Reviews
Chondroit[n Sulphate A as an Adherence Receptor for Plasmodium faiciparum.infected Erythrocytes S.J, Kogerson and G.V, Brown Until recently, the sequestration of erythrocytes infected with Plasmodium falciparum has been thmtght to be due to one of a mmt#er of protein-proteh~ interactions, hl this artick,, Stephe, Ro~erson and Graham Brow,i summarize the emerging evidence that, in vitro, i':f,-cted erythrocytes can al~ adhere to the ~h/cosaminoglycan chondwitin sulphate A (CSA) expressed on the sur}'~ce of cells and immobilized on plastic. In vivo, binding of infected erythrocytes to CSA co;:ld Ire crucial to the development Of malarial infection of the placenta, and possihly to sequestraSon in the lung and brain. The consequences of this may include maternal morbhlity and mortality, 16,zobirth weight in the fllfant, pulmo,ary oedema and cerebral malarh~. They discuss the need io characterize the molecular basis of this interaction, and to investigate the possible therapeutic role of CSA in malaria. Chondroitin sulphates are nontoxic compounds already iu use for other diseases in hmnans. Vaccim~ ba~d on inhibithlg this receptor-ligand h#eraction could al~ be appropriate. Sequestration of mature parasites in deep tissues is a key feature of the replication of Plasmodim, falciporum infection. Sequestration of parasitized red blood cells (PRBC) in the brain and in the placei-tta appears to be crucial to the development of cerebral ~alaria and complications of malaria in pregnancy. It is clear that several different host molecules can act as receptors for PRBC in vitro. These include thrombospondin (TSPp, CD36 (Re/. 2), intercellular adhesion molecule 1 (ICAM-1) 3, E-selectha and vascular cell adhesion molecule 1 (Ref. 4). When isolates from patients have been studied, it has born difficult to correlate adhesion to any of these receptors with disease severity, and only one study has found an association between adhesion to C32 melanoma cells and severe disease s. Evidence has pointed to the existence of other receptors, on microvascular endothelial cells6, C32 melanoma cellsz, human umbilical vein endothelial cells (HUVEC) s and on 5aimiri monkey brain endothelial ceils ($BEC) 9. A n o v e l receptor for P. [aiciparum
We recently described adhesion of PRBC to tl, e glycosaminoglycan (GAG) chondroitin sulphate A (CSA) m. When we selected laboratory lines by panning on wild-t3/pe Chinese hamster ovary (CHO) Stephen Rolers~-~ and Graham Brown are at the Im~qunoparasitology UniL The Walter and Eliza Hall Institute of Pieaical Research,Post Office Royal ,~-lelboume Hospital. Victona 3050. Austra!ia. S.]. Roger~on is also at the Wellcome Centre for Research in Clinical Tropical Medione, University of Liverpool, Liverpool, UK L69 3BX. G.V. Rrown is also at the Victorian Infectious DiseasesService,Royal Plelboume Hospital, Victoria 3050. Australia.Tel: +1"13 9345 2555, Fax: +613 9347 08S2s e-rnzil: ro~,
[email protected] 70
cells, high levels of binding resulted (Fig. Is), and this binding was inhibited by free CSA at low concentration (1 ~g ml-1), but not by other GAGs or glycoconjugates. By scanning electron microscopy, attachment was found to be mediated by processes from the CHO cell surface making contact with the PRBC (Fig. 2), similar to those observed for attachment to C32 melanoma cells n. Like many human cells, CHO cells have significant amounts of proteoglycan on their surface, containing CS or heparan sulphate (HS) 12 and binding was found to be dependent on the presence of CS on the cell surface. These selected lines of PRBC hound well to C32 melanoma ceils (which express a CSA containing M~ 250000 protein on their surface B) and this binding was also inhibited by CSA. Inhibition was specific, as free CSA did not inhibit binding to CD36 or to ICAM-1 (S.J. Rogerson, unpublished). When CSA was conjugated to a lipid and immobilized on plastic it supported binding of PRBC, while an HS control did notre(Fig, lb). GAGs are heteropolysaccharides composed of alternating N-acetylated amino sugars and hexuronic acids with variable degrees of sulphation, forming chains of up to 100 or more saccharide units. GAGs are normally attached to a protein backbone