Cerebral malaria: gamma-interferon redux

Abstract

There are two theories that seek to explain the pathogenesis of cerebral malaria, the mechanical obstruction hypothesis and the immunopathology hypothesis. Evidence consistent with both ideas has accumulated from studies of the human disease and experimental models. Thus, some combination of these concepts seems necessary to explain the very complex pattern of changes seen in cerebral malaria. The interactions between malaria parasites, erythrocytes, the cerebral microvascular endothelium, brain parenchymal cells, platelets and microparticles need to be considered. One factor that seems able to knit together much of this complexity is the cytokine interferon-gamma (IFN-γ). In this review we consider findings from the clinical disease, in vitro models and the murine counterpart of human cerebral malaria in order to evaluate the roles played by IFN-γ in the pathogenesis of this often fatal and debilitating condition.

Keywords: CD8+T lymphocyte; blood-brain barrier; cerebral malaria; immunopathology; interferon-gamma; kynurenine pathway; microparticles; platelets.

Figure 1

Representative post-mortem histopathology findings in H & E stained brain sections from (i) wild-type and (ii) IFN-γ^−/− C57BL/6 mice on day 6 post-inoculation with 1 × 10⁶ PbA-PRBC. As no difference was evident between uninfected mice and infected IFN-γ^−/− mice, only the latter are shown. (A) Olfactory bulb; (B) Meningeal vessel; (C) Cerebellum. The brains of PbA-infected w/type mice showed hemorrhage and leukocyte adhesion to the cerebral vasculature (arrows), whereas no pathological findings were evident in any IFN-γ^−/− mouse. In this and later Figures (where appropriate) the work was carried out according to national and State legislation on animal experimentation, with approval from the University of Sydney Animal Ethics Committee.

Figure 2

Brain edema and blood-brain barrier compromise after PbA infection. Water content was calculated from wet and dry weight. Evans blue, a dye that binds to circulating albumin, was injected intravenously 2 h before mice were euthanased; the brain was perfused with saline, removed, photographed, and water-extracted; the Evans blue content was measured spectrophotometrically at 510 nm. (A) PbA-infected wild-type mouse brains had significantly greater water content compared with infected IFN-γ^−/− mice at days 6 and 7 post-inoculation (^*p < 0.001, Two-Way ANOVA with Bonferroni post-test). (B) PbA-infected wild-type mice had significantly greater extravasation of Evans Blue dye into the brain parenchyma on day 7 post-inoculation compared to infected IFN-γ^−/− mice on day 7 or 21 post-inoculation (^*p < 0.001, One-Way ANOVA with Bonferroni post-test). Above each bar of the graph is shown a representative brain from that experimental group. Columns and vertical bars are mean ± s.e.m. (n = 5 per group).

Figure 3

Blood-brain barrier compromise during PbA infection, as determined by immunohistochemical detection of fibrinogen within the parenchyma of the olfactory bulb. (A) Uninfected control mouse; (B) PbA-infected wild-type mouse at day 6 post-inoculation; (C) PbA-infected IFN-γ^−/− mouse at day 6 post-inoculation and (D) day 20–22 post-inoculation. Blood-brain barrier permeabilization to protein is clearly evident within the wild-type mouse, in which edematous changes also can be seen (arrows). These changes were not seen in IFN-γ^−/− mice at any stage of infection.

Figure 4

Enhancement of platelet-mediated endothelial cell apoptosis after IFN-γ stimulation. HBEC were stimulated with IFN-γ overnight prior to addition of platelets and RBC. Taxol treatment of HBEC was used as the positive control. FITC-BrdU nuclear fragmentation was quantified using the APO-Direct Kit (BD Biosciences) and an EPICS-XL flow cytometer (Beckman-Coulter). Results are expressed as percentages of cells undergoing apoptosis. HBEC, human brain endothelial cells; nRBC, normal red blood cells; pRBC, parasitized red blood cells; PA, Palo Alto strain of Plasmodium falciparum. In this and later Figures (where appropriate) the work was carried out according to national and State legislation on human experimentation, with approval from the University of Sydney Human Ethics Committee.

Figure 5

Effect of hypoxia-reoxygenation on TNF-induced ICAM-1 upregulation in human brain microvascular endothelial cells. Human brain endothelial cells (HBEC) were exposed to 1% O₂ for 18 h, then returned to normoxia, stimulated or not with 50 ng/mL TNF and ICAM-1 mRNA was quantified at the designated time points using a PhosphorImager® SI (Molecular Dynamics). TNF, tumor necrosis factor; ICAM-1, intercellular cell adhesion molecule-1.

Figure 6

IFN-γ mRNA in various brain regions in murine cerebral malaria. C57BL/6 mice were inoculated with 2 × 10⁵ PbA-PRBC, their brains removed on day 6 post-inoculation and dissected into regions prior to homogenization. RT-PCR was performed as described elsewhere (50). Horizontal lines and vertical bars are mean ± s.e.m. of fold differences vs. equivalent samples from uninfected mice.

Figure 7

Summary of processes relevant to cerebral malaria that are stimulated by IFN-γ, as derived from experimental models. PRBC, parasitized red blood cell; BBB, blood-brain barrier.

Figure 8

Activation of endothelial indoleamine dioxygenase-1 by IFN-γ does not affect growth of co-cultured Plasmodium falciparum. Pf-PRBC or uninfected RBC (uRBC) were cultured together with human brain endothelial cells (line HBEC-5i) for 72 h. Parasite growth as Plasmodium falciparum histidine rich protein-2 (PfHRP-2) was determined by ELISA. Under the same conditions, the IFN-γ treatment previously had been demonstrated to deplete tryptophan and cause kynurenine formation (data not shown), indicating expression and activity of IDO-1. Values are mean ± s.e.m. of triplicate determinations in a single experiment.

Figure 9

IFN-γ and the immunopathology of cerebral malaria. Schematic based on intervention studies in experimental cerebral malaria. For discussion of possible relevance to human cerebral malaria see text. Solid lines indicate direct actions (e.g., release of IFN-γ) or transitions (e.g., activation of CD8⁺T cells to CD8^*), broken lines indicate influences of secreted factors (IFN-γ and CXCL10). Ag, malaria antigen; BBB, blood-brain barrier; ICAM-I, intercellular adhesion molecule-1; μparticles, microparticles; NK, Natural Killer cell; PLT, platelets; TGF-β1, transforming growth factor-β1; VCAM-1, vascular cell adhesion molecule-1.