After invasion of red blood cells RBCs by merozoitesparasites matur - PDF document

After invasion of red blood cells (RBCs) by merozoites,parasites mature over an ~48-h period through ring stageto pigmented trophozoites and finally divide into daugh-continue further cycles of asexual replication. Some par-asites at the ring stage develop into male and female ga-metocytes, which may be taken up by a feeding mosquitowhere sexual reproduction occurs. Developmental timein the mosquito varies according to ambient tempera-tures, but is typically in the order of 7Ð14 days.Clinical features of P.falciparumP. falciparumfever, malaise, headache, and lethargy followed by spon-taneous recovery, which may occur even without drugtherapy. In a proportion of cases, severe disease results.Following repeated exposure to malaria, immune re-sponses eventually develop that protect against seriousdisease but do not convey sterile immunity; low-grade in-fections occur with few or no clinical symptoms.Clinical manifestations of severe P. falciparumare variable in nature and severity, and it can be difficultto clearly define syndromes in many cases [2]. A cerebralmalaria deaths and is characterised by unrousable comaoften with convulsions, but any degree of impaired con-sciousness may indicate cerebral involvement [2]. Severenormocytic anaemia is probably the second most com-mon presentation of severe P. falciparumprobably results from increased RBC destruction and re-duced erythropoiesis [3]. Respiratory distress carries avery poor prognosis and may be a consequence of fluidretention, but has been observed in individuals with anormal or negative fluid balance status, and metabolicacidosis [2, 4]. Renal dysfunction or failure, hypogly-caemia, circulatory collapse and shock, disseminated in-travascular coagulation and spontaneous bleeding, andacidosis can also occur [2]. Severe haemoglobinuria, orÔblack water feverÕ, as a consequence of severe intravas-cular haemolysis, is sometimes observed and can resultfrom immune lysis of quinine-sensitised erythrocytes, orfrom antimalarial therapy in individuals with glucose-6-phosphate dehydrogenase deficiency [2].Among malaria-exposed adults, pregnant women are par-ticularly susceptible to malaria, despite substantial im-munity prior to pregnancy, and the risk is highest in firstpregnancies [5, 6]. The major complications of infectionare maternal anaemia, which in turn increases maternaldeaths, and reduced infant birthweight from a combina-tion of intrauterine growth retardation and premature de-livery leading to excess infant mortality [5Ð7]. In someCMLS, Cell. Mol. Life Sci.Vol. 59, 2002Review Articlesettings maternal malaria may also cause spontaneousabortion or stillbirth [6].Histopathological findings in severeP.falciparum infection:parasite sequestrationPostmortem examination of individuals who die of cere-bral malaria typically reveals gross swelling of the brainwith ring or petechial haemorrhages associated with thein small blood vessels on histopathology [8Ð10] (fig.1A). Several studies have shown that in cases definedwith cerebral malaria on clinical grounds, parasite se-questration at postmortem examination is greater in the Figure1.()Histological section of tissue collected post-mortemfrom a Malawian child. RBCs infected with mature stages ofP. fal-(arrows) are packed tightly within a small vessel in thebrain. (Image courtesy of Dr Terrie Taylor, Wellcome Trust Re-search Laboratories, College of Medicine, Blantyre, Malawi.) (Large numbers of infected RBCs (long arrow) and inflammatorycells (short arrow) in the vascular spaces of an infected placentafrom a Malawian woman. (P. falciparum mobilised on a plastic surface. 260J. G. Beeson and G. V. BrownPathogenesis of P. falciparumbrain than other organs, and parasite density in cerebralvessels is higher than in the peripheral blood [8Ð12].However, studies linking sequestration in vessels withcerebral malaria have faced a number of practical diffi-culties such as clinical case definition and the delay be-tween death and autopsy, which may influence the find-ing of sequestered parasites given the cyclical nature ofthe development of P. falciparumblood stages. In someindividuals believed to have died from cerebral malaria,sequestered parasites have not been found. Parasite se-questration is also observed in many other organs andvascular beds such as the lung, liver, intestine and skin [9,13]. However, most interest has focussed on sequestrationin the brain in regard to understanding the pathology andpathogenesis of cerebral malaria.Parasite-infected RBCs also concentrate in large numbersin the placenta (fig. 1B) during pregnancy [14, 15], insome cases associated with prominent infiltrates ofmonocytes and macrophages [16], and are deleterious tofoetal development [5, 6]. The placenta provides a uniqueopportunity to study sequestration, because viable para-sites can be isolated relatively easily from infected tissuefollowing delivery rather than relying on examination oftissue post-mortem. Recent studies on

placental malariahave given us important insights into how organ-specificparasite sequestration might occur and is discussed later.The link between parasite sequestration and other formsof severe malaria is not clear.Role of sequestration in the pathogenesisof severe diseaseThe observation of large numbers of parasites accumu-lated in specific organs, such as the brain and the placenta,associated with adverse clinical outcomes does suggestthat organ-specific accumulation of parasites is importantin the pathogenesis of malarial disease. The downstreameffects of sequestration might include mechanical ob-struction of blood flow [17], leading to hypoxia, the focalrelease of parasite toxins and inflammatory mediators,y mediators,(fig. 2). Persistent focal neurological deficits similar tothose following a stroke, such as hemiparesis and ataxia,have been recorded in about 10% of children recoveringfrom cerebral malaria [19, 20], and more subtle effectsmay be more common [21].Toxic mediators,inflammatory responsesand metabolic disturbancesIndividuals with profound coma can regain conscious-ness very rapidly, and may show no neurological seque-lae, suggesting that the pathology of cerebral malaria hasa metabolic component [22]. Furthermore, organ-spe-cific sequestration of parasites does not account for allfeatures of severe malaria, such as severe malarialanaemia, which may result from parasite and immune-mediated RBC destruction and reduced eythropoiesis as-sociated with increased turmour necrosis factor and reduced interleukin (IL)-10 levels [3]. Se-questration of parasites may concentrate the metabolicdisturbances and inflammatory responses triggered bymalaria parasites to a particular organ, such as the brain,contributing to the development of severe clinical dis-ease. The role of inflammatory and metabolic distur-bances to severe malarial disease is only briefly reviewedhere, but has been extensively studied.Studies in animals and humans suggest an important rolefor cytokines in the development of severe malaria. TNF-appears to be involved in the development of severe murinemalaria [23], and higher TNF-levels were recorded inchildren that died of severe malaria in two African popula-tions [24, 25]. Nitric oxide (NO) has also been proposed tobe important in cerebral malaria, and the expression of in-ducible NO synthase (iNOS) is increased by proinflamma-tory cytokines and hypoxia [26]. However, studies of ni-trate levels as a marker for NO production in the peripheralblood and urine of infected individuals have revealed anegative association between nitrate levels and severemalaria in some studies [27, 28], and positive associationsin others [29]. Studies of NO production and iNOS ex-pression in postmortem tissue from the brain and other or-gans may help clarify the role of NO. Figure2.Model of proposed mechanisms involved in the pathogenesis of severe malarial syndromes. P. falciparum-infected RBCs maysequester in small blood vessels though a combination of direct adhesion to endothelial cells and the formation of rosettes or prising infected and uninfected RBCs. The resultant sequestration may obstruct blood flow in vessels, leading to focal hypoxia, and trig-ger inflammatory responses and metabolic disturbances. CMLS, Cell. Mol. Life Sci.Vol. 59, 2002Review ArticleThe existence of a malaria toxin that leads to severe dis-ease in the host is an attractive hypothesis that has led tothe identification of P. falciparumglycosylphosphatidyli-nositol (GPI) as a candidate toxin [30]. GPI purified fromcultured parasites was shown to induce cytokine release,fever and hypoglycaemia in mice, or death in primedmice, effects