to form proteoglycans, although they may be shed in some inflammatory conditions. The most extensively studied GAGs are HS and closel~ related heparin, which is used as aniicoagulant ther4py for . . . . . . . . . . . . . . . . . . . H$ is attached t,, a variety of proteins, including several that are important in the vascular system14. HS proteoglyeans mediate adhesion of a variety of infectious agents including Trypanosoma cruzi is, Leishmania donovani 16, herpes simplex virus 17 and cytomegalovirus 8. Most interestingly, HS may also act as a receptor for plasmodie~| sporozoltes on the hepatocyte surtfa,~ e~19. mncfing ~. -of the circumsporozoite protein to the hepatocyte can be inhibited by HS and other glycoconjugates, and a similar range of compounds can prevent malarial infection when administered to mice prior to sporozoite challenge2°. Or.'r unde~tanding of the role of CS in adhesive proce ~ses is less comprehensive. CS has not previously been described as mediating adhesion of infectious agents. Three forms of CS [CSA, CSB (dermatan sulphate) and CSC] have different basic structures. The classical structure of CSA is one of repeating units of alternating glucuronic acid and N-acety! galactosamine (Adi-45), although it is now clear that GAGs are much more heterogeneous21 than previously suspected. In particular, variations in the number and the positions of sulphate residues added and in the degree of epJmerization between glucuro;~ic acid and iduronic acid can convey structural sptn ificity. Specific oligosaccharide ligands have been d~scovered in heparin for antithrombin III and for fibroblast growth
Cop~;~t© 1~97Et'~,~rSc~c~LtctAll~ght,~.~O 0t69475BFgy/~MLO0 ~I;$0169-4758(9~10081-8 Parasitology Today, vol. 13..0. 2, 1997
Reviews factor.. Determination of the CSA motif interacting with PRBC would facilitate the possible development of CSA-based therapy and the investigation of the fine details of the receptor-ligand interaction. Heparin, HS and CS are found circulating in the blood of normal individuals, and appear to be largely protein bound ~. CS was detected at a concentration of approximately 1 ~.g m1-1, and 40%' was Adi-4S, the principal component of CSA, with a peak of M r 15000-17000 (ReL 22). These levels a~'e below those described as inhibitlng adherence of PRBC to CSA 1°~. We do not know how much levels increase during malaria infection, or what the effects of this host response would be on adherence. Some patient isolates bind to CSA It was clear from our initial examination of isolates from patients that at least a proportion of patients exhibit PRBC that can adhere to CSA l°. About half of the patients had PRBC that bound .~o CSA, at similar levels to binding to ICAM-1. In more detailed studies on isolates from Thai adults with severe or uncomplicated malaria, Chaiyaroj et el. found that five of 36 isolates bound to CSA, similar findings to those observed for ICAM-1 binding 24. All but one of the isolates boun~ to CD36, and that isolate bound in a CSA-dependent manner to C32 melanoma cells. There wa'no correlation between binding to CD36 and to C32 cells in this study, or in an earlier one25, possibly indicating a component of CSAdependent binding to these cells. No association was seen between adherence to any receptor and disease severity in this relatively small study 24. CSA as a receptor on Saimiri m o n k e y brain endothelial cells CSA was found to be an important receptor for PRBC on SBEC cloned lines developed by Gay and colleagues9,26. Binding to five SBEC clones by two of three laborato~ lines and three of five patient isolates was significantly inhibited by free CSA, and almost totally inhibited by treatment of SBEC with chondroitinase ABC, suggesting that virtually all binding to SBEC is dependent on the presence of surface CSA. Cerebral sequestration is infrequent in this model, and the Parasitology Today. vol. 13. no. 2. 1997
Fig. I. Binding of PRBC to witd-tTpe Chinese hamster ovary (CHO) cells (x 10~) after ~ve cycles of selec[ion on C l i o cells (a); PRBC binding to immobilixed CSA (b); PRBC bind to AS.O ceils (c). This binding was inhibited >9S% by free CSA at IO ttg ml -I, (S,J.Roger'son, unpublished).
