T cells . However, our results clearly show that the MUC2 promoter is associated with active chromatin marks at high concentrations of butyrate. Therefore the dual effect observed at high concentrations, is most likely due to dramatic butyrate-induced changes in the global chromatin landscape , rather than direct histone modifications at the MUC2 promoter. Numerous studies showing that expression and post-translational modifications of factors known to positively or negatively regulate MUC2 transcription (including, among others, Sp3, CDX-2 or p53 transcription factors [44–46]) is dramatically affected by butyrate, support this hypothesis.
Taken together, we have shown that butyrate stimulates MUC2 expression at low concentrations, but has no effect on MUC2 expression at moderate and high concentrations. We therefore hypothesize that low concentrations of butyrate could have a protective effect on intestinal barrier function by increasing mucus production, whereas moderate to high concentrations may decrease intestinal barrier function by decreasing MUC2 production. This effect might partially explain the difference in incidence of NEC between the formula-fed and human milk-fed newborn infants. Manipulation of the SCFA profile can be established by influencing the composition of microbiota, for instance by treatment with prebiotics, probiotics and/or human milk. This approach seems to be promising in the treatment of IBD and NEC.
In summary, the in vitro results in the present study indicate that low concentrations of butyrate stimulate MUC2 mucin expression, which in vivo would lead to an increased intestinal epithelial barrier function. In contrast, high concentrations of butyrate decrease MUC2 expression which might diminish intestinal barrier function. Moreover, identification of AP-1 and histone modifications as mechanisms involved in MUC2 regulation by butyrate may represent pathways to target prevention of IBD and NEC by influencing SCFA production by the intestinal flora.
This work was supported by unrestricted grants from The Nutricia Research Foundation and The Association François Aupetit (to I. V. S.). A. V. is the recipient of an Inserm-région Nord-Pas de Calais PhD fellowship.
We thank Dr Avery (Pennsylvania State University, University Park, PA, U.S.A.) for the gift of the AP-1-Luc construct.
Abbreviations: ChIP, chromatin immunoprecipitation; EMSA, electrophoretic mobility-shift assay; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HDAC, histone deacetylase; IBD, inflammatory bowel disease; NEC, necrotizing enterocolitis; qRT-PCR, quantitative real-time PCR; SCFA, short-chain fatty acid; UC, ulcerative colitis
- © The Authors Journal compilation © 2009 Biochemical Society