Interferon Regulatory Factor 7 Is Activated by a Viral Oncoprotein through RIP-Dependent Ubiquitination^▿

Cell lines.

HEK293 cells were maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum (FBS). RIP^+/+ TNFR^−/− and RIP^−/− TNFR^−/− immortalized fibroblasts were maintained in Dulbecco's modified Eagle's medium supplemented with 10% FBS and 2 mM glutamine. All other cell lines were maintained in RPMI with 10% FBS. SVT35 Jurkat T lymphoma cells (parental) and RIP-deficient 35.3.13 Jurkat cells have been described previously and were a gift from Brian Seed (Massachusetts General Hospital, Boston) (36). Raji is a Burkitt's lymphoma B-cell line. DG75-TA and DG75-LMP1 cell lines have been described elsewhere (9).

Plasmids and reagents.

IRF7 in pcDNA3 and IRF7 in pcLMP1 have been described previously (46, 47). Flag-IRF7 in pCMV2 and the IFN-β promoter construct were gifts of John Hiscott (McGill University, Canada) and have been described previously (22). Wild-type RIP and kinase-dead RIP (kd-RIP) (K45A) were provided by William Cance (University of Florida, Gainesville). Hemagglutinin-tagged Ub (HA-Ub) and HA-K63-only Ub (see below) in pcDNA3 were generously provided by David Boone and Averil Ma (University of California, San Francisco). Dominant negative TRAF6 (TRAF6DN) was provided by Arnd Kieser (GSF National Research Center for Environment and Health, Germany). The IFN-stimulated response element (ISRE)-Luc construct and pRL-TK (where TK is thymidine kinase) were purchased from Clontech and Promega, respectively. IRF7 and RIP antibodies were purchased from Santa Cruz; LMP1 antibody (cs 1-4) was purchased from DAKO; and the Flag-M2, Ub, and HA antibodies were purchased from Sigma.

Transfections and luciferase assays.

HEK293 cells in 12-well plates were transfected with 0.1 μg of IFN-β promoter reporter plasmid or ISRE-Luc construct, 0.1 μg of pcIRF7, and 0.1 μg of LMP1 or RIP unless otherwise noted. Empty vector was used to equalize the total amount of DNA in the transfections. pRL-TK (0.025 μg) served as an internal transfection control. SVT35 and RIP-deficient Jurkat cells (8.5 × 10⁵) were transfected with 1.5 μg of IFN-β promoter reporter plasmid, 2 μg pcIRF7, 2 μg pcLMP1, and 1.5 μg of pRL-TK. All transfections were done with Fugene 6 transfection reagent (Roche). Luciferase assays were performed 18 to 24 h posttransfection with a dual luciferase assay kit (Promega).

Immunoprecipitations and immunoblot analysis.

Raji, DG75-TA, and DG75-LMP1 cells were lysed in immunoprecipitation (IP) lysis buffer (10 mM Tris, pH 7.5, 150 mM NaCl, 5 mM EDTA, 0.5% NP-40, 5 mM dithiothreitol, 0.2 mM Na orthovanadate, and Complete protease inhibitors). Cell extracts (1 mg) were incubated with 2 μg IRF7 or RIP antibodies overnight at 4°C. Protein A/G-Sepharose beads (Santa Cruz) were added and samples incubated for 1 h at 4°C. Precipitates were washed five times with 10 volumes of IP lysis buffer and eluted by boiling the beads for 5 min in 2× sodium dodecyl sulfate (SDS) loading buffer. The eluted proteins were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes followed by immunoblotting with the indicated antibodies. HEK293 cells in 60-mm dishes and RIP^+/+ or RIP^−/− immortalized fibroblasts in 100-mm dishes were transfected with combinations of Flag-IRF7, pcLMP1, and HA-Ub expression vectors with the use of Lipofectamine 2000 (Invitrogen). Cells were harvested 24 to 48 h posttransfection and lysed in IP lysis buffer. Cell extracts were incubated with 2 μg Flag M2 antibody overnight at 4°C. Protein A/G-Sepharose beads were then added, and samples were incubated for 1 h at 4°C and washed five times with 10 volumes IP lysis buffer. For sequential immunoprecipitations, the washed beads were resuspended in 50 μl 1% SDS and 1 mM dithiothreitol and heated to 95°C for 5 min. After removal of the beads by centrifugation, the eluted proteins were diluted 20-fold with IP lysis buffer and incubated with 2 μg IRF7 antibody overnight at 4°C. Protein A/G-Sepharose beads were added for 1 h at 4°C, followed by five washes with 10 volumes of IP lysis buffer. The immunoprecipitated proteins were eluted by boiling for 5 min in 2× SDS loading buffer and separated by SDS-PAGE followed by immunoblotting with the indicated antibodies.

