Human Fetal Astrocytes Infected with Zika Virus Exhibit Delayed Apoptosis and Resistance to Interferon: Implications for Persistence

Abstract

Zika virus (ZIKV) infection and persistence during pregnancy can lead to microcephaly and other fetal neurological disorders collectively known as Congenital Zika Syndrome. The immunological and virological events that contribute to the establishment of persistent ZIKV infection in humans are unclear though. Here we show that human fetal astrocytes (HFAs), the most abundant cell type in the central nervous system, become persistently infected with ZIKV resulting in continuous viral shedding for at least one month; a process that is facilitated by TIM/TAM receptors. HFAs are relatively resistant to ZIKV-induced apoptosis, a factor that may be important for chronic infection of these cells. Once infection was established, interferon treatment did not reduce virus replication. Moreover, the fact that the innate immune system was highly activated in persistently infected HFAs indicates that the virus can thrive in the presence of a sustained antiviral response. RNAseq analyses of persistently infected cells revealed that ZIKV alters host gene expression in a manner that could affect developmental processes. Conversely, data from sequencing of ZIKV genomes in persistently infected HFAs suggest that adaptive mutations were not required for establishing chronic infection. Based on these results, we postulate that HFAs are reservoirs for ZIKV in the fetal brain and that moderate apoptosis combined with inefficient antiviral response from these cells may contribute to the establishment of chronic brain infection associated with the ZIKV neurodevelopmental abnormalities.

Keywords: Zika virus; apoptosis; astrocytes; interferon; persistence.

Figure 1

HFAs are highly permissive for ZIKV infection. (A) HFAs and A549 cells were infected with PLCal ZIKV (MOI = 3) for 1–5 days and virus replication was quantitated by measuring genomic RNA in cells using qRT-PCR. (B–E) Cells were infected with ZIKV that had been pre-incubated with duramycin for 1 h or first treated with R428 for 1 h followed by infection with ZIKV (MOI = 0.5). One day later, cells were processed for indirect immunofluorescence using antibodies against flavivirus envelope protein or total RNA was harvested for qRT-PCR. (B,C) Data from automated quantitation of ZIKV-positive cells. (D,E) Quantification of ZIKV replication by qRT-PCR after drug treatment. (F) HFAs were pre-incubated with anti-AXL blocking antibody for 1 h followed by ZIKV infection (MOI = 10). After 24 h, cells were collected and viral RNA was quantified by qRT-PCR. The means of three independent experiments are shown. Error bars represent standard error of the mean. * p 0.05, ** p 0.01, *** p 0.001 (Student’s t-test).

Figure 1

HFAs are highly permissive for ZIKV infection. (A) HFAs and A549 cells were infected with PLCal ZIKV (MOI = 3) for 1–5 days and virus replication was quantitated by measuring genomic RNA in cells using qRT-PCR. (B–E) Cells were infected with ZIKV that had been pre-incubated with duramycin for 1 h or first treated with R428 for 1 h followed by infection with ZIKV (MOI = 0.5). One day later, cells were processed for indirect immunofluorescence using antibodies against flavivirus envelope protein or total RNA was harvested for qRT-PCR. (B,C) Data from automated quantitation of ZIKV-positive cells. (D,E) Quantification of ZIKV replication by qRT-PCR after drug treatment. (F) HFAs were pre-incubated with anti-AXL blocking antibody for 1 h followed by ZIKV infection (MOI = 10). After 24 h, cells were collected and viral RNA was quantified by qRT-PCR. The means of three independent experiments are shown. Error bars represent standard error of the mean. * p 0.05, ** p 0.01, *** p 0.001 (Student’s t-test).

Figure 2

ZIKV can persistently infect primary HFAs. (A–C) HFAs were infected with ZIKV (MOI = 3) and media were collected at the indicated intervals for up to 28 days. Viral titers were determined by plaque assay. The percentages of infected cells at indicated time periods were determined by indirect immunofluorescence using antibodies against flavivirus envelope protein. (A) ZIKV titers in persistently infected HFAs. The average titers obtained from two experiments using three independent donors are shown. (B) Measurement of viral replication by qRT-PCR. The average values obtained from two experiments using two independent donors are shown. (C) Percentage of infected HFAs over time is shown. (D) Importance of viral spread in maintaining chronic infection. HFAs infected with ZIKV for 14 days were treated with anti-AXL (4 µg/mL) for 4 days and viral replication was determined by qRT-PCR. The average values obtained from three experiments using two independent donors are shown. Error bars represent standard error of the mean. * p 0.05 (Student’s t-test).