Reviews well to CSA) specifically agglutinate CS2 and not its parent FAF-EA8 (which binds poorly to CSA), and vice versa. Parasite line T4 ,adhered to an uncharacterized receptor on C32 cells and had a different form of PfEMP1 from FAF-EA8 from which it was derived 7. We have now found T4 to bind to CSA on C32 or on plastic at much higher levels than does FAFEA8 (K. Davern, unpublished). If PfEMPI is indeed the ligand for CSA binding, it will be important to identify the binding motif within these large proteins. Although PfEMPt sequences differ markedly from isolate to isolate, functional domains encoding binding may be conserved. If such conserved domains can be identified, they can then be assessed as poss~bte vaccine compouents a n d / o r ther~,veutic agents. role of CSA binding in the development of disease in the monkey requires further investigation. CSA-dependent binding under conditions of physiological flow
Does adherence to CSA occur under conditions resembling those in the sites where s~quesh'ation occurs? When PRBC flow over immobilized receptors a~ ~'~i~ ~i~,~r ~tresses equivalent to those in post carpellary venub:~, they bind well to CD36 and ICAM-1 but poorly ~.o TSp27.Using a p~ra~Ic~-p!ate flow chamber and contro~ed flow ra.~es, we have now also found CSA to bi~d P~L3C M~m flow at wall shear stresses of 0.05-'/).i Pascal (PaL similar to those found in post capillary ven~!es, avd to resist detachment of immobilized PRBC ex'~,~i whe~ stresses are increased to high levels~. A wall shear stress of 1.2 Pe :;'as zt:qtiired to detach 50% of bo,.~nd PRBC 2~. fhis ctn'tfirms the biological relevance of finding~ from the static assays for adherence ;o CSA, CD36 and ICAM-1, but not to TSP~~. What is the ligand for adhesion? The principal protein mediating adherence of PRBC to endothe!ial ceils i~ P ta;ciparum erythrocyte membrane protei~. 1 (P~MI"I), a variable, M, 200000-350~0 protein syn;h,:sized b) mature trophozoites and expressed at th. red blocxl cell surface. Convincing evidence has recently been presented that PfEMP1 is responsible for binding to CD36, ICAM-1 and TSP3°. PfEMP1 is encoded by the var genes, a large multigene family 3~-33 and appears to be the dominant immunogen on the PRBC ~urface, with antibodies recognizing PfEMPI causing agglutination of infected erythrocytes~i. Furthermore, we have shown that there is an association between antigenicity and phenotype, as selection for particular binding phenotypes produces parasites with PfEMPI antigenically distinct from the parental line3~. Available evidence suggests that PfEMP1 is also responsible for binding to CSA. Antibodies raised to parasite line C~2 (which binds
CSA proteaglycan linkage GAGs such as CSA are linked by a protein backbone to the cell surface. Although many HS proteoglycans have been described, fewer CS proteoglycans • 14. In the vascular syste,~ the only have been identified abundant CSA containing proteoglycan is thrombomodulin (TM), a transmembrane protein of Mr 75000, which may have a CSA chain attached at one of two sites36,37. CSA contributes to the biological role of TM, principally binding and inactivating the circulating coagulant thrombin and activating the anticoagulant protein C. TM is found widely on vascular endotheliom (~100000 molecules per cell), and is expressed at high levels on the syncytiotrophoblasts lining the maternal blood sinusoids-', a site to which PRBC have recently bee//shown to bind in the ].alacenta23. No other known adherence receptors for PRBC, and no other abundant CS proteoglycans, have been identified on these placental cells. We have recently found that PRBC bind to TM, and that binding is dependent on the presence of the C:~A chain of TM 39. Free CSA or TM with it~ CSA attached could inhibit binding to TM/CSA, and minima! binding was seen to TM that lacked CSA, or that had CSA cleaved off by chondroitinase ABC• PRBC also bound to A549 lung carcinoma cells, which express CSApositive TM, and binding was inhibited by free CSA. Binding to TM/CSA and to A549 cells could be reversed by free CSA at a concentration of 10 tzg m l - ' . TM may therefore be. the natural receptor for CSA binding of PRBC, both from its distribution and from ~he interactions of PRBC with TM on cells and on plastic.
