Storage-Dependent Generation of Potent Anti-ZIKV Activity in Human Breast Milk

Abstract

Zika virus (ZIKV) causes congenital neurologic birth defects, notably microcephaly, and has been associated with other serious complications in adults. The virus has been detected in human breast milk and possible transmissions via breastfeeding have been reported. Breast milk is rich in nutrients and bio-active substances that might directly affect viral infectivity. Thus, here, we analyzed the effect of human breast milk on ZIKV infection. We observed that fresh human breast milk had no effect on ZIKV, but found that upon storage, milk effectively suppressed infection. The antiviral activity is present in the fat-containing cream fraction of milk and results in the destruction of the structural integrity of viral particles, thereby abrogating infectivity. The release of the factor is time dependent but varies with donors and incubation temperatures. The viral titer of milk that was spiked with ZIKV decreased considerably upon storage at 37 °C for 8 h, was lost entirely after 2 days of 4 °C storage, but was not affected at -20 °C. This suggests that cold storage of milk inactivates ZIKV and that the antiviral factor in milk may also be generated upon breastfeeding and limit this transmission route of ZIKV.

Keywords: breast milk; breastfeeding; transmission; zika virus.

Figure 1

Effect of long-term frozen breast milk on ZIKV infection. (a) ZIKV MR766, GWUH or PRV (MR766: 1.58, GWUH: 2.51, PRV: 1.58 × 10⁷ TCID₅₀/mL) were mixed 1:1 with thawed breast milk at indicated concentrations and incubated for 10 min at room temperature. Milk was stored for more than 6 months at −20 °C. ZIKV/milk mixtures were then used to inoculate Vero E6 cells resulting in a 10-fold dilution of the samples. Then, 2 h later, the medium was changed and, 2 days later, the infection rates were determined by a cell-based immunodetection assay that enzymatically quantifies the flavivirus protein E. (b) Vero E6 cells were incubated with thawed pooled breast milk at indicated concentrations for 2 h. The medium was then replaced, and the cellular viability was determined 2 days later by CellTiter-Glo Assay. Data are normalized to viability in the absence of milk. (c) Thawed breast milk from four individual donors was incubated with ZIKV MR766 (1.58 × 10⁷ TCID₅₀/mL) at indicated concentrations before inoculation of Vero E6 cells as described in (a). Infection data are normalized to infection rates in the absence of the respective sample. Data are represented as average values obtained from triplicate infections ± standard deviations.

Figure 2

Mechanism of ZIKV inhibition by long-term freeze-stored breast milk. (a) For cell treatment (cell), Vero E6 cells were incubated with indicated concentrations of breast milk, and thereafter inoculated with ZIKV MR766. For virion treatment (virion), ZIKV MR766 (1.58 × 10⁷ TCID₅₀/mL) was incubated 1:1 for 10 min with 10-fold higher breast milk concentrations before the mix was diluted onto Vero E6 cells. Concentrations shown correspond to the final concentrations of milk (v/v) in cell culture. Then, 2 h after inoculation, the medium was changed and, 2 days later, the infection rates were determined by a cell-based immunodetection assay that enzymatically quantifies the flavivirus protein E. (b) ZIKV MR766 was incubated with PBS (90%), SDS (0.5%), Triton X-100 (1%) or milk (90%) for 1 h at 37 °C. Free (and released) RNA was incubated in buffer or degraded by RNase A (10 U) for 1 h at 37 °C. Remaining ZIKV RNA copy numbers were determined by RT-qPCR. (c) PBS or breast milk (5%) were incubated with ZIKV MR766 for the indicated time at room temperature before the mixture was inoculated onto Vero E6 cells. After 2 h, the medium was changed and, 2 days later, the infection rates were determined as described in (a). Infection data are normalized to infection rates in the absence of the respective sample. All data are represented as average values obtained from triplicates ± standard deviations.

Figure 3

Effect of fractions derived from long-term freeze-stored breast milk on ZIKV infection. (a) Pooled long-term stored breast milk was defatted by centrifugation, the cream collected and resuspended in PBS at the original volume. The remaining skim milk was further ultracentrifuged to separate caseins from serum proteins. The casein-containing pellet was resuspended in PBS to the original volume and the supernatant (whey) collected. The breast milk and its components were then incubated on Vero E6 cells at indicated concentrations for 2 h. The medium was then replaced, and the cellular viability was determined 2 days later by CellTiter-Glo Assay. (b) Breast milk and its components were mixed 1:1 with ZIKV MR766 (1.58 × 10⁷ TCID₅₀/mL) for 10 min before 10-fold dilution onto Vero E6 cells resulting in indicated concentrations. Then, 2 h after inoculation, the medium was changed and, 2 days later, the infection rates were determined by a cell-based immunodetection assay that enzymatically quantifies the flavivirus protein E. All data are normalized to values in the absence of the respective sample. Data are represented as average values obtained from triplicate infections ± standard deviations.

Figure 4

Effect of fresh milk on ZIKV infection. Fresh breast milk was received from three donors and, after 30 min, was mixed 1:1 with ZIKV MR766 (1.58 × 10⁷ TCID₅₀/mL) at indicated concentrations and incubated for 10 min at room temperature. Mixtures were then diluted 10-fold onto Vero E6 cells and incubated for 2 h before the medium was changed. Infection rates were determined 2 days later by an E protein immunodetection assay. Infection data are normalized to infection rates in the absence of the respective breast milk sample and represent average values obtained from triplicate infections ± standard deviations.

Figure 5

Effect of storage and temperature on the anti-ZIKV activity of milk. Fresh breast milk was received from three donors and, after 30 min, was mixed 1:1 with ZIKV MR766 (1.58 × 10⁷ TCID₅₀/mL) at indicated concentrations and incubated for 10 min at room temperature. Mixtures were then diluted 10-fold onto Vero E6 cells and incubated for 2 h before the medium was changed. Additionally, the breast milk was incubated at (a) 37 °C, (b) 22 °C, (c) 4 °C, or (d) −20 °C for indicated time points before mixing with ZIKV and inoculation of Vero E6 cells. Infection rates were determined 2 days later by a cell-based immunodetection assay that enzymatically quantifies the flavivirus protein E. Infection data are normalized to infection rates in the absence of the respective breast milk sample and represent average values obtained from triplicate infections ± standard deviations. See also Figures S3–S6.

Figure 6

Determination of the ZIKV titer after the incubation of the virions in breast milk at different temperatures. ZIKV MR766, GWUH, or PRV were spiked into buffer or fresh breast milk that was received from two donors resulting in 90% buffer or milk, respectively. Samples were then incubated at (a) 22 °C, (b) 37 °C, (c) 4 °C or (d) −20 °C for indicated time points up to 8 h (a+b) or 10 days (c+d), before the virus titer was determined by TCID₅₀ titration onto Vero E6 cells.