Envelope Protein Glycosylation Mediates Zika Virus Pathogenesis

Abstract

Zika virus (ZIKV) is an emerging mosquito-borne flavivirus. Recent ZIKV outbreaks have produced serious human disease, including neurodevelopmental malformations (congenital Zika syndrome) and Guillain-Barré syndrome. These outcomes were not associated with ZIKV infection prior to 2013, raising the possibility that viral genetic changes could contribute to new clinical manifestations. All contemporary ZIKV isolates encode an N-linked glycosylation site in the envelope (E) protein (N154), but this glycosylation site is absent in many historical ZIKV isolates. Here, we investigated the role of E protein glycosylation in ZIKV pathogenesis using two contemporary Asian-lineage strains (H/PF/2013 and PRVABC59) and the historical African-lineage strain (MR766). We found that glycosylated viruses were highly pathogenic in Ifnar1^-/- mice. In contrast, nonglycosylated viruses were attenuated, producing lower viral loads in the serum and brain when inoculated subcutaneously but remaining neurovirulent when inoculated intracranially. These results suggest that E glycosylation is advantageous in the periphery but not within the brain. Accordingly, we found that glycosylation facilitated infection of cells expressing the lectins dendritic cell-specific intercellular adhesion molecule-3-grabbing nonintegrin (DC-SIGN) or DC-SIGN-related (DC-SIGNR), suggesting that inefficient infection of lectin-expressing leukocytes could contribute to the attenuation of nonglycosylated ZIKV in mice.IMPORTANCE It is unclear why the ability of Zika virus (ZIKV) to cause serious disease, including Guillain-Barré syndrome and birth defects, was not recognized until recent outbreaks. One contributing factor could be genetic differences between contemporary ZIKV strains and historical ZIKV strains. All isolates from recent outbreaks encode a viral envelope protein that is glycosylated, whereas many historical ZIKV strains lack this glycosylation. We generated nonglycosylated ZIKV mutants from contemporary and historical strains and evaluated their virulence in mice. We found that nonglycosylated viruses were attenuated and produced lower viral loads in serum and brains. Our studies suggest that envelope protein glycosylation contributes to ZIKV pathogenesis, possibly by facilitating attachment to and infection of lectin-expressing leukocytes.

Keywords: CD209; CD209L; DC-SIGN; DC-SIGNR; Ifnar1−/− mouse; L-SIGN; Zika virus; flavivirus; glycosylation.

FIG 1

E glycosylation is not required for ZIKV replication. (A) ZIKV envelope protein, depicting the nucleotide and amino acid residues of the glycosylation site, and N154Q mutation. (B) Sequence chromatograms of E protein glycosylation site of wild-type (WT) and N154Q viruses. (C) Vero cells were infected at an MOI of 0.01 with ZIKV H/PF/2013 isolate, WT clone, or N154Q mutant. Viruses in culture supernatants were titrated by focus-forming assay. Data shown are the mean values ± standard errors of the means (SEM) of 9 samples from 3 independent experiments. (D and E) E proteins were immunoprecipitated with MAb 1M7 from lysates of Vero cells infected with ZIKV H/PF/2013 isolate, WT clone, N154Q mutant, or DENV. (D) Lysates were treated with PNGase F, separated by nonreducing SDS-PAGE, and probed with MAb 4G2. (E) Lysates were separated by nonreducing SDS-PAGE and probed with biotinylated concanavalin A to detect glycans. FT, flow-through; IP, immunoprecipitate.

FIG 2

ZIKV E N154Q is attenuated upon subcutaneous but not intracranial inoculation. Five- to six-week-old Ifnar1^−/− or wild-type (WT) mice were inoculated with 1 × 10³ FFU of ZIKV strain H/PF/2013 WT clone or N154Q mutant by a subcutaneous (A and B) or intracranial (C to E) route. Mice were weighed daily, and weights are expressed as percentages of body weight prior to infection. Results shown are the mean values ± SEM of 6 to 8 Ifnar1^−/− mice or 3 WT mice per virus from two or three independent experiments. Lethality was monitored for 14 days.

FIG 3

N154 glycosylation mediates ZIKV infection in mice. Five- to six-week-old Ifnar1^−/− mice were inoculated with 1 × 10³ FFU of ZIKV strain H/PF/2013 WT clone or N154Q mutant by a subcutaneous route in the footpad. (A) Blood was collected at 2, 4, and 6 days after infection, and ZIKV RNA in serum was measured by qRT-PCR. (B to E) Mice were euthanized 6 days after infection and perfused, and tissues were harvested. ZIKV RNA in tissue was measured by qRT-PCR. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant (unpaired 2-tailed t test).

