IL-10 promoter transactivation by the viral K-RTA protein involves the host-cell transcription factors, specificity proteins 1 and 3

Kaposi's sarcoma-associated herpesvirus (KSHV)/human herpesvirus-8 (HHV-8) causes a persistent infection, presenting latent and lytic replication phases during its life cycle. KSHV-related diseases are associated with deregulated expression of inflammatory cytokines, including IL-6 and IL-10, but the mechanisms underlying this dysregulation are unclear. Herein, we report a molecular mechanism for KSHV-induced IL-10 gene expression. KSHV replication and transcription activator (K-RTA) is a molecular switch for the initiation of expression of viral lytic genes, and we describe, for the first time, that K-RTA significantly activates the promoter of the human IL-10 gene. Of note, mutations involving a basic region of K-RTA reduced the association of K-RTA with the IL-10 promoter. Moreover, the host-cell transcription factors, specificity proteins (SP) 1 and 3, play a pivotal cooperative role in K-RTA–mediated transactivation of the IL-10 promoter. K-RTA can interact with SP1 and SP3 directly in vitro, and electrophoresis mobility shift assays (EMSAs) revealed co-operative interaction involving K-RTA, SP1, and SP3 in binding to the IL-10 promoter. As DNase I footprinting assays indicated that K-RTA did not affect SP3 binding to the IL-10 promoter, SP3 can function to recruit K-RTA to the IL-10 promoter. These findings indicate that K-RTA can directly contribute to IL-10 up-regulation via a functional interplay with the cellular transcription factors SP1 and SP3.

scriptional mRNA processing of specific genes (25,26). In addition, IL-10 inhibits Ets2 expression and LPS-induced miR-155 expression during the inflammatory response (27). Similarly, the IL-10 pathway might enable the virus to evade host immunity (28). Interestingly, another gammaherpesvirus, EBV, and a betaherpesvirus, CMV, encode viral homologs of IL-10, which contribute to evasion from virus-specific immune responses and to the establishment of latent infections (29,30). In KSHVassociated diseases, such as PEL and MCD in patients with HIV/AIDS, elevated levels of viral cytokine IL-6 and of cellular cytokines IL-6, IL-10, stromal cell-derived factor 1, and vascular endothelial growth factor are thought to contribute to viral persistence in and pathogenesis of these virus-associated lymphoproliferative disorders (31)(32)(33)(34).
IL-10 is produced by a number of cells of the immune system, such as T-, B-, and dendritic cells and macrophages, and the IL-10 promoter is regulated by various transcriptional networks in different cell types (35). Recent interesting studies have demonstrated that infection by KSHV or expression of the viral miRNAs miR-K12-3 and miR-K12-7 induces IL-10 in macrophages through targeting the negative regulator of IL-10 transcription, C/EBP␤ (36). STAT3 activation contributes to IL-10 production by dendritic cells exposed to UV-inactivated KSHV (37). These observations support the idea that lytic replication caused by intermittent minor reactivation during latency may be associated with IL-10 production by KSHVinfected cells.
In this study, we examined the effect of K-RTA on IL-10 up-regulation and explored the possible mechanism of K-RTAmediated IL-10 promoter activation. We show, for the first time, that K-RTA significantly activates the human IL-10 promoter through association with SP1 and SP3 in KSHV-infected and non-infected human cells. We also show that specific residues such as Lys-152, Lys-154, and His-145 on K-RTA play a role in K-RTA-mediated IL-10 promoter activation. These data indicate that the functional activity of K-RTA can contribute to IL-10 induction in KSHV-associated diseases.

