To study the capacity of butyrate to inhibit production of tumor necrosis factor α (TNFα) in macrophage-like synoviocytes (MLS) from patients with rheumatoid arthritis (RA), in human peripheral monocytes, and in murine RAW264.7 macrophages.
The concentrations of TNFα in culture supernatants of these cells were measured using enzyme-linked immunosorbent assay. The expression levels of various messenger RNAs (mRNA), such as those for TNFα, the mRNA-binding protein TIS11B, and luciferase, were measured using real-time quantitative polymerase chain reaction. The in vitro effects of butyrate on transcriptional regulation were evaluated by transfection with various reporter plasmids in RAW264.7 macrophages. The effects of TIS11B on TNFα expression were examined using an overexpression model of TIS11B in RAW264.7 cells.
Butyrate suppressed TNFα protein and mRNA production in MLS and monocytes, but paradoxically enhanced transactivation of the TNFα promoter. Expression of the AU-rich element (ARE)–binding protein TIS11B was up-regulated by butyrate. Induction of TNFα mRNA by lipopolysaccharide was significantly inhibited when TIS11B was overexpressed. Butyrate facilitated the degradation of luciferase transcripts containing the 3′-untranslated region (3′-UTR) of TNFα, and this effect was dependent on the ARE in the 3′-UTR that is known to be involved in the regulation of mRNA degradation.
These results indicate that butyrate suppresses TNFα expression by facilitating mRNA degradation mediated through a cis-acting ARE. Butyrate has the ability to regulate TNFα at the mRNA level and is therefore a potential therapeutic drug for RA patients.
Rheumatoid arthritis (RA) is a chronic inflammatory disease characterized by cartilage destruction and extracellular matrix degradation in multiple joints (1). The pathogenesis of RA is not clearly understood; however, tumor necrosis factor α (TNFα) is involved in its development, a conclusion which is supported by successful treatments with anti-TNFα reagents (2). The production of TNFα was found to be increased in rheumatoid synovium, followed by the induction of other proinflammatory cytokines, including interleukin-1β (IL-1β), IL-6, and IL-8, as well as matrix metalloproteinases involved in cartilage and bone destruction in RA (3–5). These cytokines are involved in synovial cell activation and proliferation, leading to generation of pannus (6).
Synovial tissue consists of heterogeneous immune and non–immune cell populations, including fibroblast-like synoviocytes, macrophage-like synoviocytes (MLS), lymphocytes, dendritic cells, and endothelial cells (7). Among these populations, MLS originating from bone marrow–derived monocytes are largely responsible, upon activation, for the production of TNFα protein (6). Monocytes can give rise to osteoclasts involved in rheumatoid bone destruction (8). The mechanisms of MLS activation have been only partially understood. Macrophages express TNFα via activation of Toll-like receptor 4–NF-κB signaling, and this type of activation may be mimicked experimentally by administration of lipopolysaccharide (LPS) (9, 10).
TNFα-mediated chronic inflammation has also been studied in inflammatory bowel diseases and in animal models of such diseases. In those studies, physiologic roles for short-chain fatty acids have been identified (11, 12). Short-chain fatty acids are a natural product of colonic anaerobic fermentation of dietary fiber by luminal microflora. These are the preferred sources of energy for the normal colonic epithelial cell, and they can modulate a variety of fundamental cellular processes to induce cell-cycle arrest, differentiation, and apoptosis in transformed cells (13, 14). These molecules, especially butyrate, have potent antiinflammatory effects and can modulate TNFα expression in colonic epithelial cells and in monocytes (15). It has also been shown that administration of butyrate suppresses experimental enteritis induced in mice by dextran sulfate sodium (16). Experiments in colonocytes revealed that butyrate down-regulates TNFα expression by modulating NF-κB–DNA binding activity (15), although the precise mechanism is not fully understood. In addition, butyrate is known to function as a histone deacetylase (HDA) inhibitor in cells, and it can induce alteration of the chromatin structure (17, 18), although the effect of this activity on TNFα down-regulation is not yet understood. We investigated whether butyrate could suppress TNFα production in activated synovial cells and macrophages, in order to evaluate short-chain fatty acids as a potential investigational new treatment for chronic inflammation in RA.