that are inhibited by specific antibodies.GPI also stimulates TNF-production and iNOS expres-sion by macrophages, and its effects on endothelial cellsinclude upregulation of parasite adhesion molecules, in-tercellular adhesion molecule (ICAM)-1, vascular cellascular cellAcidosis and hypoglycaemia have been associated withworse outcomes in individuals with severe malaria infec-tions [33Ð35]. Acidosis is largely metabolic, attributableto increased lactate production by the host and parasitebiomass [33, 36]. Hypoglycaemia may arise from re-duced hepatic gluconeogenesis, increased glucose con-sumption by a large parasite biomass [36] or triggered byparasite toxins such as GPI [30].Parasite biomassSome investigators favour the proposal that the total par-asite biomass is important in the development of severemalaria, which may manifest in a variety of syndromes[36, 37]. The total biomass of parasites within the bodymay influence concentrations of parasite toxins and in-flammatory mediators and result in metabolic and bio-chemical disturbances contributing to severe disease. Se-questration of parasites, at any site, would contribute to agreater biomass of parasites. Using a combination ofmathematical models and clinical findings, associationshave been found between the estimated or predicted ma

ssof infected RBCs in an infected individual and severe dis-ere dis-Cytoadherence of P.falciparumThe observed sequestration of infected RBCs in the mi-crovasculature of various organs, and its association withclinical disease, has led to intense investigation of the ad-hesive interactions between infected RBCs and host en-dothelial cells. These phenomena form the focus of this re-view. Mature-stage infected RBCs can adhere to endothe- Mature-stage infected RBCs can adhere to endothe-have been identified that can act as receptors for the adhe-sion of parasitised erythrocytes in vitro (table 1). Exami-nation of postmortem tissue by electron microscopy sug-gests infected RBCs directly adhere to endothelial cellsvia electron-dense knobs on the RBC surface [8, 10].It is predominantly mature-stage parasites that can adhereat substantial levels in vitro, consistent with the observa-tion that the great majority of sequestered parasites aretrophozoites or schizonts [10, 12] (fig. 1). These formsare typically not seen in the peripheral blood of infectedindividuals [38]. As circulating parasites mature, antigensexpressed on the RBC surface enable adhesion to en-dothelial cells and sequestration in vascular beds [40].Table1.Host cell adhesion molecules for P. falciparum-infected RBCs and their possible clinical significance. ReceptorClinical significanceComments References CD36uncomplicated malariaÐnearly all isolates from nonpregnant individuals can adhere60, 61, 67severe malariaÐlittle expression in cerebral vessels13Ðsynergy with ICAM-1 to augment adhesion62Ðno clear association between adhesion to CD36 and disease severity ICAM-1cerebral malariaÐcerebral ICAM-1 expression and parasite sequestration colocalised13Ðboth positive and negative associations reported between adhesion 60, 61of clinical isolates and disease severityRosetting receptorssevere malariaÐrosette formation associated with complicated malaria in some 43Ð48CSAplacental infectionÐadhesion to CSA is a feature of placentally sequestered parasites55, 56other forms of diseaseÐCSA-dependent adhesion to brain and lung endothelial cells reported78HAplacental infectionÐadhesion to HA is a feature of placentally sequestered parasites58E-selectin?Ðlittle or no adhesion of clinical isolates in two studies60, 104Ðlittle expression in the cerebral vasculature, but increased in 13VCAM-1?Ðlittle adhesion of clinical isolates in two studies60, 104PECAM/CD31?Ðlow levels of adhesion of clinical isolates in one large study60Ðcerebral CD31 expression and parasite sequestration not colocalised13P-selectin?Ðno studies of adhesion in the context of clinical disease yet published 262J. G. Beeson and G. V. BrownPathogenesis of P. falciparumThose infected RBCs failing to express adhesive pheno-types are thought to be cleared from the circulation by thespleen. Recently, it has been demonstrated that the earlyring forms of infected RBCs can adhere to host cells atlow levels in vitro, and this may contribute to parasite se-y contribute to parasite se-Mature-stage infected RBCs can also adhere to unin-fected RBCs to form spontaneous rosettes in vitro. Ap-parent rosettes have also been observed in vivo by histo-logical examination of infected tissue collected post-mortem [9, 42], and this property is thought to contributeto microvascular obstruction and severe disease. Somestudies [43Ð45], but not all [46Ð48], have found a posi-tive association between rosette formation of peripheralblood isolates, or negative association with rosette-dis-rupting ability of serum, and severe clinical disease.Further adhesive phenotypes of infected RBCs have beendefined in vitro. Autoagglutination or clumping is theterm used to describe the adhesion of infected RBCs toeach other [49, 50] and a recent African study found anassociation between autoagglutination of peripheralblood isolates and severity of clinical disease [51]. Para-sitised RBCs may also adhere to leucocytes [52].As will be discussed below, infected RBCs are antigeni-cally diverse and clonally variant, enabling repeat and re-crudescent infections. Changes in antigenic phenotypeare associated with changes in adhesive properties (fig. 3), and these properties may vary greatly from oneindividual to the next, or from one infection to the next.Furthermore, the antigenic and adhesive properties of cir-culating parasites may be quite different to those se-questered at a particular site. These features have made itdifficult to reliably identify parasite virulence factors as-sociated with specific clinical syndromes.Intraerythrocytic development of parasites markedlyreduces the deformability of the RBC membrane, which may also contribute to the sequestration of matureparasite stages [53, 54], particularly in combination with adhesive processes. The early stages or ring formsof infected RBCs show reduced cellular deformability,but by the mature stage infected cells are very rigid, having lost the properties that normally enable RBCs

to pass through the narrow lumen of capillaries [53].However, histological studies of postmortem tissue stered in postcapillary venules [10] rather than arteriolesformability do not account for the bulk of parasiteReceptors for parasite adhesionNumerous host molecules have been identified that can(table 1), but the role of each of these in the pathogene-sis of malarial disease remains largely unclear. Severalstudies have implicated adhesion to chondroitin sul-fate A (CSA) in placental malaria [55Ð57], and more re-cently hyaluronic acid (HA) has also been identified as areceptor that is probably involved in this process [58, ICAM-1 [60, 61], which are both widely distributed invascular beds and likely to be important in infection andSequestration in an organ probably involves multiple re-ceptors, and different combinations of specific receptorsfor adhesion may determine the site at which parasites ad-here and accumulate. For example, the phenotypes of par-asites sequestered in the placenta are quite different fromparasites infecting nonpregnant adults and children [55,56, 59]. Synergistic effects that augment adhesion havebeen observed when multiple receptors are expressed ona cell surface [62]. Sequestration is probably a multistepprocess involving initial attachment and rolling beforetethering occurs, akin to adhesion of leukocytes. Thisprocess has recently been directly observedin vivo insubcutaneous vessels of human tissue grafted onto SCID Figure3.