Role of CREB1 and NFκB-p65 in the Down-regulation of Renin Gene Expression by Tumor Necrosis Factor α*

Tumor necrosis factor-α (TNFα) is a potent inhibitor of renin gene expression in renal juxtaglomerular cells. We have found that TNFα suppresses renin transcription via transcription factor NFκB, which targets a cAMP responsive element (CRE) in the renin promoter. Here we aimed to further clarify the role of NFκB and the canonical CRE-binding proteins of the CRE-binding protein/activating transcription factor (CREB/ATF) family in the inhibition of renin gene expression by TNFα in the juxtaglomerular cell line As4.1. TNFα caused a moderate decrease in the binding of CREB1 to its cognate CRE DNA binding site. On the other hand, NFκB-p65 transcriptional activity was substantially reduced by TNFα, which targeted a trans-activation domain at the very C terminus of the p65 molecule. Our results suggest that TNFα inhibits renin gene expression by decreasing the transactivating capacity of NFκB-p65 and partially by attenuating CREB1 binding to CRE.

and CREB1 coding sequences were cloned into pcDNA3.1 V5-His TOPO vector (Invitrogen), verified by sequencing, and linearized with EcoRV. The proteins were in vitro transcribed/translated using TNT-coupled wheat germ system (Promega). Two to five microliters from a single reaction were used for DNA binding studies.
EMSA-Nuclear protein extraction and shift/supershift assays were described elsewhere (10). The sequence of the Gal4 probe is 5Ј-AGTCG-GAGTACTGTCCTCCGAGCTGC-3Ј (the Gal4 binding site is underlined). The native renin CRE in the m30 sequence (renCRE probe) of the mouse renin promoter (10) was replaced by consensus CRE (TGACGTCA) or NFB binding sequence (GGGACTTTCC) to construct the CRE and the NFB probe, respectively.
Transcription Factor Enzyme-linked Immunosorbent Assay-CREB1 binding activity was measured with the Trans-AM TM transcription factor binding kit from Active Motif using the protocol described in Todorov et al. (10).
Transient Transfection-Transfection was performed using Fu-GENE TM 6 transfection reagent (Roche Applied Science) essentially as described (5,10). The firefly luciferase vector (Luc1) was co-transfected with the pRL-SV40 plasmid (Luc2) and where indicated with the corresponding Gal4 vector or pcDNA3.1 V5-His TOPO (blank). 0.5 g of 4.2-kb construct, 0.2 g of pFR-Luc, 0.1 mg of 4.2-kb Gal, 0.6 g of Gal4 vector or blank, and 0.005 g of pRL-SV40 were used in a single transfection reaction. For preparation of nuclear extracts containing the Gal4:p65 fusion protein, 3 ϫ 10 6 As4.1 cells were split in a 75-cm 2 flask. Four micrograms of the Gal4:p65 construct were transfected on the next day. Twenty-four hours after transfection cells were split into two 25-cm 2 flasks. After overnight incubation, TNF␣ was applied (10 ng/ml for 20 h), or cells were left untreated (control). This protocol was performed to avoid differences in the transfection efficiency in the absence of an internal control.
Luciferase Assay-Luciferase activity was measured with the dual luciferase assay kit (Promega) according to the manufacturer's instructions. Relative luciferase activity was calculated as firefly luciferase to renilla luciferase ratio (Luc1/Luc2).
Statistics-Experiments were carried out in triplicate with three samples per condition unless otherwise indicated. Representative EM-SAs or Western blots were shown. Levels of significance were estimated by analysis of variance followed by Student's unpaired t test. p Ͻ 0.05 was considered significant.

