Studies have identified IFN-inducible Ifi202 gene as a lupus susceptibility gene (encoding p202 protein) in mouse models of lupus disease. However, signaling pathways that regulate the Ifi202 expression in cells remain to be elucidated. We found that steady-state levels of Ifi202 mRNA and protein were high in mouse embryonic fibroblasts (MEFs) from E2F1 knockout (E2F1−/−) and E2F1 and E2F2 double knockout (E2F1−/−E2F2−/−) mice than isogenic wild-type MEFs. Moreover, overexpression of E2F1 in mouse fibroblasts decreased expression of p202. Furthermore, expression of E2F1, but not E2F4, transcription factor in mouse fibroblasts repressed the activity of 202-luc-reporter in promoter-reporter assays. Interestingly, the E2F1-mediated transcriptional repression of the 202-luc-reporter was independent of p53 and pRb expression. However, the repression was dependent on the ability of E2F1 to bind DNA. We have identified a potential E2F DNA-binding site in the 5′-regulatory region of the Ifi202 gene, and mutations in this E2F DNA-binding site reduced the E2F1-mediated transcriptional repression of 202-luc-reporter. Because p202 inhibits the E2F1-mediated transcriptional activation of genes, we compared the expression of E2F1 and its target genes in splenic cells from lupus-prone B6.Nba2 congenic mice, which express increased levels of p202, with age-matched C57BL/6 mice. We found that increased expression of Ifi202 in the congenic mice was associated with inhibition of E2F1-mediated transcription and decreased expression of E2F1 and its target genes that encode proapoptotic proteins. Our observations support the idea that increased Ifi202 expression in certain strains of mice contributes to lupus susceptibility in part by inhibiting E2F1-mediated functions.
Systemic lupus erythematosus (SLE)3 is an autoimmune disease that predominantly affects women of childbearing age (1, 2, 3). The disease is characterized by the production of pathogenic autoantibodies, particularly IgG Abs, to nuclear Ags and development of lupus nephritis (1, 2, 3). Based on genetic studies in human SLE patients and in mouse models of SLE, there is considerable evidence that SLE is a polygenic disease (1, 4, 5). Interestingly, studies have revealed that PBMC from lupus patients exhibit an IFN signature: expression levels of mRNAs that are encoded by the IFN-inducible genes are up-regulated (6). Consistent with a role for IFN-inducible genes in human lupus patients, generation of congenic mice, such as B6.Nba2 (7) and B10.Yaa.Bxs2/3 (8), coupled with gene expression analyses have identified the IFN-inducible Ifi202 gene as a lupus susceptibility gene. Furthermore, a study has revealed that increased expression of Ifi202 in spleen and kidney cells of MRL/lpr mice is associated with development of lupus disease (9). Although these studies have suggested that increased expression of Ifi202 in certain strains of mice is associated with the development of lupus or lupus-like disease, signaling pathways that regulate expression of the Ifi202 gene in immune cells remain to be elucidated. Moreover, it remains unclear how increased expression of Ifi202 in certain strains of mice contributes to lupus susceptibility.
The E2F family of transcription factors comprises at least six structurally-related E2Fs (E2F1–6) (10, 11, 12). Based on their ability to complex with the pocket family proteins (pRb, p107, and p130) and their expression patterns, these transcription factors have been grouped into three groups. E2F1, E2F2, and E2F3 have been grouped together, whereas E2F4 and E2F5 are grouped together. The E2F6 is grouped separately. These E2F factors heterodimerize with members of the DP family (DP1 and DP2) (10) of proteins to form an active transcription factor. The E2F and DP protein heterodimer is kept transcriptionally silent and acts as a repressor by binding to a pocket protein. Cyclin/Cdk-mediated phosphorylation of pocket proteins results in the release of “free” E2F/DP dimer, allowing transcriptional activation of E2F target genes (10, 11, 12). Transcriptional activation of genes by E2F in response to mitogenic stimulation and the identification of E2F DNA-binding sites in a number of genes critical to the regulation of DNA synthesis implicate E2F regulation as an important step in mammalian cell cycle progression (10, 11, 12). Consistent with this idea, the E2F family of transcription factors contributes to the regulation of G1-to-S phase transition of T and B cells in response to mitogenic stimulus (11, 12). Importantly, cell growth-inhibitory cytokines, such as IFNs, inhibit cell cycle progression in part by inhibiting the E2F-mediated transcription of genes (13, 14).
