Experimental Colitis Is Associated with Transcriptional Inhibition of Na+/Ca2+ Exchanger Isoform 1 (NCX1) Expression by Interferon γ in the Renal Distal Convoluted Tubules*

Background: Defective renal Ca2+ reabsorption contributes to impaired systemic Ca2+ homeostasis and loss of bone density in IBD. Results: During colitis, IFNγ represses NCX1 expression and activity in a Stat1-dependent manner. Conclusion: IFNγ contributes to the reduced renal Ca2+ reabsorption during colitis by repressing basolateral NCX1. Significance: IFNγ inhibits renal NCX1 to contribute to negative systemic Ca2+ balance and increased bone resorption in IBD patients. NCX1 is a Na+/Ca2+ exchanger, which is believed to provide a key route for basolateral Ca2+ efflux in the renal epithelia, thus contributing to renal Ca2+ reabsorption. Altered mineral homeostasis, including intestinal and renal Ca2+ transport may represent a significant component of the pathophysiology of the bone mineral density loss associated with Inflammatory Bowel Diseases (IBD). The objective of our research was to investigate the effects of TNBS and DSS colitis and related inflammatory mediators on renal Ncx1 expression. Colitis was associated with decreased renal Ncx1 expression, as examined by real-time RT-PCR, Western blotting, and immunofluorescence. In mIMCD3 cells, IFNγ significantly reduced Ncx1 mRNA and protein expression. Similar effects were observed in cells transiently transfected with a reporter construct bearing the promoter region of the kidney-specific Ncx1 gene. This inhibitory effect of IFNγ is mediated by STAT1 recruitment to the proximal promoter region of Ncx1. Further in vivo study with Stat1−/− mice confirmed that STAT1 is indeed required for the IFNγ mediated Ncx1 gene regulation. These results strongly support the hypothesis that impaired renal Ca2+ handling occurs in experimental colitis. Negative regulation of NCX1- mediated renal Ca2+ absorption by IFNγ may significantly contribute to the altered Ca2+ homeostasis in IBD patients and to IBD-associated loss of bone mineral density.

Na ϩ /Ca 2ϩ exchange is a major mechanism for cellular calcium extrusion by DCT and CNT (11,12). The sodium-calcium exchanger 1 (Na ϩ /Ca 2ϩ exchanger 1 or NCX1) is a basolateral plasma membrane transport protein that belongs to solute carrier family 8 (SLCA1) and plays a critical role in maintaining intracellular Na ϩ and Ca 2ϩ homeostasis. There are three mammalian NCX isoforms: NCX1 is widely expressed in the heart, kidney, brain, blood vessels, and other organs; expression of NCX2 and NCX3 is limited mainly to the brain and skeletal muscle. Generally, NCX1 mediates Ca 2ϩ efflux in nonpolarized cells as well as in polarized epithelia. Ncx1 mRNA is downregulated in vitamin D deficiency (in 1␣-OHase Ϫ/Ϫ mice) and corrected by 1,25(OH) 2 D 3 supplementation (13). Huybers et al. (14) showed that experimental Crohn's-like ileitis developed spontaneously by TNF ⌬ARE mice is associated with reduced expression of renal Ncx1, although the cytokine(s) involved and mechanisms responsible for this regulation have not been described. This decrease was accompanied by decreased serum 1,25(OH) 2 D 3 levels and reduced trabecular and cortical bone thickness (14).
Characterization of transcriptional regulation of Ncx1 gene expression is complicated due to multiple tissue-specific variants of Ncx1 resulting from alternative promoter usage (H1, K1, and Br1) and alternative splicing, which also differ in their ionic and kinase-dependent regulation (15)(16)(17)(18). The mouse kidneyspecific promoter has not been characterized previously. In this study, we map the mouse kidney-specific promoter and the related transcription start site. We also show that renal Ncx1 mRNA expression is reduced in two models of experimental colitis. We further show that IFN␥ down-regulates Ncx1 expression in vivo and in vitro via a STAT1-dependent mechanism. Collectively, our data provide another layer of complexity which explains the mechanism of defective renal Ca 2ϩ reabsorption in IBD, and its potential contribution to systemic Ca 2ϩ imbalance and the associated loss of bone mineral density.