implications for studies of cytoadherence Given that CSA is expressed on ~oth CHO cells and C32 melanoma cells, the interpretation of studies using these cell types is more complex than previously perceived. CD36- and ICAM-l-tranafected CHO cells developed by Hasler et al. 4° will also support CSA-dependent binding (K. Devem and G. Brown, Parasitotogy Today, vol. 13. no 2. 1997
Reviews i
unpublished), and binding to C32 melanoma cells may be mediated by CD36, CSA or ICAM-1. This may, in part, explain the lack of clear correlation between binding phenotype and disease severity seen in studies using these cells. Adherence models such as these mav be useful in identifying novel receptors and in investigating cooperativity between receptors under controlled conditions, but there are constraints on the use of such artificial systems in field studies or in the laboratory. Chondroitin sulphate A as a placental receptor Pregnancy is associated with increz/sed stisc~:ptibility to clinical malaria in women living in endemic areas. Malaria in pregnancy may cause ir:a~ernal morP,idity and mortality, and placental infection is associated with an increased risk of low birth weight and mortality in the neonate 41,42.Clinical problems and placental changes are most pronounced in first pregnancies. A number of immunological explanations have been proposed for these observations, but none to date has provided a convincing rationale for the particular predisposition of primigravid women to malaria-related complications. The recent findings of Fried and Duffy that PRBC from the placenta tended to bind to CSA and not to other receptors, and that pregnant women had circulating PRBC that could bind to CSA a n d / o r CD36, whereas PRBC from nonpregnant women bound only to CD36, suggest that adhesion to CSA is critical in the evolution of placental parasitization 23. Indeed, PRBC from infected placentae could bind to trophoblasts from a normal placenta. These findings are consistent with our hypothesis that parasites expressing particular forms of PfEMP1 can bind to particular substrates allowing emergenc~ of different binding phenotypes as parasites undergo antigenic variafion35. CSA binding parasites sequester in the placenta, but sequester poorly, or not at all, elsewhere in the body. When a woman is first pregnant: she presents novel or more abundant receptors that select for PRBC of this phenotype, which are expressing particular antigenic variants binding to CSA. Such variants may not sequester efficiently or be s~lected to establish ongoing infection in the nonpregnant host, possibly because of a lack of appropriate receptors. Earlier encounters with malaria in the nonpregnant state would not have favoured parasites of this phenotype, hence, prior to the first pregnancy, it is unlikely that specific immune responses to this variant would have developed. Over the course of a pregnancy, and with subsequent pregnancies, the degree of placental infection tends to fall43; this is consisten~ with the gradual development of variant specific immvnity Lo CSA binding PRBC. This hypothesis to explain the i ole of placental sequestration in maternal malaria was first articulated by Prof. Marcel Hommel, in the Interne: Malaria Discussion Group. Chondroitin sulphate as an endothelial recel:.tor Although this is intriguing, there is also ~ eidence that PRBC may bind to CSA on endothel~um elsewhere. Some parasite lines bind to primary human lung endothelial cells (HLEC) isolated from a single donor, and binding is markedly inhibited by CSA, consistent with a role for this receptor in ~questration in the lung2~'4'L There is also evidence of CSAPcrasit.olo~, Tnday. vol. 13. no. 2. 1997
II
dependent binding to brain capillary" endothelial cells45 (I.W. Sherman, pers. commun.). Adherence to these cells was decreased by chondroitinase treatment (by 30%) or by free CSA (by 50%) when a CSA-binding line was used, but not when a CD36-binding line was used. We have found CSA-dependent binding to HUVEC ~c,n-~some sources s9 (Fig. 3L We need to know more abort the distribution of CSA in the vascular space, and ~hether there is significant CSA on brain and lung endoth~Jial cells in most or all people. Further, do all women express CSA accessible to PRBC in the placenta, and what happens to CSA expression with malaria infection? We know the glycosylation and level of expression of TM can vary between different vessels4.. Host differences in receptor expression also could affect malaria susceptibility. Therapeutic uses of chondroitin sulphates Chondroitin sulphates have been administered to humans orally, intramuscularly and intravenously without detectable r;dverse effects and .J~d ~n treatment trials for osteoarthritis. CSA concentrations of 10 pg ml -~ or more were seen for several hours after oral admimstration of 3 g of CS, and for approxir..