Bovine Foamy Virus Transactivator BTas Interacts with Cellular RelB To Enhance Viral Transcription

Abstract

Viruses are obligate intracellular parasites that depend on cellular machinery for their efficient transcription and replication. In a previous study we reported that bovine foamy virus (BFV) is able to activate the nuclear factor κB (NF-κB) pathway through the action of its transactivator BTas to enhance viral transcription. However, the mechanism used by NF-κB to enhance BFV transcription remains elusive. To address this question, we employed a yeast two-hybrid assay to screen for BTas-interacting proteins. We found that RelB, a member of NF-κB protein family, interacts with BTas. We confirmed the putative RelB-BTas interaction in vitro and in vivo and identified the protein regions responsible for the RelB-BTas interaction. Using a luciferase reporter assay, we next showed that RelB enhances BFV transcription (BTas-induced long terminal repeat [LTR] transactivation) and that this process requires both the localization of the RelB-BTas interaction in the nucleus and the Rel homology domain of RelB. The knockdown of the cellular endogenous RelB protein using small interfering RNA (siRNA) significantly attenuated BTas-induced LTR transcription. The results of chromatin immunoprecipitation (ChIP) analysis showed that endogenous RelB binds to the viral LTR in BFV-infected cells. Together, these results suggest that BFV engages the RelB protein as a cotransactivator of BTas to enhance viral transcription. In addition, our findings indicate that BFV infection upregulates cellular RelB expression through BTas-induced NF-κB activation. Thus, this study demonstrates the existence of a positive-feedback circuit in which BFV utilizes the host's NF-κB pathway through the RelB protein for efficient viral transcription.

Foamy viruses (FVs) form the only genus in the Spumaretrovirinae subfamily of the Retroviridae. They possess a complex genome and a special gene expression regulatory mechanism. FVs have been isolated from different species, including primates, bovines, equines, and felines (4, 11, 13, 24, 31). Although FVs have extensive cellular tropism and their infection of cells in tissue culture causes the rapid formation of syncytia, these viruses are nonpathogenic in their natural hosts or in experimentally infected animals (14, 19, 21). FVs encode a regulatory protein named Tas that binds directly to a DNA sequence in the viral LTR (long terminal repeat) to enhance viral gene transcription (6, 25). Tas of bovine foamy virus (BFV), referred to as BTas, is a 249-amino-acid (aa) protein. BTas does not have a classical nuclear localization signal, yet it is found mainly in the nucleus. Two domains have been identified within BTas: a DNA-binding domain and a transactivation domain. Although the two domains have not been characterized in detail, the extreme N and C termini of BTas have been found to be indispensable for its DNA-binding and transactivation functions, respectively (our unpublished data). In addition, BTas binds to the BFV LTR at nucleotides 601 to 841, termed a Tas-responsive element, through its DNA-binding domain and thereby stimulates viral gene transcription.

In mammals, nuclear factor κB (NF-κB) belongs to a family of transcription factors that plays a pivotal role in regulating the expression of various genes related to the immune response, cell survival, and development (1, 10, 16, 17, 29). NF-κB is found as a heterodimer of the p50 (NF-κB1) and p65 (RelA) subunits. In quiescent cells, NF-κB dimers are retained in the cytoplasm due to their interaction with the inhibitory IκB protein (12). When the NF-κB pathway is activated by stimuli such as tumor necrosis factor alpha (TNF-α), interleukin-1 (IL-1), or double-stranded RNA (dsRNA), IκB undergoes phosphorylation, ubiquitination, and degradation, and ultimately, the NF-κB dimers are transported into the nucleus, leading to the transactivation of NF-κB target genes (2, 9, 32). Besides this conventional pathway, an additional one that involves signaling through the NF-κB-inducing kinase (NIK) and IKKα (IKK complex α-subunit) has been described. This results in the processing of p100-RelB into a mature p52-RelB heterodimer and its subsequent release into the nucleus, where it binds to the promoters of its target genes to regulate their transcription, similar to the action of the p50-p65 dimer (5, 8, 28).

Although RelB is a member of the NF-κB family of transcription factors (RelA, c-Rel, NF-κB1, and NF-κB2), it is structurally and functionally different from the other two transcriptionally active members, RelA and c-Rel. While possessing both a Rel homology domain (RHD) and a transactivation domain (TAD) like the other two proteins, RelB also has an N-terminal leucine zipper domain that is not present in other NF-κB family members (27). In addition, no exclusive DNA-binding activity of RelB had been discovered so far; however, it is able to bind to target promoters through interactions with other proteins (e.g., p52 and p50). Under physiological conditions, p52-RelB is poorly sequestered by IκB molecules (20), allowing the heterodimer to translocate freely to the nucleus (7) and modulate the rate of transcription of its specific target genes. Compared to other members of the NF-κB family, the biological mode of action of the RelB subunit has remained elusive. The regulation of RelB is highly dependent on its own transcription, which may ultimately influence its nuclear translocation, and on NF-κB-inducing stimuli such as lipopolysaccharide (LPS) and TNF, which can induce RelB nuclear translocation by increasing its transcription through the specific κB site region within the promoter of RelB (3).

