Early cytokine induction by Plasmodium falciparum is not a classical endotoxin-like process.

We have investigated the widely held view that malaria parasites induce pro-inflammatory cytokines primarily through an endotoxin-like stimulatory effect on macrophages. We report that the pattern of cytokine production by non-immune human peripheral blood mononuclear cells following stimulation by Plasmodium falciparum-infected erythrocytes (Pfe) in vitro differs considerably from that induced by bacterial endotoxin. The Pfe-induced TNF response at day 1 is associated with a much higher level of IFN-gamma production and a much lower level of IL-12 p40 and IL-10 expression than a comparable endotoxin-induced TNF response. Both CD3(+) and CD14(+) populations are required for this early TNF response to Pfe, whereas the endotoxin-induced response is unaffected by depletion of the CD3(+) population. Pfe fails to stimulate the monocyte-like cell line MonoMac6 to express pro-inflammatory cytokines. These findings suggest that the early inflammatory response to malaria is critically dependent on lymphocyte subpopulations that play a lesser role in the response to bacterial endotoxin.

The malarial fever response--pathogenesis, polymorphism and prospects for intervention.

It is estimated that over 200 million people each year suffer debilitating attacks of malarial fever, and roughly 2 million of these episodes are fatal. The fever is caused by tumour necrosis factor (TNF) and other pyrogenic cytokines that are released by the host immune system response to products of schizont rupture. TNF has anti-parasitic properties but excessive TNF production is thought to play an important role in the pathogenesis of cerebral malaria. This review summarizes recent attempts to achieve molecular characterization of the parasite components that stimulate the host TNF response, and to define the host and parasite factors that affect the level of TNF production. Of particular interest are host polymorphisms that may regulate TNF gene expression, and naturally acquired antibodies that prevent the parasite from inducing TNF, both of which correlate with the clinical severity of infection. Our understanding of these processes, which are potentially of considerable therapeutic relevance, remains very limited at both the molecular and the epidemiological level.

In vitro induction of nitric oxide by an extract of Plasmodium falciparum.

Malarial illness and pathology is generally accepted to be caused by material released when the infected red cells burst at schizogony. The released material has been partially purified and shown to stimulate macrophages to make TNF. We have extended this work to show that these same preparations, isolated from parasitized erythrocytes, induce the mouse macrophage cell line RAW 264.7 to produce inducible nitric oxide synthase and release nitric oxide. By using cytokine-specific antisera we have found that this induction is independent of TNF and IL-1 alpha and partly independent of IL-1 beta.

Malaria: toxins, cytokines and disease.

In this review the old concept of severe malaria as a toxic disease is re-examined in the light of recent discoveries in the field of cytokines. Animal studies suggest that the induction of TNF by parasite-derived molecules may be partly responsible for cerebral malaria and anemia, while hypoglycaemia may be due to direct effects of similar molecules on glucose metabolism. These molecules appear to be phospholipids and we suggest that when fully characterized they might form the basis of antitoxic therapy for malaria.

A monoclonal antibody that recognizes phosphatidylinositol inhibits induction of tumor necrosis factor alpha by different strains of Plasmodium falciparum.

The clinical symptoms of human malaria are mediated, at least in part, by the release of tumor necrosis factor alpha (TNF) by monocytes and macrophages. We have found that lysates of Plasmodium falciparum-infected erythrocytes stimulate the secretion of TNF from human mononuclear cells, and we have generated several immunoglobulin M monoclonal antibodies (MAbs) that inhibit the induction of TNF by such lysates. Here we describe the properties of MAb 5AB3-11, which causes dose-dependent inhibition of the TNF-inducing factors derived from P. falciparum-infected erythrocytes, with a 50% reduction in TNF secretion at nanomolar concentrations (1 to 2 micrograms/ml). The inhibitory effect appears to be malaria specific in that the induction of TNF by either lipopolysaccharide or lipoteichoic acid is not affected. MAb 5AB3-11 binds to liposomes containing phosphatidylinositol but not to other phospholipid liposomes, showing that it recognizes a phosphatidylinositol-like epitope. MAb 5AB3-11 inhibits the induction of TNF by whole lysates from several strains of P. falciparum which originated from different parts of the tropics, indicating that all of the major TNF-inducing factors derived from Plasmodium-infected erythrocytes contain a common epitope. A phosphatidylinositol-like epitope expressed by Plasmodium-infected erythrocytes may be a suitable immunological target for the prevention or treatment of severe malaria.

Stimulators of tumour necrosis factor production released by damaged erythrocytes.

