The single-minded 2 (SIM2) protein is a basic helix-loop-helix transcription factor regulating central nervous system (CNS) development in Drosophila. In humans, SIM2 is located within the Down syndrome critical region on chromosome 21 and may be involved in the development of mental retardation phenotype in Down syndrome. In this study, knockout of SIM2 expression in mice resulted in a gas distention phenotype in the gastrointestinal tract. We found that SIM2 is required for the expression of all cryptdins and numerous other antimicrobial peptides (AMPs) expressed in the small intestine. The mechanism underlying how SIM2 controls AMP expression involves both direct and indirect regulations. For the cryptdin genes, SIM2 regulates their expression by modulating transcription factor 7-like 2, a crucial regulator in the Wnt/β-catenin signaling pathway, while for other AMP genes, such as RegIIIγ, SIM2 directly activates their promoter activity. Our results establish that SIM2 is a crucial regulator in controlling expression of intestinal AMPs to maintain intestinal innate immunity against microbes.
- single-minded 2 protein
- antimicrobial peptide
- T cell factor 7-like 2
- intestinal innate immunity
the single-minded (SIM) transcription factor was first identified from Drosophila (11). The sim gene encodes a nuclear protein with DNA-binding and protein dimerization [basic helix-loop-helix (bHLH)] domains. In addition, SIM contains two Per-ARNT-Sim domain (PAS) dimerization domains juxtaposed to the bHLH region (11, 37). Accordingly, SIM belongs to the bHLH-PAS family whose members include the hypoxia sensor hypoxia-inducible factor-1 and the circadian rhythm regulator Period gene.
It has been demonstrated that sim is a master regulatory gene for midline development in Drosophila (37). The specialized midline cells of both vertebrates and invertebrates play crucial roles in central nervous system (CNS) development. In Drosophila, the midline cells comprise both neurons and glia and form a cellular structure essential for the proper elaboration of the axon scaffold during CNS formation (25). In mammals, two SIM homologs, SIM1 and SIM2, have been identified and are encoded by two different genes (8, 12, 17). SIM1 is involved in the regulation of brain development and energy homeostasis (22, 23, 33), whereas the Sim2 gene is located within the Down syndrome (DS) critical region of human chromosome 21, corresponding to chromosome 16 in mice, which in triplication is associated with some of the diverse phenotypic characteristics of DS (8, 12).
Unlike other members of the bHLH/PAS family, SIM2 reportedly acts as a transcriptional repressor (34). In mice, overexpression of SIM2 elicits many of the mental characteristics seen in DS patients (9). Although SIM2 may contribute to many of the physiological abnormalities associated with DS, its role in growth and development outside the CNS is largely unknown. It is known, however, that Sim2 may play an important role in body development, since Sim2 null mice die shortly after birth due to multiple abnormalities, including cleft palate, improper diaphragm development, and rib defects, suggesting that SIM2 is crucial in normal development of not only CNS but also other tissues (20, 43). Recently, it was found that human SIM2 might be involved in the growth of colon, prostate, and pancreatic carcinomas. A splice variant of SIM2, SIM2s, is selectively expressed in these tumors, and knock down of human SIM2 results in inhibition of growth and progression of breast cancer cells, leading to the speculation that SIM2 may be a useful marker and possible target for tumor monitoring and therapy (14, 15). Controversially, SIM2 is downregulated in human breast cancer samples, and loss of SIM2 can enhance tumorigenesis in vivo (28, 29), which suggests that SIM2 may serve as a tumor suppressor in breast cells.
To study the role of SIM2 in tissues, we have generated a SIM2 knockout mouse strain in which the first exon of the Sim2 gene can be deleted to abolish the DNA-binding bHLH domain and gene expression. We found that the Sim2 knockout homozygous mice survive at birth but developed severely bloated gastrointestinal tracts shortly after birth regardless of nursing. Some 1-day-old homozygotes with severe gas distention in their gastrointestinal tracts died within 24 h, whereas other 1-day-old homozygotes with less severe gas distention survived and continued to thrive. In the present study, we investigated the mechanism by which SIM2 deficiency causes gas distention in the gastrointestinal tract. We found that SIM2 is the master regulator in controlling expression of numerous crucial antimicrobial peptides (AMPs), such as lysozyme and all of the cryptdins, in the intestine. Lack of SIM2 resulted in the decrease of intestinal AMP levels, which caused overgrowth of microbes in the gastrointestinal tract and led to the development of the gas distention phenotype.