() Schematic representation of clonal antigenic varia-P. falciparuminfected-RBCs. A single infected RBC mayundergo spontaneous antigenic variation by switching between theexpression of different genes, encoding PfEMP1, conveyingdifferent antigenic and adhesive properties to the cell. The names ofgenes used are examples only. () Predicted domain organisa-tion of PfEMP1 encoded by genes. All PfEMP1 species arethought to have a head structure comprised by a DBL1domain, followed by a variable number of additional DBLdomains, or less commonly, additional CIDRs. Domains are namedin numerical order from the N-terminus and according to sequencehomology with other known genes classified as types. DBL, Duffy-binding-like; CIDR, cysteine-rich interdomainregion; ATS, acidic terminal sequence. Parasitised RBCs can bind to CD36 in vitro [64Ð66], andP. falciparum falciparumdisease pathogenesis is not established. In two largeAfrican studies, there was no association between highlevels of adhesion to CD36 and severe malarial syn-ere malarial syn-in cerebral vessels, but is expressed in other organs suchas lung, kidney, liver and muscle, and no association hasbeen found between parasite sequestration and CD36 im-munolabelling in postmortem tissue of brains from indi-viduals who died of severe malaria [13]. However, CD36does act synergistically with other receptors to augmentadhesion [62]. One African study found that mutations inthe CD36 gene were associated with increased risk of se-vere malaria [68], whereas another African study foundAfrican study foundInfected RBCs can also adhere to CD36 on platelets toform clumps [50], and to CD36 on monocytes [70] anddendritic cells [71]. Parasite interaction with CD36 onmonocytes appears important in phagocytosis and clear-ance of parasitised RBCs [70], whereas the adhesion ofparasites to dendritic cells appears to have immunosup-pressive effects [71].ICAM-1 is a member of the immunoglobulin superfam-ily and supports parasite adhesion in vitro and in vivo [63, 72]. It is widely expressed in vascular beds in vivo,such as brain, liver, kidney and lung, and expression is in-creased in malaria [13]. The parasite binding region onICAM-1 has been mapped to the junction of the first andsecond immunoglobulin-like domains [73]. ICAM-1 cansynergise with CD36 to augment adhesion when the tworeceptors are coexpressed on the surface of endothelialace of endothelialSome evidence supports a role for ICAM-1 in the devel-opment of severe malarial disease, particularly cerebralmalaria. Studies of postmortem tissue found significantlevels of ICAM-1 expression in cerebral vessels and par-belling [13]. A large study conducted in Kenya, using iso-lates from the peripheral blood of children with well-defined clinical syndromes, found some association be-tween severe malaria in children and parasite binding toICAM-1 in vitro [60]. However, a similar large study inMalawian children found a negative association betweensevere disease and ICAM-1 adhesion [61]. Interestingly,a specific mutation in the ICAM-1 gene was associatedwith increased disease severity in an East African popu-lation [74], but in West Africa the opposite associationwas found [75].Studies in several populations in Africa have implicatedCSA as a key receptor for adhesion of infected RBCs inthe placenta, discussed in more detail below. Parasitesfrom infected placentas typically adhere to CSA, and in-pendent manner [55Ð57]. CSA can act as a cell adhesionmolecule for infected RBCs in

static and flow-based as-says [76Ð78] and parasite adhesion is strongly depen-sulfation of the saccharide chains [76, 79,80]. The CS proteoglycan thrombomodulin, present on arange of vascular surfaces, can support parasite adhesionand may be an important receptor in vivo [81, 82]. Al-though CSA and thrombomodulin have been detected invascular beds such as the brain [83], there is no consis-tent association between adhesion to CSA and severeclinical disease in children and nonpregnant adults [61,HA has only recently been identified as a receptor forparasite adhesion [58] and appears to be important for se-in this review. Parasite isolates from infected placentastypically bind to HA, in addition to CSA [58]. Findingsficity of binding to HA and CSA [58], both of which areexpressed on the placental lining [86Ð88]. The speci-ficity of adhesion to the two GAGs was established by theuse of defined oligosaccharide fragments and enzymaticdegradation of the GAG receptors and parasite ligand[58, 89]. HA is also expressed on microvascular endothe-lial cells [90] and may therefore act as a receptor for se-questration in a number of organs. However, parasite iso-lates from children with and without severe malaria ad-here to HA less frequently and at lower levels than toels than toRosetting receptorsRosettes of infected and uninfected RBCs are observedP. falciparumto involve several RBC molecules. Heparan sulfate (HS)proteoglycans on the surface of uninfected RBCs can actmation of rosettes [91], and with many isolates, rosettescan be disrupted by heparin or HS [92Ð94]. HS is thoughtto be present on many, if not all, cell surfaces and hasbeen identified on endothelial cells. It remains anuntested possibility that HS proteoglycans in the mi-crovasculature might act as receptors for the sequestra-tion of infected RBCs in vivo.Rosette formation may also involve complement receptor1 (CR1) on the surface of RBCs. Red cells deficient inCMLS, Cell. Mol. Life Sci.Vol. 59, 2002Review Article CR1 do not form rosettes, and soluble CR1 or antibodiescan inhibit rosette formation in a range of isolates [95,96]. Furthermore, blood group antigens A and B can alsoact as coreceptors in rosette formation, and isolates showdifferent rosetting rates and rosette sizes when culturedwith erythrocytes of different blood groups [97, 98].Other receptorsA number of other host molecules have been identified invivo, including VCAM-1, E-selectin, P-selectin, CD31/integrin [99Ð103].InfectedRBCs have also been shown to bind normal im-munoglobulins and may play a role in rosette formationmationAvailable data do not generally support a role forVCAM-1 and E-selectin in disease pathogenesis. New-bold et al. reported that adhesion to E-selectin andVCAM-1 was uncommon and generally occurred at lowlevels in vitro, and was not associated with clinical syn-dromes among clinical isolates from African donors[60].Similarly, a Thai study reported no adhesion of 19clinical isolates to VCAM-1 and little binding to E-se-lectin [104]. The expression of E-selectin in the cerebralvasculature is sparse, but expression is increased inmalaria and was associated with parasite sequestration inone study [13]. The role of P-selectin in pathogenesis re-mains unknown, but adhesion of clinical isolates hasbeen reported [100]. Parasite adhesion moleculesP.falciparum erythrocyte membraneprotein 1 (PfEMP1)PfEMP1 appears to be the principal adhesive ligand of in-fected RBCs, having recently been shown by several in-dependent groups to mediate adhesion to a range of can-CSA and CR1 [91, 95, 105Ð107]. It is a clonally variantprotein encoded by genes, it is important in antigenicvariation and immune evasion, and is clustered on thesurface of infected RBCs in knoblike structures compris-and other proteins. Targeted disruption of the KAHRPgene abolishes knob formation [108], and knob-negativeinfected RBCs show reduced adhesion to receptors underconditions of physiologically relevant flow [108].Vargenes encoding PfEMP1 comprise a variable numberof cysteine-rich extracellular domains (fig. 3) that havehomology to the P. falciparummolecule EBA-175, whichbinds glycophorin A during invasion of erythrocytes bymerozoites, and the Duffy blood group antigen-binding264J. G. Beeson and G. V. BrownPathogenesis of P. falciparumP. vivax [109]. The intracellular region, termedthe acidic terminal sequence (ATS), is highly conserved.The first Duffy binding-like (DBL) domain and the adja-cent cysteine-rich interdomain region form a head struc-ture, which is relatively conserved between different genes. Recent findings have implicated the head region inmultiple adhesive properties [110]. A variable number ofadditional DBL domains follow downstream, having lowdegrees of homology with one another or between differ-genes. There is also substantial sequence diversityin the interdomain regions. In many genes were first sequenced, predicted DBL do-mains were named according to their sequential positionend