CREB1 Is Involved in the Regulation of Basal Renin
Transcription-CREB/ATF transcription factors are known to be involved in the regulation of renin transcription. However, the function of the different members of the family in the control of renin gene is not exactly known. Mouse renin CRE was found to bind almost exclusively CREB1 (14). Therefore, we tried to characterize the precise role of CREB1 in the regulation of renin gene expression in the mouse JG cell line As4.1. To this end we knocked-down CREB1 gene expression by RNA interference. CREB1 was effectively down-regulated by sequence-specific siRNA (Fig. 1A). The changes in CREB1 abundance in siRNA-treated versus control cells were paralleled by decreased renin mRNA as well as by decreased activity of renin gene promoter (Fig. 1, B and C). Moreover, there was a straight correlation between renin and CREB1 mRNA abundance (Fig.  1D). These findings show that CREB1 is capable of regulating the basal transcriptional activity of renin gene. Therefore, we studied the potential role of CREB1 in the TNF␣-mediated suppression of renin gene expression. CREB1 is known to bind to DNA in a constitutive manner (24). Its trans-activation is regulated by phosphorylation on serine 133, located in the kinase-inducible domain of the CREB1 molecule (25). Treatment with TNF␣ induced transitory phosphorylation of CREB1 during the first hour of incubation (Fig. 2). This was a surprising finding because CREB1 phosphorylation stimulates, whereas TNF␣ inhibits renin transcription (10,15,16). This apparent discrepancy could be explained with the transient nature of CREB1 phosphorylation, which peaks during the first 5 min of treatment with TNF␣ (Fig. 2). Because the renin mRNA half-life in As4.1 cells is about 8 h (15, 20), the short term activation of CREB1 would not be expected to have a FIG. 1. CREB1 regulates basal renin transcription. CREB1 was knocked-down by RNA interference as described under "Materials and Methods." A, efficacy of the CREB1 knock-down. Nuclear protein extracts were probed with anti-CREB1 antibody. B, total RNA was isolated, and renin mRNA and ␤-actin mRNA (used as an internal control) were quantified. C, As4.1 cells were additionally transfected with the proximal 4.2 kb of the mouse renin promoter, and luciferase activity was measured. Luc1, firefly luciferase activity; Luc2, renilla luciferase activity. Data are the means Ϯ S.E.; *, p Ͻ 0.05. D, total RNA was isolated from As4.1 cells, and CREB1 mRNA and renin mRNA were quantified. mRNA abundance is presented in arbitrary units. Correlation analysis was performed using GraphPad Prism software; p Ͻ 0.0001, 95% confidence interval, 0.8087-0.9846. remarkable effect on renin transcription upon prolonged treatment with TNF␣ (see also Fig. 6D).
Next, we tested whether TNF␣ has an effect on the binding of CREB1 to the mouse renin CRE. We used enzyme-linked immunosorbent assay-based DNA binding assay, which is presently the most sensitive method of studying DNA-protein interactions (26). There was a small decrease (about 25%) in the binding of CREB1 to a CRE sequence in TNF␣-treated cells (Fig. 3). This partially diminished binding of CREB1 to mouse renin CRE could contribute to a certain extent to the inhibition of renin gene expression by TNF␣. However, the total amount of the protein complex bound to mouse renin CRE seemed not to be reduced after TNF␣ treatment (10), suggesting that CREB1 is partially replaced by another protein(s).
NFB was found to bind to mouse renin CRE (10). The consensus NFB binding sequence (GGGACTTTCC) is not related to CRE (TGACGTCA). It is, therefore, unclear how NFB could interact with CRE. To address this problem we studied the DNA binding properties of in vitro translated NFB-p65, which is assumed to confer the trans-activation properties of NFB (16,17) in shift assays. Both in vitro translated CREB1 and p65 bound to their corresponding consensus DNA motif (CRE or NFB binding site, respectively), confirming the efficiency of the method (Fig. 4, lanes 4 and 10). There was additional CRE binding activity in the lysate used for the in vitro translation (marked with NR (not related)), which could be clearly distinguished from the CREB1-dependent shift of the CRE probe (Fig. 4, lanes 2-6). Neither CREB1 nor p65 could bind to an unrelated sequence (NFB binding site or CRE, respectively) ( Fig. 4, lanes 5 and 11). Similarly to the consensus CRE sequence, renCRE was found to bind CREB1 but not p65 (Fig. 5A, lanes 4 and 5). Thus, p65 seems to interact with CRE indirectly. Data suggests that p65 is capable of forming heterodimers with basic-domain-leucine-zipper (bZip) transcription factors (27)(28)(29)(30), which also include the CREB/ATF proteins. Thus, CREB1 could mediate the binding of p65 to mouse renin CRE. To test the hypothesis that p65 may heterodimerize with CREB1, we added increasing amounts of p65 to CREB1 in shift assays with renCRE probe. (Fig. 5B). The total amount of in vitro translated protein in each sample was kept constant with firefly luciferase. The addition of p65 to CREB1 appeared not to increase the amount of protein bound to renCRE (Fig.  5B). On the other hand, the binding of CREB1 to CRE also seemed not to be affected upon the addition of p65 (data not shown). These results provide evidence against possible dimerization of CREB1 and p65.