EXPERIMENTAL PROCEDURES
Experimental Colitis-TNBS colitis was induced in mice (129S1 background) by intracolonic single dose administration of trinitrobenzene sulfonic acid (TNBS, Sigma Aldrich) (2.5 mg/mouse dissolved in 50% ethanol). Control mice received 50% ethanol in PBS enema. All mice were sacrificed 5 days later. DSS colitis (129S1 background) was induced by 3% (w/v) dextran sodium sulfate (DSS; molecular weight 40 to 50 kDa; Affymetrix) added to the drinking water for 7 days, after which the mice were sacrificed and the tissues collected for analysis. Both TNBS and DSS are widely used chemically induced models of intestinal inflammation (19). In both models of colitis, the body weight changes were monitored daily after the induction of colitis. The mice were sacrificed when they lost 15-20% of their initial body weight. Colonic inflammation was assessed based on histological changes as well as levels of proinflammatory cytokines. The animal use protocol was reviewed and approved by the University of Arizona Institutional Animal Care and Use Committee. Tissue samples including kidneys and colons were harvested for histology (H&E staining), as well as protein and RNA isolation. Since the distal convoluted tubules (DCT) and the connecting tubules (CNT) play a major role in Ca 2ϩ reabsorption, we used P TRPV5 -EGFP transgenic mice (C57Bl/6), which express eGFP under the control of a 3.6-kb fragment of the mouse TRPV5 promoter in the epithelium of the late DCT, CNT, and the initial part of the cortical collecting duct, for the NCX1 functional assay. These mice were generously provided by Drs Jeppe Praetorius and Marlene Vind Hofmeister (Aarhus University, Denmark) (20). STAT1 Ϫ/Ϫ (129S6/SvEv) mice were purchased from Taconic Farms (Hudson, NY). All qPCR reactions were performed using pre-designed Taqman assay primer/probe sets (Life Technologies, CA), and ran on a Bio-Rad CFX96 real-time PCR detection system. Cycling parameters were determined and resulting data were analyzed by using the comparative Ct method as a means of relative quantification, normalized to an endogenous reference (TATA box-binding protein; TBP) and relative to a calibrator (normalized Ct value obtained from control mice) and expressed as 2 Ϫ⌬⌬Ct (Applied Biosystems User Bulletin #2: Rev B "Relative Quantification of Gene Expression"). Serum and tissue levels of mouse interferon ␥ (mIFN␥) were analyzed using an ELISA kit (eBioscience). The assay was performed according to manufacturer's instructions. 45 Ca 2ϩ Uptake Assay-The P TRPV5 -EGFP transgenic mice were injected with PBS or recombinant mIFN␥ (50,000 units/ mouse) intraperitoneal for 3 days. On the fourth day, the kidneys were collected, minced, and digested with RPMI 1640 medium containing 0.1% collegenase and 5% FBS. Tubule cells were strained with a nylon mesh and washed two times with RPMI 1640 medium containing 10% FBS plus penicillin and streptomycin. GFP ϩ cells were sorted using FACAria cell sorter and were used for the NCX1 activity as described earlier (21). Briefly, the cells were washed twice with washing buffer (10 mM MOPS (pH 7.4), 140 mM NaCl). Cells were then loaded with Na ϩ by incubation with 10 mM MOPS (pH 7.4), 140 mM NaCl, 1 mM MgCl 2, in the presence of 0.4 mM ouabain, and 25 M nystatin for 10 min at room temperature. The nystatin was removed from the cells by two washes with wash buffer containing 0.4 mM ouabain. Uptake was initiated by resuspending the cell pellet in assay buffer containing 10 mM MOPS (pH 7.4), 140 mM KCl (or NaCl as blank), 25 M CaCl 2 , 0.4 mM ouabain, and 5 Ci/ml 45 Ca 2ϩ . After 10 min of incubation, the reaction was terminated by adding 1 ml of ice-cold stop buffer (140 mM KCl and 1 mM EGTA). Cells were washed twice with stop buffer, and cell pellets were then lysed by freeze/thaw in 1N NaOH at 80°C for 20 min. Aliquots were subjected to scintillation counting and protein assay.
Fractional Urinary Ca 2ϩ Excretion-Plasma (P) and urine (U) creatinine (Cr) and Ca 2ϩ were analyzed by commercial assays (Abcam, Cambridge, MA; and BioVision, Milpitas, CA, respectively) from the samples collected at the time of sacrifice. Fractional excretion of Ca 2ϩ (FECa 2ϩ ) was calculated as described earlier (6).
Immunofluorescence-Kidneys were harvested and fixed in 10% neutral buffered formalin (Fisher Scientific). Fixed tissues were embedded in paraffin, and cut into 5-m-thick tissue sections. After deparaffinization and rehydration, antigen retrieval was performed by heating the slides in citrate buffer (10 mM sodium citrate, 0.05% Tween 20, pH 6.0). After washing in phosphate-buffered saline (PBS), the slides were blocked for 1 h in 10% normal chicken serum in PBS containing Tween 20 (0.1%), (PBST). Next, sections were incubated with primary antibodies directed against NCX1 (AB3516P, 1:100 dilution, rabbit polyclonal, Millipore) and NCCT (sc-21554, 1:50 dilution, goat polyclonal, Santa Cruz Biotechnology) in PBST containing 3% BSA, overnight at 4°C. After three washes in PBST, the slides were incubated with the chicken anti-goat and antirabbit antibodies conjugated with Alexa Fluor (6 g/ml in PBST, Invitrogen). After three washes with PBST, slides were mounted with a drop of mounting medium with DAPI (Molecular Probes, Life Technologies). The slides were visualized under the microscope (EVOS FL Auto, Life Technologies).
