Zika virus (ZIKV) and other flaviviruses are dependent on the cellular pathways for every step of their replication cycle including replication of viral genome, assembly of new viral particles, maturation and release. The viruses alter host cell protein production and interaction network by hijacking and competing with host proteins. The outcome of infection is determined by complex host-virus interactions with a large number of altered transcriptional and translational rates, and functional kinetics of participating genes. Therefore, the analysis of these virus-host PPIs plays a critical role in understanding the interplay between Zika and its hosts, revealing the major pathways and mechanisms involved in Zika neuropathogenesis. Importantly, the identified host factors serve as potential candidates for drug targets.
High throughput yeast two-hybrid (Y2H) system is used in this study to identify the host factors that are targeted by ZIKV for its survival and replication. Y2H is a well-established method and has been widely used in many PPI studies, particularly to study virus-host interactions. The Matchmaker Gold Yeast Two-Hybrid System (Clontech) used in this study has been optimized and improved by the manufacturer for a remarkable reduction of false positivity 54. A number of studies have revealed novel protein-protein interaction using the Matchmaker Gold Yeast Two-Hybrid System 55–57.
In this study, the Brazil-ZKV2015 strain (Genbank KU497555) was chosen for the ZIKV-host PPI analysis using Y2H. Phylogenetic analysis of all the 2015 ZIKV isolates in American and Asian countries including the Brazilian strain used in this study are clustered closely within the Asian lineage. All the circulating strains exhibited the highest sequence identity to the French Polynesia isolate H/PF/2013 58,59. Notably, a panel of mutations have emerged during the spread from Asia to the Americas. The epidemic Brazilian strains have shown 14 to 18 nucleotide variations when compared to other human strains within the same lineage 59. The first one is the S139N substitution in the prM protein, which has been reported to possess an enhanced capacity to induce severe microcephalic phenotype. In a mouse model, contemporary American strains resulted in an increased neurovirulence and a more severe microcephalic phenotype than a historical Asian strain 60. Similarly, another study has shown that various strains of ZIKV exhibit distinct neuropathological effects. The Paraiba strain (Brazil, 2015; Paraiba/2015) exhibited greater virulence and was more likely to induce significant gross morphological changes consistent with microcephaly in the developing mouse brain 61. Existing ZIKV-host protein-protein interaction studies 13–15,19 have shown that different strains result in the identification of varying host factors. Hence, to decipher the ZIKV-host interactome and gain comprehensive understanding of ZIKV neural invasion, neurovirulence, and neuropathogenesis, it is imperative to use the ZIKV strain associated with the microcephaly phenotype.
The Brazil-ZKV2015 strain (Genbank KU497555) used in this study, which was associated with microcephaly cases in Brazil 2. The strain has been shown to infect human cortical neural progenitor cells (NPCs), inducing apoptotic cell death. In human brain organoids, infection with the strain resulted in reduction of proliferative zones and disruption of cortical layers. Newborns from mice infected with the Brazilian strain, displayed evidence of intra-uterine growth restriction. In addition, the brains of the mice infected with the strain exhibited cortical malformations with reduced cell number and cortical layer thickness, signs associated with microcephaly 23. These findings strongly support the neurotropic nature of the Zika Brazilian strain.
Yeast two-hybrid screening for Capsid, prM, Env, NS2B, NS3, NS4A, NS4B and NS5 proteins were conducted. The screening identified a total of 85 Zika virus-interacting human proteins that are involved in key biological processes including immune response, inflammatory pathways, mitotic process, cell cycle regulation, protein ubiquitination, neurogenesis, neural development and brain disorders. The virus-host interactome in this study discovered several previously unknown interactions as well as numerous new host factors, bringing us closer to completing the Zika virus-host protein-protein interactome. Among these 85 proteins, 6 were found to interact with more than one ZIKV protein. In particular COPS5 interacted with Env, prM, NS4B and NS5. Similarly, this pattern of promiscuity among host factors in their interactions with viral counterparts were also observed in previous studies 14,19.
Through integration analysis using BioVenn, it is found that 18 ZIKV interacting host factors from this study overlaps with the host factors identified in previous individual ZIKV-host PPI studies 13–15,19. There are several factors that could contribute to the low level of overlap observed between studies. The two main methods used among these studies are Y2H and affinity purification followed by mass spectrometry. In general, AP-MS tends to identify a higher number of interactions compared to Y2H screening. AP-MS has the advantage of being able to detect both direct and indirect interactions. Due to its sensitivity and ability to analyse complex mixtures, AP-MS is capable of identifying hundreds to thousands of potential interactors simultaneously. On the other hand, Y2H is based on the interaction between proteins in a yeast system, and while it can identify specific interactions, it may not capture the full range of interactions that occur in a cellular environment 62. In addition to AP-MS, Coyaud et al. used BioID (a proximity-dependent labeling approach) as an additional complementary method, which may also contribute to the differences between the studies. Differences in methods such as expression vectors, affinity tags, and the number of viral proteins (all viral proteins or only non-structural proteins) can contribute to differences observed across studies. Additionally, the differences in ZIKV strains used in experiments can also impact the outcomes and results of protein interaction studies. Other than that, the different prey cDNA libraries used in Y2H studies lead to identification of different sets of genes. CEP192 was one of proteins that was identified in 3 different ZIKV PPI studies.