MATERIALS AND METHODS
Preparation of primary synoviocytes and culture.
Primary synoviocytes were obtained from surgically resected synovial tissue from Japanese patients with RA. Informed consent was obtained from each patient. Tissue specimens were minced and dissociated in Hanks' balanced salt solution (Invitrogen, Carlsbad, CA) containing 5 mg/ml type I collagenase (Sigma, St. Louis, MO) and 0.15 mg/ml DNase I (Sigma) for 2 hours at 37°C. Samples were then passed through a metal mesh and a nylon mesh, each with a 100-μm pore size. Cells were collected by centrifugation and resuspended at 0.5 × 106/ml in Iscove's modified Dulbecco's medium (Invitrogen) containing 10% heat-inactivated fetal bovine serum (FBS) and antibiotics. The resulting synoviocytes were cultured in a 6-well tissue culture plate (Becton Dickinson, Mountain View, CA) at 37°C in a humidified atmosphere with 5% CO2 for 24–48 hours. Our primarily rheumatoid synoviocyte cultures contained ∼10–35% of a CD14-positive subpopulation as assessed by fluorescence-activated cell sorting (FACS Calibur system; Becton Dickinson) (data not shown). The proportion of CD14-positive rheumatoid synoviocytes varied depending on the patient's background, such as duration or activity of the disease and treatment. Before the stimulation assay, nonadherent cells were removed by washing twice with phosphate buffered saline (PBS).
Cell culture and stimulation.
Human monocytes were enriched from whole blood obtained from healthy Japanese volunteers. The mononuclear cell fraction was prepared by density-gradient centrifugation over Ficoll-Paque (Amersham Biosciences, Uppsala, Sweden), followed by plating in culture medium at 37°C for 1 hour in a humidified incubator to obtain the adherent cells. After washing the cells twice with PBS, adherent cells were resuspended in RPMI 1640 medium (Invitrogen) containing 10% heat-inactivated FBS and antibiotics, then counted and diluted to 0.5 × 106/ml. Cells were processed for the stimulation assay on the same day that the blood was collected. Purity of CD14-positive cells was 80–90% as assessed by fluorescence-activated cell sorting (data not shown). The murine macrophage cell line RAW264.7 (no. TIB-71; American Type Culture Collection, Rockville, MD) was grown in Dulbecco's modified Eagle's medium (DMEM; Invitrogen) containing 10% heat-inactivated FBS and antibiotics at 37°C in a humidified atmosphere with 5% CO2 (19). Cells were resuspended at 0.25 × 106/ml, then 2 ml/well of the cell suspension was transferred to 6-well tissue culture plates, followed by 24 hours of culture before stimulation. After preincubation, cells were stimulated with LPS (no. L4391; Sigma) or were used without stimulation. In some experiments, various concentrations of sodium butyrate (Sigma) were included in the culture. Where indicated, actinomycin D (Sigma) was also included in the culture to restrict transcriptional events.
RNA preparation and real-time quantitative polymerase chain reaction (PCR) analysis.
Total RNA was obtained by using TRIzol RNA Reagent (Invitrogen) according to the manufacturer's instructions. In some experiments, total RNA was extracted from transfected cells using the Concert Cytoplasmic RNA Reagent (Invitrogen) in combination with DNase I Amplification Grade (Invitrogen) to eliminate residual plasmid DNA. RNA samples were reverse-transcribed using oligo(dT) primers and ReverTra Ace (Toyobo, Osaka, Japan) according to the manufacturer's instructions. Real-time quantitative PCR analyses were performed using 100 nM TaqMan probe and 200 nM forward and reverse primers in a final volume of 30 μl using 2×PCR reagent (Applied Biosystems, Chiba, Japan) in an ABI PRISM 7000 Sequence Detection System instrument (Applied Biosystems) based on dual-labeled fluorogenic probe technology (20).