We sought to characterize factors released by sonicated human erythrocytes that stimulate peripheral blood mononuclear cells (PBMC) to release tumour necrosis factor-alpha (TNF). This response is not inhibited by polymyxin B, indicating that contaminating lipopolysaccharide (LPS) is not responsible. When erythrocyte lysates are fractionated by reverse-phase chromatography using a gradient of n-propanol on Sep-Pak C18 cartridges, the TNF-inducing activity elutes as a single peak. The erythrocyte-derived TNF-inducing activity is unaffected by digestion with proteases but is destroyed by mild base hydrolysis or digestion by lipases, indicating that compounds containing ester-linked acyl chains may be essential. These properties are similar to those of TNF stimulators that we have previously identified in erythrocytes infected with malaria parasites, except that the TNF-inducing activity per cell is about 200 times higher in parasitized erythrocytes than in uninfected erythrocytes. Lipase-digested erythrocyte lysates inhibit the TNF-inducing factors of both normal and malaria-infected erythrocytes, suggesting that lipase digestion creates partial structures which compete with active components for macrophage receptors. Such receptors may recognize a common structure that contains an inositol monophosphate (IMP)-like component, as IMP also inhibits the TNF response to erythrocyte-derived factors and to parasite lysates whereas it does not affect the response to LPS. We conclude that lysed erythrocytes release specific cytokine-inducing factors that may contribute to the fever response to non-infectious tissue injury.

Inhibitory immunoglobulin M antibodies to tumor necrosis factor-inducing toxins in patients with malaria.

Cytokines such as tumor necrosis factor alpha (TNF) appear to play an important role in the pathogenesis of malaria. We have previously shown that TNF is produced in response to substances released at schizont rupture, which we have called malaria toxins. In mice these toxins stimulate a T cell-independent antibody response, generating short-lived immunoglobulin M (IgM) antibodies that inhibit the TNF-inducing activity of the toxins. We report here that a similar antibody response is seen in humans. Serum from a European adult infected with Plasmodium falciparum inhibited the induction of TNF by malaria toxins derived from P. falciparum-infected erythrocytes. We found that IgM antibodies were responsible for the inhibitory activity. These inhibitory antibodies could not be detected in convalescent-phase serum collected from the same patient 6 weeks later or in sera from healthy European and African controls. The antibodies appeared to be malaria specific in that they inhibited TNF induction by a variety of P. falciparum isolates but failed to inhibit TNF induction by bacterial lipopolysaccharide or lipoteichoic acid. The inhibitory antibodies bound to liposomes containing phosphatidylinositol but not other phospholipids. Serum from a European adult infected with P. vivax also inhibited the activity of toxins derived from P. falciparum-infected erythrocytes, and this too was mediated by IgM antibodies which were malaria specific and bound to phosphatidylinositol liposomes.

Phospholipids coupled to a carrier induce IgG antibody that blocks tumour necrosis factor induction by toxic malaria antigens.

Phospholipid-containing antigens of malaria parasites stimulate macrophages to secrete tumour necrosis factor (TNF), induce hypoglycaemia and are toxic to mice. This TNF induction is inhibited by antisera made against the antigens, the inhibitory activity of which can be removed specifically by adsorption to phosphatidylinositol (PI) liposomes. Although the same was true of antisera made against PI, the inhibitory activity of antisera made against some other phospholipids appeared to be directed against a common determinant, probably the phosphate ester head group. We have shown previously that the activity of all the antisera was associated mainly with IgM and was not boosted by repeated injections of the antigens. To try and induce a secondary response against the parasite antigens using non-toxic molecules, mice were immunized with various phosphorylated compounds coupled to keyhole limpet haemocyanin (KLH). Three injections of PI-KLH or of phosphatidylserine (PS) coupled to KLH induced significantly higher titres of inhibitory antibody than one; furthermore, the inhibitory activity was mainly in the IgG fraction. The antisera did not inhibit TNF induction by lipopolysaccharide (LPS) or lipoteichoic acid. However, antisera against PS-KLH, though not PI-KLH, inhibited the induction of TNF by the phospholipid, platelet-activating factor (PAF). These antisera, and antisera from mice immunized with phospho-threonine or galactosamine-1-phosphate conjugated to KLH, contained inhibitory antibodies of differing specificities. Mice immunized with PI-KLH, PS-KLH or phospho-threonine-KLH did not develop hypoglycaemia when challenged with the parasite toxic antigens. These results indicate that the antigenicity of non-toxic analogues can be dramatically enhanced by coupling to a protein carrier.

