Characterization of Short Range DNA Looping in Endotoxin-mediated Transcription of the Murine Inducible Nitric-oxide Synthase (iNOS) Gene*

The local structural properties and spatial conformations of chromosomes are intimately associated with gene expression. The spatial associations of critical genomic elements in inducible nitric-oxide synthase (iNOS) transcription have not been previously examined. In this regard, the murine iNOS promoter contains 2 NF-κB binding sites (nt –86 and nt –972) that are essential for maximal transactivation of iNOS by LPS. Although AP-1 is commonly listed as an essential transcription factor for LPS-mediated iNOS transactivation, the relationship between AP-1 and NF-κB in this setting is not well studied. In this study using a model of LPS-stimulated ANA-1 murine macrophages, we demonstrate that short range DNA looping occurs at the iNOS promoter. This looping requires the presence of AP-1, c-Jun, NF-κB p65, and p300-associated acetyltransferase activity. The distal AP-1 binding site interacts via p300 with the proximal NF-κB binding site to create this DNA loop to participate in iNOS transcription. Other geographically distant AP-1 and NF-κB sites are certainly occupied, but selected sites are critical for iNOS transcription and the formation of the c-Jun, p65, and p300 transcriptional complex. In this “simplified” model of murine iNOS promoter, numerous transcription factors recognize and bind to various response elements, but these locales do not equally contribute to iNOS gene transcription.

The nucleotide sequences of inducible nitric-oxide synthase (iNOS) 2 promoters of different species exhibit homologies to binding sites for numerous transcription factors known to be involved in the endotoxin (LPS) and/or cytokine-mediated induction of transcription, such as activating protein-1 (AP-1), CCAAT-enhancer box-binding protein (C/EBP), cAMP-responsive element-binding protein (CREB), interferon regulatory factor-1 (IRF-1), NF-B, and STAT-1␣ (1-7) NF-B is generally accepted to be the critical transcription factor for LPS-mediated iNOS expression, while Stat1 plays a similar role in interferon-␥-mediated iNOS expression. In contrast, the potential role of AP-1 and the interplay among these various transcription factors is not clear. One explanation that may draw together these disparate but parallel observations is that transcription factors mediate their effects via recruitment of coactivators, which may remodel local chromatin structure, form a preinitiation complex and initiate transcription (8). The transcriptional coactivator p300 interacts with several transcription factors involved in iNOS expression, such as NF-B and AP-1. p300 coactivates transcription via its histone acetyltransferase activity, mediates interactions with the basal transcription machinery, and is a key scaffolding protein in the formation of enhanceosomes (9). p300 has been shown to be essential for transactivation of the iNOS promoter (10).
In this regard, the murine iNOS promoter contains 2 NF-B binding sites (nt Ϫ86 and nt Ϫ972) that are essential for maximal transactivation of iNOS by LPS (1). Although AP-1 is commonly listed as an essential transcription factor for LPS-mediated iNOS transactivation, the relationship between AP-1 and NF-B in this setting are not well studied. In addition, the potential epigenetic relationships between these two entities have not been previously described. In this study, using LPSstimulated ANA-1 murine macrophages, we examine the interplay of p300, AP-1, and NF-B in short range DNA looping of the iNOS promoter. Our results indicate that p300 bridges AP-1, c-Jun (nt Ϫ1069), and NF-B p65 (nt Ϫ86) to form a DNA loop to initiate iNOS gene transcription. The spatial associations of critical genomic elements in iNOS transcription have not been previously examined. This reductionist model using the murine iNOS promoter indicates that while numerous transcription factors recognize and bind to various response elements, they do not equally contribute to gene transcription. Rather, a hierarchy exists among these sites that is determined by the activating factor, requirement for spatial association, and interaction among the various transcription factors and coactivators.

EXPERIMENTAL PROCEDURES
Materials-ANA-1 macrophages were a gift from Dr. George Cox (Uniformed Services University of the Health Sciences, Bethesda, MD). Wild-type RelA and its mutant expression vectors were gifts from Dr. Jeremy M. Boss (Emory University of Medicine, Atlanta, Georgia). The expression vectors containing full-length p300 and its acetyltransferase (AT) deletion derivative (m-p300) were constructed by inserting a XhoI-NotI fragment carrying the N-terminal Flag-tagged p300 open reading frame or its acetyltransferase deletion derivative (⌬ nt 1472-1522) into pCl vector (Promega, Madison, WI) and were provided by Dr. Joan Boyes (Institute of Cancer Research, London, UK). E1A plasmid was received from Dr. Debabrata Chakravarti, University of Pennsylvania, School of Medicine.
