Plasmodium falciparum Erythrocyte Membrane Protein-1 Specifically Suppresses Early Production of Host Interferon-γ

Cell Host & Microbe

Short Article Plasmodium falciparum Erythrocyte Membrane Protein-1 Specifically Suppresses Early Production of Host Interferon-g Marthe C. D’Ombrain,1,2 Till S. Voss,1 Alexander G. Maier,1 J. Andrew Pearce,1 Diana S. Hansen,1 Alan F. Cowman,1 and Louis Schofield1,* 1

Infection and Immunity Division, The Walter and Eliza Hall Institute of Medical Research, 1G Royal Parade, Parkville, Victoria 3050, Australia 2 Department of Medical Biology, The University of Melbourne, Parkville, Victoria 3010, Australia *Correspondence: [email protected] DOI 10.1016/j.chom.2007.06.012

SUMMARY

Plasmodium falciparum erythrocyte membrane protein-1 (PfEMP-1) is a variable antigen expressed by P. falciparum, the malarial parasite. PfEMP-1, present on the surface of infected host erythrocytes, mediates erythrocyte binding to vascular endothelium, enabling the parasite to avoid splenic clearance. In addition, PfEMP-1 is proposed to regulate host immune responses via interactions with the CD36 receptor on antigen-presenting cells. We investigated the immunoregulatory function of PfEMP-1 by comparing host cell responses to erythrocytes infected with either wild-type parasites or transgenic parasites lacking PfEMP-1. We showed that PfEMP-1 suppresses the production of the cytokine interferon-g by human peripheral blood mononuclear cells early after exposure to P. falciparum. Suppression of this rapid proinflammatory response was CD36 independent and specific to interferon-g production by gd-T, NK, and ab-T cells. These data demonstrate a parasite strategy for downregulating the proinflammatory interferon-g response and further establish transgenic parasites lacking PfEMP-1 as powerful tools for elucidating PfEMP-1 functions. INTRODUCTION The multidomain variant antigen Plasmodium falciparum Erythrocyte Membrane Protein-1 (PfEMP-1), expressed on the surface of P. falciparum-infected erythrocytes, can mediate binding to CD36, intercellular adhesion molecule-1, vascular cell adhesion molecule, E-selectin, P-selectin, CD31, and chondroitin sulfate A (CSA) on vascular endothelial cells, thus enabling the parasite to avoid splenic clearance (Newbold, 1999). PfEMP-1 is encoded by the var gene family, which has been divided into three major groups (A, B, and C) based on genomic location

and upstream sequence (Ups). Group B and C var genes typically bind CD36, whereas group A var genes generally do not (Robinson et al., 2003). Furthermore, group B and C var genes tend to be associated with asymptomatic malaria, while group A var genes have been implicated in cerebral malaria (Kyriacou et al., 2006). The extensive var repertoire and allelic diversity within PfEMP-1 enable the parasite to evade humoral immunity while maintaining cytoadherence to blood vessel walls (Hisaeda et al., 2005). Although successful in evading antibody responses, PfEMP-1 is also proposed to modulate host immune responses via CD36-dependent interactions with antigen-presenting cells. Phagocytosis of nonopsonized infected erythrocytes by macrophages was shown to be dependent on CD36 ligation and was inhibited by the proteolytic removal of PfEMP-1 (McGilvray et al., 2000). Furthermore, infected erythrocytes inhibited DC maturation, reducing the capacity of DC to activate T cells (Urban et al., 1999). This was proposed to result from PfEMP-1 binding to CD36, thus implicating the CD36-binding cysteine interdomain region-1a (CIDR-1a) domain of PfEMP-1 as mediating this effect (Urban et al., 2001b). In contrast, however, addition of recombinant CIDR-1a to cultures of naive human PBMC resulted in the polyclonal activation of naive CD4 T cells and NK cells and induced IFN-g production, leading to the suggestion that this molecule may contribute to inflammatory immunopathogenesis (Ndungu et al., 2006). These inconsistencies may reflect methodological constraints that have previously restricted immunological analyses of PfEMP-1 to use of either nonisogenic parasites selected for specific adhesion phenotypes (Urban et al., 1999) (whereby uniform expression of one PfEMP-1 allele in the population is not assured); or on recombinant PfEMP-1 subdomains expressed in E. coli (Allsopp et al., 2002; Joergensen et al., 2006; Ndungu et al., 2006), which may be confounded by misfolding, glycosylation, LPS contamination, presentation out of cellular context, and application in physiologically inappropriate concentrations. Clearly, PfEMP-1 constitutes a major interface between parasite and host, and the immunoregulatory activities of this molecule are likely to be important (Urban et al., 2005). Furthermore, as PfEMP-1 domains may be developed as

130 Cell Host & Microbe 2, 130–138, August 2007 ª2007 Elsevier Inc.

Cell Host & Microbe PfEMP-1 Suppresses Early IFN-g Production

humoral vaccines (Baruch et al., 2002), understanding their potential role in the promotion or prevention of immunoregulation, immunopathology, and cell-mediated immunity is a priority. Therefore, to investigate the immunological function of endogenous, whole PfEMP-1, expressed in situ in the correct context and concentration on the parasite surface, we used transgenic parasite lines that lack surface expression of PfEMP-1 (Maier et al., 2006; Voss et al., 2006). Using these parasite lines, we demonstrate that PfEMP-1 suppresses innate IFN-g production by naive gd-T cells and NK cells, as well as early IFN-g production by naive ab-T cells. No evidence for PfEMP-1-mediated polyclonal activation of these cell populations was found. The interaction of CD36 with CIDR-1a, proposed variously to mediate immunosuppression (Urban et al., 1999, 2001a) or polyclonal activation of naive T cells and NK cells (Ndungu et al., 2006), was not required for the suppression of IFN-g production. Therefore, PfEMP-1-mediated suppression of innate IFN-g production occurs via a CIDR-1a/CD36-independent pathway. Furthermore, PfEMP-1 was neither a suppressor nor an activator of malaria-induced NK cell- or T cellmediated cytolysis. These data demonstrate a strategy employed by the parasite to downregulate the proinflammatory IFN-g response and validate the use of PfEMP-1 transgenic parasites as powerful tools for dissecting the functional and immunological roles of this important virulence protein in human malaria. RESULTS AND DISCUSSION