Interestingly, the E2F1 transcription factor autoregulates expression of the E2F1 gene (15), and the lack of E2F1 expression is associated with defects in apoptosis of cells (16). The E2F1 transcriptional target genes encode proteins with functions in cell cycle progression and apoptosis (10, 11, 12). The E2F1 target genes encode proteins, such as cyclin E, DHFR, and E2F1, that function in the G1-to-S phase transition during the cell cycle progression (10, 11, 15). Moreover, E2F1 target genes also encode proteins, such as p73, PUMA, Noxa, and Bim, that are known to have proapoptotic functions in immune cells (12, 17).
IFN-inducible Ifi202 gene encodes a 52-kDa phosphoprotein p202 (18, 19, 20, 21). Overexpression of p202 protein in a variety of mouse fibroblasts retards cell proliferation and increases cell survival, in part by inhibiting E2F-mediated transcription (20, 21). Furthermore, increased expression of p202 in splenic cells of the B6.Nba2 congenic mice (congenic for Nba2 locus derived from New Zealand Black (NZB) lupus-prone mice) is associated with increased production of IgG Abs, splenomegaly, and defects in apoptosis of B cells in vitro (7, 21). Importantly, generation of B6.Nba2 mice that are deficient for the IFN-α/β receptor resulted in more than a 2-fold reduction in Ifi202 mRNA levels in splenic cells (22). Consistent with this observation, others have reported (23) that basal levels of Ifi202 mRNA are 3–4-fold higher in NZB/W (White) mice than BALB/c or MRL/lpr mice. Although these observations are consistent with the regulation of Ifi202 gene by IFNs (IFN-α or IFN-β) in immune cells, our studies (24) revealed that type I IFN receptor deficiency reduced lupus-like disease in NZB mice, but did not result in decreased levels of p202 protein.
Our previous studies (25, 26, 27, 28) revealed that expression of Ifi202 is also regulated by IFN-independent signaling pathways. For example, we found that IL-6 up-regulates expression of the Ifi202 gene in mouse fibroblasts and B6.Nba2 splenocytes through activation of transcription factor STAT3 (28). Furthermore, we noted that serum growth factors, such as platelet-derived growth factor, basic fibroblast growth factor, and TGF-β, negatively regulate the expression of Ifi202 (25). However, the molecular mechanisms remain unclear.
Because the E2F family of transcription factors contributes to transition from G1-to-S phase of cell cycle in response to serum growth factors (10, 11) and because certain E2F family members, such as E2F1 and E2F2, are implicated in the regulation of autoimmunity (29, 30), we investigated whether the E2F family members regulate Ifi202 expression. Herein we report that the E2F family members differentially regulate expression of the Ifi202 gene.
Materials and Methods
Cells, cell culture, and reagents
Mouse embryonic fibroblasts (MEFs) from p53-null (p53−/−) and isogenic wild-type mice (31) were generously provided by Dr. L. Donehower (Baylor College of Medicine, Houston, TX). Rb-null (Rb−/−) and isogenic wild-type (Rb+/+) MEFs (32) were generously provided by Dr. Ed Harlow (Harvard Medical School, Boston, MA). MEFs from E2F1-null (E2F1−/−) and wild-type (E2F1+/+) mice (16) were generously provided by Dr. L. Yamasaki (Columbia University, New York, NY). MEFs and splenocytes from E2F1 and E2F2-double null (E2F1−/−E2F2−/−) and wild-type (E2F1+/+E2F2+/+) age-matched mice (30) were generously provided by Dr. J. DeGregori (University of Colorado, Denver, CO). MEFs and splenocytes from age-matched C57BL/6 (B6) and B6.Nba2 congenic mice (7) were generously provided by Drs. S. J. Rozzo and B. L. Kotzin (University of Colorado Health Sciences Center, Denver, CO). NIH3T3 fibroblasts stably overexpressing E2F1 (3T3/E2F1 cl3, Ref. 33) were generously provided by Dr. H. Tanaka (Osaka University, Osaka, Japan). Mouse NIH3T3 fibroblasts were purchased from the American Type Culture Collection. MEFs (between passages 3 and 5) and NIH3T3 fibroblasts were maintained at low density in DMEM containing high glucose, supplemented with 10% calf serum and antibiotics in an incubator with 5% CO2. Antiserum to p202 has been described previously (19).