MATERIALS AND METHODS
All studies were conducted in accordance with the Institutional Animal Care and Use Committee guidelines of Academia Sinica for the use of laboratory animals. The Sim2 mutant mice, Sim2−/− and Sim2Lp/Lp, that are homozygous for the Sim2 gene null and the floxed alleles, respectively, were generated using gene targeting and Cre-loxP DNA recombination technology detailed in Fig. 1. The resultant mutant mice were backcrossed to C57BL/6. Afterward, sister-brother crossing was employed to maintain the Sim2 null allele in the stock and to obtain Sim2−/− mice for experiments described here. The adenomatous polyposis coli (Apc)min/+, Vil-cre, and R26R-LacZ mice were from Jackson Laboratory. The Sim2−/− mice carrying an Apcmin allele were generated by breeding male Sim2−/− mice with female Apcmin/+ mice. Likewise, the Sim2Lp/− mice carrying the Vil-cre transgene were produced by breeding male Sim2−/− with female Vil-cre mice to generate Sim2+/−, Vil-cre mice that were then bred with Sim2Lp/Lp mice to obtain mice with the desired genotypes. The double-transgenic mice, Vil-cre/R26R-LacZ, were generated by breeding R26R-LacZ mice with Vil-cre mice. For all the experiments described here, only littermates (from female Sim2+/− parents) carrying different Sim2 mutant alleles were used. All mice were maintained in an specific pathogen-free animal facility with 12:12-h light-dark cycles.
Quantification of intestinal gas volume.
The whole gastrointestinal tracts of Sim2−/− mice were removed and placed in a 1- or 5-ml graduated syringe, depending on the size of gastrointestinal tracts. The desired volume of PBS was then drawn up through the needle. The gastrointestinal tract along with the PBS contents were then transferred from the syringe to a petri dish, and the gastrointestinal tract was cut open longitudinally to release gas. Last, the contents of the petri dish were loaded back to the syringe to record the difference in volume.
Intestinal peristalsis test.
The Sim2−/− mice (12-wk-old males) were fasted overnight before 0.2 ml of 25% BeSO4 suspended in a 6% glucose solution was orally administered using a feeding tube. After 60 min, the mice were killed, and their gastrointestinal tracts were removed for X-ray images. The intestinal peristalsis activity was assessed as the ability of BeSO4 to be translocated from stomach to colon.
d-Xylose absorption test.
To measure intestinal carbohydrate absorption (1), the 30% d-xylose solution (0.8 g/kg body wt) was administered using a feeding tube to the overnight-fasted Sim2−/− mice (5 wk old). Blood samples (50 μl) were drawn from the retro-orbital venous plexus immediately after (0 min) and 60 min (60 min) after xylose administration. Serum xylose concentrations were measured using a d-xylose assay kit (Megazye).
Gastrointestinal emptying test.
The overnight-fasted Sim2−/− mice (6 wk old) were intragastrically given 0.2 ml of 0.2% phenol red in 1.5% hydroxypropylmethylcellulose solution. After 45 min, mice were killed, and their gastrointestinal tracts were quickly clamped at proper sites to separate stomach, duodenum, jejunum, and ileum from each other before removal. Each part was then isolated, cut into pieces, and homogenized with 10 ml of 0.1 M NaOH at room temperature. Later (1 h), the homogenate was spun at 3,000 g for 10 min, and 5 ml of the supernatant were then added to 0.5 ml of 20% trichloroacetic acid and centrifuged at 3,000 g for 20 min to remove precipitate. One milliliter of this pink-colored supernatant was added to 4 ml of 0.5 M NaOH, and the absorbance at 560 nm was measured. Phenol red recovered from each part of the gastrointestinal tract of an individual mouse was summed up and used as the input level to calculate the percentage of levels remaining in each region.
The Sim2−/− mice (4 wk old) were treated with antibiotic by administration of drinking water containing 200 μg/ml ketoconazole and 1 mg/ml streptomycin for 3 days. Control mice received only water. After treatment, mice feces were collected for 3 h and immersed in sterile water for 1 h to release microbes. The fecal solution was then plated on Luria broth and thioglycollate/resazurine agar plates and incubated at the appropriate O2 levels at 37°C for 24 h to measure microbial colony-forming units (cfu) of aerobic and anaerobic microbes, respectively.