(i.e., DBL1, DBL2, DBL3 and so on) andcharacteristic signature sequences [109]. Subsequently,many more genes have been sequenced from differentisolates which has enabled a reclassification of DBL do-mains in to five categories on the basis of sequence ho-mology (gy (used here.Various domains of PfEMP1 from a number of parasiteisolates have been implicated in adhesion to several re-ceptors (table 2). Adhesion to CD36 appears to be medi-ated by the CIDR1domain [105, 112Ð114], and a re-gion of this domain that shows substantial sequence ho-mology between different parasite isolates that bindCD36 has been implicated in adhesion [105]. However,domain does not appear to bind CD36 ex-clusively, as recent studies have suggested a role for thisregion in adhesion to CSA, CD31 and HS in some para-site isolates [106, 110, 115, 116]. Adhesion to CSA ap-[106, 117], but in some isolates adhesion may also in-volve the CIDR1e the CIDR1Table 2.Parasite ligands implicated in adhesion to host receptors ReceptorPfEMP1 domainOther ligandsReferences CD36CIDR1band 3105, 114, 122clag9125sequestrin121ICAM-1DBL2/typeno107CSADBL3/typeno106, 117no106, 115, 116HSDBL1no91CR1DBL1no95CD31/CIDR1no110PECAMDBL2/typeno110TSPunknownband 3118, 123 Note: The binding properties of the PfEMP1 domains vary fromone isolate to the next. Therefore any single domain from a partic-ular PfEMP1 species may not bind to all of the receptors listed. Antigenic variation of P.falciparumP. falciparum-infected RBCs have been shown to undergoclonal antigenic variation in vitro and can potentiallyswitch to different antigenic phenotypes at rates of up to2% per generation [49, 126]. Although direct evidencefor antigenic variation in vivo is lacking for P. falciparumPlasmodium knowlesiwlesiPlasmodium fragile[128] in monkeys, and Plasmodium chabaudi[129]. An interesting aspect of studies of antigenic varia-P. falciparumwas the finding that switching ofantigenic types was associated with changes in adhesiveproperties. When cloned isolates were selected for adhe-sion to endothelial cells, changes in were observed in theantigenic type as detected by agglutination assays [49,130]. The observed comodulation of antigenic type andparasite adhesion was subsequently explained by theidentification and sequencing of immunodominant and adhesive protein PfEMP1 on theerythrocyte surface.Clonally variant antigens on the RBCsurface:PfEMP1,Rif and StevorPrior to the identification of genes, substantial evidencesuggested the involvement of PfEMP1 in antigenic varia-tion and adhesion of infected RBCs. PfEMP1 was first de-fined as a high molecular weight protein that could be la-belled on intact infected RBCs, or degraded by trypsincleavage, suggesting a cell surface location [131]. PfEMP1can be immunoprecipitated by human hyperimmune serumconsistent with its location on the red cell surface where itwould be exposed to host immune responses. Switchingantigenic types of clonal isolates invitro was associatedof specific PfEMP1 types by agglutinating sera [126]. Sup-porting a role for PfEMP1 in cytoadherence, selection ofinfected RBCs for increasing levels of adhesion or for dif-ferent adhesive phenotypes resulted in variations in proteinsize, and trypsin cleavage of PfEMP1 was associated withloss of cytoadherence [49, 130, 132].The molecular basis for antigenic variation and adhesionP. falciparumwas revealed by the identification of genes, a large family of up to 50 genes distributedthroughout the genome [109, 112, 133] (fig. 3). Antiseravar-derived proteins reacted with PfEMP1on Western blots or by immunoprecipitation and labelledthe surface of infected RBCs in a variant-specific man-ner. Analysis of a series of cloned isolates with differentadhesive and antigenic properties showed expressionto be specific for the different variants [133]. Other in-vestigators have also demonstrated that changes in anti-genic and/or adhesive phenotypes are correlated with theexpression of specific ic CMLS, Cell. Mol. Life Sci.Vol. 59, 2002Review ArticlePfEMP1extracted from whole infected RBCs has beenshown to directly bind ICAM-1 and may be mediated by-type domains [107, 118]. The DBL1of PfEMP1 has been shown to bind to the surface of un-ace of un-Currently it is unclear how the presence of consensus mo-tifs in DBL and CIDR domains relates to parasite biology,such as immune recognition and cytoadhesion. For exam-lates form rosettes or adhere to HS. Similarly, not all iso-lates expressing to CD36, and this domain has been linked to other adhe-sive properties. The presence of a DBL does not necessarily indicate an ability to bind CSA. Re-cently, it was shown that a change of only three aminoacid residues in the prosed active region of the CIDR1domain can be sufficient ablate binding to CD36 [120],suggesting that a detailed knowledge of binding interac-tions is needed before it is possible to predict adhesiveproperties on the basis of known sequence.Other ligandsOther molecules have been propose

d to act as ligands forcellular adhesion. Their role remains somewhat unclearin light of the substantial evidence implicating PfEMP1in adhesion, they may act in association with PfEMP1.oclonal antibody led to the identification of a large mole-cular weight protein associated with adhesion to CD36,termed sequestrin [121]. To date, a direct interaction be-tween sequestrin and CD36 has not been reported, nor hasPeptides derived from parasite-modified band 3, a redcell protein, have been shown to inhibit adhesion to CD36[122], and antibodies targeting two putative exofacialexpressing CD36. Band 3 has also been suggested to bindTSP [123], but involvement in adhesion to other receptorshas not been described. The relationship to PfEMP1-me-diated CD36 adhesion is unknown.P. falciparummaintained in vitro may spontaneously lose[124], and this activity was mapped to a region termedclag9(cytoadherence-linked asexual gene) [125]. Tar-geted disruption of clag9by transfection resulted in lossclag9may encode a surface proteinthat is directly involved in mediating adhesion to CD36,clag9protein were found to inhibitadhesion to CD36 and labelled the surface of infected Recently, additional multigene families, termed stevor(repetitiveinterspersed family), have been identified [136Ð138], buttheir possible roles in antigenic variation or adhesion arenot known. The family comprises about 200 genes en-coding proteins of 30Ð40 kDa that can be labelled on thesurface of infected RBCs and do appear to be targeted byyVariant-specific immunityObservations in human studies support an important role for variant-specific immunity in protection frominfection and clinical disease. In several settings, it has been shown that convalescent serum, collected P. falciparumagglutinate the infecting parasite isolate, but generallynot other isolates. This suggests that specific antibodiesdevelop following infections that target variant antigenson the infected RBC surface [139Ð142]. Sera fromadults in endemic areas generally demonstrate a largerepertoire of variant-specific agglutinating antibodies,whereas children, who are more susceptible, possess an-tibodies that agglutinate a limited number of differentvariants. By association, this suggests that variant-spe-cific agglutinating antibodies, which appear to targetPfEMP1, may be involved in protection from clinicalSeveral African studies have directly tested this hypothe-sis [142Ð144]. In the Gambia, a number of potential in-dicators of immunity to blood-stage antigens were mea-sured, but only agglutinating antibodies were predictiveof protection from clinical malaria in children [143]. Amore recent large study in Kenya also found a significantassociation between agglutinating antibodies targetingantigens on the surface of infected red cells and protec-tion from clinical disease [142]. Children were generallyinfected with variant types against which they did notAn unresolved question is to what extent immunity isvariant specific and achieved by exposure to a finite setof antigenic variants, or reliant on the development of im-munity to conserved, and possibly poorly immunogenic,epitopes on the infected erythrocyte surface. Support forthe latter comes from the report of pan-agglutinating an-tibodies able to agglutinate virtually all variant types orariant types ories have not reported this phenomenon [140, 141]. Inter-estingly, sera collected from one geographic area are ableto agglutinate some isolates from quite separate regions[145], suggesting that there may be conserved epitopesamong different variants or that some antigenic variantsmay have a wide distribution. It was recently shown thatcross-reactivity to a number of different variants, sug-gesting the presence of conserved epitopes that may besuitable vaccine targets [146].Malaria during pregnancyThe pathogenesis of maternal malaria is worth specialmention, as recent studies have provided significant in-sights that illustrate how adhesion and antigenic variationP. falciparum-infected RBCs are important in the devel-opment of specific forms of malaria disease, and is re-viewed in detail elsewhere [59]. Two observations are par-ticularly striking about malaria during pregnancy. First, in-fection typically leads to the accumulation of vast numbersof infected RBCs in the placental blood spaces (fig. 1B), atblood and, second, pregnant women are much more sus-ceptible to malaria than their nonpregnant counterparts,suffering more frequent and severe malaria infections.Substantial findings now point to the importance of spe-cific receptor-mediated adhesion in the sequestration ofinfected RBCs in the placental blood spaces, althoughother mechanisms may also be involved [59, 147]. Sev-eral African studies have shown that parasites accumu-lated in the placenta typically adhere to CSA, which is ex-pressed on the placental vascular surface, and infectedmanner [55Ð57, 148]. By contrast, infected RBCs iso-lated from nonpregnant adults or children