Assay of NO Production-NO released from cells in culture was quantified by measurement of the NO metabolite, nitrite. After stimulation, 50 l of culture medium was mixed with 50 l of 1% sulfanilamide in 0.5 N HCl. After a 5-min incubation at room temperature, an equal volume of 0.02% N-(1-naphthyl)ethylenediamine was added. Following incubation for 10 min at room temperature, the absorbance of samples at 540 nm was compared with that of NaNO 2 standard on a MAXLINE microplate reader.
Immunoprecipitation Studies-Nuclear protein from ANA-1 cells was incubated with anti-p65 or anti c-Jun Ab (R&D Systems, Minneapolis, MN) and protein G-agarose in Co-IP buffer (10 mmol/liter Tris-HCl, pH 7.5, 3 mmol/liter EGTA, 20 mmol/ liter NaCl, 0.02% Triton X-100, 1ϫ protease inhibitors mixture, 0.2 mmol/liter dithiothreitol, 1 mmol/liter phenylmethylsulfonyl fluoride at 4°C for 4 h. Protein G-agarose beads were collected by centrifugation and washed three times with the Co-IP buffer. The immune complexes were resuspended with SDS sample buffer and then loaded onto 4 -20% SDS-PAGE gel, separated, and electrotransferred to polyvinylidene difluoride membranes for Western blot analysis. The membrane was blocked with 5% skim milk in PBS-0.05% Tween for 1 h at room temperature. After washing three times, blocked membranes were incubated with anti-p65 Ab, anti-c-Jun Ab or anti-p300 Ab (Upstate Biotechnology, Waltham, MA), washed, and incubated with horseradish peroxidase-conjugated anti-rabbit-IgG (Santa Cruz Biotechnology, Santa Cruz, CA). Bound peroxidase activity was detected by the West Pico chemiluminescent kit (Pierce).
ChIP Assay and ChIP Quantitative PCR-ANA-1 cells were grown in 10-cm dishes and DNA-protein cross-linked by addition of 1% formaldehyde at room temperature for 10 min. Assays were performed using the ChIP assay kit (Upstate Biotechnology, Waltham, MA) following the manufacturer's instructions. Anti-p65 Ab, anti-p300 Ab, and anti-c-Jun Ab were used for each immunoprecipitation. The DNA was recovered and subjected to analysis by PCR. The primers for the distal NF-B binding site had the following sequence: 5Ј-acacgaggctgagctgacttt-3Ј (sense strand) and 5Ј-tgggctagcctggtctacag-3Ј (antisense strand) to yield a PCR product of 277 bp, and the sequences of primers for the proximal NF-B binding site were 5Ј-cctagtgagtcccagttttgaagt-3 (sense strand) and 5Ј-catcaggtatttatacccctccag-3Ј (antisense strand); their product was 273 bp. The PCR program was 94°C for 5 min, followed by 94°C for 30 s, 55°C for 30 s, and 72°C for 40 s for a total of 30 cycles, and then 72°C for 10 min. The amplified DNA was visu-alized by electrophoresis on 1% agarose gel in 1ϫ TAE (Tris acetate/EDTA) buffer after staining with ethidium bromide. For ChIP quantitative PCR analysis, anti-p65, anti-RNA Pol2, or anti-c-Jun (Santa Cruz Biotechnology) were used for each immunoprecipitation, and normal rabbit IgG was used as endogenous control. The specific PCR primers for the iNOS promoter gene were designed using Primer 3 and mfold software. Forward primer: 5Ј-taaaaaggcttcactcagcaca-3Ј and reverse primer: 5Ј-agtgaggttagatggtgccaat-3Ј were for the middle AP-1 binding site. Forward primer: 5Ј-tcccagttttgaagtgactacg-3Ј and reverse primer: 5Ј-cataactgttcccaaagggaga-3Ј were for distal NF-B and AP-1 binding sites, forward primer: 5Ј-tattgaggccacacactttttg-3Ј and 5Ј-attgacagtgttaggggaaaagg-3Ј were for proximal NF-B and AP-1 binding sites. Real-time PCR was performed with iQ SYBR Green super mix (Bio-Rad), using iCycler iQ Real-time PCR Detection System (Bio-Rad) with the following amplification parameters: denaturation at 95°C for 3 min, followed by 40 cycles of 95°C for 30 s and 55°C for 35 s. Quantification was normalized using endogenous control and calculated using the 2 Ϫ⌬⌬Ct method (11).
