Unexpected patterns of Epstein–Barr virus transcription revealed by a High throughput PCR array for absolute quantification of viral mRNA

Rosemary J Tierney,Claire D Shannon-Lowe,Leah Fitzsimmons,Andrew I Bell,Martin Rowe

doi:10.1016/j.virol.2014.10.030

Abstract

We have validated a flexible, high-throughput and relatively inexpensive RT-QPCR array platform for absolute quantification of Epstein–Barr virus transcripts in different latent and lytic infection states. Several novel observations are reported. First, during infection of normal B cells, Wp-initiated latent gene transcripts remain far more abundant following activation of the Cp promoter than was hitherto suspected. Second, EBNA1 transcript levels are remarkably low in all forms of latency, typically ranging from 1 to 10 transcripts per cell. EBNA3A, -3B and -3C transcripts are likewise very low in Latency III, typically at levels similar to or less than EBNA1 transcripts. Thirdly, a subset of lytic gene transcripts is detectable in Burkitt lymphoma lines at low levels, including: BILF1, which has oncogenic properties, and the poorly characterized LF1, LF2 and LF3 genes. Analysis of seven African BL biopsies confirmed this transcription profile but additionally revealed significant expression of LMP2 transcripts.

Highlights

Primary infection of resting B cells in vitro with EBV leads to the activation of the multimerised Wp promoter and expression of EBNA2, EBNA-LP and latent BHRF1, followed by activation of a second promoter Cp and eventually expression of all six EBNAs (EBNA1, 2, 3A, 3B, 3C and LP) and, from their own promoters, three latent membrane proteins (LMP1, 2A and 2B) (Amoroso et al, 2011; Kelly et al, 2009; Rickinson and Kieff, 2007; Tierney et al, 2011)
We have optimized and validated the high throughput Fluidigm Dynamic Array system to simultaneously quantify 48 EBV and cellular transcripts in up to 48 RNA samples. This platform is flexible such that individual QPCR assays can be substituted according to the needs of a particular experiment and expanded to a 96 Â 96 format if screening for additional transcripts is required in the future
Due to developments in sequencing technologies, RNA-Seq is rapidly becoming the gold standard method for gene expression profiling, and two recent RNA-Seq studies have characterized in detail the EBV transcriptome in Latency I (Lat I) Burkitt lymphoma (BL) cells (Concha et al, 2012; Lin et al, 2010)

Summary

Introduction

Primary infection of resting B cells in vitro with EBV leads to the activation of the multimerised Wp promoter and expression of EBNA2, EBNA-LP and latent BHRF1, followed by activation of a second promoter Cp and eventually expression of all six EBNAs (EBNA1, 2, 3A, 3B, 3C and LP) and, from their own promoters, three latent membrane proteins (LMP1, 2A and 2B) (Amoroso et al, 2011; Kelly et al, 2009; Rickinson and Kieff, 2007; Tierney et al, 2011). The non-coding EBERs (Arrand and Rymo, 1982; Lerner et al, 1981), BARTs (Chen et al, 1999; Gilligan et al, 1990; Sadler and Raab-Traub, 1995) and a series of miRNAs (Amoroso et al, 2011; Pfeffer et al, 2004) are transcribed This pattern of latent gene expression, which drives B cell growth transformation and the establishment of permanently growing lymphoblastoid cell lines (LCLs), has been classified as Latency III, or Lat III (Rickinson and Kieff, 2007; Rowe et al, 1992). These latency definitions are not absolute, but represent points on a spectrum of EBV gene expression which may occur at different times or different anatomical sites during EBV latency in vivo (Thorley-Lawson et al, 2013)

Methods

Results

Conclusion