typically ad-y ad-56]. Accumulation of parasites in the placenta also ap-pears to involve adhesion to HA. A recent African studyshowed that placentally sequestered infected RBCs typi-cally bound to HA, as well as CSA, rather than other re-ceptors [58]. Similar to observations with CSA, adhesionto HA was not typical of parasites collected from chil-dren, and was less common among circulating parasitesthan those sequestered in the placenta. Therefore, circu-lating parasites that express an ability to adhere to CSAand/or HA, and possibly other receptors yet to be identi-fied, would preferentially accumulate in the placentalblood spaces. The ability of parasites to adhere to multi-ple receptors may augment adhesion and sequestration,or may be essential in determining precisely in which vas-CSA, HA or syncytiotrophoblasts using in vitro assays[58, 148], suggesting that other mechanisms may be in-volved.Of further interest is that these placental-binding parasitetypes appear to be antigenically different from serotypesor variants that typically infect nonpregnant adults andchildren [55, 149]. As discussed earlier, variant-specificantibodies form an important component of the naturallyacquired protective response to malaria infections.266J. G. Beeson and G. V. BrownPathogenesis of P. falciparum Adults, including pregnant women, have a broad reper-toire of variant-specific antibodies to parasite variants in-fecting children and nonpregnant adults [55, 139, 140].However, antibodies to parasites infecting pregnantwomen are uncommon or rare among men or women whoare in their first pregnancy [55, 149].Following severalpregnancies, many women develop antibodies specifi-cally against placental parasite isolates [55, 149], or par-asites selected for adhesion to CSA [150]. These variantspecific antibodies appear to primarily target PfEMP1[59]. It appears that a switch to a placental-binding phe-notype is associated with switching to novel antigenicphenotypes to which individuals are not exposed prior topregnancy. Thus, at first pregnancy a woman is most sus-ceptible, but following exposure to these variants duringpregnancy, immune responses may develop that protect insubsequent pregnancies. Aside from the development ofvariant-specific antibodies, women also develop antibod-ies that can block the adhesion of parasites to CSA, andthis was associated with protection from malaria in onestudy [149]. As specific domains of PfEMP1 have beenshown to be responsible for mediating the adhesion of in-fected RBCs to CSA, there may be opportunities to de-velop therapeutic or preventative strategies, such as vac-cination, targeting the adhesion and subsequent accumu-P. falciparumcan lead to a diverse range ofclinical syndromes, and the pathogenesis of disease ap-pears to involve a combination of host and parasite fac-tors. Substantial evidence points to the importance of re-ceptor-specific adhesion and resultant accumulation ofinfected RBCs in the vasculature of organs such as thebrain and placenta. Clinical and laboratory studies haveidentified a number of potential host receptors, such asimplicated in cerebral malaria and CD36. Further inves-tigation is needed to clarify the role of other known adhe-sion receptors in the development of disease and whetheradhesion to specific receptor combinations is an impor-tant determinant of organ-specific sequestration and dis-ease. A greater knowledge of these events in vivo and theidentification of parasite ligands and specific adhesivemotifs may lead to opportunities for interventions to treatsevere malaria and its complications, or to prevent malar-ial disease through vaccination.Acknowledgements.Thanks to Stephen Rogerson for providingcomments on the manuscript, and to colleagues in the Departmentof Medicine, University of Melbourne, for their help and advice. Fi-nancial support was provided by the National Health and MedicalResearch Council of Australia, and the Royal Australasian Collegeof Physicians through a Cottrell Fellowship to J. G. B.1World Health Organisation (1997) World malaria situation.Weekly Epidemiol. Rec. 269Ð2742Warrell D. A., Molyneux M. E. and Beales P. F. (1990) Severeand complicated malaria. Trans. R. Soc. Trop. Med. Hyg. 1Ð653Menendez C., Fleming A. F. and Alonso P. L. (2000) Malaria-related anaemia. Parasitol. Today 469Ð4764English M., Waruiru C., Amukoye E., Murphy S., Crawley J.,Mwangi I. et al. (1996) Deep breathing in children with severeAm. J. Trop. Med. Hyg. 521Ð5245Brabin B. J. (1983) An analysis of malaria in pregnancy inAfrica. Bull. WHO 1005Ð10166McGregor I. A. (1984) Epidemiology, malaria and pregnancy.Am. J. Trop. Med. Hyg. 517Ð5257Granja A. C., Machungo F., Gomes A., Bergstrom S. and Bra-bin B. (1998) Malaria-related maternal mortality in urbanMozambique. Ann. Trop. Med. Parasitol. 257Ð2638Aikawa M., Iseki M., Barnwell J. W., Taylor D., Oo M. M. andHoward R. J. (1990) The pathology of human cerebral malaria.Am. J. Trop. Med. Hyg. 30Ð379Pongp

onratn E., Riganti M., Punpoowong B. and Aikawa M.(1991) Microvascular sequestration of parasitized erythro-cytes in human falciparum malaria: a pathological study. Am.J. Trop. Med. Hyg. 168Ð17510MacPherson G. G., Warrell M. J., White N. J., LooareesuwanS. and Warrell D. A. (1985) Human cerebral malaria. A quan-titative ultrastructural analysis of parasitized erythrocyte se-questration. Am. J. Pathol. 385Ð40111Oo M. M., Aikawa M., Than T., Aye T. M., Myint P. T., IgarashiI. et al. (1987) Human cerebral malaria: a pathological study.J. Neuropathol. Exp. Neurol. 223Ð23112Silamut K., Phu N. H., Whitty C., Turner G. D. H., LouwrierK., Mai N. T. H. et al. (1999) A quantitative analysis of the mi-crovascular sequestration of malaria parasites in the humanbrain. Am. J. Pathol. 395Ð41013Turner G. D. H., Morrison H., Jones M., Davis T. M. E., Looa-reesuwan S., Buley I. D. et al. (1994) An immunohistochemi-cal study of the pathology of fatal malaria. Evidence for wide-spread endothelial activation and a potential role for intercel-lular adhesion molecule-1 in cerebral sequestration. Am. J.Pathol. 1057Ð106914Walter P. R., Garin Y. and Blot P. (1982) Placental pathologicchanges in malaria. A histologic and ultrastructural study. Am.J. Pathol. 330Ð34215Yamada M., Steketee R., Abramowsky C., Kida M., Wirima J.,Heymann D. et al. (1989) placental pathology: a light and electron microscopic and im-munohistologic study. Am. J. Trop. Med. Hyg. 161Ð16816Ordi J., Menendez C., Ismail M. R., Ventura P. J., Palacin A.,Kahigwa E. et al. (2001) Placental malaria is associated withcell-mediated inflammatory responses with selective absenceof natural killer cells. J. Infect. Dis. 1100Ð110717Berendt A. R., Turner G. D. H. and Newbold C. I. (1994) Cere-bral malaria: the sequestration hypothesis. Parasitol. Today412Ð41418Newton C. R. and Krishna S. (1998) Severe falciparummalaria in children: current understanding of pathophysiologyand supportive treatment. Pharmacol. Ther. 1Ð5319Brewster D. R., Kwiatkowski D. and White N. J. (1990) Neu-rological sequelae of cerebral malaria in children. Lancet 1039Ð104320Molyneux M. E., Taylor T. E., Wirima J. J. and Borgstein A.(1989) Clinical features and prognostic indicators in paedi-atric cerebral malaria: a study of 131 comatose Malawian chil-dren. Q. J. Med. 441Ð45921Holding P. A., Stevenson J., Peshu N. and Marsh K. (1999)Cognitive sequelae of severe malaria with impaired con-sciousness. Trans. R. Soc. Trop. Med. Hyg. 529Ð534CMLS, Cell. Mol. Life Sci.Vol. 59, 2002Review Article 22Clark I. A. and Rockett K. A. (1994) The cytokine theory ofhuman cerebral malaria. Parasitol. Today 410Ð41223Grau G. E., Fajardo L. F., Piguet P. F., Allet B., Lambert P. H.and Vassalli P. (1987) Tumor necrosis factor (cachectin) as an1210Ð121224Grau G. E., Taylor T. E., Molyneux M. E., Wirima J. J. and Vas-salli P. (1989) Tumor necrosis factor and disease severity inmalaria. N. Engl. J. Med. 1586Ð159125Kwiatkowski D., Hill A. V. S., Sambou I., Twumasi P., Castra-cane J., Manogue K. R. et al. (1990) TNF concentration in fa-tal cerebral, non-fatal cerebral and uncomplicated 1201Ð120426Clark I. A. and Cowden W. B. (1999) Why is the pathology offalciparum worse than that of vivax malaria? Parasitol. Today458Ð46127Anstey N. M., Weinberg J. B., Hassanali M. Y., Mwaikambo E.D., Manyenga D., Misukonis M. A. et al. (1996) Nitric oxidein Tanzanian children with malaria: inverse relationship be-tween malaria severity and nitric oxide production/nitric oxidesynthase type 2 expression. J. Exp. Med. 557Ð56728Cot S., Ringwald P., Mulder B., Miailhes P., Yap-Yap J., Nuss-ler A. K. et al. (1994) Nitric oxide in cerebral malaria. J. In-1417Ð141829Al-Yaman F. M., Mokela D., Genton B., Rockett K. A., AlpersM. P. and Clark I. A. (1996) Association between serum levelsof reactive nitrogen intermediates and coma in children withcerebral malaria in Papua New Guinea. Trans. Roy. Soc. Trop.270Ð27330Schofield L. and Hackett F. (1993) Signal transduction in hostcells by a glycosylphosphatidylinositol toxin of malaria para-sites. J. Exp. Med. 145Ð15331Schofield L., Novakovic S., Gerold P., Schwarz R. T., Mc-Conville M. J. and Tachado S. D. (1996) Glycosylphos-phatidylinositol toxin of upregulates intercellularadhesion molecule-1, vascular cell adhesion molecule-1, andE-selectin expression in vascular endothelial cells and in-creases leukocyte and parasite cytoadherence via tyrosine ki-nase-dependent signal transduction. J. Immunol. 1886Ð189632Tachado S. D., Gerold P., McConville M. J., Baldwin T.,Quilici D., Schwarz R. T. et al. (1996) Glycosylphosphatidyli-nositol toxin of induces nitric oxide synthase ex-pression in macrophages and vascular endothelial cells by apendent signaling pathway. J. Immunol. 1897Ð190733Taylor T. E., Borgstein A. and Molyneux M. E. (1993) Acid-malaria. Q. J.99Ð10934Taylor T. E., Molyneux M. E., Wirima J. J., Fletcher A. andMorris K. (1988) Blood glucose levels in Malawian childrenbefore and during the adminis

