Cytoadhesion of Plasmodium falciparum-infected erythrocytes (IE) to activated endothelial cell (EC) surfaces plays a key role in the pathophysiology of cerebral malaria. Moreover, recent evidence suggests that platelet adhesion and aggregation are critical in facilitating this cytoadhesion process. We have recently reported marked increased plasma VWF and VWF propeptide levels in severe P. falciparum infection, indicative of acute EC perturbation. Furthermore, plasma VWF:Ag levels in patients with malaria inversely correlated with platelet count, and plasma VWF propeptide levels correlated with other established biochemical markers of malaria severity, including plasma lactate. Nevertheless, it remains unclear whether VWF plays any direct role in mediating IE cytoadhesion, or whether the increased plasma levels of both VWF and propeptide merely serve as a marker of acute EC activation. To address this question, we collected plasma samples from a cohort of African children presenting with cerebral malaria (CM), or non-cerebral severe malaria (SM). In children with CM and SM, plasma VWF:Ag levels were significantly elevated (medians 3.1 and 3.4 IU/ml) as before. VWF collagen binding (VWF:CB) levels were also markedly increased (medians 7.6 and 7.0 IU/ml). Furthermore, the relative rise in VWF:CB was much greater, so that the ratio of CB:Ag was consistently increased. VWF multimer analysis demonstrated abnormal circulating ULVWF multimers in children with CM and SM, consistent with acute, regulated VWF secretion. To characterize mechanisms responsible for the markedly increased VWF:CB and circulating ULVWF in malarial plasma, we investigated ADAMTS13 antigen and activity levels in CM and SM plasma. Plasma ADAMTS13 activity levels (FRETS-VWF73 assay) and antigen levels were both significantly reduced (medians = 0.63 U/ml and 0.56 U/ml; p< 0.001) in children with CM and SM compared to controls. Classical mixing studies of malaria and normal plasma demonstrated no evidence of immediate ADAMTS13 inhibition. However, ADAMTS13 activity in normal plasma was significantly reduced (~60%) following 30 min incubation with malarial plasma (75%:25% mix). This inhibitory ability of malaria plasma was further confirmed by spiking malarial plasmas with recombinant human ADAMTS13. Significant time-dependent inhibition of FRETS-VWF73 activity was again observed in the plasmas of children with severe P. falciparum, but not in normal control plasmas. Potential inhibitors of ADAMTS13 in vivo include interleukin-6 (IL-6), thrombospondin-1 (TSP-1), thrombin and plasmin, free plasma haemoglobin, and reduced FVIII levels. Although plasma IL-6 levels were significantly elevated in children with either CM (mean 240 pg/ml; p<0.001) or SM (mean 217 pg/ml; p=0.01) compared to normal controls, levels did not approach those previously reported necessary to inhibit ADAMTS13 activity. Disseminated intravascular coagulation (DIC) is associated with enhanced thrombin generation, and consumption of FVIII, both of which can inhibit VWF proteolysis by ADAMTS13. However, in children with CM or SM respectively, we observed significantly increased plasma FVIII:C levels. Although intravascular haemolysis is also a recognised complication of malarial infection, we observed only minor increased plasma haemoglobin concentrations, again well below that previously described to significantly inhibit ADAMTS13 activity. Finally, in contrast to the increased plasma levels of IL-6 and FVIII:C, we found that plasma TSP-1 levels were not significantly elevated in children with either CM or SM compared to pooled normal plasma. In conclusion, we demonstrate that the presence of ULVWF multimers in the plasma of children with severe P. falciparum malaria is the result of acute EC activation and release of ULVWF from WP bodies; significantly reduced plasma ADAMTS13 antigen levels a circulating but unidentified inhibitor of human ADAMTS13 activity. In view of the critical role played by VWF in mediating platelet adhesion/aggregation, and the accumulating evidence suggesting that platelet adhesion/aggregation also facilitate cytoadhesion of IE, we propose a novel role for ULVWF multimers in the pathophysiology of severe P. falciparum malaria.

