Effects of the NEDD8-Activating Enzyme Inhibitor MLN4924 on Lytic Reactivation of Kaposi's Sarcoma-Associated Herpesvirus

Abstract

The switch of Kaposi's sarcoma-associated herpesvirus (KSHV) from latency to lytic replication is a key event for viral dissemination and pathogenesis. MLN4924, a novel neddylation inhibitor, reportedly causes the onset of KSHV reactivation but impairs later phases of the viral lytic program in infected cells. Thus far, the molecular mechanism involved in the modulation of the KSHV lytic cycle by MLN4924 is not yet fully understood. Here, we confirmed that treatment of different KSHV-infected primary effusion lymphoma (PEL) cell lines with MLN4924 substantially induces viral lytic protein expression. Due to the key role of the virally encoded ORF50 protein in the latent-to-lytic switch, we investigated its transcriptional regulation by MLN4924. We found that MLN4924 activates the ORF50 promoter (ORF50p) in KSHV-positive cells (but not in KSHV-negative cells), and the RBP-Jκ-binding elements within the promoter are critically required for MLN4924 responsiveness. In KSHV-negative cells, reactivation of the ORF50 promoter by MLN4924 requires the presence of the latency-associated nuclear antigen (LANA). Under such a condition, LANA acts as a repressor to block the ORF50p activity, whereas MLN4924 treatment relieves LANA-mediated repression. Importantly, we showed that LANA is a neddylated protein and can be deneddylated by MLN4924. On the other hand, we revealed that MLN4924 exhibits concentration-dependent biphasic effects on 12-O-tetradecanoylphorbol-13-acetate (TPA)- or sodium butyrate (SB)-induced viral reactivation in PEL cell lines. In other words, low concentrations of MLN4924 promote activation of TPA- or SB-mediated viral reactivation, whereas high concentrations of MLN4924, conversely, inhibit the progression of TPA- or SB-mediated viral lytic program.IMPORTANCE MLN4924 is a neddylation (NEDD8 modification) inhibitor, which currently acts as an anti-cancer drug in clinical trials. Although MLN4924 has been reported to trigger KSHV reactivation, many aspects regarding the action of MLN4924 in regulating the KSHV lytic cycle are not fully understood. Since the KSHV ORF50 protein is the key regulator of viral lytic reactivation, we focus on its transcriptional regulation by MLN4924. We here show that activation of the ORF50 gene by MLN4924 involves the relief of LANA-mediated transcriptional repression. Importantly, we find that LANA is a neddylated protein. To our knowledge, this is the first report showing that neddylation occurs in viral proteins. Additionally, we provide evidence that different concentrations of MLN4924 have opposite effects on TPA-mediated or SB-mediated KSHV lytic cycle activation. Therefore, in clinical application, we propose that MLN4924 needs to be used with caution in combination therapy to treat KSHV-positive subjects.

Keywords: KSHV; Kaposi's sarcoma-associated herpesvirus; LANA; MLN4924; ORF50; lytic reactivation.

FIG 1

MLN4924 disrupts viral latency in PEL cell lines. Three PEL cell lines, BCP1 (A), BCBL1 (B), and BC3 (C), were treated with different concentrations (0.3, 1.0, 2.0, and 5.0 μM) of MLN4924. In parallel, cells treated with TPA or sodium butyrate (SB) were included as positive controls. At 24 or 48 h posttreatment, the expression of viral lytic proteins, including ORF50, K8, and ORF45, was detected by Western blotting.

FIG 2

Effects of MLN4924 treatment on the expression of viral and cellular transcriptional regulators in PEL cell lines. Increasing amounts of MLN4924 were used to treat BCP1 and BCBL1 cells for 24 and 48 h, respectively. The expression levels of viral (ORF50 and LANA) and cellular (RBP-Jκ, KAP1, Nrf2, HIF-1α, c-Fos, c-Jun, phosphorylated c-Jun, SP1, and actin) transcriptional regulators were determined by immunoblotting. Ub, ubiquitin.