Plasmids and expression vectors
Dr. Peggy Farnham (University of California–Davis, Davis, CA) generously provided pCMV-mE2F1 expression vector. Dr. J. Nevins (Duke University, Durham, NC) generously provided plasmids encoding various mutants of E2F1 (E2F1 (E132), E2F1 Δ206–220, E2F1 Δ1–88, and E2F1 (411/421)). The 202-luc-reporter plasmid (202-luc), containing the 5′-regulatory sequence (∼0.8 kb) of the Ifi202 gene (derived from the EAT cells), has been described previously (28). Mutations in the E2F1 DNA-binding site (indicated as 202-E2F-BS) in the 202-luc-reporter plasmid were introduced using the QuikChange site-directed mutagenesis kit (from Stratagene) as suggested by the supplier. The sequencing of the mutant reporter plasmid confirmed that the nucleotide sequence in the 202-luc-reporter plasmid was changed from AAAGGGCGCGAAA to AAAGGTTTCGAAA.
For reporter assays, subconfluent cultures of cells (in 6-well plates) were transfected with the reporter plasmids 202-luc (2.5 μg) and pRL-TK (0.5 μg) using a calcium phosphate transfection kit (from Invitrogen), as suggested by the supplier. Unless otherwise indicated, cells were harvested between 43 and 48 h after transfections. Cells were lysed, and the firefly and Renilla luciferase activities were determined as described previously (28).
Isolation of RNA from splenocytes and RT-PCR
Total single-cell splenocytes were isolated from female B6 or B6.Nba2 mice (8–10 wk of age). After lysis of RBC, splenocytes were resuspended in RPMI 1640 with 10% FBS, 1% penicillin/streptomycin/glutamate, and 1× MEM, nonessential amino acids/sodium pyruvate. Splenocytes (5 × 106 cells) were used to isolate total RNA using TRIzol (Invitrogen). Total RNA was digested with DNase I (to remove any DNA in the preparation), and 0.5 μg of RNA was used for RT-PCR reaction using a pair of the Ifi202-specific primers (forward primer: 5′-ggtcatctaccaactcagaat-3′; reverse primer: 5′-ctctaggatg ccactgctgttg-3′). For RT-PCR, we used the SuperScript one-step RT-PCR system from Invitrogen. Primers for other genes were as follows: mE2F1 (118 bp), forward: 5′-GATCGAAGCTTTAATGGAGCG-3′, reverse: 5′-CCCTTGCTTCA GAGAACAG-3′; PUMA (211 bp), forward: 5′-AGCACTTAGAGT CGCCCGT-3′, reverse: 5′-GAGGAGT CCCATGAAGAGATTG-3′; Bim (244 bp), forward: 5′-TAAGTTCTGAG TGTGACAGAGA AGG-3′, reverse: 5′-CAGTTGTA AGATAACCATTTGA GGGTGG-3′.
Splenocytes, MEFs, or NIH3T3 cells were collected in PBS and resuspended in a modified radioimmunoprecipitation assay (RIPA) lysis buffer (50 mM Tris-HCl (pH 8.0), 250 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS), supplemented with protease inhibitors (leupeptin, 50 μg/ml; pepstatin A, 50 μg/ml; and PMSF, 1 mM) and incubated at 4°C for 30 min. Cell lysates were sonicated briefly before centrifugation at 14,000 rpm in a microcentrifuge for 10 min at 4°C. The supernatants were collected, and the protein concentration was measured by Bio-Rad protein assay kit. Equal amounts of protein were processed for immunoblotting as described previously (28).
Nuclear extract was prepared from mouse fibroblasts as described previously (34). The protein concentration was measured with the Bio-Rad protein assay reagents. An oligonucleotide (25 nucleotides long) containing the potential E2F DNA-binding site (202-E2F-BS) present in the 5′-regulatory region of Ifi202, or containing an E2F DNA-binding site consensus sequence (purchased from Santa Cruz Biotechnology), was end-labeled with [γ-32P]ATP and polynucleotide kinase. Gel-purified labeled oligonucleotide (probe, 1 ng/assay) was incubated with nuclear extracts containing equal amounts of protein in binding buffer (20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 1 mM EDTA, 10% glycerol, 0.1% Nonidet P-40, 1 mM DTT, and 100 μg of BSA) for 20 min at room temperature in the presence of 0.5 μg of poly(dI:dC). Nuclear extracts were treated with detergent sodium deoxycholate as described previously (34). In the competition assays, the nuclear extracts were preincubated with excess unlabeled oligonucleotides at room temperature for 15 min. Samples were analyzed on a 5% native gel in 1× TBE (90 mM Tris-borate, 2 mM EDTA (pH 8.0)) buffer. Gel was dried and exposed to x-ray film at −80°C.