~ately one hour after intravenous administration of 0.5 g (Ref. 42j, suggesting that levels of free CSA that are able to reverse adhere~.ce it1 vitro are achievabie ii'~ rive. If binding to CSA by isolates from pregiiant women correlates with disease severity, then studies will be required to determine whether CSA could prove beneficial in the treatment of malaria complications, in areas where maternal a n d / o r foetal morbidity associated with malaria is high. The effect of such therapy in the development of immunity to placental infection, in particular, would require close monitoring. Further questions, &ature directions The di~,covery of CSA as a novel adhesion receptor, and evidence that it may be involved in adherence of PRBC to cells from placenta, umbilical vessels, lung and brain (Fig. 3) gives us a new tool to investigate the pathoge~esis of malaria complications related to pregnancy and possibly the development of cerebral malaria and respiratory, distress and pulmonar:., oedema associated with malaria. Future studies should focus on several areas: (1) The newly described role for CSA binding in the development of placental infection23 will need to be confirmed in other settings. Is CSA-mediated placental sequestration crucial to the development of malaria in pregnancy? Does bindivg of placental or maternal P R B C to CSA correlate w~.h love hi:h,' weisht in the neonate or with mater:M morbidity? PRBC from some pregnant wome~ appear i~'t to adhere to CSA. This suggests that sequestration o:Itside the placenta may also be important. Evaluation of the contribution of this finding to the development of malaria complications in pregnancy will lead to greater understanding of the possible benefits and consequences associated with therapy targeting this interaction. (2) Does placental sequestration occur throueh adherence to CSA alone, or are humeral factors important, as they may be in rosetting~? Studies of placental 73
Reviews
...
Brain: endothelial
cells
Lung: endothelial cells
Umbilical vein: endothelial cells
Placenta: trophoblast ~
@ I
I
I
;
~ig. 3. Possible roles of CSA in sequestration of Plasmodium folciparurn.infected erythrocytes. There is now evidence ef CSA-dependent binding to endothelial cell lines from (I) lung44, (2) umbilical vein (S.J. Rogerson, unpublished), and 0 ) trophoblast cells from placenta~. as well as brain capillary endothelial cells (see text). The r~la~ive importancc ~ ~hese sites is still not known.
pathology have suggested that most PRBC seen in the placenta are not in fact adherent to the trophobiast surface but are within the villous spaces49. Is this an
(4) Can we identify a specific ligand within PfEI~,~P1 (or another protein) responsible for binding to CSA? Identification of such a ligand could pave the way for a vaccine based on preventing complications of malaria that are associated with CSA binding. (5) Could CSA therapy reverse sequestration in affected individuals? Reversal of binding is seen in vitro, and CSA appears to be a safe and non-toxic therapeutic agent. Obtaining animal safety and efficacy data is a necessary first step ~n il~,'estigating this possibility. The effects of CSA on maternal immunity would need to be closely monitured. (6) Might CSA binding represent a link between placental infection and p u l m o n a r y disease such as the p u l m o n a r y oedema sometimes witnessed in n o n - i m m u n e or semii m m u n e pregnant w o m e n with malariaS°? PRBC m a y bind to HLEC using CSA, which could be invo!ved in p u l m o n a r y sequestration. Post-mortem studies of w o m e n with fatal malarial pulmonary oedema should address this question. Conclusion The discovery that CSA can act as a receptor for adherence of PRBC, and emerging evidence that parasitized cells can adhere via CSA to endothelial cells of the placenta, lung a n d brain, open n e w avenues in our exploration of the m e c h a n i s m s of sequestration of P. falciparum. Understanding the molecular m e c h a n i s m s of adherence to CSA m a y give rise to n e w approaches to the treat~:qent of complications of malaria i~ pregnancy and childhood. Acknowledgements
SiR is a Wellcome Trust Research Career
Development Fellow in Clinical Tropical Medicine. The work of the Immunoparasitology Unit ~ssupporXedby the National Health and Hedical Research Council of Australia. We tha, k RosemaW van D6el for performing the electron microscoDy. EmanuelaHandman and John Reederfor reviewingthe manuscript. ~nd Irwin Shear.anand J0rgGys,n for"sharingresultswith us.