The following forward and reverse primers and TaqMan probes, designed by Primer Express software (Applied Biosystems), were used for analyses: mouse TNFα, forward primer 5′-CAGACCCTCACACTCAGATCATCT-3′, reverse primer 5′-GCACCACTAGTTGGTTGTCTTTGA-3′, TaqMan probe 5′-CAAGCCTGTAGCCCACGTCGTAGCA-3′; mouse TIS11B, forward primer 5′-TTGTTGGTAGCTTCTGGCTTGA-3′, reverse primer 5′-GGCATCTACTGACAAAAGATGGAA-3′, TaqMan probe 5′-TCCATTTCATAGCCCACTTAACCACGCA-3′; luciferase, forward primer 5′-TGACCGCCTGAAGTCTCTGA-3′, reverse primer 5′-ACACCTGCGTCGAAGATGTTG-3′, TaqMan probe 5′-CCGCTGAATTGGAATCCATCTTGCTC-3′; AU-rich element (ARE) mutation, forward primer 5′-ATGCACAGCCTTCCTCACAG-3′, reverse primer 5′-CCCGGCCTTCCAAATAAATAC-3′, minor groove binder (MGB) TaqMan probe 5′-TATCCATTATCCATCCATTATCCATC-3′. The first 3 TaqMan probes were labeled on the 5′ end with FAM reporter dye and on the 3′ end with TAMRA quencher dye; the ARE mutation MGB TaqMan probe was labeled on the 5′ end with FAM reporter dye and on the 3′ end with conjugated MGB.
To measure the gene copy number of the various transcripts, the purified artificial gene product containing an amplification sequence was serially diluted to achieve a standard curve. Data for each messenger RNA (mRNA) quantity were normalized based on the mRNA copy number of GAPDH obtained using the TaqMan rodent GAPDH control reagents (Applied Biosystems).
Enzyme-linked immunosorbent assay (ELISA).
The concentrations of human and mouse TNFα proteins were determined using specific sandwich ELISA kits (no. 656227 from Cosmo Bio [Tokyo, Japan] and no. 10019 from Genzyme [Cambridge, MA], respectively) according to the manufacturers' instructions.
The luciferase reporter plasmids pNFκB-Luc and pGL3-BASIC were purchased from Stratagene (La Jolla, CA) and Promega (Madison, WI), respectively. Murine genomic DNA extracted from a normal C57BL/6 mouse using a standard protocol (21) was used for amplification of a genomic DNA fragment of the mouse TNFα gene. A 0.9-kb DNA fragment corresponding to the 5′-untranslated region (5′-UTR) of the mouse TNFα gene was PCR-amplified with the sense and antisense primers 5′-CTCAAGCTTATCAGAGTGAAAGGAGAAGGC-3′ and 5′-CTCAAGCTTAGTGAAAGGGACAGAACCTGC-3′, respectively. The product was inserted into the Hind III cloning site of pGL3-BASIC and designated pGL-mTNFα. The 0.8-kb 3′-UTR of the mouse TNFα gene was similarly amplified using PCR, and the Xba I and Bam HI restriction sites were introduced with the PCR primers 5′-TCTAGAGGGAATGGGTGTTCATCC-3′ and 5′-GGATCCCATGCCCCAGGGCAAA-3′, respectively. The resulting DNA fragment was used to replace the 3′-UTR of the pGL-mTNFα, and the resulting plasmid was designated pGL-mTNFα-UTR.
A pGL-CMV-UTR plasmid, in which the luciferase gene was constitutively expressed under the control of the cytomegalovirus (CMV) promoter, was generated by replacing the Hind III fragment corresponding to the TNFα promoter region in the pGL-mTNFα-UTR vector with a DNA fragment containing the CMV promoter sequences from pFLAG-CMV-2 (Sigma). In addition, the AT repeat (ARE) in pGL-CMV-UTR corresponding to the TNFα 3′-UTR sequence (+1299 to +1332) was mutated to yield an ARE mutation plasmid (pGL-CMV-UTR/mARE) using standard recombinant techniques and the mutagenic oligonucleotide 5′-ACAGCCTTCCTCACAGAGCCAGCCCCCCTCTATTTATATTTGCACTTATTATCCATTATCCATCCATTATCCATCCATTTGCTTATGAATGTATTTATTTGGAAGGCCG-3′. The sequences of all the DNA fragments obtained by PCR amplification were confirmed by DNA sequencing and completely matched the reported mouse TNFα genomic sequence (GenBank accession no. Y00467) (22).