FIG 3

Mapping of the MLN4924-responsive element in the ORF50 promoter. (A) Schematic diagram of the full-length and deleted ORF50 promoters in reporter constructs. Several known binding sites for transcription factors in the ORF50 promoter are indicated. (B) Transcriptional activation of the full-length ORF50 promoter (−3801/+10) by MLN4924 in PEL cells. The pORF50p(−3801/+10)/luc reporter construct was transfected into different PEL cell lines (BCP1, BCBL1, and BC3), and the transfected cells were left untreated or treated with different concentrations of MLN4924 for 24 h. The relative luciferase activity of the reporter construct in untreated or MLN4924-treated cells was measured as described in Materials and Methods. Asterisks indicate significant differences in results versus those with the untreated control (P < 0.05). (C) Responses of the ORF50p deletion constructs to MLN4924. BCP1 and BCBL1 cells were transfected with indicated reporter plasmids, and the transfected cells were left untreated or treated with MLN4924 (0.3 μM). Activation of each deleted ORF50p reporter construct by MLN4924 was determined at 24 h after MLN4924 treatment. *, P < 0.05, for results compared to those with pGL3-Basic; #, P < 0.05, for results compared to those with the indicated controls.

FIG 4

The RBP-Jκ-binding motifs in the ORF50 promoter critically confer MLN4924 responsiveness. (A) Responses of 1×RBP-Jκ- and 3×RBP-Jκ-containing reporter constructs to MLN4924. One or three copies of a RBP-Jκ element or its mutant element (mt) were constructed into pE4luc (E4). The indicated reporter plasmids were individually transfected into BCP1 and BCBL1 cells, and the relative reporter activation by MLN4924 (0.3, 1.0, and 2.0 μM) was measured at 24 h posttreatment. Asterisks indicate significant difference in results versus those with the untreated control (P < 0.05). (B) MLN4924 responsiveness of the reporter plasmids containing an HIF-1α-binding element from the ORF50 promoter or a consensus HIF-1α-responsive element (cHIF-1α). 3 × HIF-1α, three copies of viral HIF-1α-binding element from the ORF50 promoter; 3×cHIF-1α, three copies of a consensus HIF-1α response element. (C) Responses of AP1- and SP1-containing reporter constructs to MLN4924. 3×AP1, three copies of an AP1-binding element from the ORF50 promoter; 3×SP1, three copies of an SP1-binding element from the ORF50 promoter. The dashed horizontal lines on the graphs indicate that the reporter activation was maintained at the uninduced level.

FIG 5

MLN4924 relieves LANA-mediated repression of the ORF50 promoter in 293T cells. (A) MLN4924 responsiveness of the ORF50p-driven reporter constructs in 293T cells. 293T cells were transfected with the indicated reporter plasmids and then left untreated or treated with MLN4924 (1.0 μM) for 24 h. The fold activation of these reporter constructs by MLN4924 was determined as the luciferase activity in the presence of MLN4924 divided by luciferase activity in the absence of MLN4924. (B) Western blot analysis of specific cellular transcription factors expressed in 293T cells after treatment with MLN4924 (0.3, 1.0, 2.0, and 5.0 μM) for 24 h. (C) Involvement of LANA in the MLN4924-mediated ORF50p activation in 293T cells. The pORF50p(−3801/+10)/luc reporter construct or the pGL3-Basic reporter vector was cotransfected with the empty vector or the LANA expression plasmid into 293T cells. The transfected cells were left untreated or treated with MLN4924 (1.0 μM) for 24 h. The luciferase activity of pORF50p(−3801/+10)/luc under different treatment conditions was measured relative to the activity of the empty reporter vector pGL3-Basic. *, P < 0.05, for results compared to those with the indicated controls. (D) MLN4924 responsiveness of the ORF50p reporter constructs in LANA-transfected 293T cells. The indicated reporter constructs were individually cotransfected with the LANA expression plasmid into 293T cells, and their MLN4924 responsiveness was determined in these transfected cells after treatment with MLN4924 (1.0 μM) for 24 h. *, P < 0.05, for results compared to those with pGL3-Basic; #, P < 0.05, for results compared to those with the indicated controls. (E) Activation of the 3×RBP-Jκ-containing reporter construct by MLN4924 in LANA-transfected 293T cells. *, P < 0.05, for results compared to those with pE4luc.