Multiple families of viruses have evolved sophisticated strategies to regulate NF-κB signaling. In turn, NF-κB is also important for the replication of many viruses. For example, the induction of the NF-κB pathway aids in HIV replication: it stimulates the transactivation of the HIV LTR and leads to enhanced viral transcription (15, 35). In a recent study, we reported that BFV is also able to activate the NF-κB pathway through the action of BTas to enhance viral transcription (34). However, our results also showed that the BFV LTR lacks a functional κB site. Thus, we speculated that NF-κB might use a mechanism other than direct binding to the viral LTR to enhance BFV transcription. Here, we show that BFV engages the cellular RelB protein as a cotransactivator of BTas to enhance its transactivation function (BTas-induced LTR transactivation). In addition, our findings reveal that BFV infection upregulates cellular RelB expression through BTas-induced NF-κB activation. Thus, we describe a positive virus-host feedback circuit in which BFV utilizes the host's NF-κB pathway through the RelB protein for its efficient transcription.

MATERIALS AND METHODS

Materials.

4′,6-Diamidino-2-phenylindole (DAPI), protein A beads, and β-actin antibody were purchased from Sigma (St. Louis, MO). Antibodies against the Myc tag and Flag tag were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Fluorescein-conjugated anti-mouse and anti-rabbit secondary antibodies were obtained from Jackson ImmunoResearch Laboratories (West Grove, PA). Horseradish peroxidase-conjugated anti-rabbit and anti-mouse secondary antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). RelB small interfering RNA (siRNA) (catalog no. sc-36402) was obtained from Santa Cruz Biotechnology. It consists of pools of three target-specific siRNAs: siRNAs A (5′-GCAACAUGUUCCCCAAUCAtt-3′), B (5′-CGUGCACUAGCUUGUUACAtt-3′), and C (5′-CUCCAGUAGGAUUCGGAAAtt-3′). Each siRNA is able to knock down RelB expression independently. The antibody against BTas was described previously (30). Anti-RelB antibody was generated in BALB/c mice using a bacterially purified human RelB protein as the immunogen. This polyclonal anti-RelB serum was used for immunoprecipitation, immunofluorescence imaging, and Western blotting (1:3,000 starting dilution). This antiserum specifically recognizes both human and bovine RelB (BRelB) proteins, which are 91% homologous.

Plasmid DNA.

pCMV-RelB and pCMV-BRelB (or pCMV-Tag3B−) were generated by inserting RelB cDNA (purchased from FulenGen, GuangZhou, China) and bovine RelB cDNA (purchased from Sangon, Shanghai, China) into the expression vectors pCDNA 3.1 (Invitrogen) and pCMV-Tag3B (Stratagene). RelB mutants were created by PCR and cloned into pcDNA3.1 or pCMV-Tag3B. The pCMV-BTas (aa 27 to 249), pCMV-BTas (aa 1 to 234), pEGFP-BTas (aa 1 to 167), and pEGFP-BTas (aa 167 to 249) DNA constructs were engineered by inserting the respective PCR fragment into the pcDNA3.1 or pEGFP-N1 vector. The pCMV-BTas, pNF-κB-Luc, pTRAF6, hemagglutinin (HA)-RelA, pCMV-p100, pCMV-p52, HA-tagged NIK, and pIκBα(AA) DNA constructs were previously described (34). Glutathione S-transferase (GST)-BTas and GST-RelB were generated by inserting BTas and RelB cDNAs into the pGEX6P-1 DNA vector. The sequences of all of the new constructs were confirmed by sequencing.

Cells, viruses, and proteins.

HeLa cells, 293T cells, and Madin-Darby bovine kidney (MDBK) cells were maintained in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% fetal bovine serum (FBS), streptomycin (50 μg/ml), and penicillin (50 U/ml) as monolayers in tissue culture plates. BFV (strain 3026) was isolated from peripheral lymphoid cells of infected bovines by us in 1993 (22). Escherichia coli BL21(DE3) was used to express the GST-BTas and GST-RelB proteins. Proteins were purified in the presence of 500 units of Benzonase nuclease (Sigma-Aldrich, St. Louis, MO) using glutathione-Sepharose 4B beads according to the manufacturer's instructions (Promega, Madison, WI). The GST tag was removed by using PreScission protease (GE Healthcare).

Western blotting.

Cell lysates were separated by 12% SDS-PAGE (polyacrylamide gel electrophoresis). Proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane (Millipore, Billerica, MA). Following incubation in 5% nonfat milk (in 1× phosphate-buffered saline [PBS]) for 45 min at room temperature, the membrane was blotted with primary antibody for 90 min at room temperature and then incubated with goat anti-mouse or goat anti-rabbit secondary antibodies conjugated with peroxidase. The protein signal was visualized on an X-ray film.

Luciferase reporter assay (Luc assay).

Cells (1 × 10⁵) were seeded in 12-well plates 20 h before transfection using polyethylenimine (PEI; Sigma, St. Louis, MO). The total DNA in each transfection mixture was adjusted to the same amount with vector DNA. pCMV-β-gal plasmid DNA was included in each transfection. Cells were harvested 48 h after transfection. The level of luciferase activity was measured by using a luciferase assay system (Promega, Madison, WI). The activity of β-galactosidase in cell lysates was also measured, and the results were used as an internal control to normalize the efficiency levels between transfections. Each experiment was performed at least three times.

Immunofluorescence assay (IFA).