tration of intravenous quininefor severe falciparum malaria. N. Engl. J. Med. 1040Ð104735Marsh K., Forster D., Waruiru C., Mwangi I., Winstanley M.,Marsh V., et al. (1995) Indicators of life-threatening malaria inAfrican children. N. Engl. J. Med. 1399Ð140436White N. J. and Ho M. (1992) The pathophysiology ofmalaria. Adv. Parasitol. 83Ð17337Gravenor M. B., van Hensbroek M. B. and Kwiatkowski D.cerebral malaria. Proc. Natl. Acad. Sci. USA 7620Ð762438Silamut K. and White N. J. (1993) Relation of the stage of par-asite development in the peripheral blood to prognosis in se-vere malaria. Trans. R. Soc. Trop. Med. Hyg. 436Ð44339Udeinya I. J., Schmidt J. A., Aikawa M., Miller L. H. andFalciparummalaria-infected erythrocytesspecifically bind to cultured human endothelial cells. Science555Ð55740Roberts D. J., Biggs B.-A., Brown G. and Newbold C. I.(1993) Protection, pathogenesis and phenotypic plasticity malaria. Parasitol. Today 281Ð41Pouvelle B., Buffet P. A., Lepolard C., Scherf A. and Gysin J.infected erythrocytes. Nat. Med. 1264Ð126842Scholander C., Treutiger C. J., Hultenby K. and Wahlgren M.(1996) Novel fibrillar structure confers adhesive property tomalaria-infected erythrocytes. Nat. Med. 204Ð20843Carlson J., Helmby H., Hill A. V. S., Brewster D., GreenwoodB. M. and Wahlgren M. (1990) Human cerebral malaria: as-sociation with erythrocyte rosetting and lack of anti-rosetting1457Ð146044Treutiger C. J., Hedlund I., Helmby H., Carlson J., Jepson A.,Twumasi P. et al. (1992) Rosette formation in isolates and anti-rosette activity of sera from Gam-bians with cerebral or uncomplicated malaria. Am. J. Trop.503Ð51045Rowe A., Obeiro J., Newbold C. I. and Marsh K. (1995) rosetting is associated with malaria sever-ity in Kenya. Infect. Immun. 2323Ð232646Ho M., Davis T., Silamut K., Bunnag D. and White N. J.(1991) Rosette formation of Plasmodiumfalciparumerythrocytes from patients with acute malaria. Infect. Immun.2135Ð213947Al-Yaman F., Genton B., Mokela D., Raiko A., Kati S., Roger-son S. et al. (1995) Human cerebral malaria: lack of signifi-cant association between erythrocyte rosetting and diseaseseverity. Trans. R. Soc. Trop. Med. Hyg. 55Ð5848Rogerson S. J., Beck H.-P., Al-Yaman F., Currie B., Alpers M.P. and Brown G. V. (1996) Disruption of erythrocyte rosettesand agglutination of erythrocytes infected with by the sera of Papua New Guineans. Trans. R. Soc.Trop. Med. Hyg. 80Ð8449Roberts D. J., Craig A. G., Berendt A. R., Pinches R., Nash G.,Marsh G. et al. (1992) Rapid switching to multiple antigenicand adhesive phenotypes in malaria. Nature 689Ð69250Pain A., Ferguson D. J., Kai O., Urban B. C., Lowe B., Marsh-infected erythrocytes is a common adhesive phe-notype and is associated with severe malaria. Proc. Natl.1805Ð181051Roberts D., Pain A., Kai O., Kortok M. and Marsh K. (2000)Autoagglutination of malaria-infected red blood cells andmalaria severity. Lancet 1427Ð142852Ruangjirachuporn W., Afzelius B. A., Helmby H., Hill A. V. S.,Greenwood B. M., Carlson J. H. et al. (1992) Ultrastructuralanalysis of fresh -infected erythro-cytes and their cytoadherence to human leukocytes. Am. J.Trop. Med. Hyg. 511Ð51953Nash G. B., OÕBrien E., Gordon-Smith E. C. and Dormandy J.A. (1989) Abnormalities in the mechanical properties of redblood cells caused by855Ð86154Miller L. H., Usami S. and Chien S. (1971) Alteration in therheologic properties of Plasmodium knowlesicells. A possible mechanism for capillary obstruction. J. Clin.Invest. 1451Ð145555Beeson J. G., Brown G. V., Molyneux M. E., Mhango C., Dzinjalamala F. and Rogerson S. J. (1999) isolates from infected pregnant women and childrenare associated with distinct adhesive and antigenic properties.J. Infect. Dis. 464Ð47256Fried M. and Duffy P. E. (1996) Adherence of to chondroitin sulfate A in the human placenta.1502Ð1504268J. G. Beeson and G. V. BrownPathogenesis of P. falciparum 57Gysin J., Pouvelle B., Fievet N., Scherf A. and Lepolard C.(1999) Ex vivo desequestration of infected erythrocytes from human placenta by chondroitinsulfate A. Infect. Immun. 6596Ð660258Beeson J. G., Rogerson S. J., Cooke B. M., Reeder J. C., ChaiW., Lawson A. M. et al. (2000) Adhesion of -infected erythrocytes to hyaluronic acid in placental86Ð9059Beeson J. G., Reeder J. C., Rogerson S. J. and Brown G. V.(2001) Parasite adhesion and immune evasion in placentalmalaria. Trends Parasitol. 331Ð33760Newbold C., Warn P., Black G., Berendt A., Craig A., Snow B.et al. (1997) Receptor-specific adhesion and clinical disease. Am. J. Trop. Med. Hyg. 389Ð39861Rogerson S. J., Tembenu R., Dobano C., Plitt S., Taylor T. E.and Molyneux M. E. (1999) Cytoadherence characteristics of-infected erythrocytes from Malaw-ian children with severe and uncomplicated malaria. Am. J.Trop. Med. Hyg. 467Ð47262McCormick C. J., Craig A., Roberts D., Newbold C. I. andBerendt A. R. (1997) Intercellular adhesion molecule-1 andCD36 synergize to mediate adherence of -infected erythrocytes to cultured human microvascularen

dothelial cells. J. Clin. Invest. 2521Ð251963Ho M., Hickey M. J., Murray A. G., Andonegui G. and KubesP. (2000) Visualization of lium interactions in human microvasculature: mimicry ofleukocyte recruitment. J. Exp. Med. 1205Ð121164Ockenhouse C. F., Tandon N. N., Magowan C., Jamieson G. A.and Chulay J. D. (1989) Identification of a platelet membraneglycoprotein as a falciparum malaria sequestration receptor.1469Ð147165Barnwell J. W., Asch A. S., Nachman R. L., Yamaya M. andAikawa M. (1989) A human 88-kD membrane glycoprotein(CD36) functions in vitro as a receptor for a cytoadherence-infected erythrocytes. J.Clin. Invest. 765Ð77266Oquendo P., Hundt E., Lawler J. and Seed B. (1989) CD36 di-rectly mediates cytoadherence of parasitized erythrocytes. Cell 95Ð10167Ockenhouse C. F., Ho M., Tandon N. N., Van-Seventer G. A.,Shaw S., White N. J. et al. (1991) Molecular basis of seques-tration in severe and uncomplicated malaria: differential adhesion of infected erythrocytes toCD36 and ICAM-1. J. Infect. Dis. 163Ð16968Aitman T. J., Cooper L. D., Norsworthy P. J., Wahid F. N., GrayJ. K., Curtis B. R. et al. (2000) Malaria susceptibility and1015Ð101669Pain A., Urban B. C., Kai O., Casals-Pascual C., Shafi J.,Marsh K. et al. (2001) A non-sense mutation in Cd36 gene isassociated with protection from severe malaria. Lancet 1502Ð150370McGilvray I. D., Serghides L., Kapus A., Rotstein O. D. andKain K. C. (2000) Nonopsonic monocyte/macrophage phago-cytosis of -parasitized erythrocytes: a3231Ð324071Urban B., Ferguson D., Pain A., Willcox N., Plebansk i. M.,Austyn J. et al. (1999) -infected ery-throcytes modulate the maturation of dendritic cells. Nature73Ð7772Berendt A. R., Simmons D. L., Tansey J., Newbold C. I. and57Ð5973Berendt A. R., McDowall A., Craig A. G., Bates P. A., SternbergM. J. E., Marsh K. et al. (1992) The binding site on ICAM-1 for-infected erythrocytes overlaps, but isdistinct from the LFA-1 binding site. Cell 71Ð8174Fernandez-Reyes D., Craig A., Kyes S., Peshu N., Snow R.,Berendt A. et al. (1997) A high frequency African codingpolymorphism in the N-terminal domain of ICAM-1 predis-posing to cerebral malaria in Kenya. Hum. Mol. Genet. 75Kun J. F., Klabunde J., Lell B., Luckner D., Alpers M., May J.et al. (1999) Association of the ICAM-1 Kilifi mutation withprotection against severe malaria in Lambarene, Gabon. Am.J. Trop. Med. Hyg. 776Ð77976Rogerson S. J., Chaiyaroj S. C., Ng K., Reeder J. C. and BrownG. V. (1995) Chondroitin sulfate A is a cell surface receptor for-infected erythrocytes. J. Exp. Med.15Ð2077Cooke B. M., Rogerson S. J., Brown G. V. and Coppel R. L.