artefact of the birth process (in which case it m a y not be seer, in placentae from w o m e n delivered by Refecen.es c~esa,r-a'~'! section) or of the w a y in which tissue w a s 1 R u b ~ , D,D. et aL (1985) lIhrombospondin binds falciparum malaria parasitized erythrocytes and may mediate cytoad:aerhand!,.--:fl,or do local placental i m m u n e response~ help ence. Nature 318,64-66 trap PRBC in ihe F]acenta? 2 Bamwell, J.W., Ockenhou.~, C.F. and Knowles, D.M. (1985) (3~ IS binding to CSA associated with ce~'ebzL' Monoc|oaal antibody OKMb inhibits the in vitro bindin8 of malaria or other forms ~f severe disease, especially in " Plasmodium falciparum-infeded erythrocytes to monocytes, endolhei[al,and C32 melanoma cells.].hnmunol. 135,~94--3497 children? Detailed studies ~,~possible association,~ be3 Berendt, A,R. et al. (1989) Intercellular adhesion molecule-1 is tween C,C..Abinding and particu!ar disease s y n d r o m e s an endothelial cell adhesion molecule for Plasmodimn falcipwill be oi ~'~!erest. atom. Nature 341, 57-,59 ?4
Parasitology Toda),: vo/. 13, no. 2. 1997
Reviews 40ckenhouse, C.F. cl aL (1992) Human vascular endothelial cell adhesion receptors for Plasmodi,m [alcipar, m-infected erythrecyles: roles for endothelial leukocyte adhesion molecule I and vascular cell adhesion molecule 1. I. Exp. Med. 176,1183-1180 5 Ho, M e: el. (19911 Clinical correlates of in vitro Plasmodium [alcip,lnim cytoadherence, hffcct, llmm/li. 59, 873~'q78 b, Johnsol~., J.K.e.~~!. (1(,~21 Cytoadherence of Plaamadiuni falcip~mmt-infected erythrocytes to microvascular endolhelium is zegulatable by cytokines and phorbol ester. ]. Infect. Dis. 167, 698-7O3 7 Chaiy~.roi, S.C. rl el. (19941 Multiple ligands for cytoadherence can be present simultaneously on the surface of Piasmodium ~alciparom-infected erylhmcytes. Pr(n" N.dl. A~id. Sci. /3. q. A. 91, 10805-10808 8 Chaiyaroj, S.C, ct el. 09941 A Plasmodium falcipar~ml isolate with a chromosome 9 deletion expresses a trypsin resistant cytoadherence molecule. Mol. Biochem. Parasitol. 67, 21-30 ~: Gay, F. et al. (19951 Isolation and characterization of brain microvascular endothelial cell~ from Saimiri monkeys: An in vitro model for sequestration of Plasmodiom [aleiparuminfected erythrocytes. J. Immmtt,L Meth. 184,15-28 I0 Rogerson, S.J. et el. (1995) Chondrnltin sulfate A is a cell surface receptor for Plasmodium [.dciparmn-infecled erythrocytes. I. Exp. Med. 182,15-20 ll Roberts, D.J. el el. (19931 Prolection, pathogenesis and phenotypic plasticity in Plasmodim:l falciparum malaria. ParasitoL Today 9, 281-285 12 Esko, I.D., Stewart, T.E. and Tayh:r, W.H. (19851 Animal cell mutants defective in glycosaminoglycan biosynthesis. ~rl)c. Natl. Acad. Sci. U. S. A. 82, 3197-3201 13 Bumol, T.F., Walker, L.E. and Reisfeld, R.A (19841 Biosynthetic studies of proteoglycans in human melanoma cells with a monoclonal antibody 1o a core glycoprotein of ehondroitin sulfate proleoglycans. J. Biol. Chem. 259, 12735-12741 14 Ihrcke, N.S. et al. (19931 Role of heparan sulfate in immune system-blood vessel interactions, hmmmol. Today 14, 500-505 15 Herrera, E.M. et el. (1994) Mediation of Trypa~io~,om~l crl~:i invasion by heparan sulfate receptors on host cells and penelfin counter-receptors on the trypanosomes. Mt,I. Bh,chem. Pm ,~itoL 65, 73-83 16 Love, D.C., Esko, J.D. and Mosser, D.M. (19931 A heparinbinding activi~' on Leishmania amastigotes which mediates adhesiolt to cellular proteoglycans. ]. Cell Biol. 123. 