Sphingosine 1-phosphate (S1P) in blood is phosphorylated, stored, and transported by red blood cells (RBC). Release of S1P from RBC into plasma is a regulated process that does not occur in plasma- or serum-free media. Plasma fractionation and incubations with isolated and recombinant proteins identified high density lipoprotein (HDL) and serum albumin (SA) as non-redundant endogenous triggers for S1P release from RBC. S1P bound to SA and HDL was able to stimulate the S1P(1) receptor in calcium flux experiments. The binding capability of acceptor molecules triggers S1P release, as demonstrated with the anti-S1P antibody Sphingomab. More S1P was extracted from RBC membranes by HDL than by SA. Blood samples from anemic patients confirmed a reduced capacity for S1P release in plasma. In co-cultures of RBC and endothelial cells (EC), we observed transcellular transportation of S1P as a second function of RBC-associated S1P in the absence of SA and HDL and during tight RBC-EC contact, mimicking conditions in tissue interstitium and capillaries. In contrast to S1P bound to SA and HDL, RBC-associated S1P was significantly incorporated by EC after S1P lyase (SGPL1) inhibition. RBC-associated S1P, therefore, has two functions: (1) It contributes to the cellular pool of SGPL1-sensitive S1P in tissues after transcellular transportation and (2) it helps maintain extracellular S1P levels via SA and HDL independently from SGPL1 activity.

These results suggest that SAA enrichment of HDL during disease conditions contributes to the decreased protective function. It is a novel finding that SAA acts as a pro-inflammatory molecule to reduce the anti-inflammatory properties of HDL.

BACKGROUND AND PURPOSEPurinergic signalling plays an important role in vascular tone regulation in humans. We have identified uridine adenosine tetraphosphate (Up4A) as a novel and highly potent endothelial-derived contracting factor. Up4A induces strong vasoconstrictive effects in the renal vascular system mainly by P2X1 receptor activation. However, other purinoceptors are also involved and were analysed here. EXPERIMENTAL APPROACHThe rat isolated perfused kidney was used to characterize vasoactive actions of Up4A. KEY RESULTSAfter desensitization of the P2X1 receptor by a,b-methylene ATP (a,b-meATP), Up4A showed dose-dependent P2Y2-mediated vasoconstriction. Continuous perfusion with Up4A evoked a biphasic vasoconstrictor effect: there was a strong and rapidly desensitizing vasoconstriction, inhibited by P2X1 receptor desensitization. In addition, there is a long-lasting P2Y2-mediated vasoconstriction. This vasoconstriction could be blocked by suramin, but not by PPADS or reactive blue 2. In preparations of the rat isolated perfused kidney model with an elevated vascular tone, bolus application of Up4A showed a dose-dependent vasoconstriction that was followed by a dose-dependent vasodilation. The vasoconstriction was in part sensitive to P2X1 receptor desensitization by a,b-meATP, and the remaining P2Y2-mediated vasoconstriction was only inhibited by suramin. The Up4A-induced vasodilation depended on activation of nitric oxide synthases, and was mediated by P2Y1 and P2Y2 receptor activation. CONCLUSIONS AND IMPLICATIONSUp4A activated P2X1 and P2Y2 receptors to act as a vasoconstrictor, whereas endothelium-dependent vasodilation was induced by P2Y1/2 receptor activation. Up4A might be of relevance in the physiology and pathophysiology of vascular tone regulation.Abbreviations a,b-meATP, a,b-methylene ATP; AngII, angiotensin II; ApnA, diadenosine-n-phophate (n: number of phosphates); ApnG, adenosine-guanosine-n-phosphate (n: number of phosphates); ApoE, apolipoprotein E; CI, confidence interval; DMSO, dimethyl sulphoxide; eNOS, endothelial NOS; GpnG, diguanosine-n-phosphate (n: number of phosphates); L-NAME, N G -nitro-L-arginine methyl ester; MAP, mean arterial blood pressure; MCP-1, monocyte chemoattractant protein-1; MRS2179, 2′-deoxy-N6-methyladenosine 3′,5′-bisphosphate; PPADS, pyridoxal-phosphate-6-azophenyl-2;4-disulphonic acid; RB2, reactive blue 2; Up4A, uridine adenosine tetraphosphate BJP British Journal of Pharmacology

Nimesulide is a nonsteroidal anti-inflammatory drug (NSAID) marketed in more than 50 countries. This drug has caused rare and idiosyncratic but severe hepatotoxicity. The mechanisms associated with and factors responsible for this toxicity remain unknown. One of the nimesulide metabolites identified in human urine is 4-amino-2-phenoxy-methanesulfonanilide (M1). In the current study, we demonstrate that M1 is a stable metabolite that is highly susceptible to facile oxidation by cytochrome P450 enzymes (P450s) to form a reactive diiminoquinone intermediate (M2). Direct detection of M2 was difficult by LC-MS. However, its formation was confirmed indirectly by identification of N-acetyl-cysteine (NAC) adducts of M2. The formation of diiminoquinone M2 was P450 mediated with 2C19 and 1A2 as the two principal P450 enzymes catalyzing M1 oxidation. M1 metabolism irreversibly inhibited 2C19 but activated 1A2 in a time-dependent manner. P450 2C19 exclusively mediated further metabolism of M1 to the amino hydroxynimesulide M3 and its diiminoquinone M4. Similar to M2, M4 is also reactive and can be observed indirectly as its NAC adduct. Nucleophilic addition to diiminoquinone M2 occurs with low regioselectivity, yielding three adducts (the peak area ratio 1:0.08:12). The three regioisomers have the same m/z for [M + H](+), presumably due to nucleophilic addition at the three possible electrophilic sites (C-3, -5, and -6 positions of the sulfonaniline ring). The primary adduct, R, was derived from the attack of the nucleophile at the C-5 position of the sulfonaniline ring and was determined by MS/MS and (1)H and (13)C NMR analyses. The structural assignments were confirmed by chemical synthesis of the adduct R. M2 demonstrated its electrophilic reactivity by selectively alkylating human serum albumin (HSA) at the only free thiol, Cys-34. This suggests the possibility that other proteins may undergo a similar conjugation to form irreversible adducts. Under oxidizing conditions in the presence of cumene hydroperoxide (CHP), the formation of M2 was enhanced, indicating that oxidative stress may accelerate the production of reactive diiminoquinone species (M2 and M4).