Identification of Mouse Renal-specific Ncx1 Gene Promoter-The RNA ligase-dependent rapid amplification of cDNA ends (5Ј-RLM-RACE) kit (FirstChoice RLM-RACE, Invitrogen) was employed to identify the mouse renal Ncx1 transcriptional start site and the gene promoter region. 10 g of total RNA isolated from mouse kidneys was treated with calf alkaline phosphatase followed by tobacco pyrophosphatase, reverse transcribed using a gene-specific primer (5Ј-AAAGATGGGTCTTGGGG-TTC-3Ј), and subsequently tailed with terminal deoxynucleotidyl transferase using a 5Ј adapter (5Ј-GCUGAUGGCGAUGA-AUGAACACUGCGUUUGCUGGCUUUGAUGAAA-3Ј). cDNA was then PCR amplified using the 5Ј-RACE inner Primer (5Ј-CGCGGATCCGAACACTGCGTTTGCTGGCTTTG-ATG-3Ј) and a gene-specific primer (as described above). The final PCR products were purified and cloned into a pGEM-T vector (Promega), and clones were sequenced at The University of Arizona Genetics Core facility, Tucson, Arizona.
Promoter Constructs-5Ј-RACE identified the transcription start site (TSS) of Ncx1 at position Ϫ375 nt relative to the ATG start codon. Guided by these results, we cloned genomic fragments extending from positions Ϫ703, Ϫ379, Ϫ269, Ϫ110 to 0 relative to the Ncx1 TSS, flanked by restriction sites GGTACC (KpnI) at 5Ј and CTCGAG (XhoI) at 3Ј. Ncx1 promoter fragments were amplified from mouse genomic kidney DNA using standard PCR with Accuprime DNA polymerase (Invitrogen). The primers amplifying the promoter region were: Ϫ703 forward 5Ј-GCGGTACCCCGAGGAAGGAAGTAGAGTC-TTC-3, Ϫ379 forward 5Ј-GCGGTACCAAGAGACAGCTTG-CGCTT-3, Ϫ269 forward 5Ј-GCGGTACCT AGAAACGAG-GAGGCCCTA-3, Ϫ110 forward 5Ј-GCGGTACCAAATACC-AATCACCTATG-3 and reverse primer 5Ј-GCCTCGAG-TGCTCCTCAGATCTGATG-3. All amplicons were first cloned into the pGEM-T cloning vector (Promega), then restriction-digested with KpnI and XhoI and ligated into the corresponding site within the multiple cloning sites in the pGL3-basic vector (Promega). Both pGEM-T and pGL3 plasmids with inserts were fully sequenced and verified. The transfections of mIMCD3 cells with the reporter constructs were carried out in 24-well plates using the TransLT1 (Mirius) transfection agent and a luciferase assay was performed with a luciferase assay kit (Promega), and a tube luminometer (FB12, Zylux).
Analysis of the mouse kidney specific promoter sequence revealed the existence of a single potential binding site for STAT1 at position Ϫ385 to Ϫ397 relative to the TSS. Therefore, we constructed plasmid with a mutant STAT response element by site-directed mutagenesis. The mutant sequence for mutagenesis of the STAT1 RE is changed from 5Ј-TTCCTGG-GAA-3Ј to 5Ј-CCTCTGGGAA-3Ј. The mutant PCR products were amplified by using two different sets of primers, Primer set I (forward, 5Ј-GCGGTACCCCGAGGAAGGAAGTAGAGT-CTTC-3Ј; reverse, 5Ј-AATTCCCAGAGGTCCTTCCTGCC-CAAG-3Ј) and primer set II (forward, 5Ј-CTTGGGCAGGAA-GGACCTCTGGGAATT-3Ј; reverse, 5Ј-GCCTCGAGT-GCTCCTCAGATCTGCATG-3Ј). The two amplified products were mixed and amplified with forward primer of set I and reverse primer of set II. The final product was ligated to pGEM-T vector and sequenced. Subsequently, the insert was digested with KpnI/XhoI and ligated to pGL3-basic vector. Reporter assays were performed as described above.