Cells were fixed with 4% (wt/vol) paraformaldehyde (in 1× PBS) for 10 min at room temperature, followed by permeabilization in 0.5% Triton X-100 (in 1× PBS) for 10 min. Cells were first incubated with 3% bovine serum albumin (BSA) (in 1× PBS) at 37°C for 30 min and then incubated with antibodies against BTas, p65, and p100 (all at a dilution of 1:500) at 37°C for 1 h. After washing with 0.5% Tween 20 (in 1× PBS) three times for 10 min at room temperature, Texas Red-conjugated goat anti-rabbit and fluorescein isothiocyanate (FITC)-conjugated rabbit anti-mouse secondary antibodies (at a dilution of 1:1,000) were added at 37°C for 30 min. Nuclei were stained with DAPI. Cells were examined with an Olympus X71 fluorescence microscope.

Coimmunoprecipitation (Co-IP).

A total of 1 × 10⁷ cells were transfected with various plasmids by using PEI reagent. Forty-eight hours after transfection, cells were harvested, lysed in 600 μl lysis buffer (50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 1% NP-40, and 1 mM phenylmethylsulfonyl fluoride), sonicated, and centrifuged at 4°C (10,000 × g for 15 min). The supernatant (500 μl) was incubated with antibodies for 2 h at 4°C. Fifteen microliters of protein A-Sepharose was then added, and the mixture was incubated for 2 h at 4°C. After five washes with lysis buffer, the Sepharose was resuspended in 35 μl of 1× SDS loading buffer and boiled for 5 min, and the protein samples were then subjected to Western blotting.

ChIP assay.

Chromatin immunoprecipitation (ChIP) assays were performed by using the Upstate Biotechnology ChIP assay kit (catalog no. 17-295). In brief, formaldehyde was added to the culture medium at a final concentration of 1% and incubated at 37°C for 10 min. Chromatin was immunoprecipitated by using anti-BTas or anti-RelB antibodies. One-tenth of the lysates was kept to measure the amount of DNA present in different samples before immunoprecipitation. DNA was purified from both the immunoprecipitated and preimmune samples, diluted from 1:10 to 1:10,000, and subjected to PCR and a real-time PCR assay to detect BFV LTR DNA using the primer pair LTR forward (5′-AGGTCAGGCGGTATGCTTTCTACT-3′) and LTR reverse (5′-GGATCCGAACCTTGTTCTCTCAGT-3′). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) promoter DNA was also amplified from a ChIP assay that was performed with anti-RNA polymerase II antibodies (catalog no. 05-623B, anti-RNA polymerase II, clone CTD4H8; Upstate Biotechnology), with the results serving as a positive control.

RT-PCR and real-time PCR.

Total RNA was isolated from cells by using an RNeasy kit (Qiagen) according to the manufacturer's instructions. Fifteen micrograms of total RNA was used for cDNA synthesis using Moloney murine leukemia virus (M-MLV) reverse transcriptase (RT) (Promega). Pilot experiments were done to determine the linear range of amplification with respect to the quantity of starting template cDNA and PCR cycles using the primer pairs GAPDH forward (5′-AACAGCGACACCCATCCTC-3′) and GAPDH reverse (5′-CATACCAGGAAATGAGCTTGACAA-3′) and RelB forward (5′-CGCCCATCGCTTGTTCATCGTG-3′) and RelB reverse (5′-CCGCAGCCCCAGCAGGTGTAT-3′). According to the above-described results, PCR was performed for 25 cycles (RelB) or 18 cycles (GAPDH) at 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s. The resulting PCR product was sequenced and confirmed to be RelB.

Real-time PCR was performed by using the IQ5 Multicolor real-time PCR detection system (Bio-Rad) and SYBR green real-time PCR master mix (Toyobo) according to the manufacturers' instructions. The primers used are described above. Cycling parameters were 94°C for 3 min followed by 40 cycles at 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s. Fluorescence measurements were performed for each cycle at the 72°C step. The specificity of the amplification reaction was verified by melting-curve analyses. Threshold cycle (C_T) values were calculated automatically as the cycle when the fluorescence of the sample exceeded a threshold level corresponding to 10 standard deviations (SD) of the mean of the baseline fluorescence. The level of RelB expression was normalized to GAPDH by using the 2^ΔΔCT method (23). The PCR product was sequenced to confirm its identity.

RESULTS

BTas interacts directly with RelB.

To identify proteins that interact with BTas, a human B-cell cDNA library was screened in a yeast two-hybrid assay with BTas as the bait. One of the positive clones contained a 1,029-bp cDNA fragment (GenBank accession no. NM_006509.2) that encodes a portion of the RelB protein (amino acids [aa] 236 to 579). We then performed experiments to validate this putative interaction between BTas and RelB. 293T cells were cotransfected with BTas and Myc-RelB DNA constructs. Following immunoprecipitation with BTas antibodies, the precipitates were assessed on Western blots by using anti-Myc tag antibodies. The results show that Myc-RelB was coimmunoprecipitated with BTas (Fig. 1 A). Similarly, BTas was coimmunoprecipitated with Myc-RelB (Fig. 1B). In support of these coimmunoprecipitation data, the results of immunofluorescence staining showed that BTas and RelB were colocalized within the nucleus (Fig. 1C). In order to determine whether BTas binds directly to RelB, we expressed RelB and GST-BTas in E. coli cells and purified the two proteins. The results of GST pulldown experiments showed that RelB was eluted together with GST-BTas, suggesting a direct association of RelB and BTas (Fig. 1D). We further tested whether BTas interacts with endogenous RelB in BFV-infected cells. Forty-eight hours after infection with BFV, HeLa cells were lysed, and endogenous RelB was immunoprecipitated with RelB antibody. The results of Western blot analysis showed that BTas was present in the precipitates, suggesting an interaction between BTas and cellular endogenous RelB during BFV infection (Fig. 1E). Taken together, these results demonstrate a direct interaction of BTas with RelB in cells that are infected with BFV.