Figure 1. PfEMP-1 Suppresses the Innate IFN-g Response of PBMC to P. falciparum In Vitro uRBC, parental 3D7, and 3D7 PfEMP-1 knockdown parasites were solubilized in 1% Triton X-100. The insoluble fraction was then solubilized in 2% SDS, and the supernatant was probed in western blot with antibodies (mAb 6H1) against the conserved intercellular region of PfEMP-1 (A). Parental 3D7 shows a distinctive band above 250 kDa, typical of PfEMP-1, whereas 3D7 PfEMP-1 iRBC and uRBC do not (asterisks denote erythrocyte proteins) (A). (B) is a vector map of the construct used to produce the 3D7 PfEMP-1 line containing the human dihydrofolate reductase (hDHFR) gene that encodes resistance to WR99210. The expression of hDHFR is under the control of a var upsC gene promoter. The blasticidin deaminase (BSD) gene encodes resistance to blasticidin-S and serves to maintain stable integration of the plasmid in the parasite chromosome. PBMC (2 3 105) from a number of malaria-naive donors were stimulated for 24 hr with 6 3 105 3D7 parental iRBC, 3D7 PfEMP-1 iRBC, or uRBC (C), or 6 3 105 3D7 parental iRBC, trypsin-digested 3D7 parental iRBC, uRBC, or trypsin-digested uRBC (D). Supernatants were harvested and assayed for cytokines. The mean IFN-g detected in supernatants is shown (C and D). Error bars are the SEM for R2 experiments. Uppercase letters indicate donors. In 9/9 donors, the mean level of IFN-g

PfEMP-1 Suppresses the Innate IFN-g Response to Malaria To assess the role of PfEMP-1 in the regulation of cellular immune responses to malaria, transgenic parasites lacking surface expression of PfEMP-1 were used (Maier et al., 2006; Voss et al., 2006). As shown in Figure 1A, P. falciparum 3D7 parental parasites express PfEMP-1, whereas the 3D7 PfEMP-1 knockdown parasites do not. This is due to the expression of a human dihydrofolate reductase gene, which confers resistance to WR99210, under the control of a var upsC gene promoter that allows infiltration of the antigenic variation program, thus silencing the gene family (Voss et al., 2006) (Figure 1B). To investigate the effect of the absence and presence of PfEMP-1 on human, innate cellular immune responses, peripheral blood mononuclear cells (PBMC) from healthy, malarianaive adult donors were stimulated for 24 hr with 3D7 parental, live P. falciparum schizont-infected red blood cells (iRBC), 3D7 PfEMP-1 iRBC, or uninfected red blood cells (uRBC), and their ability to induce proinflammatory and anti-inflammatory cytokines was observed. Loss of produced by PBMC following stimulation with 3D7 PfEMP-1 iRBC exceeds that for 3D7 parental iRBC (p < 0.01; nonparametric sign test) (C). Similarly, in 7/7 donors, the mean level of IFN-g produced by PBMC following stimulation with trypsin-digested 3D7 parental iRBC exceeds that for nondigested 3D7 parental iRBC (p < 0.01; nonparametric sign test) (D).

Cell Host & Microbe 2, 130–138, August 2007 ª2007 Elsevier Inc. 131

Cell Host & Microbe PfEMP-1 Suppresses Early IFN-g Production

PfEMP-1 resulted in substantially higher IFN-g production in all donors compared to parental iRBC, with some donors producing up to four times more IFN-g in the absence of PfEMP-1 (Figure 1C), thus suggesting that PfEMP-1 suppresses the host innate IFN-g response to malaria. Parental 3D7 iRBC that were digested with trypsin to remove PfEMP-1 proteolytically also induced significantly higher IFN-g production compared to nondigested iRBC (Figure 1D), confirming that the differences in the ability of the parental and transgenic iRBC to induce IFN-g are not due to differences in parasite maturation or a consequence of the introduction of the plasmid. In addition, knockdown of PfEMP-1 does not affect expression or trafficking of any other known infected erythrocyte surface proteins (T.S.V. and A.F.C., unpublished data). Furthermore, loss of PfEMP-1 had no effect on TNF-a, IL-6, IL-10, IL-2, or IL-4 production by PBMC (Figure S1 in the Supplemental Data available with this article online). Therefore, the suppressive effect of PfEMP-1 appears to be specific for pathways responsible for innate IFN-g output. PfEMP-1 Negatively Regulates IFN-g Production by gd-T, NK, and ab-T Cells We and others have previously shown that IFN-g is rapidly produced by gd-T cells, NK cells, and ab-T cells following stimulation of naive human PBMC with iRBC (ArtavanisTsakonas and Riley, 2002; D’Ombrain et al., 2007; Hensmann and Kwiatkowski, 2001). We have further demonstrated that, in the majority of donors, gd-T cells are the predominant source of early IFN-g, while NK cells and ab-T cells make only minor contributions (D’Ombrain et al., 2007). To identify which early IFN-g-producing cell types are suppressed by PfEMP-1, PBMC were stimulated for 24 hr with 3D7 parental iRBC, 3D7 PfEMP-1 iRBC, or uRBC, harvested, and stained for intracellular IFN-g in combination with CD3, gdTCR, and CD56. gd-T cells (CD3+gdTCR+), NK cells (CD3 CD56+), and to a lesser extent ab-T cells (CD3+gdTCR ), were each negatively regulated by PfEMP-1 with respect to early IFN-g production (Figure 2). These results contrast with previous studies that found recombinant CIDR-1a (a subdomain of PfEMP-1) caused the polyclonal activation of NK cells and CD4 T cells to produce IFN-g (Allsopp et al., 2002; Ndungu et al., 2006). Furthermore, recombinant CIDR-1a was also found to induce IL-10 production by PBMC-derived DC (Allsopp et al., 2002; Ndungu et al., 2006); however, the PfEMP-1 null transgenic iRBC had no effect on IL-10 output from PBMC (Figure S1). The use of recombinant proteins expressed in either bacterial or mammalian cells may be subject to effects arising from glycosylation, LPS contamination, molar ratio/concentration dependence effects, misfolding, and expression out of appropriate cellular context. In contrast, the isogenic parasite lines used in the present study differ only in the presence or absence of surface PfEMP-1, which is expressed in situ in the correct context and concentration on the iRBC surface. Furthermore, recombinant CIDR-1a was used at 5 mg/ml (Allsopp et al., 2002; Ndungu et al., 2006), or 1.8 3 10 7 M. Assum-