Tumour necrosis factor and interleukin-6 production induced by components associated with merozoite proteins of Plasmodium falciparum.

P. falciparum merozoite antigens, merozoite surface protein-1 (MSP-1) and rhoptry associated protein-1 (RAP-1), were shown to be liberated into the supernatant of in vitro parasite cultures and to be included in the endotoxin-like exoantigen complex, previously designated Ag7. Material affinity purified from culture supernatants, using immobilized monoclonal antibodies specific for RAP-1 or MSP-1, stimulated normal human mononuclear cells to produce TNF and IL-6 in vitro. However, stimulation of TNF was absent, and that of IL-6 was reduced, when the antigens were purified from detergent extracts of infected erythrocytes. These results indicate that the RAP-1 and MSP-1 proteins themselves do not stimulate the production of TNF. Instead, other components associating with these exoantigens may be responsible for the TNF production. Mouse antisera blocking TNF production stimulated by P. yoelii exoantigens also blocked TNF production stimulated by material affinity purified from P. falciparum culture supernatants using RAP-1 specific monoclonal antibody, indicating the conserved structure of the TNF inducing component.

Phospholipid-containing toxic malaria antigens induce hypoglycaemia.

Hypoglycaemia is associated with severe malaria and is an important prognostic indicator. Molecules liberated during overnight incubation of erythrocytes infected with Plasmodium yoelii induce marked hypoglycaemia in normal mice, with a delayed time course compared with insulin; some, though weaker, activity could also be obtained by overnight incubation of uninfected erythrocytes. The active component shares many properties with the phospholipid-containing molecules which we have previously shown to be toxic and to induce the release of tumour necrosis factor (TNF) from macrophages. However a MoAb which neutralizes the cytotoxicity of tumour necrosis factor in vitro did not prevent this induction of hypoglycaemia, whereas antiserum against the toxic antigens did, as did immunization of normal (but not the immunoglobulin-deficient SCID) mice with the same material. Furthermore, normal mice injected with the antigens after immunization with phosphatidyl inositol or inositol monophosphate did not develop hypoglycaemia; the latter compound was also inhibitory when mixed with the antigens before injection. These compounds were previously shown to block the induction of TNF by the antigens and to induce the production of inhibitory antibodies. The role of these molecules in the etiology of the hypoglycaemia of malaria is discussed.

Detoxified exoantigens and phosphatidylinositol derivatives inhibit tumor necrosis factor induction by malarial exoantigens.

We have previously shown that malaria parasites liberate exoantigens which, through a phospholipid component, stimulate mouse macrophages to secrete tumor necrosis factor (TNF), which are toxic to D-galactosamine-sensitized mice, and which therefore might be involved in pathology. Plasmodium yoelii exoantigens detoxified by dephosphorylation or digestion with lipases do not induce TNF production. However, these partial structures inhibited its production in response to the exoantigens, although not to bacterial lipopolysaccharide (LPS). When pure phospholipids were tested in a macrophage assay, none stimulated the production of TNF, but phosphatidylinositol (PI) inhibited TNF induction by P. yoelii exoantigens. Moreover, inositol monophosphate (IMP) was the only one of a number of monophosphate saccharides tested which was inhibitory; inositol was not. Macrophages pretreated with PI, IMP, or detoxified exoantigens and then incubated with parasite exoantigens also yielded much less TNF. PI, IMP, and lipase-digested exoantigens of P. yoelii similarly inhibited the TNF-inducing activity of exoantigens of the human parasites Plasmodium falciparum and Plasmodium vivax. Neither PI nor IMP diminished TNF production in response to LPS, in contrast to a platelet-activating factor antagonist [1-O-hexadecyl-2-acetyl- sn-glycero-3-phospho(N,N,N-trimethyl hexanolamine)] which inhibited both exoantigen- and LPS-induced production of TNF. We conclude that at least two different parts of the molecule are involved in the induction of TNF secretion by parasite exoantigens: one requires the presence of a phosphate bound to inositol, and, since dephosphorylated exoantigens were also inhibitory, one does not. It would seem that both affect interactions between parasite-derived exoantigens and the macrophage receptors.

Antibodies against phosphatidylinositol and inositol monophosphate specifically inhibit tumour necrosis factor induction by malaria exoantigens.