A mammalian expression plasmid, pFLAG-TIS11B, that encodes mouse TIS11B complementary DNA (cDNA) (23) was generated by introducing a mouse full-length TIS11B cDNA fragment into the pFLAG-CMV-2 plasmid (Sigma). The TIS11B cDNA fragment was obtained by PCR amplification of cDNA from RAW264.7 cells and the specific sense and antisense primers 5′-GAATTCGATGACCACCACCCTCGT-3′ and 5′-TCTAGAGGAGAGGTGAAGGAGGCATG-3′, respectively. After subcloning into pCR-blunt and confirmation by DNA sequencing, the 1.2-kb fragment corresponding to the TIS11B cDNA (GenBank accession no. BC016621) was excised and ligated into pFLAG-CMV-2 at the Xba I and Eco RI cloning sites.
Transfection and luciferase assay.
RAW264.7 cells (0.25 × 105) were transferred to 8-cm tissue culture plates (Becton Dickinson) and incubated for 24 hours to 70% confluence. Transfections were achieved using Transfectam reagent (Promega), according to the manufacturer's instructions, with the plasmids described above. Plasmid DNA (10 μg) was mixed with 20 μl of Transfectam reagent in 3 ml of FBS-free DMEM and transferred to the plates. After 2 hours of incubation, complete culture medium was added to the cells, followed by further incubation to semiconfluence. The transfection efficiency in the current protocol was ∼5% when cells transfected with green fluorescent protein–expressing plasmid were analyzed by fluorescence-activated cell sorting (data not shown). Transfected cells were collected and distributed into new 6-well tissue culture plates (Becton Dickinson) before the stimulation assay. This step was performed to equalize the number of transfected cells and eliminate any differences in transfection efficiency between plates.
Where indicated, the luciferase assay was carried out according to the procedure described previously with minimal modifications (24). Briefly, RAW264.7 cells were transfected with various reporter plasmids as described above. After incubation and stimulation, cells were lysed using lysis buffer (Promega). Cell lysates were stored at −80°C until the luciferase assay was performed. The protein content of each sample was determined using a protein assay reagent (Bio-Rad, Hercules, CA), and the luciferase activities of the samples were normalized according to the protein content of the samples. Data are reported as relative luciferase units.
Cell proliferation assay.
In order to exclude the possibility that butyrate directly affects cell viability, a proliferation assay was performed using MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) tetrazolium assay (Cell Titer96 AQueous One Solution Cell Proliferation Assay; Promega). RAW264.7 cells (0.25 × 105) were transferred into microtiter-plate wells in 100 μl of DMEM containing various concentrations of butyrate, then MTS reagent was added to the wells. Optical density was read using a microplate autoreader (Microplate Reader Model 3550; Bio-Rad) at a wavelength of 490 nm after addition of MTS reagent.
Statistical analyses were calculated by Student's t-test with the use of the Excel program (Microsoft, Redmond, WA).
Suppression by butyrate of TNFα protein production in cultured primary synoviocytes.
We first analyzed the effect of butyrate on TNFα production in rheumatoid synoviocytes during stimulation with LPS. When primary synoviocytes were stimulated with LPS, significant amounts of TNFα protein were detected in the culture supernatants (Figure 1A). Addition of butyrate to the culture medium led to a significant, dose-dependent decrease in LPS-stimulated TNFα production. MLS, major producers of TNFα in the rheumatoid inflammatory synovium, play an important role in the initiation of synovitis and bone destruction (25). We paid attention to the subpopulation of monocyte/macrophages, and thus we repeated the experiments on the effect of butyrate on TNFα production both by peripheral monocytes and by the macrophage cell line RAW264.7. Human peripheral monocytes and RAW264.7 cells both secreted significant amounts of TNFα upon LPS stimulation (Figures 1B and C). As was the case in the primary cell culture, there was a significant decrease in LPS-stimulated TNFα secretion with increasing butyrate concentrations in the medium. In all experiments, similar GAPDH mRNA expression levels were confirmed in each cell preparation after treatment with LPS/butyrate (data not shown).
RAW264.7 cells were exposed to various concentrations of butyrate, and MTS tetrazolium assay was performed to assess cell proliferation. The results revealed that butyrate had no significant effect on cell proliferation (Figure 2B).