FIG 6

LANA is naturally modified with NEDD8 in cells. (A) Neddylation of LANA in 293T cells. 293T cells were transfected with the indicated plasmids expressing HA-NEDD8 or F-LANA and left untreated or treated with MLN4924 (2.0 μM) for 24 h. Cell samples were then subjected to immunoprecipitation (IP) and immunoblot (IB) analysis. (B) Neddylation of LANA in 293T(BAC16) cells. 293T(BAC16) cells were transfected with the expression plasmid for HA-NEDD8 and cultured in medium with or without MLN4924 (2.0 μM) for 24 h. Cell lysates were immunoprecipitated using anti-LANA antibody, and the resultant immunoprecipitates were analyzed by immunoblotting using antibodies against LANA, HA, and NEDD8. (C) Neddylation of LANA in BCBL1 cells. BJAB and BCBL1 cells were left untreated or treated with MLN4924 (2.0 μM) for 24 h. After cell lysates were immunoprecipitated with anti-LANA antibody, the immunoprecipitated proteins were probed with anti-LANA and anti-NEDD8 antibodies. Arrows indicate the positions of the neddylated LANA.

FIG 7

Treatment of PEL cells with MLN4924 triggers the expression of early lytic proteins but not the late lytic protein K8.1. MLN4924 (MLN) at 0.3 μM or 2.0 μM was used to treat BCP-1 (A), BCBL1 (B), and BC3 (C) cells. At different time points (1, 2, and 3 days) after treatment, the expression levels of viral lytic proteins in these PEL cell lines were determined by Western blotting. In parallel, cell samples treated with TPA or SB were also included as positive controls. Bar graphs show densitometry quantification of the maximal expression level of each viral lytic protein induced by MLN4924 (0.3 or 2.0 μM) relative to that induced by TPA or SB. As noted, K8.1 protein expression was undetectable in BCP1 cells even after treatment with TPA. The dashed horizontal lines on the graphs represent the average expression levels of ORF50, ORF45, and K8 induced by MLN4924 in PEL cells. Asterisks indicate significant difference between K8.1 induction and induction of other viral lytic proteins after MLN4924 treatment (P < 0.05). ND, not detected.

FIG 8

The blockage of KSHV lytic progression by MLN4924 is not due to cellular cytotoxicity. (A to C) Western blot detection of K8.1 as well as cleaved PARP and caspase-3 in BCP1, BC3, and BCBL1 cells that were treated with various concentrations of MLN4924 (0.1, 0.3, 1, 2, or 5 μM) for 3 days. In the experiments, PEL cells treated with TPA or SB served as controls. As noted, there are two cleaved PARP fragments (89 and 50 kDa; asterisks) produced in treated cells. (D to F) Quantitative analysis of virus particles from culture supernatants of BCP1, BC3, and BCBL1 cells after treatment with MLN4924, TPA, or SB for 3 days. (G to I) Effects of different concentrations of MLN4924 on cell proliferation of BCP1, BC3, and BCBL1 cells. After PEL cells were left untreated or treated with MLN4924 at the indicated concentrations or with TPA or SB, cell proliferation was measured by XTT assay. DMSO, dimethyl sulfoxide.

FIG 9

MLN4924 exhibits concentration-dependent biphasic actions in viral lytic-cycle activation in TPA-treated BCBL1 cells and in SB-treated BC3 cells. (A and B) BCBL1 and BC3 cells were treated with TPA and SB, respectively, in combination with various concentrations (0, 0.1, 0.3, 1, and 2 μM) of MLN4924. After 3 days of treatment, treated samples were analyzed for K8.1 expression and both cleaved PARP and caspase-3. As noted, there are two cleaved PARP fragments (89 and 50 kDa; asterisks) detected in treated cells. (C and D) The amounts of virus particles released from the above treated cell samples were measured by quantitative PCR. (E) Kinetics of viral lytic protein expression in BCBL1 cells that were treated with MLN4924 (0.1 or 2.0 μM), TPA, or the combination of MLN4924 (0.1 or 2.0 μM) and TPA. The expression levels of viral lytic proteins were examined by Western blotting using the indicated antibodies. (F) Western blot analysis of viral lytic protein expression in BC3 cells that were treated with MLN4924 (0.1 or 2.0 μM), SB, or the combination of MLN4924 (0.1 or 2.0 μM) and SB. (G and H) Densitometry quantification of viral lytic protein expression under different treatment conditions in BCBL1 cells and in BC3 cells. The dashed vertical lines on the graphs are used to separate the expression patterns of viral lytic proteins induced by 2.0 μM MLN4924 along with TPA or SB in BCBL1 cells or in BC3 cells from those induced by other treatment combinations.