Figure 3

(A) Human recombinant IFN treatment and innate immune gene activation in persistently infected HFAs. Relative levels of IFIT1 and OAS transcripts (compared to actin mRNA) in chronically infected HFAs were determined by qRT-PCR. The average values obtained from two experiments using two independent donors are shown. (B–E) HFAs were treated before (12 and 0 h) or (F–I) after (24 and 48 h) ZIKV infection (MOI = 0.3) with the indicated amounts of IFN-α and γ. Two days later, supernatants and cell lysates were collected for viral determination by plaque assays (C,E,G,I) and viral genome quantitation by qRT-PCR (B,D,F,H) respectively. The average values obtained from three experiments using three independent donors are shown. Error bars represent standard error of the mean. * p 0.05, ** p 0.01, *** p 0.001 (Student’s t-test).

Figure 3

(A) Human recombinant IFN treatment and innate immune gene activation in persistently infected HFAs. Relative levels of IFIT1 and OAS transcripts (compared to actin mRNA) in chronically infected HFAs were determined by qRT-PCR. The average values obtained from two experiments using two independent donors are shown. (B–E) HFAs were treated before (12 and 0 h) or (F–I) after (24 and 48 h) ZIKV infection (MOI = 0.3) with the indicated amounts of IFN-α and γ. Two days later, supernatants and cell lysates were collected for viral determination by plaque assays (C,E,G,I) and viral genome quantitation by qRT-PCR (B,D,F,H) respectively. The average values obtained from three experiments using three independent donors are shown. Error bars represent standard error of the mean. * p 0.05, ** p 0.01, *** p 0.001 (Student’s t-test).

Figure 4

Apoptosis is delayed in ZIKV-infected HFAs. HFAs were infected with ZIKV (MOI = 3) and apoptosis determined by flow cytometry at indicated time points. (A) The percentages of HFAs and A549 cells expressing ZIKV envelope protein were determined after 1–5 days post-infection. (B) The percentages of viable HFAs and A459 cells remaining after 1–5 days post-infection with ZIKV were determined using an automated cell counter. (C) The percentages of HFAs and A549 cells with active caspase-3 are shown. (D) The percentages of cells with active caspase-3 among ZIKV-infected HFAs and A549 cells are shown. Values are expressed as the mean of three independent experiments. Error bars represent standard error of the mean.

Figure 4

Apoptosis is delayed in ZIKV-infected HFAs. HFAs were infected with ZIKV (MOI = 3) and apoptosis determined by flow cytometry at indicated time points. (A) The percentages of HFAs and A549 cells expressing ZIKV envelope protein were determined after 1–5 days post-infection. (B) The percentages of viable HFAs and A459 cells remaining after 1–5 days post-infection with ZIKV were determined using an automated cell counter. (C) The percentages of HFAs and A549 cells with active caspase-3 are shown. (D) The percentages of cells with active caspase-3 among ZIKV-infected HFAs and A549 cells are shown. Values are expressed as the mean of three independent experiments. Error bars represent standard error of the mean.

Figure 5

(A) Analysis of transcripts differentially expressed in HFAs persistently infected (28 days) with ZIKV. MA plots representing transcripts that were deregulated in cells persistently infected with ZIKV are shown. Red dots represent transcripts that were significantly upregulated with a fold change 2; green dots represent transcripts that were significantly downregulated with a fold change –2; blue dots represent genes that were significantly deregulated with a fold change 2; black dots represent genes not differentially expressed. (B) Hierarchical clustering identified two major clusters of differentially expressed genes in persistently infected HFAs (28 days), with relatively consistent expression across samples (i.e., consistently up or downregulated in ZIKV-infected or mock samples). Z-scores of 400 differentially expressed genes exhibiting the largest variance between mock and ZIKV-infected samples are plotted. (C) The fold-change and the corrected P value of the top up- and downregulated transcripts are presented for persistently infected HFAs (28 days). (D) Mutations are not accumulated in the ZIKV genome after persistent infection. Single nucleotide polymorphisms (SNPs) and length polymorphisms (LPs) were determined in the ZIKV genome in acute (2 days) and persistent (28 days) infection of HFAs using the software Vphaser-2. Kolmogorov–Smirnov test was applied on the frequency of each type of mutations consolidated in segments of 500 nt.