Ebola virus-induced eye sequelae: a murine model for evaluating glycoprotein-targeting therapeutics

Abstract

Background: Ebola virus disease (EVD) survivors experience ocular sequelae including retinal lesions, cataracts, and vision loss. While monoclonal antibodies targeting the Ebola virus glycoprotein (EBOV-GP) have shown promise in improving prognosis, their effectiveness in mitigating ocular sequelae remains uncertain.

Methods: We developed and characterized a BSL-2-compatible immunocompetent mouse model to evaluate therapeutics targeting EBOV-GP by inoculating neonatal mice with vesicular stomatitis virus expressing EBOV-GP (VSV-EBOV). To examine the impact of anti-EBOV-GP antibody treatment on acute retinitis and ocular sequelae, VSV-EBOV-infected mice were treated with polyclonal antibodies or monoclonal antibody preparations with antibody-dependent cellular cytotoxicity (ADCC-mAb) or neutralizing activity (NEUT-mAb).

Findings: Treatment with all anti-EBOV-GP antibodies tested dramatically reduced viremia and improved survival. Further, all treatments reduced the incidence of cataracts. However, NEUT-mAb alone or in combination with ADCC-mAb reduced viral load in the eyes, downregulated the ocular immune and inflammatory responses, and minimized retinal damage more effectively.

Interpretation: Anti-EBOV-GP antibodies can improve survival among EVD patients, but improved therapeutics are needed to reduce life altering sequelae. This animal model offers a new platform to examine the acute and long-term effect of the virus in the eye and the relative impact of therapeutic candidates targeting EBOV-GP. Results indicate that even antibodies that improve systemic viral clearance and survival can differ in their capacity to reduce acute ocular inflammation, and long-term retinal pathology and corneal degeneration.

Funding: This study was partly supported by Postgraduate Research Fellowship Awards from ORISE through an interagency agreement between the US DOE and the US FDA.

Keywords: Anti-EBOV-GP antibodies; BSL-2 model; EBOV glycoprotein (EBOV-GP); EBOV-GP pseudotyped vesicular stomatitis virus (VSV-EBOV); Ebola virus (EBOV); Ocular sequelae; Retinitis.

Fig. 1

Therapeutic effects of anti-EBOV-GP antibodies on weight change, survival and lymphopenia in VSV-EBOV-infected mice. (a) Virus neutralization by antibodies: SAB-139 (polyAb), REGN3478 (ADCC-mAb), and REGN3481 (NEUT-mAb) or the combination of ADCC-mAb and NEUT-mAb using VERO E6 cells infected with replicating VSV-EBOV (MOI 0.1). Data are representative of three independent experiments. (b–d) P3 C57BL/6 mice were subcutaneously (s.c.) infected with 1000 TCID50 of VSV-EBOV and intraperitoneally (i.p.) treated with human IgG isotype controls (IgG control; 100 mg/kg, (b) n = 9, (c) n = 17, or (d) n = 6, respectively), polyAb (100 mg/kg, (b) n = 3, (c) n = 12, or (d) n = 6), ADCC-mAb (100 mg/kg, (b) n = 10, (c) n = 13, or (d) n = 6), NEUT-mAb (100 mg/kg, (b) n = 8, (c) n = 12, or (d) n = 8) or the combination of ADCC-mAb and NEUT-mAb (50 mg/kg, each, (b) n = 11, (c) n = 10, or (d) n = 6) at 3 dpi. Mice were monitored for weight changes (b) and survival (c). Data are shown as mean ± standard deviation (SD). Statistical significance was determined using a two-way ANOVA (mixed-effects model with the Geisser-Greenhouse correction) (b) and the log-rank test (c), respectively. ∗∗∗P < 0.001 compared to infected mice treated with IgG control. (d) Hematological assessment at 9 dpi. Table shows incidence of lymphopenia by 9 dpi.

Fig. 2

Viral burden in the eyes of VSV-EBOV-infected mice treated with anti-EBOV-GP antibodies. P3 C57BL/6 mice were infected s.c. with 1000 TCID50 of VSV-EBOV and treated IgG control (100 mg/kg, (a) n = 5 or (b) n = 6), polyAb (100 mg/kg, (a) n = 6 or (b) n = 3), ADCC-mAb (100 mg/kg, (a) n = 3 or (b) n = 6), NEUT-mAb (100 mg/kg, (a) n = 3 or (b) n = 4) or the combination of ADCC-mAb and NEUT-mAb (50 mg/kg, each, (a) n = 3 or (b) n = 5) at 3 dpi. (a) Viral loads in the eyes of mice were evaluated by TCID50 at 3, 6 and 9 dpi. Data are shown as mean + SD. N.D., not detectable. (b) Viral antigen in the eyes of uninfected mice (n = 5) and VSV-EBOV-infected mice treated as above at 3 dpi and sacrificed at 9 dpi. The images show representative immunofluorescence images of eye sections stained for EBOV-GP antigen (red), NF-160 (green) and DAPI (blue). Scale bar: 50 μm. GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer; IS/OS, inner/outer segment junction; RPE, retinal pigment epithelium. Images of the full eye can be found in Supplementary Figure S6.