The active component of the exoantigens of malarial parasites which stimulates macrophages to secrete tumour necrosis factor (TNF) has been shown to depend upon a phospholipid, the activity of which was blocked by phosphatidylinositol (PI) and inositol monophosphate (IMP) in competitive inhibition studies. Antisera made against the exoantigens of Plasmodium yoelii, which inhibited their induction of TNF, were found by an ELISA assay to contain antibody against several other phospholipids. However, the inhibitory antibody was removed specifically by adsorption with liposomes containing PI, but not other phospholipids. Furthermore, PI was the only phospholipid in non-liposomal form which induced the production of inhibitory antisera. Mice immunized with IMP, but not inositol, also produced inhibitory antisera. When incorporated into liposomes several other phospholipids did give rise to inhibitory antibodies but, in contrast to the antisera against parasite exoantigens, PI and IMP, the inhibitory activity was removed by adsorption with heterologous phospholipid liposomes, suggesting that it was directed against a common determinant, presumably the phosphate ester head group. Inhibitory antibodies in the antisera tested were predominantly IgM and titres were not increased after repeated injections. Antisera raised against PI, IMP or the cross-reacting phospholipid liposomes also inhibited TNF secretion by macrophages stimulated by exoantigens of the human parasites P. falciparum and P. vivax, but not by bacterial lipopolysaccharide. These findings confirm our conclusion that exoantigens from these different species contain phosphate bound to inositol in their TNF-inducing moiety.

Serological relationship of tumor necrosis factor-inducing exoantigens of Plasmodium falciparum and Plasmodium vivax.

Exoantigens of Plasmodium vivax-parasitized erythrocytes stimulated macrophages to secrete tumor necrosis factor, and antisera raised against the exoantigens inhibited this secretion. The antisera also inhibited the activity of Plasmodium falciparum and Plasmodium yoelii exoantigens, and conversely, antisera against the latter cross-reacted with the exoantigens of P. vivax.

Tumour necrosis factor induction by malaria exoantigens depends upon phospholipid.

In patients with malaria, the clinical manifestations of the disease are associated with the presence of high concentrations of tumour necrosis factor (TNF) in the serum. Blood-stage parasites of human and rodent malarial parasites release serologically related exoantigens which induce the production of TNF in vitro and in vivo and which can kill mice made hypersensitive to TNF by pretreatment with D-galactosamine. They also elicit the production of T-independent antibody, which blocks these effects. The capacity of the exoantigens to stimulate macrophages to secrete TNF does not require the presence of protein or carbohydrate, but is associated with a lipid whose activity can be abolished by treatment with phospholipase C. Treatments of the exoantigens which destroyed their activity in vitro also abrogated their immunogenicity and their toxicity for mice. No TNF-inducing activity could be detected in preparations of parasitized erythrocytes that was not associated with phospholipid, and the TNF-inducing properties of the malarial phospholipids are quite distinct from those of bacterial lipopolysaccharide. We conclude that release of potentially toxic phospholipids by parasites may be responsible for some of the pathology of malaria.

Don't kill the parasite: control the disease.

It is clear from both laboratory and clinical studies that the blood-stage malaria parasite does not itself directly cause most of the serious complications of the disease, with the possible exception of anaemia. For example, T cell- deprived mice with lethal infections survive longer and mice can be protected against early death by vaccines that appear not to affect parasitaemia. In certain cases antibodies to TNF have the same effect. Clinically it has been known for over 50 years that children in endemic areas develop immunity to the serious toxic aspects of malaria several years before their parasitaemias start to fall. Recent work on the induction of cytokines such as tumour necrosis factor (TNF) by exoantigens of the blood-stage parasite and on the role of cytokines in this and other toxic diseases suggests that an appropriate vaccine might induce antibody that blocks the effect of the exoantigens, thus conferring on young children the anti-disease immunity that normally takes years to appear. Such vaccines might be less hampered by the antigenic variation that makes anti-parasite immunity slow to develop. Characterisation of the molecules involved is a high priority.

Two soluble antigens of Plasmodium falciparum induce tumor necrosis factor release from macrophages.