Chromatin immunoprecipitation assays
National Institutes of Health 3T3 or NIH3T3-E2F1 cells (8 × 106) were cross-linked for 10 min at room temperature by directly adding 1/10th medium volume of cross-linking reagent (11% formaldehyde, 100 mM NaCl, 0.5 mM EGTA, 50 mM HEPES (pH 8.0)) to the plate. The cross-linking reaction was quenched by adding 1/10th volume of 1.25 M glycine and incubating cells at room temperature for 5 min. Cells were centrifuged and washed (twice) with PBS. Cells were resuspended in 500 μl of lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl (pH 8.0), supplemented with the protease inhibitors) and tubes were incubated on ice for 10 min with occasional mixing. The cell lysates were sonicated (2 times, 15 s each) to break chromatin and processed for the reversal of cross-linking. The cell lysates were diluted to a final volume of 1 ml with a mixture of 9 parts of the dilution buffer (1% Triton X-100, 150 mM NaCl, 2 mM EDTA, 20 mM Tris-HCl (pH 8.0), plus protease inhibitors) and 1 part of cell lysis buffer. The diluted cell lysates were first precleared by incubating with protein A/G beads (Pierce Biotechnology) for 30 min at 4°C. After preclearing, the cell lysates (containing equal amounts of proteins) were divided into two fractions: one fraction was subjected to immunoprecipitation using mouse mAb to E2F1, and another fraction was immunoprecipitated with the same amount of mouse normal IgG Ab as a control. The immunoprecipitates were collected using protein A/G beads. Beads were centrifuged, washed (five times) in 1 ml of washing buffer (1% Triton X-100, 0.1% SDS, 150 mM NaCl, 2 mM EDTA, 20 mM Tris-HCl (pH 8.0)). The final wash was with a buffer containing 1% Triton X-100, 0.1% SDS, 500 mM NaCl, 2 mM EDTA, 20 mM Tris-HCl (pH 8.0). Immune complexes were eluted from beads by adding 450 μl of elution buffer containing proteinase K and RNase A (500 μg/ml each), and by incubating tubes at 37°C for 30 min. Tubes were centrifuged and the supernatants containing DNA were collected. The DNA was purified, precipitated, washed, and dissolved in water. Equal volumes of DNA were used to perform PCR to detect the promoter region of the Ifi202 gene. The following primers were used: region 383–402 forward primer (5′-gtgtctagtggccagtgtac-3′) and region 804–784 reverse primer (5′-tctgcagtgatgtacagatcc-3′) (see Fig. 5; Ref. 18). The PCR amplification conditions were: incubation at 94°C for 5 min, followed by 94°C for 30 s, 45°C for 30 s, and 72°C for 45 s. The amount of amplified DNA was compared after 32 or 36 amplification cycles by agarose gel electrophoresis to determine whether the amplification of the input DNA was in the linear range.
Transcription factor E2F1 negatively regulates expression of Ifi202
Our previous studies (25) revealed that serum growth factors, such as platelet-derived growth factor, basic fibroblast growth factor, or TGF-β, negatively regulate expression of Ifi202 in mouse fibroblasts. Moreover, we found (35) that levels of Ifi202 mRNA and protein decrease in mouse fibroblasts when cells exit from quiescence (the G0/G1 phase of cell cycle) to enter into S phase of cell cycle after serum stimulation. Because serum growth factors stimulate transition of cells from G0/G1 phase of cell cycle to S phase, in part by activating the E2F1-mediated transcription of S-phase genes (10, 11), we explored whether E2F1 could negatively regulate expression of the Ifi202. For this purpose, we compared basal levels of Ifi202 mRNA and protein between E2F1-null (E2F1−/−) and isogenic wild-type (E2F+/+) MEFs. As shown in Fig. 1,A, basal levels of Ifi202 mRNA were not detectable in the wild-type MEFs. However, the mRNA levels were readily detectable in E2F-null MEFs. Consistent with detectable mRNA levels in E2F1-null MEFs, p202 protein levels were detectable in E2F1-null MEFs, but not in isogenic wild-type cells (Fig. 1,B). Furthermore, overexpression of E2F1 in NIH3T3 cells resulted in reduced levels of p202 protein (Fig. 1 C). Taken together, these observations suggested that expression of E2F1 in MEFs negatively regulated expression of the Ifi202.
E2F1-mediated transcriptional repression of Ifi202 is p53 and pRb independent
Our previous studies revealed that p53 transcriptionally represses expression of Ifi202 (26). This observation made it conceivable that E2F1 represses the expression of Ifi202 through stabilization of p53 (through transcriptional activation of the alternative reading frame of p19; see Ref. 36). Therefore, we tested whether E2F1-mediated negative regulation of Ifi202 expression is p53 dependent. For this purpose, we performed promoter-reporter assays. As shown in Fig. 2,A, expression of mouse E2F1 (mE2F1) in p53-null (p53−/−) or wild-type (p53+/+) MEFs reproducibly decreased the activity of the 202-luc-reporter ∼50–60% in three independent experiments. This observation suggested that the E2F1-mediated transcriptional repression of 202-luc-reporter activity was independent of p53 expression. Furthermore, we noted that transfection of p53-null MEFs with increasing amounts of mE2F1 encoding plasmid decreased the activity of the 202-luc-reporter in a dose-dependent manner, and the decrease in the activity of the reporter was ∼70% (Fig. 2 B).