Chromosome Conformation Capture (3C) Assay-ANA-1 cells (1 ϫ 10 7 ) were cultured in 10-cm dishes and fixed with 1% formaldehyde for 10 min at room temperature. The reaction was quenched by the addition of 0.125 M glycine for 5 min at room temperature and 10 min at 4°C. The cells were washed with phosphate-buffered saline and lysed in cold lysis buffer (10 mM Tris-HCI, pH 8.0, 10 mM sodium chloride, 0.2% Igepal, 0.8 M aprotinin, 50 M bestatin, 20 M leupeptin, 10 M pepstatin A, 25 M p-bromotetramisole oxalate, 5 M cantharidin, and 5 nM microcystin-LR (microcystin-leucine arginine)). The nuclei were harvested and suspended in digestion buffer with 0.1% SDS and 1% Triton X-100. The DNA was digested with restriction enzymes SspI and/or RsaI overnight at 37°C. Samples were then diluted with ligase buffer with 0.1% SDS and 1% Triton X-100. T4 DNA ligase was added with incubation at 16°C overnight, followed by overnight incubation at 65°C in the presence of 10 g/ml proteinase K to reverse the cross-links. The DNA was isolated by phenol-chloroform extraction and ethanol precipitation. The purified DNA concentration was determined and was used as a PCR template with the following primer pair: 5Ј-ggggattttccctctctctg-3Ј and 5Ј-ccagttgggtgtgcaagtta-3Ј. PCR was performed in 25-l reaction mixtures with an initial denaturing step for 4 min at 94°C and then 30 cycles including 30 s at 94°C, 30 s at 55°C and 30 s at 72°C. The PCR products were run on 1% agarose gels. 3C analyses were performed in three independent experiments. PCR products were cloned and sequenced to confirm the presence of the predicted junction in the iNOS DNA SspI site.
ChIP Loop Assay-ANA-1 cells were cross-linked for 10 min with 1% formaldehyde at room temperature. After cell lysis, chromatin was sheared and digested by the restriction enzyme SspI. Chromatin fragments were immunoprecipitated using normal rabbit IgG, anti-p65 Ab, anti-c-Jun Ab, or anti-p300 Ab with protein G-agarose beads at 4°C overnight, followed by overnight ligation with T4 ligase at 16°C. After reverse crosslinking with proteinase K digestion at 65°C overnight, the DNA was isolated by phenol-chloroform extraction and ethanol pre-iNOS Promoter DNA Looping cipitation. PCR and real-time PCR were performed as described above.
Immunoprecipitation of Acetylated Proteins-ANA-1 cells were transfected with p300 wild-type or m-p300 expression vectors using JetPEI-Macrophage reagent (Genesee Sci, San Diego, CA), and incubated at 37°C, 5% CO 2 for 24 h. The cells were incubated in Dulbecco's modified Eagle's medium (Invitrogen, Carlsbad, CA) containing 50 ng/ml LPS for 30 min before the cells were collected for nuclear protein. The nuclear extract was immunoprecipitated overnight at 4°C with acetylated-lysine antibody (Cell Signaling, Danvers, MA) crosslinked to agarose beads. After washing the beads with cold TBS (50 mM Tris-HCl, pH 7.5, 150 mM NaCl), protein was eluted from the beads by heating for 10 min at 95°C, followed by SDS-PAGE (4 -20% polyacrylamide) and fluorography. Western blot analysis was performed using Ab directed against NF-B p65, AP-1, c-Jun, or p300 (Upstate Inc, Billerica, MA).
Statistical Analysis-All data are presented as mean Ϯ S.D. Analysis was performed using a Student's t test. Values of p Ͻ 0.05 were considered significant.