FIG 4

Generating an infectious clone of ZIKV PRVABC59 and nonglycosylated mutants. (A) An infectious clone of ZIKV strain PRVABC59 was generated using a system that divides the viral genome into 4 fragments flanked by the indicated restriction endonuclease sites. Numbers above fragments indicate nucleotide position in the viral genome. T7 promoter and a hepatitis delta virus (HDV) ribozyme sequences flank the genome. (B) Sequence chromatograms of E protein glycosylation site of WT, N154Q, or T156I clone. (C) Vero cells were infected at an MOI of 0.01 with ZIKV PRVABC59 isolate, WT clone, N154Q mutant, or T156I mutant. Viruses in culture supernatants were titrated by focus-forming assay. Data shown are the mean values ± SEM of 9 samples from 3 independent experiments. (D and E) E proteins were immunoprecipitated with MAb 1M7 from lysates of Vero cells infected with ZIKV PRVABC59 isolate, WT clone, N154Q mutant, or T156I mutant. (D) Lysates were treated with PNGase F, separated by nonreducing SDS-PAGE, and probed with MAb 4G2. (E) Lysates were separated by nonreducing SDS-PAGE and probed with biotinylated lectin concanavalin A.

FIG 5

E glycosylation mediates ZIKV PRVABC59 infection in mice. Five- to six-week-old Ifnar1^−/− mice were inoculated with 1 × 10³ FFU of ZIKV strain PRVABC59 isolate, WT clone, N154Q mutant, or T156I mutant by a subcutaneous route in the footpad. (A) Blood was collected at 2, 4, and 6 days after infection, and ZIKV RNA levels in serum were measured by qRT-PCR. (B to E) Mice were euthanized 6 days after infection and perfused, and tissues were harvested. ZIKV RNA in tissues was measured by qRT-PCR. Data are combined from 2 independent experiments. *, P < 0.05; ***, P < 0.001; ****, P < 0.0001; ns, not significant (ANOVA).

FIG 6

E glycosylation mediates ZIKV MR766 infection in mice. (A and B). Seven- to 10-week-old Ifnar1^−/− mice were inoculated with 1 × 10³ FFU of ZIKV strain MR766 isolates containing the E glycan (+gly) or lacking the E glycan (−gly) by a subcutaneous route in the footpad. Mice were weighed daily, and weights are expressed as percentages of body weight prior to infection and censored once one mouse in a group died. Results shown are the mean values ± SEM of 14 to 16 Ifnar1^−/− mice per virus. Lethality was monitored for 21 days. (C to G) Five-week-old Ifnar1^−/− mice were inoculated with 1 × 10³ FFU of ZIKV strain MR766 +gly clone, −gly clone 4-amino-acid deletion (4AAΔ), or T156I clone by a subcutaneous route in the footpad. (C) Blood was collected at 2 and 4 days after infection, and ZIKV RNA levels in serum were measured by qRT-PCR. (D to G) Mice were euthanized 4 days after infection and perfused, and tissues were harvested. ZIKV RNA in tissues was measured by qRT-PCR. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, not significant (ANOVA).

FIG 7

E glycosylation facilitates ZIKV infection of DC-SIGN- and DC-SIGNR-expressing cells. (A) Representative flow cytometry plots of Raji, Raji-DC-SIGN, and Raji-DC-SIGNR cells stained for cell surface expression of DC-SIGN and DC-SIGNR. (B) Representative flow cytometry plots of Raji, Raji-DC-SIGN, and DC-SIGNR cells infected at an MOI of 5 with ZIKV PRVABC59 WT, N154Q, T156I, or UV-inactivated WT virus. Cells were stained at 24 hpi with Alexa Fluor 488-conjugated ZIKV MAb 4G2 to detect intracellular E protein. Values indicate the proportions of cells staining positive. (C) Percentages of infected (E-positive) Raji, Raji-DC-SIGN, or DC-SIGNR cells combined from 3 independent experiments performed in triplicate.

FIG 8

E glycosylation facilitates ZIKV infection of A549 but not Vero cells. (A to D) Vero and A549 cells were infected at an MOI of 5 with ZIKV PRVABC59 WT, N154Q, or T156I clones. Cells were stained at 24 hpi with Alexa Fluor 488-conjugated ZIKV MAb 4G2 to detect intracellular E protein. (A and C) Representative flow cytometry plots of infected Vero or A549 cells. (B and D) Percentages of infected (E positive) Vero or A549 cells combined from 3 independent experiments performed in triplicate. (E and F) A549 cells were infected at an MOI of 0.01 with ZIKV PRVABC59 WT clone, N154Q mutant, or T156I mutant or with ZIKV H/PF/2013 WT clone or N154Q mutant. Viruses in culture supernatants were titrated by focus-forming assay. Data shown are the mean values ± SEM of 9 samples from 3 independent experiments.