RNA interference.

HEK293 cells were transfected with RIP small interfering RNA (siRNA) oligomer, purchased from Santa Cruz, with the use of Lipofectamine 2000 according to the manufacturer's instructions. The specific sequence is not supplied by the manufacturer. The control siRNA, which is a scrambled sequence that does not lead to the specific degradation of any known cellular mRNA, was also purchased from Santa Cruz. After 48 h, siRNA-treated cells were transfected again with 0.1 μg reporter plasmid, 0.1 μg pcIRF7, and 0.1 μg pcLMP1 with Lipofectamine 2000. Cells were harvested 24 h later. Luciferase assays were performed as described above.

Retroviral production and infection.

pBabe/RIP1 was made by PCR, subcloning RIP1 into the pBabe/Hygro vector at the EcoRI/SalI site. 293T cells in 100-mm dishes were cotransfected with 7 μg pVSV-G and pGAG-pol and either 7 μg pBabe/GFP (where GFP is green fluorescent protein) or pBabe/RIP1 with the use of a CalPhos mammalian transfection kit (BD Biosciences, Bedford, MA). Supernatant fluids containing GFP or RIP1 retroviruses were collected 48 h later. For retroviral infections, RIP^−/− cells were plated in 100-mm dishes and infected with GFP or RIP1 retroviruses in the presence of Polybrene (8 μg/ml). The cells were selected in hygromycin (Invitrogen) for 2 weeks (40).

RIP1 activates IRF7.

Although we have demonstrated previously that LMP1 activates IRF7 and enhances its transcriptional activity (45), the mediators of this activation of IRF7 have not been identified. RIP interacts with LMP1 CTAR2 (14), but a specific role for RIP in LMP1 signaling has remained unassigned. However, RIP mediates activation of IRF3 in response to dsRNA (2), and thus we thought that RIP may mediate activation of IRF7 by LMP1. To test this hypothesis, we first examined whether expression of RIP enhances the transcriptional activity of IRF7. HEK293 cells, which do not express endogenous IRF7 under our culture conditions, were transfected with IRF7, RIP, and an IFN-β promoter reporter plasmid, a target of IRF7 and commonly used index of IRF7 activity. Cotransfection of IRF7 and increasing concentrations of RIP activated the IFN-β promoter in a dose-dependent manner. RIP's effect on this promoter was dependent on IRF7; very little if any activity was detected upon transfection of RIP alone (Fig. 1A). The results suggest that RIP activates IRF7.

We next used a kd-RIP (K45A) mutant to examine whether RIP's enzymatic activity is required for activation of IRF7. The combinations of IRF7 and wild-type RIP or kd-RIP both activated the IFN-β promoter (Fig. 1B). Thus, RIP's kinase activity is not required for the activation of IRF7.

RIP1 is involved in LMP1-mediated activation of IRF7.

To test further whether RIP is implicated in activation of IRF7 by LMP1, we examined IRF7 transactivation of the IFN-β promoter reporter in response to LMP1 in RIP-deficient Jurkat cells. Whereas very little activity was detected upon transfection of IRF7 or LMP1 alone into parental Jurkat cells, promoter activity was substantially increased upon cotransfection of IRF7 and LMP1, indicating that LMP1 activates IRF7 in these cells. However, the LMP1-plus-IRF7-dependent IFN-β promoter activity was reduced in RIP-deficient Jurkat cells compared with that in parental Jurkat cells (Fig. 2A). In addition, we used siRNA targeting RIP to suppress its expression in 293 cells (Fig. 2B) and then examined the ability of IRF7 to transactivate an ISRE-luciferase reporter, another common index of IRF7 activity, in response to LMP1. LMP1-plus-IRF7-dependent ISRE activity was reduced significantly after treatment with RIP siRNA oligomer compared with activity after treatment with control siRNA (Fig. 2C). Together these data demonstrate that RIP mediates activation of IRF7 by LMP1.

RIP1 and IRF7 interact.