K-RTA expression activates IL-10 gene promoter
KSHV-infected PEL cells produce various cytokines, including IL-10 (38,39), and some PEL cells induce lytic viral replication under latent conditions (34). The KSHV-infected PEL cell line JSC-1 is reported to show higher basal and induced expression of lytic phase genes and to produce higher titers of infectious viruses than KSHV-infected BC3 cells (40). In line with those results, luciferase reporter assays confirmed that promoter activities for the lytic phase initiator ORF50/K-RTA and human IL-10 genes were higher in the KSHV-infected PEL cell line JSC-1 than in BC3 cells (Fig. 1A). Quantitative real-time PCR showed that the relative expression of IL-10 mRNA was higher in JSC-1 cells than in BC3 cells and that 12-O-tetradecanoylphorbol-13-acetate (TPA), a chemical initiator of lytic viral replication (41,42), caused an increase in IL-10 mRNA expression in both cell lines (Fig. 1B). These results suggest that lytic replication and IL-10 mRNA expression are correlated.
Various transcription factors are involved in IL-10 mRNA induction in different cell types and conditions. We examined the influence of KSHV lytic replication on IL-10 gene promoter activity, by luciferase reporter assays, using reporter plasmids (p-265) containing IL-10 promoter regions corresponding to the 5Ј-upstream sequence of the human IL-10 gene open reading frame (position Ϫ265 to Ϫ1 from the first ATG codon). The activity of the IL-10 promoter was increased after TPA treatment (a stimulus for lytic viral replication) in KSHV-infected JSC-1 cells (Fig. 1C). Expression of FLAG-K-RTA activated IL-10 promoter activity from reporter plasmids in JSC-1, BC3, and HCT 116 cells (Fig. 1D).
The N-terminal portion of K-RTA (amino acids (aa) 1-490), lacking its C-terminal transactivation domain, can bind directly to the K-RTA-responsive element (43). Transactivation domain is also responsible for binding to other transcriptional co-factors (8,11). FLAG-K-RTA (F490, containing aa 1-490) did not activate p-265 in KSHV-infected JSC-1 or non-infected HCT 116 cells (Fig. 1E). In addition, transcription of IL-10 mRNA was induced by full-length K-RTA but not by F490 in HCT 116 cells (Fig. 1F). Thus, the transactivation domain of K-RTA is required for K-RTA-mediated IL-10 mRNA induction. To examine direct binding of K-RTA to the IL-10 promoter DNA, electrophoresis mobility shift assay (EMSA) was performed. The results showed that direct binding of FLAG-K-RTA (1-531) with IL-10 promoter sequences (Ϫ265/Ϫ1) was not detected, although FLAG-K-RTA (1-531) bound to PAN promoter DNA in a dose-dependent manner (Fig. 1, G and H). These data suggested that K-RTA has the potential to transactivate the IL-10 promoter by cooperating with other factors.