AP-1 and NF-B Promote LPS-mediated iNOS Gene
Transcription-The human iNOS promoter contains five putative NF-B binding sites upstream of Ϫ4.7 kb that contribute to human iNOS expression (12). In contrast, the murine iNOS promoter contains only two requisite NF-B binding sites and thus, was chosen as a simpler model for investigation of DNA looping that may occur among the various NF-B and AP-1 sites (1). The relevant AP-1 and NF-B response elements in the murine iNOS promoter are depicted in Fig. 1a. To determine the role of AP-1 and NF-B in iNOS transcription, ChIPreal time PCR was performed to determine LPS-mediated iNOS transcriptional activity at the mouse iNOS promoter (GenBank TM accession no. L09126) after 2 h of incubation. Purified chromatin was immunoprecipitated using anti-RNA polymerase II (RNA Pol II) or as a control, rabbit immune serum. Murine iNOS promoter DNA coimmunoprecipitating with RNA Pol II was then quantified by real-time PCR using primers specific for the promoter region inclusive of the TATA start site (Fig. 1b). In selected instances, siRNA to AP-1, c-Jun, and/or NF-B p65 was administered. Our results show that LPS (50 ng/ml) treatment increased iNOS transcription activity. When c-Jun-siRNA or p65-siRNA was added 24 h prior to administration of LPS, the extent of iNOS transcription was decreased by 50 and 75%, respectively (p Ͻ 0.01 versus LPS alone for c-Jun-siRNA or p65-siRNA). In the presence of both c-Jun-siRNA and p65-siRNA, iNOS transcription was decreased by 95% (p Ͻ 0.01 versus LPS alone). Mismatch siRNA control did not significantly alter LPS-mediated iNOS transcription. In a parallel series of studies, NO expression following 6 h of LPS stimulation was determined by the extent of production of the NO metabolite, nitrite (Fig. 1c). Similar to that noted with iNOS gene transcription, NO production was increased in the presence of LPS. When c-Jun-siRNA or p65-siRNA was added 24 h prior to administration of LPS, the extent of NO production was decreased by 60 and 80%, respectively (p Ͻ 0.01 versus LPS alone for c-Jun-siRNA or p65-siRNA). In the presence of both c-Jun-siRNA and p65-siRNA, NO synthesis was decreased by 97% (p Ͻ 0.01 versus LPS alone). Mismatch siRNA control did not significantly alter LPS-mediated NO production. These results indicate that AP-1, c-Jun, and NF-B p65 contribute to maximal LPS-mediated iNOS transcription and NO synthesis.
Quantitative ChIP PCR assays were performed to determine relative c-Jun and p65 binding to their respective response elements following LPS stimulation. The AP-1 sites at nt Ϫ1069, nt Ϫ488, and nt Ϫ231 were labeled as distal AP-1 (d-AP-1), middle AP-1 (m-AP-1), and proximal AP-1 (p-AP-1), respectively; the NF-B sites at nt Ϫ972 and nt Ϫ86 were labeled as distal NF-B (d-NF-B) and proximal NF-B (p-NF-B), respectively. As determined by 2 Ϫ⌬⌬Ct , the relative distribution of AP-1 c-Jun binding among the distal, middle, and proximal sites following LPS stimulation was 217:1:7, respectively. The ratio of NF-B binding between the proximal and distal sites was 415:1. In total, our results suggest that LPS-mediated iNOS gene transcription requires both AP-1 and NF-B binding; these are skewed toward the d-AP-1 and p-NF-B sites.
AP-1 and NF-B Mediate DNA Looping of the iNOS Promoter via p300-To determine the extent of potential DNA looping, 3C assays were first performed using the relevant restriction enzyme, SspI (Fig. 2). If DNA looping occurred, the selected PCR primers would generate a 344-bp oligonucleotide following treatment with Ssp1 and T4 ligase. In Fig. 2a, ANA-1 macrophages were treated with LPS (50 ng/ml) for 10, 30, and 60 min, and 3C assays performed. The presence of a band at ϳ300 bp indicates the presence of DNA looping. These bands were sequenced and found to be identical to the region of interest. All treatment conditions produced a 1000-bp band of varying intensities corresponding to the PCR product predicted in the absence of Ssp1 treatment (data not shown). Controls included 3C assays performed in the absence of ligase, and a product representing an unaffected portion of the iNOS gene. As TNF, interferon ␥ (IFN), and interleukin-1␤ (IL-1) are known activating agents for iNOS transcription, we examined the relative amount of 3C product generated in response to these various agents (Fig. 2b). 3C assays were performed in ANA-1 cells that were exposed to LPS (50 ng/ml), IFN (200 units/ml), TNF (2 ng/ml), or IL-1 (1000 units/ml) for periods of 10, 30, and 90 min. The resulting gels demonstrate that DNA looping was detected in LPS-and IL-1-treated cells as early as 10 min, peaking at 30 min, and then significantly diminished by 90 min.