Chromatin Immunoprecipitation (ChIP)-The chromatin immunoprecipitation assay was performed using the Simple-ChIP Enzymatic Chromatin IP Kit (Magnetic Beads, Cell Signaling Technology) according to the manufacturer's protocol. Chromatins from IFN␥-treated and untreated mIMCD3 cells were immunoprecipitated with normal rabbit IgG (negative control) or anti-STAT1 antibodies (sc-346X, rabbit polyclonal, Santa Cruz Biotechnology). Stat1 binding to Ncx1 promoter DNA was quantified by real-time PCR using primers (forward, 5Ј-CTCACCTGTCAGGGGTCTGT-3Ј; reverse, 5Ј-ACTCG-TCATGGCTGTCCTCT-3Ј) designed to amplify a 233-bp region spanning the putative STAT1 cis-element of the Ncx1 promoter using SYBR green. Chromatin incubated with beads alone or bead plus normal rabbit IgG was used to control for nonspecific binding. The products were run on 1.5% agarose gel to confirm the specificity of PCR amplification.

RESULTS
Assessment of Colitis-Both TNBS and DSS administered mice experienced progressive weight loss and the colon showed typical gross signs of inflammation, i.e. increased wet weight, increased weight/length ratio, and colonic wall thickening (data not shown). Histological examination of the colon, revealed hyperplasia, loss of crypts, focal ulcerations, and loss of epithelial cells, those were consistent with other studies with these models from our laboratory (6,(22)(23)(24)(25). Colonic expression of proinflammatory cytokines, IFN␥, TNF, and IL-1␤ mRNA were elevated in both models of experimental colitis. Significant elevation of IL-17 expression was documented in DSS-, but not in TNBS-treated mice (Fig. 1). Serum level of IFN␥ showed a tendency toward an increase in both TNBS and DSS models (data not shown), an observation reflected by IFN␥ protein concentration in the renal tissue (harvested without perfusion). Importantly, its brief half-life (1.1 min) and the rapid systemic elimination of IFN␥ (26) make measurement of systemic IFN␥ levels unreliable and do not represent the true local (colon) and peripheral exposure to the cytokine. This is further emphasized by the observation that even after injection of recombinant IFN␥ (as described later in the report) no increase in the tissue concentration of the cytokine was observed (data not shown), despite robust activation of pTyr 701 -STAT1 and total STAT1 expression ( Fig. 4D and Fig. 6B). Moreover, we have previously convincingly demonstrated that colitis-associated cytokines are associated with extra-intestinal manifestations in tissues such as the bone (23,27), and the kidneys (6,10).
Decreased Renal Expression of Ncx1 mRNA and Protein in Experimental Colitis-Previously, we showed that TNF and IFN␥ negatively affected the expression and activity of renal Klotho gene (10), which in turn led to impaired renal Ca 2ϩ transport by TRPV5 in experimental colitis (6). In this study, we examined the expression of Ncx1, a basolateral Na ϩ /Ca 2ϩ exchanger, in the distal convoluted tubules. Both the TNBS and DSS models of colitis showed an almost 40 -50% decrease in Ncx1 mRNA levels ( Fig. 2A). Consistently, a marked decrease in renal NCX1 protein abundance was also observed in colitic mice, as demonstrated by Western blotting. TNBS-induced colitis resulted in almost no detectable amount of NCX1 protein and a similar trend was observed in the DSS model of colitis (Fig. 2B).
Previous immunohistochemistry studies indicated that NCX1 expression in the epithelial cells is not limited to the basolateral membrane, but is also abundant in other intracellular compartments (28 -30). We confirmed these observations and showed that consistent with Ncx1 mRNA and protein expression, there was a dramatic decrease in NCX1-specific immunofluorescence in the DCT of colitic mice, as compared with healthy controls (Fig. 2C).
IFN␥ Down-regulates NCX1 Expression and Activity in Vitro and in Vivo-As IFN␥ and other proinflammatory cytokines are typically up-regulated in colitis, including our experimental models, we tested their effects on NCX1 expression in cultured murine IMCD3 cells in vitro. This cell line expresses Ncx1 mRNA and protein at levels easily detectable by both qPCR and Western blotting. We tested the effects of IFN␥, TNF, and IL1␤ individually (Fig. 3A) and in combination (data not shown) on Ncx1 mRNA expression and showed that of the three major cytokines, IFN␥ (100 units/ml) most profoundly reduced Ncx1 transcript (Fig. 3A). Similar cytokine effects were observed on the cloned mouse renal Ncx1 gene promoter (Fig. 3B). Use of cytomix (combination of IFN␥, TNF, and IL1␤) had a similar effect to IFN␥ used individually (data not shown). Time course analysis of the effects of mIFN␥ in mIMCD3 cells showed a significant and progressive decrease in the Ncx1 mRNA level by qPCR starting from 12 h (Fig. 4A). The effect of mIFN␥ on NCX1 expression was also reproduced at the protein level in a time-dependent fashion (Fig. 4B). Similarly, mIFN␥ injected P TRPV5 -EGFP transgenic mice showed significantly decreased expression of NCX1 both at the level of mRNA and protein, in the renal lysates (Fig. 4, C and D). The mIFN␥ signaling activation was abundant as evidenced by the increased levels of pTyr 701 -STAT1 expression both in vitro and in vivo (Fig. 4, B  and D). The GFP ϩ renal distal tubule epithelial cells from the mIFN␥-injected P TRPV5 -EGFP transgenic mice were flow- sorted and used for the NCX1 functional Ca 2ϩ uptake assay. Consistent with decreased Ncx1 mRNA and protein expression, the GFP ϩ cells from mIFN␥-injected mice showed a nearly 70% reduction in NCX1-mediated Na ϩ -dependent Ca 2ϩ uptake (Fig. 4E). Furthermore, compared with PBS-injected controls, exogenous IFN␥ significantly reduced renal Ca 2ϩ reabsorption in vivo, as indicated by increased fractional urinary Ca 2ϩ excretion (FECa 2ϩ ) (Fig. 4F).
