Platelets potentiate brain endothelial alterations induced by Plasmodium falciparum

Abstract

Brain lesions of cerebral malaria (CM) are characterized by a sequestration of Plasmodium falciparum-parasitized red blood cells (PRBC) and platelets within brain microvessels, as well as by blood-brain barrier (BBB) disruption. In the present study, we evaluated the possibility that PRBC and platelets induce functional alterations in brain endothelium. In a human brain endothelial cell line, named HBEC-5i, exhibiting most of the features demanded for a pathophysiological study of BBB, tumor necrosis factor (TNF) or lymphotoxin alpha (LT-alpha) reduced transendothelial electrical resistance (TEER), enhanced the permeability to 70-kDa dextran, and increased the release of microparticles, a recently described indicator of disease severity in CM patients. In vitro cocultures showed that platelets or PRBC can have a direct cytotoxic effect on activated, but not on resting, HBEC-5i cells. Platelet binding was required, as platelet supernatant had no effect. Furthermore, platelets potentiated the cytotoxicity of PRBC for TNF- or LT-alpha-activated HBEC-5i cells when they were added prior to these cells on the endothelial monolayers. This effect was not observed when platelets were added after PRBC. Both permeability and TEER were strongly affected, and the apoptosis rate of HBEC-5i cells was dramatically increased. These findings provide insights into the mechanisms by which platelets can be deleterious to the brain endothelium during CM.

FIG. 1.

HBEC-5i morphology. Shown is an HBEC-5i confluent monolayer after 3-day culture, exhibiting closely associated cells at confluence with cobblestone-like morphology (A). Magnification, ×400. Scanning electron microscopy revealed in vitro a monolayer of thin and extremely spread EC in resting conditions (B). In mock inflammatory conditions, however (i.e., after TNF or LT-α activation, C and D, respectively), HBEC-5i morphology was completely different. Cells exhibited a more compact and fusiform shape, in accordance with other microvascular EC under similar conditions. Transmission electron microscopy analysis of several randomly selected HBEC-5i monolayer areas showed the presence of surface villi (E, arrows), numerous pinocytotic vesicles and submembrane Weibel-Palade bodies (F, arrowheads, and “w,” respectively), and finally, peripheral electron-dense tight junctions (G and H, arrowheads).

FIG. 2.

Immunofluorescence analysis of endothelial markers on HBEC-5i subconfluent monolayers. The figure shows evidence for the expression of several typical endothelial markers on HBEC-5i surface or submembrane, ICAM-1, VE-cadherin, vWF, and occludin. The presence of other molecules with a role in cell-cell interactions, CD40 and CSA, was also demonstrated. Magnification, ×600.

FIG. 3.

Induction of HBEC-5i vesiculation and PRBC cytoadherence by platelets, LT-α, and TNF. HBEC-5i cells were cultured and left unstimulated or stimulated with TNF (100 ng/ml) or LT-α (100 ng/ml) for 6 h before scanning electron microscopy (A to C, increasing magnification). Micrographs are representative of several randomly selected fields for HBEC-5i cells exposed to both cytokines. MP production was quantified by flow cytometry for each stimulation condition (D) and after a coincubation with platelets in resting conditions (E). Results are expressed in numbers of MP per 1,000 cells (HBEC-5i). For cytometric identification and enumeration, MP were labeled with annexin V-FITC (D; labels phosphatidylserine at the surface of MP) or double-labeled with annexin V-FITC and anti-CD54-phycoerythrin (E; allows specific identification of MP of endothelial origin, excluding thereby platelet MP), respectively, both extracted from culture supernatants of 10³ HBEC-5i cells (Mann-Whitney U test; **, P < 0.01; ***, P < 0.001). The effect of platelets on PRBC cytoadherence was evaluated by a prior incubation of the HBEC-5i cells with or without platelets before the cytoadherence assay was run (F). Results are expressed in bound PRBC per mm². Bars represent standard deviations of four determinations in four experiments.

FIG. 4.

HBEC-5i monolayer alteration by cytokines, platelets, and PRBC. TEER (A) and 70-kDa dextran permeability (B) of confluent HBEC-5i monolayers were measured in resting or mock inflammatory conditions (TNF or LT-α, 50 ng/ml) and after coculture with platelets, NRBC, PRBC, or platelets and PRBC. Results are expressed in ohms per square centimeter for electrical resistance and in optical density for permeability. Bars represent standard deviations of four experiments for both TEER and dextran permeability assays (Kruskal-Wallis and Dunn pairwise tests; *, P < 0.05; **, P < 0.01; ***, P < 0.001).

FIG. 5.

HBEC-5i apoptosis induction by cytokines, platelets, and PRBC and abrogation by a pancaspase inhibitor. Shown are results of flow cytometry analysis of apoptotic HBEC-5i cells, prestimulated with TNF or LT-α (50 ng/ml) before sequential cocultures with platelets, PRBC, or platelets and PRBC (A); cells were detached and stained with FITC-coupled dUTP, as described in Materials and Methods. Histograms shown here are representative graphs of five experiments, and LT- or TNF-stimulated HBEC-5i cells without any other cell incubation (in gray) are shown as controls (7.2 and 9.3%, respectively, for the four coincubation conditions). Also shown are results of flow cytometry analysis of HBEC-5i cells, preincubated or not with zVAD-fmk before TNF stimulation (50 ng/ml) and sequential cocultures with platelets and then PRBC (B); cells were detached and stained with FITC-coupled dUTP, and results are expressed as a percentage of apoptotic EC. Bars represent standard deviations of three experiments. Dot plots presented here are representative of three experiments.