Varicella zoster virus productively infects human natural killer cells and manipulates phenotype

Abstract

Varicella zoster virus (VZV) is a ubiquitous human alphaherpesvirus, responsible for varicella upon primary infection and herpes zoster following reactivation from latency. To establish lifelong infection, VZV employs strategies to evade and manipulate the immune system to its advantage in disseminating virus. As innate lymphocytes, natural killer (NK) cells are part of the early immune response to infection, and have been implicated in controlling VZV infection in patients. Understanding of how VZV directly interacts with NK cells, however, has not been investigated in detail. In this study, we provide the first evidence that VZV is capable of infecting human NK cells from peripheral blood in vitro. VZV infection of NK cells is productive, supporting the full kinetic cascade of viral gene expression and producing new infectious virus which was transmitted to epithelial cells in culture. We determined by flow cytometry that NK cell infection with VZV was not only preferential for the mature CD56dim NK cell subset, but also drove acquisition of the terminally-differentiated maturity marker CD57. Interpretation of high dimensional flow cytometry data with tSNE analysis revealed that culture of NK cells with VZV also induced a potent loss of expression of the low-affinity IgG Fc receptor CD16 on the cell surface. Notably, VZV infection of NK cells upregulated surface expression of chemokine receptors associated with trafficking to the skin -a crucial site in VZV disease where highly infectious lesions develop. We demonstrate that VZV actively manipulates the NK cell phenotype through productive infection, and propose a potential role for NK cells in VZV pathogenesis.

Fig 1. VZV infects human peripheral blood NK cells, CD3⁺CD56⁺ lymphocytes, and T cells.

Healthy human donor PBMCs were inoculated with mock or VZV infected ARPE-19 epithelial cells for 2 days then analysed for infection by flow cytometry. (A) Representative flow cytometry plots of mock or VZV-S infection, examining surface VZV gE:gI expression on live T cells (CD3⁺CD56^–), CD3⁺CD56⁺ lymphocytes, and NK cells (CD3^–CD56⁺). (B) Frequencies of live gE:gI⁺ lymphocytes in total (shaded), compared to specific populations: T cells, CD3⁺CD56⁺ lymphocytes, and NK cells (n = 19). Symbols represent individual donors consistent across lymphocyte populations, and bars indicate mean. Statistical analysis was performed between specific lymphocyte populations. **p < 0.01, ****p < 0.0001 (RM one-way ANOVA with the Greenhouse-Geisser correction and Tukey’s multiple comparisons test). (C) Representative flow cytometry plots of vOka infection, examining surface gE:gI expression on live T cells (CD3⁺CD56^–), CD3⁺CD56⁺ lymphocytes, and NK cells (CD3^–CD56⁺) (n = 3).

Fig 2. IL-2 stimulation of NK cells, CD3⁺CD56⁺ lymphocytes, and T cells enhances VZV infection.

(A) Healthy human donor PBMCs were infected with VZV by cell-associated infection with or without IL-2 (200 U/ml) for 2 days, then analysed by flow cytometry. Plots show surface VZV gE:gI expression from one representative donor and graphs show frequency of live gE:gI⁺ NK cells (CD3^–CD56⁺) (top panels), CD3⁺CD56⁺ lymphocytes (middle panels), and T cells (CD3⁺CD56^–) (bottom panels). Symbols represent individual donors consistent across lymphocyte populations, and bars indicate mean (n = 8). ***p < 0.001, ****p < 0.0001 (two-tailed paired t test). (B & C) Healthy human donor CD56⁺-selected lymphocytes were infected with VZV by cell-associated infection with or without IL-2 (200 U/ml) for 2 days, then analysed by flow cytometry. Plots show surface gE:gI expression from one representative donor and graphs show frequency of live gE:gI⁺ NK cells (B) or CD3⁺CD56⁺ lymphocytes (C). Symbols represent individual donors, consistent across (B & C) (n = 7). *p < 0.05 (two-tailed Wilcoxon matched-pairs signed rank test).

Fig 3. NK cells are productively infected by VZV and support virus transmission.

