Avatar Mice Underscore the Role of the T Cell-Dendritic Cell Crosstalk in Ebola Virus Disease and Reveal Mechanisms of Protection in Survivors

Abstract

Ebola virus disease (EVD) is a complex infectious disease characterized by high inflammation, multiorgan failure, the dysregulation of innate and adaptive immune responses, and coagulation abnormalities. Evidence accumulated over the last 2 decades indicates that, during fatal EVD, the infection of antigen-presenting cells (APC) and the dysregulation of T cell immunity preclude a successful transition between innate and adaptive immunity, which constitutes a key disease checkpoint. In order to better understand the contribution of the APC-T cell crosstalk to EVD pathophysiology, we have developed avatar mice transplanted with human, donor-specific APCs and T cells. Here, we show that the transplantation of T cells and APCs from Ebola virus (EBOV)-naive individuals into avatar mice results in severe disease and death and that this phenotype is dependent on T cell receptor (TCR)-major histocompatibility complex (MCH) recognition. Conversely, avatar mice were rescued from death induced by EBOV infection after the transplantation of both T cells and plasma from EVD survivors. These results strongly suggest that protection from EBOV reinfection requires both cellular and humoral immune memory responses. IMPORTANCE The crosstalk between dendritic cells and T cells marks the transition between innate and adaptive immune responses, and it constitutes an important checkpoint in EVD. In this study, we present a mouse avatar model in which T cell and dendritic cell interactions from a specific donor can be studied during EVD. Our findings indicate that T cell receptor-major histocompatibility complex-mediated T cell-dendritic cell interactions are associated with disease severity, which mimics the main features of severe EVD in these mice. Resistance to an EBOV challenge in the model was achieved via the transplantation of both survivor T cells and plasma.

Keywords: Ebola; Ebola virus; T cells; avatar; dendritic cells; humanized mice; mouse models.

FIG 1

Generation and characterization of avatar mice. (A) Schematic indicating the procedure by which to generate mouse avatars. HuPBLs (CD14 negative PBMC fraction) are transplanted at day 0. CD14⁺ cells from the same donor are incubated with GM-CSF and IL-4 to generate immature monocyte-derived DCs, which are then infected with EBOV at an MOI of 1 and transplanted at day 5. (B) Frequency of human hematopoietic cells (hCD45⁺) in the peripheral blood of avatar mice at the indicated time points. Data are shown as the mean ± the standard error the mean (SEM). (C) Representative flow cytometry plots of the data shown in panel B. (D) Frequency of CD8⁺ central memory, effector memory, naive and effector T cells (TEMRA) at the indicated time points in the peripheral blood of avatar mice, as assessed by flow cytometry. (E) Frequency of CD4⁺ central memory, effector memory, naive and effector T cells (TEMRA) at the indicated time points in the peripheral blood of avatar mice, as assessed by flow cytometry.

FIG 2

The EVD course in avatar mice is donor-specific. (A and E) Body weight loss curves of avatar mice transplanted with huPBL and DCs from donor 1 (D1) and donor 2 (D2), respectively. Mock-mice were transplanted with noninfected DCs, whereas EBOV-mice were transplanted with DCs infected with EBOV for 60 min at an MOI of 1. (B and F) Kaplan-Meier survival curves for D1 and D2-avatar mice. (C and G) Levels of viremia as determined by the assessment of EBOV infectious particles via focus-formation assay at the indicated time points. Shaded areas (gray) mark the limit of detection of the assay. (D and H) Levels of serum aspartate transaminase (AST) in D1 and D2 avatars. In panels A, E, D, and H, statistical significance was determined using a two-way analysis of variance (ANOVA) followed by Dunnett’s multiple comparison test. In panels B and F, statistical evaluation was performed via the Mantel-Cox test. Across the figure, significance levels are presented as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001, ****; and P ≤ 0.0001. Data are shown as mean ± SEM. D1 (Mock n = 5, EBOV n = 5), D2 (Mock n = 5, EBOV n = 4).