To determine the role of AP-1, c-Jun, NF-B p65, and p300 in DNA looping at the iNOS promoter, loss-of-function studies were performed using siRNA to c-Jun, p65, and/or p300. Additional negative controls included transient transfection of S276A, a p65 mutant in which serine 276 in the Rel homology domain has been mutated to an alanine, E1A, a direct inhibitor of p300, and/or m-p300, a p300 containing an acetyltransferase (AT) deletion. Western blot analysis confirmed that siRNAs directed against c-Jun, p65, and p300 effectively ablated expression of the respective proteins in LPS-stimulated cells (data not shown). 3C assays were performed (Fig. 2c). In each instance, ablation of c-Jun, p65, or p300 resulted in loss of DNA looping in the iNOS promoter of LPS-stimulated ANA-1 macrophages. Mutation of the AT domain in p300 resulted in the loss of DNA iNOS Promoter DNA Looping looping. These data suggest that AP-1, c-Jun, NF-B p65, and p300 with intact AT activity are all required for DNA looping and subsequent iNOS expression in LPS-stimulated ANA-1 macrophages.
We performed IP assays to determine whether p65, c-Jun, or p300 were potential targets of p300 AT activity in LPS-stimulated ANA-1 cells (Fig. 2d). Nuclear proteins were immunoprecipitated using Ab directed against acetylated lysine targets; immunoblots were performed thereafter. ANA-1 cells treated with LPS in the presence of m-p300 demonstrated near ablation of NF-B p65 acetylation. Neither p300 nor c-Jun was found to be acetylated. These results indicate that NF-B p65 acetylation may be required for formation of the p-65/p300/c-Jun complex and DNA looping found in the iNOS promoter.

AP-1, c-Jun, and NF-B p65
Interact with p300-The transcriptional coactivator, p300, has been demonstrated to be essential for iNOS expression. It interacts with NF-B and AP-1 and is a key scaffolding protein in the formation of enhanceosomes (9). Examination of p300 indicates the presence of individual binding sites for NF-B p65 at the CREB-binding domain and AP-1, c-Jun at the N-terminal zinc finger domain; this would suggest that p300 does not interact with two NF-B (or two AP-1) proteins simultaneously (9). In this regard, we examined the potential interactions between c-Jun, p65, and p300, using IP-Western blot studies. (Fig. 3a) Our IP-Western blot studies of nuclear protein demonstrate that a c-Jun, p65, and p300 protein complex exists during LPS stimulation of ANA-1 macrophages. We then performed ChIP-Loop Q PCR FIGURE 2. a, 3C assay of murine iNOS promoter using SspI. ANA-1 macrophages were treated with LPS (50 ng/ml) for 10, 30, and 60 min. The presence of a band at ϳ300 bp indicates the presence of DNA looping. These bands were sequenced and found to be identical to the region of interest. Controls include assays performed in the absence of ligase and a product representing an unaffected portion of the iNOS gene. The gel is representative of three experiments. b, 3C assay using SspI. ANA-1 cells were exposed to LPS (50 ng/ml), IFN (200 units/ml), TNF (2 ng/ml), or IL-1 (1000 units/ml) for periods of 10, 30, and 90 min. The gel is representative of three experiments. c, 3C assay using SspI. Loss-of-function studies were performed in LPS-stimulated ANA-1 cells using siRNA to c-Jun, p65, and/or p300, S276A, a p65 mutant in which serine 276 in the rel homology domain has been mutated to an alanine, E1A, a direct inhibitor of p300, and/or m-p300, a p300 containing an acetyltransferase deletion. Controls included unstimulated cells, MM siRNA, and WT p300. Gel is representative of three experiments. d, immunoprecipitation assays for acetylated proteins. To determine whether p65, c-Jun, or p300 were potential targets of p300 AT activity in LPS-stimulated ANA-1 cells, nuclear proteins were immunoprecipitated using Ab directed against acetylated lysine targets; immunoblots were performed for p65, c-Jun, and p300. In selected instances, cells were transfected with m-p300, a p300 containing an acetyltransferase deletion, or WT p300. Control cells were unstimulated. The blot is representative of three experiments.