GO analysis of the identified cellular targets of ZIKV revealed significant enrichment of genes involved in regulation of cell cycle. Among these genes, CUL1, CHORDC1 and COPS5 were targeted by the ZIKV Env protein. Notably, prior studies have found that ZIKV Env protein induces G2/M cell cycle 63. ZIKV infection of NPCs has been shown to cause centrosome disruption and alterations in division plane. This subsequently results in premature differentiation of neural progenitors, which leads to their depletion, impaired neurogenesis, and cortical thinning. 64. Centrosomal abnormalities lead to impairment of mitosis, which is a hallmark of autosomal primary recessive microcephaly (MCPH) 65. A study found that microcephaly and intellectual disability genes are significantly enriched for genes annotated with the centrosome 66. In this study, centrosome cycle, centrosome duplication and microtubule organizing center organization were among the significantly enriched GO terms for biological process. The identified ZIKV cellular targets involved are ROCK2, PPP1R12A, CKAP5, CEP192 and CHORDC1. In this context, CEP192 (identified as an ZIKV NS3 interaction partner) plays a central role in the initial steps of centriole duplications through the interaction and recruitment of CEP152 (MCPH9) and PLK4, respectively, which is necessary for the proper recruitment of SAS6 (MCPH14), STIL (MCPH7), and CENPJ (MCPH6).
Importantly, Y2H screening in this study identified interaction of ZIKV proteins with microcephaly genes such as FBLN5, PDHA1 and CTCF. In addition to microcephaly associated genes, Y2H screening also identified ZIKV-interacting host factors that are assoociated with neurogenesis, neural development and brain disorders including OPA1, ANK3, OMGp, NEDD9 and LanCL1. Consistent with its role in virus replication and disruption of immune signaling, NS5 was found to interact with several genes that are associated with immune response and antiviral signaling such as DCAF6, PPM1A and PI3K. Another gene, tripartite motif containing 36 (TRIM36) that plays an essential role in viral replication by regulating the innate immune response was found to be targeted by ZIKV NS2B. TRIM36 has been shown to act as a negative regulators of type I IFN responses 67. Overall, Y2H screening of ZIKV proteins revealed interaction with host factors associated with cell cycle, microcephaly, neural development, immune response and antiviral signaling.
Among the interactions, ZIKV NS5-ROCK2 was selected for further evaluation to determine its role in ZIKV infection and pathogenesis. ROCK2 protein, which is primarily expressed in the brain was found to interact with ZIKV NS5. As ROCK signalling pathway plays an essential role in the CNS and is a key regulator of neuronal growth, differentiation, development, migration, metabolism, apoptosis and cell cycle 68–70, it is worth for further exploration. Gene ontology analysis revealed enrichment of ROCK2 in several biological process category including regulation of cellular response to stress, centrosome cycle, regulation of cell cycle, microtubule organizing center organization, embryonic morphogenesis and regulation of establishment of endothelial barrier. In this study, ROCK2 was found to co-localize with ZIKV NS5 in the nucleus of SK-N-SH cells. Consistent with the existing findings 71, ZIKV NS5 formed discrete spherical shell-like structures in the nucleus. In addition to ZIKV, NS5 of several other flaviviruses such as DENV, WNV, YFV and JEV also has been reported to accumulate in the nucleus, despite its role in virus RNA replication that occurs in the cytoplasm. The subcellular localization of ZIKV NS5 in the nucleus is mediated by the nuclear localization signal (NLS) in the αβNLS region. Ng et al. demonstrated that the subcellular localization of ZIKV NS5 in the nucleus is associated with its role in the modulation of the host immune response 71. Coincidentally, it has been found that ROCK2 is involved in neuroinflammation and cell death by regulating the secretion of inflammatory cytokines such as tumor necrosis factor-alpha (TNFα/TNF), interleukin-1 beta (IL-1b), interleukin-2 (IL-2), and CXC chemokines 72. Therefore, the interaction between ZIKV NS5 and ROCK2 may be involved in the induction of ZIKV-induced inflammatory effects.
Additionally, ROCK2 mediated gene silencing significantly inhibited ZIKV RNA copy number in HEK293FT cells. The role of ROCK2 in virus infection has been explored in several studies. Qisheng Li et al. demonstrated that siRNA-mediated silencing of ROCK2 prior to HCV infection blocked HCV entry. In addition, both pre- and post-treatment with a ROCK inhibitor, Fasudil significantly reduced HCV RNA titers in infected cells 73. A recent study, demonstrated that ROCK1/2 act as the downstream of LPA1 signalling to supress IFN-I/III production and antiviral responses. The study showed that inhibition of LPA1, promoted IFN-induced clearance of SARS-CoV-2 and ZIKV. Furthermore, it was also shown that treatment with Y27632, a ROCK inhibitor improved survival of vesicular stomatitis virus (VSV)-infected mice though increased enhanced production of IFN-I and IFN-III 41. In contrast to these findings, several studies have demonstrated that ROCK inhibition enhances virus replication. Das et al. found that inhibition of ROCK using Y-27632 enhanced virus host fusion and viral release through decrease in phosphorylation of RLC 74. Eliyahu et al. showed that ROCK activity restricts HCMV replication and its inhibition leads to increase in viral titers. The study proposed that the ROCK antiviral activity against HCMV may be associated to activation of actomyosin network and inhibition of capsid egress out of the nucleus 75. Therefore, the role of ROCK2 protein in ZIKV infection as well as other RNA viruses require infection require further