Figure 2. PfEMP-1 Negatively Regulates Early IFN-g Production by gd-T, NK, and ab-T Cells PBMC (2 3 105) from eight malaria-naive donors were stimulated for 24 hr with 6 3 105 3D7 parental iRBC, 3D7 PfEMP-1 iRBC, or uRBC. PBMC were harvested and stained for intracellular IFN-g, gdTCR, CD3, and CD56. IFN-g+ cells were gated (A) and identified as CD3 CD56+ NK cells (B), CD3+gdTCR+ gd-T cells (C), or CD3+gdTCR ab-T cells (C). FACS plots in (A)–(C) are from one representative donor. FACS data from all eight donors is displayed graphically in (D) (uppercase letters indicate donors).

ing PfEMP-1 has 5000 copies/cell, the concentration or ligand density of native PfEMP-1 in standard costimulation assays is approximately 2.5 3 10 11 M, 72,000 times less than that used in the recombinant protein studies. Importantly, the change in phenotype upon loss of PfEMP-1 in the present study indicates bioactivity at physiological levels of expression. PfEMP-1 Modulates Innate IFN-g Production via a CD36-Independent Pathway The simultaneous negative regulation of IFN-g production by gd-T cells, NK cells, and ab-T cells (Figure 2) suggests

132 Cell Host & Microbe 2, 130–138, August 2007 ª2007 Elsevier Inc.

Cell Host & Microbe PfEMP-1 Suppresses Early IFN-g Production

that PfEMP-1 might signal via a common mechanism. Binding of the CIDR-1a domain of PfEMP-1 to CD36 has been proposed to mediate the reduction in the capacity of DC to stimulate secondary T cell responses following parasite stimulation (Urban et al., 1999, 2001b). Furthermore, phagocytosis of nonopsonized infected erythrocytes by macrophages was shown to be dependent on CD36 (McGilvray et al., 2000). Thus, we sought to determine whether the PfEMP-1-mediated negative regulation of IFN-g production by gd-T cells, NK cells, and ab-T cells requires the CIDR-1a/CD36 receptor-ligand interaction. This hypothesis was tested using the CS2 strain of P. falciparum, which, unlike 3D7, expresses the var2CSA PfEMP-1, which lacks a CIDR-1a domain and binds to CSA but not CD36 (Figure 3). Consequently, CD36-dependent pathways are not initiated by CS2 parasites. PBMC were stimulated as previously with CS2 parental iRBC, CS2 Skeletal Binding Protein-1 (SBP-1) KO iRBC (which lack expression of PfEMP-1 on the infected erythrocyte surface) (Maier et al., 2006), or uRBC. As shown in Figure 3, the CS2 parental and SBP-1 KO PfEMP-1 iRBC recapitulate the results obtained with the 3D7 parental and 3D7 PfEMP-1 iRBC (Figure 1), with one donor producing 19.6-fold more IFN-g in the absence of PfEMP-1 (Figure 3). Clearly, suppression of IFN-g production by PfEMP-1 occurs via a pathway independent of CIDR-1a/CD36 interactions. The PfEMP-1 domain that mediates suppression of IFN-g is likely to be conserved between 3D7 and CS2 and thus may involve a Duffy binding-like (DBL) domain. We are currently creating PfEMP-1 null parasites expressing defined PfEMP-1 domains to identify the molecular basis of PfEMP-1 immunoregulatory activity and to identify the corresponding host receptor. PfEMP-1 Is Not a Ligand for NK Cell or T Cell Cytotoxicity In addition to cytokine production, NK cells, CD8+CD4 gd-T cells, and CD8+ ab-T cells also possess cytotoxic capabilities. gd-T cells have been shown to be cytotoxic for iRBC, while CD8 T cells are not, due to the lack of MHC class I expression on iRBC (Troye-Blomberg et al., 1999). LAMP-1 (CD107a) appears on the surface of effector cells following degranulation and can be used as a marker of cytotoxic activation (Alter et al., 2004). NK cells have previously been reported to lyse iRBC (Chaicumpa et al., 1983; Orago and Facer, 1991) and induce expression of LAMP-1 following exposure to iRBC (Korbel et al., 2005). Slight changes in NK cell activation status were detected after incubation with peptidic DBL-1a (a PfEMP-1 subdomain) which was proposed to bind the Natural Cytotoxicity Receptors NKp30 and NKp46 and induce NK cell cytotoxicity (Mavoungou et al., 2007). In that study, DBL-1a peptides were used at 40 mg/well, or around 2 3 10 4 2 3 10 5 M, assuming a peptide molecular mass of approximately 2200 Daltons. This is approximately 2–10 million times greater than the ligand density of native PfEMP-1 on the surface of iRBC. Therefore, we sought to test the role of PfEMP-1 in NK cell cytotoxicity further using isogenic parasites differing only in the pres-

Figure 3. PfEMP-1 Modulates Innate IFN-g Production via a CD36-Independent Pathway The western blot (prepared as for Figure 1A) confirms that parental CS2 expresses PfEMP-1 on the infected erythrocyte surface, whereas DSBP-1KO PfEMP-1 does not (A). The typical PfEMP-1 extracellular domain from a CD36-binding 3D7 parental parasite is shown in (B) (Robinson et al., 2003). The CIDR-1a domain of PfEMP-1 mediates binding to CD36 (Baruch et al., 1999). The extracellular domain of var2CSA, the predominant PfEMP-1 protein expressed by CS2 parental parasites that mediates CSA binding, is illustrated in (C) (Gardner et al., 2002). Note the absence of a CIDR-1a domain in var2CSA PfEMP-1 (C), rendering this parasite unable to bind CD36 (Gardner et al., 2002). PBMC (2 3 105) from ten malaria-naive donors were stimulated for 24 hr with 6 3 105 CS2 parental iRBC, DSBP-1KO PfEMP-1 iRBC, or uRBC. Supernatants were harvested and assayed for IFN-g. The mean IFN-g detected in supernatants from ten donors is shown (D). Error bars are the SEM for R2 experiments. In 10/10 donors, the mean level of IFN-g produced by PBMC following stimulation with DSBP1KO PfEMP-1 iRBC exceeds that for CS2 parental iRBC (p < 0.001; nonparametric sign test). Upper case letters indicate donors.

