Reviewers: 1st-year graduate students in the Biological and Biomedical Sciences Graduate program, Cornell, Ithaca, NY (course BioAP7100)
Fontaine et al. present an interesting investigation of how de novo Zika virus (ZIKV) infection modulates the nonsense-mediated mRNA decay (NMD) pathway in an effort to better understand the molecular mechanism underlying ZIKV-induced microcephaly. The NMD pathway was chosen in part because it has previously been shown to affect brain size in mice. Additionally, Hepatitis C virus, also in Flaviviridae, has been shown to disrupt the NMD pathway. The authors build upon these observations and hypothesize that the NMD pathway may be a possible target for ZIKV and hence a candidate mechanism for causing microcephaly associated with ZIKV infections. In particular, the authors probe whether the ZIKV capsid protein is able to bind to up-frameshift protein 1 (UPF1), a factor that is known to block the NMD pathway. The authors present clear evidence that ZIKV infection increases expression of NMD substrates in multiple cell lines in a dose-dependent manner. Using viral capsid protein as a bait, affinity purification followed by mass-spectroscopy identified several host NMD factors as potential interactors, some of which were further confirmed by co-immunoprecipitation. Interestingly, UPF1 function appears to be down-regulated by the viral capsid protein post-transcriptionally affecting specifically UPF1’s nuclear subpopulation, however the exact mechanism of regulation remained elusive. Finally, using UPF1 depletion studies, evidence is presented to indicate UPF1 as a restriction factor for ZIKV infections in neural progenitor cells (NPCs), suggesting a competition between ZIKV and the host’s surveillance system ultimatley determining infectivity and/or pathogenesis. This study lays the foundation for more mechanstic studies linking ZIKV infection, NMD function and potentially microcephaly. Furthermore, the results could inform new antiviral therapies targeting this pathway.
The authors use an impressive amalgamation of experiments to indicate that ZIKV proteins affect the NMD pathway. We feel the following points would make the manuscript even stronger.
Major points the authors could consider:
1. The authors set out to test the hypothesis that ZIKV infection affects the cellular NMD pathway. They motivate this hypothesis by stating that impairment of the NMD pathway has been linked to microcephaly, which is one of the most severe neuropathological manifestations of ZIKV infection. Importantly, they propose that ZIKV may be post-transcriptionally downregulating UPF1. However, a direct investigation into the relationship between downregulation of UPF1 and growth or proliferation of neuronal cells in the context of microcephaly is missing. This issue appears to represent a disconnect in the relationship between the authors’ findings and microcephaly, and it could be addressed by either providing a stronger, more direct link or by toning down statements regarding the study’s implications in ZIKV infection-associated microcephaly.
2. The authors use multiple cell lines and viral strains, which in principle is a strength regarding to the rigor and reproducibility of the work. However, the authors vary cell lines and strains between experiments even though the experiments are presented as if they build off of one another. This is especially important given the apparent differences between the different cells lines. Also, little to no justification is given as to why these cell lines in particular were chosen. Ideally, all cell lines and strains would be included in the experiments covered in Figure 1 to established direct comparisons, followed by consistent use of a justified selection of host cells and viral strains throughout all future experiments. This information is critical in interpreting the results of the RNA-Seq datasets in Figure 1b (comparing HeLa and NPCs for UPF1 knock-down and ZIKV infection, respectively), as well as the Western blot analysis in Figure 3a/b (comparing Huh7 and NPCs infected with ZIKV). Given that these cell lines likely exhibit different gene expression profiles, it remains unclear whether the similarities and differences observed across experiments are generated by the different cell lines alone or from specific conditions such as ZIKV infection. Furthermore, one possible confounding variable is that these cell lines may react and modulate their NMD pathways differently following infection.
3. Correlations between functional NMD and virus propagation were tested exclusively by knock-down of NMD factor UPF1. One key experiment we would have liked to see is the reverse experiment, using overexpression of UPF1. This experiment could further strengthen the claims that the NMD pathway inhibits ZIKV infection made in the title and discussion are not fully supported. Showing that overexpressing of UPF1 is protective against ZIKV infection would be a valuable addition to the manuscript and support the overall claims regarding an “arms race” between NMD and the virus, and could highlight therapeutic avenues.
4. The authors use predominantly a single time point of 48h post infection for their assays. It was not clear to us why this specific time point was chosen, and whether there is an effect on cell morphology and mortality, which could complicate the analysis. Can compensatory responses that may affect the results be ruled out? These questions are particularly interesting considering that the NMD pathway and UPF1 specifically is discussed as a restriction factor that competes with viral proteins. A time course experiment surveying different times post infection would establish how UPF1 alters viral replication efficiency in host cells and address the uncertainties mentioned here.
Minor points to consider:
1. We are concerned that UPF1 knock-down might cause cells to be more susceptible to stress or any virus. It would be prudent to include cell viability assays as a control and/or infect cells with other viruses.
2. The rationale for selecting UPF1 as opposed to other exon junction complex (EJC) subunits identified in the protein-protein interaction assay is unclear. Please clarify why only UPF1 and not other EJC subunits were targeted in this study.
3. A discussion regarding the link between ZIKV infection and microcephaly, which do not always correlate, may be a valuable addition to the manuscript. In particular, there is a big discrepancy between microcephaly linked to ZIKV infection during the 2015 and 2016 outbreaks. It seems premature to postulate a strong correlation between ZIKV with microcephaly. At he same time, the study presented here has the potential to provide specific mechanistic hypotheses to explain the underlying reasons for the differences in microcephaly occurrences between outbreaks.
4. Please consider providing the rationale for choosing the particular NMD substrates/genes shown in Fig 1a &1b. Likewise, it is not clear why the authors choose UPF3B and UPF1 out of the many NMD factors identified in the pull-down experiment shown in Fig. 2a. Is UPF1 known to play a role in regulating any other viral infections? If so, please mention.
In summary, this manuscript describes an exciting study that establishes a link between the host’s NMD pathway and ZIKV infection, suggesting a potential mechanism for the development of microcephaly in ZIKV-infected individuals. The study paves the way for interesting future experiments exploring the potential to manipulate this pathways in therapeutic approaches and investigating its effects on ZIKV infection in vivo.