Real-time PCR quantitation of intestinal bacterial 16S rDNA gene.
In addition to the cfu culture method, qPCR with SYBR green system (Applied Biosystems) was used to quantitate intestinal total bacteria rDNA levels. The primers used were forward primer 5′-GCAGGCCTAACACATGCAAGTC and reverse primer 5′-CTGCTGCCTCCCGTAGGAGT (7). Briefly, the entire gastrointestinal tracts of newborn mice were isolated and homogenized in 200 μl H2O. Homogenate was briefly spun to remove debris before addition of an equal volume of 1% SDS/200 mM NaOH to lyse enteric bacteria. DNA was then precipitated by adding 2.5× volume of ethanol, dissolved in TE buffer at 50°C, and used directly for qPCR assay.
Recombinant adenovirus generation.
To generate the recombinant adenovirus carrying the mouse transcription factor 7-like 2 (TCF7L2) expression cassette, Tcf7l2 cDNA was introduced to the Adeno-X-ZsGreen genome using an In-Fusion system (Adeno X expressionsystem-3; Clontech). The recombinant adenoviruses generated from this system express both mouse TCF7L2 and ZsGreen1 fluorescent proteins. The green fluorescent protein tag can be used to easily monitor transduction and virus production. To package adenoviruses, the Pac I-linearized plasmid DNA was used to transfect human embryonic kidney 293 cells using Lipofectamine 2000 (GIBCO). About one week later, cells and conditioned media were collected when the late cytopathic effect appeared. The titer (plaque-forming units/ml) of the recombinant adenovirus was determined by the plaque assay.
Paneth cell count.
The small intestines from 8-wk-old mice were divided into duodenum, jejunum, and ileum segments, and each was fixed in buffered formalin and embedded in paraffin to prepare 5-μm thin sections. Intestinal sections were dewaxed and stained with hematoxylin and phloxine-tartrazine. For Paneth cell counts, 20 crypts of each segment were evaluated only if aligned along the longitudinal axis such that the lumen of the crypt could be seen along its length (as shown in Fig. 5B, left). The results were expressed as Paneth cells per crypt. To determine whether Paneth cell counts differ along the small intestine, counts were performed in sections of normal mucosa taken from duodenum, jejunum, or ileum.
Promoter-secreted alkaline phosphatase reporter constructs.
A secreted alkaline phosphatase (SEAP) reporter system (Clontech) utilizing a secreted form of human placental alkaline phosphatase as a reporter molecule was used to construct the indicated AMP gene promoter-SEAP reporter plasmids (see Fig. 7A). For the Tcf7l2 promoter-SEAP reporter plasmids, an 1,853-bp DNA fragment containing the sequence from −1,469 to +384 of the Sim2 gene was subcloned into pSEAP (TATA box-less derived from pTAL-SEAP from Clontech). A series of p-SEAP plasmids containing deletions or a mutation of this 1,853-bp fragment was then constructed (depicted in Fig. 7B). The mouse SIM2 expression vector was constructed by subcloning the coding region of SIM2 cDNA into the pDNR.CMV plasmid (Clontech) and named pCMV.mSIM2. The sequences of all the plasmid constructs were confirmed before use.
Transient cotransfection and promoter activity assay.
The 293 cells were grown to 50% confluence in 60-mm cell culture dishes. Three micrograms of promoter-SEAP plasmid DNA, 1 μg of internal standard, the β-galactosidase reporter pCH110, and 1 μg of expression vector DNA for SIM2, pCMV.SIM2, were mixed with 200 μl of 0.25 M CaCl2, and a solution of 200 μl of 50 mM HEPES, 280 mM NaCl, and 1.75 mM NaH2PO4 (pH 7.1) was added dropwise. The mixture was added to the cells after 15 min of incubation. Cell media were changed 16 h after transfection, and both the cells and media were harvested 72 h later. Twenty-five microliters of the harvested media were then used directly for SEAP activity assays. Cells were lysed, and the cell lysates were assayed directly for β-galactosidase activity. The SEAP- and β-galactosidase-initiated light signals were measured in a microplate luminometer (model TR717; PE Biosystems). The relative SEAP activity was calculated based on the activity of β-galactosidase internal standard.
