Research Article: CAGE-seq analysis of Epstein-Barr virus lytic gene transcription: 3 kinetic classes from 2 mechanisms

Date Published: June 4, 2018

Publisher: Public Library of Science

Author(s): Reza Djavadian, Mitchell Hayes, Eric Johannsen, Erik K. Flemington.


Epstein-Barr virus (EBV) lytic replication proceeds through an ordered cascade of gene expression that integrates lytic DNA amplification and late gene transcription. We and others previously demonstrated that 6 EBV proteins that have orthologs in β- and γ-, but not in α-herpesviruses, mediate late gene transcription in a lytic DNA replication-dependent manner. We proposed a model in which the βγ gene-encoded viral pre-initiation complex (vPIC) mediates transcription from newly replicated viral DNA. While this model explains the dependence of late gene transcription on lytic DNA replication, it does not account for this dependence in α-herpesviruses nor for recent reports that some EBV late genes are transcribed independently of vPIC. To rigorously define which transcription start sites (TSS) are dependent on viral lytic DNA replication or the βγ complex, we performed Cap Analysis of Gene Expression (CAGE)-seq on cells infected with wildtype EBV or EBV mutants defective for DNA replication, βγ function, or lacking an origin of lytic replication (OriLyt). This approach identified 16 true-late, 32 early, and 16 TSS that are active at low levels early and are further upregulated in a DNA replication-dependent manner (leaky late). Almost all late gene transcription is vPIC-dependent, with BCRF1 (vIL10), BDLF2, and BDLF3 transcripts being notable exceptions. We present evidence that leaky late transcription is not due to a distinct mechanism, but results from superimposition of the early and late transcription mechanisms at the same promoter. Our results represent the most comprehensive characterization of EBV lytic gene expression kinetics reported to date and suggest that most, but not all EBV late genes are vPIC-dependent.

Partial Text

Epstein-Barr virus (EBV) is a γ-herpesvirus that infects more than 95% of the human adult population. If acquired early in life, EBV infection is generally asymptomatic, however infection in adolescence may lead to infectious mononucleosis [1,2]. EBV infection is associated with a wide-spectrum of malignancies, such as Burkitt lymphoma, Hodgkin lymphoma, diffuse large B cell lymphoma, NK/T-cell lymphoma, post-transplant lymphoproliferative disease, nasopharyngeal carcinoma, and gastric carcinoma [3–6]. Although these tumors are characterized by latent EBV infection, an increasing body of evidence indicates that lytic infection is also important for the emergence of EBV-associated malignancies [7–10].

In this study, we present a comprehensive analysis of the kinetics of EBV lytic gene transcription based on dependence on lytic DNA replication and expression of BDLF4 (a βγ gene encoding an essential component of the viral late gene pre-initiation complex (vPIC) [30,31]). We chose CAGE-seq to accurately quantify transcription start sites in a strand-specific manner. By using this approach, we were able to avoid measurement of overlapping transcripts which confounds most other genome-wide attempts to analyze EBV lytic gene expression. CAGE-seq identified 32 TSS clusters that were expressed independently of lytic DNA replication (i.e. early kinetics) and 16 TSS that exhibited late kinetics defined by their strict dependence upon lytic DNA replication. In addition, 16 TSS were identified that exhibited partial dependence on lytic DNA replication (referred to here as leaky late, but also termed early-late or γ1). Although BALF2 has been suggested to play a role in recruiting Zta to the BHLF1 promoter [47], we consistently found that genes requiring BALF2 for their expression also required OriLyt (S1 Table). Thus, for each TSS, BALF2-dependence reflects lytic DNA replication dependence.