Analysis of the Host Stress Response to Ebola Virus Infection and Generation of a Recombinant Marburg Virus Expressing EGFP to Study Viral Spread
Ebola virus (EBOV) and Marburg virus (MARV) belong to the filovirus family and cause outbreaks with case fatality rates up to 90%. Currently there is no approved vaccine or antiviral treatment available. Therefore filoviruses are classified as priority A select agents, which can only be handled in h...
Saved in:
Main Author: | |
---|---|
Contributors: | |
Format: | Doctoral Thesis |
Language: | English |
Published: |
Philipps-Universität Marburg
2012
|
Subjects: | |
Online Access: | PDF Full Text |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Ebola virus (EBOV) and Marburg virus (MARV) belong to the filovirus family and cause outbreaks with case fatality rates up to 90%. Currently there is no approved vaccine or antiviral treatment available. Therefore filoviruses are classified as priority A select agents, which can only be handled in high containment biosafety level 4 laboratories.
Generation of an infectious Marburg virus clone expressing EGFP
The generation of recombinant viruses expressing enhanced green fluorescent protein (EGFP) has significantly improved the study of their replication cycle and opened up the possibility for the rapid screening of antiviral drugs. The goal of this part of the work was to generate a recombinant MARV expressing EGFP from an additional transcription unit inserted between the second and third genes, encoding VP35 and VP40, respectively. The recombinant MARV containing the EGFP gene (rMARV-EGFP) was successfully rescued and used in live-cell imaging to follow viral spread in real time, revealing EGFP expression at 32 hours post infection (hpi), and infection of neighboring cells at 55 hpi. A slight growth restriction of rMARV-EGFP compared to the wt rMARV was observed which might be due to the additional gene insertion.
During filovirus infection characteristic viral inclusions are formed in the cytoplasm of infected cells, induced by self-aggregation of the nucleoprotein (NP). Immunofluorescence analysis of rMARV-EGFP-infected cells revealed an accumulation of EGFP in these viral inclusions. This was reproduced by transient expression of both EGFP and other fluorescent proteins (FPs) along with filovirus nucleocapsid proteins, which further showed that NP-induced inclusion formation was sufficient for the recruitment. In contrast, ectopic GFP fusion proteins containing a localization signal were not relocated into inclusions formed by MARV NP and VP35. Taken together, the observed relocalization of ectopically expressed, untagged FPs suggests an unspecific recruitment to the viral inclusions based on a weak interaction. Interestingly, EGFP aggregates observed by autofluorescence were undetected by antibody-based immunofluorescence. This indicates that antibodies might not be able to penetrate viral inclusions.
Cellular stress response to Ebola virus infection
EBOV is known to antagonize various antiviral signaling pathways including the interferon response. Here, the interaction of EBOV with another host antiviral defense mechanism, the cellular stress response, was analyzed. During environmental stress a small range of kinases phosphorylate the eukaryotic initiation factor 2α (eIF2α), which plays a central role in the control of translational arrest. Cellular stress is also accompanied by the formation of cytoplasmic stress granules (SGs) and processing bodies (PBs) containing stalled messenger ribonucleoprotein bodies. Since viruses depend on the cellular translation apparatus for viral protein synthesis, many viruses have evolved strategies to antagonize cellular stress response mechanisms. Here we demonstrate that EBOV interferes with the cellular stress response. Since, SGs and PBs are highly dynamic structures, which cannot be isolated from cells, most of the studies presented here were performed by microscopic analysis. Formation of endogenous SGs was not observed in EBOV-infected cells, suggesting that filovirus infection per se does not induce a stress response. This raised the question of whether SG formation is actively inhibited in EBOV-infected cells. The oxidative stressor sodium arsenite (As) induces eIF2α phosphorylation and SG formation. After As-
treatment, SGs formed in EBOV-infected cells, albeit in fewer cells compared to non-infected cells, indicating that EBOV is able to inhibit As-induced SG formation to a certain level.
