Interaction of host and viral microRNAs with infectious laryngotracheitis virus transcripts
Infectious Laryngotracheitis Virus (ILTV) is an Alphaherpesvirus of the domesticated chicken and other economically important fowl such as pheasants, peafowl and turkeys. It causes an upper respiratory disease that is clinically characterised by dyspnoea, rales and expulsion of a thick, sometimes hemorrhagic, tracheal exudate. Incidences of mortality range from 10 – 70 % whilst morbidity ranges from 50 – 100 %. The disease causes significant financial losses to the poultry industry through bird death, stunted growth and a marked decrease in egg production. Due to its economic importance, attenuated live-vaccines have been developed by serial passage of virus either in eggs or tissue culture. These have the ability to protect birds against ILTV however they do not stop latent infection which can result in reactivation of the virus termed ‘vaccinal laryngotracheitis’. The molecular biology underlying virus-host interactions for ILTV is poorly understood and there are large gaps in knowledge regarding the pathogenesis of ILTV infection. MicroRNAs (miRNAs) are short, non-coding RNAs that post-transcriptionally regulate gene expression through targeting of specific mRNAs. Several herpesviruses have been shown to encode miRNAs that have the ability to regulate both viral and cellular gene expression which can impact virus-host interactions. Previous work in the literature has shown that ILTV encodes for 10 miRNAs with sparse data on what they may be regulating. It was hypothesised that the virus-encoded miRNAs may have an effect upon the pathogenesis of the virus by targeting both cellular and viral mRNAs. To investigate this hypothesis initially, the biochemical technique CLASH (Cross-Linking and Sequencing of Hybrids) was attempted however a lack of suitable reagents such as physiologically relevant cell lines of chicken origin made this technically challenging and this approach was halted. Instead, a bioinformatic approach was developed and split into two avenues of research. Firstly, it was hypothesised that miRNAs encoded by ILTV would target virally derived transcripts. As the virus genome is poorly annotated, transcripts for all 79 open reading frames (ORFs) were created manually using an arbitrary system of 1000 bp upstream of the ATG start site and 50 bp downstream of the designated PolyA tail. These were then fed into the online algorithm RNA Hybrid alongside sequences for all 10 virus-encoded miRNAs. Results from the bioinformatic predictions were then sorted and filtered using pre-defined conditions. This left a total of 227 predicted interactions. These were then filtered again leaving 28 novel targets that were screened in a reporter based system. Three of the predicted interactions showed a decrease in luciferase-reporter activity compared to the siRNA control (UL24, UL29 and UL46/48), however only the latter two showed statistically significant decreases in activity of 15 % and 20 % respectively. Mutation of the seed sequences in both UL29 and UL46/48 targets abrogated the effects of the miRNA mimic. Further work on UL29 and its interaction with ILTV-miR-I2 looked at validating this interaction by western blotting however these results were inconclusive. Investigations into the interaction between UL46/48 and ILTV-miR-I6-5p first confirmed by RT-PCR that UL46 was targeted by ILTV-miR-I6-5p. Validation of the interaction between UL46 and ILTV-miR-I6-5p by western blotting was inconclusive. Investigations into the interplay between UL46, UL48 and the ICP4 promoter were also characterised with UL46 able to negatively modulate the effects of UL48 on ICP4 promoter activity in a reporter-based system. Secondly, the same viral transcripts were then used in conjunction with high confidence chicken miRNAs as per MiRBase (Release 21, Jun 2014). The sorting and filtering of results mirrored that of the viral transcript study giving a final list of 103 predicted targets. From the list, three targets were picked that were all targeted by the cellular miRNA ggamiR- 133a-3p and tested using the same reporter system. Two targets, one in UL20 and one in the coding region of ICP4 showed no statistical difference between the miRNA mimic and siRNA control. In contrast, one target, located in the 5’UTR of ICP4 and confirmed by RTPCR to be within the expressed mRNA transcript was found to cause a 55 % reduction in luciferase activity. This effect was then abrogated upon mutation of the miRNA seed sequence. Further investigations found that this miRNA can cause an apparent reduction in virus titer and a statistically significant decrease in plaque size morphology when virus is harvested from cells transfected with the miRNA mimic and used to infect naïve cells. Moreover, a combination RT-qPCR and sequencing was used to confirm the sequence of gga-miR-133a-3p in several tissues of the chicken including the Dorsal Root Ganglia (DRG) and Harderian gland (HG). These are of importance to ILTV biology as the DRG is a site of latent infection and the HG is a secondary lymphoid organ (SLO) in the bird which monitors the upper respiratory tract, the site of lytic replication/clinical symptoms. Finally, CRISPR-Cas9 genome editing was used to delete a cluster of five miRNAs from the viral genome. Guide RNAs (sgRNAs) were designed to target the miRNA cluster and shown to efficiently direct cleavage of target DNA in an in vitro system. Following transfection/infection of cells, virus was harvested and subsequent sequencing showed that this approach was successful in creating a recombinant ILTV. This was detectable after passage of the virus through naïve cells although a pure population of recombinant virus was not obtained due to a lack of time.