FIG 3

The onset of EVD in avatar mice depends on T cell-DC interactions. (A) Body weight loss curves of avatar mice transplanted with EBOV-infected DCs at the indicated inputs. Numbers between brackets indicate the number of surviving mice out of n = 5 mice in each group. (B) The Kaplan-Meier survival curves of avatar mice transplanted with infected DCs are as follows. Mock-mice were transplanted with noninfected moDCs, and moDCs indicates monocyte-derived DCs. MUTZ3-DCSign indicates MUTZ-3 cells derived into DC populations enriched in interstitial DC-Sign-like DCs. MUTZ3-Langerin indicates MUTZ-3 cells derived into DC populations enriched in Langerhans-like cells. The bottom panel indicates viremia, as assessed via focus forming assay at the indicated time points. This experiment was done once with n = 5 mice/group. (C) Avatar mice were transplanted with EBOV-infected moDCs in the absence of a previous infusion of autologous huPBLs. The Kaplan-Meier survival curves in comparison with the mock are shown in the upper panel, while the viremia and AST serum levels are shown in the lower panels. Gray areas indicate the limit of detection. (D) Avatar mice were generated by the transplantation of huPBLs followed by the transplantation of EBOV-infected moDCs from HLA-matched (HLA-A*02:01) or mismatched donors. Survival curves, viremia and AST levels in serum are shown. Across the figure, significance levels are presented as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; and ****, P ≤ 0.0001. The survival curves were assessed via Mantel-Cox. All other analyses were conducted with a two-way ANOVA followed by Dunnett’s multiple comparison test. Data are shown as mean ± SEM.

FIG 4

Avatar mice recapitulate EVD but not Lassa fever. (A) Model schematic. EBOV, Ebola virus; RESTV, Reston virus; and LASV, Lassa virus. (B) Kaplan-Meier survival curves of avatar mice transplanted with EBOV-infected moDCs (red), RESTV-infected moDCs (blue), or LASV-infected DCs (green). (C) Body weight loss curves of avatar mice infected with each one of the three viruses. (D) Levels of serum AST at the indicated time points in all three models. (E) Levels of viremia. The gray areas represent the limit of detection of the assay. (F) Virus titers in organs at the time of necropsy. Organ titers were calculated via immunofocus assay as described in Materials and Methods section. Across the figure, significance levels are presented as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; and ****, P ≤ 0.001. The survival curves were assessed via Mantel-Cox, and the comparison of virus titers in organs was assessed via a Kruskal-Wallis nonparametric analysis. All other analyses were conducted with a two-way ANOVA followed by Dunnett’s multiple comparison test. Data are shown as mean ± SEM.

FIG 5

EVD in avatar mice transplanted with survivor huPBLs. (A) Body weight loss curves of avatar mice transplanted with control huPBLs plus noninfected autologous moDCs (Mock, black), control huPBLs plus HLA-matched EBOV-infected moDCs (Controls, blue), or EVD survivor huPBLs plus HLA-matched EBOV-infected moDCs (Survivors, red). (B) Kaplan-Meier curves. (C) Levels of viremia at the indicated time points in survivor-avatars (red) and control-avatars (blue). The gray area indicates the assay limit of detection. (D) Levels of serum AST in survivor-avatars (red) and control-avatars (blue) at the indicated time points. Across the figure, significance levels are presented as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; and ****, P ≤ 0.0001. The survival curves were assessed via Mantel-Cox, and all other analyses were conducted with a two-way ANOVA followed by Dunnett’s multiple comparison test. Data are shown as mean ± SEM.

FIG 6

Transplantation of both EVD survivor huPBLs and plasma rescued avatar mice from the EBOV challenge. (A) Model schematic. (B) Avatar mice body weight loss after transplantation with EBOV-infected HLA-matched moDCs in mice that received control huPBLs (EBOV, red), control huPBLs plus anti-EBOV MAb KZ52 (EBOV+KZ52, green), EBOV plus low neutralizing survivor plasma (EBOV+lowNab, blue), or EBOV plus high neutralizing survivor plasma (EBOV+hiNab, pink). (C) Body weight loss curves as indicated in panel B, but in avatar mice that received EVD-survivor huPBLs and HLA-matched DCs. (D) Kaplan-Meier curves of mice that received control huPBLs. (E) Kaplan-Meier curves of mice that received survivor huPBLs. Across the figure, significance levels are presented as follows: *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; and ****, P ≤ 0.0001. The survival curves were assessed via Mantel-Cox, and all other analyses were conducted with a two-way ANOVA followed by Dunnett’s multiple comparison test. Data are shown as mean ± SEM.