Because E2F transcription factors in complex with the pRb (or other pocket proteins, such as p107 and p130) repress transcription of their target genes (10, 11), we also tested whether E2F1-mediated transcriptional repression of Ifi202 gene is pRb dependent. Our promoter-reporter assays using pRb-null MEFs revealed that E2F1 repressed the activity of 202-luc-reporter in pRb-null MEFs (Fig. 2 C). These observations suggested that E2F1-mediated transcriptional repression of Ifi202 is pRb independent.
E2F1 DNA-binding activity is required for transcriptional repression of Ifi202
The E2F1-mediated transcriptional repression of genes is known to require DNA-binding activity (10, 11). Therefore, we tested whether E2F1-mediated transcriptional repression of Ifi202 gene requires the DNA-binding domain in E2F1. As shown in Fig. 2 D, expression of the wild-type E2F1 in p53-null cells decreased the activity of 202-luc-reporter (compare column 2 with column 1). Interestingly, transfection of cells with a deletion mutant of E2F1 (E132) in which the DNA-binding domain is deleted stimulated the activity of the reporter (compare column 3 with column 2). However, expression of deletion mutants of E2F1 in which either the N terminus (Δ1–88 amino acids) or the leucine zipper (Δ206–220) region was deleted repressed the activity of 202-luc-reporter. Moreover, expression of the E2F1 double-point mutant 411/421, which has previously been shown (37) to be selectively defective in overcoming p107 while maintaining wild-type pRb repressor function, inhibited the activity of 202-luc, suggesting that the p107 repressor function is not required for E2F1-medaited repression of Ifi202. Taken together, these observations suggested that the DNA-binding domain in E2F1 is required for transcriptional repression of Ifi202.
Differential regulation of Ifi202 expression by the E2F family members
Because the E2F family of transcription factors, such as E2F1, E2F2, or E2F3, but not E2F4, is thought to participate in the G1-to-S phase transition of cells (10, 11), we compared the ability of these transcription factors to repress transcription of Ifi202 gene. As shown in Fig. 3, transfection of NIH3T3 cells with a plasmid encoding either E2F1 or E2F2 transcription factor repressed the activity of the 202-luc-reporter to a similar extent. Interestingly, transfection of plasmid that encodes the E2F3 transcription factor only partially (40%) inhibited the activity of the reporter. However, expression of the E2F4 transcription factor did not repress transcription of 202-luc-reporter in several independent experiments.
Furthermore, consistent with our above observations that both E2F1 and E2F2 repressed transcription of the Ifi202 gene, we found that basal levels of p202 protein were detectable in splenocytes from E2F1−/−E2F2−/− double knockout mice, but not in age-matched wild type (E2F1+/+E2F2+/+) mice (data not shown). Moreover, the basal activity of 202-luc-reporter was consistently ∼2-fold higher in MEFs from the double knockout mice than from the wild-type mice in two independent experiments (data not shown). These observations suggested that structurally related E2Fs, such as E2F1, E2F2, and E2F3, repress the transcription of the Ifi202 gene, albeit to different extents.
E2Fs bind to a potential DNA-binding site in the 5′-regulatory region of the Ifi202 gene
Because E2F-mediated transcriptional repression of 202-luc-reporter was dependent on the ability of E2F to bind DNA (Fig. 2,D), we analyzed the 5′-regulatory region of Ifi202 gene for the presence of potential E2F DNA-binding sites using the Matinspector 2.0 software program. The sequence analyses revealed a potential E2F DNA-binding site (5′-TTTCCCTG-3′; 202-E2F-BS). A comparison of the 202-E2F-BS in the Ifi202 5′-regulatory region with other E2F DNA-binding consensus sites (5′-TTTCCCGC-3′; E2F-CS), which are bound by most (90%) E2F1/DP1 complexes (38), revealed at least two variations or mismatches between the 202-E2F-BS and E2F-CS sequence (see Table I).
|E2F1/DP1 DNA-binding consensus sequence||T T T C C C G C|
|202-E2F1-BS||T T T C C C T G|
|202-E2F-BS mutant||T T T C C T T T|
|E2F1/DP1 DNA-binding consensus sequence||T T T C C C G C|
|202-E2F1-BS||T T T C C C T G|
|202-E2F-BS mutant||T T T C C T T T|
Boldface letters indicate nucleotides that appear in >90% of recovered E2F1/DP1 DNA-binding sites (38 ). Underlined letters indicate variations in the nucleotides from the E2F DNA-binding consensus sequence.