(1996) Adhesion of malaria-infected red blood cells to chondroitin sulfate A under flow conditions. Blood 4040Ð404478Robert C., Pouvelle B., Meyer P., Muanza K., Fukioka H.,Aikawa M. et al. (1995) Chondroitin-4-sulphate (proteogly--infected erythro-cyte adherence on brain microvascular endothelial cells. Res.383Ð39379Beeson J. G., Chai W., Rogerson S. J., Lawson A. M. andBrown G. V. (1998) Inhibition of binding of malaria-infectederythrocytes by a tetradecasaccharide fraction from chon-droitin sulfate A. Infect. Immun. 3397Ð340280Alkhalil A., Achur R. N., Valiyaveettil M., Ockenhouse C. F.and Gowda D. C. (2000) Structural requirements for the ad--infected erythrocytes tochondroitin sulfate proteoglycans of human placenta. J. Biol.40357Ð4036481Rogerson S. J., Novakovic S., Cooke B. M. and Brown G. V.-infected erythrocytes adhereto the proteoglycan thrombomodulin in static and flow-basedsystems. Exp. Parasitol. 8Ð1882Gysin J., Pouvelle B., Le Tonqueze M., Edelman L. and BoffaM.-C. (1997) Chondroitin sulfate of thrombomodulin is aninfected ery-throcytes. Mol. Biochem. Parasitol. 267Ð27183Wong V. L. Y., Hofman F. M., Ishii H. and Fisher M. (1991)Regional distribution of thrombomodulin in human brain.84Chaiyaroj S. C., Angkasekwinai P., Buranakiti A., Looa-reesuwan S., Rogerson S. J. and Brown G. V. (1996) Cytoad-from Thailand: evidence for chondroitin sulfate A as a cy-toadherence receptor. Am. J. Trop. Med. Hyg. 76Ð8085Traore B., Muanza K., Looareesuwan S., Supavej S., Khu-isolates in Thailand using an invitro human lung endothelial cells model. Am. J. Trop. Med.38Ð4486Sunderland C. A., Bulmer J. N., Luscombe M., Redman C. W.G. and Stirrat G. M. (1985) Immunohistological and bio-chemical evidence for a role for hyaluronic acid in the growthand development of the placenta. J. Reproduct. Immunol. 197Ð21287Martin B. J., Spicer S. S. and Smythe N. M. (1974) Cyto-chemical studies of the maternal surface of the syncytiotro-phoblast of human early and term placenta. Anat. Rec. 769Ð78688Parmley R. T., Takagi M. and Denys F. R. (1984) Ultrastruc-tural localization of glycosaminoglycans in human term pla-centa. Anat. Rec. 477Ð48489Chai W., Beeson J. G., Kogelberg H., Brown G. V. and Lawson-infected erythrocytes by structurally defined hyal-420Ð42590Mohamadzadeh M., DeGrendele H., Arizpe H., Estess P. andSiegelman M. (1998) Proinflammatory stimuli regu

late en-CMLS, Cell. Mol. Life Sci.Vol. 59, 2002Review Article dothelial hyaluronan expression and CD44/HA-dependentprimary adhesion. J. Clin. Invest. 97Ð10891Chen Q., Barragan A., Fernandez V., Sundstrom A.,Schlichtherle M., Sahlen A. et al. (1998) Identification oferythrocyte membrane protein 1P. fal-. J. Exp. Med. 15Ð2392Carlson J., Ekre H. P., Helmby H., Gysin J., Greenwood B. M.and Wahlgren M. (1992) Disruption oferythrocyte rosettes by standard heparin and heparindevoid of anticoagulant activity. Am J. Trop. Med. Hyg. 595Ð60293Rogerson S. J., Chaiyaroj S. C., Reeder J. C., Al-Yaman F. andBrown G. V. (1994) Sulfated glycoconjugates as disrupters oferythrocyte rosettes. Am. J. Trop.198Ð20394Barragan A., Spillmann D., Kremsner P. G., Wahlgren M. andCarlson J. (1999) ground to strain-specific rosette disruption by glycosamino-glycans and sulfated glycoconjugates. Exp. Parasitol. 133Ð14395Rowe J. A., Moulds J. M., Newbold C. I. and Miller L. H.P. falciparumrosetting mediated by a parasite-varianterythrocyte membrane protein and complement-receptor 1.292Ð29596Rowe J. A., Rogerson S. J., Raza A., Moulds J. M., Ka-zatchkine M., Marsh K. et al. (2000) Mapping of the region rosetting and demonstration of the importance ofCR1 in rosetting in field isolates. J. Immunol. 6341Ð634697Carlson J. and Wahlgren M. (1992) erythrocyte rosetting is mediated by promiscuous lectin-likeinteractions. J. Exp. Med. 1311Ð131798Barragan A., Kremsner P., Wahlgren M. and Carlson J. (2000)Blood group A antigen is a coreceptor in 2971Ð297599Ockenhouse C. F., Tegoshi T., Maeno Y., Benjamin C., Ho M.,Kan K. E. et al. (1992) Human vascular endothelial cell adhe-Plasmodiumfalciparum-infected erythro-cytes: roles for endothelial leukocyte adhesion molecule 1and vascular cell adhesion molecule 1. J. Exp. Med. 1183Ð1189100Ho M., Schollaardt T., Niu X., Looareesuwan S., Patel K. D.and Kubes P. (1998) Characterization of -infected erythrocyte and P-selectin interaction underflow conditions. Blood 4803Ð4809101Roberts D. D., Sherwood J. A., Spitalnik S. L., Panton L. J.,Howard R. J., Dixit V. M. et al. (1985) Thrombospondin bindsfalciparum malaria parasitized erythrocytes and may mediatecytoadherence. Nature 64Ð66102Treutiger C. J., Heddini A., Fernandez V., Muller W. A. andWahlgren M. (1997) PECAM-1/CD31, an endothelial recep--infected erythrocytes.1405Ð1408103Siano J. P., Grady K. K., Millet P. and Wick T. M. (1998) : cytoadherence to crovascular endothelial cells. Am. J. Trop. Med. Hyg. 77Ð79104Udomsangpetch R., Taylor B. J., Looareesuwan S., White N.J., Elliott J. F. and Ho M. (1996) Receptor specificity of clini-bound E-selectin and vascular cell adhesion molecule-1.2754Ð2760105Baruch D. I., Ma X. C., Singh H. B., Bi X., Pasloske B. L. andHoward R. J. (1997) Identification of a region of PfEMP1 thatinfected ery-throcytes to CD36: conserved function with variant sequence.3766Ð3775106Reeder J. C., Cowman A. F., Davern K. M., Beeson J. G.,Thompson J. K., Rogerson S. J. et al. (1999) The adhesion of-infected erythrocytes to chondroitinsulfate A is mediated by PfEMP1. Proc. Natl. Acad. Sci. USA5198Ð5202107Smith J. D., Craig A. G., Kriek N., Hudson-Taylor D., Kyes S.,Fagen T. et al. (2000) Identification of a Natl. Acad. Sci. USA 1766Ð1771108Crabb B. S., Cooke B. M., Reeder J. C., Waller R. F., Car-uana S. R., Davern K. M. et al. (1997) Targeted gene dis-ruption shows that knobs enable malaria-infected red cells tocytoadhere under physiological shear stress. Cell 287Ð296109Su X. Z., Heatwole V. M., Wertheimer S. P., Guinet F., Her-rfeldt J. A., Peterson D. S. et al. (1995) The large diverse genefamily encodes proteins involved in cytoadherence andantigenic variation of -infected ery-throcytes. Cell 89Ð100110Chen Q., Heddini A., Barragan A., Fernandez V., Pearce S. andWahlgren M. (2000) The semiconserved head structure oferythrocyte membrane protein 1 me-diates binding to multiple independent host receptors. J. Exp.1Ð10111Smith J. D., Subramanian G., Gamain B., Baruch D. I. andMiller L. H. (2000) Classification of adhesive domains in theerythrocyte membrane protein 1family. Mol. Biochem. Parasitol. 293Ð310112Baruch D. I., Pasloske B. L., Singh H. B., Bi X., Ma X. C.,Feldman M. et al. (1995) Cloning the P. falciparumcoding PfEMP1, a malarial variant antigen and adherence re-ceptor on the surface of parasitized human erythrocytes. Cell77Ð87113Cooke B. M., Nicoll C. L., Baruch D. I. and Coppel R. L.(1998) A recombinant peptide based on PfEMP1 blocks andreverses adhesion of malaria-infected red blood cells to CD36under flow. Mol. Microbiol. 83Ð90114Smith J. D., Kyes S., Craig A. G., Fagan T., Hudson-Taylor D.,Miller L. H. et al. (1998) Analysis of adhesive domains fromthe A4VAR erythrocyte membraneprotein-1 identifies a CD36 binding domain. Mol. Biochem.Parasitol. 133Ð148115Reeder J. C., Hodder A. N., Beeson J. G. and Brown G. V.(2000) Identification of glycosaminoglycan binding domainserythrocyte membrane protein 1 ofa chondroitin sulfat