The production of cytokines such as tumor necrosis factor (TNF) may contribute to the pathology of malaria. We showed previously that crude preparations of heat-stable exoantigens from parasite cultures induce the release of TNF in vitro and in vivo. When separated from the culture medium by affinity chromatography, in which immune immunoglobulin G was used as ligand, the mixture of exoantigens of Plasmodium falciparum retained the capacity to induce the secretion of TNF, both by human monocytes from Gambian children and by mouse macrophages. Two individual antigens, Ag1 and Ag7, further purified by affinity chromatography and identified by crossed immunoelectrophoresis, also stimulated TNF production by both types of cell but differed in other functional properties. Thus, the activity of Ag7, but not that of Ag1, was inhibited by polymyxin B, and antisera made against boiled exoantigens of the rodent parasite Plasmodium yoelii which blocked the ability of these antigens to induce the production of TNF also inhibited the activity of Ag7 without affecting Ag1. Since the prevalence of antibody against Ag7 in sera from children in endemic areas appears to correlate with the development of immunity against the manifestations of the disease, this antigen may be one cause of pathology, perhaps through its ability to induce the production of TNF. Its serological relationship with rodent exoantigens suggests that it might be a candidate for an anti-disease vaccine which has the advantage that its active moiety is not subject to significant antigen polymorphism.

Malaria exoantigens induce TNF, are toxic and are blocked by T-independent antibody.

The production of cytokines, including tumour necrosis factor (TNF), may be involved in the pathology of malaria, as well as in protection against the parasite. We have shown that parasite exoantigens induce the secretion of TNF in vitro and in vivo and kill mice made hypersensitive to TNF. They elicit T-independent antibody that inhibits their capacity to stimulate TNF production and protects against toxicity in vivo, and those of human and rodent parasites are serologically related. Their active component does not appear to be protein. Here we review their properties and consider the epidemiological significance of our findings and their possible contribution to the development of an "anti-disease" vaccine.

Malaria exoantigens induce T-independent antibody that blocks their ability to induce TNF.

Much of the pathology of malaria may be due to the interactions of cytokines, especially tumour necrosis factor (TNF), with various cell types, including endothelial cells, with consequent widespread systemic effects. It has been shown previously that heat-stable exoantigens in the supernatants of blood-stage parasite cultures induced the release of TNF in vitro from activated macrophages and behaved like toxins in vivo, that mice immunized with the antigens are protected from the toxic effect and that their serum specifically blocks the ability of the antigens to stimulate the production of TNF. It is reported here that the inhibitory antibody is mainly IgM and that it appears to be T-independent, as the titres of antisera from T-deficient and immunologically intact mice were similar. Antisera raised against exoantigens from two species of rodent parasite inhibited TNF production by those of the human parasite Plasmodium falciparum, and vice versa, indicating that the TNF-inducing moieties of the exoantigens cross-react and therefore presumably contain common epitopes. Thus vaccination with these exoantigens might provide a means of protection against the clinical effects of malaria and of generating anti-disease immunity by reducing cytokine production. However, these findings imply that it will be necessary to confer on these antigens the ability to stimulate T cells and generate memory before they can provide a useful basis for an anti-disease vaccine. The results obtained are discussed in relation to some of the epidemiological features of malaria.

The malaria vaccine: anti-parasite or anti-disease?

There are three major difficulties hindering the development of a vaccine for malaria. First, for all three stages of the parasite life cycle there is an incomplete understanding of the precise type of immune response to aim for. Second, only a handful of the many hundreds of parasite-derived antigens have been explored, and though several have been shown to be protective in animal models, it is not known if they are the most potent. Third, there is strong evidence that the parasite can evade host immunity, for example by antigenic variation. In this brief article, John Playfair and his colleagues address mainly the first issue and suggest that complete resistance to infection is probably not feasible, and that attention should be directed not so much at vaccines designed to eliminate one or other stage of the parasite, but rather towards the possibility of an 'antitoxic' vaccine that prevents the serious pathological complications of the disease.

Human and murine macrophages produce TNF in response to soluble antigens of Plasmodium falciparum.

Heat-stable antigens of rodent malarial parasites induce the release of tumour necrosis factor (TNF) from mouse macrophages, in vitro and in vivo. We report here that analogous antigens of Plasmodium falciparum trigger the release of TNF from human monocytes in vitro, in conditions that exclude the effects of any contaminating endotoxin. These antigens also induced TNF release from a murine monocytic cell line and from the peritoneal macrophages of several strains of mice, including the LPS-hyporesponsive C3H/HeJ mice. Similarly, boiled soluble antigens from the rodent parasites P. yoelii and P. berghei also stimulated human monocytes. Antisera made by immunizing mice with boiled antigens of P. falciparum or P. yoelii inhibited the stimulation of TNF secretion by P. falciparum antigens. They did not block the induction of TNF by bacterial lipopolysaccharide. Thus mouse macrophages provide a convenient system for investigating the nature, cross-reactions and antigenic variation of human malarial soluble antigens. Since these are known to occur in the circulation of patients with malaria, they may be responsible for excess production of TNF, a mediator that is thought to play a central role in the pathogenesis of the disease.