|Home | About | Journals | Submit | Contact Us | Français|
The chemokine RANTES (regulated upon activation normal T cell expressed and secreted) is expressed “late” (3 to 5 days) after activation in T lymphocytes. In order to understand the molecular events that accompany changes in gene expression, a detailed analysis of the interplay between transcriptional machinery and chromatin on the RANTES promoter over time was undertaken. Krüppel-like factor 13 (KLF13), a sequence-specific DNA binding transcription factor, orchestrates the induction of RANTES expression in T lymphocytes by ordered recruitment of effector molecules, including Nemo-like kinase, p300/cyclic AMP response element binding protein (CBP), p300/CBP-associated factor, and Brahma-related gene 1, that initiate sequential changes in phosphorylation and acetylation of histones and ATP-dependent chromatin remodeling near the TATA box of the RANTES promoter. These events recruit RNA polymerase II to the RANTES promoter and are responsible for late expression of RANTES in T lymphocytes. Therefore, KLF13 is a key regulator of late RANTES expression in T lymphocytes.
After T lymphocytes are triggered by specific antigen, they enter a several-day maturation period involving commitment, proliferation, and terminal differentiation (10). Commitment depends upon binding of the antigen-specific T-lymphocyte receptor to peptide in the context of self-major histocompatibility complex (HLA) and is accompanied by immediate induction of a number of genes, including proto-oncogenes and interleukins (25). By 2 days after stimulation, T lymphocytes synthesize new DNA and proliferate. By 3 to 5 days after activation, additional genes involved in terminal differentiation, such as those mediating cytolytic effector functions of cytotoxic T lymphocytes, are expressed. Although a great deal is understood about early events after T-cell activation, mechanisms regulating the later events associated with terminal differentiation are much less clear. Using a variety of techniques, we identified a number of genes, including 519 (granulysin) (19), WP34 (22), Tactile (CD99) (58), and RANTES (46), that are expressed 3 to 5 days after activation of T lymphocytes.
RANTES (regulated upon activation normal T cell expressed and secreted) was identified by subtractive hybridization and noted to be a member of a “new” gene family, now designated chemokines (chemoattractant cytokines) (46). RANTES is a C-C chemokine (46) that attracts and activates a myriad of inflammatory cells, including T cells, monocytes (45), eosinophils (23), basophils (11), and natural killer cells (53). RANTES is also associated with human immunodeficiency virus resistance, as its receptor, CCR5, is a coreceptor for human immunodeficiency virus uptake into inflammatory cells (9, 37). Thus, RANTES is an important therapeutic target for inflammatory disease and AIDS.
Although RANTES is expressed within minutes after activation of fibroblasts, epithelial cells, and monocytes/macrophages under the control of Rel proteins, p50, and p65 (43), T lymphocytes do not express RANTES until 3 to 5 days after activation (42, 46). In order to define the molecular basis of “late” expression of RANTES in T lymphocytes, we characterized the RANTES promoter (41) and identified a novel transcription factor regulating RANTES expression in T lymphocytes (48). This factor, originally designated RANTES factor of late activated T lymphocytes 1 (RFLAT-1), is now called Krüppel-like factor 13 (KLF13). Human KLF13, a 288-amino-acid polypeptide of the KLF family, contains three C2H2 zinc fingers responsible for DNA binding (49). Members of the KLF family regulate target gene expression by recruitment of coactivators and corepressors into large molecular complexes (11, 12, 31). Recently, p300/cyclic AMP response element binding protein (CBP) and p300/CBP-associated factor (PCAF), two proteins with histone acetyltransferase (7, 30, 32, 39) activity, were shown to physically interact with the zinc finger domain of KLF13 and to stimulate the DNA binding activity of KLF13 both singly and cooperatively in vitro (50, 51). Based on this information, the present study was undertaken to begin to delineate the molecular interplay between transcriptional machinery and chromatin in regulating “late” expression of RANTES in T lymphocytes.
Anti-KLF13 (48) and anti-Nemo-like kinase (anti-NLK) were generated as polyclonal antisera in rabbits. Additional antibodies were anti-p300/CBP (05-267), anti-p300 (05-257), anti-PCAF (07-141), anti-acetyl-histone H3 (Lys14) (06-760), anti-phospho-histone H3 (Ser10) (06-570), anti-histone H3 (05-499), and anti-α-actinin (05-384) (Upstate, Lake Placid, NY); anti-CBP (sc-369), anti-KLF13 (sc-9605), anti-Brg-1 (sc-10768), and anti-RNA polymerase II (sc-9001x) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA); anti-acetyl-histone H4 (Lys 8) (ab1760-100) (Abcam, Cambridge, MA); anti-V5 (A7345) (Sigma-Aldrich, St. Louis, MO); and anti-RANTES (Pierce, Rockford, Ill.).
Plasmids used included the following: pGL3-RP-luc, constructed by inserting bp −195 to +54 of the RANTES promoter into pGL3 (Promega, Madison, WI); pREP4-RP-luc, constructed by inserting bp −195 to +54 of the RANTES promoter into the pREP4 plasmid (a gift from Keji Zhou at National Institutes of Health, Bethesda, Md.); pREP4-ΔA-RP-luc, constructed by subcloning the insert from the previously described mutant of the A site of the RANTES promoter (38) into the XhoI-HindIII sites of pREP4-luc; pBJ5-Brg-1 plasmid (a gift from Jerry Crabtree, Stanford University); pcDNA3.1(+) (Invitrogen, Carlsbad, CA); pcDNA-KLF13, described previously (49); and pcDNA/V5/His-NLK, constructed by inserting the full-length cDNA of NLK into pcDNA 3.1/V5/His (Invitrogen).
Human peripheral blood T lymphocytes were prepared from LeukoPacs (Stanford Blood Bank, Stanford, CA) by negative selection (RosetteSep) according to the manufacturer's protocol (StemCell Technologies, Vancouver, BC, Canada). Purified T lymphocytes or Jurkat cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, 1 mM glutamine, and 100 U/ml penicillin-streptomycin and stimulated with 5 μg/ml phytohemagglutinin (PHA) for up to 7 days at 37°C, 5% CO2. HeLa cells and SW13 cells (ATCC) were maintained in Dulbecco's modified eagle medium containing 10% bovine calf serum, 1 mM glutamine, and 100 U/ml penicillin-streptomycin at 37°C, 5% CO2.
RANTES protein secreted into the cell culture supernatant was quantified by enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's protocol (Pierce, Rockford, IL).
Total RNA was prepared using the RNeasy kit (QIAGEN, Valencia, CA). cDNA was made using Superscript II with random primers (Invitrogen). Primers for human RANTES (forward primer, 5′-AGTCGTCTTTGTCACCCGAAA-3′; reverse primer, 5′-AGCTCATCTCCAAAGAGTTGATGTAC-3′) were purchased from Elim Biopharmaceuticals, Inc. (Hayward, CA). Primers for β-glucuronidase (Applied Biosystems, Foster City, CA) were used as an internal control. Primers for gamma interferon (IFNG; Applied Biosystems) were used as a negative control in the small interfering RNA (siRNA) experiment. PCR was performed in triplicate using SYBR Green or TaqMan Universal PCR master mix (Applied Biosystems) with a GeneAmp 7900 sequence detection system (Applied Biosystems) for 40 cycles of PCR under the following conditions: 2 min at 50°C and 10 min at 95°C for 1 cycle, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C. The expression of the RANTES gene was represented as the fold increase (2−ΔΔCt), where ΔΔCt = [ΔCt(stimulated)] − [ΔCt(unstimulated)] and ΔCt = [Ct(sample)] − [Ct(Gus)].
The RANTES transcription rate was measured by nuclear run-on analysis. Resting and PHA-activated T lymphocytes (2 × 107 cells) were harvested, and nuclei were isolated in hypotonic buffer (10 mM Tris, pH 7.6, 10 mM NaCl, 3 mM MgCl2, 0.5% NP-40). Isolated nuclei were incubated with 25 mM HEPES, pH 7.5, 2.5 mM MgCl2, 2.5 mM dithiothreitol (DTT), 75 mM KCl, 5% glycerol, 2.8 mM ATP, 2.8 mM GTP, 2.8 mM CTP, 0.003 mM UTP, and 500 μCi [γ-32P]UTP (New England Nuclear, Wilmington, DE) for 30 min at 37°C. After in vitro transcription, total RNA was isolated and hybridized using a slot blot to RANTES and actin cDNA and to pUC18 plasmid DNA that had been immobilized on nylon membranes using ExpressHyb (Clontech). After washing, the membranes were exposed to X-ray film (Kodak Biomax MS).
Double-stranded siRNA oligonucleotides directed against KLF13 mRNA (5′-CCUCAGGUGUCAAAGUAAAdTdT-3′) and nonsilencing siRNA (5′-UUCUCCGAACGUGUCACGUdTdT-3′) were purchased from QIAGEN. A total of 5 × 106 T lymphocytes were nucleofected with 1.5 μg KLF13 or nonsilencing siRNA using the Amaxa nucleofector system (Cologne, Germany) according to the manufacturer's protocol (program U-14). Cells were plated in six-well plates in RPMI plus 10% fetal bovine serum and incubated at 37°C, 5% CO2. After 5 h, PHA (5 μg/ml) was added and cells were incubated for an additional 48 h and harvested for use in Western blotting assays, real-time quantitative PCR, and ELISAs.
Dimethyl sulfate (DMS) was added to cell cultures (5 μl of DMS/ml of culture medium) and incubated for 2 min at room temperature. After washing cells with phosphate-buffered saline, 2 ml of stop buffer (20 mM Tris-HCl, pH 8.0, 20 mM NaCl, 20 mM EDTA, 1% sodium dodecyl sulfate, 600 μg/ml proteinase K) was added to each tube and incubated at 37°C for 3 h. Genomic DNA was purified and analyzed by ligation-mediated PCR (LM-PCR) as described elsewhere (40) using the following primer set: primer 1 (located at bp −332), 5′-TAACTGCCACTCCTTGTTGTCC-3′, for the first-strand reaction; primer 2 (located at bp −311), 5′-CCCAAGAAAGCGGCTTCCTGCTCTC-3′, for 15 cycles of PCR amplification; primer 3 (located at bp −294), 5′-CTGAGGAGGACCCCTTCCCTGGAAGGTA-3′, for the labeling reaction. The products were analyzed on 8% acrylamide-urea gels. After electrophoresis, the gel was dried and exposed to film overnight at −80°C using intensifying screens.
A chromatin immunoprecipitation (ChIP) assay was performed using the ChIP-IT kit (Active Motif, Carlsbad, CA), following the manufacturer's instructions using 2 × 107 human T lymphocytes per condition and specific antibody. For the chromatin reimmunoprecipitation (re-ChIP) assays, the chromatin complexes were eluted from the first ChIP with 10 mM DTT at 37°C for 30 min and diluted 20 times with ChIP dilution buffer (1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl, pH 8.1, and 150 mM NaCl) and reimmunoprecipitated with a second antibody. Input DNA and DNA immunoprecipitated with either specific antibodies or immunoglobulin G (IgG) were PCR amplified using primers flanking the proximal RANTES promoter and the TATA box from bp −209 to +100 (−209 primer, 5′-CACCATTGGTGCTTGGTCAAAGAGG-3′; +100 primer, 5′-GCAGTAGCAATGAGGATGACAGCGA-3′). Reaction mixtures were cycled with an initial melt step at 94°C for 5 min and then 24 to 30 cycles of 94°C for 45 seconds, 56°C for 30 seconds, and 72°C for 60 seconds, followed by 72°C for 10 min. Products were analyzed by electrophoresis on a 2% gel. As a negative control, primers corresponding to a genomic region distal to the RANTES promoter from −3789 to −3459 were used (5′ primer, −3789, 5′-GCAGATTACGAGGTCAGGAG-3′; 3′ primer, −3459, 5′-TTATGCTTTTCAACAGTCT-3′).
Nuclear extracts were prepared according to the manufacturer's protocol (Transfactor extraction kit; Clontech, Mountain View, CA). Western blots with nuclear extracts or immunoprecipitates were detected using ECL reagent (Amersham Pharmacia Biotech, Piscataway, NJ).
In vivo cross-linking and immunoprecipitation were performed as described previously (28) with the following modifications: 5-day-activated T lymphocytes were cross-linked by the addition of 0.6 mg dithio-bis(succinimidylpropionate) [DSP; Pierce]/ml for 15 min at room temperature. For IP, 1 mg of nuclear extract was mixed with 2 or 5 μg of antibody and rotated overnight at 4°C. Protein A/G beads were added and incubated for an additional 4 h at 4°C with rotation. The beads were then pelleted and washed five times with 1 ml of IP wash buffer (20 mM Tris-HCl, pH 7.5, 0.5 M NaCl, 1 mM EDTA, 0.2% Triton X-100). The beads were incubated in 40 μl of 2× sodium dodecyl sulfate (SDS) loading buffer (0.1 M Tris, pH 6.8, 4% SDS, 10% glycerol, 0.2 M DTT) for 10 min at 95°C. After electrophoresis, samples were analyzed by Western blotting.
HeLa cells were transiently transfected with either pcDNA3.1/V5/His vector or pcDNA/V5/His-NLK using Lipofectamine 2000 (Invitrogen) following the manufacturer's instructions. Cells were lysed 48 h after transfection, and 500 μg of lysate was subjected to immunoprecipitation with anti-V5 antibody (Sigma). To determine whether NLK phosphorylates histone H3 protein, aliquots of immunoprecipitant were incubated at 30°C with 10 mM HEPES (pH 7.4), 5 mM MgCl2, 5 mM DTT, 1 μCi [γ-32P]ATP, and 1 μg of recombinant histone H3 (Upstate) for 15 min in a final volume of 25 μl. The reaction was stopped by addition of SDS-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer, and the samples were subjected to SDS-PAGE. The γ-32P-labeled histone H3 substrate was visualized by autoradiography. To identify the specific phosphorylation sites of NLK on histone H3, 1 μg of synthetic H3 peptide and 0.2 mM unlabeled ATP were used instead of recombinant histone H3 and [γ-32P]ATP in the kinase reaction mixture. Histone H3 peptides (United Biochemical Research, Inc., Seattle, WA) that included residues 1 to 27 (ARTKQTARKSTGGKAPRKQLASKAARK) or residues 1 to 35 (ARTKQTARKSTGGKAPRKQLASKAARKSAPATGGV) were used to determine serine 10 and serine 28 phosphorylation, respectively. Samples were separated by 20% SDS-PAGE and subjected to Western blotting with anti-phospho-histone H3 (serine 10 or serine 28) antibody following the in vitro kinase reaction.
Histone acetyltransferase (HAT) activity was assayed in nuclear extracts using a HAT assay kit (Upstate).
Transient transfection was carried out using Superfect (QIAGEN) transfection reagent according to the manufacturer's protocols. Luciferase was measured with the Dual Luciferase assay kit (Promega) following the manufacturer's instructions. Luciferase activity was measured over 30 s in an EG&G Lumat LB 9507 luminometer.
T lymphocytes were purified from peripheral blood and activated with the mitogen PHA. RANTES mRNA (Fig. (Fig.1A)1A) and protein (Fig. (Fig.1B)1B) were measured in resting T lymphocytes and through 7 days after activation. Since most RANTES is secreted, an ELISA was used to measure RANTES protein levels. In resting and day 1 activated T lymphocytes, small amounts of RANTES mRNA and protein were observed (Fig. 1A and B), attributable to circulating memory T lymphocytes (56). In order to differentiate between the accumulation of stable mRNA transcripts over time versus new transcription, nuclear run-on assays were performed. As shown in Fig. Fig.1C,1C, RANTES transcripts were detected at day 3 and increased through 7 days after activation compared to negative pUC18 and positive actin controls. Thus, both mRNA and protein levels increased significantly on days 3 to 7 after activation, confirming transcriptional regulation of RANTES (48). The RANTES promoter includes subregions G, C, D, E, A, and B within the first 219 bp upstream of the transcriptional start site (Fig. (Fig.1D).1D). These regions were named based upon their order of discovery (38, 48). Deletion or replacement of the A site results in a 90% loss of RANTES promoter activity by reporter gene assay in Jurkat T lymphocytes (38). We previously showed that KLF13 binds to this site and regulates RANTES expression in vitro (48).
In order to further demonstrate the role of KLF13 in regulating RANTES expression in T lymphocytes, KLF13-specific siRNA was transfected into T lymphocytes and expression of RANTES was then measured. As shown in Fig. Fig.1E,1E, both KLF13 mRNA and protein were suppressed at 48 h after transfection of KLF13-specific siRNA and activation of T lymphocytes. As a result, RANTES mRNA and protein, as measured by ELISA, were also suppressed (Fig. (Fig.1E).1E). However, KLF13-specific siRNA did not suppress IFNG expression (Fig. (Fig.1E),1E), demonstrating the specificity of the KLF13 siRNA. Thus, KLF13 is an important transcription factor for RANTES expression in T lymphocytes.
To further demonstrate the in vivo significance of KLF13 and its binding to the A site of the RANTES promoter (Fig. (Fig.2A),2A), in vivo DMS-mediated footprinting coupled with LM-PCR (Fig. (Fig.2B)2B) and a ChIP assay (Fig. (Fig.2C)2C) were performed. Individual bands were resolved in the sequencing gel by electrophoresing DMS and piperidine-treated LM-PCR samples (Fig. (Fig.2B).2B). Enhanced DMS reactivities were identified at guanines at bp −58 and −59 upstream from the start site of transcription. The enhanced bands at bp −58 and −59 change after activation. Because the KLF13 and p50 binding sites overlap (Fig. (Fig.2A),2A), a ChIP assay was required to determine which protein(s) binds to this site at various days after activation. Furthermore, since DMS reacts with purines to form N3-methyladenine in addition to N7-methylguanine, the increased DMS reactivity of adenine bands at the TATA Box (Fig. (Fig.2B)2B) suggests that RNA polymerase II can bind to the sites. Therefore, ChIP was performed with antibodies recognizing KLF13, p50, and RNA polymerase II (Fig. (Fig.2C).2C). DNA was amplified with primers from bp −209 to +100, which includes both the A site and the TATA box. KLF13, p50, and RNA polymerase II bind to the RANTES promoter over time after activation, as demonstrated in Fig. Fig.2C.2C. Some KLF13 is detectable on day 1, but levels markedly increase on days 3 to 7. p50 binds strongly on day 1, but levels are markedly decreased by day 3. RNA polymerase II is detectable on days 3 to 7, coincident with the kinetics of RANTES expression (Fig. (Fig.1A).1A). No PCR products were amplified using control antibody (IgG) or distal primers (Fig. (Fig.2C),2C), demonstrating specificity.
Thus, in vivo footprinting and ChIP assays demonstrated that KLF13 interacts with the A site of the RANTES promoter in vivo and that the pattern of binding changes over time after activation. This indicates that the interaction of KLF13 with the RANTES promoter is dynamic and involves binding of other factors over time. Further experiments were undertaken to identify additional factors associating with KLF13 at the RANTES promoter as well as the molecular mechanisms underlying cofactor accessibility.
We evaluated the interaction of KLF13 with known mitogen-activated protein (MAP) kinases and found that NLK interacts with KLF13 (S. Chou and A. M. Krensky, unpublished data). Expression of NLK in nuclear extracts from resting and PHA-activated (days 1, 3, 5, and 7) T lymphocytes was evaluated by Western blotting (Fig. (Fig.3A).3A). NLK is expressed in resting T lymphocytes and at all times after activation but is increased in expression 3 to 7 days after activation. Reciprocal in vivo coimmunoprecipitations of KLF13 and NLK demonstrated an interaction between these two proteins in T lymphocytes (Fig. (Fig.3B).3B). This suggests that NLK can be recruited to the RANTES promoter in activated T lymphocytes. To demonstrate that KLF13 and NLK are present as part of a complex on the RANTES promoter, PHA-activated T lymphocytes were subjected to sequential ChIP, known as re-ChIP, first with an anti-KLF13 antibody and next with either an NLK antibody or rabbit IgG and vice versa (Fig. (Fig.3C).3C). Both NLK and KLF13 are present on the RANTES promoter as soon as 1 day after activation. No PCR products were amplified using control antibody (IgG) or distal primers, demonstrating specificity.
To determine a role for NLK recruited to the RANTES promoter, an in vitro kinase assay using [γ-32P]ATP and recombinant histone H3 substrate was performed (Fig. (Fig.3D,3D, upper panel). Lysates of HeLa cells transiently transfected with a V5-tagged NLK expression vector or empty vector were subjected to immunoprecipitation using anti-V5 antibody. Recombinant human histone H3 was phosphorylated by immunoprecipitated V5-NLK but not by control immunoprecipitates (Fig. (Fig.3D,3D, upper panel). Since there are two phosphorylation sites on histone H3, Western blotting was used to demonstrate that serine 10, and not serine 28, is phosphorylated by NLK. The in vitro kinase reaction was repeated as above, except that unlabeled ATP and synthetic histone H3 peptides were used. A band was detected with antibody specific for phosphorylated serine 10 of histone H3 [P-H3 (S10)] (Fig. (Fig.3D,3D, middle panel) but not with an antibody specific for serine 28 (not shown). Equal protein loading was demonstrated by stripping and reprobing the blot with an anti-histone H3 antibody (Fig. (Fig.3D,3D, bottom panel). Furthermore, a ChIP assay performed using the P-H3 (S10) antibody specific for phosphorylated serine 10 of histone H3 indicated phosphorylation of histone H3 at serine 10 (Fig. (Fig.3E).3E). The ratio of PCR products obtained from immunoprecipitated chromatin and input chromatin reflects the relative level of histone phosphorylation. Highly phosphorylated histone was observed on days 1 and 7 after activation, with lower levels detected on days 3 and 5. Phosphorylation on day 7 may be due to NLK or to another as-yet-unidentified kinase. Thus, KLF13 is associated with a MAP kinase family member (NLK or others) that phosphorylates histone H3 near the RANTES promoter on day 1 after activation.
Since HAT activity is often associated with opening of chromatin structure and initiation of transcription, we investigated the interaction of KLF13 with p300/CBP and PCAF. Expression of p300, CBP, and PCAF in nuclear extracts from resting and PHA-activated (days 1, 3, 5, and 7) T lymphocytes was evaluated by Western blotting (Fig. (Fig.4A).4A). Both p300 and CBP were markedly increased in expression in nuclear extracts from T lymphocytes 3 to 7 days after activation. PCAF expression progressively increased from days 1 through 7 after T-lymphocyte activation compared to resting cells. A HAT-linked ELISA was used to directly measure HAT activity in the same samples (Fig. 4B and C). HAT activity peaked on day 5 and then decreased, using both histone H3 (Fig. (Fig.4B)4B) and histone H4 (Fig. (Fig.4C)4C) as substrates. KLF13 coprecipitates p300/CBP and PCAF in nuclear extracts from day 5 PHA-activated T lymphocytes (Fig. 4D and E, upper panels). Reciprocal experiments using antibodies specific for p300/CBP or PCAF coprecipitated KLF13 (Fig. 4D and E, lower panels). Thus, KLF13 interacts with both p300/CBP and PCAF in nuclear extracts of activated T lymphocytes.
ChIP and re-ChIP assays were conducted to determine whether KLF13, p300, CBP, and PCAF form a complex at bp −209 to +100 of the RANTES promoter (Fig. 5A and B). In resting T lymphocytes, no PCR products were detected following immunoprecipitation with antibodies specific for KLF13, p300, CBP, or PCAF, indicating that none of these factors is associated with this region of the RANTES promoter in unactivated cells (Fig. (Fig.5B).5B). After activation, only KLF13 is present at day 1 and throughout activation, while p300, CBP, and PCAF are only present on days 3 to 7 (Fig. (Fig.5B).5B). No PCR products were amplified using control antibody (IgG) or distal primers (Fig. (Fig.5B),5B), demonstrating specificity. These results suggest that KLF13 may recruit these cofactors to the RANTES promoter.
Since acetylation of histone H4 and histone H3 by p300/CBP and PCAF recruits protein complexes required for transcriptional initiation (1, 47), acetylation of histones associated with bp −209 to +100 of the RANTES promoter was evaluated by ChIP assay using antibodies specific for acetylated lysine 14 in histone H3 [Ac-H3 (K14)] (Fig. (Fig.5C)5C) or acetylated lysine 8 in histone H4 [Ac-H4 (K8)] (Fig. (Fig.5D).5D). The ratio of PCR products obtained from immunoprecipitated chromatin and input chromatin reflects the relative level of histone acetylation. Although lysine 14 of histone H3 was most strongly acetylated on day 5 after activation, there was also significant acetylation detected on days 1, 3, and 7 (Fig. (Fig.5C).5C). Lysine 8 of histone H4 was most strongly acetylated on day 3, although acetylation was also evident in resting T lymphocytes and at other days after activation (Fig. (Fig.5D).5D). These results indicate that KLF13 binds and/or recruits both p300/CBP and PCAF to the RANTES promoter, where one or both may participate in acetylation of associated histone proteins.
Brg-1, an ATPase involved in chromatin remodeling, is recruited to chromatin by direct interactions with DNA binding proteins (3, 4). Expression of Brg-1 in nuclear extracts from resting and PHA-activated (days 1, 3, 5, and 7) T lymphocytes was evaluated by Western blotting (Fig. (Fig.6A).6A). Brg-1 was present at all time points, but expression increased slightly on days 5 and 7 after T-lymphocyte activation compared to resting cells. To determine whether KLF13 could recruit initiators of ATP-dependent chromatin remodeling to the RANTES promoter, the interaction of KLF13 and Brg-1 was evaluated by coimmunoprecipitation assay. KLF13 and Brg-1 reciprocally coprecipitate each other (Fig. (Fig.6B).6B). To demonstrate that the KLF13-Brg-1 complex is present on the RANTES promoter, PHA-activated T lymphocytes were subjected to ChIP and re-ChIP (Fig. (Fig.6C).6C). In vivo-cross-linked chromatin was first immunoprecipitated with either anti-KLF13 or anti-Brg-1 antibody, and then these complexes were eluted and reprecipitated with the reciprocal antibody.Immunoprecipitation with rabbit IgG was used as a control (Fig. (Fig.6C).6C). KLF13 was evident at days 1 to 7, while Brg-1 was present weakly on day 3 and increased on days 5 to 7, suggesting that KLF13 may recruit Brg-1 to the RANTES promoter. No PCR products were amplified using control antibody (IgG) or distal primers (Fig. (Fig.6C),6C), demonstrating specificity.
Since Brg-1-mediated ATP-dependent chromatin remodeling requires “proper” chromatin structure, a 249-bp segment of the RANTES promoter was cloned into pREP4 plasmid (pREP4-RP-luc). This vector forms “proper” chromatin structure when transiently transfected into SW13 cells (28, 29, 55). The same 249-bp segment of the RANTES promoter was also cloned into the pGL3-basic vector (pGL3-RP-luc) and transiently transfected into SW13 cells as a control. As shown in Fig. Fig.6D6D (left panel), Brg-1 can only activate the pREP4-RP-luc plasmid and not pGL3-RP-luc, indicating that the RANTES promoter requires “proper” chromatin structure to activate transcription of RANTES. In order to test synergy of Brg-1 and KLF13 in transcription, reporter gene assays were performed with pREP4-RP-luc in which Brg-1 and KLF13 expression plasmids were transfected into SW13 cells alone or together (Fig. (Fig.6D,6D, right panel). Activation of pREP4-RP-luc was significantly increased only when both KLF13 and Brg-1 were expressed, indicating synergy between these two proteins. These data strongly indicate that KLF13 is involved in ATP-dependent chromatin remodeling of the RANTES promoter. In addition, to demonstrate that KLF13 binding to the RANTES promoter is essential for KLF13-Brg-1 synergy, the A site of the RANTES promoter was mutated and tested in reporter gene assays in SW13 cells (Fig. (Fig.6D,6D, right panel). Reporter constructs containing mutated KLF13 binding DNA sequences showed no activation with KLF13 or Brg-1 alone or in combination, indicating that the KLF13 binding element is a bridge for the KLF13-Brg-1 effect on ATP-dependent chromatin remodeling at the RANTES promoter.
RANTES is expressed within minutes after activation of fibroblasts, monocytes, and other cells but is not expressed until 3 to 5 days after activation in T lymphocytes. This late expression of RANTES in T lymphocytes serves a kinetic bridging function in the generation of an inflammatory response. Stromal and/or other cells resident in tissues sense danger or damage and express Rel proteins that induce RANTES expression in these cell types. In contrast, Rel proteins alone are not enough to induce expression of RANTES in T lymphocytes. Chemokines, including RANTES, draw T lymphocytes into tissues at the site of inflammation. T lymphocytes become activated if and when they sense specific antigen in the context of self-major histocompatibility complex via their clonally expressed T-cell receptors. Nevertheless, it takes three to five days for T lymphocytes to express RANTES. This serves to amplify the immune response in time and space, but only if specific antigen capable of activating a T lymphocyte is present. We previously showed that KLF13 is required for late expression in T lymphocytes (48). In order to better understand the molecular mechanism of late expression of RANTES in T lymphocytes, here we further explored the function of KLF13 as the lynchpin for recruitment of a number of coactivators and other factors involved in chromatin remodeling and gene transcription.
We first identified KLF13 (also known as RFLAT-1, FKLF2, and BTEB3), a member of the Krüppel-like transcription factor family, by expression cloning through its binding to the A site of the RANTES promoter (48). Subsequently, others have shown that KLF13 also can activate the human γ-globin promoter and other erythroid-specific genes, simian virus 40 (SV40), and SM22α promoters in vitro (2, 33). In vitro DNA binding studies demonstrated that KLF13 binds to the A site of the RANTES promoter, a consensus basic transcription element, and to the CACCC box of the γ-globin promoter (2, 48). Nevertheless, since in vitro assays may not reflect the in vivo cellular events actually involved in gene regulation, ChIP assay and in vivo footprinting were performed to verify the in vivo significance of KLF13 binding to the A site of the RANTES promoter. Song et al. reported that acetylation of KLF13 by p300/CBP disrupts KLF13 DNA binding activity (50). However, PCAF inhibits this p300/CBP-mediated disruption of KLF13 DNA binding (50). These previous observations may explain the changes at bp −58 and −59 in Fig. Fig.2B.2B. Binding is dynamic, with other proteins entering and exiting the complex on the RANTES promoter over time after activation. Because KLF13 and p50 (subunit of NF-κB) share the same binding site (CTCCC) within the A region of the RANTES promoter, other factors, or signal transduction pathways, may determine which protein(s) binds the site. Since Notch signaling reduces NF-κB DNA binding activity in T lymphocytes (57) and recruits p300/CBP after T-lymphocyte activation (13, 15), it may impact late RANTES gene expression in T lymphocytes. The RANTES promoter contains three putative binding sites around bp −650 for CSL [CBF1/RBP-Jκ, Su(H), Lag-1], a nuclear effector of Notch signaling (5). The relationship between the Notch signaling pathway and RANTES expression in T lymphocytes is currently under investigation. The in vivo footprinting and ChIP/re-ChIP results demonstrate that KLF13 and p50 bind to DNA as early as day 1 after activation, while other proteins interact with KLF13 3 to 7 days after activation. Thus, we hypothesize that KLF13 binds the A site of the RANTES promoter in the packaged nucleosome and recruits additional proteins that change chromatin structure.
KLF13 consists of three zinc fingers as well as activation and repression domains (49) through which interactions with coactivators, such as p300/CBP and PCAF, or corepressors, such as Sin3A and histone deacetylase 1, have previously been shown to occur (20, 51). In order to investigate the positive role of KLF13 in late expression of RANTES in T lymphocytes, we focused on its interaction with the known histone acetyltransferase coactivators p300/CBP and PCAF. The zinc finger domain of KLF13 appears to have multiple functions, including nuclear localization, DNA binding, and interaction with coactivators p300/CBP and PCAF (50). p300/CBP and PCAF have HAT activity (7, 30, 32, 39), which transfers an acetyl group to the -amino group of a lysine residue. The acetylation state of chromatin has been established as a key mechanism for opening the silent chromatin structure, which leads to increased transcription (6). Sequential ChIP (re-ChIP) assays and in vivo coimmunoprecipitation assays show that p300/CBP and PCAF form a complex on the RANTES promoter via KLF13 binding, thereby acetylating histone proteins and opening compact chromatin structure, permitting RANTES expression. In addition, acetylation of KLF13 by p300/CBP and PCAF regulates KLF13 DNA binding activity (50). Thus, p300/CBP and PCAF control KLF13 binding to the RANTES promoter and help open compact chromatin structure.
The repression domains of KLF13 have a high content of hydrophobic residues and can interact with the paired amphipathic helix 2 domain of Sin3A (20, 31). Sin3A recruits proteins with histone deacetylase and histone methyltransferase, resulting in more compact chromatin structure. This decreases accessibility for transcriptional activators and basal promoter factors, leading to an inhibition of transcription. Thus, KLF13 can function either as a transcriptional activator or repressor, depending upon its protein associations. Of particular note, in recent in vivo studies using whole-animal gene disruption, we showed that KLF13 can function as either a positive or negative regulator for different genes (M. Zhou, D. Feng, A. Song, C. Dong, L. McPherson, S.-C. Lyu, L. Zhou, X. Shi, Y.-T. Ahn, D. Wang, C. Clayberger, and A. M. Krensky, submitted for publication). KLF13 also represses low-density lipoprotein receptor promoter activity (36). In addition, we also showed that NLK, a MAP kinase that functions downstream of transforming growth factor β-activated kinase 1 in the Wnt pathway (18, 34), can interact with KLF13 in vivo in T lymphocytes. NLK is a coactivator or corepressor in Wnt signaling, depending on the partner proteins it binds (16, 54, 59). Wnt signaling is associated with proliferation and survival of T lymphocytes and activates T-cell factor/lymphocyte enhancer factor 1 (TCF-1/Lef-1) transcription (52). The RANTES promoter contains three putative TCF-1/Lef-1 binding sites at around bp −300. Therefore, Wnt signaling may modify RANTES expression.
Finally, we have demonstrated that there is an interaction between KLF13 and Brg-1, a component of the SWI/SNF family of ATP-dependent chromatin remodeling complexes (35). Since Brg-1 lacks a sequence-specific DNA binding domain, the selective recruitment of the ATP-dependent remodeling complex to target genes requires interaction with certain gene-specific transcription factors that bind Brg-1 (21), such as KLF family members, erythroid Krüppel-like factor (KLF-1) (27), and SP1 (28), β-catenin (3), CCAAT/enhancer binding protein β (24), and glucocorticoid receptor (14). These interactions facilitate ATP-dependent chromatin remodeling as a prerequisite for transcriptional activation. KLF13 is a good candidate for a similar interaction with Brg-1 and the ATP-dependent chromatin remodeling apparatus. ChIP/re-ChIP assays and coimmunoprecipitation of Brg-1 and KLF13 suggest that KLF13 regulates ATP-dependent chromatin remodeling of the RANTES promoter, implying a role for KLF13 in chromatin configuration for active transcription. Of note, deletion of the A site of the RANTES promoter severely reduces luciferase activity in KLF13-meditated transcription assays. These data provide the link between KLF13 and ATP-dependent chromatin remodeling.
Covalent histone modifications (e.g., acetylation, methylation, phosphorylation, and ubiquitination) play a vital role in regulating chromatin functional states (4). These types of posttranslational modifications result in changes in the electrostatic charge of histones and/or conformational changes that expose binding surfaces containing protein recognition modules such as bromo- and chromo-domains (17). Phosphorylation of histone H3 at serine 10 can promote acetylation of lysine 14 of the same histone H3 by p300/CBP or PCAF (30, 60). Our data indicate that NLK, like other MAP kinases, phosphorylates histone H3 at serine 10 on the RANTES promoter (8, 44). In this manner, NLK could have a positive regulatory role in RANTES expression. NLK also induces methylation of histone H3 at lysine 9 at Myb-bound promoter regions (26).
Acetylation of lysine 14 on histone H3 and acetylation of lysine 8 on histone H4 are involved in recruitment of other activators and coactivators for transcription initiation (1). The ChIP assay demonstrated an increase of acetylated histone H3 at lysine 14 on day 5 as well as acetylated histone H4 at lysine 8 on days 5 and 7 after T-lymphocyte activation. Therefore, p300/CBP and PCAF, recruited by KLF13, appear to acetylate histone proteins on the RANTES promoter. In addition, HAT expression and activity in nuclear extracts from activated T lymphocytes parallel the observed hyperacetylation of histone proteins on days 5 and 7 after activation.
Based on the nuclear run-on assays for RANTES transcription, new transcripts are made through day 7 and the rate of transcription is actually greatest on day 7. Although histone acetylation levels are decreased between days 5 and 7, increased levels of RANTES transcripts in T lymphocytes suggest that other modifications or mechanisms not yet identified may also be involved in RANTES transcription.
Figure Figure77 summarizes our current model of the events regulating RANTES transcription in T lymphocytes. First, KLF13 binds its core element within the A site of the RANTES promoter in the compacted nucleosome. NLK binds KLF13 at the RANTES promoter, leading to phosphorylation of histone H3 at serine 10. This phosphorylation enables acetylation of lysine 14 on histone H3 by p300/CBP or PCAF, which are recruited by direct binding to KLF13. This acetylation leads to the recruitment of additional factors required for transcriptional initiation. p300/CBP or PCAF also acetylates histone H4 at lysine 8 and initiates ATP-dependent chromatin remodeling at the A site of the RANTES promoter with association of KLF13 and Brg-1. ATP-dependent chromatin remodeling twists and deforms the chromatin, exposing the adjacent TATA box and enabling initiation of transcription. Thus, ordered recruitment of factors to the RANTES promoter by KLF13 and associated proteins leads to posttranslational modification and ATP-dependent chromatin remodeling regulating expression of RANTES “late” after activation in T lymphocytes.
This study was supported by NIH grant R37 DK35008. A.M.K. is the Shelagh Galligan Professor.
Published ahead of print on 30 October 2006.