S1P is associated with protection in human and experimental cerebral malaria

Abstract

Cerebral malaria (CM) is associated with excessive inflammatory responses and endothelial activation. Sphingosine 1-phosphate (S1P) is a signaling sphingolipid implicated in regulating vascular integrity, inflammation and T-cell migration. We hypothesized that altered S1P signaling during malaria contributes to endothelial activation and inflammation, and show that plasma S1P levels were decreased in Ugandan children with CM compared with children with uncomplicated malaria. Using the Plasmodium berghei ANKA (PbA) model of experimental CM (ECM), we demonstrate that humanized S1P lyase (hS1PL)(-/-) mice with reduced S1P lyase activity (resulting in increased bio-available S1P) had improved survival compared with wild-type littermates. Prophylactic and therapeutic treatment of infected mice with compounds that modulate the S1P pathway and are in human trials for other conditions (FTY720 or LX2931) significantly improved survival in ECM. FTY720 treatment improved vascular integrity as indicated by reduced levels of soluble intercellular adhesion molecule (sICAM), increased angiopoietin 1 (Ang1) (regulator of endothelial quiescence) levels, and decreased Evans blue dye leakage into brain parenchyma. Furthermore, treatment with FTY720 decreased IFNγ levels in plasma as well as CD4(+) and CD8(+) T-cell infiltration into the brain. Finally, when administered during infection in combination with artesunate, FTY720 treatment resulted in increased survival to ECM. These findings implicate dysregulation of the S1P pathway in the pathogenesis of human and murine CM and suggest a novel therapeutic strategy to improve clinical outcome in severe malaria.

Figure 1

Plasma S1P is decreased in children with cerebral malaria. (A) S1P is decreased in children with cerebral malaria. Plasma S1P levels were assayed in Ugandan children with malaria. S1P levels were significantly lower in children with cerebral malaria (CM, n = 29) versus uncomplicated malaria (UM, n = 27), P < 0.0001, MW. (B) Hemoglobin levels are decreased in children with cerebral malaria. Hemoglobin levels were significantly lower in children with cerebral malaria (CM, n = 29) versus uncomplicated malaria (UM, n = 27), P < 0.0001, MW. (C) Platelet counts are decreased in children with cerebral malaria. Counts were measured in Ugandan children with malaria, and were shown to be significantly lower in children with cerebral malaria (CM, n = 29) versus uncomplicated malaria (UM, n = 27), P < 0.0001, MW.

Figure 2

Modulating S1P levels improves survival in mice with P. berghei ANKA-induced ECM. (A) Survival (left panel) is improved, while parasite burden remains unchanged (right panel), in mice with decreased S1PL activity in P. berghei ANKA-induced ECM. Survival of hS1PL^−/− mice infected with 1 × 10⁶ parasites i.p. was compared to their wild-type litter-mates. Decreased S1PL activity (resulting in increased bioavailable S1P) provided a significant survival advantage compared with control animals (two pooled independent experiments [IE], n = 10/group/IE, LRt, P < 0.0001). (B,C,D) Mice were infected with 1 × 10⁶ parasites i.p., and treated daily by gavage with FTY720 (0.3 mg/kg). Treatment with FTY720 provided a significant survival advantage if administered (B) 1 d prior to (two pooled IE, n ≥ 10/group/IE, LRt, P < 0.0001), (C) 1 d p.i. (three pooled IE, n ≥ 10/group/IE, LRt, P < 0.05) or (D) 5 d p.i. in combination with a single dose of artesunate administered on d 5 of infection (five pooled IE, n ≥ 10/group/IE, FEt, P = 0.05). Parasite burden remain unchanged between groups for all these experiments (right panels).

Figure 3

Treatment of ECM with FTY720 decreases endothelium activation and improves blood-brain barrier integrity. (A) Plasma levels of sICAM-1, a marker of en-dothelial activation, are decreased in FTY720-treated animals (P < 0.001, KW): for (FTY720 [−1], P < 0.001, FTY720 [+1] P < 0.05 and FTY720 [+3], P < 0.001; DMC, three pooled IE, n ≥ 3/group/IE). (B) Plasma levels of Ang1, a marker of endothelial integrity, are increased in animals treated with FTY720 prophylactically (P < 0.01, KW and DMC, two pooled IE, n ≥ 3/group/IE) but not in animals treated during infection. (C) Brains stained with Evans blue were harvested on day 7 p.i. Brains displayed are from animals with the following parasitemia: Wild type: 12.7%; FTY720 (−1): 20.6%; FTY720 (+1): 17.8%; FTY720 (+3): 11.9%. Brains from FTY720-treated animals show improved endothelium integrity (less blue dye leakage from the blood into brain parenchyma) as compared with the wild-type animal (two pooled IE, n ≥ 3/group/IE).

Figure 4

FTY720 treatment of ECM decreases plasma IFNγ and TNF. (A) TNF plasma levels trended toward a decrease in the plasma of all animals treated with FTY720 (P > 0.05, ANOVA, three pooled IE, n ≥ 5/group/IE). (B) IFNγ plasma levels were decreased in the plasma of animals treated with FTY720 1 d before infection and 3 d p.i. (P < 0.01 and P < 0.5 respectively, KW and DMC, three pooled IE, n ≥ 5/group/IE).

Figure 5

FTY720 treatment of ECM decreases lymphocyte numbers and granzyme A mRNA in the brain. (A) Relative CD8⁺ and (B) CD4⁺ T-cell numbers isolated from the brains of mice infected for 6 d. Numbers were decreased in animals treated with FTY720 1 d before infection and 1 d p.i. (CD4⁺: P < 0.5 and P < 0.001, CD8⁺: P < 0.5 and P < 0.01 respectively, KW and DMC, two pooled IE, n ≥ 4/group/IE) (C) GZMA mRNA levels were decreased in the brains of FTY720-treated animals compared with control animals: for (FTY720 [−1], P < 0.05, FTY720 [+1] P > 0.05 and FTY720 [+3], P < 0.01; KW and DMC, two pooled IE, n ≥ 3/group/IE).