Mapping of the Renal Ncx1 Transcription Start Site-Originally characterized and cloned as a cardiac Na ϩ /Ca 2ϩ exchanger, NCX1 is also highly expressed in the brain and kidney, and at lower levels in almost all other tissues. The presence of different non-coding initiating exons is a consequence of three independent promoters selectively driving and regulating expression of Ncx1 in different tissues: one specific to heart, one to kidney, and one for other tissues. The Ncx1 gene tran-script in mammals has been shown to undergo alternative splicing of six exons: A, B, C, D, E, and F (31). Exons A and B were shown to be mutually exclusive and the other four exons were cassette-type exons. Renal tissues specifically express exons B, C, D, E, F (31). The mouse renal Ncx1 alternative promoter and transcription start sites (TSSs) were evaluated by 5Ј Rapid Amplification of cDNA Ends (RACE) using total RNA from mouse kidneys. After reverse transcription using primers provided with the RACE-PCR kit, the initial PCR reaction was performed using a reverse gene specific primer complementary to a sequence located in Exon C and the forward upstream primer provided by the kit. The products of this reaction were then subjected to nested PCR and yielded a 600-bp product representative of the predominant transcript present in the renal tissue (Fig. 5A). Following sequencing of this product, our findings were largely consistent with other species (rabbit, rat, and cat) Ncx1 expression was normalized to that of TBP relative to a calibrator (normalized Ct value obtained from untreated healthy mice) and expressed as 2 Ϫ⌬⌬Ct (mean Ϯ S.D.; *, p Ͻ 0.05 colitis versus controls; Student t test). B, NCX1 Western blot was performed with 20 g of kidney lysate protein control and TNBS-treated (left panel) or DSS-treated (right panel) mice. Loading was normalized to expression of ␤-actin. C, immunofluorescent labeling of NCX1 in the kidney of mice with TNBS-induced (top panels) or DSS-induced colitis (bottom panels). NCX1 immunoreactivity is depicted in green; Na/Cl transporter NCCT (encoded by SLC12A3 gene) was used as a marked of DCT and is depicted in pink; nuclei were counterstained with DAPI, blue. Magnification ϫ400. (17,31,32), and showed that mouse kidney Ncx1 transcript variant starts with exon B (variant 1.3). 5Ј-RACE analysis also showed a splicing region in between exon B and exon C, which is separated by a 21.5 kb intronic region. The ATG translation start site in exon C was identified 29-bp downstream of the 5Ј splicing site.
IFN␥ Inhibits Renal Ncx1 Promoter Activity-Based on the available public databases and analysis of 5Ј-RACE products, we had identified the TSS of Ncx1 at position Ϫ375 nt relative to the start codon. Therefore, we cloned genomic fragments extending from positions Ϫ110, Ϫ269, Ϫ379, Ϫ703 to ϩ1 nt relative to TSS into a pGL3-basic luciferase reporter vector, and we examined transcriptional activity following transient transfection into mIMCD3 cells. As shown in Fig. 5B, the transfected Ϫ269, Ϫ379, Ϫ703 nt Ncx1 promoter constructs demonstrated basal activity more than 15-fold higher than baseline (promoterless pGL3 basic); the Ϫ110 nt promoter was insufficient to drive luciferase expression in mIMCD3 cells. Similar to the endogenous Ncx1 gene in mIMCD3 cells, mIFN␥ was most potent among the three tested cytokines in inhibiting the activ-ity of the longest cloned promoter fragment (Ϫ703 nt; Fig. 5B). In response to mIFN␥, Ϫ269 nt and Ϫ379 nt constructs did not show any change in the promoter activity, whereas Ϫ703 nt promoter was significantly inhibited (Fig. 5B), thus suggesting that an IFN␥-dependent repressive cis-element was positioned between Ϫ380 and Ϫ703 nt relative to the TSS. Interestingly, alignment of the previously characterized rat renal Ncx1 gene promoter showed high homology to mouse promoter and showed similar response to IFN␥ in transiently transfected mIMCD3 cells (data not shown). Further analysis of the Ϫ380/ Ϫ703 nt promoter region showed a putative STAT1 response element present at position Ϫ387/Ϫ397 nt (TTCCTGGGAA), which bears a strong similarity to the ␥ interferon-activated site (GAS, 5Ј-TTCN(2-4)GAA-3Ј) known to bind STAT1 homodimers in response to IFN␥ stimulation (33). This site was conserved in the rat renal Ncx1 gene promoter. To further identify the requirement for this sequence, we eliminated this putative GAS element via site-directed mutagenesis. Contrary to the wild-type Ϫ703 nt promoter, the same construct with mutated GAS showed no decrease in response in transiently transfected and mIFN␥-treated mIMCD3 cells (Fig. 5C), thus demonstrating the requirement of an intact GAS site at Ϫ387/ Ϫ397 nt relative to the TSS in IFN␥-driven inhibition of the mouse renal Ncx1 promoter activity.