No abstract

2119 Poster Board II-96 ABO(H) blood group antigen expression on platelets varies widely among normal blood donors. An ABO ‚High Expresser' phenotype (HXP) that exhibits significantly increased A and/or B antigen expression on platelets has been identified in ∼7% of normal donors. HXP has been implicated in both platelet-refractoriness and neonatal alloimmune thrombocytopenic purpura, however, the underlying molecular and genetic elements that mediate this phenomenon are not well-defined. To investigate the mechanisms underlying HXP, blood samples were collected from 231 group A (180 A1and 51 A2) and 310 group O individual apheresis platelet donors. Quantitative expression of platelet A and H antigen were then assessed by flow cytometry of platelet-rich plasma. In total, 10 A1 donors (5.6%) exhibited HXP. Analysis of the platelet A antigen expression in these individuals identified 8 HXP donors who exhibited ‚type I' HXP (normal bimodal population of platelets, but with predominant A antigen expression) whereas 2 individuals exhibited ‚type II' HXP (a single uniform population of platelets, strongly positive for blood group A expression). Both types of HXP were found to be a stable donor characteristic. ABO(H) determinants have also been identified on the N-linked glycans of the plasma von Willebrand factor (VWF), and influence plasma VWF levels and susceptibility to proteolysis by ADAMTS13. To determine whether HXP was platelet-specific, blood group A antigen expression on plasma VWF from group A donors was determined. Interestingly, blood group A antigen expression on plasma VWF was concordantly increased in donors with type I and type II HXP, indicative of increased glycosyltransferase expression in HXP individuals. To ascertain whether increased glycosyltransferase expression contributes to HXP, ABO genotype was determined for all 231 group A donors by PCR-RFLP analysis. Genotype at the ABO locus on 9q34 exerts a dosage effect on glycosyltransferase expression. 80% HXP (all type I) donors were genotyped A1A1. suggesting increased A transferase activity contributes to type I HXP. Despite this, the majority of A1A1individuals (67%) did not exhibit HXP, and 2 HXP donors were found to possess the A2 allele, which expresses limited A transferase enzymatic activity. Collectively, this data clearly demonstrates the contribution of additional factors to ABO genotype that contribute to HXP. To identify additional HXP modifiers, potential enhancer repeat elements upstream of the ABO gene were examined in group A donors, including those with HXP. Typically, A1alleles contain a single 43-base pair repeat within a minisatellite positive regulatory region upstream of the ABO gene. In contrast, A2and O1alleles contain four 43bp repeats, which are associated with a 100-fold enhancement of transcriptional activity. Analysis of this enhancer region demonstrated two HXP donors with A1alleles containing four copies of the 43-base pair repeat. Consequently, this allele would be predicted to modulate A transferase expression via enhanced ABO gene transcription. In conclusion, we have demonstrated the multi-factorial nature of the regulatory elements mediating platelet type I and type II HXP. A1alleles containing novel upstream enhancer repeats identified in donor individuals may represent a novel genetic mediator of HXP, and contribute to the pathophysiology associated with this phenomenon independently of ABO genotype. Disclosures: No relevant conflicts of interest to declare.

2234 Von Willebrand Factor (VWF) is extensively glycosylated with both N- and O-linked carbohydrates. Moreover, these complex glycan structures influence VWF functional properties, including susceptibility to ADAMTS13 proteolysis, and plasma clearance. The molecular mechanisms through which VWF glycosylation (including ABO blood group antigens) act to influence VWF physiology remains unexplained. However, recent data suggest that VWF circulates in normal plasma bound to various carbohydrate-binding proteins, including specific members of the galectin family. In addition, galectin-3 binding has been reported to influence VWF cleavage by ADAMTS13. In this context, we sought to elucidate the role of specific VWF glycan determinants in modulating galectin interaction. VWF was purified from human plasma (pdVWF) by cryoprecipitation and gel filtration. VWF glycosylation was then modified using exoglycosidases and quantified by specific lectin ELISAs. Blood group specific VWF was also purified from pooled group AB, O, or Bombay plasmas. Galectins-1 and -3 were transiently expressed in competent E-coli cells with an N-terminal histidine tag, and purified by nickel chromatography. Finally, binding interactions were characterized via modified immunosorbant assay. In keeping with the previous report of Lenting et al, human pdVWF bound to both galectin-1 and galectin-3 in a dose-dependent manner. Enzymatic desialylation of pdVWF with α2-3,6,8,9 neuraminidase (Neu-VWF) markedly enhanced binding to galectin-1 (231±6%, p<0.0001). Similarly, removal of terminal sialic acid also increased binding to galectin-3, albeit to a lesser extent (136±6%, p<0.05). To further define the role of VWF glycans in regulating galectin binding, pdVWF