Neuroprotective interventions targeting detrimental host immune responses protect mice from fatal alphavirus encephalitis

Abstract

Systemic treatment with the tetracycline derivative, minocycline, attenuates neurologic deficits in animal models of amyotrophic lateral sclerosis, hypoxic-ischemic brain injury, and multiple sclerosis. Inhibition of microglial activation within the CNS is 1 mechanism proposed to underlie the beneficial effects of the drug in these systems. Given the widening scope of acute viral encephalitis caused by mosquito-borne pathogens, we investigated the therapeutic effects of minocycline in a murine model of fatal alphavirus encephalomyelitis in which widespread microglial activation is known to occur. We found that minocycline conferred significant protection against both paralysis and death, even when started after viral challenge and despite having no effect on CNS virus replication or spread. Further studies demonstrated that minocycline inhibited early virus-induced microglial activation and that diminished CNS production of the inflammatory mediator, interleukin (IL)-1beta, contributed to its protective effect. Therapeutic blockade of IL-1 receptors also conferred significant protection in our model, validating the importance of the IL-1 pathway in disease pathogenesis. We propose that interventions targeting detrimental host immune responses arising from activated microglia may be of benefit in humans with acute viral encephalitis caused by related mosquito-borne pathogens. Such treatments could conceivably act through neuroprotective rather than antiviral mechanisms to generate these clinical effects.

Figure 1

Systemic minocycline treatment protects mice from otherwise lethal NSV encephalomyelitis in a measurable therapeutic window. (A, B) Intraperitoneal minocycline (25 mg/kg/every 12 hours, n=20 per group) started at the time of viral challenge potently protected animals from both paralysis and death compared to parallel mice treated with a saline vehicle control. (C) This protective minocycline regimen exerted partial clinical protection even when the start of therapy was delayed up to 72 hours after the onset of infection (n=20 per group). The significance of survival differences between the various treatment groups and untreated controls in (C) was calculated by Kaplan-Meier analysis.

Figure 2

Treatment of mice with a protective dose of minocycline (25 mg/kg/every 12 hours) has no effect on the replication, spread, or clearance of NSV from the CNS of infected animals. (A) Identical viral titers were found in the brains of NSV-infected mice over time without or with minocycline treatment (n=4 animals at each time point). (B) Similarly, the amount of infectious virus present in the spinal cords of NSV-infected mice showed no differences between the two treatment groups (n=4 animals at each time point).

Figure 3

The clinical protection conferred by minocycline in NSV-infected mice is associated with reduced destruction of uninfected neurons in regions of both the brain and spinal cord where many cells become infected. (A, B) Minocycline had no effect on the numbers of infected neurons identified in either the hippocampus or the lumbar spinal cord (n=3 animals at each time point). (C, D) Drug treatment, however, reduced the loss of uninfected neurons at both of these sites (n=3 animals at each time point) (*p<0.05). (E) The same therapeutic regimen did not alter the number of TUNEL-positive hippocampal neurons over time (n=3 animals at each time point).

Figure 4

Minocycline treatment reduces the number of CD45+ microglia and lymphocytes found in the spinal cords of mice with NSV encephalomyelitis, albeit with different kinetics. It also inhibits microglial expression of tomato lectin binding epitopes, another indicator of activation (20). (A) Fewer CD45+ microglia (identified by their morphology, see insert) were found in tissue after 2 days of drug treatment, and their numbers returned back to baseline by day 4 of infection (n=3 animals at each time point) (*p<0.05). (B) Conversely, CD45+ lymphocytes (also identified by their staining pattern, see insert) infiltrated the spinal cords of vehicle- and minocycline-treated mice over the first 4 days of infection to the same degree. By day 6, however, the total number of lymphocytes was reduced in tissue from drug-treated animals compared to vehicle-treated controls (n=3 animals at each time point) (*p<0.05). (C) Tomato lectin immunostaining (see insert) confirmed an inhibitory effect of minocycline on microglial cell activation (n=3 animals at each time point). Insert magnifications: (A) 125x; (B) 125x; (C) 125x.

Figure 5

Minocycline has no effect on the reactivity of astrocytes in the brain following NSV infection. (A, B) Extensive GFAP immunoreactivity is seen in the hippocampus of animals on day 6 of NSV infection; similar numbers and intensity of positive cells are seen either without (A) or with (B) daily minocycline treatment (100x).

Figure 6

The i.n. administration of recombinant IL-1β overcomes much of the protection conferred by minocycline against NSV infection. (A) Exogenous IL-1β (10 μg i.n./animal/day, n=20 animals) started at the time of infection had little effect on the survival of otherwise untreated mice (n=24 animals) infected with NSV. (B) This same i.n. IL-1β dosing regimen, however, significantly reversed the clinical protection conferred by systemic minocycline treatment. The significance of survival differences between groups was calculated by Kaplan-Meier analysis.

Figure 7

The exogenous administration of a specific IL-1 receptor antagonist (IL-1ra) protects mice from otherwise lethal NSV encephalomyelitis. Parallel groups of mice were treated with IL-1ra (10 μg i.n./animal/day, n=6 animals) or a saline vehicle control (n=6 animals) starting at the time of viral challenge. (A) Hind limb paralysis and (B) Survival was monitored on a daily basis over a 14-day study interval, and the significance of paralysis and survival differences between the two groups was calculated by Kaplan-Meier analysis.