Zika virus causes placental pyroptosis and associated adverse fetal outcomes by activating GSDME
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
The manuscript addresses an important question with broad relevance to the fields of virology, reproductive biology and immunology, and cell death. The primary conclusion of viral-induced cell death is well supported. Some mechanistic details of the pathway remain unclear.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Microbiology and Infectious Disease (eLife)
Abstract
Zika virus (ZIKV) can be transmitted from mother to fetus during pregnancy, causing adverse fetal outcomes. Several studies have indicated that ZIKV can damage the fetal brain directly; however, whether the ZIKV-induced maternal placental injury contributes to adverse fetal outcomes is sparsely defined. Here, we demonstrated that ZIKV causes the pyroptosis of placental cells by activating the executor gasdermin E (GSDME) in vitro and in vivo. Mechanistically, TNF-α release is induced upon the recognition of viral genomic RNA by RIG-I, followed by activation of caspase-8 and caspase-3 to ultimately escalate the GSDME cleavage. Further analyses revealed that the ablation of GSDME or treatment with TNF-α receptor antagonist in ZIKV-infected pregnant mice attenuates placental pyroptosis, which consequently confers protection against adverse fetal outcomes. In conclusion, our study unveils a novel mechanism of ZIKV-induced adverse fetal outcomes via causing placental cell pyroptosis, which provides new clues for developing therapies for ZIKV-associated diseases.
Article activity feed
-
-
Author Response
Reviewer #3 (Public Review):
In "Zika virus causes placental pyroptosis and associated adverse fetal outcomes by activating GSDME," Zhao et al. investigated the mechanism of fetal growth restriction caused by maternal Zika virus infection.
Strengths:
The in vitro studies (knockouts) are clear in showing a role for GSDME in cell death. They show that GSDME may be functioning similarly in several cell types in addition to placental cells. They also show that RIG-I recognition of the viral 5' UTR is critical for the cellular pyroptotic response. Using a pregnant mouse model, they show that GSDME knockout prevents disease in fetuses.
Weaknesses:
Given that the authors describe pyroptosis in other cell types, it seems possible that the effects of GSDME knockout on the fetus could be indirect and due to decreased …
Author Response
Reviewer #3 (Public Review):
In "Zika virus causes placental pyroptosis and associated adverse fetal outcomes by activating GSDME," Zhao et al. investigated the mechanism of fetal growth restriction caused by maternal Zika virus infection.
Strengths:
The in vitro studies (knockouts) are clear in showing a role for GSDME in cell death. They show that GSDME may be functioning similarly in several cell types in addition to placental cells. They also show that RIG-I recognition of the viral 5' UTR is critical for the cellular pyroptotic response. Using a pregnant mouse model, they show that GSDME knockout prevents disease in fetuses.
Weaknesses:
Given that the authors describe pyroptosis in other cell types, it seems possible that the effects of GSDME knockout on the fetus could be indirect and due to decreased pyroptosis in elsewhere in the dams. How did GSDME knockout alter the clinical signs of disease (weight loss, histopathology) in the dams?
The Gsdme-/- mice used in our study were kindly provided by Dr. F Shao, and it was demonstrated that deletion of GSDME has no effect on the development and immune system of mice (Wang et al., 2017). In our in vivo model, no clinical signs of disease and weight loss were observed in both ZIKV-infected WT and GSDME KO dams. This is consistent with the results of previous studies which revealed that ZIKV-infection didn’t lead to any clinical symptoms in immunocompetent pregnant mice except placental damage and adverse fetal outcomes (Szaba et al., 2018; Barbeito-Andres et al., 2020). Our data showing extremely low viral loads in spleens, serum and brains of infected dams and no difference of viral titers in tissues of WT and Gsdme-/- dams (Figure 5G) also support that GSDME knockout didn’t alter the clinical signs of disease. The figure showing the weight change of dams and relative discussion (Figure S7A) were included in the revised manuscript. In addition, the vast majority of affected embryos underwent resorption, leaving only the placental residues and embryonic debris, so it’s hard to evaluate the function of GSDME by histopathological methods in embryos. In the remaining embryo samples, no obvious clinical signs were found in both WT embryos and Gsdme-/- embryos. To address whether the effects of GSDME knockout on the fetus are due to decreased pyroptosis in placenta or dams, we tried to cross Gsdme+/- and Gsdme-/- mice and compare pathology in +/- to -/- littermates. However, the pregnancy and litter rates were too low, even though we spent a lot of time and tried many times, we still could not get enough data to draw conclusions, so we added a discussion of related issues and the limitations of our in vivo experiments in revised manuscript (Line 463 to 482).
Figure 5D/E/F and Figure 6C/D- how are the authors distinguishing between apoptosis and pyroptosis as the cause of cell death in the placental tissue?