FIG. 1.

BTas interacts directly with RelB. (A) Co-IP of BTas and RelB. 293T cells were cotransfected with Myc-RelB together with an empty vector (lane 1) or BTas (lane 2). BTas was immunoprecipitated with anti-BTas (α-BTas) agarose beads, and the immunoprecipitates were detected by Western blotting using anti-Myc (top) or anti-BTas (middle). The amount of Myc-RelB in each cell lysate was checked by using anti-Myc antibody (bottom). (B) Reciprocal Co-IP of BTas and RelB. 293T cells were cotransfected with BTas together with an empty vector (lane 1) or Myc-RelB (lane 2). RelB was immunoprecipitated with anti-Myc agarose beads, and the immunoprecipitates were detected by Western blotting using anti-BTas (top) or anti-Myc (middle) antibody. The amount of BTas in each cell lysate was checked by using anti-BTas antibody (bottom). (C) HeLa cells were cotransfected with BTas and Myc-RelB. Indirect IFA was used to localize RelB (with mouse anti-RelB antibody and FITC-conjugated goat anti-mouse secondary antibody) and BTas (with rabbit anti-BTas antibody and Texas Red-conjugated goat anti-rabbit secondary antibody). Nuclei were visualized with DAPI staining. (D) In vitro interaction between BTas and RelB. The purified RelB protein was incubated with GST or GST-BTas agarose beads. GST pulldown was then performed, and precipitates were detected by Western blotting using an anti-RelB antibody. Purified GST and GST-BTas used in the pulldown assay were detected by Coomassie blue staining. (E) Lysate of BFV-infected HeLa cells was subjected to immunoprecipitation with control mouse IgG (lane 1) and anti-RelB (lane 2) agarose beads, and the immunoprecipitates were detected by Western blotting using anti-BTas (top) or anti-RelB (middle). The amount of BTas in the cell lysate was checked by using anti-BTas (bottom).

Identification of the protein sequences responsible for the RelB-BTas interaction.

We next performed experiments to identify the BTas-binding domain in RelB. To this end, five truncated mutants of RelB were constructed and tested in a GST pulldown assay for their ability to bind BTas (Fig. 2 A). The results show that the RHD of RelB is essential for binding to BTas (located between aa 236 and 285), whereas the TAD of RelB does not play a role (Fig. 2B). We also created four BTas mutants and assessed their associations with RelB in coimmunoprecipitation experiments (Fig. 2C). The results show that the region between aa 167 and 249 in BTas is responsible for the interaction with RelB (Fig. 2D), whereas the DNA-binding domain is dispensable. In addition, neither the extreme N-terminal nor C-terminal region of BTas, which are essential for the DNA-binding and transactivation capacities of BTas, is required for the RelB interaction (Fig. 2E). Taken together, these results indicate that RelB interacts with BTas through the RHD of RelB and aa residues 167 to 234 of BTas. The RHD is responsible for the interaction of RelB with its partner proteins (e.g., p52), which leads to the binding of RelB to the promoter of its target genes and their subsequent transactivation. Therefore, this prompted us to examine whether the RelB-BTas interaction is required for BTas-mediated transcription.

FIG. 2.

Identification of the protein sequences responsible for the RelB-BTas interaction. (A) Schematic overview of the Myc-tagged truncation forms of RelB. (B) 293T cells were transfected with Myc-RelB or the indicated Myc-tagged truncations. After 48 h, cell lysates were incubated with GST or GST-BTas agarose beads. A GST pulldown was then performed, and precipitates were detected by Western blotting using an anti-Myc antibody. Purified GST and GST-BTas used in the pulldown assay were detected by Coomassie blue staining. (C) Schematic representation of the truncation forms of BTas. (D) 293T cells were cotransfected with an empty vector, BTas, or the indicated truncations together with Myc-RelB; BTas was immunoprecipitated with anti-BTas agarose beads; and the immunoprecipitates were detected by Western blotting using anti-Myc (top) or anti-BTas (middle) antibody. The amount of RelB in each cell lysate was checked by using anti-Myc antibody (bottom). (E) 293T cells were cotransfected with an enhanced GFP (EGFP)-C1 vector, EGFP-BTas, or the indicated EGFP-fused truncations together with Myc-RelB; EGFP-BTas was immunoprecipitated with anti-EGFP agarose beads; and the immunoprecipitates were detected by Western blotting using anti-Myc (top) or anti-EGFP (middle) antibody. The amount of RelB in each cell lysate was checked by using anti-Myc antibody (bottom).

RelB enhances BTas-induced LTR transcription.