ence or absence of PfEMP-1 expressed in situ in the correct context and concentration on the iRBC surface. A role for PfEMP-1 in gd-T cell cytotoxicity was also explored. Accordingly, we compared the ability of 3D7 parental iRBC or 3D7 PfEMP-1 iRBC to induce LAMP-1 expression on NK cells and T cells following 6 hr of coculture with PBMC from healthy, malaria-naive adult donors. The K562 cell line, a well-defined target for NK cell cytotoxicity, was used as a positive control, while uRBC served as the negative control. LAMP-1 was expressed at high levels on NK cells and at modest levels on T cells,

Cell Host & Microbe 2, 130–138, August 2007 ª2007 Elsevier Inc. 133

Cell Host & Microbe PfEMP-1 Suppresses Early IFN-g Production

Figure 4. PfEMP-1 Is Not a Ligand for Malaria-Induced NK Cell or T Cell Cytotoxicity PBMC (2 3 105) were cultured for 6 hr in the presence of 6 3 105 K562 cells, 3D7 parental iRBC, 3D7 PfEMP-1 iRBC, or uRBC. PBMC were harvested and stained for CD3, CD56, and the cytotoxicity marker LAMP-1. CD3 CD56+ NK cells and CD3+ T cells were gated (A) and analyzed for LAMP-1 expression in response to each treatment (B). FACS plots in (A) and (B) are from one representative donor. Mean LAMP-1 expression on NK cells and T cells from ten donors is shown in (C) and (D), respectively. Error bars are the SEM for ten donors. p values from Student’s t tests show no statistically significant differences between LAMP-1 expression on NK cells or T cells stimulated with 3D7 parental iRBC or 3D7 PfEMP-1 iRBC (C and D).

following incubation with K562 target cells (Figure 4). Furthermore, LAMP-1 expression was detected above background levels on both NK cells and T cells following incubation with 3D7 parental iRBC and 3D7 PfEMP-1 iRBC (Figure 4). Importantly, no statistically significant differences between LAMP-1 expression levels on NK cells or T cells stimulated with 3D7 parental or 3D7 PfEMP-1 iRBC were detected (Figure 4), thus demonstrating that PfEMP-1, inclusive of a DBL-1a subdomain and expressed under physiologically relevant conditions, is not a ligand for NK cell-mediated or T cell-mediated cytotoxicity of iRBC. Conclusions Plasmodium falciparum causes rapid activation of macrophages and DC through the production of pathogenassociated molecular patterns, such as glycosylphosphatidylinositol (Schofield and Hackett, 1993) and malarial

DNA (associated with haemozoin) (Parroche et al., 2007), which act through Toll-like receptors (Krishegowda et al., 2005; Parroche et al., 2007; Zhu et al., 2005). Subsequently, NK receptor- and gdTCR-dependent innate and ‘‘intermediate’’ adaptive pathways become activated, resulting in rapid production of the proinflammatory cytokine IFN-g (Schofield and Grau, 2005). Following 24 hr stimulation of human PBMC with live P. falciparuminfected erythrocytes, gd-T cells expressing NK receptors represent the major cellular source of innate IFN-g, with minor contributions from NK cells and ab-T cells (D’Ombrain et al., 2007). The activating ligand for NK cells remains to be identified, while gd-T cells are known to be activated by malarial phosphoantigens (Behr et al., 1996; Pichyangkul et al., 1997) and rapid activation of naive ab-T cells by P. falciparum is thought to be due to the presence of cross-reactive antigens that are expressed on a wide range of common microbes (Currier et al.,

134 Cell Host & Microbe 2, 130–138, August 2007 ª2007 Elsevier Inc.

Cell Host & Microbe PfEMP-1 Suppresses Early IFN-g Production

1992; Zevering et al., 1992). Here we have identified an immunoregulatory mechanism of P. falciparum whereby the early IFN-g response of gd-T cells, NK cells, and ab-T cells is strongly suppressed by the presence of PfEMP-1. The host receptor(s) utilized by PfEMP-1 to suppress IFN-g production by gd-T, NK, and ab-T cells remains to be identified; however, this pathway is independent of CIDR-1a/CD36 interactions and thus occurs via a different mechanism than that proposed for PfEMP-1 interactions with DC and macrophages (McGilvray et al., 2000; Urban et al., 1999, 2001b). As plasmacytoid DC promote early gd-T cell activation (Pichyangkul et al., 2004), myeloid DC stimulate malaria-specific ab-T cells (Urban et al., 1999), and both macrophages and myeloid DC are required for NK cell activation (Baratin et al., 2005; Newman et al., 2006), it is unlikely that PfEMP-1 suppresses early IFN-g production via a common antigen-presenting cell. Furthermore, the effect of PfEMP-1 is specific to IFN-g, with no effect on a range of proinflammatory and antiinflammatory cytokines, including those typically associated with macrophage and DC activation. Therefore, it is plausible that PfEMP-1 may signal directly to gd-T cells, NK cells, and ab-T cells via a common receptor. Interestingly, in humans, malaria-responsive gd-T and NK cells more frequently express the inhibitory NK receptor NKG2A (Artavanis-Tsakonas et al., 2003; D’Ombrain et al., 2007). Thus, PfEMP-1 may engage NKG2A (and/ or other inhibitory NK receptors), resulting in suppression of IFN-g production. Alternatively, the host receptor may be a more general, non-NK receptor that negatively regulates leukocyte activation. The suppression of IFN-g by PfEMP-1 appears to be due to a PfEMP-1 domain that is conserved between 3D7 and CS2 parasite lines and thus may involve a DBL domain. We are currently creating PfEMP-1 null parasites expressing defined PfEMP-1 domains to identify both the PfEMP-1 subdomain and the corresponding host receptor that mediates innate IFN-g suppression. The suppression of early IFN-g production by PfEMP-1 was quantitatively highly significant (up to 20-fold). Given the dual role of IFN-g in malaria protection and pathogenesis, these data raise the possibility that this process might impact either IFN-g-dependent antiparasite immunity or IFN-g-dependent immunopathogenesis. For example, protection against reinfection with P. falciparum malaria has been associated with higher plasma levels of IFN-g (Luty et al., 1999), possibly resulting from increased macrophage activation and enhanced parasite clearance (Hisaeda et al., 2005). Furthermore, in the P. chabaudi murine model of malaria, ifn-g / and ifn-g receptor / mice exhibit higher mortality rates (Favre et al., 1997; Su and Stevenson, 2000), while in the P. berghei-K173 murine model of malaria, a 24 hr IFN-g peak, not observed in the closely related P. berghei-ANKA cerebral malaria model, was associated with protection from the cerebral syndrome (Mitchell et al., 2005). The suppression of early IFN-g production by PfEMP-1, therefore, may represent an immune evasion strategy employed by P. falciparum, which serves to downregulate the antiparasitic IFN-g