We next examined whether RIP and IRF7 interact. Raji is a Burkitt's lymphoma cell line that expresses RIP and due to latent EBV infection also expresses LMP1 and IRF7. Therefore, these cells are useful for examining the interaction of endogenously expressed RIP and IRF7. RIP and IRF7 were immunoprecipitated from Raji cell lysates with RIP and IRF7 antibodies, followed by immunoblotting for RIP. Two bands, neither detected with the control antibody, were observed upon immunoprecipitating RIP (Fig. 3A). The 74-kDa band corresponded to the expected size of RIP, while the other was a more slowly migrating, most likely modified, form of RIP. RIP is known to be ubiquitinated in response to TNF-α (6, 20, 49), and it is likely that RIP is ubiquitinated in this system as well. Therefore, to determine if ubiquitinated proteins immunoprecipitate with RIP, the samples were immunoblotted with an antibody to Ub, and bands, corresponding to the size observed for the modified form of RIP, were detected (Fig. 3B). These results indicate that RIP is ubiquitinated in Raji cells. Interestingly, ubiquitinated RIP specifically coimmunoprecipitated with IRF7 while very little if any unconjugated RIP was associated with IRF7 (Fig. 3A). Together, these data suggest that IRF7 preferentially interacts with ubiquitinated RIP in a physiological setting.

LMP1 stimulates ubiquitination of RIP.

LMP1 signaling is constitutively active. The detection of substantial amounts of ubiquitinated RIP in Raji cells raises the question of whether LMP1 signaling promotes the ubiquitination of RIP. To examine the effect of LMP1 on RIP, we utilized DG75 cells stably transfected with LMP1 or control vector. DG75 is an EBV-negative Burkitt's lymphoma cell line that endogenously expresses IRF7 and RIP. RIP was immunoprecipitated from DG75-LMP1 or DG75-TA control cell lysates, and the amounts of unconjugated and ubiquitinated RIP were compared by Western blotting. Although levels of unconjugated RIP were similar in the immunoprecipitates from both cell lines, more ubiquitinated RIP was consistently detected in the LMP1-positive cells (Fig. 3C). These data indicate that LMP1 promotes ubiquitination of RIP and suggest that some signals transduced through RIP by LMP1 may be Ub dependent.

LMP1 stimulates the interaction of RIP with IRF7.

The facts that RIP contributes to activation of IRF7 by LMP1 and that LMP1 promotes the ubiquitination of RIP and its association with IRF7 in cells suggest that interaction of RIP with IRF7 is part of the mechanism for activation of IRF7. Therefore, we next used DG75 cells stably transfected with LMP1 to ask specifically whether LMP1 promotes the association between RIP and IRF7. Immunoprecipitation of IRF7 from DG75-LMP1 or DG75-TA cell lysates followed by immunoblotting for RIP revealed that ubiquitinated RIP interacted with IRF7 in DG75-LMP1 cells (Fig. 3D, left), similarly to what was observed in Raji cells (Fig. 3A). RIP immunoprecipitated with RIP antibody from DG75-LMP1 cells was included to provide a size marker for unconjugated and ubiquitinated RIP. However, we detected little if any RIP coimmunoprecipitating with IRF7 from DG75-TA cells (Fig. 3D, left), although similar amounts of IRF7 were immunoprecipitated from both the LMP1-positive and the LMP1-negative cell lines (Fig. 3D, right). These data suggest that LMP1 stimulates the interaction between endogenous IRF7 and ubiquitinated RIP.

LMP1 stimulates ubiquitination of IRF7.

The interaction between endogenous RIP and IRF7 in Raji cells was detected in both experimental sequences: immunoprecipitation of RIP followed by immunoblotting for IRF7 revealed that a small amount of IRF7 immunoprecipitated nonspecifically with the control antibody. However, much more IRF7 coimmunoprecipitated with RIP (Fig. 4A). Interestingly, high-molecular-weight, most likely modified, forms of IRF7 were detected upon immunoprecipitation of IRF7, and these modified forms of IRF7 also interacted with RIP (Fig. 4A). IRF7 has recently been shown to be activated in a Ub-dependent manner (16), and the diffuse nature of the band is consistent with ubiquitination. Moreover, ubiquitinated proteins were immunoprecipitated with IRF7 from Raji cell lysates (Fig. 3B).