STAT1 Binds Directly to the Ncx1 Promoter-To confirm IFN␥-inducible recruitment of STAT1 to the mouse Ncx1 promoter in vivo (in mIMCD3 cells), we performed a quantitative ChIP assay with primers designed to span the promoter region of interest and amplify 233 bp (Ϫ279 to Ϫ511 nt). To visualize the specificity of amplification, Fig. 5D (left panel) depicts a representative DNA gel with 233-bp amplicon from the no-antibody controls (negligible background), anti-STAT1 ChIP, and input (2%) controls. The visibly increased recruitment of STAT1 to the Ncx1 gene promoter was further quantified using SYBRGreen qPCR and showed at least 10-fold increase in STAT1/Ncx1 promoter association in mIFN␥-treated mIMCD3 cells (Fig. 5D, right panel). STAT1 Is Required for Down-regulation of Ncx1 by IFN␥ in Vivo-Signaling through STAT1 transcription factor is critical and specific for the biological functions of IFN␥ (34). To verify our promoter construct and ChIP data in mice, and to further implicate STAT1-mediated inhibition of Ncx1 gene expression by IFN␥, we utilized Stat1 Ϫ/Ϫ mice. Wild-type (WT) and Stat1 Ϫ/Ϫ mice on the same genetic background (129S6/SvEv-Stat1Ͻtm1RdsϾ) were injected with PBS or recombinant mIFN␥ (50,000 units/mouse) intraperitoneal daily for 3 days. On the fourth day, kidneys were collected and analyzed for pTyr 701 -STAT1, total STAT1, and NCX1 protein and mRNA expression. Consistent with the proposed key role of STAT1 in IFN␥-mediated inhibition of Ncx1 gene transcription, no change in renal Ncx1 mRNA and protein expression was observed in IFN␥-injected Stat1 Ϫ/Ϫ mice (Fig. 6A). In WT mice, exogenous IFN␥ strongly stimulated STAT1 protein expression in the kidney (Fig. 6B), suggesting that a sensitization mechanism, previously reported in macrophages (35), also exists in the kidneys. This was accompanied by strong expression of pTyr 701 -STAT1 and, as anticipated, a significant decrease in NCX1 protein expression (Fig. 6B).

DISCUSSION
IBD patients are at higher risk of developing metabolic bone disease and low bone mineral density (BMD) than healthy subjects. Patients with inflammatory bowel disease have a more than 40% greater incidence rate of bone fracture than that of the general population (36 -38). The prevalence of osteopenia and osteoporosis in IBD ranges from 22 to 77% and from 17 to 41%, respectively (37,39,40). The etiology of low bone mass in IBD patients is likely multifactorial, including corticosteroid use, decreased physical activity and muscle strength, poor nutritional status, and malabsorption (particularly affecting vitamin D, vitamin K, and Ca 2ϩ homeostasis), as well as inflammationassociated inflammatory soluble mediators. The latter, such as IL-1␤, IL-6, TNF, affect the bone directly to alter the modeling and remodeling processes (3).
Bone is highly dynamic and it undergoes constant remodeling throughout life. The remodeling involves coupled resorption of existing bone and the formation of new bone. The supply of calcium and inorganic phosphate available for bone formation is a result of highly regulated and complex homeostatic mechanism involving paracellular and transcellular intestinal and renal (re)absorption. In the intestine, diminished Ca 2ϩ supply is believed to be the result of vitamin D 3 insufficiency, avoidance of milk and dairy food consumption and relatively common lactose intolerance in IBD patients (41), and the direct effect of inflammatory cytokines on the expression of the key components of transcellular Ca 2ϩ absorption, TRPV6, calbindin D 9K , and calcium pump PMCA1b (14). The latter report also described decreased expression of renal calbindin-D 28K and NCX1, although the consequences or mechanisms of this phenomenon were not described.