NFB-p65 Plays a Major Role in the Down-regulation of Renin Transcription by TNF␣-The strong effect of TNF␣ on renin transcription could not be explained just by the marginal changes in the binding of CREB1 and p65. Because TNF␣ influenced just transiently the trans-activation potential of CREB1, we tested whether it has an effect on the transactivation potential of p65. In contrast to CREB1, NFB is known to regulate transcription basically by signal-dependentinducible DNA binding (31). Therefore, to assess the transcriptional activity of p65 independently of its binding activity, we used a fusion protein-reporter gene system (21). The fusion protein consists of the yeast transcription factor Gal4 DNA binding domain linked to the full-length p65. When co-transfected with a reporter gene driven by a minimal Gal4 binding sequence, the fusion protein should bind to this sequence in a constitutive manner through its Gal4 domain. Thus, changes in the expression of the reporter gene would reflect solely the trans-activation status of p65 but not its binding capacity. Using firefly luciferase as a reporter, we found that TNF␣ inhibits the transcriptional activity of p65 by 70% in As4.1 cells (Fig. 6A). We tested whether TNF␣ has an impact on the DNA binding activity of the Gal4 fusion protein, since this may also provide an explanation for the alterations in the expression of Gal4-dependent reporter gene. The binding of Gal4:p65 fusion protein remained unchanged after treatment with TNF␣, indicating that the down-regulation Gal4-driven reporter gene expression by TNF␣ was due principally to impaired transactivation potential of p65 (Fig. 6B).
The activity of a Gal4:cRel fusion protein was also inhibited by TNF␣ (Fig. 6C), demonstrating that the effect of TNF␣ is not restricted to a single Rel protein of the NFB family. However, cRel appeared to possess less than 10% of the basal transactivation potential of p65 (Fig. 6C), suggesting that it plays a To test if the inhibition of p65 transcriptional activity by TNF␣ may directly suppress renin transcription, we replaced the CRE at Ϫ2697 to Ϫ2690 in the 4.2-kb mouse renin promoter construct with the minimal Gal4 binding sequence (construct 4.2-kb Gal). The introduction of the Gal4 binding sequence in the native context of the renin promoter should provide for a gene-specific effect since TNF␣ was found to target the CRE at Ϫ2697 to Ϫ2690 (10). TNF␣ did not influence the activity of the 4.2-kb Gal reporter in the absence of Gal4 fusion protein (Fig. 6D), consistent with our earlier data (10), which have shown that the CRE at Ϫ2697 to Ϫ2690 is necessary for the effect of TNF␣. Co-transfection of Gal4:p65 increased the basal activity of the 4.2-kb Gal construct about 2-fold in As4.1 cells (Fig. 6D). This finding fits also to the well known fact that p65 is a transcriptional activator (16,17,31). Importantly, TNF␣ decreased the activity of the 4.2-kb Gal reporter back to its basal level in the presence of Gal4:p65 (Fig.  6D). This finding demonstrated that TNF␣ could down-regulate renin transcription acting through p65, bound in its native context within the renin promoter. To demonstrate also that the TNF␣ effect is specific for NFB, we have co-transfected the 4.2-kb Gal construct with CREB1:Gal4 fusion protein. CREB1: Gal4 increased the basal activity of the 4.2-kb Gal construct more than 3-fold in As4.1 cells (Fig. 6D), consistent with earlier findings as well as with the data in Fig. 1, showing that CREB1 determines to a considerable extent the basal level of renin gene expression. However, TNF␣ did not affect the activity of the 4.2-kb Gal luciferase in the presence of the CREB1:Gal4 fusion protein (Fig. 6D). This result clearly demonstrated that the effect of TNF␣ on renin transcription is specific for p65.