For better visualization SG formation was further analyzed in U2OS cells expressing an EGFP-tagged SG maker protein, ras-GAP SH3 domain binding protein 1 (G3BP). In EBOV-infected G3BP-EGFP-expressing cells, SG formation was induced by As and Hippuristanol (Hip). Hip leads to SG formation by inhibiting eIF4A-dependent translation initiation, a process, which does not involve phosphorylation of eIF2α. In some EBOV-infected cells treated with As, G3BP-EGFP was diffusely distributed and SG formation was impaired. This was not seen in Hip-treated cells, suggesting that EBOV inhibits SG formation to a certain level in response to phospho-eIF2α-mediated stress. This inhibition seemed to be dependent on the size of the viral inclusions, suggesting that the level of viral protein expression is important for the inhibition.
Intriguingly, G3BP-EGFP SG-like aggregates were observed within the viral inclusions in unstressed and in stressed EBOV-infected cells. Transient expression of viral nucleocapsid proteins leading to inclusion formation was not sufficient for the aggregation of G3BP granules inside the inclusions, suggesting that other viral components, including the viral RNA, are needed to sequester SG components in the inclusions. Together this data suggest a mechanism of EBOV to interfere with SGs by sequestering SG components in the viral inclusions.
To further understand how EBOV interacts with SGs, the EBOV RNA-binding proteins NP, VP30, and VP35 were individually examined for their ability to interact with SGs. While VP30 colocalized with SGs without affecting their structure, the double-strand (ds) RNA-binding protein VP35 inhibited SG formation induced by phospho-eIF2α-mediated stress when expressed at high levels. However, when VP30 and VP35 were coexpressed with NP, both proteins were relocated into NP-derived viral inclusions and colocalization with SGs was strongly reduced or absent, which was also observed in EBOV-infected G3BP-EGFP cells. This indicates that only free VP35 and VP30 are able to colocalize with SGs and that both proteins are preferentially relocated to viral inclusions. It is conceivable that the amount of free VP35 and VP30 in infected cells is too low to be detected by IFA.
In As-treated cells the dsRNA-dependent protein kinase (PKR) is activated by the cellular protein, PKR activating protein (PACT). Activated PKR phosphorylates eIF2α, inducing SG formation. In EBOV-infected cells, activation of PKR by dsRNA has been shown to be efficiently inhibited by VP35. Our data revealed that EBOV was not able to block As-induced PKR activation. This indicates that PKR activation by PACT cannot be antagonized by VP35. Previous studies have reported that VP35 binds to PACT in unstressed cells (Fabozzi et al., 2011) Here we show that the VP35-PACT interaction was disrupted in stressed cells. This indicates that VP35 loses the ability to sequester PACT from binding to PKR, such that PKR can be activated. This further explains the presence of SGs in EBOV-infected cells but the questions remains, how eIF2α-induced SG formation is inhibited in EBOV-infected and VP35 expressing cells? Possible mechanisms include the inhibitory function of VP35 on SG formation or the sequestering of SG components.
To analyze the role of PBs in EBOV infection, a U2OS cell line expressing the PB marker protein mRNA-decapping enzyme 1A (DCP1a) fused to mRFP was used. In EBOV-infected cells, As-induced PBs were observed to surround viral inclusions, suggesting a recruitment of PBs. In transient transfection experiments, PBs colocalized with VP35 but not with VP30 or NP. Expression of VP35 in U2OS cells containing both G3BP-EGFP and DCP1a-mRFP resulted in diffusely aggregated SGs that colocalized with VP35 and intermingled with PBs. This suggests that VP35 interacts and links constituents of PBs and SGs. The results indicate that EBOV exhibits antiviral control strategies at the level of the host stress response, where VP35 functions as a key player, since it directly interacts with cellular stress components of both SGs and PBs. Furthermore, the binding between VP35 and PACT seems to play an important role in the control of PKR. |
---|---|
DOI: | 10.17192/z2012.0492 |