was exposed to sequential neuraminidase and galactosidase digestions to remove terminal sialic acid and sub-terminal galactose residues (NeuGal-VWF). In contrast to the enhanced binding of Neu-VWF, binding of NeuGal-VWF to both galectin -1 and -3 was significantly reduced (51±5% and 52±6% compared to pdVWF; p<0.005). Cumulatively these findings suggest that loss of capping sialic acid and exposure of sub-terminal galactose critically regulates VWF-galectin binding. Treatment with PNGase F to completely remove N-linked carbohydrate structures (PNG-VWF) markedly decreased binding to galectin -1 and -3 (13±1% and 57±2%, p<0.001). Moreover, combined PNGase F and O-glycosidase digestions further attenuated galectin-3 binding (21±1%, p<0.001), suggesting that both the N- and O-linked glycans are involved in mediating the VWF-galectin interaction. ABO(H) blood group antigens are expressed on both the N-linked and O-linked glycans of human VWF. Moreover, ABO(H) determinants influence VWF susceptibility to ADAMTS13 proteolysis and plasma VWF half-life, through unknown mechanisms. Purified VWF from normal group AB individuals bound to both galectin-1 and galectin-3 significantly better than group O VWF (146±8% and 483±19%; p<0.01). Conversely, no significant difference in binding was observed between Group O and Bombay VWF. Consequently, although terminal A (GalNAc) and B (Gal) sugar moieties promote galectin binding, expression of terminal α1–2 fucose residues is not important. The glycosylation profile of platelet-VWF differs from that of pdVWF. In particular, platelet-VWF expresses reduced levels of both capping sialic acid and sub-terminal galactose residues (∼50%), and lacks AB blood group antigens. To characterize the effects of this differential sugar expression on galectin binding, platelet-derived VWF was isolated and purified (platelet freeze-thawing followed by immuno-affinity chromatography with monoclonal CLB-Rag20). In keeping with the reduction in Gal and AB blood group antigen expression, platelet VWF bound less well to galectin-1 and galectin-3 (72±6% and 67±7% versus pdVWF; p<0.05). These novel data demonstrate that both the N- and O-linked oligosaccharide structures of VWF are involved in mediating galectin binding. In particular, expression of terminal AB blood group antigens, and expression of sub-terminal galactose moieties following loss of capping sialic acid, both markedly enhance galectin binding affinity. Further studies will be required to define how galectin binding is involved in mediating the functional consequences of variation in VWF glycans. Disclosures: No relevant conflicts of interest to declare.

Background: In addition to their etsblished anticoagulant activity, unfractionated heparin (UFH) and low molecular weight heparin (LMWH) are known to possess clinically important immuno-modulatory properties. However different studies have reported conflicting pro- and anti-inflammatory effects in association with heparin. Moreover, the molecular basis for these heparin effects on inflammation remains unclear. In view of the wide and diverse clinical indications for heparin, it is clearly of direct translational relevance to define how UFH and LMWH differentially regulate inflammatory responses to LPS in-vivo. Objectives: To determine how UFH and LMWH regulate lipopolysaccharide (LPS)-induced activation of human mononuclear cells in whole blood, and define the role of lipopolysaccharide binding protein (LBP) in mediating this effect. Methods: Whole blood was pre-treated with UFH or LMWH (0.1–200 IU/ml), prior to stimulation with LPS (10ng/ml). After 6 hours, monocyte pro-inflammatory cytokine (interleukin (IL)-1b, IL-6, IL-8, and TNF-a) secretion was determined by plasma ELISA. Parallel experiments using THP-1 cell line and primary monocytes were performed under serum-free conditions, in the presence or absence of varying doses of recombinant human LBP (range: 50–100nM). Results: Under serum-free conditions, heparin demonstrated dose-dependent anti -inflammatory effects, significantly reducing secretion of pro-inflammatory cytokines (IL-1b, IL-6, IL-8, and TNF-a) in response to LPS-stimulation of THP-1 cells and primary monocytes. In contrast, in the presence of LBP, both UFH and LMWH demonstrated dose-dependent pro-inflammatory effects at all heparin concentrations. In ex-vivo whole blood experiments, pro-inflammatory effects (increased IL-1b and IL-8 following LPS-stimulation) of heparin were also observed, but only at supra-therapeutic doses (10–200IU/ml). Conclusion: In keeping with previous reports, we have demonstrated that both UFH and LMWH can significantly down-regulate cytokine (TNF-a, IL-1b, IL-6 and IL-8) secretion in response to LPS-activation in-vitro. However our novel data demonstrate that the effect of heparin on monocyte activation by LPS is significantly more complex in the setting of whole blood. Firstly, in contrast to the anti-inflammatory effects observed under serum-free conditions, we found that in whole blood, high concentrations of heparin exerted marked pro-inflammatory effects. Secondly we have also demonstrated that the effects of heparin in whole blood are entirely dependent upon heparin concentration and LBP concentration.