We next asked whether RelB modulates the function of BTas in transactivating BFV LTR transcription. To answer this question, HeLa cells were transfected with pLTR-Luc reporter plasmid DNA and increasing amounts of RelB DNA in the absence or presence of pCMV-BTas plasmid DNA (transfection with pNF-κB-Luc, pCMV-RelB, and pCMV-p52 was used as a positive control [Fig. 3 A, right]). The results of the luciferase assay showed that BTas increased BFV LTR transcription by 20-fold (Fig. 3A, left). RelB alone did not affect BFV LTR transcription, yet it enhanced BTas-induced LTR transcription by 3- to 10-fold in a dose-dependent manner (Fig. 3A, left). These results indicate that RelB augments the transactivation activity of BTas.

FIG. 3.

RelB enhances BTas-induced LTR transcription. (A) HeLa cells were cotransfected with LTR-Luc plasmid (0.05 μg), the indicated amount of RelB expression plasmid with or without BTas, and an empty vector to keep the total DNA concentration in each transfection mixture constant (left). Transfection with pNF-κB-Luc (0.05 μg), pCMV-RelB (0.2 μg), and pCMV-p52 (0.2 μg) was used as a positive control (right). After 48 h the relative luciferase activities were analyzed. The relative expression of BTas in the transfected cells was monitored by a Western blot assay (bottom). Error bars represent SD of the means from three independent experiments. (B) Co-IP of BTas and RelA. 293T cells were cotransfected with HA-RelA together with an empty vector (lane 1) or BTas (lane 2). BTas was immunoprecipitated with anti-BTas agarose beads, and the immunoprecipitates were detected by Western blotting using anti-HA (top) or anti-BTas (middle) antibody. The amount of Myc-RelB in each cell lysate was checked by using anti-HA antibody (bottom). (C) HeLa cells were cotransfected with LTR-Luc plasmid (0.05 μg), the indicated amount of RelA expression plasmid with or without BTas, and empty vector to keep the total DNA concentration in each transfection mixture constant. After 48 h the relative luciferase activities were analyzed. The relative expression of BTas in the transfected cells was monitored by a Western blot assay (bottom). Error bars represent SD of the means from three independent experiments. (D) Co-IP of BTas and BRelB. MDBK cells were cotransfected with Myc-BRelB together with an empty vector (lane 1) or BTas (lane 2). BTas was immunoprecipitated with anti-BTas agarose beads, and the immunoprecipitates were detected by Western blotting using anti-Myc (top) or anti-BTas (middle) antibody. The amount of Myc-BRelB in each cell lysate was checked by using anti-Myc antibody (bottom). (E) MDBK cells were cotransfected with LTR-Luc plasmid (0.05 μg), BRelB (0.2 μg) expression plasmid with or without BTas, and an empty vector to keep the total DNA concentration in each transfection mixture constant. After 48 h the relative luciferase activities were analyzed. Error bars represent SD of the means from three independent experiments.

Because both RelA and RelB belong to the Rel protein family, we asked whether RelA could also interact with BTas and enhance BTas-induced BFV LTR transcription. The results of the coimmunoprecipitation assay showed that RelA is able to interact with BTas (Fig. 3B). However, RelA did not exert any effect on the transactivation activity of BTas (Fig. 3C). Therefore, RelB, and not RelA, specifically stimulates BTas-induced transcription of the BFV LTR, suggesting that RelB might engage a unique mechanism to modulate the transactivation activity of BTas, which is absent from other Rel family transactivators.

Next, we repeated the same experiments with bovine kidney cells (MDBK) using bovine RelB (BRelB), which is 91% homologous to human RelB. We first examined the interaction between BRelB and BTas. The results of a Co-IP assay confirm that BRelB interacts with BTas as well (Fig. 3D). Furthermore, Luc analysis showed that BRelB enhances BTas-induced LTR transcription in MDBK cells (Fig. 3E). These results suggest that BRelB modulates BTas-mediated BFV transcription in bovine cells, as human RelB does in human cells.

Both the BTas interaction domain and TAD of RelB are required for enhancing the transactivation function of BTas.

Based on the above-described results, we asked whether RelB depends on its BTas interaction domain to augment the transactivation function of BTas. To this end, a portion of the RHD (aa 236 to 285) that participates in the interaction with BTas (Fig. 2) was deleted to create the RelB(Δ236-285) mutant (Fig. 4 A, top). The subsequent coimmunoprecipitation assay verified that RelB(Δ236-285) did not interact with BTas (Fig. 4A, bottom). In contrast with full-length RelB, RelB(Δ236-285) lacks the ability to enhance BTas-induced LTR transcription (Fig. 4B). These results suggest that the RelB-BTas interaction is essential for RelB to regulate the transactivation activity of BTas.

FIG. 4.

Both the BTas interaction domain and the TAD of RelB are required for enhancing the transactivation function of BTas. (A, top) Schematic representation of the BTas-binding domain-deleted form of RelB. (Bottom) 293T cells were cotransfected with the Myc-RelB deletion together with an empty vector (lane 1) or pCMV-BTas (lane 2). BTas was immunoprecipitated with anti-BTas agarose beads, and the immunoprecipitates were detected by Western blotting using an anti-Myc or anti-BTas antibody. The amount of RelB in each cell lysate was checked by using an anti-Myc antibody. (B) HeLa cells were cotransfected with LTR-Luc plasmid (0.05 μg), wild-type RelB or the deletion form of RelB (0.3 μg) with or without pCMV-BTas (0.3 μg), and an empty vector to keep the total DNA concentration in each transfection mixture constant. After 48 h the relative luciferase activities were analyzed. The expression of RelB and its deletion mutant in the transfected cells was monitored by Western blot assay (bottom). Error bars represent SD of the means from three independent experiments. (C) HeLa cells were cotransfected with LTR-Luc plasmid (0.05 μg), wild-type RelB or the indicated RelB truncation expression plasmid (0.3 μg) with or without pCMV-BTas (0.3 μg), and an empty vector to keep the total DNA concentration in each transfection mixture constant. After 48 h the relative luciferase activities were analyzed. The expression of RelB and its truncations in the transfected cells was monitored by Western blot assay (bottom). Error bars represent SD of the means from three independent experiments.