Mechanism of the Down-regulation of p65 Transcriptional Activity by TNF␣-Next we tried to identify the domains in the p65 molecule which are targeted by TNF␣ to inhibit p65 transactivation using the Gal4-dependent system. We used the pFR but not the 4.2-kb Gal luciferase as a reporter, since we aimed to characterize the mechanism of inhibition of p65 trans-activation rather than the regulation in the specific context of a single promoter. It was reported that serine residues 276, 311, 529, and 536 in the p65 molecule are targeted by specific kinases to modulate its transcription potential (22,(32)(33)(34)(35)(36)(37)(38). We mutated each of these serines to alanine in the Gal4:p65 fusion protein (constructs Gal4:p65/276A, Gal4:p65/311A, Gal4:p65/529A, and Gal4:p65/536A) and tested the effect of TNF␣. The activities of all the constructs with a single mutated serine were down-regulated by TNF␣, suggesting that the known phosphorylation sites seemed not to be involved in the mechanism of inhibition of p65 trans-activation (Fig. 7). When serine 276 was changed to alanine in construct Gal4:p65/276A, the rate of down-regulation by TNF␣ was even more pronounced than that of the native Gal4: p65 construct (Fig. 7). Protein kinase A is known to phosphorylate serine 276 of p65 and also to stimulate renin transcription (13)(14)(15)32). Moreover, stimulation of protein kinase A by forskolin/3-isobutyl-1-methylxanthine combination increased the transcriptional activity of p65 (data not shown), suggesting that the positive effect of protein kinase A on renin gene expression is at least in part mediated by p65. Therefore, stronger inhibition of the transcriptional activity of p65 lacking its protein kinase A target would fit with the general model of the regulation of renin transcription. Furthermore we tested the effect of TNF␣ on Gal4: p65 deletion constructs. TNF␣ suppressed the activity of a Gal4: p65 N-terminal half deletion (Gal4:p65 286 -551 ) construct more effectively than the full-length Gal4:p65 1-551 and apparently did not affect the activity of a Gal4:p65 C-terminal half deletion construct (Gal4:p65 1-285 , Fig. 8). The major trans-activation domain (TA 1 , amino acids 521-551) (21), which is located at the very end of the C terminus of p65, was sufficient to transmit the inhibitory effect of TNF␣ (construct Gal4:p65 521-551 , Fig. 8). The last 18 amino acids in the TA 1 domain are believed to form an amphipathic ␣-helix (21) in which four serine residues (in positions 536, 543, 547, and 550) are clustered on one side. Amphipathic helical structures are known to play an important role in trans-activation (21,39). Moreover, serines 529, 536, 543, and 550 are arranged in a heptade repeat suggested to form a zipperlike structure that was proposed to mediate putative proteinprotein interactions (21). Therefore, to further narrow down the target region for the TNF␣ effect on the p65 molecule, we introduced serial deletions of five amino acids starting from the C terminus of p65 in the Gal4:p65 521-551 vector. The first generated construct, Gal4:p65 521-546 , had lower trans-activating capacity compared with Gal4:p65 521-551 , but it was still clearly downregulated by TNF␣ (Fig. 9). The same was also true for construct Gal4:p65 521-546/529,535,536,543A , in which all serines were changed to alanines (Fig. 9). Further deletions from the p65 C terminus yielded constructs with no transcriptional activity (data not shown). These data suggest that the serines at the very C terminus of p65 are important only for the basal trans-activation potential of p65 but not for the inhibitory effect of TNF␣. The suppressive effect of TNF␣ seemed to be transmitted by amino acids 521-546 of the p65 molecule which comprise 13 of the 18 amino acids forming the ␣-helical structure. DISCUSSION Renin synthesis in renal JG cells is regulated by many neural and humoral factors (3-5, 20, 40). The broad variety of signals, which modulate renin production, is reflected by a complicated mechanism controlling renin gene transcription at the cellular level. There are multiple transcription factors identified to date that are linked to the regulation of renin gene expression. CREB/ATFs, USF1/2, Ear2, NF-Y, NF1, and Sp1/ Sp3 were reported to be involved in the control of basal renin transcription (14,(41)(42)(43)(44). Remarkably, all these transcription factors bind to a 242-bp enhancer located ϳ2.8 kb upstream the transcription starting site of mouse renin gene (45). A highly homologous sequence is identified also upstream in the human renin gene (42,46). Although more proximal sequences are required for the proper function of renin promoter (45), it appears that the 242-bp enhancer is the key cis-acting element in the transcriptional machinery driving the expression of renin gene. In support of this hypothesis, most of the factors mediating the signal-specific stimuli to renin promoter also target the enhancer sequence. Thus, CREB/ATFs, which are activated by the cAMP-protein kinase A pathway, bind to CRE at Ϫ2697 to Ϫ2690 (14,15). LXR␣ and Pit-1 transcriptional regulators seem to be involved in the stimulation of renin transcription by cAMP as well (47)(48)(49). Retinoic acid up-regulates renin transcription through RXR␣/RAR␥, which also bind to sequences within the enhancer (50). As to the negative regulation of renin gene, it was found that oncostatin M suppresses renin gene expression in a STAT5-dependent manner (51). Calcium, which is a well known potent inhibitor of renin secretion in native JG cells, inhibited renin transcription by a mechanism involving transcription factor Ref-1 in chorio-decidual cells (52). However, the relevance of this finding for the JG cell-specific regulation of renin gene remains unclear.