Suppression by butyrate of TNFα expression at the mRNA level in monocytes and in the macrophage cell line RAW264.7.
To determine how butyrate regulates TNFα production in monocytes, we next examined the effect of butyrate on TNFα mRNA expression. We performed semiquantitative reverse transcriptase (RT)–PCR analysis using specific primers for initial screening and found that butyrate suppressed TNFα mRNA induction in response to stimulation with LPS in human monocytes (data not shown). Butyrate also suppressed LPS-stimulated TNFα production in RAW264.7 cells in a dose-dependent manner, confirmed by real-time quantitative PCR (Figure 2A). The macrophage cell line RAW264.7 is responsive to butyrate; therefore, we used the RAW264.7 cells as an appropriate system for transfection assay.
No suppression by butyrate of transcriptional activity driven through the TNFα promoter.
To test whether butyrate could affect TNFα mRNA expression via transcriptional repression, RAW264.7 cells were transfected with reporter plasmids containing the consensus NF-κB binding sequence or the full-length TNFα promoter sequence. When cells were transfected with pNFκB-Luc or pGL-mTNFα, however, butyrate did not suppress the transcriptional activities (Figures 3A and B). Instead, butyrate showed a dose-dependent enhancement of transactivation, which was inconsistent with its effects on the mRNA and protein expression of TNFα. These results indicate that butyrate regulates TNFα mRNA levels via a mechanism other than transcriptional repression in RAW264.7 cells.
TIS11B as a candidate molecule for butyrate-mediated inhibition of TNFα mRNA expression.
TNFα gene expression involves posttranscriptional regulation, including the ARE in the 3′-UTR that is thought to play a critical role in the regulatory mechanism (26). AREs are found in multiple cytokine genes including TNFα, and the cis-regulatory effects of these elements are mediated by several polypeptides that can bind AREs. Tristetraprolin (TTP) is an ARE-binding protein that facilitates TNFα mRNA degradation, and TIS11B and TIS11D are representative members of the TTP family of proteins that are also capable of interacting with AREs. We hypothesized that the effects of butyrate were related to the posttranscriptional regulation mediated through the AREs and ARE-binding proteins. To search for the butyrate-mediated induction of ARE-binding proteins, we first used semiquantitative RT-PCR to screen for the expression of TTP family proteins such as TTP, TIS11B, and TIS11D in RAW264.7 cells treated with butyrate. Among the factors examined, butyrate induced only TIS11B mRNA but neither TTP mRNA nor TIS11D mRNA (data not shown). Real-time quantitative PCR revealed that butyrate induced TIS11B mRNA in a dose-dependent manner (Figure 4A).
We determined that TIS11B had a suppressive effect on levels of TNFα mRNA. RAW264.7 cells were transfected with TIS11B expression plasmid or with a control mock vector, then stimulated with LPS. The expression level of TNFα mRNA decreased significantly in RAW264.7 cells transfected with TIS11B expression plasmid (Figure 4B). These results suggested that butyrate down-regulated levels of TNFα mRNA via binding of TIS11B to an ARE.
Facilitation by butyrate of mRNA degradation via the function of an ARE in the TNFα 3′-UTR.
We next studied whether the inhibitory effects of butyrate on TNFα specifically depended on an ARE. To determine whether butyrate affects the amounts of gene transcripts carrying the TNFα 3′-UTR, we designed a real-time TaqMan RT-PCR–based quantitative assay that detects mRNA turnover (Figure 5). RAW264.7 cells were transfected with pGL-CMV-UTR reporter plasmid and treated with actinomycin D to repress generation of newly transcribed mRNA. To equalize the difference in transfection efficiency between plates, all transfected cells were collected and distributed into new plates before butyrate treatment. The time course of decrease of mRNA levels was in direct proportion to the rate of degradation. The amount of cDNA transcripts containing the 3′-UTR was serially quantified by real-time PCR. We observed a significant decrease in cDNA transcripts derived from pGL-CMV-UTR in cells treated with butyrate, both at the 10-minute time point and thereafter (Figure 5). This suggested that butyrate facilitated the degradation of mRNA transcripts of TNFα carrying the 3′-UTR sequence.