iNOS Promoter DNA Looping SEPTEMBER 12, 2008 • VOLUME 283 • NUMBER 37 for NF-B p65, AP-1, c-Jun, and p300 binding using PCR primers previously described as specific for iNOS DNA looping (Fig. 3b). Using 2 ⌬⌬Ct analysis, the relative binding of NF-B/ AP-1/p300 within the DNA loop was 2. To confirm that these relevant sites combine to form the DNA loop, we performed 3C assays using combinations of Ssp1 and Rsa1 restriction endonucleases. Primer 1 was chosen to be complementary to the distal iNOS promoter DNA between the Rsa1 and Ssp1 sites upstream from the d-NF-B site; primer 2 was at the proximal end of the iNOS promoter downstream from the p-NF-B site (Fig. 4a). If DNA looping occurred between p-AP-1 and d-NF-B sites, a 3C product would be detected; this is represented as Option 1. However, if the interaction occurred between d-AP-1 and p-NF-B sites, no 3C product would be formed as shown in Option 2. When genomic DNA from LPS-treated ANA-1 macrophages was exposed to Ssp1 and Rsa1 during the 3C assay, there were no 3C products (Fig. 4b). This result indicates that short range DNA looping in LPS-treated ANA-1 cells occurs via interactions between the d-AP-1 and p-NF-B sites.

DISCUSSION
In LPS-mediated sepsis and inflammation, pro-inflammatory cytokines are elaborated, and iNOS is systemically expressed in multiple cell types, including macrophages (13)(14)(15)(16). The sustained production of NO in high concentration regulates multiple cellular and biochemical functions, including inotropic and chronotropic cardiac responses, systemic vasomotor tone, intestinal epithelial permeability, endothelial activation, and microvascular permeability (17)(18)(19). The cellular and biochemical consequences of NO production in the setting of sepsis and inflammation are exemplified by DNA singlestrand breakage, ADP-ribosylation of nuclear proteins, nitrosative stress-mediated alteration in gene transcription, and inhibition of mitochondrial respiration. In experimental endotoxemia, NO production results in cytotoxicity, inhibition of mitochondrial respiration, platelet aggregation, neutrophil adhesion, hypotension, and vasoplegia (13, 20 -25). In systemic inflammation induced by LPS, the macrophage is responsible for the majority of the circulating NO metabolites. Macrophage iNOS expression is central to many of the systemic effects associated with LPS stimulation. However, while the molecular pathways which regulate iNOS transcription have been extensively studied in multiple cell types, including the macrophage, little is known of epigenetic pathways in which secondary and tertiary structural issues are considered.
In this study using a model of LPS-stimulated ANA-1 murine macrophages, we demonstrate that short range DNA looping occurs at the iNOS promoter. This looping requires the presence of AP-1, c-Jun, NF-B p65, and p300-associated acetyltransferase activity. The distal AP-1 binding site interacts via p300 with the proximal NF-B binding site to create this DNA loop to participate in iNOS transcription. Although AP-1, NF-B, and p300 are commonly listed as essential transcription factors and cofactors for LPS-mediated iNOS transactivation, the potential epigenetic relationship among these three entities in this setting has not been well studied. Other geographically distant AP-1 and NF-B sites are certainly occupied, but selected sites are critical for iNOS transcription and the formation of the c-Jun, p65, and p300 transcriptional complex. This multiprotein complex is not unique and has been previously described by Brockmann et al. (26) in their description of an AP-1, p300, NF-B complex that is critical to expression of major histocompatibility complex genes. Interestingly, although not addressed by these authors, they describe the crit- iNOS Promoter DNA Looping SEPTEMBER 12, 2008 • VOLUME 283 • NUMBER 37 ical need for two geographically separated NF-B and AP-1 binding sites for maximal cytokine inducible function of their enhancer. In a manner similar to that seen with our iNOS data, DNA looping could certainly be relevant to their observations. Given that transcription supersedes the linear order of elements to emphasize the critical nature of secondary and tertiary structure, our results confirm the importance of DNA loop formation for LPS-mediated iNOS expression. These observations have not been previously described for iNOS transcription.