β-Galactosidase staining of whole mount intestine and sectioning.
Intestines of Vil-cre/R26R-LacZ mice were isolated and stained for β-galatosidase activity as described previously (19). Before processing, the isolated intestines were cut open longitudinally and then cut into sixths, each about 2 cm in length. After β-galactosidase staining, intestine strips were washed in PBS and used first for counting the positive villi under a stereomicroscope. After villi counting, the intestine strips were incubated in 30% sucrose in PBS overnight at 4°C in the dark and then embedded in optimum cutting temperature medium (TissueTek) for cryosectioning at 15 μm, −20°C.
Isolation and culture of intestinal epithelial cells.
Small intestinal epithelial cells were isolated and cultured essentially according to the protocol described previously (6). Briefly, the small intestines of 2-wk-old mice were isolated, sliced into small pieces (0.5 cm long), and washed five times in HBSS buffer supplemented with antibiotics. The intestine pieces were further minced and digested for 2 h at 37°C, 7% CO2 in DMEM containing 1% FBS, collagenase XI (75 U/ml), and dispase (0.2 U/ml) to release crypts. The isolated crypts were then collected and cultured at 37°C, 7% CO2 in DMEM supplemented with 2% FBS, insulin (0.25 U/ml), sodium pyruvate (6.8 mM), glucose (4.5 g/l), and antibiotics to obtain primary epithelial cells. To confirm whether the cells grown in the primary culture were epithelial cells, the antibody against cytokeratin (keratin) 18 (ab668; Abcam) was used to discriminate between epithelial cells and other cells in the immunocytochemistry analysis.
Real-time PCR analysis of mRNA levels.
Total RNAs were extracted from frozen tissues using the TRIzol RNA reagent (GIBCO-BRL) followed by further purification through an RNAeasy column (Qiagen). The purified total RNAs were reverse transcribed using a high-capacity cDNA reverse transcription kit (Applied Biosystems) with random hexamers as the primers. The real-time PCR measurement of the individual cDNAs was performed in triplicate using 25 ng cDNA and SYBR green dye to measure the double-stranded DNA formation with the ABI 7500 real-time PCR system (Applied Biosystems). Optimization of reactions was performed according to the manufacturer's instructions. The expression levels of 18S rRNA were used to calibrate the expression of other genes in each sample. Oligonucleotide sequences used as primers for real-time PCR are listed in Table 1.
All data are presented as means ± SE. The significant difference of single data points against the indicated control data point was assessed by the Student's two-tailed t-test.
Sim2−/− mice develop bloated gastrointestinal tract after birth.
The SIM2 knockout mouse strain (Sim2−/−) was generated using gene targeting technology in conjunction with the Cre/loxP DNA recombination system as shown in Fig. 1A (41). With the use of this strategy, the first exon of the Sim2 gene along with the selection marker neo gene cassette were excised by eIIa.CRE at the zygote stage to generate the Sim2 knockout allele (Fig. 1B). Upon deletion of exon 1, Sim2 expression was abolished because of the removal of its 5′ sequences containing the start codon for translation and of splicing sequences for proper mRNA editing between exons 1 and 2. In addition, the function of SIM2, if the truncated Sim2 mRNA is still translated, would be expected to be abolished because of the lack of the bHLH domain needed for DNA binding and transcriptional regulation. The knockout of SIM2 expression was confirmed by RT-PCR of the Sim2 transcript using a pair of primers targeting the Sim2 coding sequences downstream of exon 1 (Fig. 1C).
The Sim2−/− mice developed a bloated gastrointestinal tract shortly after birth (Fig. 1D). This knockout phenotype had been described previously in a study using another Sim2 knockout strain and was thought to result from a defect in the palate (15). However, we did not find any craniofacial abnormalities or defects in the palates of 1-day-old Sim2−/− mice (Fig. 2A), indicating that the bloated gastrointestinal tract is not the result of defects in craniofacial structure as suggested previously.