response. Furthermore, IFN-g inhibits the infectivity of P. berghei sporozoites for hepatocytes in mice and rats and is a critical effector mechanism in immunity to liver stage infection (Schofield et al., 1987a, 1987b). It is thus plausible that the suppression of IFN-g by PfEMP-1 may also favor reinfection upon subsequent sporozoite challenge in humans. However, although critical in protective antimalarial immunity, IFN-g may also play an important role in pathogenesis. In humans, polymorphisms in both IFN-g and IFN-g Receptor 1 genes are associated with severe malaria (Kwiatkowski, 2005), and high levels of circulating IFN-g were found to be indicators of poor prognostic outcome (Hunt and Grau, 2003). Similarly, in the P. berghei-ANKA model of cerebral malaria, neutralization of IFN-g in vivo prevents fatalities (Grau et al., 1986), and ifn-g / and ifn-g receptor / mice fail to develop the cerebral syndrome (Amani et al., 2000). The suppression of IFN-g by PfEMP-1 may thus also be beneficial for the host, possibly serving to reduce the severity of clinical symptoms and immunopathogenesis. Clearly, the role of IFN-g in malaria protection and pathogenesis is complex and needs to be addressed in the context of adequate epidemiological study designs. Therefore, to allow the impact of PfEMP-1-mediated immune regulation to be assessed in relation to relevant parasitological and clinical outcome variables, we are currently conducting longitudinal cohort field studies in Papua New Guinea (Michon et al., 2007).

EXPERIMENTAL PROCEDURES PBMC Preparation Blood from healthy nonimmune adults was obtained from the Australian Red Cross Blood Service under appropriate ethical approval. Blood was diluted 1:1 in PBS, and PBMC were separated by density centrifugation with Ficoll-Paque PLUS (Amersham Biosciences, Sweden). PBMC were washed and resuspended in complete medium (RPMI 1640, 5% heat-inactivated FBS, 2 mM L-glutamine, 25 mM HEPES, 50 mM 2-ME, 100 U/ml penicillin, and 100 mg/ml streptomycin sulfate) for use in cytokine induction and LAMP-1 cytotoxicity assays.

Cultivation of P. falciparum 3D7, CS2, and transgenic P. falciparum parasites were cultivated at 37 C with 5% CO2, 1% O2, and 1% N2 in 150 cm2 flasks at 4% hematocrit using O+ human erythrocytes (Australian Red Cross Blood Service [ARCBS]) in RPMI 1640 buffered with 25 mM HEPES and supplemented with 0.5% Albumax II, 2 mg/ml glucose, 28 mM sodium bicarbonate, 25 mg/ml gentamycin, and 100 mg/ml hypoxanthine. Parasite development was monitored by light microscopy using methanol-fixed, Giemsa-stained thin blood films. The construct used to produce the transgenic 3D7 PfEMP-1 parasite line was made by transfecting 3D7 with pHBupsCR (Voss et al., 2006). The blasticidin deaminase gene serves to maintain stable integration of the plasmid in the parasite chromosome. 3D7 PfEMP-1 parasites were cultured with 4 mg/ml blasticidin-S and 4 nM WR99210 (Voss et al., 2006). Trypsin-digested 3D7 parental iRBC were included as an additional control for the absence of PfEMP-1. Trypsin digestion of live iRBC was carried out as previously described (Beeson et al., 2000). Parasites were sorbitol synchronized, knob selected, and screened for mycoplasma negativity. iRBC were purified by magnetic cell sorting CS columns (Miltenyi Biotec, CA) and resuspended in complete medium for use in cytokine induction and LAMP-1 cytotoxicity assays.