Next, we performed gel-mobility shift assays to determine whether 202-E2F-BS (Fig. 4,A) in Ifi202 gene can bind to E2F transcription factors. As shown in Fig. 4 B, in nuclear extracts from NIH3T3 cells, we could detect binding of proteins to an oligonucleotide (probe) containing the 202-E2F-BS sequence (compare lane 2 with lane 1). Importantly, treatment of nuclear extracts with detergent deoxycholate, which has been previously known to dissociate E2F proteins from the pocket family proteins in NIH3T3 cells (34), resulted in “free” E2F (a dimmer between E2F and DP proteins). Moreover, competition with 50-fold excess of unlabeled oligonucleotide containing the E2F-CS greatly reduced the binding of proteins to the oligonucleotide with the 202-E2F-BS sequence (compare lane 4 with lanes 2 or 3).
Because the 202-E2F-BS sequence contains at least two variations (or mismatches) in nucleotides from the E2F-CS (see Table I), we sought to compare the relative affinity of 202-E2F-BS with E2F-CS. For this purpose, we performed gel-mobility shift assays using nuclear extracts from NIH3T3 cells and labeled oligonucleotide with E2F-CS from the DHFR, a well-known E2F target gene (10). As shown in Fig. 4 C, proteins in nuclear extracts bound to E2F-CS, and a competition with 50-fold excess of cold E2F-CS reduced the binding of proteins to the probe (compare lane 3 with lane 2). Furthermore, as expected from the presence of two variations in the 202-E2F-BS sequence, competition with 20- or 50-fold excess of cold oligonucleotide with 202-E2F-BS did not reduce binding of labeled E2F-CS with proteins (compare lane 4 or 5 with lane 2). However, competition with 100-fold excess cold oligonucleotide containing the 202-E2F-BS sequence reduced the binding of the probe >50%. These observations suggested that potential E2F-BS in the 5′-regulator region of Ifi202 gene can bind to E2F protein complexes in gel-mobility shift assays with relatively low affinity relative to the E2F DNA-binding sites found in most E2F-responsive genes.
To confirm binding of E2F1 to 202-E2F-BS in vivo, we performed chromatin immunoprecipitation assays. As shown in Fig. 5,B, some binding of E2F1 to Ifi202 regulatory region was detected in asynchronous cultures of NIH3T3 cells (compare lane 7 with lane 5). However, overexpression of E2F1 in NIH3T3 cells, which decreased the expression of p202 (see Fig. 1,C), increased binding of E2F1 to the regulatory region of the Ifi202 (compare lane 8 with lane 7). This observation is consistent with our observation that the E2F1 DNA-binding activity is required to repress transcription of 202-luc-reporter (Fig. 2,D) and that E2Fs can bind to 202-E2F-BS sequence in gel-mobility shift assays (Fig. 4).
E2Fs represses transcription of Ifi202 gene through the E2F-BS in the regulatory region
Our observation that E2F1 expression repressed expression of the Ifi202 and E2F1 bound to a potential E2F DNA-binding site in the 5′-regulatory region of Ifi202 gene prompted us to determine whether this E2F DNA-binding site is needed for E2F1-mediated transcriptional repression. Therefore, we mutated the 202-E2F-BS in the 202-luc-reporter plasmid by changing a critical nucleotide from C to T (see Table I) by site-directed mutagenesis and compared the activity of the wild-type reporter with the mutant reporter. As shown in Fig. 6, expression of mE2F1 repressed the activity of wild-type 202-luc-reporter. Interestingly, introduction of a point mutation in the 202-E2F biding sequence resulted in a 60% decrease in the E2F1-mediated transcriptional repression in two independent experiments. These observations suggested that E2F1 represses the expression of Ifi202 in part through the 202-E2F-BS in the 5′-regulatory region of the Ifi202 gene.