In various cytokine genes, including TNFα, an ARE in the 3′-UTR has been shown to affect stability of the gene transcripts (26). To test whether the inhibitory effects of butyrate on TNFα are mediated by the action of this cis ARE, we generated pGL-CMV-UTR/mARE containing a mutated ARE from pGL-CMV-UTR (Figure 6A). The amounts of transcripts derived from these plasmids were quantified and compared using RAW264.7 cells and the TaqMan PCR method. In this assay, total amounts of transcripts derived from mixed plasmids were determined by reactions using a luciferase TaqMan probe, and, at the same time, amounts of transcripts carrying the ARE mutation derived from these mixed plasmids were determined using an ARE mutation MGB TaqMan probe (Figure 6A). The quantification of these 2 transcripts was made possible by using a common gene standard, R-luc-AU, that contained both the mutated ARE and part of the common luciferase open reading frame sequence, and changes in the expression of these 2 transcripts were described as a ratio and calculated as follows: transcripts with mutated ARE:transcripts with intact ARE = (M/G):([A − M]/G) = M:(A − M), where A is the total amount of luciferase transcripts, M is the amount of transcripts with mutated ARE, and G is the internal control GAPDH.
As shown in Figure 6B, the amount of transcripts carrying mutated ARE and the amount carrying intact ARE were almost identical at the time point at which actinomycin D had stopped the transcription, when cells were treated (or not treated) with butyrate. It was shown that the relative amount of transcripts carrying a mutated ARE increased significantly during the time course. Thus, transcripts carrying the mutated ARE were more stable than those carrying intact ARE, and butyrate's ability to facilitate the degradation of transcripts was lost when the ARE was mutated. This indicates that the suppressive effect of butyrate on TNFα expression is mediated through the specific ARE in the 3′-UTR of the TNFα transcript.
In this report, we have shown that butyrate strongly down-regulated the production of TNFα in primary synoviocytes, peripheral monocytes, and murine RAW264.7 macrophages at the mRNA level. Our results also indicated that butyrate induced an ARE-binding protein, TIS11B, that could facilitate TNFα mRNA degradation. The collective data led us to believe that the down-regulation of TNFα expression by butyrate was mediated through the TIS11B protein that bound to a specific ARE and facilitated mRNA degradation.
TNFα is a key cytokine in the pathogenesis of RA. Therefore, inhibition of the action of TNFα may improve the clinical course of patients with RA. Biologic agents designed to interfere with the action of TNFα have been approved for the treatment of RA, and they provide dramatic efficacy for the affected patients. However, the disadvantages of biologic agents, such as their high cost, their instability, the difficulty of producing them, and the immune reaction of the host toward them, remain to be addressed. Therefore, new chemical agents that down-regulate TNFα production have been researched extensively. Our current findings may provide a framework for the therapeutic down-regulation of TNFα production by butyrate or its analogs.
Butyrate facilitated the degradation of luciferase transcripts when the TNFα 3′-UTR cDNA sequence was included in the reporter construct. This effect appeared to be dependent on an ARE, since a mutation in the ARE in the reporter plasmids resulted in the loss of the butyrate-induced changes in the transcript turnover. Thus, butyrate could down-regulate TNFα mRNA expression by facilitating TNFα mRNA degradation through a mechanism that was dependent on the specific ARE.
AREs are several repetitive sequences of AUUUA or UUAUUUAUU present in the 3′-UTRs of various mRNA. The AREs are clustered into several categories and have been identified in a panel of genes other than TNFα (27–30). These genes largely encode proteins that regulate cellular growth and the body's response to exogenous agents such as microbes and inflammatory and environmental stimuli. The mRNA containing AREs are commonly short lived, rapidly expressed in response to stimuli, and rapidly degraded once their critical role in gene regulation ceases.