Cell Host & Microbe 2, 130–138, August 2007 ª2007 Elsevier Inc. 135

Cell Host & Microbe PfEMP-1 Suppresses Early IFN-g Production

Western Blot Immunoblot analysis was performed using iRBC purified by magnetic cell sorting CS columns (Miltenyi Biotech, CA). iRBC (1 3 107) were extracted with 150 ml of 1% (v/v) Triton X-100 in PBS and protease inhibitors (Roche). The Triton X-100 insoluble pellet was then extracted with 2% SDS (w/v) and loaded onto the gel in reducing SDS sample buffer (Baruch et al., 2002). Cytokine Induction Assay In 96-well plates, 2 3 105 PBMC were stimulated with 6 3 105 live parental, trypsin-digested parental, or transgenic iRBC, or 6 3 105 uRBC per well. Cultures were incubated for 24 hr (37 C, 5% CO2). Golgistop (BD PharMingen) was added at 20 hr. Cells were harvested and analyzed by flow cytometry. Culture supernatants were analyzed for cytokine content by ELISA and Cytometric Bead Array (BD PharMingen). IFN-g ELISA IFN-g was detected in culture supernatants by standard sandwich ELISA. Purified mouse anti-human IFN-g (NIB42) mAb was used for capture, and biotinylated mouse anti-human IFN-g (4S.B3) was used for detection (BD Biosciences, CA). Detection of TNF-a, IL-6, IL-10, IL-2, and IL-4 TNF-a, IL-6, IL-10, IL-2, and IL-4 were detected in supernatants from the cytokine induction assays of five donors using the BD Cytometric Bead Array Human Th1/Th2 Cytokine Kit II (BD Biosciences, CA) according to the manufacturer’s instructions. Samples were analyzed in a FACSCalibur flow cytometer (BD Biosciences, CA). Data analysis was performed with the BD Cytometric Bead Array Software (BD Biosciences, CA). LAMP-1 Cytotoxicity Assay K562 cells (ATCC, Manassas, VA) were maintained at 37 C and 5% CO2 in IMDM supplemented with 4 mM L-glutamine, 1.5g/l sodium bicarbonate, and 10% FBS. In 96-well plates, 2 3 105 PBMC were stimulated with 6 3 105 parental or transgenic iRBC, K562, or uRBC cells per well. Cultures were incubated for 6 hr (37 C, 5% CO2). Golgistop (BD PharMingen) was added after 1 hr. Cells were harvested and analyzed for LAMP-1 expression by flow cytometry. Flow Cytometry Anti-TCR-g/d FITC, CD3-APC, CD56-biotin (conjugated with Streptavidin-PerCP-Cy5.5), IFN-g-PE, IgG1l isotype control-PE, CD107a (LAMP-1)-biotin (conjugated with Streptavidin-PerCP-Cy5.5), and CD56-PE (BD Biosciences, CA) were used for surface, intracellular, or isotype control staining according to the manufacturer’s recommendations. PBMC were surface stained for 40 min with appropriate combinations of labeled mAbs. For biotin-labeled antibodies, there was a second 40 min incubation for conjugation with fluorescently labeled streptavidin. PBMC from cytokine induction assays were stained for intracellular IFN-g using the BD Cytofix/Cytoperm Kit (BD Biosciences, CA) according to the manufacturer’s instructions, with appropriate isotype-matched controls. Finally, cells were resuspended in 200 ml PBS and analyzed in a FACSCalibur flow cytometer (BD Biosciences, CA). Dead cells were gated out by forward and side scatter, and 100,000 live cell events were collected. Data analysis was performed with Weasel v2.2.3 software. Statistical Analyses Nonparametric sign tests were used to determine statistically significant differences in IFN-g production between parental and transgenic iRBC as well as between nontreated parental and trypsin-digested parental iRBC at the population level (Lehmann, 1975). As the Central Limit Theorem applied, Student’s t tests were used to determine statistically significant differences in LAMP-1 expression between parental and transgenic iRBC.

Supplemental Data The Supplemental Data include one supplemental figure and can be found with this article online at http://www.cellhostandmicrobe.com/ cgi/content/full/2/2/130/DC1/. ACKNOWLEDGMENTS We thank anonymous donors and the Australian Red Cross Blood Service for providing blood samples. This study was supported by grants from the National Health and Medical Research Council of Australia and from the National Institutes of Health. A.F.C. and L.S. are International Research Scholars of the Howard Hughes Medical Institute. Received: April 2, 2007 Revised: June 12, 2007 Accepted: June 28, 2007 Published: August 15, 2007 REFERENCES Allsopp, C.E., Sanni, L.A., Reubsaet, L., Ndungu, F., Newbold, C., Mwangi, T., Marsh, K., and Langhorne, J. (2002). CD4 T cell responses to a variant antigen of the malaria parasite Plasmodium falciparum, erythrocyte membrane protein-1, in individuals living in malariaendemic areas. J. Infect. Dis. 185, 812–819. Alter, G., Malenfant, J.M., and Altfield, M. (2004). CD107a as a functional marker for the identification of natural killer cell activity. J. Immunol. Methods 294, 15–22. Amani, V., Vigario, A.M., Belnoue, E., Marussig, M., Fonseca, L., Mazier, D., and Renia, L. (2000). Involvement of IFN-gamma receptor-medicated signaling in pathology and anti-malarial immunity induced by Plasmodium berghei infection. Eur. J. Immunol. 30, 1646–1655. Artavanis-Tsakonas, K., and Riley, E.M. (2002). Innate immune response to malaria: Rapid induction of IFN-gamma from human NK cells by live Plasmodium falciparum-infected erythrocytes. J. Immunol. 169, 2956–2963. Artavanis-Tsakonas, K., Eleme, K., McQueen, K.L., Cheng, N.W., Parham, P., Davis, D.M., and Riley, E.M. (2003). Activation of a subset of human NK cells upon contact with Plasmodium falciparum-infected erythrocytes. J. Immunol. 171, 5396–5405. Baratin, M., Roetynck, S., Lepolard, C., Falk, C., Sawadogo, S., Uematsu, S., Akira, S., Ryffel, B., Tiraby, J.-G., Alexopoulou, L., et al. (2005). Natural killer cell and macrophage cooperation in MyD88dependent innate responses to Plasmodium falciparum. Proc. Natl. Acad. Sci. USA 102, 14747–14752. Baruch, D.I., Ma, X.C., Pasloske, B., Howard, R.J., and Miller, L.H. (1999). CD36 peptides that block cytoadherence define the CD36 binding region for Plasmodium falciparum-infected erythrocytes. Blood 94, 2121–2127. Baruch, D.I., Gamain, B., Barnwell, J.W., Sullivan, J.S., Stowers, A., Galland, G.G., Miller, L.H., and Collins, W.E. (2002). Immunization of Aotus monkeys with a functional domain of the Plasmodium falciparum variant antigen induces protection against a lethal parasite line. Proc. Natl. Acad. Sci. USA 99, 3860–3865. Beeson, J.G., Rogerson, S.J., Cooke, B.M., Reeder, J.C., Chai, W., Lawson, A.M., Molyneux, M.E., and Brown, G.V. (2000). Adhesion of Plasmodium falciparum-infected erythrocytes to hyaluronic acid in placental malaria. Nat. Med. 6, 86–90. Behr, C., Poupot, R., Peyrat, M.A., Poquet, Y., Constant, P., Dubois, P., Bonneville, M., and Fournie, J.J. (1996). Plasmodium falciparum stimuli for human gammadelta T cells are related to phosphorylated antigens of mycobacteria. Infect. Immun. 64, 2892–2896. Chaicumpa, W., Chatkaeomorakot, A., Patarapotikul, J., and Looareesuwan, S. (1983). Effects of human natural killer (NK) cells on Plasmodium falciparum infected red cells. Asian Pac. J. Allergy Immunol. 1, 135–139.

