Inflammation associated with monocyte/macrophage activation and recruitment corresponds with lethal outcome in a mouse model of Crimean-Congo haemorrhagic fever1

Abstract

Crimean-Congo haemorrhagic fever virus (CCHFV) causes human disease ranging from subclinical to a fatal haemorrhagic syndrome. Determinants of CCHF pathogenesis are largely unknown and animal models that recapitulate human disease are limited. A recently described mouse model uses a monoclonal antibody (mAb 5A3) targeting the interferon (IFN) alpha/beta receptor to suppress type I IFN responses, making animals transiently susceptible to infection. To advance utility of this model, we investigated effects of challenge route, timing of 5A3 delivery, mouse sex and age, and virus strain on clinical course and outcome. C57BL/6J mice received mAb 5A3 -1, 0, or -1/+1 days post-infection (dpi). Subsets were challenged with CCHFV strain Turkey04 or IbAr10200 subcutaneously or intraperitoneally, and serially euthanized 3- and 7-dpi, when meeting euthanasia criteria or at study completion (14 dpi). CCHFV-IbAr10200-infected mice almost uniformly succumbed to infection, whereas CCHFV-Turkey04-infected mice transiently lost weight but survived. These results were consistent regardless of mAb timing or route of challenge. Viral replication and dissemination were comparable between the two strains at 3 dpi. However, in the plasma and livers of non-survivors, expression of proinflammatory cytokines/chemokines that correspond with macrophage activation and recruitment were significantly elevated. Lethal disease was also associated with elevated levels of macrophage activation marker CD163 in plasma. Further, mouse macrophages were more permissive to IbAr1200 infection in vitro, suggesting tropism for these cells may influence pathogenesis. Our data suggest that early inflammation may be a critical determinant of CCHF outcome and therapeutics to control inflammation may be worthwhile targets for future investigation.

Keywords: CCHF; Crimean-Congo haemorrhagic fever virus; Inflammation; Macrophage activation; Mice; Tropism.

Figure 1.

Lethal outcome in transiently immunosuppressed mice is dependent on CCHFV strain. In a pilot study, groups of C57BL/6J mice (4/group, mixed male [animal # 1–2] and female [animal # 3–4], 6 weeks of age) were infected either (A) subcutaneously (SC) or (B) intraperitoneally (IP) with 100 TCID₅₀ of CCHFV strain IbAr10200 (red) or Turkey04 (blue). Mice received anti-IFNAR1 monoclonal antibody (mAb) 5A3 via IP injection on the following days post infection (dpi): −1/+1 (2.0 mg/0.5 mg), 0 (2.5 mg), or +1 (2.5 mg). Control animals (4/group; mixed male and female) were included for each experimental group (each challenge route [SC and IP] and mAb timing group [−1/+1, 0, +1 dpi]) and received an isotype control (IgG1) mAb rather than 5A3 (orange). Control animals were challenged with Turkey04. After challenge, weight loss (% change from baseline) and clinical scores (see Methods) were assessed daily for each animal until reaching euthanasia criteria (see Methods) or study completion (14 dpi). Individual dots and connecting lines represent the daily mean and error bars represent the standard deviation (SD). (C–D) Subsequent studies were performed using 0 dpi as the designated timing for 5A3 delivery (mixed male and female C57BL/6J mice, ages ranging from 6 to 83 weeks), resulting in high rates of lethality using both (C) SC and (D) IP routes of challenge. (E) Age-stratified cohorts of C57BL/6J mice at 6 (n = 12), 10 (n = 12), 16–19 (n = 14), 25–29 (n = 14), 32–36 (n = 17), 45–47 (n = 13), 55 (n = 6), 62 (n = 11), or 83 (n = 10) weeks of age were challenged (SC) with 100 TCID₅₀ CCHFV IbAr10200 and 2.5 mg mAb 5A3 (0 dpi) and followed until meeting euthanasia criteria.

Figure 2.

Divergent clinical outcomes are not the result of differences in viral load. Tissues including liver, spleen, ovary/testis (gonad), kidney, lung, heart, eye, brain, and whole blood, as well as oral and rectal swabs, were collected from all mice at the time of euthanasia at 3, 7, or 14 dpi (Turkey04) or 3 dpi and terminal timepoint (5–6 dpi; IbAr10200). Isotype control animals were challenged with Turkey04. (A) Viral RNA (vRNA) was isolated and quantified via RT-qPCR using primers/probes specific for the CCHFV S gene segment. Bars indicate the mean and error bars the standard deviation (SD). Statistics were calculated using two-way ANOVA with Tukey’s multiple comparison test; **** p < 0.0001. Only significant results are reported. (B) Immunohistochemistry (IHC) for CCHFV was performed using anti-CCHFV NP antibody with Fast Red chromogen. CCHFV antigen (red) amount and distribution is similar in liver and spleen tissues from animals infected with Turkey04 and IbAR10200 at 3 dpi but increases in terminal IbAr10200-infected tissues while decreasing in Turkey04-infected tissues at 7 dpi, and becoming undetectable by 14 dpi. Original magnifications: ×40 (liver); ×20 (spleen).

Figure 3.

