Myeloperoxidase and Septic Conditions Disrupt Sphingolipid Homeostasis in Murine Brain Capillaries In Vivo and Immortalized Human Brain Endothelial Cells In Vitro

Abstract

During inflammation, activated leukocytes release cytotoxic mediators that compromise blood–brain barrier (BBB) function. Under inflammatory conditions, myeloperoxidase (MPO) is critically involved in inflicting BBB damage. We used genetic and pharmacological approaches to investigate whether MPO induces aberrant lipid homeostasis at the BBB in a murine endotoxemia model. To corroborate findings in a human system we studied the impact of sera from sepsis and non-sepsis patients on brain endothelial cells (hCMEC/D3). In response to endotoxin, the fatty acid, ceramide, and sphingomyelin content of isolated mouse brain capillaries dropped and barrier dysfunction occurred. In mice, genetic deficiency or pharmacological inhibition of MPO abolished these alterations. Studies in metabolic cages revealed increased physical activity and less pronounced sickness behavior of MPO−/− compared to wild-type mice in response to sepsis. In hCMEC/D3 cells, exogenous tumor necrosis factor α (TNFα) potently regulated gene expression of pro-inflammatory cytokines and a set of genes involved in sphingolipid (SL) homeostasis. Notably, treatment of hCMEC/D3 cells with sera from septic patients reduced cellular ceramide concentrations and induced barrier and mitochondrial dysfunction. In summary, our in vivo and in vitro data revealed that inflammatory mediators including MPO, TNFα induce dysfunctional SL homeostasis in brain endothelial cells. Genetic and pharmacological inhibition of MPO attenuated endotoxin-induced alterations in SL homeostasis in vivo, highlighting the potential role of MPO as drug target to treat inflammation-induced brain dysfunction.

Distinct effects of calcineurin dependent and independent immunosuppressants on endotoxaemia‐induced nephrotoxicity in rats: Role of androgens

“…TNF‐α is a multifunctional cytokine that follows TLR‐4 activation and causes subsequent upregulation of other cytokines in the proinflammatory cascade, e.g. interleukins, adhesion molecules, and MPO 20,41‐43 . Overall, our findings favour the idea that implicates calcineurin inhibition in the heightened inflammatory and nephrotoxic effects of endotoxaemia upon concurrent exposure to cyclosporine or tacrolimus.…”

The lipid biology of sepsis

“…Treating endothelial cells with serum from septic patients resulted in endothelial cell SM loss accompanied by a sharp and rapid production of Cer, which then is further metabolized to glycosylated ceramides (122). However, these results were not observed by Sattler et al, who observed decreases in brain endothelial cell levels of Cer and an associated loss of barrier function after treatment with septic patient serum (123). Decreased S1P levels in human sepsis is likely the result of reduced sphingosine kinase (Sphk1 and J o u r n a l P r e -p r o o f Sphk2) activity (112).…”

Review of Design Considerations for Brain-on-a-Chip Models

“…Depending on the origin of the cell types of interest, the composition of ECM incorporated into the models should also be considered. BM coating considerations for endothelial cell phenotypes should be cell type-dependent, as immortalized cell lines have similar phenotypes across different ECM substrates or without coatings [ 270 , 271 , 272 , 273 , 274 , 275 ], primary cells have shown promising results with fibronectin, collagen IV and laminin as the BM [ 231 , 276 ] and iPSC-derived BMECs show physiological behavior when cultured with collagen IV/fibronectin coatings, Matrigel ® , and possibly laminin-221 [ 44 , 279 ]. Any investigators studying barrier function should consider testing an array of BM compositions to find a suitable option for their specific cell source.…”

“…Since HCMEC/D3 cells are the most commonly used BMEC cell line, we will focus on coatings compatible with these cells. Given their generally robust phenotype (reviewed in [ 79 ]), most studies that do use a coating for these cells select generic rat-tail collagen-I (10 ug/cm 2 ) [ 272 , 273 , 274 , 275 ] due to its low cost, even though collagen-I is not found in the cerebral microvasculature. Several combinations of coatings that are more physiologically relevant have been tested in an attempt to optimize the barrier formation of HCMEC/D3, with little success.…”