Our recent work has demonstrated that, consistent with the clinical data (5), experimental murine colitis is also associated with increased urinary fractional Ca 2ϩ excretion and decreased bone mineral density (6). We have postulated that IBD-associated TNF and IFN␥ lead to transcriptional inhibition of renal  's t test). B, representative Western blot showing NCX1, pTyr 701 -STAT1 and GAPDH (as loading control) in mIFN␥-treated mIMCD3 cells. C, Ncx1 mRNA expression of in the kidneys of mIFN␥ injected P TRPV5 -EGFP transgenic mice was quantified by real-time RT-PCR as described for panel A. D, representative Western blot analysis of NCX1, pTyr 701 -STAT1, GFP, and GAPDH proteins in renal lysates of PBS-and mIFN␥-injected P TRPV5 -EGFP transgenic mice. E, NCX1-mediated Ca 2ϩ uptake in GFP ϩ epithelial cells of the renal DCT isolated from mIFN␥-treated P TRPV5 -EGFP transgenic mice. PBS or mIFN␥ were injected intraperitoneal daily for 3 days in P TRPV5 -EGFP transgenic mice, which express GFP expression under the control of the TRPV5 promoter (C57Bl/6 background). GFP ϩ cells were flow-sorted using FACAria and used for the NCX1 activity assay as described under "Experimental Procedures." Data from three independent experiments are presented as Na ϩ -dependent Ca 2ϩ uptake relative to PBS-injected mice (*, p Ͻ 0.05, Student's t test). F, fractional urinary Ca 2ϩ excretion (FECa 2ϩ ) in IFN␥-injected mice. Mean Ϯ S.D., *, p Ͻ 0.05; Student's t test. ϪTAP refers to negative control lacking Tobacco Acid Pyrophosphatase (TAP) used for removing the capping nucleoside. B, luciferase reporter gene activity with mouse kidney specific Ncx1 promoter constructs: pGL3-Basic, pGL3-110, pGL3-269, pGL3-379, and pGL3-703 (promoter length relative to the identified transcription start site) were transiently transfected in to mIMCD3 cells treated with PBS or 100 units/ml mIFN␥ for 24 h. Luciferase activity (RLU) was normalized to mg protein and expressed relative to the activity of the promoterless PGL3-basic (pGL3) vector in PBS-treated cells (mean Ϯ S.E., *, p Ͻ 0.05 PBS versus IFN␥, Student's t test; n ϭ 3). C, effects of IFN␥ on the pGL-703 WT promoter, or construct with mutated core GAS element. Data calculated as above and normalized to the activity of pGL-703 promoter in PBS-treated cells (mean Ϯ S.E., *, p Ͻ 0.05 PBS versus IFN␥, Student's t test; n ϭ 3). D, inducible STAT1 recruitment to Ncx1 gene promoter in IFN␥-treated mIMCD3 cells. ChIP assay was performed using a control rabbit IgG or an anti-STAT1 antibody as described under "Experimental Procedures." Representative DNA gel and the summary of real-time PCR analysis are depicted in left and right panels, respectively. ϪAb indicates control IgG, ϩAb indicates anti-STAT1 antibody. Bar graph (mean Ϯ S.E.) indicates an over 10-fold difference in STAT1 recruitment (*, p Ͻ 0.05 PBS versus IFN␥, Student's t test; n ϭ 3). expression of Klotho (10), which leads to hyper-sialylation of the apical TRPV5 channel, its endocytosis, UBR4-mediated ubiquitination, and primarily proteasomal degradation (6), thus providing a compelling and comprehensive mechanism of impaired apical renal Ca 2ϩ transport in IBD. Decreased expression of NCX1, a Na ϩ /Ca 2ϩ exchanger responsible for the basolateral Ca 2ϩ extrusion in the renal DCT, as reported by Huybers et al. (14) in TNF ⌬ARE mice, would add another mechanism further contributing to the diminished transcellular Ca 2ϩ reabsorption in the kidney during IBD. Such a decrease would likely be independent of potentially altered vitamin D 3 metabolism, as renal Ncx1 gene expression is not affected by dietary Ca 2ϩ or vitamin D 3 supply (42,43).
TNF ⌬ARE mice are considered a valuable model of Crohn'slike ileitis. They bear an endogenous 69 bp deletion of the 3Ј-AUrich region in the gene encoding TNF and, as a result, have high circulating levels of TNF protein. However, the development of intestinal inflammation in this model depends on Th1-type cytokines, such as IL-12 and IFN␥, and TNF ⌬ARE mice on IFN␥ Ϫ/Ϫ background develop attenuated inflammation (44). Utilizing two alternative models of chemically induced colitis associated with significantly increased colonic IFN␥ production, we showed a similar decrease in renal expression of Ncx1 mRNA and protein.
