Zika convalescent macaques display delayed induction of anamnestic cross-neutralizing antibody responses after dengue infection

Abstract

Structural similarities between Zika (ZIKV) and dengue virus (DENV) leads to the induction of cross-reactive responses. We have previously demonstrated that ZIKV exposed macaques significantly enhance DENV viremia. Here we show that this enhancement of DENV infection occurred in the presence of high levels of DENV cross-reactive IgG1 subclass of binding antibodies (bAb) with low DENV neutralizing antibody (nAb) activity (<1:10). The DENV-2 nAb titres after ZIKV infection were, however, higher than those induced in DENV-2 only infected animals suggesting that ZIKV induced low titres of cross-nAb against DENV. Surprisingly, DENV-2 infection of animals previously infected with ZIKV was not accompanied by an anamnestic increase in cross-nAb titres till about 1 week after DENV-2 infection. This delay coincided with enhanced DENV-2 viremia indicating that high levels of cross-bAb in the absence of high nAb contributes to enhancement of DENV infection. Serum collected 8 weeks after DENV-2 infection had high levels of nAb and showed delayed antibody dependent enhancement (ADE) of infection (1:100 dilution) as compared with serum that was collected from ZIKV infected animals prior to DENV-2 infection (1:10 dilution). Examination of serum from macaques that were simultaneously infected with both ZIKV and DENV-2 showed high levels of nAb and delayed ADE responses raising the possibility that the low levels of cross-nAb induced by ZIKV infection could be overcome by co-immunization against ZIKV and DENV infection. Taken together, our results provide additional insights into the nature and kinetics of cross-reactive antibody responses and identify a critical correlate that could potentially prevent enhancement of DENV infection during ZIKV convalescence.

Fig. 1. Kinetics of cross-reactive antibody responses.

Relative binding antibody (bAb) levels against whole ZIKV and DENV-2 were examined using ELISA. The kinetics of binding IgM in serum that was collected longitudinally from (a) DENV-2 only (ZIKV naïve) infected animals (n = 4), (b) ZIKV-infected animals prior to DENV-2 infection (n = 5). The kinetics of binding IgG in serum that was collected longitudinally from (c) DENV-2 only (ZIKV naïve) infected animals (n = 4), (d) ZIKV-infected animals prior to DENV-2 infection (n = 5). Statistical significance was determined using multiple unpaired t-tests and corrected for multiple comparisons using the Holm-Sidak method. A p < 0.05 was considered significant and * indicates significant difference between groups

Fig. 2. ZIKV-infected animals display little or no cross-neutralizing antibody responses against DENV-2.

Percentage neutralization of (a) ZIKV and (b) DENV-2 by serum that was collected longitudinally from DENV-2 only (ZIKV naïve) (n = 4) and ZIKV-infected animals prior to DENV-2 infection (n = 5). Line represents PRNT₅₀ titres

Fig. 3. ZIKV immune animals display delayed anamnestic cross-neutralizing antibody responses after DENV-2 infection.

Kinetics of neutralizing antibody responses against DENV-2 using serum that was collected longitudinally from DENV-2 only (ZIKV naïve) and ZIKV-infected animals at (a) day 0, (b) day 1, (c) day 4, (d) day 7, (e) day 14, and (f) day 56 after DENV-2 infection (n = 5). Line represents 50% neutralization

Fig. 4. ZIKV infected animals display delayed binding antibody responses after DENV-2 infection.

Kinetics of binding antibody (bAb) responses against DENV-2 in serum showing relative levels of (a) IgM and (b) IgG against DENV-2 using serum that was collected longitudinally from ZIKV-infected animals 1, 4, and 7 days after DENV-2 infection (n = 5). Percentage OD₄₅₀ values relative to each animals day 56 post-ZIKV infection (day 0 DENV-2) values are shown. Line represents day 0 vales. Statistical significance was determined using multiple unpaired t-tests and corrected for multiple comparisons using the Holm-Sidak method. A p < 0.05 was considered significant and * indicates significant difference between groups

Fig. 5. IgG1 is the primary subclass of IgG in serum from ZIKV naïve and infected animals prior to and after infection with DENV-2.

Relative levels of (a) DENV-2 and (b) ZIKV binding IgG1, IgG2, and IgG3 in serum that was collected at day 14 and day 56 post-infection from DENV-2 only (n = 4) infected animals. Relative levels of (c) DENV-2 and (d) ZIKV binding IgG1, IgG2, and IgG3 in serum that was collected at day 14 and day 56 post-infection from ZIKV-infected animals prior to DENV-2 infection (n = 5). Relative levels of (e) DENV-2 and (f) ZIKV binding IgG1, IgG2, and IgG3 in serum that was collected at day 1, 4, 7, and 56 post-infection from ZIKV-infected animals after DENV-2 infection (n = 5). Statistical significance was determined using one-way ANOVA and differences between time-points were determined by post-hoc analysis using Tukey’s multiple comparisons test. A p < 0.05 was considered significant. Error bars represent standard error

Fig. 6. Serum from ZIKV infected animals collected after infection with DENV-2 demonstrate antibody dependent enhancement at a dilution of 1:100.

Fold enhancement of DENV-2 infection in K562 cells using serum that was collected (a) from ZIKV-infected animals (n = 5) prior to DENV-2 infection and (b) at day 56 after DENV-2 infection. Statistical significance was determined using One-way ANOVA and differences between time-points were determined by post-hoc analysis using Tukey’s multiple comparisons test. A p < 0.05 was considered significant. (c) Percentage neutralization of DENV-2 using serum that was collected at day 56 post-infection from animals that was simultaneously infected with ZIKV and DENV-2 (n = 5). (d) Fold enhancement of DENV-2 infection in K562 cells using serum that was collected at 56 days from rhesus macaques that were simultaneously infected with ZIKV and DENV-2 (n = 5). Statistical significance was determined using One-way ANOVA and differences between time-points were determined by post-hoc analysis using Tukey's multiple comparisons test. A p < 0.05 was considered significant. Error bars represent standard error