Teaching a new mouse old tricks: Humanized mice as an infection model for Variola virus

Abstract

Smallpox, caused by the solely human pathogen Variola virus (VARV), was declared eradicated in 1980. While known VARV stocks are secure, smallpox remains a bioterrorist threat agent. Recent U.S. Food and Drug Administration approval of the first smallpox anti-viral (tecovirimat) therapeutic was a successful step forward in smallpox preparedness; however, orthopoxviruses can become resistant to treatment, suggesting a multi-therapeutic approach is necessary. Animal models are required for testing medical countermeasures (MCMs) and ideally MCMs are tested directly against the pathogen of interest. Since VARV only infects humans, a representative animal model for testing therapeutics directly against VARV remains a challenge. Here we show that three different humanized mice strains are highly susceptible to VARV infection, establishing the first small animal model using VARV. In comparison, the non-humanized, immunosuppressed background mouse was not susceptible to systemic VARV infection. Following an intranasal VARV challenge that mimics the natural route for human smallpox transmission, the virus spread systemically within the humanized mouse before mortality (~ 13 days post infection), similar to the time from exposure to symptom onset for ordinary human smallpox. Our identification of a permissive/representative VARV animal model can facilitate testing of MCMs in a manner consistent with their intended use.

Fig 1. Study design, survival table and survivorship curves for mice and dosage groups.

A schematic overview of the study design (A). The number of the three types of humanized mice per group based on days post infection and organized by dose group and mouse strain (B). If a sample day is not shown, there was no change in animal number per group. If a mouse was found deceased in-between PM and AM checks, the animal number was decreased the day the animal was found to have succumbed. Survival of three types of humanized mice (hu-BLT, hu-CD34+, and hu-PBMC) were assessed for low dose and high dose independently (C). *Hu-PBMC mice were significantly more likely to survive variola infection compared to hu-BLT and hu-CD34+ mice in both dose groups (high dose p = 0.004, low dose p = 0.008). There were no significant differences between hu-BLT and hu-CD34+ mice, when controlling for dose.

Fig 2. Representative histopathology and immunohistochemistry of VARV infection in humanized mice.

A, B: Chronic skin ulceration over the hock, with epidermal hyperplasia (arrow, A) and VARV immunostaining in skin, tendon, and periosteum (hu-PBMC-11; low dose, day of death 21 dpi). C, D: Multifocal adrenal gland necrosis (*, C) with VARV immunostaining (hu-BLT-5; high dose, day of death 13 dpi). E, F: Diffuse splenic necrosis with lymphoid depletion, fibrin, and hemorrhage; extensive VARV immunostaining in reticuloendothelial and mesenchymal cells (hu-CD34+-4; high dose, day of death 16 dpi). G, H: Nasal mucosa with submucosal edema and mild inflammation (open arrows, G); extensive VARV immunostaining (hu-PBMC-6; high dose, day of death 21 dpi). I, J: Tooth with dental pulp necrosis (*, I); VARV immunostaining of pulp and periodontal ligament (hu-CD34+-9; low dose, day of death 20 dpi). K, L: Ovary with multifocal stromal and follicular necrosis (*, K) and VARV immunostaining (hu-PBMC-10; low dose, day of death 21 dpi). M, N: Uterus with transmural necrosis and VARV immunostaining (PBMC-5; high dose, day of death 13 dpi). O, P: Focal necrosis (*, O) in renal subcapsular human fetal thymic graft, with extensive VARV immunostaining (hu-BLT-5; high dose,). A, C, E, G, I, K, M, O (hematoxylin-eosin); B, D, F, H, J, L, N, P (VARV immunohistochemistry, viral antigen labeling in red). Original magnifications: A, B, O, P (x50); C-J, M, N (x100); K, L (x200).

Fig 3. Representative differences in pathologic findings among three strains of humanized mice with VARV infection.

Hu-CD34+ and hu-BLT mice had similar findings in liver, bone marrow, and lung, which contrasted those seen in hu-PBMC mice. Hu-CD34+ and hu-BLT livers (A, G) had confluent and lobular hepatocyte necrosis with very rare eosinophilic globular inclusions (A, inset), and VARV immunostaining of hepatocytes, Kupffer cells, and occasional endothelial cells (B, H). Hu-CD34+ and hu-BLT bone marrow specimens similarly showed extensive necrosis and hemorrhage associated with VARV immunostaining (C, D, I, J). Hu-PBMC liver (M) and bone marrow (O) showed minimal inflammation (open arrows, M) and no necrosis, and only very rare VARV immunostaining in these tissues (N, P). Conversely, hu-CD34+ and hu-BLT lung tissues showed minimal inflammation (E, K) and VARV immunostaining (F, L), while hu-PBMC lungs (Q) had more prominent peribronchiolar and perivascular inflammation (arrowheads, Q) with VARV immunostaining (R). Hu-CD34+ and hu-BLT mice had disseminated intravascular bacteria (arrows in A, E, G, K), which were not present in hu-PBMC mice. A, B, E, F (hu-CD34⁺-4, high dose, day of death 16 dpi); C, D (hu-CD34⁺-6, high dose, day of death 17 dpi); G, H (hu-BLT-5, high dose, day of death 13 dpi); I, J (hu-BLT-9, low dose, day of death 17 dpi); K,L (hu-BLT-4, high dose, day of death 13 dpi); M, N, Q, R (hu-PBMC-11, low dose, day of death 21 dpi); O, P (hu-PBMC-5, high dose, day of death 19 dpi). A, C, E, G, I, K, M, O, Q (hematoxylin-eosin); B, D, F, H, J, L, N, P, R (VARV immunohistochemistry, viral antigen labeling in red). Original magnifications: A, B, C, D, G, H, I, J, O, P (x100); E, F, K, L, Q, R (x200); M, N (X400).

Fig 4. Electron microscopic images of VARV in a humanized mouse.

Subset of select tissues from VARV challenged humanized-mice (hu-BLT, hu-CD34+ and hu-PBMC) were examined by transmission electron microscopy. (A) Hepatocyte with multiple spherical, diffuse immature particles (arrowhead) and a single condensed mature particle (arrow). Bar, 500 nm. (B) Sinusoidal endothelial cell in the liver with immature (arrowhead) and mature (arrow) VARV particles. RBC, red blood cell. Bar, 1 um. (C) Human fetal thymic allograft containing mature virions with a characteristic dumbbell-shaped nucleoid (arrow). (BLT-8, low dose, day of death 13 dpi). Bar, 100 nm.

Fig 5. High viable virus loads were detected in all three types of humanized mice.

Humanized-mice (hu-BLT (A), hu-CD34+ (B) and hu-PBMC (C)) were inoculated via the intranasal route with 7x10³ or 7x10⁵ plaque forming units (pfu) of VARV (VARV_JAP51_hrpr (primary clade I)) (n = 4 per group). Animals that succumbed to VARV infection, or were euthanized at 21 dpi, had tissues collected and processed for viral titration (plaque assay). The mean with SEM is shown. An * indicates one or more of that sample had cell culture monolayer destroyed or plaques were present but below the limit of detection for this assay.