Comparative analysis of different cell systems for Zika virus (ZIKV) propagation and evaluation of anti-ZIKV compounds in vitro

Highlights

• ZIKV replicates in human, mosquito and monkey cell lines.

• Huh7 cells yield maximal production of infectious ZIKV.

• Human A549 are effective as the reference nonhuman VERO cells in plaque assay test.

• ZIKV RNA and PFU titers best correlated in Huh7-producing, A549-readout cells.

• Comparable drug EC50 obtained by qRT-PCR and plaque assay in the Huh7/A549 system.

Abstract

A strong correlation between Zika virus (ZIKV) infection and severe neurological disease in newborns and occasionally adults has emerged in the Brazilian outbreak. Efficient human cell-based assays are required to test candidate inhibitors of ZIKV replication. The aim of this work was to investigate ZIKV propagation and quantification in different cell lines. The human (U87, A549, Huh7), mosquito (C6/36) and monkey (VERO E6) cell lines tested were all permissive to ZIKV infection. When assessed by plaque forming units (PFU) in three different target cell lines, the maximal production of ZIKV was achieved in Huh7 at day 3 post-infection (6.38 ± 0.44 log₁₀ PFU/ml). The C6/36 cell line showed a low and slow production of virus when compared with other cell lines. A549 readout cells generated a larger number of plaques compared to Huh7 but not to VERO E6 cells. ZIKV PFU and RNA titers showed the highest correlation when Huh7 and A549 were used as the producer and readout cells, respectively. Also, U87 cells produced ZIKV RNA titers which were highly correlated with PFU independently from the readout cell line. Using the best virus-cell system, sofosbuvir and ribavirin EC₅₀ were 1.2 μM and 1.1 μM when measured through plaque assay, and 4.2 μM and 5.2 μM when measured by quantitative real time PCR (qRT-PCR), respectively. In summary, ZIKV can efficiently infect different human cell lines and rapidly reach peak viral titers. Overall, A549 cells appear to be as efficient as the VERO E6 gold standard for plaque assay allowing the use of human, rather than simian, cells for evaluating candidate anti-ZIKV compounds by the reference assay. The possibility to replace the labor-intensive plaque assay with the more rapid and easy-to-perform qRT-PCR is appealing and warrants further investigation.

Introduction

Zika virus (ZIKV) is a member of the Flaviviridae family, which includes several agents of clinical significance, such as Dengue, Hepatitis C (HCV), West Nile and Japanese encephalitis viruses. ZIKV is assumed to be propagated in nature through an enzootic cycle involving mammalian hosts, such as primates, and insect vectors, primarily mosquitoes of the Aedes genus (Akoua-Koffi et al., 2001, Fagbami, 1979, Marchette et al., 1969, McCrae and Kirya, 1982). ZIKV can be transmitted to humans by the bite of female Aedes mosquitoes, through a sexual route, from a pregnant mother to her unborn fetus, by blood transfusion, or other bodily fluids (Barzon et al., 2016, Miner et al., 2016, Petersen et al., 2016, Sharma and Lal, 2017). Three geographically distinct lineages of ZIKV have been reported, including the East African, West African, and Asian (Baronti et al., 2014). This last one seems to have evolved from the African lineage and has been shown to have increased pathogenicity and epidemic potential (Lazear and Diamond, 2016, Wang et al., 2016, Weaver et al., 2016). Up to 80% of human ZIKV infections appear to be asymptomatic, with a small subset of cases presenting with mild flu-like symptoms associated with rash, arthralgia, and conjunctivitis (Lessler et al., 2016, Marano et al., 2016, Shah and Kumar, 2016). However, the recent spread across the Brazilian territory and the neighboring countries in 2015 has renewed the interest in ZIKV due to the association between infection acquired during pregnancy and congenital malformations, including microcephaly (Brasil et al., 2016a, de Araújo et al., 2016, Paploski et al., 2016, Rasmussen et al., 2016), spontaneous abortion and intrauterine growth restriction (Carteaux et al., 2016, Miner et al., 2016). In addition, a broad range of neurological disorders have been reported in adults, including Guillain-Barré syndrome (GBS) (Brasil et al., 2016b, Cao-Lormeau et al., 2016, Oehler et al., 2014). Phylogenetic analysis revealed that the ZIKV strains currently circulating in South America share >99% identity with ZIKV isolates from French Polynesia outbreaks (Asian Lineage) (Imperato, 2016) and seem to be associated with neurologic damage, including few cases of bilateral macular and perimacular lesions (Mlakar et al., 2016, Ventura et al., 2016), globally proving that ZIKV genetic evolution has resulted in expanded virus tropism and increased pathogenicity. The emergence of ZIKV associated pathology in Brazil led the WHO to declare the ZIKV outbreak to be a public health emergency of international concern (WHO, 2016).

Despite the clinically relevant consequences of ZIKV in humans, currently there are no countermeasures to halt transmission, control infection or mitigate disease (Bollati et al., 2010, Yun and Lee, 2017). Some broad-spectrum antivirals, such as interferons (IFNs), ribavirin and favipiravir, cannot be used against ZIKV because they can be harmful to pregnant women (Barrows et al., 2016, Chutaputti, 2000, Xu et al., 2016). Recently identified crosstalk pathways between cancer and viral replication have drawn the attention on FDA-approved cancer drugs which show anti-ZIKV activity (Cheng et al., 2016). The repurposing of these “old” drugs to identify compounds with novel activity against ZIKV infection can be a useful tool to overcome the high cost of antiviral drug-discovery pipeline. In addition, the high degree of conservation of viral RNA polymerase among the Flaviviridae family has recently prompted the evaluation of compounds approved for treatment of HCV infection. Indeed, the nucleotide analog sofosbuvir, approved in 2013, has shown anti-ZIKV activity in vitro, although with effective concentration significantly higher than those reported with HCV (Bullard-Feibelman et al., 2017, Murakami et al., 2010, Sacramento et al., 2017). To evaluate candidate anti-ZIKV compounds, robust methods for ZIKV propagation and quantification are required. Monkey (VERO E6) or insect (C6/36) cell lines have been most frequently used to propagate ZIKV and test candidate antivirals (Adcock et al., 2017, Goebel et al., 2016, Pascoalino et al., 2016). Interestingly, cell-dependent inhibition of ZIKV replication has been shown for sofosbuvir (Eyer et al., 2016, Mumtaz et al., 2017, Sacramento et al., 2017). This advises for investigating multiple ZIKV propagation protocols to represent the variability of human cells to process candidate anti-ZIKV compounds and impact their activity.

In this work, we have evaluated the kinetics of viral replication in several cell lines, as measured by both ZIKV RNA and replication competent virus through plaque assay, to define the best standards for ZIKV antiviral screening. As a proof of concept, the inhibitory effect of sofosbuvir and ribavirin on ZIKV has been assessed through a yield reduction assay.