While in vivo neutralization of IFN␥ could not be used due to the anticipated confounding effect of reduced colitis, our in vitro studies with mIMCD3 cells showed that IFN␥ was more potent than TNF or IL-1␤ and sufficient to reduce the Ncx1 expression to the same extent as cytomix (all three cytokines combined). IFN␥ is abundantly produced by cultured colonic mucosal tissue and intestinal lamina propria mononuclear cells obtained from IBD patients (45,46). Even in remission, CD8 ϩ lymphocytes from IBD patients show an increased capacity for IFN␥ induction (47). However, circulating levels of IFN␥ are not associated with the disease activity or serum C-reactive protein levels and cannot be reliably measured due to a very short half-life of the cytokine, its rapid binding to heparan sulfate, and its quick degradation (26). In our studies, experimental colitis was associated only with a minor trend toward increased serum or renal tisue IFN␥ concentrations. FIGURE 6. STAT1 is required to down-regulate the expression of NCX1 by IFN␥. WT and Stat1 Ϫ/Ϫ mice (129S6/SvEv background) were injected with PBS or mIFN␥ through i.p for 3 days and kidneys were collected for RNA and protein extraction. A, real-time RT-PCR analysis. Ncx1 expression was normalized to that of TBP relative to a calibrator (normalized Ct value obtained from untreated WT or Stat1 Ϫ/Ϫ mice) and expressed as 2 Ϫ⌬⌬Ct ; (*, p Ͻ 0.05 treated versus control, Student's t test; n ϭ 5). B, representative Western blot analysis of NCX1, pTyr 701 -STAT1, and GAPDH (loading control) in renal lysates of PBS and mIFN␥ injected WT and Stat1 Ϫ/Ϫ mice.
Moreover, in mice repeatedly injected with the recombinant cytokine, we were also unable to show an increased level of serum IFN␥. Despite this, we saw a dramatic increase in renal Stat1 and pTyr 701 -STAT1 expression in the kidney (Fig. 6B), thus cofirming that circulating IFN␥ reaches and exerts its signaling effects in the kidney. Most importantly, colitis or recombinant IFN␥ injection resulted in decreased expression of NCX1 in the renal distal convoluted tubules, as demonstrated by Western blotting, qPCR, and immunofluorescence. We also confirmed decreased NCX1-mediated Na ϩ -dependent Ca 2ϩ uptake in DCT epithelial cells isolated from healthy or colitic mice.
To address the molecular mechanism of transcriptional repression of Ncx1 by IFN␥ in mice, we characterized the primary mouse renal transcript of Ncx1 as Ncx1.3, identified the transcription start site, cloned the kidney-specific mouse Ncx1 gene promoter, and identified the GAS element in the proximal promoter. The effects of IFN␥ on Ncx1 promoter activity in vitro and gene transcription in vivo were dependent on the presence and inducible recruitment of Stat1 to the GAS element at position Ϫ387/Ϫ397 nt, a site also identified in silico in the previously characterized rat renal Ncx1 gene promoter. A human renal-specific NCX1 promoter has not been characterized at this time. Since IFN␥ can activate additional signaling pathways and can regulate gene expression by STAT1-independent pathways (48), the importance of STAT1 in mediating the inhibitory effects of IFN␥ was further demonstrated in vivo, whereby IFN␥ did not affect renal Ncx1 expression in Stat1 Ϫ/Ϫ mice .
The most frequent biological effect of the stimulation of the IFN␥/STAT1 pathway is transcriptional activation of a vast network of IFN␥-inducible genes, while transcriptional repression is a much less common phenomenon. Interestingly, and most likely related to the alternative promoter use and the transcriptional context, IFN␥ may repress, as in the case of renal Ncx1 gene, or activate the expression of Ncx1, as described for the brain-specific promoter in microglia by Nagano et al. (49). These discrepancies will have to be addressed mechanistically in the future, ideally in the context of human NCX1 gene regulation.
In conclusion, our findings showed that renal NCX1 expression and function is down-regulated over the course of experimental colitis, and pointed to a dominant role for IFN␥, which inhibits Ncx1 gene transcription via a STAT1-dependent mechanism. Since NCX1 is one of the key transporters responsible for renal Ca 2ϩ reabsorption and systemic Ca 2ϩ homeostasis, our observations provide a new mechanism contributing to the reduced renal Ca 2ϩ reabsorption in IBD, which together with reduced expression of KLOTHO (10) and TRPV5 (6), as well as the postulated reduced intestinal Ca 2ϩ absorption (14), likely contributes to a negative systemic Ca 2ϩ balance and increased bone resorption in IBD patients.