Around one-half of the 1-day-old Sim2−/− mice with less severely bloated gastrointestinal tracts survived the neonatal stage and continued to thrive, although they grew slowly during the first few weeks possibly because of reduced nutrient absorption (Fig. 2, B and C), likely caused by gas distention in their gastrointestinal tracts. Interestingly, despite having normal craniofacial structure, the surviving Sim2−/− mice still exhibited significant gas retention in their gastrointestinal tracts, and this phenotype persisted throughout their lifetimes (Fig. 2D), suggesting that the gas distention in the gastrointestinal tract of surviving Sim2−/− mice might be caused by a direct effect of SIM2 knockout in the gastrointestinal tract.
Overgrowth of intestinal microbes causes bloated gastrointestinal tracts in Sim2−/− mice.
Two other major causes that lead to intestinal gas and flatulence are the reduction of intestinal peristalsis and microbial overgrowth (27). Accordingly, we examined the intestinal peristalsis of the SIM2 knockout mice by performing the gastric emptying assay (44). After loading the stomach with 1 ml of BeSO4 via a feeding tube, X-ray was used to monitor its movement (Fig. 2E, shown as white area in the X-ray film) along the gastrointestinal tract. In Sim2−/− mice, in spite of the numerous air-interrupting areas along their gastrointestinal tracts, the BeSO4 movement was not delayed compared with that in the heterozygous control mice. Furthermore, using phenol red to monitor transport of the ingested contents along the intestine, we found that the levels of phenol red transported in each intestinal segment were comparable between Sim2−/− mice and their control littermates (Fig. 2F). These results indicate that intestinal peristalsis is unaffected by the loss of SIM2 function.
We then examined both the aerobic and anaerobic bacterial population in the intestine of 1-day-old Sim2−/− mice and found that their gastrointestinal tracts harbored significantly higher cfu of bacteria than the gastrointestinal tracts of heterozygous control mice (Fig. 3A) and that the bacterial population increased with the severity of distention (Fig. 3B). In addition, the real-time PCR analysis for the total intestinal bacterial 16S rDNA levels was performed to corroborate the cfu results (Fig. 3A). Bacteria can easily enter mouse neonates via the oral route during the intensive contact with mother's intestinal or vaginal microflora during birth. The significantly increased aerobic bacterial population in the bloated gastrointestinal tracts of Sim2−/− newborns suggests that overgrowth of microbes might be a cause for the gas distention in the gastrointestinal tracts of Sim2−/− newborn mice.
Accordingly, we treated 9-wk-old Sim2−/− mice with streptomycin/ketoconazole for 3 days. This treatment significantly reduced the intestinal bacterial population in Sim2−/− mice as determined by both the reduced bacterial cfu numbers and 16S rDNA levels from the facets of Sim2−/− mice (Fig. 3C). In addition, after treatment, there were fewer and smaller intestinal gas bubbles in the gastrointestinal tracts of Sim2−/− gastrointestinal tracts (Fig. 3D), strongly suggesting that microbial overgrowth is a prominent cause for the excessive intestinal gas in Sim2−/− gastrointestinal tracts.
Innate immunity is impaired in sim2−/− gastrointestinal tracts.
Because the bloated gastrointestinal tract is developed shortly after birth in Sim2−/− mice, we examined whether the gastrointestinal innate immunity is defective in the knockout mice. For the status of gastric acid production, the stomach contents of Sim2−/− neonates was washed out for pH value measurement. We found that the pH value in the stomach of Sim2−/− neonate (5.09 ± 0.32) was not significantly different from that in the stomach of control neonate (5.28 ± 0.51), indicating that the gastric acid production is not altered in Sim2−/− mice. The Paneth cell-derived cryptdins (α-defensins) are the predominant AMPs and the key molecules of innate immunity in mammalian intestines (4, 30). Other AMPs, such as lysozyme, RegIIIγ, and phospholipase A2, are also crucial in defense against intestinal microbes (3, 45). There are at least six cryptdins in the mouse gastrointestinal tract (42). Although Paneth cells and the significant expression of these cryptdins occur about 2 wk after birth when intestinal crypts are developed, some cryptdins are detectable at the neonatal stage (13, 32). We analyzed the expression of cryptdins 1 and 6, and the matrix metalloproteinase 7 (MMP-7) that processes and activates cryptdin precursors, and found that their levels were significantly reduced in the intestines of Sim2−/− neonates (Fig. 4A). In addition, the expression of cathelicidin-related antimicrobial peptide (CRAMP), an important antimicrobial factor during the first two postnatal weeks (3), was also significantly reduced in the Sim2−/− neonate intestine. Similarly, in the intestines of 15-day-old Sim2−/− mice, in which intestinal crypts had developed (Fig. 5A), expression of MMP7, all the cryptdins, and several other antimicrobial factors examined was still significantly reduced (Fig. 4, B, C, and D). On the other hand, the crypt structure and number of Paneth cells appeared unaffected in the intestines of Sim2−/− mice (Fig. 5B), indicating that the decreased mRNA expression of these antimicrobial factors is not due to a defect in the development of crypts or a reduction in the number of Paneth cells but is likely caused by a dysregulation of these factors at the transcriptional level. Taken together, our results have suggested that SIM2 may play a crucial role in maintaining intestinal AMP levels.