136 Cell Host & Microbe 2, 130–138, August 2007 ª2007 Elsevier Inc.

Cell Host & Microbe PfEMP-1 Suppresses Early IFN-g Production

Currier, J., Sattabongkot, J., and Good, M.F. (1992). ‘Natural’ T cells responsive to malaria: Evidence implicating immunological crossreactivity in the maintenance of TCR alpha beta+ malaria-specific responses from non-exposed donors. Int. Immunol. 4, 985–994.

Michon, P., Cole-Tobian, J.L., Dabod, E., Schoepflin, S., Igu, J., Susapu, M., Tarongka, N., Zimmerman, P.A., Reeder, J.C., Beeson, J.G., et al. (2007). The risk of malarial infections and disease in Papua New Guinean children. Am. J. Trop. Med. Hyg. 76, 997–1008.

D’Ombrain, M.C., Hansen, D.S., Simpson, K.M., and Schofield, L. (2007). gd-T cells expressing NK receptors predominate over NK cells and conventional T cells in the innate IFN-g response to Plasmodium falciparum malaria. Eur. J. Immunol. 37, 1864–1873.

Mitchell, A.J., Hansen, A.M., Hee, L., Ball, H.J., Potter, S.M., Walker, J.C., and Hunt, N.H. (2005). Early cytokine production is associated with protection from murine cerebral malaria. Infect. Immun. 73, 5645–5653.

Favre, N., Ryffel, B., Bordmann, G., and Rudin, W. (1997). The course of Plasmodium chabaudi chabaudi infections in interferon-g receptor deficient mice. Parasite Immunol. 19, 375–383.

Ndungu, F.M., Sanni, L., Urban, B., Stephens, R., Newbold, C.I., Marsh, K., and Langhorne, J. (2006). CD4 T cells from malarianonexposed individuals respond to the CD36-Binding Domain of Plasmodium falciparum erythrocyte membrane protein-1 via an MHC class II-TCR-independent pathway. J. Immunol. 176, 5504–5512.

Gardner, M.J., Hall, N., Fung, E., White, O., Berriman, M., Hyman, R.W., Carlton, J.M., Pain, A., Nelson, K.E., Bowman, S., et al. (2002). Genome sequence of the human malaria parasite Plasmodium falciparum. Nature 419, 498–511. Grau, G.E., Piguet, P.F., Engers, H.D., Louis, J.A., Vassalli, P., and Lambert, P.H. (1986). L3T4+ T lymphocytes play a major role in the pathogenesis of murine cerebral malaria. J. Immunol. 137, 2348–2354. Hensmann, M., and Kwiatkowski, D. (2001). Cellular basis of early cytokine response to Plasmodium falcipaum. Infect. Immun. 69, 2364–2371. Hisaeda, H., Yasutomo, K., and Himeno, K. (2005). Malaria: Immune evasion by parasites. Int. J. Biochem. Cell Biol. 37, 700–706. Hunt, N.H., and Grau, G.E. (2003). Cytokines: Accelerators and brakes in the pathogenesis of cerebral malaria. Trends Immunol. 24, 491–499. Joergensen, L., Turner, L., Magistrado, P., Dahlback, M.A., Vestergaard, L.S., Lusingu, J.P., Lemnge, M., Salanti, A., Theander, T.G., and Jensen, A.T. (2006). Limited cross-reactivity among domains of the Plasmodium falciparum clone 3D7 erythrocyte membrane protein 1 family. Infect. Immun. 74, 6778–6784. Korbel, D.S., Newman, K.C., Almeida, C.R., Davis, D.M., and Riley, E.M. (2005). Heterogeneous human NK cell responses to Plasmodium falciparum-infected erythrocytes. J. Immunol. 175, 7466–7473. Krishegowda, G., Hajjar, A.M., Zhu, J., Douglass, E.J., Uematsu, S., Akira, S., Woods, S.A., and Gowda, D.C. (2005). Induction of proinflammatory responses in macrophages by the glycosylphosphatidylinositols of Plasmodium falciparum. J. Biol. Chem. 280, 8606–8616. Kwiatkowski, D.P. (2005). How malaria has affected the human genome and what human genetics can teach us about malaria. Am. J. Hum. Genet. 77, 171–192. Kyriacou, H.M., Stone, G.N., Challis, R.J., Raza, A., Lyke, K.E., Thera, M.A., Kone, A.K., Doumbo, O.K., Plowe, C.V., and Rowe, J.A. (2006). Differential var gene transcription in Plasmodium falciparum isolates from patients with cerebral malaria compared to hyperparasitaemia. Mol. Biochem. Parasitol. 150, 211–218. Lehmann, E.L. (1975). Nonparametrics: Statistical Methods Based on Ranks (San Francisco: McGraw-Hill International Book Company). Luty, A.J., Lell, B., Schmidt-Ott, R., Lehman, L.G., Luckner, D., Greve, B., Matsusek, P., Herbich, K., Schmid, D., Migot-Nabias, F., et al. (1999). Interferon-g responses are associated with resistance to reinfection with Plasmodium falciparum in young African children. J. Infect. Dis. 179, 980–988. Maier, A.G., Rug, M., O’Neill, M.T., Beeson, J.G., Marti, M., Reeder, J., and Cowman, A. (2006). Skeleton Binding Protein 1 functions at the parasitophorous vacuole membrane to traffic PfEMP1 to the Plasmodium falciparum-infected erythrocyte surface. Blood 109, 1289–1297. Mavoungou, E., Held, J., Mewono, L., and Kremsner, P.G. (2007). A Duffy binding-like domain is involved in the NKp30-mediated recognition of Plasmodium falciparum-parasitized erythrocytes by natural killer cells. J. Infect. Dis. 195, 1521–1531. McGilvray, I.D., Serghides, L., Kapus, A., Rotstein, O.D., and Kain, K.C. (2000). Nonopsonic monocyte/macrophage phagocytosis of Plasmodium falciparum-parasitized erythrocytes: A role for CD36 in malarial clearance. Blood 96, 3231–3240.