Increased expression of p202 in B6.Nba2 mice is associated with inhibition of E2F1 expression and down-regulation of the E2F1 target genes
Our previous studies (34, 39) revealed that increased levels of p202 in cells inhibit E2F1- and E2F4-mediated transcription of target genes. These observations raised the possibility that increased expression of p202 in B6.Nba2 mice contributes to lupus susceptibility, in part by down-regulating the expression of E2F1 and its target genes. Therefore, we tested whether increased expression of p202 in MEFs derived from the B6.Nba2 mice affects the transcriptional activity of E2F. For this purpose, we compared the activity of a well-known E2F-responsive reporter (E2F-luc) in MEFs derived from B6 and B6.Nba2 mice. As shown in Fig. 7,A, basal activity of the E2F-responsive reporter was 70% lower in unsynchronized B6.Nba2 than in B6 MEFs. This observation prompted us to compare the expression of E2F1 mRNA and protein between B6 and B6.Nba2 splenocytes in vivo. As shown in Fig. 7,B, levels of E2F1 mRNA were ∼80% lower in B6.Nba2 splenocytes than in B6 splenocytes. Additionally, we found that mRNA levels of known E2F1 target genes, such as PUMA and Bim, were also lower in splenocytes derived from the B6.Nba2 mice. We also compared the protein levels of E2F1 and its transcriptional target, cyclin E, between B6 and B6.Nba2 splenocytes (Fig. 7,C). We noted that levels of both E2F1 and cyclin E were lower in B6.Nba2 splenocytes than in B6. Additionally, we also compared the expression levels of p73, a transcriptional target of the E2F1, between age-matched B6 and B6.Nba2 female mice. As shown in Fig. 7 D, levels of p73 were lower in splenocytes derived from two B6.Nba2 female mice than in B6 female mice. These observations provide support to the idea that increased levels of p202 in B6.Nba2 mice contribute to lupus susceptibility by decreasing the expression of E2F1 transcription factor and inhibition of E2F1-mediated transcription of target genes.
Studies involving SLE patients (6) and mouse models of lupus disease (9, 23) have revealed a role for IFN-inducible genes in the development of lupus disease. Because IFN-inducible proteins that regulate cell proliferation and survival are likely to contribute to the development of autoimmunity, identification of such IFN-inducible proteins will allow us to elucidate their role in immune regulation. Furthermore, it is important to understand how cellular levels of IFN-inducible proteins are regulated by other signaling pathways in immune cells.
Through regulation of a set of genes that regulate transition from G1-to-S phase of cell cycle, the transcription factor E2F1 regulates cell proliferation, differentiation, and apoptosis of B and T cells (29, 30, 36, 40, 41, 42, 43, 44, 45, 46, 47). Moreover, defects in cell cycle progression and apoptosis of immune cells that result from the lack of E2F1 or E2F1 and E2F2 function contribute to the development of lupus-like disease (29, 30). Consistent with this idea, the transcription factor E2F1 is important for activation-induced cell death in T cells (44). Furthermore, E2F1 and E2F2 transcription factors determine the threshold for Ag-induced T cell proliferation (30), and the activity of E2F2 is needed for suppression of T cell proliferation and immunologic self-tolerance (29). Consequently, disruption of the gene for the E2F1 transcription factor causes significant increases in T cell number and splenomegaly (16). Similarly, mice that are null for E2F2 transcription factor develop a lupus-like disease (29). Importantly, the E2F2 transcription factor represses the transcription of the E2F1 gene, whose activity is required in cells for normal S phase entry during cell cycle progression (29). Because serum growth factors down-regulate the expression of Ifi202 (25) and p202 inhibits the E2F (E2F1 and E2F4)-mediated transcription (34, 39), we tested whether E2F transcription factors could regulate expression of the Ifi202.
Our experiments revealed that: 1) steady-sate levels of Ifi202 mRNA and protein were higher in E2F-null than in wild-type MEFs and overexpression of E2F1 decreased p202 levels (Fig. 1); 2) splenic cells from E2F1 and E2F2 double knockout mice expressed higher levels of p202 than did isogenic wild-type cells (Fig. 3); 3) in promoter-reporter assays, expression of E2F1, E2F2, and E2F3, but not E2F4, repressed transcription of the 202-luc-reporter (Fig. 3); 4) the 5′-regulatory region of the Ifi202 gene contains a potential E2F DNA-binding site (202-E2F-BS) that bound to E2F protein complexes in gel-mobility shift assays (Fig. 4) and in chromatin immunoprecipitation assays (Fig. 5), and a point mutation in this E2F DNA-binding site decreased the E2F-mediated transcriptional repression of 202-luc-reporter (Fig. 6); and 5) increased expression of Ifi202 in splenic cells from B6.Nba2 congenic mice was associated with reduced expression of E2F1 and its target genes (Fig. 7). Taken together, these observations demonstrated that: 1) E2Fs differentially regulate expression of the Ifi202 gene, and 2) increased expression of p202 in splenic cells from B6.Nba2 mice negatively regulates E2F-medaited transcription.