Due to the lack of effective commercial antibodies that can detect the activation of GSDME by immunohistochemistry or by Western blot within mouse tissues, we can only use PI staining to exclude early apoptosis, because PI cannot pass through the membrane structure of early apoptotic cells. But till now, there is still a lack of effective methods to distinguish between GSDME-mediated pyroptosis and the late stage of apoptosis completely, and some studies also called GSDME-mediated pyroptosis as secondary necrosis (Rogers et al., 2017). In addition, our results showed that ZIKV could induce GSDME-mediated pyroptosis in primary mouse trophoblast cells (Figure S6), which supports the conclusion of our in vivo experiments. However, we realized that this is a limitation of our in vivo model and additional research efforts are required to address this issue in the future. A discussion relative to this issue has been included in the revised manuscript (Line 463 to 482).
-
Evaluation Summary:
The manuscript addresses an important question with broad relevance to the fields of virology, reproductive biology and immunology, and cell death. The primary conclusion of viral-induced cell death is well supported. Some mechanistic details of the pathway remain unclear.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)
-
Reviewer #1 (Public Review):
In the study "Zika virus causes placental pyroptosis and associated adverse fetal outcomes by activating GSDME", the authors report that Zika infection induces GSDME mediated cell death. The significance of this finding is high, as the impact of Zika on cell death in the placenta is probably key to pathogenesis. The data implicating GSDME in vitro and in vivo is impressive. However, the pathways proposed that lead to GSDME cleavage and pyroptosis are not supported by the data. Much additional work is needed here. If this pathway is mapped as suggested, the impact on the field would be significant.
-
Reviewer #2 (Public Review):
Zhao et al. addressed the possibility that Zika virus (ZIKV)-induced pyroptosis (a lytic form of cell death) contributes to ZIKV pathogenesis and adverse outcomes during pregnancy. Using the human choriocarcinoma cell line JEG-3 authors observe that ZIKV infection induced a Gasdermin E (GSDME)-dependent pyroptosis, which they demonstrated by comparing JEG-3 WT and monoclonal GSDME-/- cells. ZIKV was also found to induce a similar cell death in a panel of immortalized cell lines derived from other tissues. A particularly surprising observation by the authors is that the ZIKV untranslated regions of the viral genome was sufficient to induce pyroptosis via the RNA sensor RIG-I. Based on known activators of GSDME, the authors use a panel of caspase or other cell death pathway inhibitors to identify a CASP8/9>CASP…
Reviewer #2 (Public Review):
Zhao et al. addressed the possibility that Zika virus (ZIKV)-induced pyroptosis (a lytic form of cell death) contributes to ZIKV pathogenesis and adverse outcomes during pregnancy. Using the human choriocarcinoma cell line JEG-3 authors observe that ZIKV infection induced a Gasdermin E (GSDME)-dependent pyroptosis, which they demonstrated by comparing JEG-3 WT and monoclonal GSDME-/- cells. ZIKV was also found to induce a similar cell death in a panel of immortalized cell lines derived from other tissues. A particularly surprising observation by the authors is that the ZIKV untranslated regions of the viral genome was sufficient to induce pyroptosis via the RNA sensor RIG-I. Based on known activators of GSDME, the authors use a panel of caspase or other cell death pathway inhibitors to identify a CASP8/9>CASP3 pathway upstream of GSDME activation. Finally, using an immunocompetent mouse model of ZIKV in utero infection, the authors observed that Gsdme-/- mice were less susceptible than WT mice to congenital ZIKV syndromes (CZS). Taken together, the authors provide considerable evidence that largely supports their interpretations for how ZIKV causes pyroptotic cell death. The primary observation - that ZIKV induces GSDME pyroptosis - is highly impactful. In some areas, surprising findings (e.g., RIG-I-dependent pyroptosis) would be more interpretable by including additional controls or testing additional predictions of the model. In other cases, the reader would benefit by the authors further contextualizing their model relative to preexisting work, and acknowledging alternative hypothesis that may (or may not) occur in other cell types or non-tumor derived cells and potential differences between mouse and human pathways. I attempt to highlight specific strengths and concerns below.
In general, the primary conclusion that ZIKV induces a GSDME-dependent pyroptosis is well supported. In my opinion, aspects of the mechanistic details of the upstream pathway remain uncertain. This stems from a few different experimental choices.
