X-box-binding protein 1 activates lytic Epstein-Barr virus gene expression in combination with protein kinase D

Abstract

Epstein-Barr virus (EBV) establishes a latent form of infection in memory B cells, while antibody-secreting plasma cells often harbor the lytic form of infection. The switch between latent and lytic EBV infection is mediated by the two viral immediate-early proteins BZLF1 (Z) and BRLF1 (R), which are not expressed in latently infected B cells. Here we demonstrate that a cellular transcription factor that plays an essential role in plasma cell differentiation, X-box-binding protein 1 (XBP-1), also activates the transcription of the two EBV immediate-early gene promoters. In reporter gene assays, XBP-1 alone was sufficient to activate the R promoter, whereas the combination of XBP-1 and protein kinase D (PKD) was required for efficient activation of the Z promoter. Most importantly, the expression of XBP-1 and activated PKD was sufficient to induce lytic viral gene expression in EBV-positive nasopharyngeal carcinoma cells and lymphoblastoid cells, while an XBP-1 small interfering RNA inhibited constitutive lytic EBV gene expression in lymphoblastoid cells. These results suggest that the plasma cell differentiation factor XBP-1, in combination with activated PKD, can mediate the reactivation of EBV, thereby allowing the viral life cycle to be intimately linked to plasma cell differentiation.

FIG. 1.

XBP-1s and PKD synergistically activate lytic EBV gene expression in HONE-1/EBV cells. HONE-1/EBV cells were transfected with empty vector (−), XBP-1s, constitutively active PKD (PKD), or the combination of XBP-1s and PKD. Immunoblot analysis of cellular extracts was performed 2 days later to quantitate the expression of two different lytic EBV proteins (Z and BMRF1), a cellular protein (β-actin), and transfected XBP-1s protein. +, present.

FIG. 2.

XBP-1s and PKD synergistically activate lytic EBV gene expression in early-passage LCLs. LCLs were transfected with empty vector (−), the XBP-1s expression vector, constitutively active PKD (PKD), the combination of XBP-1s and PKD, or a Z expression vector. Immunoblot analysis of cellular extracts was performed 2 days later to quantitate the expression of the lytic EBV proteins (BMRF1, R, and Z) and cellular β-actin. +, present; LCL/EBV, EBV-infected LCLs.

FIG. 3.

XBP-1s and PKD activation of Rp versus Zp. HeLa cells were transfected with vectors containing the luciferase (Luc) gene linked to either the R (Rp-LUC) or Z (Zp-LUC) promoter in combination with either a control plasmid (−), an XBP-1s expression vector, a constitutively active form of PKD (PKD), or both XBP-1s and PKD, as indicated. Luciferase activity was measured 2 days later. Results are presented as the increase (n-fold) in the amount of luciferase activity after transfection with XBP-1s, PKD, or both compared with the amount of activity after transfection with the control vector, as indicated. +, present.

FIG. 4.

XBP-1-responsive regions in Rp. Various 5′ deletions in Rp-LUC were constructed as indicated, and HeLa cells were transfected with the constructs in combination with either empty vector (−) or the vector expressing XBP-1s (+). Results are presented as the increase (n-fold) in the amount of luciferase activity after transfection with XBP-1s versus the vector control for each promoter construct.

FIG. 5.

MEF2D and CRE motifs contribute to XBP-1s-PKD activation of Zp. (A) HeLa cells were cotransfected with reporter gene constructs containing the wild-type Zp (−221 to +12) inserted upstream of the CAT gene or containing site-directed mutations in the Zp altering the MEF2D or CRE motifs and either a vector control or the XBP-1s expression vector. CAT activity was quantitated 2 days later. Results are presented as the increase (n-fold) in the amount of CAT activity after transfection with XBP-1s compared with the amount of activity after transfection with the control vector as indicated for each promoter construct. +, present; −, absent. (B) EMSA was performed using in vitro-translated XBP-1s (+) or untranslated rabbit reticulocyte lysate (−) in combination with either the positive control oligonucleotide probe containing a consensus XBP-1s-binding motif or a probe containing the Zp CRE motif in the absence (−) or presence (+) of anti-XBP-1 antibody (Ab). (C) HeLa cells were transfected with Zp-LUC constructs with either wild-type Zp or Zp containing a site-directed mutation in the ZEB binding motif (ΔZEB) along with either empty vector or vector expressing XBP-1s. Luciferase activity was measured 2 days later.

FIG. 6.

XBP-1s and HDAC inhibitors CaMKIV and valproic acid activate lytic EBV gene expression. (A) HONE-1/EBV cells were transfected with empty vector (−), XBP-1s, or constitutively active CaMKIV either alone or in combination, as indicated. Cell extracts were examined by immunoblotting for the expression of the lytic EBV proteins (BMRF1, R, and Z) and cellular β-actin 2 days later. +, present. (B) HONE-1/EBV cells were transfected with empty vector or XBP-1s in the presence (+) or absence (−) or valproic acid (VPA; 1 mM). Extracts were analyzed by immunoblotting 2 days later for the lytic EBV proteins (BMRF1, R, and Z) and cellular β-actin.

FIG. 7.

XBP-1 is required for lytic EBV gene expression in LCLs. EBV-transformed B cells were transfected with an siRNA directed against XBP-1 or equal amounts of a control siRNA. Results from two separate experiments are shown. (A) The level of XBP-1 expression was examined by RT-PCR analysis 2 days after siRNA delivery. PCR amplifications were performed using undiluted cDNA, as well as various dilutions of the cDNA, as indicated. +, present; −, absent. (B) The expression level of the lytic viral Z and R genes was examined by RT-PCR. (C) The expression of the β₂-microglobulin (β₂-M) message was examined by RT-PCR. (D) The expression of the early lytic EBV BMRF1 protein, as well as cellular actin, was examined by immunoblot analysis 2 days after siRNA delivery.