That modified—probably ubiquitinated—IRF7 interacts with RIP in Raji cells suggests that IRF7 may be ubiquitinated in response to LMP1 signaling. Therefore, we determined whether LMP1 promotes ubiquitination of IRF7. After transfection of 293 cells with HA-Ub and Flag-tagged IRF7 expression vectors alone or together with an LMP1 expression vector, cells were lysed and IRF7 was immunoprecipitated with Flag antibody followed by immunoblotting for HA or Flag. Cotransfection of LMP1 and IRF7 produced additional slowly migrating forms of IRF7 not readily detected in the absence of LMP1 (Fig. 4B, left). Furthermore, consistently more Ub was immunoprecipitated from lysates containing both IRF7 and LMP1 (Fig. 4B, right). Together these data suggest that LMP1 promotes ubiquitination of IRF7.

Several types of polyubiquitin chains linked through distinct lysines of Ub may have different functional outcomes. Polyubiquitination linked through K48 targets proteins for proteasomal-dependent degradation. However, polyubiquitination linked through other Ub lysines, such as K63, can have proteasome-independent functions, which include regulating protein-protein interactions and activation of signaling molecules, such as IKK (35). Therefore, we next asked whether LMP1 can promote K63-linked ubiquitination of IRF7. A Ub vector in which all lysine residues except K63 are mutated to arginine (HA-K63-only Ub) (4) along with Flag-IRF7 alone or in combination with LMP1 was transfected into 293 cells. Sequential immunoprecipitations were used to favor detection of ubiquitinated IRF7 but not ubiquitination of factors associating with IRF7. Proteins immunoprecipitated first with Flag antibody were heated to 95°C for 5 min in 1% SDS, followed by a second immunoprecipitation with IRF7 antibody. Interestingly, immunoblotting for HA revealed low levels of spontaneous ubiquitination of IRF7 upon cotransfection of K63-only Ub. However, addition of LMP1 clearly increased the amount of ubiquitinated IRF7 (Fig. 4C, right). Moreover, immunoblotting for Flag revealed a high-molecular-weight smear, consistent with an increase in ubiquitinated protein, in the LMP1-positive sample that was barely detectable in the LMP1-negative sample (Fig. 4C, left). These results indicate that LMP1 promotes K63-linked ubiquitination of IRF7.

Ub enhances IRF7 activity.

That LMP1 promotes ubiquitination of IRF7 at least in part through K63-linked Ub suggests that Ub may regulate the activity of IRF7. To test further whether ubiquitination affects IRF7 transcriptional activity, 293 cells were transiently transfected with expression vectors for IRF7 and HA-tagged Ub together with an ISRE-luciferase reporter construct. The coexpression of IRF7 and Ub resulted in a dose-dependent increase in ISRE activity (Fig. 5A). Moreover, coexpression of IRF7 and the K63-only Ub mutant also increased ISRE activity (Fig. 5B). The results indicate that Ub enhances the activation of IRF7.

RIP contributes to LMP1-promoted ubiquitination of IRF7.

Because RIP interacts with unconjugated and ubiquitinated IRF7 and contributes to activation of IRF7 by LMP1, we next asked whether RIP contributes to the ubiquitination of IRF7 promoted by LMP1. RIP^+/+ and RIP^−/− immortalized fibroblasts were transfected with HA-Ub and Flag-IRF7 with or without LMP1. Flag-IRF7 was immunoprecipitated with Flag antibody. The samples were boiled and reprecipitated with IRF7 antibody before detection of ubiquitination by immunoblotting for HA. Ubiquitinated IRF7 detected in small amounts in RIP^+/+ cells in the absence of LMP1 was greatly increased by LMP1 (Fig. 6A, top). In striking contrast, ubiquitination of IRF7 was not detected in RIP^−/− cells even in the presence of LMP1, although similar amounts of IRF7 were present in each sample (Fig. 6A, bottom). Finally, reconstitution of RIP^−/− cells with RIP restored ubiquitination of IRF7 (Fig. 6B). The results indicate that RIP mediates LMP1-promoted ubiquitination of IRF7.

TRAF6 may be involved in LMP1-mediated activation of IRF7.

TRAF6, a Ub ligase that catalyzes K63-linked Ub chains, interacts with and activates IRF7 in a Ub-dependent manner in response to TLR9 signaling (16). Therefore, we next examined whether TRAF6 is involved in LMP1-mediated activation of IRF7. After transfection of IRF7 and LMP1 or RIP with and without TRAF6DN into 293 cells, we assessed activation of an ISRE-luciferase reporter by IRF7. TRAF6DN blocked IRF7 activation by both LMP1 and RIP (Fig. 7), a result consistent with the idea that TRAF6 is involved in LMP1/RIP-mediated IRF7 activation.