Because the RelB protein contains two functional domains, the RHD and TAD, we next examined whether the TAD is required for RelB to enhance BTas-induced LTR transcription as well. Two RelB mutants, RelB(1-285) and RelB(1-377), were tested. Both mutants lack the TAD but still interacted with BTas (Fig. 2). The results of the luciferase assay showed that RelB(1-285) and RelB(1-377) were unable to enhance BTas-induced LTR transcription, indicating the requirement of the TAD for RelB to enhance the transactivation activity of BTas (Fig. 4C). Together, these results imply that RelB is recruited to the BFV LTR through an interaction with BTas, which further increases the transcriptional efficiency of the LTR.

RelB depends on nuclear localization to enhance the transactivation of BTas.

As RelB interacts with BTas and enhances the transactivation activity of BTas, this function should be dependent on the nuclear localization of RelB. To verify this, p100, a member of the Rel protein family, was used; p100 binds to and sequesters RelB within the cytoplasm. Consistent with this expectation, the results of immunofluorescence staining showed that Myc-p100 and Flag-RelB were located within the cytoplasm and nucleus, respectively (Fig. 5 A). The coexpression of Myc-p100 and Flag-RelB resulted in a redistribution of Flag-RelB from the nucleus to the cytoplasm (Fig. 5A). We next tested whether the p100-mediated cytoplasmic retention of RelB impaired the ability of RelB to regulate the transactivation activity of BTas. The results showed that p100 prevented RelB from enhancing BTas-induced LTR transcription in a dose-dependent manner (Fig. 5B). This result indicates that p100 sequesters RelB in the cytoplasm, thereby preventing RelB from entering the nucleus and enhancing BTas-induced LTR transcription. Taken together, the data suggest that nuclear RelB acts as a cotransactivator of BTas during BFV transcription.

FIG. 5.

RelB depends on nuclear localization to enhance the transactivation function of BTas. (A) Indirect IFA was used to localize Flag-RelB (with mouse anti-Flag antibody and FITC-conjugated goat anti-mouse secondary antibody) and Myc-p100 (with rabbit anti-Myc antibody and Texas Red-conjugated goat anti-rabbit secondary antibody). Nuclei were visualized with DAPI staining. (Top and middle) Individual expression of Flag-tagged RelB and Myc-tagged p100 in HeLa cells. (Bottom) Coexpression of Flag-tagged RelB and Myc-tagged p100 in HeLa cells. (B) HeLa cells were transfected with LTR-Luc plasmid (0.05 μg), pCMV-RelB (0.3 μg), the indicated amount of a p100 expression plasmid in the presence or absence of pCMV-BTas (0.3 μg), and an empty vector to keep the total DNA concentration in each transfection mixture constant. After 48 h the relative luciferase activities were analyzed. Error bars represent SD of the means from three independent experiments.

BTas depends on endogenous RelB to transactivate the BFV LTR.

Interestingly, as shown in Fig. 5B, transfection with a large amount of the p100 expression plasmid (600 ng) not only completely eliminated the RelB-enhanced transactivation of BTas but also partially inhibited the basal level of BTas-induced LTR transcription. We therefore proposed that a connection may exist between BTas transactivation and the endogenous RelB protein, i.e., the noncanonical NF-κB pathway in host cells. To verify this connection, HeLa cells were transfected with pLTR-Luc and pCMV-BTas, together with the stimulator (NIK) or the inhibitor (p100) of the noncanonical NF-κB pathway. The results of the luciferase assay showed that NIK augmented BTas-induced LTR transcription by 4-fold, whereas p100 led to a 20% decrease (Fig. 6 A). These data suggest that modulating the noncanonical NF-κB pathway alters BTas-induced LTR transcription, which possibly relates to cellular endogenous RelB. To address this possibility, we assessed to which extent endogenous RelB contributes to BTas transactivation. We depleted endogenous RelB in HeLa cells by using a mixture of siRNA oligonucleotides and then measured the transactivation of the BFV LTR by BTas in these cells. The results of Western blot analyses and RT-PCR showed that the siRNA oligonucleotides effectively knocked down RelB in HeLa cells (Fig. 6B). In comparison to control siRNA, RelB siRNA (siRelB) decreased luciferase expression from the BFV LTR in the presence of BTas but did not exert any effect on BFV LTR transcription in the absence of BTas (Fig. 6C). Together, these data suggest that BTas depends on endogenous RelB to efficiently transactivate the BFV LTR promoter.

FIG. 6.