e A-adherent parasite. Infect. Immun. 3923Ð3926116Degen R., Weiss N. and Beck H. P. (2000) : cloned and expressed CIDR domains of PfEMP1 bindto chondroitin sulfate A. Exp. Parasitol. 113Ð121117Buffet P. A., Gamain B., Scheidig C., Baruch D., Smith J. D.,Hernandez-Rivas R. et al. (1999) main mediating adhesion to chondroitin sulfate A: a receptorfor human placental infection. Proc. Natl. Acad. Sci. USA 12743Ð12748118Baruch D. I., Gormley J. A., Ma C., Howard R. J. and PasloskeB. L. (1996) erythrocyte membraneprotein 1 is a parasitized erythrocyte receptor for adherence to1. Proc. Natl. Acad. Sci. USA 3497Ð3502119Barragan A., Fernandez V., Chen Q., von Euler A., WahlgrenM. and Spillmann D. (2000) The Duffy-binding-like domain 1erythrocyte membrane protein 1(PfEMP1) is a heparan sulfate ligand that requires 12 mers for3594Ð3599120Gamain B., Smith J. D., Miller L. H. and Baruch D. I. (2001)Modifications in the CD36 binding domain of the Plasmod-ium falciparum variant antigen are responsible for the inabil-270J. G. Beeson and G. V. BrownPathogenesis of P. falciparum 136Cheng Q., Cloonan N., Fischer K., Thompson J., Waine G.,stevormulticopy gene families which potentially encode vari-ant antigens. Mol. Biochem. Parasitol. 161Ð176137Fernandez V., Hommel M., Chen Q., Hagblom P. andWahlgren M. (1999) Small, clonally variant antigensexpressed on the surface of ed erythrocytes are encoded by the gene family and are tar-gets of human immune responses. J. Exp. Med. 1393Ð138Kyes S. A., Rowe J. A., Kriek N. and Newbold C. I. (1999)Rifins: a second family of clonally variant proteins expressedon the surface of red cells infected with . Proc. Natl. Acad. Sci. USA 9333Ð9338139Marsh K. and Howard R. J. (1986) Antigens induced on ery-throcytes by P. falciparum: expression of diverse and con-served determinants. Science 150Ð153140Reeder J. C., Rogerson S. J., Al-Yaman F., Anders R. F., Cop-pel R. L., Novakovic S. et al. (1994) Diversity of agglutinatingphenotype, cytoadherence, and rosette-forming characteris-isolates from Papua NewGuinean children. Am. J. Trop. Med. Hyg. 45Ð55141Forsyth K. P., Philip G., Smith T., Kum E., Southwell B. andBrown G. V. (1989) Diversity of antigens expressed on the sur-face of the erythrocytes infected with mature parasites in Papua New Guinea. Am. J. Trop. Med.259Ð265142Bull P. C., Lowe B. S., Kortok M., Molyneux C. S., NewboldC. I. and Marsh K. (1998) Parasite antigens on the infected redcell surface are targets for naturally acquired immunity to358Ð360143Marsh K., Otoo L., Hayes R. J., Carson D. C. and GreenwoodB. M. (1989) Antibodies to blood stage antigens of in rural Gambians and their relation to protec-tion against infection. Trans. R. Soc. Trop. Med. Hyg. 293Ð303144Giha H. A., Staalsoe T., Dodoo D., Roper C., Satti G. M.,Arnot D. E. et al. (2000) Antibodies to variable -infected erythrocyte surface antigens are associ-ated with protection from novel malaria infections. Immunol.117Ð126145Aguiar J. C., Albrecht G. R., Cegielski P., Greenwood B. M.,Jensen J. B., Lallinger G. et al. (1992) Agglutination of -infected erythrocytes from East and WestAfrican isolates by human sera from distant geographic re-gions. Am. J. Trop. Med. Hyg. 621Ð632146Gamain B., Miller L. H. and Baruch D. I. (2001) The surfacevariant antigens of active epitopes. Proc. Natl. Acad. Sci. USA 2664Ð2669147Rogerson S. J. and Beeson J. G. (1999) The placenta inmalaria: mechanisms of infection, disease and fetal morbidity.Ann. Trop. Med. Parasitol. S35ÐS42148Maubert B., Fievet N., Tami G., Boudin C. and Deloron P.erythrocytes in the human placenta. Parasite Immunol. 191Ð199149Fried M., Nosten F., Brockman A., Brabin B. J. and Duffy P. E.(1998) Maternal antibodies block malaria. Nature 851Ð852150Ricke C. H., Staalsoe T., Koram K., Akanmori B. D., Riley E.M., Theander T. G. et al. (2000) Plasma antibodies frommalaria-exposed pregnant women recognize variant surface-infected erythrocytes ina parity-dependent manner and block parasite adhesion tochondroitin sulfate A. J. Immunol. 3309Ð3316CMLS, Cell. Mol. Life Sci.Vol. 59, 2002Review Articleity of chondroitin sulfate A adherent parasites to bind CD36.3268Ð3274121Ockenhouse C. F., Klotz F. W., Tandon N. N. and Jamieson G.A. (1991) Sequestrin, a CD36 recognition protein on malaria-infected erythrocytes identifiedby anti-idiotype antibodies. Proc. Natl. Acad. Sci. USA 3175Ð3179122Crandall I., Collins W. E., Gysin J. and Sherman I. W. (1993)protein inhibit cytoadherence/sequestration of the malaria. Proc. Natl. Acad. Sci. USA4703Ð4707123Lucas J. and Sherman I. (1998) thrombospondin mediates parasitized erythrocyte band 3-re-lated adhesin binding. Exp. Parasitol. 78Ð85124Day K. P., Karamalis F., Thompson J., Barnes D. A., PetersenC., Brown H. et al. (1993) Genes necessary for expression ofa virulence determinant and for transmission of are located on a 0.3-megabase region of chromo-some 9. Proc. Natl. Acad. Sci. USA 8292Ð8296125Trenholme K., Gardiner D., Holt D., Thomas E.

, Cowman A.and Kemp D. (2000) clag9: a cytoadherence gene in essential for binding of parasitized erythro-cytes to CD36. Proc. Natl. Acad. Sci. USA 4029Ð4033126Biggs B. A., GoozŽ L., Wycherley K., Wollish W., SouthwellB., Leech J. H. et al. (1991) Antigenic variation in iumfalciparum. Proc. Natl. Acad. Sci. USA 9171Ð9174127Brown K. N. and Brown I. N. (1965) Immunity to malaria:Antigenic variation in chronic infection of knowlesi1286Ð1288128Handunnetti S. M., Mendis K. N. and David P. H. (1987) Anti-genic variation of cloned Plasmodiumfragile. Sequential appearance of successive vari-ant antigenic types. J. Exp. Med. 1269Ð1283129McLean S. A., Pearson C. D. and Phillips R. S. (1982) modium chabaudi: relationship between the occurrence of re-crudescent parasitaemias in mice and the effective levels ofacquired immunity. Exp. Parasitol. 213Ð221130Biggs B. A., Anders R. F., Dillon H. E., Davern K. M., MartinM., Petersen C. et al. (1992) Adherence of infected erythro-cytes to venular endothelium selects for antigenic variants of. J. Immunol. 2047Ð2054131Leech J. H., Barnwell J. W., Miller L. H. and Howard R. J.(1984) Identification of a strain-specific malarial antigen ex-posed on the surface of -infected ery-throcytes. J. Exp. Med. 1567Ð1575132Magowan C., Wollish W., Anderson L. and Leech J. (1988)Cytoadherence by -infected erythro-cytes is correlated with the expression of a family of variableproteins on infected erythrocytes. J. Exp. Med. 1307Ð1320133Smith J. D., Chitnis C. E., Craig A. G., Roberts D. J., Hudson-Taylor D. E., Peterson D. S. et al. (1995) Switches in expres-var genes correlate withchanges in antigenic and cytoadherent phenotypes of infectederythrocytes. Cell 101Ð110134Chen Q., Fernandez V., Sundstrom A., Schlichtherle M., DattaS., Hagblom P. et al. (1998) Developmental selection of gene expression in 392Ð395135Scherf A., Hernandez-Rivas R., Buffet P., Bottius E., BenatarC., Pouvelle B. et al. (1998) Antigenic variation in malaria: switching, relaxed and mutally exclusive transcription ofgenes during intra-erythrocytic development in . EMBO J. 5418Ð5426 ReviewPathogenesis of Plasmodium falciparumthe roles of parasite adhesion and antigenic variationJ.G.Beeson* and G.V.BrownDepartment of Medicine, University of Melbourne, Post Office Royal Melbourne Hospital, Victoria 3050 (Australia),Fax +61 3 9347 1863, e-mail: beeson@unimelb.edu.auReceived 25 June 2001; received after revision 22 August 2001; accepted 24 August 2001Abstract.nually, with young children and pregnant women at great-est risk. The great majority of severe disease is caused by. A characteristic feature of in-P. falciparumtration of parasite-infected red blood cells (RBCs) in var-ious organs, such as the brain, lung and placenta, and to-gether with other factors is important in the pathogenesisof severe forms of malaria. Sequestration results from ad-hesive interactions between parasite-derived proteins ex-pressed on the surface of infected RBCs and a number ofhost molecules on the surface of endothelial cells, pla-CMLS, Cell. Mol. Life Sci. 59 (2002) 258Ð271© BirkhŠuser Verlag, Basel, 2002 parasite adhesion have been implicated in particularcule 1 in cerebral malaria and chondroitin sulfate A andhyaluronic acid in placental infection. The principal par-asite ligand and antigen on the RBC surface, P. falci-erythrocyte membrane protein 1 encoded by amultigene family termed , is clonally variant, enablingevasion of specific immune responses. An understandingof these host-parasite interactions in the context of clini-cal disease and immunity may reveal potential targets toprevent or treat severe forms of malaria.Key words.Malaria; pathogenesis; adhesion; antigenic variation; cerebral malaria; placenta;IntroductionAround 40% of the worldÕs population lives in malaria-cal and subtropical regions [1]. Malaria infection resultsin 300Ð500 million clinical cases and 1.5Ð2.7 milliondeaths annually, with ~1 million deaths in children under5 years of age [1]. About 90% of cases and most deathsoccur in tropical Africa.is responsible for the majority ofsevere clinical disease, most affecting young children,nonimmune adults and pregnant women, and is the focusof this review. It is the predominant species in tropicalAfrica, eastern Asia, Oceania and the Amazon basin oftial amount of clinical malaria and is widely distributedgeographically; however, severe clinical disease is rarelyP. vivax Plasmodium ovaleThe life cycle of P. falciparum species) involves several stages in both human and mos-quito hosts. Following the bite of an infected femalesectÕs salivary glands enter the bloodstream and travelquickly to the liver. Invasion of hepatocytes ensues, andan 8- to 12-day incubation period enables asexual repli-cation to generate many daughter merozoites. Infectedhepatocytes then rupture, releasing merozoites into thecirculation, thus commencing the blood stage of the in-fection during which clinical malaria may develop. Corresponding author.