Vascular endothelial cells cultured from patients with cerebral or uncomplicated malaria exhibit differential reactivity to TNF

Abstract

Plasmodium falciparum malaria is a major cause of morbidity and mortality in African children, and factors that determine the development of uncomplicated (UM) versus cerebral malaria (CM) are not fully understood. We studied the ex vivo responsiveness of microvascular endothelial cells to pro-inflammatory stimulation and compared the findings between CM and UM patients. In patients with fatal disease we compared the properties of vascular endothelial cells cultured from brain tissue to those cultured from subcutaneous tissue, and found them to be very similar. We then isolated, purified and cultured primary endothelial cells from aspirated subcutaneous tissue of patients with CM (EC(CM) ) or UM (EC(UM) ) and confirmed the identity of the cells before analysis. Upon TNF stimulation in vitro, EC(CM) displayed a significantly higher capacity to upregulate ICAM-1, VCAM-1 and CD61 and to produce IL-6 and MCP-1 but not RANTES compared with EC(UM) . The shedding of endothelial microparticles, a recently described parameter of severity in CM, and the cellular level of activated caspase-3 were both significantly greater in EC(CM) than in EC(UM) . These data suggest that inter-individual differences in the endothelial inflammatory response to TNF may be an additional factor influencing the clinical course of malaria.

Fig. 1

Immunofluorescence analysis of endothelial markers on subcutaneous fat and brain-derived EC^CM monolayers. The figure shows evidence for the expression of several typical endothelial markers on submembrane or cell type surface, including vWF, CD31, ZO-1, ICAM-1, CD36 and CD62-E. Micrographs presented here are from patient PM91-derived cells in passage 3 and are representative of the results obtained with all the other CM patients. Magnification: × 100 for general morphology micrographs, × 400 for immunofluorescence micrographs.

Fig. 2

Comparison of TNF-induced adhesion molecule upregulation between EC^UM and EC^CM of subcutaneous fat origin. Cells were seeded in a 96-well plate and stimulated by increasing concentrations of TNF. EC^UM and EC^CM were then fixed and the expression of ICAM-1, VCAM-1 and CD61 was measured by cell-based ELISA (A, B and C respectively). Results are expressed as mean ± SD of two experiments (Dunn's test, *P < 0.05, **P < 0.01 and ***P < 0.001). As a control, each experiment was confirmed by flow cytometry (D–F). The results of one representative reading obtained with the greatest concentration of TNF are shown here.

Fig. 3

Quantification of TNF-induced endothelial MP release by subcutaneous fat-derived EC^UM and EC^CM. Cells were cultured and left resting or stimulated with TNF for 6 h before analysis. MP production was quantified by flow cytometry for each condition. Results (two determinations in three experiments) are expressed in numbers of MP labelled with annexin V-FITC, extracted from culture supernatants of 1000 EC (Dunn's test, **P < 0.01).

Fig. 4

Comparison of TNF-induced production of MCP-1, RANTES, IL-6 and cleaved Caspase-3 between EC^UM and EC^CM. Subcutaneous fat-derived EC from both patient categories were stimulated with different doses of TNF. In the case of MCP-1, RANTES and IL-6, supernatants were collected after 12 h of incubation and concentrations were measured by ELISA. For activated Caspase-3, supernatants and cell extracts were pooled before analysis by ELISA. Results are expressed as mean ± SD of three determinations in two experiments (Dunn's test, *P < 0.05 and ***P < 0.001).

Fig. 5

Schematic representation of needle aspiration biopsy. The diagram shows the technique used to isolate subcutaneous fat-derived fat EC by inserting a hollow bore ‘menghini’ needle in the upper external thigh of patients with malaria (UM and CM) after application of a local anaesthetic for 2 h. A strong manual aspiration force represented by the arrow allowed the sampling of approximately 5 mm³ of adipose tissue, from which EC were isolated, selected and cultured before analysis.

Fig. 6

Method for sampling and isolating EC^CM from brain and subcutaneous fat tissue. EC^CM were collected at autopsy and isolated from both subcutaneous fat and brain tissue from patients who died of CM (n = 7). Brain EC^CM were only used for a phenotypical comparison with subcutaneous fat-derived EC^CM in the first part of our study to assess the relevance of the latter as an accurate model of cerebral endothelium.