BTas depends on endogenous RelB to transactivate the BFV LTR. (A) HeLa cells were transfected with LTR-Luc plasmid (0.05 μg), NIK or p100 expression plasmid (0.3 μg) in the presence or absence of pCMV-BTas (0.3 μg), and an empty vector to keep the total DNA concentration in each transfection mixture constant. After 48 h the relative luciferase activities were analyzed. Error bars represent SD of the means from three independent experiments. (B) HeLa cells were cotransfected with 1 μg of control siRNA or RelB siRNA together with RelB. After 24 h, the same amounts of cell lysates were subjected to immunoprecipitation using RelB antibody followed by Western blotting using anti-RelB and anti-actin (left). HeLa cells were transfected with 1 μg of control siRNA and RelB siRNA. After 24 h, the mRNA levels of RelB were analyzed by RT-PCR (right). (C) HeLa cells were cotransfected with LTR-Luc plasmid (0.05 μg), control siRNA or RelB siRNA (1 μg) with or without pCMV-BTas (0.3 μg), and an empty vector to keep the total DNA concentration in each transfection mixture constant. After 48 h the relative luciferase activities were analyzed.

Endogenous RelB binds to the BFV LTR in BFV-infected cells.

To further examine whether the endogenous RelB protein is a component of the viral transcriptional complex in BFV-infected cells, a ChIP assay was performed to determine the binding of endogenous RelB to BFV LTR DNA in BFV-infected HeLa cells. The ChIP assay was first validated by the successful immunoprecipitation of GAPDH promoter DNA with RNA polymerase II antibody (Fig. 7). We then performed a ChIP assay using antibodies against BTas or RelB. The results of the PCR and real-time PCR analyses demonstrate a significant difference in the LTR DNA content in BTas or RelB samples versus IgG controls, suggesting that BFV LTR DNA was immunoprecipitated by both BTas and RelB antibodies (Fig. 7). These results indicate that RelB, together with BTas, associates with BFV LTR DNA and enhances the transactivation activity of BTas during BFV replication.

FIG. 7.

RelB is a component of the viral transcriptional complex in BFV-infected cells. A total of 10⁷ HeLa cells were infected by BFV. At 24 h after infection, these cells were subjected to a ChIP assay. Subsequently, PCR and real-time PCR assays were carried out to detect LTR DNA in the immunoprecipitated chromatin fragments.

BFV infection enhances expression of RelB through the NF-κB pathway in host cells.

Since RelB is involved in BFV transcription, we asked whether BFV infection, in turn, affects RelB expression in host cells. It is known that the RelB promoter region contains two tandem NF-κB-binding sites (3) (Fig. 8 A). Given that BFV infection, through BTas, activates the NF-κB pathway (34), we predicted that BFV infection might upregulate the transcription of relB. To test this possibility, we measured RelB mRNA levels in HeLa cells that had been infected by BFV. The results of the real-time RT-PCR analysis show that the level of RelB mRNA increased by 1.4-fold and 2.4-fold 6 h and 72 h post-BFV infection, respectively (Fig. 8B). We next tested whether BTas alone is able to stimulate relB expression. To this end, HeLa cells were transfected with TRAF6 (an inducer of the NF-κB pathway), BTas, or an empty vector. Twenty-four hours after transfection, the level of RelB mRNA and protein was determined by real-time RT-PCR and Western blotting. The results show that that both TRAF6 and BTas increased felB mRNA expression levels by 1.6- and 2.5-fold (Fig. 8C). Western blot analysis further verified the RT-PCR results, showing that RelB expression is upregulated by BTas (Fig. 8D). To determine whether BTas enhanced RelB expression by activating the NF-κB pathway, we employed an IκBα(AA) mutant to block the activation of NF-κB. This mutant has two serine residues at positions 32 and 36 mutated to alanine, which is able to efficiently inhibit the cellular NF-κB pathway. The results of the real-time RT-PCR analysis show that the IκBα(AA) mutant markedly attenuated the ability of BTas to increase RelB expression (Fig. 8E). Taken together, these results suggest that BFV infection upregulates RelB expression in HeLa cells as a result of the activation of the NF-κB pathway by BTas. Furthermore, this result suggests a close association between the cellular NF-κB pathway and BTas-mediated BFV transcription. To confirm this association, HeLa cells were transfected with pLTR-Luc and pCMV-BTas, together with a stimulator (TRAF6) or inhibitor [IκBα(AA)] of the NF-κB pathway. The results of the luciferase analysis show that BTas-induced LTR transcription was augmented by TRAF6 and intensely inhibited by IκBα(AA) (Fig. 8F). Taken together, the results suggest that BFV infection (BTas) upregulates the expression of RelB by activating the NF-κB pathway in host cells and that the increased amount of the RelB protein, in turn, enhances BFV transcription.

FIG. 8.