Fig. 3

Therapeutic effects of anti-EBOV-GP antibodies on acute retinal inflammation in VSV-EBOV-infected mice. P3 C57BL/6 mice were infected, treated as above at 3 dpi, and sacrificed at 9 dpi. (a) Histopathology of the eyes of uninfected mice (n = 4) and VSV-EBOV-infected treated with IgG control (n = 10), polyAb (n = 7), ADCC-mAb (n = 6), NEUT-mAb (n = 4) or the combination of ADCC-mAb and NEUT-mAb (n = 5). Left images show representative H&E staining of eyes. Right graph shows the retinal inflammation score. Data are shown as mean + SD. (b) Left images show representative immunofluorescence staining for TUNEL (green) and DAPI (blue) in eye sections from uninfected mice (n = 5) and VSV-EBOV-infected mice treated with IgG control (n = 6), polyAb (n = 3), ADCC-mAb (n = 6), NEUT-mAb (n = 4) or the combination of ADCC-mAb and NEUT-mAb (n = 5). Scale bar: 50 μm. Images of the full eye can be found in Supplementary Figure 6. Right graph shows the quantification of TUNEL + cells per field of view (3 fields per eye section). Data are shown as mean + S.D. (n = 9–18 fields for each group). (c) The images show representative immunofluorescence staining for CD45 (green), IBA (red) and DAPI (blue) in eye sections from uninfected mice (n = 5) and VSV-EBOV-infected mice treated with IgG control (n = 6), polyAb (n = 3), ADCC-mAb (n = 6), NEUT-mAb (n = 4) or the combination of ADCC-mAb and NEUT-mAb (n = 5). The bottom close-up images show representative microglia morphology observed in eyes. Scale bar: 50 μm. Images of the full eye can be found in Supplementary Figure S6. (d) Flow cytometry analysis of cells isolated from the eyes collected at 9 dpi. The graphs show the percentage of infiltrating immune cells (CD45⁺) and T cells (CD45⁺CD3⁺NK1.1⁻) in the eyes from uninfected mice (n = 3) and VSV-EBOV-infected mice treated with IgG control (n = 5), polyAb (n = 4), ADCC-mAb (n = 7), NEUT-mAb (n = 5) or the combination of ADCC-mAb and NEUT-mAb (n = 5). Data are shown as mean + SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001 compared to uninfected mice. #P < 0.05 and ##P < 0.01 compared to VSV-EBOV-infected mice treated with IgG control.

Fig. 4

Pattern of inflammation in the eyes of VSV-EBOV-infected mice treated with IgG control or anti-EBOV-GP-antibodies. Heat map of mRNA expression genes selected from a 591 gene array in the eyes at 9 dpi from VSV-EBOV-infected mice treated with IgG control (n = 8), polyAb (n = 7), ADCC-mAb (n = 3), NEUT-mAb (n = 3) or the combination of ADCC-mAb and NEUT-mAb (n = 3). RNA expression was assessed by NanoString using the nCounter mouse Immunology panel. Expression levels were normalized to the mean value of the uninfected group (n = 3). Results from the full gene array can be found in Supplementary Figure S7.

Fig. 5

Sequelae of VSV-EBOV-infected mice treated with IgG control or anti-EBOV-GP antibodies. (a and b) P3 neonatal mice were uninfected (n = 19) or infected s.c. with 500 TCID50 of VSV-EBOV (n = 42) and monitored for survival (a) for 60 days. Statistical significance was determined using the log-rank test. ∗∗∗P < 0.001 compared to uninfected mice. (b) Comparison of viral load in the eyes from mice challenged with 500 TCID50 (n = 3) or 1000 TCID50 (n = 5) of VSV-EBOV by TCID50 analysis at 9 dpi. Data are shown as mean + SD. (c–f) P3 C57BL/6 mice were uninfected (n = 8) or infected with 500 TCID50 of VSV-EBOV and treated with IgG control (n = 6), polyAb (n = 6), ADCC-mAb (n = 8), NEUT-mAb (n = 3) or the combination of ADCC-mAb and NEUT-mAb (n = 4) at 3 dpi, and sacrificed at 3–6 months post infection (mpi). RNA from eyes of uninfected and convalescent mice was tested for the expression of VSV-N (c), VSV-EBOV-positive sense RNA (d, left) and VSV-EBOV-negative sense RNA (d, right) in the eyes. Ct values normalized to housekeeping gene (GAPDH). Data are shown as mean + SD. The red dashed lines indicate the limit of detection. (e) Histopathology of the eyes at 3–6 mpi. Left images show representative H&E staining of eyes from uninfected mice (n = 9) and VSV-EBOV-infected mice treated with IgG control (n = 10), polyAb (n = 7), ADCC-mAb (n = 9), NEUT-mAb (n = 11) or the combination of ADCC-mAb and NEUT-mAb (n = 8). Right graph shows the retinal inflammation score. Data are shown as mean + SD. (f) Fundus images of the eyes showing cataract development at 3–6 mpi. Right table shows cataract incidence by eyes.