759-766 17 WuDurm, 13 and Spear, P. (19891 luitial interaction of berpes simplex virus with cells is binding to hepara~l sulfate. I. Vieol. 63, 52-58 18 Neyts, J. el al. (19921 Sulfated polymers inhibit the interaction of human cyiomega~ovirus with cell surface heparan sulfate. 'diroh,gy 18;9, 48-58 19 Cerami, C et el. t19921 The basolateral domain of the hep.~iocyte plasma ~.~mbrane bears receptors for the ciicumsporozoite protein o1 Plasmodium falcipar.,n..sporozoiles. Cell 70, 1021-1033 20 Pancake, S.I e# al. (1992) Malaria sporozoites and circumsporuzoile prolelos bind specifically to sulfated glycoconjugates. J. Cell Bh)l. 117,1351-1357 2i garamanos, N.K. et aL (19941 Determination of 24 variously sulfated galactosaminoglycan- and hyaluronan-derived disaecharides by high-performance liquid chromatography. A'taL Biochem. 221,189-199 22 Volpi, N., Cusmano, M. and Venturelli. T. (19q51 Qua|iialive and quantitative studies of hepario and ehondroitia sulfates in normal human plasma. Biochim. Bivphys. Acts 1243, 49-58 23 Fried, M. and Duffy, P.E. (19961 Adherence of Plasmodium fi;t:-iparum to chondroilin sulfate A in the human placenta. Science 272,1502-15~;4 24 Chaiyaroj, S.C. et el. (19961 Cytoadherence characler'.'siics of Plasmodimn falciparom isolates fmrd Thailand: evidence t,~r cbondroirin sulfate A as a cytoadherence receptor. Am. I Trap. Med. Hyg. 55, 76-~0 25 RelY.let, j.C. at at. (19941 Divetmity of ag~!.ltinafin.~ phenotype, eytoadherence and rosette-to .rm.h~g t~araeteristks of Plasmodium falciparutn isolates from Papua New Guinean children. Am. [. Trop. Meal. Hy3,. 5l, 45-55 26 Robert, C. et aL (19951 Chondroitin-4-sulphate (proteoglycanL a receplor for Plasntodiunt falciparuot-infected eryihrocyie adherence on brain microvascular endo,helial cells. R<.s Imm:,lol. 146, 383-393 Porosaology Today. vol. 13. no. 2. ;097
27 Coek~., BM. and Coppel, R.L. (19951 Cytoadhesion and fa!civ. arum malaria: going with the flow. Porasitol. Taday 91, 282-287 28 Cooke, B.M. et al. ~,dhesion of malaria-infected red blood celts to chondroitin sulfate A under flow conditions. Bha)d tin press) 29 Cooke, B.M. at al. (19941 Rolling and stationary cytoadhesion of r~d biuod ceils parasitized by Plasmodium falciparum: separate roles for 1CAM-l, CD36 and thrombospondin./~r. ]. H,t'm~o:H. 8,', 162-170 3!) B,,uch, ~. ct el. (1~161 Plasmv, dium yalcipan,n erythrocyte membrane protein 1 is a parasitized erythrocyte receptor for adherence In CD36, thrombospondin, and intercellular adhesion molecule 1. Proc. Natl. Acad. Sci. U. S. A. 93, 3497-3502 3, Su, X. ct aL (19951 The large diverse gene family var encodes proteins involved in cytoadherence and antigenic variation of Plasmodium [alcipa.nmt-infecled erythmcytes. Cell 82, 89 IOO 