BFV infection enhances expression of RelB through the NF-κB pathway in host cells. (A) Schematic representation of the promoter of RelB showing the NF-κB-binding sites. (B) RNA was extracted from HeLa cells at the indicated times after infection with BFV and subjected to a real-time RT-PCR assay. The level of RelB expression was normalized to GAPDH by using the 2^ΔΔCT method. Error bars represent SD of the means from three independent experiments. (C) HeLa cells were transfected with an empty vector, pTRAF6, or pCMV-BTas (0.3 μg), and the mRNA level of RelB was analyzed by a real-time RT-PCR assay using the 2^ΔΔCT method 24 h after transfection. Error bars represent SD of the means from three independent experiments. (D) HeLa cells were transfected with an empty vector or pCMV-BTas (0.3 μg). After 24 h, the same amounts of cell lysates were subjected to immunoprecipitation using RelB antibody followed by Western blotting using anti-RelB and anti-β-actin antibodies. (E) HeLa cells were cotransfected with an empty vector or BTas (0.3 μg) and with an empty vector or pIκBα(AA) (0.5 μg), and the mRNA level of RelB was analyzed by a real-time RT-PCR assay using the 2^ΔΔCT method 24 h after transfection. Error bars represent SD of the means from three independent experiments. (F) HeLa cells were transfected with LTR-Luc plasmid (0.05 μg), a TRAF6 or IκBα(AA) expression plasmid (0.3 μg) in the presence or absence of pCMV-BTas (0.3 μg), and an empty vector to keep the total DNA concentration in each transfection mixture constant. After 48 h, the relative luciferase activities were analyzed. Error bars represent SD of the means from three independent experiments.

DISCUSSION

Viruses are obligate intracellular parasites and depend on cellular machinery for their efficient replication. For example, the NF-κB pathway is hijacked by a number of viruses, such as HIV-1 and herpes simplex virus type 1 (HSV-1). Our previous studies showed that BFV also activates the NF-κB pathway to enhance viral transcription (34). However, in contrast to HIV-1 and HSV-1, whose promoters contain NF-κB-binding sites, the BFV LTR does not have a functional NF-κB-binding site. The results of the present study demonstrate that the activated NF-κB pathway first increases the level of expression of RelB that subsequently interacts with BTas and enhances transcription from the BFV LTR promoter. This is the first report that viruses use RelB as a cotransactivator to promote viral transcription. Thus, in comparison to HIV-1 and HSV-1, BFV has evolved a different strategy to exploit the NF-κB pathway to support efficient viral transcription.

Although our findings cannot reveal the complete molecular details of the interaction between the NF-κB pathway and BFV, they provide an important clue to explain how the NF-κB pathway is involved in BFV transcription. According to the data shown here, we speculate that in BFV-infected cells, BTas recruits the cellular RelB protein to form a transcriptional complex, which recognizes and binds to the LTR through the DNA-binding domain of BTas and facilitates the recruitment of other coactivators and the displacement of repressors through the TAD of RelB to ultimately initiate LTR transcription. RelB does not bear a DNA-binding domain. In order to modulate gene transcription, RelB usually forms a heterodimer with other DNA-binding proteins to exert its function. For example, RelB forms a heterodimer with p52, and p52 recruits RelB to the target gene through the specific recognition of promoter sequences. In addition, RelB interacts with the aryl hydrocarbon receptor (AhR) to activate the transcription of chemokines such as interleukin-8 (IL-8) (33). In this study, we showed that RelB is recruited to the BFV LTR through interactions with BTas and enhances the BTas-mediated transactivation of the BFV LTR promoter.

The cellular RelB protein is regulated at two levels: the transcriptional level and the posttranscriptional level. Both layers of regulation depend on the activation of the NF-κB pathway. The RelB promoter contains two NF-κB-binding sites. Therefore, the activation of NF-κB leads to an increased level of transcription of RelB. Although the transcriptional level of RelB is tightly regulated, its nuclear translocation is based largely on the activation of noncanonical NF-κB (i.e., p100 processing). In quiescent cells, at the posttranscriptional level, a majority of the RelB protein exists within the cytoplasm as a heterodimer with p100. In the presence of activated NF-κB-inducing kinase (NIK) and IKKα (IKK complex α-subunit), p100 is processed to become p52. The RelB/p52 heterodimer then translocates into the nucleus and activates gene expression. Some viral proteins, such as Tax (human T-cell leukemia virus type 1 [HTLV-1]) and LMP1 (Epstein-Barr virus [EBV]), are able to stimulate RelB nuclear translocation through the activation of the noncanonical NF-κB pathway (18, 26). In addition, our previous study demonstrated that BTas can activate the noncanonical NF-κB pathway by inducing the processing of p100 (34). Taken together, BFV regulates RelB at both the transcriptional and posttranscriptional levels through diverse mechanisms, which implies that RelB is a convergent node for the interaction between BFV and its host.

In summary, we propose a model to illustrate the interplay between BFV and the NF-κB pathway through the RelB protein (Fig. 9). Following BFV infection, BTas is expressed as an immediate-early viral protein. Together with cellular cofactors such as RelB, BTas enhances viral gene expression. Meanwhile, BTas also activates the NF-κB pathway, which in turn increases RelB expression levels and stimulates RelB nuclear translocation through the processing of p100 to p52. These events elevate the amount of nuclear RelB protein, which is exploited by BTas to further enhance viral gene expression. A positive virus-host feedback circuit is thereby described in this study, in which BFV hijacks the host's signaling pathway for its efficient transcription. Although our data demonstrate that RelB plays a critical role in BFV transcription, it is not clear whether some other cellular components specifically participate in RelB-enhanced BTas transactivation, which may function as inhibitors or stimulators of this process. Hence, the identification of other components of the viral transcriptional complex in response to NF-κB activation will further enhance our current understanding of this virus-host cross talk.

FIG. 9.

Model depicting the interaction between BFV and the host cell through the NF-κB pathway and RelB.