Table of Contents:
The Ebola virus is a human pathogenic RNA virus which can cause hemorrhagic fever with high case fatality rates up to 90%. Unpredictable and sporadic outbreaks occurring mainly in Africa are characteristic for this virus. Owing to the recent outbreak with over 28 000 infected people and about 11 000 deaths, the Ebola virus has attracted increasing public interest.
The Ebola virus harbors a negative single strand RNA genome of 19 kb with seven protein coding genes. For replication of the viral genome three of the viral proteins are necessary: the RNA-dependent RNA polymerase (L), the viral protein 35 (VP35) and the nucleoprotein(NP). In contrast, for transcription a fourth protein is required, namely the viral protein 30(VP30). Phosphorylation of two serine clusters in the N-proximal region of VP30 regulates the protein´s ability to support viral transcription; in its dephosphorylated state it functions as a transcription factor and interacts with VP35; by phosphorylation VP30 becomes transcriptionally inactive and loses the contact to VP35 while gaining affinity for the nucleoprotein. This results in a switch of the RNA polymerase to the replicative mode. The strict VP30 dependency of transcription is mediated by a stable hairpin structure on the viral genome/antigenome (nt 56-78) and can be abolished by weakening the structural stability of this hairpin. VP30 is known to bind RNA and to form a hexamer consisting of three dimers. Moreover, it possesses an unconventional zinc finger motif (Cys3His) and was shown to act as suppressor of RNA silencing in addition to VP35 and VP40.
The central goal of this work was to investigate the RNA binding ability of VP30 with regard to its function as transcription factor. This gave rise to four part projects dealing with (i) the establishment of an appropriate in vitro assay to analyze RNA:VP30 complex formation,(ii) the characterization of the RNA binding properties of VP30, (iii) the possible correlation between transcriptional activation and RNA binding of VP30 and (iv) a comprehensive
transcriptome analysis to deepen the understanding of the transcriptional regulation by VP30.
The RNA:VP30 interaction studies were mainly based on an EMSA (Electrophoretic Mobility Shift Assay) procedure tailored to VP30, which enabled us to measure binding affinities (in terms of apparent Kd values) and to visualize different types of RNA:VP30 complexes under native conditions. The results suggested that the RNA:VP30 interaction is strongly governed by electrostatic interactions. The preferential substrates were found to be single stranded RNAs with an optimal length of ~ 37 nt. Initially, we analyzed the 3´-terminal ~ 150 nt of the viral genome as well as the corresponding complementary antigenomic sequence for potential structures. A deletion analysis based on these two RNA substrates confirmed the preference of VP30 for single stranded regions. In addition, an affinity-enhancing effect of the hairpin structure (nt 56-78) could be observed. The comparison of different 5´-modifications(5´-Mono- or Triphosphate, 5´-Cap(0)) in an RNA substrate mimicking the 5´-terminal
sequence of the NP mRNA revealed a strong negative influence on VP30 binding by the presence of a 5´-Cap modification. To explore the correlation between RNA binding and transcriptional activation of VP30, the transcriptionally inactive VP30 variants VP30_DD (mimicking the phosphorylated state), VP30_5LA (hexamerization-defective) and VP30_C72S (mutated zinc finger motif) as well as
the transcriptionally active wild type and VP30_AA (mimicking the unphosphorylated state) were produced recombinantly. These VP30 variants were tested for their ability to bind RNA using the newly established EMSA procedure. Variants VP30_5LA and VP30_C72S - in line with their transcriptional inactivity - were essentially unable to bind RNA. Similarly, the two
counterparts VP30_AA and VP30_DD showed RNA binding affinities that correspond to their ability to support viral transcription: VP30_AA (transcriptionally active) was able to bind RNA with a 4.3-fold lower Kd value than VP30_DD (transcriptionally inactive). The influence of the arginine cluster (overlapping with the two serine clusters) on RNA binding ability and
transcriptional activation was investigated using three single mutants (VP30_R26A, _R28A,_R40A) as well as one triple mutant (VP30_3RA). All single mutants had little effect on RNA binding or transcriptional activation. Solely the triple mutant showed a twofold reduced RNA binding affinity, an altered RNA binding mode and a substantially impaired transcriptional activation. Additionally, we found that the complex formation between VP30 and VP35 strictly
depends on the presence of RNA. In summary, we observed a strong correlation between RNA binding and transcriptional activation using different VP30 variants, thus pinpointing the physiological link between VP30´s RNA binding ability and its function as transcription factor.
In the course of a transcriptome analysis using Next Generation Sequencing (NGS) we were able to detect short transcripts in the mRNA start regions whose 3´-end coincided with the predicted hairpin structures there. Furthermore a leader transcript and two additional short transcripts were assigned to the intergenic region between the VP30 and VP24 genes. The 3´-leader transcript preceding the NP gene was confirmed experimentally by Northern Blot and its ratio to the first mRNA (NP) determined using RT-qPCR (mRNA/leader transcript:
~8.7:1). This is the first experimental evidence for the existence of a 3´-leader transcript of Ebola virus, which may exert - analogous to leader transcripts found in other viruses - a key function in the regulatory switch between viral transcription and replication.