Intestinal epithelium-specific knockout of SIM2.
To determine if the aforementioned defect in intestinal immunity is caused by a direct action of SIM2 in intestinal cells, we examined the immunity in the intestines of mice (Sim2Lp/−/Vil-cre) in which the Sim2 floxed allele is deleted specifically in the intestinal epithelial cells by the expression of a villin promoter-driven cre transgene (31). We found that the intestines of Sim2Lp/−/Vil-cre mice, at 9 wk old, indeed develop numerous gas bubbles, whereas the intestines of control littermates, Sim2Lp/−/Vil-cre, harbor none to very few gas bubbles (Fig. 6A), indicating that the intestinal SIM2 is directly involved in the development of bloated gastrointestinal phenotype. However, the severity of gas bubbles, both in number and size, was reduced in Sim2Lp/−/Vil-cre mice compared with those in the complete knockout Sim2−/− mice (Fig. 6A). Similarly, the expressions of antimicrobial factors, such as cryptdin 4 and RegIIIγ, in the intestines of Sim2Lp/−/Vil-cre mice were not reduced to the extent found in those of Sim2−/− mice (Fig. 6B). We examined the efficiency of the Vil-CRE in mediating DNA deletion in intestines using an R26R-LacZ reporter mouse strain and found that Vil-CRE did not completely reactivate the R26R-LacZ reporter gene in all the villi of intestine for the β-galatosidase activity. There were 13.8 ± 2.1, 11.4 ± 1.9, and 17.4 ± 2.3% of villi in the duodenum, jejunum, and ileum, respectively, not stained by the β-galactosidase (Fig. 6C, left), and crypts were not stained within the area of unstained villi (Fig. 6C, right). Accordingly, the reduction in severity of gas distention in Sim2Lp/−/Vil-cre mice is likely caused by an incomplete CRE-mediated excision of the Sim2 floxed allele in the intestine cells due to the mosaic expression of Vil-cre or inefficient CRE-mediated DNA excision. A similar situation was previously observed with the use of the same Vil-cre transgenic mice (26, 40). Indeed, we found that the SIM2 mRNA levels were only one-half reduced in the intestines of Sim2Lp/−/Vil-cre mice and that their TCF7L2 mRNA levels were not reduced to the levels found in the complete knockout Sim2−/−/Vil-cre littermates (Fig. 6D). Nevertheless, these results showing that Sim2Lp/−/Vil-cre mice also develop gas distention and express less antibacterial factors in their gastrointestinal tracts support that SIM2 is able to act directly on intestinal cells to regulate expression of antimicrobial factors and maintain intestinal immunity.
SIM2 regulates the expression of major intestinal AMPs through both direct and indirect mechanisms.
Because SIM2 is a transcription regulator that interacts directly with a DNA sequence on the proximal promoter region of target genes to control their transcription, we suspected that the SIM2-binding sequences might be present on the proximal promoter regions of these AMP genes to mediate SIM2 regulation. Accordingly, we searched the SIM2-binding sequence 5′-(G/A)(T/A)ACGTG (49) within at least 5 kbp of the proximal promoter of each AMP gene and found strong SIM2-binding sites on the promoters of the CRAMP (5′-AACGTG, at −1,379 to −1,374 bp), RegIIIγ (5′-AAACGTG, at −945 to −951 bp), and phospholipase A2 (5′-ATACGTG, at −1,195 to −1,201 bp) genes, suggesting that SIM2 may directly bind to and regulate transcriptional activity of these genes. Indeed, when the promoter-reporter constructs for these genes were assayed in cultured cells, the activity of each promoter fragment containing the SIM2-binding site was significantly enhanced if the SIM2-expression vector was cotransfected (Fig. 7A). Interestingly, however, the 5-kb proximal promoter regions of lysozyme and all the cryptdin genes do not possess a binding sequence for SIM2. This suggests that SIM2 may not directly control the expression of these factors, and it raises the question of how SIM2 regulates cryptdin expression to maintain intestinal immunity.