DNA looping refers to a conformation of DNA in which cis elements of a DNA strand are physically in proximity to one another to bring a locally high concentration of transcription factors, cofactors, and chromatin-modifying factors near the transcriptional start site of genes and activate transcription (27). This mechanism has been used to explain why enhancer elements located hundreds to thousands of base pairs away from the transcriptional start site can contribute to gene activation. Using the 3C assay, looping of this nature has been demonstrated to provide proximity between the ␤-globin locus control region and the ␤-major globin promoter, which are located ϳ50 kilobase pairs apart (28). However, in a setting similar to ours, Babu et al. (29) study of insulin gene regulation indicates that looping can occur over shorter spans of hundreds of base pairs in relatively nucleosome-free segments of DNA. In this context, our results indicate that p300 acetyltransferase activity is critical for preferential targeting of distinct transcription factors at specific locations along the iNOS promoter to form a transcriptional protein complex. Beyond this, we are left with quandaries regarding studies demonstrating iNOS promoter binding sites for numerous transcription factors, such as AP-1, CCAAT-enhancer box-binding protein (C/EBP), cAMP-responsive element-binding protein (CREB), interferon regulatory factor-1 (IRF-1), NF-B, and STAT-1␣ (15). If these cisacting elements are truly critical for iNOS expression, how do they contribute to secondary and tertiary structural regulation of iNOS expression? While there will be activator-mediated specificity, the overall picture is extremely complex.
With respect to iNOS, the contribution of chromatin remodeling and transcriptional coactivators are just beginning to be examined in the context of "straight-forward" iNOS expression. Histone acetylation and chromatin remodeling are known to regulate iNOS gene expression. Inhibition of histone deacetylase activation suppresses iNOS expression in pancreatic ␤-islets, macrophages, and intestinal epithelial cells (30,31). Using in vivo DNA footprinting/ligation-mediated PCR and DNase hypersensitivity assays, Mellott et al. (32) have demonstrated cytokine-induced changes in chromatin structure at the iNOS promoter in lung and liver cells. In RAW264.7 cells, Deng and Wu (10) have demonstrated that coactivator p300 AT activity regulates LPS-induced iNOS expression by increasing NF-B binding and transactivation. Cao et al. (8) have subsequently corroborated the role of p300 in iNOS expression using E1A, a specific adenovirus-derived inhibitor of CBP/p300. With these exceptions, there have been no other studies that address the role of epigenetics in iNOS transcription. In this study, our results demonstrate for the first time that an AP-1, c-Jun, NF-B p65, p300 protein complex requires p300 acetyltransferase activity for short range DNA looping in the setting of LPS stimulation. The target of this p300 acetyltransferase activity is currently unknown, but our data suggest that among the potential candidates are histones and NF-B p65.
Our data provide a framework for modeling the molecular events that lead to LPS-induced iNOS gene activation in ANA-1 murine macrophages. Central to this model is the formation of a DNA loop that brings into proximity a distal region at about Ϫ1000 bp near the coding region of the gene. Certainly, our data support the mechanism that ensures that iNOS remains in an "off" until the correct activation signal is received. The off-state is maintained by an inaccessible chromatin structure at the promoter region. Separation of the distal and proximal regulatory regions allows the iNOS gene to maintain two distinct chromatin structures: an accessible structure that can bind the requisite transcription factors and initiate transactivation and a closed structure to prevent the random binding of RNA polymerase or a factor that can recruit RNA polymerase. The differentiation between these two states depends upon a variety of signals including acetylation of targets such as histones and the interacting transcription factors.
The spatial associations of critical genomic elements in iNOS transcription have not been previously examined. The "simplified" model of murine iNOS promoter indicates that while numerous transcription factors recognize and may bind to various response elements, these locales do not equally contribute to gene transcription. Rather, a hierarchy exists among these sites that is determined by the activating factor, requirement for spatial association, and interaction among the various transcription factors and coactivators. Further application of this principle to the more "complex" model of the human iNOS promoter will be challenging. Functionally important NF-Blike sequences have been identified at Ϫ5.2, Ϫ5.5, Ϫ5.8, Ϫ6.1 kb, and Ϫ8.2 kb in the human iNOS promoter (12,15). In addition, inducible AP-1 binding sites have been reported at Ϫ5.1 and Ϫ5.3 kb in the hiNOS promoter. The potential exists for a multitude of DNA loops to exist in this context. Nevertheless, consideration of the relationship between regulatory elements and genes must go beyond the linear order of elements to account for secondary and tertiary considerations. Local structural properties and spatial conformations clearly play a major role in regulation of gene expression.