We reported that TNF␣, which seems to be a physiologically relevant regulator of renin production (5), inhibited renin transcription in a mouse JG cell line acting through NFB (10). Interestingly, NFB targeted not its consensus sequences in the renin promoter but the CRE at Ϫ2697 to Ϫ2690. Thus, it is well possible that the CREB/ATF transcription factors are also involved in the suppression of renin transcription by TNF␣. Our results imply that CREB1 substantially determines the level of basal renin transcription (Fig. 1). Although it was described that CREB/ATF proteins generally control renin gene expression (13)(14)(15), we show for the first time that CREB1 is a necessary factor for the basal transcription of renin gene. We found only a short term transitory effect of TNF␣ on CREB1 trans-activation, which seemed not to be relevant for the suppression of renin gene expression (Refs. 5 and 10 and Fig. 6D). CREB1 binding to CRE seemed to be moderately reduced after incubation with TNF␣ (Fig. 3). Because crosscoupling of NFB-p65 and bZip proteins such as Fos/Jun, ATF2, and C/EBP transcription factors is well documented (27)(28)(29)(30) and since CREB1 also contains a bZip domain, we tested if CREB1 may interact with NFB-p65. Moreover, NFB proteins generally have increased binding activity upon treatment with TNF␣ (16,17,31). Thus, TNF␣ might decrease the amount of CRE-bound CREB1 by shifting the CREB1 ho- modimers to CREB1:p65 heterodimers. Using in vitro translated p65 and CREB1, we could not establish direct interaction between these two proteins in EMSA (Fig. 5B). Consistently, we could not find CREB1 bound to NFB DNA binding sequence in the nuclear extracts of As4.1 cells (data not shown). Thus, it seems that p65 binds to renCRE as a constituent of the multiprotein trans-activation complex interacting with an unidentified co-factor.
Although the changes in the binding capacity of CREB1 to CRE was minor, TNF␣ is one of the most potent naturally occurring inhibitors of renin transcription (5,10). Therefore, we looked for additional effects that could further elucidate the mechanism of action of TNF␣ on renin gene expression. We have already established that the recruitment of NFB is the crucial event in the down-regulation of renin transcription by TNF␣ (10). The classical concept of NFB signaling postulates that NFB dimers are kept in an inactive state in the cytoplasm by the inhibitor proteins IB-␣, -␤, and -⑀, Bcl-3, p100, and p105 (31). Upon stimulation, these inhibitors are targeted to degradation in the 26 S proteasome, whereas the released NFB translocates to the nucleus to target multiple genes. NFB is a dimeric complex of five proteins: RelA (p65), RelB, cRel, p50, and p52 (31). Depending on the cell type and the incoming signal, it may display differential gene-specific transactivation properties. During the last years accumulating evidence suggested that translocation to the nucleus and binding to the regulatory DNA sequences per se could not account for the various effects of NFB on gene expression (53,54). Therefore, possible post-translational modifications of the p65 subunit, which essentially holds the trans-activation potential of NFB, were studied intensively. It was found that at least four serine residues in the p65 molecule could be phosphorylated by different kinases, which would lead to changes in its transactivation properties (22,(32)(33)(34)(35)(36)(37)(38). Thus, it was demonstrated that NFB may regulate the expression of target gene(s) without dramatic changes in its DNA binding properties. This seems to be the case for the regulation of renin gene by TNF␣. NFB mediates the TNF␣ signal to renin gene, but the binding of p65 to renCRE sequence seemed not to be induced (10). Using a well established Gal4:p65-reporter gene system (21,22,53,54), we demonstrated that the effect of TNF␣ on renin gene expression in As4.1 cells is mediated by a repression of the transcriptional activity of p65 acting through the renin promoter CRE at Ϫ2697 to Ϫ2690. This effect was highly specific since TNF␣ did not affect the long term transactivation of CREB1, which is also known to bind to renin CRE (Fig. 6D). Our results demonstrate for the first time a cytokine-mediated inhibition of NFB transcriptional activity. TNF␣ suppressed the transcriptional activity of p65, as measured