32 Baruch, D.I. ('t al. (19951 Cloning th~ 11. ~alcipantm gene encoding PfEMPI, a malarial variant antigen aud adherence receptor on the surface 0f parasitized hu man erythrocytes. Cell 82, 77-87 33 Smith, J.D. et aL (19951 Switches in expression of Plasmodium [alciparum vat genes correlate w."th changes in antigenic and cytoadherent phenotypes of infected erythrocytes. Cell 82, 101-110 34 Biggs, B.A. et al. (19911 Anfige~,'.",~arlalion in Plasmodhvn falciparum. Proc. Nail. Acad. Sci. U. S. A. 88, 9171-9174 35 Biggs, B.A. t't aL (19921 Adherence of infected erythrocyte~ to venular endolhelium selects for antigenic variants of Plasotodium f~lciparum. I. hmmmoL 149, 2047-2054 36 Salem, H.H. et al. (19841 Isolation and characterization of Ihrombomodulin from human placenta. I. Biol. Chem. 259, 12246-12251 37 Gerlitz, B. e ~. al. i1993) identification of the pred,aminant glycosamlnoglycan-attachment site in soluble recombinant human thrombomodulia: polenth'! regulation of funclionalfly by glycosyltransferase competition for serine4rL Biachem. L 295,131-140 38 Maruyama, 1., Bell, C.E. and Majer,s, P.W. (19851 Thrombomodulin is found on endothelium ef arteries, veins, capillaries, and iy.,nphalics, and on syncyiio'~rophoblast of human placenta. I. CeB BioL I01,363-371 3~ Rogerson, S.1. el al. Plasmodium falciparum: infected erythrocytes adhere to the proteoglycan thrombomodulin in static and flow based systems. Exp. Parasitol. (in press) 40 Hasler, T. et a]. (19931 An improved microassay for Plasm o d i , , , [alciparmn cytoadherence using stable transformauls of Chinese hamster ovary cells expressir.g CD36 or intercellular adhesion molecule-L Am. L Trap. Med. Hyg. 48 332-347 41 Brabin, 13J. (1983) An analysis of malaria in pregnancy in Africa. 81dl WHO 61,1005-1016 42 Menendez, C. (19951 Malaria durin~ pregnancy: a priority area Of malaria research and con~:rol. Para~itoL Today 11, 178-183 43 grabiu, B.J. I ,qt;-~) Epidemiology of infection in pregnancy. Roy. Infect. Drs. 7, 579-603 44 Muanza, K. et al. (1996l Primary culture of human lung microvessel endothelial cells: a useful in vitro model for studying Plasmodium /a!ciparun#-infected erythrocytes cyloadherence. Res. hmmmol. 147,149-163 45 Prudhomme, J.G. et aL (19961 Studies of Plasmodium falciparum cyto~dherence usir.g immortalized human brain capillary endothdial cells. Int. L Parasih)L 26, 647-655 46 Lin, J-It. et el. (1994) Modulation of glycosaminoglyean addition in naturally expressed and recombinant human thrombomoduli~. I, Biol. Ch:',t 76~, 2,~21-25030 47 Conte, A. et ,1L (1',911 Metabolic fate of exogenous chondroitin sulfate in man. Arzneimittd[orschunq 41,768-772 .18 Scholander, C. et el. (1996) Novel fibrillar structure confers adhesive property In malaria-infected eryihrocyles. Nat. Meal. 2, 21)4-2f)8 49 Yamada, M. el el. (19891 Plasmodium falciparum associated placental pathology: a light and electron microscopic and immunohistologic sludy. Am. J. Trop. Mcd. Hyg 41,161-168 50 Warrell, D.A., Molyr, eux, M.E. and Bealcs, P.F. (19901Severe and complicated malaria. 7runs. R. Soc. Trap. M,',/. Hyg. 84, 1-65
75