Newbold, C.I. (1999). Antigenic variation in Plasmodium falciparum: Mechanisms and consequences. Curr. Opin. Microbiol. 2, 420–425. Newman, K.C., Korbel, D.S., Hafalla, J.C., and Riley, E.M. (2006). Cross-talk with myeloid accessory cells regulates human natural killer cell interferon-g responses to malaria. PLoS Pathogens 2, e118. 10.1371/journal.ppat.0020118. Orago, A.S., and Facer, C.A. (1991). Cytotoxicity of human natural killer (NK) cell subsets for Plasmodium falciparum erythrocytic schizonts: Stimulation by cytokines and inhibition by neomycin. Clin. Exp. Immunol. 86, 22–29. Parroche, P., Lauw, F.N., Goutagny, N., Latz, E., Monks, B.G., Visintin, A., Halmen, K.A., Lamphier, M., Olivier, M., Bartholomeu, B.C., et al. (2007). Malaria hemozoin is immunologically inert but radically enhances innate responses by presenting malaria DNA to Toll-like receptor 9. Proc. Natl. Acad. Sci. USA 104, 1919–1924. Pichyangkul, S., Saengkrai, P., Yongvanitchit, K., Stewart, A., and Heppner, D.G. (1997). Activation of gammadelta T cells in malaria: Interaction of cytokines and a schizont-associated Plasmodium falciparum antigen. J. Infect. Dis. 176, 233–241. Pichyangkul, S., Yongvanitchit, K., Kum-arb, U., Hemmi, H., Akira, S., Krieg, A.M., Heppner, D.G., Stewart, V.A., Hasegawa, H., Looareesuwan, S., et al. (2004). Malaria blood stage parasites activate human plasmacytoid dendritic cells and murine dendritic cells through a tolllike receptor 9-dependent pathway. J. Immunol. 172, 4926–4933. Robinson, B.A., Welch, T.L., and Smith, J.D. (2003). Widespread functional specialization of Plasmodium falciparum erythrocyte membrane protein 1 family members to bind CD36 analysed across a parasite genome. Mol. Microbiol. 47, 1265–1278. Schofield, L., Ferreira, A., Altszuler, R., Nussenzweig, V., and Nussenzweig, R.S. (1987a). Interferon-gamma inhibits the intrahepatocytic development of malaria parasites in vitro. J. Immunol. 139, 2020–2025. Schofield, L., Villaquiran, J., Ferreira, A., Schellekens, H., Nussenzweig, R., and Nussenzweig, V. (1987b). Gamma interferon, CD8+ T cells and antibodies required for immunity to malaria sporozoites. Nature 330, 664–666. Schofield, L., and Hackett, F. (1993). Signal transduction in host cells by a glycosylphosphatidylinositol toxin of malaria parasites. J. Exp. Med. 177, 145–153. Schofield, L., and Grau, G.E. (2005). Immunological processed in malaria pathogenesis. Nat. Rev. Immunol. 5, 722–735. Su, Z., and Stevenson, M.M. (2000). The central role of endogenous IFN-g in protective immunity against acute blood-stage Plasmodium chabaudi AS infection. Infect. Immun. 68, 4399–4406. Troye-Blomberg, M., Worku, S., Tangteerawatana, P., Jamshaid, R., Soderstrom, K., Elghazali, G., Moretta, L., Hammarstrom, M., and Mincheva-Nilsson, L. (1999). Human gamma delta T cells that inhibit the in vitro growth of the asexual blood stages of the Plasmodium falciparum parasite express cytolytic and proinflammatory molecules. Scand. J. Immunol. 50, 642–650. Urban, B.C., Ferguson, D.J., Pain, A., Willcox, N., Plebanski, M., Austyn, J.M., and Roberts, D.J. (1999). Plasmodium falciparum-infected

Cell Host & Microbe 2, 130–138, August 2007 ª2007 Elsevier Inc. 137

Cell Host & Microbe PfEMP-1 Suppresses Early IFN-g Production

erythrocytes modulate the maturation of dendritic cells. Nature 400, 73–77. Urban, B.C., Mwangi, T., Ross, A., Kinyanjui, S., Mosobo, M., Kai, O., Lowe, B., Marsh, K., and Roberts, D.J. (2001a). Peripheral blood dendritic cells in children with acute Plasmodium falciparum malaria. Blood 98, 2859–2861. Urban, B.C., Willcox, N., and Roberts, D.J. (2001b). A role for CD36 in the regulation of dendritic cell function. Proc. Natl. Acad. Sci. USA 98, 8750–8755. Urban, B.C., Ing, R., and Stevenson, M.M. (2005). Early interactions between blood-stage plasmodium parasites and the immune system. Curr. Top. Microbiol. Immunol. 297, 25–70.

Voss, T.S., Healer, J., Marty, A.J., Duffy, M.F., Thompson, J.K., Beeson, J.G., Reeder, J.C., Crabb, B.S., and Cowman, A.F. (2006). A var gene promoter controls allelic exclusion of virulence genes in Plasmodium falciparum malaria. Nature 439, 1004–1008. Zevering, Y., Amante, F., Smillie, A., Currier, J., Smith, G., Houghten, R.A., and Good, M.F. (1992). High frequency of malaria-specific T cells in non-exposed humans. Eur. J. Immunol. 22, 689–696. Zhu, J., Krishegowda, G., and Gowda, D.C. (2005). Induction of proinflammatory responses in macrophages by the glycosylphosphatidylinositols of Plasmodium falciparum. J. Biol. Chem. 280, 8617–8627.

138 Cell Host & Microbe 2, 130–138, August 2007 ª2007 Elsevier Inc.