with a reporter gene, driven by synthetic Gal4 binding sequences (Fig. 6A), suggesting that the effect of TNF␣ is not due to sequencespecific interactions within the renin promoter, but it is characteristic feature for the As4.1 cells. Renin-producing cells are featured by some further "paradoxes," such as protein kinase C-or Ca 2ϩ -dependent inhibition of transcription (20,52). The cellular mechanisms responsible for these "paradoxes" are almost completely unknown. Therefore we were interested in studying the mode of action of TNF␣ on p65. TNF␣ seemed not to have an impact on the Gal4:p65 1-285 deletion construct, which carries only the DNA binding and dimerization domain of p65. This finding was important since it excludes the possibility that endogenous NFB proteins might squelch Gal4:p65. TNF␣ seemed not to employ the known kinase signaling pathways, which target serines 276, 311, 529, or 536 in p65 molecule (Fig. 7). Rather, TNF␣ was shown to target the main trans-activation domain TA 1 to impair the transcriptional activity of p65 (Figs. 8 and 9). The TA 1 domain of p65 is featured by an 18-amino acid-long amphipathic ␣-helix and a putative serine-zipper. Deletion analysis proved that the intact helix is required for maximal trans-activation and that at least 13 amino acids of the helix are necessary to transmit the effect of TNF␣ as well as to initiate transcription. It was already reported that an artificial 15-amino acid peptide with ␣-helical structure is an efficient trans-activator (39). Such domains are suggested to interact with a general transcription factor or FIG. 7. Effect of TNF␣ on p65 mutants. As4.1 cells were transfected with Gal4:p65, Gal4:p65/276A, Gal4:p65/311A, Gal4:p65/529A, or Gal4: p65/536A, and the effect of TNF␣ (10 ng/ml for 20 h) was tested on a Gal4-driven firefly luciferase (pFR-Luc). Control cells remained untreated. Luc1, firefly luciferase activity; Luc2, renilla luciferase activity. Data are the means Ϯ S.E.; *, p Ͻ 0.05 against the corresponding control; †, p Ͻ 0.05 against TNF␣-treated Gal4:p65.
FIG. 8. TNF␣ targets the C-terminus of p65. As4.1 cells were transfected with Gal4:p65 1-551 , Gal4:p65 1-285 , Gal4:p65 286 -551 , or Gal4: p65 521-551 , and the effect of TNF␣ (10 ng/ml for 20 h) was tested on a Gal4-driven firefly luciferase (pFR-Luc). Control cells remained untreated. Luc1, firefly luciferase activity; Luc2, renilla luciferase activity. Data are the means Ϯ S.E.; *, p Ͻ 0.05 against the corresponding control; †, p Ͻ 0.05 against TNF␣-treated Gal4:p65 1-551 . with RNA polymerase. Thus, TNF␣ may interfere with the interaction of p65 TA 1 domain with the general transcriptional machinery to suppress transcription. However, the zipper-like arranged serines in the TA 1 domain seem not to mediate the suppressive effect of TNF␣, although it was believed that they may play role in protein-protein interactions, similar to the leucine zippers. Our data could not entirely rule out the possibility that other trans-activation domains in the p65 molecule are also involved in mediating the TNF␣ effect. However, the TA 1 sequence holds the full transactivating capacity of p65 in kidney epithelioid cell lines such as COS7 or As4.1 (21, data not shown), and TA 1 alone was found to be sufficient to mediate the effect of TNF␣ (Fig. 8), suggesting that it plays central role in the repression of p65 trans-activation.
Recently it was reported that p65, constitutively bound to a serum response element, is required for the full activity of the c-fos promoter (55). Although it was proposed that p65 may co-operate with other transcription factors and co-activators, the precise function of p65 in the regulation of c-fos gene expression remained unclear. These findings would support the concept, based on our earlier results (10) and on the present data, that p65, constitutively bound to an NFB-unrelated sequence, may play a substantial role in the regulation of transcription.
In this paper we provide evidence that the effect of TNF␣ on renin gene expression is mediated by CREB1 and NFB-p65 in a mouse JG-like cell line. Attenuated binding of CREB1 was found to contribute in part to the inhibition of renin transcription. However, the principal mechanism of down-regulation of renin gene by TNF␣ appears to be the suppression of the NFB-p65 transcriptional activity.