Transcriptional analysis of the unfolded protein response (UPR) and lymphoma microenvironment during Marek's disease virus (MDV) infection
Date
2015
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Marek’s disease virus (MDV) is a highly cell-associated alphaherpesvirus, capable of replicating in multiple cell types (B cells, T cells, macrophages, epithelial and fibroblast cells) and specifically transforming activated CD4+ T helper cells, polarizing them into a T regulatory (TREG)-like immunophenotype. Marek’s disease (MD) is caused by MDV and is one of the most prevalent diseases in poultry production, worldwide. MDV is associated with profound immune suppression and the rapid formation of T cell lymphomas. Real time RT-PCR is widely used in the field of MD research to measure transcriptional responses to infection and/or vaccination. Studies in the past have either used cellular β-actin or GAPDH as internal reference genes, although the stability of their expression in the context of MD infection was never investigated. We investigated the suitability of five housekeeping genes (β-actin, 28S RNA, 18S RNA, GAPDH, Peptidyl-prolyl-isomerase B aka cyclophilin B or PPIB ) as standard internal controls during lytic infection in vitro in CEFs and latent infection in vivo in TK strain-induced tumors. Upon Bestkeeper® and Normfinder® analysis of stability, we found that β-actin is the least stable reference gene, while both PPIB and 28S RNA displayed equally higher stability in the context of MD infection both in vitro and in vivo. MDV serves as an excellent model to study how herpesvirus lytic replication triggers cellular stress activation and how the virus modifies the malefic consequences of cellular stress (translation attenuation, ERAD and apoptosis) while allowing beneficial effects (chaperone induction) that support viral replication. In a preliminary study, we investigated the induction of ER stress and activation of the unfolded protein response pathways (UPR) during the course of MDV1 (mildly virulent CU-2) lytic replication in CEFs. We observed a lack of induction of UPR signaling until day 4 post-infection, suggesting that UPR signaling was maintained in a repressed state during this initial phase of infection. However by day 5, we observed a significant transcriptional induction of ATF6 (GRP78/BiP) and IRE1 (XBP(S)) pathways while the PERK (ATF4) pathway was still maintained in a repressive state. Based on the transcriptional responses observed on day 5 with mildly virulent CU-2, we followed up our investigation with MDV1 pathotypes of higher virulence, RB1B (very virulent), MD5 (very virulent) and TKING (very virulent plus). Among the different MDV pathotypes, the vv+ MDV strain (TK) induced the highest level of UPR gene expression compared to other pathotypes at 5 days post-infection, despite a more limited replication in these cells (typically, vv+MDVs, require adaptation or a higher number of passages to replicate in CEF). UPR induction seen in vvMDV (RB1B and MD5) infection is relatively lower than that of vv+MDV infected cells indicating that UPR pathways might be more tightly regulated by vvMDVs. Tumor cells are subjected to severe conditions such as hypoxia, glucose deprivation, proto-oncogene activation, rapid proliferation and increased cytokine secretion. Each of these is capable of inducing ER stress and UPR activation in the cancer cells. UPR plays a paradoxical role in tumor cells and the tumor microenvironment, either promoting adaptation and cell survival under acute stress, or triggering apoptosis upon failure of adaptation and chronic ER stress. However, many tumor cells also possess the ability to block apoptosis under chronic ER stress. Prolonged activation of PERK (Protein kinase R-like ER Kinase) and IRE1-JNK (Inositol requiring kinase 1- Jun N terminal Kinase) pathways induce apoptosis while IRE1-XBP1(s) and ATF6 (activating transcription factor 6) and IRE1 pathways are anti-apoptotic. Upon investigation of the UPR pathways in TK strain-induced tumors (which are latently infected with the TK virus) we found a significant transcriptional induction of the ATF6 pathway target, GRP94 (glucose regulatory protein 94). Although not significant, there was an increased transcriptional induction of ATF6 and IRE1 pathway targets, GRP78 (glucose regulatory protein 78) and XBP1(s) (spliced X-box binding protein 1) respectively, while we observed no induction of the PERK pathway target, ATF4 (activating transcription factor 4). Overall, anti-apoptotic mechanisms appear to be induced in the TK-induced tumors. Anti-apoptotic mechanisms mediated by cellular bZIP transactivators, ATF6 and XBP1(s) could be independent of Meq, as bZIP domains of human ATF6 and XBP1(s) were found not to interact with Meq bZIP domain either on a coiled coil array surface or in solution (130). Based on the induction of these targets, MD lymphomas can serve as an excellent model for the targeting of ATF6 and IRE1-XBP1 pathways in order to specifically ablate tumor cells. While previous studies that analyzed cytokine gene expression (at transcriptomic or proteomic level) in transformed lymphomas have compared sorted CD30hi (transformed) and CD30lo cells populations, we profiled the cytokines at the transcriptional level in the TK induced lymphomas by comparing frank lymphomatous masses to surrounding visibly normal and putatively non-transformed tissue in affected spleens. This approach would define responses in lymphoma tissue environment that contributes to tumor initiation and progression. We found a significant transcriptional induction of cytokines in lymphomas belonging to the following cell signatures, TH1 (IFN-γ, IL-2, T-bet), TH2 (IL-4), TH17 (IL-17, IL-21, IL- 6, TGF-β) and TREG immunophenotypes (IL-4, IL-6, IL-10, TGF-β, IL-2).