Divergent clinical outcomes are not the result of differences in tissue tropism. Tissues including brain, eye, heart, lung, kidney, adrenal gland, pancreas, and lymph node were collected from all mice at the time of euthanasia at 3, 7, or 14 dpi (Turkey04) or 3 dpi and terminal timepoint (5–6 dpi; IbAr10200). Immunohistochemistry (IHC) for CCHFV was performed using anti-CCHFV NP antibody with Fast Red chromogen within the brain, eye, heart, lung, kidney, adrenal gland, pancreas, and lymph node. CCHFV antigen (red) amount and tissue distribution are similar in animals infected with Turkey04 and IbAR10200. CCHFV antigen (red) is seen primarily within endothelial cells, intravascular leukocytes, and tissue macrophages, with either inoculation route and at both timepoints. Original magnifications: ×200.

Figure 4.

Lethal outcome with IbaR10200 is associated with progressive liver and spleen pathology. (A) Haematoxylin-eosin-stained liver and spleen from mice that were inoculated with mAb 5A3 and CCHFV SC or IP on day 0 and sacrificed 3, 7, or 14 dpi (Turkey04), or 3 dpi and at terminal timepoints (5–6 dpi; IbAr10200). Livers from both Turkey04 and IbAr10200-inoculated mice showed similar findings, with moderate hepatocellular necrosis (arrowheads) and neutrophilic inflammation at 3 dpi. At 7 dpi, livers from Turkey04-infected survivors showed decreased necrosis and increased inflammation (arrows), with addition of mononuclear infiltrates, and by 14 dpi, livers were normal. In contrast, at terminal timepoints, IbAr10200-infected livers showed continued progression and severity of changes. Spleens showed mild to moderate reactivity, with expansion of plasma cells and few macrophages and neutrophils in red pulp at 3 dpi in Turkey04-infected survivors. Lymphocyte necrosis/apoptosis was mild and more prominent in IP-inoculated animals. At 7 dpi, spleens from Turkey04-infected survivors showed marked lymphoid reactivity, with abundant plasma cells without lymphocyte necrosis/apoptosis. By 14 dpi, reactivity was present, but with fewer plasma cells compared to 7 dpi. For IbAr10200-inoculated animals, 3 dpi spleens were similar to spleens of Turkey04-infected animals at the same timepoint, but with more severe lymphoid reactivity and more prominent lymphocyte apoptosis/necrosis, especially in IP-inoculated animals. In terminal spleens of IbAr10200-infected animals, these changes were sustained, with more pronounced lymphoid necrosis/apoptosis also in SC-inoculated animals. Original magnifications: ×20 (spleen); ×40 (liver). (B) Liver enzymes (ALT, alanine aminotransferase; AST, aspartate aminotransferase) from each animal were assessed using whole blood (lithium heparin) at 3 (IbAr10200 and Turkey04), 7, or 14 (Turkey04 only) dpi. Bars indicate the mean and error bars the standard deviation (SD). Statistics were calculated using multiple t-tests (Mann–Whitney); no significant differences between IbAr10200 and Turkey04 were found at 3 dpi.

Figure 5.

Inflammation associated with monocyte/macrophage recruitment and activation corresponds with lethal outcome. (A) Cytokine/chemokine responses in all mice euthanized 3 dpi were analysed using the ProcartaPlex Mouse Th1/Th2 Cytokine and Chemokine panel and 25 µL mouse plasma. Statistics compare IbAr10200- to Turkey04-infected animals. (B) Plasma CD163 levels were measured in all animals euthanized 3 dpi as well as from (n = 4) naive age-matched C57BL/6J mice to determine baseline levels in healthy mice. (C) Gene expression profiles in liver and spleen of all mice euthanized 3 dpi were analysed via RT-qPCR and are reported as fold change compared to tissue from age-matched uninfected C57BL/6J control animals. Isotype control animals were infected with CCHFV Turkey04. Bars indicate the mean and error bars the standard deviation (SD). Statistics were calculated using multiple t-tests (Mann–Whitney); ** p < 0.01; *** p < 0.001; **** p < 0.0001. Only significant results are reported. Iso, isotype control; Tur, CCHFV Turkey04; 10200, CCHFV IbAr10200; NMP, normal mouse plasma.

Figure 6.

Mouse macrophages are more permissive to IbAr10200 infection in vitro. CCHFV IbAr10200 and Turkey04 viruses were titered via immunofluorescent TCID₅₀ assays in BSRT7/5 (baby hamster kidney cell stably expressing T7 polymerase), PMJ2-R (C57BL/6J peritoneal macrophage), and THP-1 (human monocyte) cells. Prior to titering, THP-1 cells were differentiated into macrophages by incubating with 5 ng/mL PMA for 24 h. Titers were also determined on polarized PMJ2-R and THP-1 cells (classical, M1 phenotype) by incubating with 250 ng/mL lipopolysaccharide (LPS) and 20 ng/mL mouse or human IFN-gamma for 48 h. Bars indicate the mean and error bars the standard deviation (SD). Statistics were calculated using multiple t-tests (Mann–Whitney); * p < 0.05; only significant results are reported.