In the intestinal epithelium, Wnt signaling is crucial not only for maintaining stem cell function and proper epithelial cell positioning but also for the development and terminal differentiation of Paneth cells, which involves the transcriptional activation activity of the β-catenin/TCF7L2 (also known as T cell specific factor 4, TCF4) complex (46, 47). The β-catenin/TCF7L2 target genes, such as SOX9, play key roles in the development of Paneth cells (2). In addition, TCF7L2 is the master regulator in controlling expression of MMP7 and all the cryptdins in Paneth cells (47). Accordingly, if SIM2 does not directly target cryptdin genes but has to depend on another factor to regulate their expression, TCF7L2 may be the best candidate to mediate SIM2 action on cryptdin expression. Indeed, the mRNA level of TCF7L2 was significantly reduced in the Sim2−/− intestine (Fig. 4), suggesting that TCF7L2 may be a target gene of SIM2. Moreover, there is a consensus SIM2-binding site (5′-ATACGTG) located at −1,250 to −1,244 bp in the promoter region of Tcf7l2 gene. Our further analysis of Tcf7l2 promoter activity confirmed that the Tcf7l2 proximal promoter can be activated by SIM2 in cell culture and that this binding site is functional and responsible for most of the SIM2-stimulated promoter activity in the cotransfection assay (Fig. 7B). Taken together, our results indicate that SIM2 directly regulates the expression of TCF7L2, and we hypothesize that TCF7L2 mediates the action of SIM2 on the expression and activation of all cryptdins for intestinal immunity.
To prove our hypothesis, we used approaches in both primary intestinal epithelial cells in culture and mutant mice to test whether the exogenous or activated TCF7L2 was able to restore the cryptdin levels in primary Sim2−/− epithelial cells or reduce gas distention in Sim2−/− mice, respectively (Fig. 8). Figure 8B shows that recombinant adenoviral vector carrying a TCF7L2 expression cassette was able to increase the levels of cryptdins, but not RegIIIγ, in the primary epithelial cells derived from Sim2−/− intestinal crypts, indicating that the reduced levels of cryptdins in the Sim2−/− intestine can be restored by exogenous TCF7L2.
The Apcmin/+ mice carry a dominant heterozygous nonsense mutation on the Apc gene. The APC protein functions as a negative regulator of β-catenin by binding to β-catenin, which leads to its proteasomal degradation (38, 39). Thus, in Apcmin/+ mice, β-catenin can escape the APC-mediated degradation, translocate into the nucleus to interact with TCF4 (TCF7L2), and then activate the expression of their target genes (36). In principle, although the TCF7L2 levels are reduced in the Sim2−/− intestine, the Apcmin allele should be able to enhance the β-catenin-TCF signaling and consequently compensate for the reduced levels of TCF7L2 to maintain cryptdin expression and in turn improve the problem of a gas-distended gastrointestinal tract. Accordingly, we introduced the Apcmin allele into Sim2−/− mice by breeding Apcmin/+ mice with Sim2−/− mice, and, indeed, we observed that, in the intestines of 1-day-old Sim2−/−, Apcmin/+ mice, the cryptdin 1 transcript level was restored (Fig. 8C) and that the severity of gas distention was greatly reduced, if not abolished (Fig. 8D). These results further support that TCF7L2 mediates the action of SIM2 on cryptdin expression to maintain intestinal innate immunity.
In this study, we have demonstrated for the first time the important regulatory role of transcription factor SIM2 in intestinal innate immunity. SIM2 is required for the expression of major intestinal AMPs, including CRAMP, lysozyme, and all the cryptdins that are crucial in preventing microbial overgrowth and invasion. Interestingly, SIM2 controls the transcriptional activities of these AMP genes through different mechanisms. For those AMP genes whose proximal promoter regions possess a binding sequence for SIM2, SIM2 regulates their expression likely only by the direct interaction with their promoter sequences. Whereas, for the cryptdin genes whose proximal promoter regions do not contain a SIM2-binding site, SIM2 controls their expression through TCF7L2, the major factor known for regulating cryptdin expression and activation. Regardless, our results establish that SIM2 is a key regulator in maintaining intestinal AMP levels.
