Protective role of brain water channel AQP4 in murine cerebral malaria

Abstract

Tragically common among children in sub-Saharan Africa, cerebral malaria is characterized by rapid progression to coma and death. In this study, we used a model of cerebral malaria appearing in C57BL/6 WT mice after infection with the rodent malaria parasite Plasmodium berghei ANKA. Expression and cellular localization of the brain water channel aquaporin-4 (AQP4) was investigated during the neurological syndrome. Semiquantitative real-time PCR comparing uninfected and infected mice showed a reduction of brain AQP4 transcript in cerebral malaria, and immunoblots revealed reduction of brain AQP4 protein. Reduction of brain AQP4 protein was confirmed in cerebral malaria by quantitative immunogold EM; however, polarized distribution of AQP4 at the perivascular and subpial astrocyte membranes was not altered. To further examine the role of AQP4 in cerebral malaria, WT mice and littermates genetically deficient in AQP4 were infected with P. berghei. Upon development of cerebral malaria, WT and AQP4-null mice exhibited similar increases in width of perivascular astroglial end-feet in brain. Nevertheless, the AQP4-null mice exhibited more severe signs of cerebral malaria with greater brain edema, although disruption of the blood-brain barrier was similar in both groups. In longitudinal studies, cerebral malaria appeared nearly 1 d earlier in the AQP4-null mice, and reduced survival was noted when chloroquine rescue was attempted. We conclude that the water channel AQP4 confers partial protection against cerebral malaria.

Fig. 1.

Brain edema in murine cerebral malaria. Uninfected C57BL/6 control mice and littermates infected with 10⁶ P. berghei parasites were evaluated 7 to 8 d postinfection when signs of experimental cerebral malaria appeared. (A) Brain water content, assessed from the wet-to-dry-weight ratio, was 2.07% higher in mice with cerebral malaria than in uninfected controls. Values are mean ± SEM (*P ≤ 0.05, Student t test.) (B) Representative immunoblot shows induction of ICAM-1 expression in brains of mice with cerebral malaria. (C) Representative immunofluorescence microscopy shows increased expression of GFAP in brains of mice with cerebral malaria. (Scale bar: 20 µm.)

Fig. 2.

Reduced brain AQP4 mRNA and protein in murine cerebral malaria. Uninfected control mice and littermates with cerebral malaria were killed 7 to 8 d postinfection. (A) Semiquantitative real-time PCR of AQP4 mRNA in cerebral malaria. Analysis showed that brain AQP4 mRNA was reduced by 57% in murine cerebral malaria. Values are mean ± SD. (B) Anti-AQP4 immunoblot of brain membranes revealed 32 to 34 kDa AQP4 splice variants and 64 kDa AQP4 dimers. (C) Densitometric analysis showed brain AQP4 protein was decreased by 20% in murine cerebral malaria. Values are mean ± SEM (*P ≤ 0.05, Student t test). Data are representative of three independent experiments.

Fig. 3.

Anti-AQP4 immunogold EM of murine cerebral malaria. (A and B) Ultrathin Lowicryl sections of perfused fixed brains from uninfected control mice or mice with cerebral malaria killed 7 to 8 d postinfection. Anti-AQP4 immunogold labeling of astrocyte perivascular end-feet in membranes facing endothelial basal lamina (arrows) or membranes facing neuropil (arrowheads). Single asterisk marks basal lamina between the pericyte and astrocyte end-foot; double asterisk marks basal lamina between the endothelial cell and pericyte. E, endothelial cell; L, capillary lumen; P, pericyte. (Scale bar: 500 nm.)

Fig. 4.

Quantification of AQP4 immunolabeling in uninfected control mice (open bars) and mice with cerebral malaria 7 to 8 d postinfection (closed bars). The values along the ordinate represent the number of immunogold particles per micrometer in the perivascular (Left) and subpial (Right) membrane domains of the astroglial end-feet. Values are mean ± SE (*P ≤ 0.05, **P ≤ 0.01, Student t test).

Fig. 5.

EM image of AQP4-null brain in murine cerebral malaria. (A and B) Ultrathin Lowicryl sections of brains from uninfected AQP4-null mice or AQP4-null mice with cerebral malaria killed 6 d postinfection. The sections were incubated together with the WT sections shown in Fig. 3. As expected, no AQP4 immunogold labeling is observed in the AQP4-null mice. Astrocyte end-feet are identified in membranes facing endothelial basal lamina (arrows) or membranes facing neuropil (arrowheads). Asterisk indicates basal lamina between the endothelium and astrocyte endfoot. E, endothelial cell; L, capillary lumen. (Scale bar: 500 nm.)

Fig. 6.

Quantification of astrocyte end-feet width in WT and AQP4-null mice without (open bars) and with cerebral malaria 6 d postinfection (closed bars). Values are mean ± SE (*P ≤ 0.01, ANOVA).

Fig. 7.

Severity of cerebral malaria in AQP4-null mice. WT and AQP4-null mice were infected with 10⁶ P. berghei parasites and their clinical scores compared on day 6 postinfection. Loss of body weight, reduced body temperature, appearance, and behavior were graded. Greater disease severity is associated with higher scores. AQP4-null mice exhibited higher clinical scores for appearance and behavior and overall clinical scores compared with WT mice. Values are mean ± SEM (*P ≤ 0.05, Student t test). Data are representative of four independent experiments.

Fig. 8.

Brain edema in AQP4-null mice with cerebral malaria. WT and AQP4-null mice were infected with 10⁶ P. berghei parasites and evaluated on day 6 postinfection. (A) Brain water content was similar in WT and AQP4-null mice in basal conditions and increased during cerebral malaria with greater increase in the AQP4-null mice compared with WT mice. (B) Blood–brain barrier disruption assessed by Evans blue dye leakage from the brain microvasculature was similar in both groups during cerebral malaria. Values are mean ± SEM for each group and are representative of three independent experiments (*P ≤ 0.05, Student t test).

Fig. 9.

Accelerated mortality from cerebral malaria in AQP4-null mice. WT and AQP4-null mice infected with 10⁶ P. berghei parasites were monitored for survival. (A) Earlier deaths were seen in the AQP4-null mice compared with WT mice during the progression of cerebral malaria. Survival data are pooled from four independent experiments (P ≤ 0.01; n = 35 WT mice; n = 37 AQP4-null mice). (B) Chloroquine was administered to mice twice per day beginning on the morning of day 6. By day 15, only 20% of AQP4-null mice were rescued from cerebral malaria vs. 70% of the WT mice. Data are survival probabilities and SEs as determined by Kaplan–Meier analysis (P ≤ 0.05; n = 10 WT mice; n = 10 AQP4-null mice).