Journal of General Virology - Current Issue

The complexity and speed of evolution in viruses with RNA genomes makes predictive identification of variants with epidemic or pandemic potential challenging. In recent years, machine learning has become an increasingly capable technology for addressing this challenge, as advances in methods and computational power have dramatically improved the performance of models and led to their widespread adoption across industries and disciplines. Nascent applications of machine learning technology to virus research have now expanded, providing new tools for handling large-scale datasets and leading to a reshaping of existing workflows for phenotype prediction, phylogenetic analysis, drug discovery and more. This review explores how machine learning has been applied to and has impacted the study of viruses, before addressing the strengths and limitations of its techniques and finally highlighting the next steps that are needed for the technology to reach its full potential in this challenging and ever-relevant research area.

Translation errors, impaired folding or environmental stressors (e.g. infection) can all lead to an increase in the presence of misfolded proteins. These activate cellular responses to their removal, including intracellular protein degradation activities. Protein ubiquitylation is involved in two major degradation pathways, the ubiquitin-proteasome system and selective autophagy. In humans, the ubiquitin-like modifier-activating enzyme 1 (UBA1) is the primary E1 enzyme in the ubiquitin conjugation cascade. Viruses have evolved to exploit protein degradation pathways to complete their infection cycles. Zika virus (ZIKV) is an emerging orthoflavivirus causing serious neurologic disorders in neonates (congenital microcephaly) and adults (Guillain–Barré syndrome). Non-structural protein 5 (NS5), the largest and most conserved protein in the orthoflaviviruses, catalyses the synthesis and capping of new viral genomes. In addition to viral RNA replication in the cytoplasm, ZIKV NS5 is translocated into the nucleus to interfere with host antiviral responses. Here, we demonstrate that ZIKV NS5 co-immunoprecipitates with cellular UBA1. Immunofluorescence assays suggest that this interaction takes place primarily in the nucleus of an infected cell, although colocalization of both proteins is also detected in the cytosol. RNA interference-mediated depletion of UBA1 leads to reduced virus titres in the infected cells, while transient overexpression of UBA1 favours faster replication kinetics, with higher virus titres and protein levels detected. Moreover, UBA1-targeting drugs cause significant drops in virus infectivity. These results support a proviral role for UBA1 during ZIKV infection and encourage the potential use of inhibitors against this enzyme or its NS5-interacting epitopes as potential therapeutic targets.

Species A rotaviruses (RVs), which belong to the family Sedoreoviridae and contain a genome of 11 segmented dsRNA segments, are a leading cause of severe acute gastroenteritis in infants and children younger than 5 years of age. We previously developed a strategy to recover rotavirus vaccine strain LLR from 11 cloned plasmids. Here, we report an improved reverse genetics system for LLR by combining two or three transcriptional cassettes in a single plasmid, which substantially enhances rescue efficiency from 66.7% (8/12) to 91.7 % (11/12). Furthermore, the recombinant LLR stably expressing NLuc was rescued based on the five-plasmid reverse genetics system. Improvements to the rotavirus reverse genetics system will enhance its applicability for studies of rotavirus biology and clinical use.

The Betacoronavirus murine hepatitis virus (MHV) is an important model system for studying coronavirus (CoV) molecular and cell biology. Despite this, few reagents for MHV are available through repositories such as ATCC or Addgene, potentially limiting the widespread adoption of MHV as a tractable model system. To overcome some challenges inherent in the existing MHV reverse genetics systems, we developed a plasmid-launched transformation-associated recombination (TAR) cloning-based system to assemble the MHV (strain A59; MHV-A59) genome. Following assembly in yeast, virus replication was launched by transfecting the fully assembled genome into HEK-293T cells. MHV-A59 recovered using this TAR cloning-based approach (WTTAR MHV-A59) replicated with kinetics identical to the virus recovered using a ligation- and T7-based approach (WTLIG MHV-A59). Additionally, WTTAR MHV-A59 can be detected at least 10 h post-transfection without requiring additional nucleocapsid (N) provided in trans. Lastly, we demonstrated the tractability of this TAR cloning-based system by recovering MHV-A59 expressing an 11 amino acid-containing HiBiT tag fused to the C-terminus of spike (S). While this virus, SC MHV-A59, replicated with reduced kinetics compared to WTTAR MHV-A59, the kinetics of virion production could be measured over time directly from the supernatant. This report represents the first plasmid-launched, TAR cloning-based system for MHV-A59. Furthermore, it describes a new reporter virus that could be used to study early steps during MHV-A59 entry and be used in the screening of antiviral compounds. To support future research with MHV-A59, we have made the necessary plasmids for this system available through ATCC.