SIM2 is a member of the bHLH-PAS transcription factor family. Although most bHLH-PAS factors are activators of transcription, Sim2 acts as a transcriptional repressor in mammalian cells (9, 35). In addition to its role in CNS development and mental retardation in DS, SIM2 is also involved in muscle and limb development (10, 16, 21). Recently, several studies have reported that, depending on the origin of tumors, SIM2 may act as a suppressor or stimulator on their growth and development (14, 15, 28, 29). For these reported functions, SIM2 acts by repressing transcription of its target genes to exert its effects. Here we show that, to maintain intestinal immunity, SIM2 exerts its effect by activating transcription of TCF7L2 and numerous AMP genes. In view of this, SIM2 is a versatile transcription factor, capable of either transcriptionally repressing or activating its target genes.
The major function of the intestine is to digest food and absorb nutrients, so the intestine is continuously defending against attacks by microbes ingested through diet and water. To efficiently digest and absorb nutrients, the intestinal epithelial layer has evolved a large surface area that consists of folds, villi, and microvilli for this purpose, which in turn provides increased surface areas for microbial attack. In addition to the dietary microbial threat, there is a continuous potential threat from a wide array of commensal microbes. Therefore, the body has to maintain effective intestinal mucosal defense and homeostasis to keep the commensal and pathogenic microbes from overgrowth in the intestine. Mucosal innate immunity provides the first line of defense. There are several noninflammatory mechanisms used to maintain this innate immunity for immediate protection, such as secretion of antimicrobial peptides, expression of innate immune receptors, and autophagy of intracellular bacteria (5, 19). Recent progress on mucosal defense has highlighted the significant role of AMPs in innate immunity. Paneth cell α-defensins, named cryptdin in mice, are the key AMPs in the small intestine. They are synthesized as larger precursor peptides and require proteolytic cleavage to become activated. In mice, there are six or more cryptdins and numerous cryptdin-related peptides expressed in the small intestine (4, 24, 47). β-Catenin/TCF7L2 (TCF4) appears to be the major regulator in controlling the expression of all the mouse cryptdins and MMP7, the protease responsible for proteolytic cleavage of cryptdin propeptides (47). Several pathways, such as alternative splicing of the TCF7L2 transcript and the stability of β-catenin, can regulate the transactivation activity of TCF7L2 (36, 48). However, the mechanism and the regulator that controls the expression of TCF7L2 have not been reported. Our study has established that SIM2 is an important regulator in activating the transcription of TCF7L2 in the intestine. As a result, through TCF7L2, SIM2 can control the expression and activation of all cryptdins. Considering that SIM2 also directly regulates the expression of numerous other AMPs and that lack of SIM2 causes microbial overgrowth in the intestine, it is clear that SIM2 plays a crucial role in maintaining intestinal mucosal AMP levels and immunity against microbes.
In addition to the intestinal mucosa, the oral mucosa and saliva also contain numerous defense factors that are involved in both innate and acquired immunity against microbes in the oral cavity (16a). Although in this study we uncovered the role of SIM2 in intestinal mucosal immunity against microbial overgrowth, we did not rule out the possibility that the gas accumulation in the stomach of Sim2−/− neonates might be caused in part by the overgrowth of microbes in the upper digestive tract due to a defect in the oral immunity against microbes, and it remains to be elucidated whether SIM2 is also involved in regulating the oral defense system.
This work was supported in part by a research grant from the Ministry of Science and Technology (NSC-101-2311-B-001-031-MY3) to Y. H. Lee.
No conflicts of interest, financial or otherwise, are declared by the authors.
K.-J.C. and Y.-H.L. conception and design of research; K.-J.C. performed experiments; K.-J.C., A.L., and Y.-H.L. analyzed data; K.-J.C. and A.L. prepared figures; A.L. and Y.-H.L. interpreted results of experiments; A.L. and Y.-H.L. drafted manuscript; Y.-H.L. edited and revised manuscript; Y.-H.L. approved final version of manuscript.
- Copyright © 2014 the American Physiological Society