Cerebral Malaria Pathogenesis

Revisiting Parasite and Host Contributions

Georges Emile Raymond Grau; Alister Gordon Craig


Future Microbiol. 2012;7(2):291-302. 

In This Article

Parasite Contributions

PfEMP1 & Cytoadherence

When Pf merozoites invade red blood cells they remodel the host cell extensively through changes to the cytoskeleton and insertion of parasite-derived proteins into and onto the erythrocyte membrane (for review see [10]). One such protein, thought to be a major virulence factor in Pf, is the major variant surface antigen PfEMP1. Owing to its expression on the surface of the IE, this protein must undergo antigenic variation to escape host defense mechanisms and a complex mechanism (i.e., antigenic variation) exists to support this.[11] This presentation of variant PfEMP1 proteins to the host results in differences in the host immune responses they provoke and in the repertoire of host receptors that erythrocytes infected with different variants can bind to (as well as the affinity of this binding).[12]

PfEMP1 is encoded by a family of var genes that can be characterized by sequence motifs at the 5' promoter region and within the first adhesive (DBL) domain. Analysis of these sequence tags has identified associations of specific subsets of var genes with severe disease,[13–17] supporting the idea that PfEMP1 is a virulence factor. For example, Warimwe et al. showed a positive correlation between Group A-like/cys2 (or Ups-A) var gene expression and impaired consciousness, reinforcing a number of studies that had suggested associations between Ups-A var gene expression in the invading isolate and severe malaria.[17] This sequence variant type is also selected from parasite populations using sera from semi-immune children,[18,19] the same category that shows greater susceptibility to CM. The relationship is not exact and many Ups-A/cys2 type infections do not lead to CM or even other aspects of severe malaria, which probably reflects the requirement of particular host–pathogen combinations to drive specific disease patterns. However, in considering potential parasite pathogenic processes, the ability of IE to undergo a range of adhesive interactions with host cells has been a major area of study.

Binding interactions between IE and host cells can be divided into three main categories:

  • Cytoadherence (interaction with endothelial cells [ECs] or, in placental malaria, with the syncytiotrophoblast cells of the placenta);

  • Rosetting (interaction with uninfected erythrocytes);

  • Clumping (interaction with other IE via platelet bridges).

The molecular basis for these interactions has largely been described (for review see [12]), although there is clearly scope for the identification of further host receptors. The picture emerging is a complex one with over 12 host receptors available to the parasite through its repertoire of PfEMP1 protein variants, used in different permutations and with varying affinities. Several studies have attempted to correlate specific binding activity with clinical outcomes through testing binding phenotypes of patient isolates. Rosetting and clumping have both been variably associated with severe malaria[20–23] but results for cytoadherence to endothelium have been variable, with only one recent study showing a statistically significant association between higher binding to ICAM-1 under flow assay conditions and CM.[24] These binding experiments also do not explain how the sequestration of IE in the brain leads to CM, and an explanation of this awaits further work on the processes leading to differential cerebral IE sequestration and the effect of these IE on the host. This is further complicated by findings that the rigidity of IE itself, most likely caused through the parasite-derived modifications to the erythrocyte, is also associated with severe disease.[25,26]

How might IE sequestration lead to pathology? Earlier considerations of this issue divided into two main hypotheses: mechanical obstruction of vessels and inflammation. More recently, the boundaries between these categories have become blurred and it seems most likely that combinations of both of these processes are responsible for disease, and support for mechanical blockage being a significant factor has come recently from elegant observations of the mucosal microvasculature performed by Dondorp et al..[27] What has become clear is that the binding interaction between IE and host endothelium is not a passive process and the engagement and adjacency for these cell types results in the activation of a range of activities ascribed to the binding interaction itself, biomechanical stimuli and the localized release of soluble or membrane-bound mediators.[28–31] Two major categories of endothelial 'damage' have been reported, namely a breakdown of the barrier function of the specialized blood–brain barrier (BBB) and apoptosis of ECs. BBB integrity has been shown in postmortem studies to be subtly disturbed during severe malaria, resulting in some leakage from cerebral vessels[32] and further studies in vitro have supported the role of IE as well as host effector cells in this process.[33–35] Much of the latter work has used measurement of transendothelial electrical resistance in EC/IE coculture models, identifying transcriptional changes caused through this interaction[33] and, in some cases, alterations to the junctional proteins (e.g., ZO-1 and occludin) at the EC–EC interface.[36,37] In terms of apoptosis, patient isolates have been shown to have variable apoptosis-inducing abilities, with a weak relationship between an increased induction of EC apoptosis by isolates taken from children with neurological manifestations.[38]In vitro work also supports the hypothesis that vascular damage may be caused by apoptotic destruction of the ECs, particularly in combination with platelets or microparticles.[39–41] It is possible that a balance exists between the triggering of apoptosis by the engagement of host receptors during cytoadherence and IE–EC apposition, and the ability of the IE to modulate this response,[30] with some of the highly localized vascular damage seen on postmortem of people dying of CM in the form of micro-hemorrhages potentially being caused through a failure to control EC apoptosis by the parasite.