Impaired DNA-binding ability of K-RTA leads to less activation of IL-10 promoter
The basic region at the N terminus of K-RTA is proposed to have a critical role in the direct binding to DNA (16). We introduced single lysine-to-arginine substitutions in the basic region of K-RTA (K94R, K124R, K142R, K152R, and K154R), and we found that Lys-152 and Lys-154 are responsible for the transactivation of IL-10 and PAN promoters (Fig. S1). Next we compared the activities of the mutant K-RTAs K152R and K154R with that of an already published mutant, H145L (44). H145L K-RTA is shown to lose the capacity to promote degradation of IRF7 as well as self-ubiquitination (44). These mutant K-RTAs were expressed at comparable levels in cells ( Fig. 2A), and luciferase reporter assays showed that the K152R, K154R, and H145L mutations caused a loss of transactivational activity of K-RTA in both PAN and IL-10 p-265 promoters (Fig. 2, B and C). Consistently, our in vitro EMSA experiments indicated that these three mutant K-RTAs lost direct DNA-binding ability to the PAN promoter DNA (Fig. S2).
To compare the loading of mutant K-RTA on IL-10 promoter DNA, extracts of cells expressing wild type or mutant K-RTAs were subjected to DNA affinity precipitation, by an avidin-biotin-coupled DNA (ABCD) assay. We found that affinity precipitation of mutant K-RTAs with p-265 DNA was clearly reduced (Fig. 2D). In addition, loading of K152R and K154R mutants on IL-10 promoter p-265 reporter plasmid was Transactivation of IL-10 promoter by K-RTA Transactivation of IL-10 promoter by K-RTA analyzed by ChIP assay. The results showed that the targeting of the K152R and K154R mutant K-RTAs to p-265 plasmid was clearly diminished (Fig. 2E). Furthermore, we compared the affinity of mutant K-RTAs for chromatin by a biochemical cell fractionation method (Fig. S3). In contrast to the case of wildtype K-RTA, significant amounts of mutant K-RTAs (K152R, K154R, and H145L) were readily extracted by low-salt buffer, and there was little mutant K-RTA in nuclear extracts (Fig. S3, B and C). Because microscopic analysis showed nuclear localization of both K152R and K154R mutant K-RTAs (Fig. 2F), the DNA-binding ability of the basic domain plays a pivotal role in chromatin retention and IL-10 promoter targeting of K-RTA, despite the fact that, in vitro, direct binding of K-RTA to the IL-10 promoter DNA was not detected (Fig. 1G). Collectively, these data suggested that K-RTA might require other co-factors for IL-10 promoter activation.
SP1 has been shown to interact with various cellular proteins (47). PBX1 also forms a complex with SP1 and activates expression of a subset of promoters containing SP1-binding sites (48). Thus, we examined the effect of the SP1/3-targeting site on K-RTA-mediated transactivation of the IL-10 promoter. The loading of K-RTA on IL-10 promoter DNA was analyzed by ABCD assay using biotin-labeled wild type or mSP IL-10 promoter DNA (fragment from Ϫ265 to Ϫ1). The results showed that co-precipitation of K-RTA was significantly decreased by mutation at the SP1/3-targeting element (from AGGAGG to GAATCC, positions Ϫ177 to Ϫ172 in mSP) (Fig. 3C). ChIP assay confirmed that the occupancy of IL-10 promoter p-265 by K-RTA was reduced by a mutation at the SP1/3-targeting element (mSP), suggesting that the SP1/3-binding site is involved in the loading of K-RTA onto the IL-10 promoter (Fig. 3D). In addition, luciferase reporter assays showed that co-expression of SP1-HA or SP3-HA with K-RTA elevated K-RTA-mediated IL-10 promoter activation (Fig. 4A), whereas siRNA-mediated knockdown of SP1 or SP3 decreased K-RTA-mediated IL-10 promoter activation (Fig. 4, B and C). Collectively, these results indicated that SP1/3 participates in K-RTA-mediated IL-10 promoter activation.
The ABCD assay using biotin-labeled wild type or mSP IL-10 promoter DNA (fragment from Ϫ265 to Ϫ1) showed that K-RTA and SP3 co-precipitated with wild-type IL-10 promoter DNA in a dose-dependent manner, but this did not occur with mSP DNA probe (Fig. 5A). In this ABCD assay, binding of SP1 to p-265 probe appeared relatively weaker than that of SP3. EMSA using immunopurified SP1-HA and SP3-HA proteins ( Fig. 5B) with an IL-10 promoter DNA probe showed that SP3, rather than SP1, preferentially binds to wild-type IL-10 promoter DNA (Fig. 5, C and D), and the formation of a specific complex of DNA probe with SP3 was decreased for mSP probe (Fig. 5, E and F). Thus, SP3, rather than SP1, binds directly to the IL-10 promoter (sequence from Ϫ265 to Ϫ1).
To study possible combinational effects of SP1, SP3, and K-RTA, EMSA was performed using immunopurified FLAG-K-RTA (aa 1-531), SP1-HA, and SP3-HA. The results showed that a specific complex of SP3 and wild-type IL-10 promoter DNA (Fig. 6A,   . Student's t test; **, p Ͻ 0.05. C, human IL-10 promoter reporter plasmid p-265 and internal control pRL-TK were transfected into JSC-1 cells. One day later, cells were treated or not with TPA (20 ng/ml) for 24 h. Luciferase activities were determined 2 days after transfection. Results are presented as the mean Ϯ S.D. (n ϭ 3). The activation was statistically significant by Student's t test; p Ͻ 0.01. D, three human cell lines, JSC-1, BC3, and HCT 116 cells, were co-transfected with FLAG-K-RTA, IL-10 promoter reporter plasmid p-265, and the internal control pRL-TK plasmid (ratio 1:1:1). Two days after transfection, luciferase activity was examined. Results are presented as the mean Ϯ S.D. (n ϭ 3). The activation was statistically significant by Student's t test; *, p Ͻ 0.01. E, human IL-10 promoter reporter plasmid p-265 and internal control pRL-TK were co-transfected into cells with either an empty plasmid (i.e. without K-RTA) or with a plasmid carrying the wild-type K-RTA (WT) or the transactivation domain-deleted K-RTA (F490). Luciferase activity was determined 2 days after transfection. Results are presented as the mean Ϯ S.D. (n ϭ 3). Student's t test; *, p Ͻ 0.01. F, HCT 116 cells were transiently transfected with either K-RTA WT or K-RTA F490 expression plasmids. Two days after transfection, mRNA was prepared, and human IL-10 gene expression was measured by RT-qPCR. GAPDH mRNA was used to normalize data. Results are presented as the mean Ϯ S.D. (n ϭ 3). Student's t test. *, p Ͻ 0.01. G and H, binding efficiencies of K-RTA to PAN (PANp) or IL-10 (Ϫ265/Ϫ1 from p-265) promoter DNA probes were analyzed by EMSA. Five-fmol biotin-labeled promoter probes were incubated with increasing amounts of FLAG-K-RTA protein (aa 1-531). The amount of FLAG-K-RTA increased from 0 to 300 ng. A complex of FLAG-K-RTA-DNA probes was examined by EMSA, indicated by the triangle (G). The asterisk indicates free probe DNA. The binding curves were calculated (H).

Transactivation of IL-10 promoter by K-RTA
EMSA disappeared when mSP DNA probes were used (Fig. S4), specific binding of the K-RTA-SP1/3 complex to the IL-10 promoter requires the SP-binding element (AGGAGG, position at Ϫ177 to Ϫ172). Next, we examined the existence of a possible interaction between K-RTA, SP1, and SP3 proteins. A GSTpulldown assay using nuclear extracts showed that both SP1 and SP3 were co-precipitated with GST-K-RTA (aa 1-531) and mutations involving the basic region did not affect such interactions (Fig. 6B). Collectively, these data suggested that SP3 binding to the element AGGAGG in the IL-10 promoter (sequence from Ϫ177 to Ϫ172) is involved in the recruitment of K-RTA to the IL-10 promoter, and both SP1 and SP3 could Transactivation of IL-10 promoter by K-RTA mediate loading of K-RTA onto the IL-10 promoter (possible models are shown in Fig. 6D).
Next, we investigated target sites for K-RTA and SP3 complex formation with DNase I footprinting experiments. The results showed that SP3 alone bound to the predicted site (around Ϫ177/Ϫ172 in p-265) (Fig. S5). Intriguingly, K-RTA alone did not show any apparent binding site, and specific SP3 binding appeared to be independent of the addition of K-RTA ( Fig. 6C and Fig. S6). Thus, we concluded that SP3 binding to the SP site (Ϫ177/Ϫ172) on the IL-10 promoter was not affected by the presence of K-RTA, and SP3 can function as a guide for K-RTA at the IL-10 promoter. Overall, our data suggested a mechanism that SP3 can recruit K-RTA to the IL-10 promoter (Fig. 6D, number 4), but transactivational activity of K-RTA would be required for IL-10 promoter activation.

Discussion
We demonstrated that K-RTA can transactivate the human IL-10 gene promoter. The N-terminal basic domain and C-terminal transactivational domain of K-RTA were required for K-RTA-mediated IL-10 promoter activation. Both SP1 and SP3 were critical regulators for K-RTA-mediated IL-10 promoter activation, and we also uncovered the functional interplay between K-RTA, SP1, and SP3. K-RTA alone did not bind to the IL-10 promoter DNA in vitro. However, ChIP and ABCD assays and EMSAs testing the combinational effect of K-RTA with SP1 and SP3 indicated that SP1 and SP3 mediate the loading of K-RTA onto the IL-10 promoter. Collectively, K-RTA can activate the IL-10 promoter via SP1 and SP3.
The IL-10 promoter contains various cellular transcription factor-targeting sites such as those for STAT3, NF-B, C/EBP␣, AP1, SP1, and SP3 (35), and previous studies suggested that K-RTA may interact with those proteins (15,19,21,42). Although we did not observe functional co-operation between K-RTA and STAT1/3 in K-RTA-mediated IL-10 promoter activation, 3 herein, we show that both SP1 and SP3 have a cooperative effect on K-RTA-mediated IL-10 promoter activation. IL-10 transcription in macrophages is positively regulated by both SP1 and SP3 (49), and two previous studies also support functional interplay between K-RTA and SP1. KSHV thymidine kinase TATAA-less promoter is co-operatively regulated by K-RTA and SP1 (50), and K-RTA-mediated transcription of KSHV viral interferon regulatory factor (vIRF) gene involves SP1-binding sites in the vIRF promoter (51). Unfortunately, however, the precise molecular mechanism of the interplay between K-RTA, SP1, and SP3 was unresolved. We failed to confirm the existence of a protein complex composed of K-RTA, SP1, and SP3 by co-immunoprecipitation assays 3 ; ChIP and ABCD assays and EMSAs revealed its presence on DNA (Fig. 6D). Although K-RTA alone did not form a specific complex with IL-10 promoter DNA, an interaction with SP1/3 enables K-RTA to target the IL-10 promoter.
In this study, we showed that basic domain mutations K152R, K154R, and H145L of K-RTA disrupted its DNA-binding ability to viral PAN promoter, and we also found that the K-RTA N-terminal mutants K152R and K154R caused a reduction of IL-10 promoter transactivation. Although we did not observe direct K-RTA binding to the IL-10 promoter in vitro, results from the ABCD assay, EMSA, and biochemical cell fractionation experiments suggested that the basic region of K-RTA can contribute to the loading of K-RTA onto the IL-10 promoter. As our data did not support the possible role of the basic region in the physical interaction between K-RTA, SP1, and SP3 (Fig. 6), it may suggest another partner protein in K-RTA-mediated IL-10 promoter activation. It should be mentioned that SP1 and SP3 are not the only transcription factors involved in K-RTA-mediated IL-10 activation. In fact, we observed that mutations at PBX1-or NF-B-targeting elements, located near the SP1/3-targeting element, resulted in a clear reduction of K-RTA-mediated IL-10 promoter activation. Intriguingly, PBX1 has been shown to interact with SP1 and SP3 competitively (52) or cooperatively (48). Thus, such PBX1-SP1/3 interplay might be the mechanism affecting K-RTA-mediated IL-10 promoter activation, and there may be possible interactions between basic domains of K-RTA with other transcription factors. It is also of some concern that most of the results were obtained from transient expression of proteins combined with reporter constructs and from in vitro experiments. More precise investigations surrounding the authentic regulation of the IL-10 promoter in its cellular context are required. The functional interaction between K-RTA and other regulators involving the IL-10 promoter should be analyzed in greater detail in the future.
Collectively, this study demonstrated a novel functional interaction between K-RTA, SP1, and SP3 in K-RTA-mediated IL-10 promoter activation. Although there is the possibility of as-yet-unknown molecules that interact with SP1, SP3, and K-RTA, overall, our observations and those in other studies suggest that SP1/3-K-RTA complex-mediated gene transcription is not restricted to IL-10 promoter activation but might be widely associated with various KSHV-associated diseases. Fur-

Transactivation of IL-10 promoter by K-RTA
ther studies should uncover the intricate mechanisms of functional associations between SP1, SP3, and K-RTA.

Quantitative RT-PCR
Expression of IL-10 mRNA was analyzed by the TaqMan probe-based real-time, reverse-transcriptase PCR (RT-qPCR) method. Total cellular RNA was extracted using an RNeasy kit (Qiagen Sciences, Germantown, MD), according to the manufacturer's protocol. RT-qPCR was performed using the QuantiTect probe RT-PCR kit (Qiagen), with TaqMan probes (IL-10, Hs00961622_m1; GAPDH, Hs03929097_g1; purchased from Applied Biosystems, Waltham, MA) on the Applied Biosystems 7300 real-time PCR system. The PCR conditions were as follows: 1 cycle at 50°C for 30 min and 95°C for 15 min, followed by 40 cycles at 94°C for 15 s and 60°C for 1 min. Relative gene expression was calculated from triplicates by the ⌬⌬Ct method, using GAPDH as a reference gene.

Luciferase reporter assay
Cells were seeded on 24-well plates (1.5 ϫ 10 5 cells/well for HCT 116 and HEK293 and 5 ϫ 10 5 cells/well for JSC-1 and BC3 cells) 1 day before transfection. FuGENE HD transfection reagents (Roche Diagnostics GmbH, Mannheim, Germany) were used for transfection of HEK293 and HCT 116 cells, and Lipofectamine TM 2000 (Life Technologies, Inc., Invitrogen) was used for JSC-1 and BC3 cells, according to the manufacturers' protocols. Cells were collected 48 h post-transfection, and the Dual-Glo TM luciferase assay system (Promega) was used to measure luciferase activity. As an internal control to normalize transfection efficiency, pRL-TK or pRL-CMV plasmids (Promega) were co-transfected. Renilla luciferase activities from the pRL plasmid were not affected by co-transfection with K-RTA. For all the transfections, the amount of DNA in each well was made constant by adding empty plasmid DNA pCMV2. Triplicate transfection was performed in each reporter assay. The ratio between firefly and Renilla luciferase activities from the same sample was calculated to obtain normalized firefly activity. Average and standard deviations (S.D.) were calculated (n ϭ 3), and reproducibility was confirmed.

Western blotting and antibodies
Proteins were resolved by SDS-PAGE and electrotransferred to Immobilon-P membrane (EMD Millipore, Billerica, MA), and membranes were blocked and incubated with antibodies. Western blotting signals were developed with the SuperSignal West Dura Extended Duration Substrate (Thermo Fisher Scientific, Pierce) and recorded with an ImageQuant LAS4000 mini image analyzer (GE Healthcare). The following primary antibodies were used: anti-FLAG M2 (Sigma); anti-HA (3F10; Roche Applied  mutant p-265). C, SP-binding site-dependent K-RTA loading onto the IL-10 promoter was determined by ABCD assay. Nuclear extracts (100 g) from myc-K-RTA-expressing HEK293 cells were incubated with biotin-labeled IL-10 promoter probe (from either wild type or mutated mSP p-265). Myc-K-RTA co-precipitating with biotinylated DNA probes was detected by Western blotting (upper panel). Signal intensity was quantified from triplicate Western blotting experiments and calculated as p-265-bound K-RTA/input K-RTA ratio (lower plot). Student's t test; **, p Ͻ 0.05 (WT versus mSP). D, WT or mutant IL-10 promoter occupancy by K-RTA was analyzed by ChIP assay. HCT 116 cells were co-transfected with FLAG-K-RTA-expressing plasmid and p-265 (either WT or mSP) reporter plasmid (ratio 1:1) and 2 days later were fixed with 3.7% formalin. Nuclei were sonicated, and FLAG-K-RTA-bound DNA was immunoprecipitated with anti-FLAG antibody. Eluted DNA was quantified by qPCR, using primers targeting p-265. Results are presented as the mean Ϯ S. D. (n ϭ 3). Student's t test; *, p Ͻ 0.01.

Immunofluorescence confocal microscopy
Cells were seeded on Scientific TM Nunc TM Lab-Tek TM II CC2 TM chamber slides (5-10 ϫ 10 4 cells/0.5 ml/well; Thermo Figure 5. Superior SP3 binding to IL-10 promoter in vitro. A, ABCD assay was employed to detect in vitro IL-10 promoter association of K-RTA. Increasing amounts of nuclear extracts (NE) from myc-K-RTA-expressing HEK293 cells were incubated with biotin-labeled DNA probes from either wild type (WT) or SP-binding site-mutated (mSP) IL-10 promoter (fragment from nucleotides Ϫ265 to Ϫ1). Myc-K-RTA and endogenous SP1 and SP3 co-precipitating with biotin-DNA probes were detected by Western blotting. Myc-K-RTA and SP3 co-precipitated with p-265 DNA probe in a nuclear extract dose-dependent manner. Two independent experiments showed similar results. B, SP1-HA and SP3-HA were expressed in HEK293 cells and immunopurified using anti-HAagarose. Purified SP1-HA and SP3-HA were visualized by silver staining after SDS-PAGE. An eluted sample from an empty plasmid-transfected lysate was used as control; the molecular mass marker is shown (left lane, Marker). C and D, binding affinities of SP1-HA and SP3-HA to the IL-10 promoter were analyzed by EMSA. The biotin-labeled IL-10 promoter probes (fragment from nucleotides Ϫ265 to Ϫ1) were incubated with increasing amounts of SP1-HA (left lanes) and SP3-HA (right lanes). The amount of SP1-HA and SP3-HA increased from 0 to 15 ng. The triangle indicates SP3-DNA complex, and the asterisk indicates free DNA probe (C). EMSA experiments were performed three times, and the binding curves and the K d values for SP3-HA on the IL-10 promoter probe were determined as in Fig. 1H (D). E and F, binding affinities of SP3-HA to wild type (WT p-265; left lanes) or SP3-binding site-mutated (mSP; right lanes) IL-10 promoter were analyzed by EMSA. The biotin-labeled IL-10 promoter probes (fragment from nucleotides Ϫ265 to Ϫ1) were incubated with increasing amounts of SP3-HA. The amount of SP3-HA increased from 0 to 15 ng. The triangle indicates SP3-DNA complex, and the asterisk indicates free DNA probe (E). EMSA experiments were performed three times, and the binding curve of WT is identical to that of SP3 in D (F).

Transactivation of IL-10 promoter by K-RTA
Fisher Scientific) and cultured overnight. Cells were transfected with the plasmid expressing FLAG-K-RTA, using FuGENE HD transfection reagent (Promega) and, on the following day, were fixed with 3.7% formalin/phosphate-buffered saline (PBS) for 10 min at room temperature. Incubation with the antibodies was performed as described previously (55). Briefly, cells were permeabilized, blocked, and incubated with primary and secondary antibodies. Anti-FLAG M2 antibody was used at a 1:4000 dilution and the Alexa-Fluor-488-conjugated antimouse antibody (Invitrogen, Molecular Probes) at 1:2000. Cells were washed three times with PBS and mounted with Prolong gold antifade reagent with DAPI (Invitrogen). Acquisition of images (640 ϫ 640 pixels) was performed using an FV1000-D IX81 confocal microscope equipped with an UPLFLN 40 ϫ NA1.3 oil objective lens (Olympus Corp., Tokyo), with identical parameter settings. Confocal two-dimensional TIFF images were merged with Adobe Photoshop CS4 extended software (Adobe Systems Inc., San Jose, CA).

Cell fractionation
Cells were fractionated into cytosolic, nuclear, and pellet fractions as described elsewhere (56), with modifications. Briefly, cells were washed with PBS and suspended at 2 ϫ 10 6 cells/ml in hypotonic buffer (HB) (10 mM HEPES (pH 7.9), 10 mM KCl, 1.5 mM MgCl 2 , 1 mM dithiothreitol, 0.1% Nonidet P-40) plus a supplement. Cells were incubated for 5 min on ice, and after a brief centrifugation, the low-salt fraction and the nuclei were recovered. Nuclei were resuspended in nuclear extract buffer (20 mM HEPES (pH 7.9), 0.42 M KCl, 1.5 mM MgCl 2 , 1 mM dithiothreitol, 25% glycerol) plus the supplement, to recover the nuclear fraction. The remaining pellet was resuspended in 1ϫ SDS buffer and sonicated.

Immunopurification
For immunopurification of SP1-HA, SP3-HA, FLAG-K-RTA (aa 1-531), and myc-K-RTA (aa 1-531), HEK293 cells were transfected with the respective plasmids using Lipofectamine 2000 (Thermo Fisher Scientific), according to the manufacturer's protocol. Alternatively, HEK293 cells were transfected by the calcium phosphate method. Two days after transfection, nuclear extracts or NEB350 extracts of transfected HEK293 cells were prepared as described (57) and incubated with 10 l of anti-HA or anti-FLAG M2 affinity gel (EZview TM red, Sigma) for 2 h, at 4°C, with constant rotation. Affinity gel was recovered by gentle centrifugation and washed six times with binding buffer. Tagged proteins were eluted by competitive peptide-containing binding buffer (100 g/ml HA peptide or 400 g/ml FLAG peptide). Competitive peptides were purchased from Sigma.

Expression and purification of GST-fusion proteins
The K-RTA-531STOP fragment was cloned into pGEX-6P-3 to express a GST-fusion protein. Recombinant K-RTA protein was induced in Escherichia coli Rosetta 2(DE3) pLysS (Novagen, Merck Millipore, Darmstadt, Germany) with 0.1 mM isopropyl D-thiogalactopyranoside or 2 h at 30°C. The cells was harvested and resuspended in lysis buffer (20 mM Tris-HCl (pH 8.0), 5 mM EDTA, 500 mM NaCl, 0.5% Nonidet P-40, 0.01% SDS, and protease inhibitor mixture from Roche Applied Science) and sonicated. The lysate was incubated with glutathione-Sepharose 4B (GE Healthcare Japan Corp.) for 2 h at 4°C, with continual rotation. Sepharose was recovered by gentle centrifugation and washed five times with lysis buffer. The GST-fusion protein was eluted using elution buffer (20 mM GSH in lysis buffer).

Knockdown experiments
For SP1 and SP3 knockdown by siRNA transfection, control (AllStars negative control siRNA from Qiagen) and SP1 and

Transactivation of IL-10 promoter by K-RTA
SP3 siRNAs (siGENOME SP1 and SP3; D-026959 and D-023096, respectively; GE Healthcare) were used. In siRNA transfection experiments, HiPerfect reagent (Qiagen) was used for transfection of HeLa cells, in accordance with the manufacturer's instructions.

Chromatin-immunoprecipitation assay
HEK293 or HCT 116 cells were co-transfected with FLAG-K-RTA and IL-10 promoter reporter plasmid p-265, and 2 days post-transfection, cells were harvested, and nuclei were collected and cross-linked with 1% formalin for 5 min. DNA bound to FLAG-K-RTA was immunoprecipitated from sonicated nuclear extracts using anti-FLAG M2 antibody, as described earlier (56). Collected DNA was subjected to qPCR, performed with the QuantiTect SYBR Green PCR kit (Qiagen) and primers for the IL-10 promoter fragment from nucleotides Ϫ265 to Ϫ1 (primer Ϫ265, 5Ј-CTCGAGATAGCTGTAATGCA-GAAGTTCATG-3Ј, and primer Ϫ1, 5Ј-GCCTTCTTTTG-CAAGTCTGTCTTGTGG-3Ј). A portion of pre-cleared lysate was separated as "input standard" and subjected to the same qPCR experiments to obtain a standard curve. The ratio of K-RTA-bound IL-10 promoter was calculated from this standard curve.

ABCD assay
HEK293 cells were transfected with myc-K-RTA expression vectors by the calcium phosphate method. Two days after transfection, nuclear extracts or NEB350 extracts of HEK293 cells were prepared as described (56). Extracts were diluted 1:3 in buffer D (20 mM HEPES (pH 7.9), 20% (v/v) glycerol, 0.5 mM PMSF, and 0.5 mM DTT) and incubated with 1 g of biotinlabeled IL-10 promoter probe (fragment from nucleotides Ϫ265 to Ϫ1), in a binding buffer consisting of 20 mM HEPES (pH 7.9), 140 mM KCl, 20% glycerol, 1.5 mM MgCl 2 , 2.5 M ZnSO 4 , 1 mM DTT, 2.5 M aprotinin, 1 mM AEBSF, and a phosphatase inhibitor mixture (Sigma). Streptavidin affinity gel (EZview TM Red, Sigma) was then added, and the reaction mixtures were incubated at 4°C for 20 h. Affinity gel was washed six times with binding buffer, and the proteins were dissolved in SDS sample buffer and subjected to Western blotting.

GST-pulldown assay
Nuclear extracts of HCT 116 cells (350 g) were incubated with 1.2 pmol of GST-fused mutant or K-RTA proteins bound to glutathione-Sepharose, in a binding buffer consisting of 20 mM HEPES (pH 7.9), 105 mM KCl, 20% glycerol, 0.4 mM MgCl 2 , 1 mM dithiothreitol, 2.5 M aprotinin, and 1 mM AEBSF. The mixtures were incubated at 4°C for 16 h. Beads were washed six times in binding buffer and eluted with 20 mM GSH in binding buffer. Proteins were analyzed by Western blotting.

DNase I footprinting analysis
A DNA probe containing the IL-10 promoter region (sequence from nucleotides Ϫ206 to Ϫ72) was amplified by PCR using the primers (5Ј-CAAGAATTCAAAACTAAGTT-TATTAGAGAGGTTAG-3Ј and 5Ј-CTCTGTCCCCCTTT-TATATTGTAAGCTCAGGGAGG-3Ј) with plasmid p-265 as template. The amplified probe was digested by EcoRI and purified. The probe was labeled at the 3Ј terminus with [␣-32 P]dATP using DNA polymerase I, large (Klenow) fragment (Life Technologies, Inc., Invitrogen), and purified. The 32 P-labeled probes (10,000 cpm/sample) were incubated with immunopurified proteins in 1.75 g of poly(dI-dC), 10 mM HEPES (pH 7.45), 50 mM NaCl, 5% glycerol, 5.7 mM MgCl 2 , 50 mM ␤-mercaptoethanol, and 0.38 mM dithiothreitol. After incubation at room temperature for 30 min, samples were treated with DNase I (Roche Applied Science) at room temperature for 15 min. Reactions were halted by adding 20 mM EDTA and subjected to ethanol precipitation. The purified DNA samples were electrophoresed on 6% denaturing polyacrylamide gel containing 8.3 M urea in 1ϫ TBE (89 mM Tris, 89 mM boric acid, 2 mM EDTA (pH 8.3)) for 120 min at 1000 V. The dried gels were analyzed using an FLA 7000 bio-image analyzer (Fujifilm, Tokyo, Japan).

Statistical analysis
In vitro experiments were repeated independently at least twice, and consistent results were observed. Quantitative results are presented as the mean Ϯ S.D. (n ϭ 3). GraphPad Prism7 software (La Jolla, CA) was applied to statistical analyses and to generate scatter plots. Error bars represent S.D. Twotailed Student's t test was used to evaluate the significance of differences between two groups of data with similar variances. Values of p Ͻ 0.05 were considered to indicate statistically significant differences.