Niemann-pick C1 is essential for ebolavirus replication and pathogenesis in vivo

Abstract

Recent work demonstrated that the Niemann-Pick C1 (NPC1) protein is an essential entry receptor for filoviruses. While previous studies focused on filovirus entry requirements of NPC1 in vitro, its roles in filovirus replication and pathogenesis in vivo remain unclear. Here, we evaluated the importance of NPC1, and its partner in cholesterol transport, NPC2, by using a mouse model of Ebolavirus (EBOV) disease. We found that, whereas wild-type mice had high viral loads and succumbed to EBOV infection, Npc1(-/-) mice were entirely free of viral replication and completely protected from EBOV disease. Interestingly, Npc1(+/-) mice transiently developed high levels of viremia, but were nevertheless substantially protected from EBOV challenge. We also found Npc2(-/-) mice to be fully susceptible to EBOV infection, while Npc1(-/-) mice treated to deplete stored lysosomal cholesterol remained completely resistant to EBOV infection. These results provide mechanistic evidence that NPC1 is directly required for EBOV infection in vivo, with little or no role for NPC1/NPC2-dependent cholesterol transport. Finally, we assessed the in vivo antiviral efficacies of three compounds known to inhibit NPC1 function or NPC1-glycoprotein binding in vitro. Two compounds reduced viral titers in vivo and provided a modest, albeit not statistically significant, degree of protection. Taken together, our results show that NPC1 is critical for replication and pathogenesis in animals and is a bona fide target for development of antifilovirus therapeutics. Additionally, our findings with Npc1(+/-) mice raise the possibility that individuals heterozygous for NPC1 may have a survival advantage in the face of EBOV infection.

Importance: Researchers have been searching for an essential filovirus receptor for decades, and numerous candidate receptors have been proposed. However, none of the proposed candidate receptors has proven essential in all in vitro scenarios, nor have they proven essential when evaluated using animal models. In this report, we provide the first example of a knockout mouse that is completely refractory to EBOV infection, replication, and disease. The findings detailed here provide the first critical in vivo data illustrating the absolute requirement of NPC1 for filovirus infection in mice. Our work establishes NPC1 as a legitimate target for the development of anti-EBOV therapeutics. However, the limited success of available NPC1 inhibitors to protect mice from EBOV challenge highlights the need for new molecules or approaches to target NPC1 in vivo.

FIG 1

NPC1 knockout mice are protected from mouse-adapted EBOV infection. Npc1^−/− (n = 28), Npc1^+/− (n = 21), and wild-type littermate control (n = 19) mice were challenged i.p. with 100 PFU of mouse-adapted EBOV. Mice were monitored for weight loss (A) and mortality (B) following challenge. On days 3 (C) and 7 (D) postchallenge, subsets of mice from each group were euthanized and viral titers were determined for the indicated tissues by plaque assay. The dotted line indicates the assay limit of detection. ****, P ≤ 0.0001 compared to wild-type mice.

FIG 2

Serum cytokine levels following mouse-adapted EBOV infection of wild-type and NPC1 knockout mice. Npc1^−/−, Npc1^+/−, and wild-type littermate control mice were challenged i.p. with 100 PFU of mouse-adapted EBOV. Serum collected at day 3 (A and C) or 7 (B and D) postchallenge was used to assess cytokine levels in a BioPlex assay. °, P ≤ 0.05; °°, P ≤ 0.01; °°°, P ≤ 0.001; °°°°, P ≤ 0.0001 compared to uninfected control mice. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001 compared to wild-type mice.

FIG 3

Cholesterol accumulation does not confer resistance to mouse-adapted EBOV infection. Npc2^−/− (n = 9), Npc2^+/− (n = 11), and wild-type littermate control (n = 10) mice were challenged i.p. with 100 PFU of mouse-adapted EBOV. Mice were monitored for weight loss (A) and mortality (B) following challenge. Npc1^−/− (n = 8), Npc1^+/− (n = 11), and wild-type littermate control (n = 11) mice were treated or not with HPBCD starting at 7 days of age and continuing every other day until sacrifice at 7 weeks of age (33 to 36 days old). (C) Filipin staining of unesterified cholesterol was evaluated in the dorsal neocortex of HPBCD-treated and control mice. Cholesterol accumulation is seen as white areas, primarily cytoplasmic, examples of which are denoted by red arrows. Images were taken at 20×. Bar, 50 µm. 2-Hydroxypropyl-β-cyclodextrin-treated mice were challenged i.p. with 100 PFU of mouse-adapted EBOV. Mice were monitored for weight loss (D) and mortality (E) following challenge. ***, P ≤ 0.001; ****, P ≤ 0.0001 compared to wild-type mice.

FIG 4

Imipramine and compound 3.47 inhibit filovirus replication and GP-mediated entry in vitro. Filovirus infected HUVEC and Vero cells were treated with or without imipramine (A) or compound 3.47 (B), respectively. Cells were fixed at 48 or 72 h postinfection, and infected cells were enumerated using MARV-, EBOV-, or SUDV-specific monoclonal antibodies and a fluorescently labeled secondary antibody. (C) Vero cells were treated with compound 3.47 at the indicated concentrations prior to infection with GFP-expressing VSV pseudoviruses. Cells were fixed 16 to 24 h postinfection, and infected cells were assessed by eGFP expression analysis. *, P ≤ 0.05; ***, P ≤ 0.001.

FIG 5

Imipramine reduces EBOV replication in vivo. Wild-type mice were challenged i.p. with 100 PFU of mouse-adapted EBOV and treated either sid (n = 39) or qod (n = 30) with 20 mg/kg of imipramine. Control mice (n = 40) received an equivalent volume of vehicle control sid. Mice were monitored for weight loss (A) and mortality (B) following challenge. On days 3 (C) and 5 (D) postchallenge, subsets of mice from control and qod groups were euthanized, and viral titers were determined for the indicated tissues by plaque assay. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001.

FIG 6

U18666a reduces EBOV replication in vivo. Wild-type mice were challenged i.p. with 100 PFU of mouse-adapted EBOV and treated sid with either 2 mg/kg of U18666a (n = 39) or an equivalent volume of vehicle control (n = 38). Mice were monitored for weight loss (A) and mortality (B) following challenge. On days 3 (C) and 5 (D) postchallenge, subsets of mice from each group were euthanized and viral titers were determined for indicated tissues by plaque assay. *, P ≤ 0.05; **, P ≤ 0.01.