NK cells (CD3^–CD56⁺) were FACS sorted from healthy human donor CD56⁺-selected lymphocytes following mock or VZV infection for 1 day. (A & B) Staining by IFA of sorted VZV cultured (left panels) or mock cultured (right panels) NK cells for IE63 (A), pORF29 (B) or respective isotype control, with DAPI (n = 3). (C) Sorted VZV cultured NK cells were added to ARPE-19 epithelial cell monolayers. Four days later monolayers were fixed and infectious centres detected with IFA by staining for IE63 and gE:gI, with DAPI. One representative experiment of five is shown.

Fig 4. CD56^dim NK cells are more effective than CD56^bright NK cells at supporting VZV infection.

(A) Healthy human donor PBMCs were infected with VZV for 2 days then analysed for infection by flow cytometry. Plots show gating strategy for CD56^bright and CD56^dim NK cells (CD3^–CD56⁺) (left panel), with respective surface VZV gE:gI expression (right panels) from one representative donor (n = >7). (B & C) CD3^–CD56^bright (B) and CD3^–CD56^dim (C) NK cells were isolated from healthy human donor PBMCs by FACS sorting and subsequently infected with VZV for 2 days before analysis by flow cytometry. Plots show surface gE:gI expression from one representative donor (n = 2).

Fig 5. NK cell markers associated with maturity influence VZV infection of NK cells and are modulated by VZV.

Healthy human donor PBMCs were mock or VZV infected with or without IL-2 (200 U/ml) for 2 days then analysed by flow cytometry. (A) Diagram describes gating strategy and tSNE analysis workflow for samples shown in (B & C). (B & C) tSNE plots show marker expression levels for single parameters on individual cells in the tSNE map for mock and VZV cultured NK cells after 2 days, either untreated (B) or in the presence of IL-2 (C). Arrowheads indicate the CD56^bright NK cell subset, and the outlined population indicates the localisation of VZV⁺ NK cells. One representative experiment of three is shown. (D & E) Plots show CD57 expression between mock and VZV cultured NK cells (D) and between bystander and VZV⁺ NK cells (E), from one representative donor. Graphs show respective frequencies of CD57⁺ NK cells when untreated or with IL-2 (shaded) for four donors. Bars indicate mean. (F) Histograms show CD16 expression for mock, bystander and VZV⁺ NK cells from one representative donor. Graph shows frequency of CD16⁺ NK cells when untreated or with IL-2 (shaded) for six donors. Bars indicate mean. *p < 0.05, **p < 0.01, ***p < 0.001 (Friedman test with Dunn’s multiple comparisons test).

Fig 6. VZV infects both CD57^– and CD57^bright NK cells and drives CD57 expression.

CD3^–CD56⁺CD57^– NK cells and CD3^–CD56⁺CD57^bright NK cells were isolated from healthy human donor PBMCs by FACS sorting and subsequently mock or VZV infected with or without IL-2 (200 U/ml) for 2 days before analysis by flow cytometry. (A) Diagram describes experimental design of isolating NK cells on CD57 expression, then infecting, and subsequently analysing for infection and phenotype changes. (B) Plots show surface VZV gE:gI expression between subsets from one representative donor. Graph shows frequency of VZV⁺ NK cell subsets when untreated or with IL-2 (shaded) for three donors. Bars indicate mean. (C) Plots show subsequent CD57 expression between mock, bystander and VZV⁺ CD57^– NK cells (left panels) and CD57 versus gE:gI expression for VZV cultured CD57^– NK cells (middle panels), from one representative donor. Graphs show frequency of CD57 expression on mock, bystander and VZV⁺ CD57^– NK cells for three donors. Bars indicate mean. *p < 0.05 (two-tailed paired t test). (D) Histograms show CD16 expression for mock, bystander and VZV⁺ CD57^– NK cells (left panel) and CD57^bright NK cells (right panel) for one representative donor (n = 3).

Fig 7. VZV upregulates expression of skin-homing chemokine receptors on NK cells.

Healthy human donor PBMCs were mock or VZV infected with or without IL-2 (200 U/ml) for 2 days then analysed by flow cytometry. (Left panels) Representative plots show CCR4 (A) or CLA (B) expression against CD56 expression for mock, bystander and VZV⁺ NK cells. (Right panels) Representative plots show CCR4 (A) or CLA (B) expression versus VZV gE:gI expression for VZV cultured NK cells. Data are representative of five donors.