First, the mechanistic workup is completed entirely in immortalized cells. Obtaining primary trophoblasts is recognized as a potentially significant challenge. However, the dependency on immortalized cell lines precludes some of the more generalized conclusions that the authors attempt to draw from their data. The RIG-I>CASP8/9>CASP3>GSDME pathway is also entirely described in the JEG-3 cells using a single ZIKV strain. With the exception of the GSDME-/- JEG-3 cells, all other experiments rely on knockdowns or chemical inhibitors. While these experiments are informative, clean knockouts in at least one other cell line for key aspects of the pathway in question would provide better support for the authors conclusions. This is particularly important for the interesting but surprising claim that the ZIKV genome UTRs are sufficient to induce GSDME pyroptosis in a RIG-I dependent manner. To my knowledge, no such pathway has been described. Hence, a few additional experiments to shore up that observation are merited. For example, one prediction is that other RIG-I agonists should similarly induce pyroptosis. Moreover, RIG-I, MDA5, and MAVS knockouts would allow for clear, genetic determination of whether or not the pathway in question operates similarly to known RIG-I signaling. Another interesting but not fully clear aspect of the pathway is how various caspases contribute to GSDME activation. In this case, authors can partly rely on prior work demonstrating that CASP3 cleaves and activates GSDME. However, the relative contributions of other caspases (e.g., CASP8 and CASP9) are more difficult to discern because the authors exclusively rely on chemical inhibitors, which although largely specific also have known off-target effects. In my opinion, this is a more minor aspect of the manuscript, but should be considered by readers with appropriate discernment.
The reduction in CZS in Gsdme-/- compared to WT mice is a particularly striking observation. This does soften some of my concerns about the generality of the JEG-3 results. However, the JEG-3-defined pathway and the in vivo results are largely correlative. For example, it is not clear from the data presented if ZIKV infection activates mouse GSDME using the same RIG-I>CASP8/9>CASP3>GSDME pathway. The mating scheme utilized by the authors also prevents a determination of the relative maternal versus fetal contributions of GSDME to the phenotype. In addition, littermate and co-housing is not described. Although the placenta is considered a sterile tissue, non-genetic contributions to the phenotype have not been controlled for. An interesting feature of the in vivo infection model is the use of immunocompetent mice. Although this falls outside of the expertise of this reviewer, my understanding was that IFN-competent mice are largely resistant to ZIKV infection. It would be helpful if the authors contextualized their model relative to those previously published.
Another area where additional context would be useful is in considering other aspects of the JEG-3 cells compared to cells used in prior ZIKV studies. Tumor-derived cell lines (like JEG-3) often lose cell death pathways. As the authors recognize, other groups have reported ZIKV activation of the NLRP3 inflammasome. Do JEG-3 cells (and other cells used in the study) accurately mimic the status of the NLRP3 pathway in primary trophoblasts? Is CASP1 expressed in these cells (almost certainly not in HeLa and 293T cells)? Thus, it remains possible that the NLRP3 pathway is active in human primary trophoblasts and contributes to ZIKV pathogenesis in vivo. And although deciphering the relative contributions of these pathways falls outside the context of this work, it again limits the generalization of the data. In addition, the authors should consider multiple reports that CASP3 can cleave and inhibit GSDMD, which seems particularly relevant to their observation.
-
Reviewer #3 (Public Review):
In "Zika virus causes placental pyroptosis and associated adverse fetal outcomes by activating GSDME," Zhao et al. investigated the mechanism of fetal growth restriction caused by maternal Zika virus infection.
Strengths:
The in vitro studies (knockouts) are clear in showing a role for GSDME in cell death. They show that GSDME may be functioning similarly in several cell types in addition to placental cells. They also show that RIG-I recognition of the viral 5' UTR is critical for the cellular pyroptotic response. Using a pregnant mouse model, they show that GSDME knockout prevents disease in fetuses.Weaknesses:
Given that the authors describe pyroptosis in other cell types, it seems possible that the effects of GSDME knockout on the fetus could be indirect and due to decreased pyroptosis in elsewhere in …Reviewer #3 (Public Review):
In "Zika virus causes placental pyroptosis and associated adverse fetal outcomes by activating GSDME," Zhao et al. investigated the mechanism of fetal growth restriction caused by maternal Zika virus infection.
Strengths:
The in vitro studies (knockouts) are clear in showing a role for GSDME in cell death. They show that GSDME may be functioning similarly in several cell types in addition to placental cells. They also show that RIG-I recognition of the viral 5' UTR is critical for the cellular pyroptotic response. Using a pregnant mouse model, they show that GSDME knockout prevents disease in fetuses.Weaknesses:
Given that the authors describe pyroptosis in other cell types, it seems possible that the effects of GSDME knockout on the fetus could be indirect and due to decreased pyroptosis in elsewhere in the dams. How did GSDME knockout alter the clinical signs of disease (weight loss, histopathology) in the dams?Figure 5D/E/F and Figure 6C/D- how are the authors distinguishing between apoptosis and pyroptosis as the cause of cell death in the placental tissue?
-