To date, several ARE-binding proteins have been reported that can stabilize or destabilize the target mRNA by binding competitively to their AREs. TTP (also known as TIS11, Nup475, and G0S24) is one of these ARE-binding proteins, and it has been characterized as a critical factor that controls TNFα mRNA turnover (31). Interestingly, genetically manipulated mice that lack TTP showed an abundance of endogenous TNFα production, an RA-like arthritis, and autoimmunity, presumably mediated by an excess of TNFα (32–34). TTP-related proteins, designated TIS11B and TIS11D, are structurally similar to TTP in their zinc-finger motifs, and they share functions with TTP, although their unique functions have yet to be characterized (35). These TTP-related proteins are rapidly expressed in response to 12-O-tetradecanoylphorbol-13-acetate and other diverse stimuli in various types of eukaryotic cells (23, 36). The expression of these proteins differs depending on the circumstances, hence they may have their own roles in controlling mRNA turnover under different circumstances (37). Among these proteins, TTP is induced by TNFα, suggesting that TTP may function as a feedback regulator of TNFα gene expression (31, 38). Notably, a TIS11B homolog expressed rapidly in response to butyrate has also been discovered in human cells and characterized as butyrate response factor 1 (BRF1) (39).
Consistently, butyrate rapidly induced the expression of TIS11B in RAW264.7 cells (Figure 4A). In addition, the induction of TNFα mRNA by LPS stimulation was strongly inhibited when TIS11B was overexpressed in these cells (Figure 4B). Stoecklin et al identified BRF1 as a regulator of ARE-dependent mRNA decay, and also showed that BRF1 can bind directly to ARE and promote the degradation of ARE-containing mRNA (40). Thus, we postulated a model whereby butyrate induced the expression of TIS11B/BRF1, followed by the binding of this ARE-binding protein to the ARE, thus facilitating TNFα mRNA degradation. Our cDNA microarray (GeneNavigator cDNA Array System; Toyobo) revealed that butyrate suppressed expression of mRNA containing AREs in their 3′-UTRs (e.g., mRNA for IL-1β, IL-15, and granulocyte–macrophage colony-stimulating factor) (data not shown). This result further supported the relevance of our hypothesis.
TIS11B has a unique character that facilitates TNFα mRNA degradation; therefore, analyzing the regulation of this molecule would provide a novel approach to the control of TNFα production. TNFα expression is activated mainly by the transcription factor NF-κB and by the MAP kinase (MAPK) pathways (the ERK, JNK, and p38 MAPK pathways). Recent studies have shown that the p38 MAPK pathway in particular plays an important role in posttranscriptional regulation that leads to mRNA stabilization (41). The p38 MAPK pathway also strongly induces and activates TTP, which down-regulates TNFα (42–45). Those studies suggested that the p38 MAPK pathway may play a crucial role in regulating the expression of TNFα (involving a TTP-dependent mechanism); however, the precise mechanism is not completely understood, and less is known about the relationship between TIS11B and the p38 MAPK pathway. Since butyrate induced TIS11B expression and has been shown to affect the p38 MAPK pathway (46), it may be that butyrate influences TIS11B expression through the p38 MAPK pathway. Analysis of the relationship between TIS11B and the p38 MAPK pathway would be important for understanding the effect of butyrate.
The effects of butyrate on TNFα gene expression, other than those involving TIS11B- and ARE-dependent mechanisms, also need to be addressed. In previous studies, it was shown that butyrate can inhibit the binding of NF-κB to DNA (12, 15, 47). In contrast, in the reporter gene assays, butyrate enhanced the transcriptional activity driven by NF-κB sites and the TNFα promoter in a dose-dependent manner (Figures 3A and B). This phenomenon may be explained in part by the HDA-inhibitory effects of butyrate. Butyrate strongly inhibits HDA activity in cells, and it can cause hyperacetylation of nucleotides and thereby nonspecifically enhance transcriptional activity (18, 48, 49). As in the case of genomic DNA, transfected plasmid DNA has been shown to be assembled with histones to form a “minichromosome” that is sensitive to histone hyperacetylation (50). Thus, butyrate can function as a nonspecific transcriptional enhancer for transfected plasmid DNA; hence, cotransfection with an internal control was not informative in the current experiments.
Analysis of the regulation of TNFα by butyrate provides information for a novel approach to the treatment of patients with RA. Further investigation of the regulation of TIS11B expression, including study of its gene promoter, promises to pave the way for therapeutic approaches that address ARE-dependent cytokine gene regulation.
We thank Ms Akiko Hirano and Ms Miki Aoto for technical assistance.