The 202-luc-reporter construct that is used in our promoter-reporter assays (Figs. 2 and 3) contains only ∼800 bp from the 5′-regulatory region of the Ifi202 gene. Therefore, it is conceivable that E2F-responsive elements upstream to this 800-bp region in Ifi202 gene also contribute to the regulation of the gene. This could account for moderate decreases in the activity of the 202-luc-reporter in our assays after expression of E2F (E2F1, E2F2, or E2F3) transcription factors and significantly increased steady-state levels of Ifi202 mRNA and protein in E2F1-null cells.
We found that E2F-mediated transcriptional repression of Ifi202 gene was independent of p53 or pRb expression (Fig. 2). Because the p53 protein family includes p63 and p73 proteins (48) and the pRb protein family includes p107 and p130 proteins (11), our observations do not rule out the possibility that E2F1-mediated transcriptional repression of Ifi202 gene depends on other p53 or pRb family proteins. Further work will be needed to test these possibilities.
The IFN-inducible p202 protein binds to E2F1 and inhibits E2F1-mediated transcription of target genes (34). Binding of p202 to E2F1 results in inhibition of specific DNA-binding activity of E2F complexes in gel-mobility shift assays (34). Moreover, p202 inhibits E2F1-mediated apoptosis (49). Taken together, these observations support the idea that increased levels of p202 in immune cells (B and T cells) increase the threshold for apoptosis by inhibiting E2F-mediated transcription of its target genes, which encode proapoptotic proteins. Consistent with the above idea we found that increased expression of Ifi202 in B6.Nba2 was associated with increases in the numbers of T and B cells and splenomegaly (7).
Stimulation of the TCR on cycling peripheral T cells causes their apoptosis by TCR activation-induced cell death (44). Furthermore, E2F1-null or p73-null primary T cells do not undergo TCR-mediated apoptosis (44). Similarly, E2F1-mediated expression of proapoptotic proteins, such as Bim and Puma, contributes to apoptosis of cells (50). Therefore, our observations that increased expression of p202 in B6.Nba2 splenic cells is associated with reduced expression of E2F1, Bim, and p73 provide support for the idea that inhibition of TCR activation-induced cell death by p202 contributes to accumulation of T cells.
Studies have suggested that E2F1 plays a role in the induction of ARF, p53, and apoptosis during thymic negative selection (36). Because increased expression of p202 in cells inhibits p53- and E2F1-mediated transcription (20, 21), our observations are consistent with the idea that increased levels of p202 in B6.Nba2 mice contribute to defects in apoptosis during thymic negative selection. Moreover, our observations suggest that transcriptional repression of Ifi202 gene by both p53 (26) and E2F1 (this study) may be required for normal thymic negative selection.
Increased expression of the p202 in cell lines inhibits the transcriptional activity of factors such as p53, NF-κB, and AP-1 (20, 21). These transcription factors, by modulating transcription of target genes, are known to regulate cell proliferation and survival (20, 21). Consistent with inhibition of the transcriptional activity of the above factors by p202, the increased expression of p202 in cell lines, depending on the cell context, either sensitizes cells to apoptosis or decrease the susceptibility to apoptosis (21). Importantly, we found that increased expression of p202 in splenic B6.Nba2 cells was associated with inhibition of p53-mediated transcription of proapoptotic genes and defects in apoptosis of cells (51). Therefore, our earlier observations that certain E2Fs, such as E2F1 and E2F2, repress the transcription of the Ifi202 gene are consistent with our above observations.
In summary, our observations provide support for our model (Fig. 8). The model predicts that mitogenic signaling pathways that activate the E2F1-mediated transcription negatively regulate Ifi202 expression. In contrast, signaling pathways (and host factors) that contribute to increased expression of Ifi202 inhibit E2F1-mediated transcription of proapoptotic genes in cells. Therefore, defects in mutual regulation of Ifi202 and E2F1 expression and functions are likely to contribute to defects in cell proliferation and/or apoptosis, resulting in autoimmunity. Our observations will serve as a basis to identify signaling pathways and molecules that contribute to the development of autoimmune diseases.
We thank Drs. L. Donehower, Ed Harlow, L. Yamasaki, J. DeGregori, S. J. Rozzo, B. L. Kotzin, H. Tanaka, P. Farnham, and J. Nevins for generously providing cells and reagents. We also thank Drs. P. Lengyel and B. L. Kotzin for thoughtful suggestions.
The authors have no financial conflicts of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by National Institutes of Health Grant AI066261 (to D.C.).
Abbreviations used in this paper: SLE, systemic lupus erythematosus; MEF, mouse embryonic fibroblast; Rb, retinoblastoma.