Differential Response Following Infection of Mouse CNS with Virulent and Attenuated Vaccinia Virus Strains

Abstract

Viral infections of the central nervous system (CNS) lead to a broad range of pathologies. CNS infections with Orthopox viruses have been mainly documented as an adverse reaction to smallpox vaccination with vaccinia virus. To date, there is insufficient data regarding the mechanisms underlying pathological viral replication or viral clearance. Therefore, informed risk assessment of vaccine adverse reactions or outcome prediction is limited. This work applied a model of viral infection of the CNS, comparing neurovirulent with attenuated strains. We followed various parameters along the disease and correlated viral load, morbidity, and mortality with tissue integrity, innate and adaptive immune response and functionality of the blood⁻brain barrier. Combining these data with whole brain RNA-seq analysis performed at different time points indicated that neurovirulence is associated with host immune silencing followed by induction of tissue damage-specific pathways. In contrast, brain infection with attenuated strains resulted in rapid and robust induction of innate and adaptive protective immunity, followed by viral clearance and recovery. This study significantly improves our understanding of the mechanisms and processes determining the consequence of viral CNS infection and highlights potential biomarkers associated with such outcomes.

Keywords: RNA-seq; brain; meningoencephalitis; neurovirulence; smallpox; vaccinia.

Figure 1

Disease progression following intracranial infection. Mice were injected intracranially with vaccinia virus-western Reserve (VACV-WR) (10² pfu), VACV-WR with the vFire cassette (VACV-WRvFire) (10² pfu), VACV-Lister (10⁶ pfu) or VACV-WyethvFire (10⁶ pfu). (A) Weight change following infection. (B) Viral loads, relative (log scale) to the initial infection dose, in brain (2, 4, 5 d.p.i.), blood (5 d.p.i.) and spleen (4 d.p.i. for VACV-WR, 5 d.p.i. for VACV-WRvFire and VACV-WyethvFire). All tissues were weight normalized for analysis. Asterisks denote p < 0.05 (t-test). (C) In-vivo whole body imaging following VACV-WRvFire at 5 d.p.i. (photon flux color scale: 10⁸–10⁹ p/s/cm²/sr). (D) In-vivo whole body imaging following VACV-WyethvFire at 5 d.p.i. (photon flux color scale: 10⁵–10⁶ p/s/cm²/sr). (E) In-vivo whole body imaging following VACV-WyethvFire at 7 d.p.i. (photon flux color scale: 10⁵–10⁶ p/s/cm²/sr). (F) Head signal intensity over the background noise at 2–5 d.p.i in VACV-WRvFire and 2–13 d.p.i in VACV-WyethvFire. Asterisks denote p < 0.05 in post-hoc t-test following 2-way ANOVA. Error bars represent standard errors.

Figure 2

Detection of VACV-WR but not VACV-Wyeth antigens in the meninges and the ventricular system. Brain sections, 5 d.p.i with VACV-WR (10² pfu) or VACV-Wyeth (10⁶ pfu) stained with Hematoxylin and Eosin (H&E) or with antibodies to VACV (positive stain in brown). (A,B) VACV-WR. (C,D) VACV-Wyeth. (E,F) Carrier-injected control. Meninges (A,C,E—serial sections) and ventricles (B,D,F—serial sections) areas are shown.

Figure 3

Macrophages/Microglia and virus-infected cells are almost mutually exclusive. Serial sections of meninges, 5 d.p.i with VACV-Wyeth (10⁶ pfu infection dose; A,C,E) or VACV-WR (10² pfu infection dose; B,D,F). Brain sections were stained with H&E (A,B), antibodies to VACV (C,D) or antibodies to iba-1 antigen of macrophages / microglia cells (E,F); positive stain in brown. White box area enlarged (×100) on the right of each ×10 picture. Cell debris are marked with arrows in panel B (×100 enlargement).

Figure 4

Infection with VACV-WR results in BBB breakdown and elevated levels of MMP-9. Mice were infected i.c. with VACV-Lister (10⁶ pfu, n = 4), VACV-Wyeth (10⁶ pfu, n = 5) or of VACV-WR (10² pfu, n = 5). Perfused brains of VACV-Lister (left; 6 d.p.i), VACV-Wyeth (middle; 5 d.p.i.) and VACV-WR (right; 5 d.p.i.) following Evans-blue peripheral administration (two representative brains from each strain). Blue color represents vasculature dysfunctional leakage, a hallmark of BBB breakdown.

Figure 5

Quantification of up- and down-regulated genes following i.c. infection. Mice were infected with VACV-Wyeth and VACV-WR and total RNA-seq was done from brains 2, 4, and 5 d.p.i. Van diagrams representing number of genes that were significantly up- or down-regulated compared to the carrier control (p < 0.05, Fold change >2.0).

Figure 6

Clearance of VACV-Wyeth from infected brain is associated with robust induction of brain immunity pathways. Ingenuity pathway analysis (IPA) analysis based on brain gene expression of i.c. infected mice with 10⁶ pfu of VACV-Wyeth or 10² pfu of VACV-WR at 2, 4, and 5 d.p.i. (A) Antigen Presentation Pathway. (B) Th1-Th2 Activation Pathway. (C) T Helper Cell Differentiation Pathway. Dashed line marks p value of 0.05.

Figure 7

Verification of gene expression analysis. Expression of selected genes was measured by quantitative real-time RT-PCR on brain RNA isolated from mice infected i.c. with VACV-Wyeth (10⁶ pfu) or VACV-WR (10² pfu) 2 (A) 4 or 5 d.p.i (B). Vertical dashed line in panel B, marks the border between the set of genes that were expected to be higher in the VACV-Wyeth compared to the VACV-WR, based on the whole genome RNA seq (left to the line) and the genes that were expected to be higher in the VACV-WR compared to VACV-Wyeth (right to the line). (C) MMP-9 protein levels in sera of naive mice, VACV-Wyeth and VACV-WR infected mice at 2, 4, and 5 d.p.i. Asterisks denote p < 0.05 (t-test).