Cytoadherence-related Signaling Pathways

Further evidence that the IE/EC interaction is not a passive one comes from various reports demonstrating the activation of signaling pathways. Receptor clustering during adhesive events is a common pathway for activation of signaling cascades and cytoadherence is able to mimic these events. IE binding via CD36 activates a Src-family kinase-dependent MAP kinase pathway and this Src-family kinase activity is required for efficient cytoadherence such that treatment of EC with specific inhibitors was able to reduce IE binding.[42] The same group also showed that ectophosphorylation of CD36 itself, through a Src-family kinase/alkaline phosphatase pathway, also controlled adhesion to this receptor.[43] This information has subsequently been used to support clinical intervention trials with the alkaline phosphatase inhibitor levamisole, which are currently underway.[44] Cerebral endothelium has low levels of CD36 that would not normally support IE adhesion so its relevance to CM has been questioned. In addition, higher levels of binding to CD36 have not been associated with CM, either being linked to non-CM severe malaria in one study[45] and to uncomplicated malaria in another.[24] However, studies have shown that platelets and microparticles are able to act as a bridge to support adhesion to cerebral EC[46,47] and even to transfer CD36 to these cells.[34] Thus it is possible that CD36 may sometimes be available for adhesion in the brain, or alternatively supports efficient exponential parasite multiplication during the asexual growth phase indirectly contributing to subsequent pathology. ICAM-1 is available for IE cytoadherence in the brain in the context of the widespread endothelial activation seen in CM and significant colocalization of this receptor (as well as CD36 and E-selectin) has been seen with IE sequestration in adults.[48] IE are also able to activate host signaling by binding to ICAM-1, via the MAP kinases ERK1/2, JNK and p38.[49] The nuclear transcription factor NF-κB is also implicated in cerebral cytoadherence, as demonstrated by transcriptional analysis of IE/brain EC coculture showing the differential expression of pathways controlled by NF-κB.[31] The full impact of these signaling changes on CM is not known but experiments have implicated a number of pathogenic pathways such as apoptosis, BBB integrity and increased receptor expression.

Systemic & Localized Inflammation

As well as interactions with endothelium, IE are also able to release mediators directly causing host effector cells (and ECs) to release a range of pro- and anti-inflammatory cytokines. Pf glycosylphosphatidylinositol is released during schizont rupture and is able to stimulate macrophages to release TNF and IL-1,[50] contributing to the systemic proinflammatory response seen during malaria infection. Pf glycosylphosphatidylinositol has been shown to increase expression of ICAM-1, potentially increasing the recruitment of IE to the cerebral vasculature,[51] as well as inducing nitric oxide (NO) production,[52] although the link between NO and CM pathology is less clear. It is also able to induce limited apoptosis in some tissues,[53,54] although these have not included brain.

The other major parasite component capable of directly stimulating the host immune system is the malaria pigment known as hemozoin (or hematin). Hemozoin is produced during the digestion of hemoglobin in the parasite digestive vacuole in order to detoxify the heme moiety and mediates a broad range of inflammatory and immunomodulatory activities (for review see [55]). The characteristic brown crystals of hemozoin are seen in a number of different phagocytic cells and appear to be able to persist despite the degradation pathways available within these cells. Hemozoin-containing monocytes and neutrophils have been found in greater numbers in adults dying of severe malaria compared with survivors[56] as well as children with CM.[9,57] Research on the role of hemozoin in inflammation has been complicated by the heterogeneity of experimental conditions, not the least being the source of hemozoin, either synthesized or natural. However it seems clear that it is able to induce the production of proinflammatory cytokines from a range of host cells contributing to the immunopathology of malaria, including CM. Hemozoin is also thought to be able to influence the adaptive immunity,[58,59] altering the ability of the host to make an effective response to infection. It is also involved in the inhibition of erythropoiesis and thereby to malarial anemia.[60] It has recently been suggested that the host fibrinogen bound to hemozoin is principally responsible for at least some of the proinflammatory stimulation seen.[61] A direct link with CM is harder to discern. The systemic proinflammatory response seen with hemozoin would directly support cytoadherence-mediated pathology and some evidence exists to support activation of matrix metalloproteinases (e.g., MMP-9[62]) that could cause vascular damage and induce morphological changes to endothelium.[63] The modulation of host protective responses could also have an indirect action via parasite multiplication and retention of parasite material in cerebral vessels.

Endothelial activation appears to be strongly associated with CM[48,64] and the basis for this is likely to be multifactorial. Von Willebrand factor (vWF) is raised in CM[65,66] at the same time that ADAMTS13 activity, which is responsible for degrading the bioactive ultra-large vWF multimers expressed on the EC surface, is reduced.[67–69] Part of the mechanism responsible for this association may be the cytoadherence of IE to platelets decorating the ultra-large vWF expressed on the EC,[47] including brain tissues. Release of ultra-large vWF requires activation and mobilization of the Weibel-Palade bodies (for review see [70]), which can occur through several mediators including histamine. The malaria parasite Pf produces a protein, PfTCTP, which is a homolog for the mammalian histamine releasing factor and can induce histamine release from basophils.[71]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: