Recent advances in understanding the mechanism of hemozoin (malaria pigment) formation

Available online at www.sciencedirect.com JOURNAL OF

Inorganic Biochemistry Journal of Inorganic Biochemistry 102 (2008) 1288–1299 www.elsevier.com/locate/jinorgbio

Recent advances in understanding the mechanism of hemozoin (malaria pigment) formation Timothy J. Egan * Department of Chemistry, University of Cape Town, Private Bag, Rondebosch 7701, South Africa Received 7 September 2007; received in revised form 19 October 2007; accepted 31 October 2007 Available online 23 December 2007

Abstract The recent literature on hemozoin/b-hematin formation is reviewed, with an emphasis on the mechanism of its formation. Recent findings from unrelated organisms that produce hemozoin, namely the malaria parasite Plasmodium falciparum, the worm Schistosoma mansoni and the kissing bug Rhodnius prolixus all of which consume human hemoglobin show that the formation of this crystalline substance occurs within or at the surface of lipids. Biomimetic experimental models of the lipid–water interface as well as computational studies indicate that these lipid environments are probably extraordinarily efficient at producing hemozoin. A rethink is now needed, with a new emphasis on Fe(III)PPIX in non-aqueous environments that mimic lipids and indeed within the lipid environment itself. These findings are explored and discussed in the context of earlier studies on b-hematin formation. Ó 2008 Elsevier Inc. All rights reserved. Keywords: Hemozoin; b-Hematin; Malaria pigment; Schistosoma pigment

1. Introduction When examined under a microscope, a striking feature of the trophozoite stage of a malaria parasite is the presence of a dark brown-black substance known as malaria pigment or hemozoin. This substance is found within a lysosome-like compartment known as a food vacuole (or digestive vacuole). The pigment is released into the blood of the host each time the parasite completes a blood cycle and eventually deposits in internal organs, causing noticeable discoloration. This phenomenon was first reported by Giovanni Maria Lancisi in a book published in 1717 [1], predating the discovery of the malaria parasite by more than 150 years. Later hemozoin played an important part in elucidating the role of mosquitoes as vectors of the parasite, acting as a visible tracer [2]. In these early studies, the pigment was assumed to be melanin, but in 1911 Wade H. Brown demonstrated that the chromophore is heme [3]. It *

Tel.: +27 21 650 2528; fax: +27 21 689 7499. E-mail address: [email protected]

0162-0134/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2007.12.004

was to be almost 80 years before Fitch and Kanjananggulpan demonstrated that hemozoin consists solely of ferriprotoporphyrin IX (Fe(III)PPIX) [4], probably identical to b-hematin, an insoluble Fe(III)PPIX precipitate first described in the 1930s [5]. Subsequently Slater et al. demonstrated by X-ray diffraction, infrared spectroscopy and solubilization studies that hemozoin is indeed identical to b-hematin and showed by EXAFS spectroscopy that bhematin contains bonds between the propionate group of one iron porphyrin and the Fe(III) center of its neighbor [6]. Any lingering doubts that hemozoin and b-hematin are identical were removed in 1997 when Bohle et al. demonstrated by high resolution synchrotron X-ray diffraction that lyophilized parasitized erythrocytes give an identical diffraction pattern to b-hematin [7]. Finally, in 2000 Pagola et al. determined the structure of b-hematin from the X-ray powder diffraction pattern (Fig. 1) [8]. Hemozoin is produced as an end product of heme released during the digestion of host hemoglobin by the malaria parasite and is believed to be a detoxification pathway in the parasite. Fe(III)PPIX produced by autoxidation

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

1289

Fig. 1. The structure of hemozoin/b-hematin. Dotted lines represent hydrogen bonds.

of heme released from hemoglobin is known to be capable of causing lipid peroxidation [9] and to destabilize membranes through a colloid osmotic mechanism [10]. Packaging Fe(III)PPIX into compact and highly insoluble hemozoin crystals decreases its pro-oxidant capacity [11] and likely also avoids colloid osmotic effects. If the parasite were to degrade Fe(III)PPIX, as mammals do using the enzyme heme oxygenase, it would be faced with sequestering the vast quantity of iron released since free Fe(III) is also highly toxic. It is handled by specialized transport and binding proteins in higher organisms [12]. Nonetheless, there has been uncertainty about whether hemozoin formation is the main fate of heme in the parasite Plasmodium falciparum [13,14]. In two studies a large discrepancy was found between the total iron content of parasitized erythrocytes on the one hand and the sum of undigested hemoglobin and hemozoin on the other. Zhang et al. claimed that about 70% of the Fe(III)PPIX is degraded in a glutathione dependent process outside of the parasite food vacuole where hemozoin is found [13]. Loria et al. claimed a similar degree of Fe(III)PPIX degradation occurs within the food vacuole [14]. In both of these studies hemozoin was quantified by dissolution in 0.1 M NaOH and measurement of the Soret band of Fe(III)PPIX in basic aqueous medium. However, when in a later study the iron content of P. falciparum parasitized erythrocytes, isolated parasites, food vacuoles and hemozoin were each separately measured colorimetrically by release of total iron followed by coordination of Fe(III) with ferrozine to form a colored complex, it was found that 70–100% (95% CI) of the parasite iron is found in the hemozoin [15]. This conclusion was unequivocally supported by Mo¨ssbauer spectra of freeze-dried intact parasites in which the only detectible iron signal was shown to be that of hemozoin and the magnitude of background scatter limits other iron species to no more than 5% of the total iron content [15]. Elemental mapping of transmission electron micrographs by electron energy loss spectroscopy further supported the conclusion that hemozoin formation is the overwhelming fate of heme released in the parasite [15]. Collectively, these techniques demonstrate that at least 95% of the heme released in the parasite is converted to hemozoin. Recently Gligorijevic et al. have studied hemozoin formation in single live intra-

erythrocytic parasites using spinning disk confocal microscopy [16]. The quantity of hemozoin determined in this work from the volume of crystals observed (15 lmol/ 1010 cells) corresponds to 88% of the heme present in the erythrocyte (17 lmol/1010 cells), confirming that heme must be almost entirely converted to hemozoin. Over the last decade, hemozoin has been discovered in a number of other blood-feeding organisms, including the insect Rhodnius prolixus (the kissing bug) [17], helminth worms Schistosoma mansoni (the causative agent of schistosomiasis or bilharzia) [18,19] and Echinostoma trivolvis [20] and the bird-infecting protozoan Haemoproteus columbae [19]. These discoveries have both broadened the interest in hemozoin and aided in understanding the mechanism of hemozoin formation. Apart from being interesting in its own right, hemozoin formation is also important as a target of antimalarials such as chloroquine. In 1992, Slater and Cerami reported that 4-aminoquinoline and quinoline methanol antimalarials inhibit parasite extract induced b-hematin formation and suggested that they inhibit an enzyme responsible for hemozoin formation (then thought to be a polymer) [21]. Later it was shown that these drugs inhibit b-hematin formation brought about under abiotic conditions and that inhibition thus occurs by direct interaction between the drugs and Fe(III)PPIX [22,23]. Later Mungthin et al. demonstrated that these drugs depend on the release of heme from hemoglobin for their activity, since inhibitors of hemoglobin digesting proteases antagonize their activity [24]. Sullivan et al. also provided direct evidence of interaction of chloroquine with hemozoin in the food vacuole of P. falciparum by electron micrographic autoradiography [25]. The interaction of antimalarials with Fe(III)PPIX has been reviewed previously [26] and several reviews are available on the role of hemozoin inhibition in the action of antimalarials, so this will not be discussed further here [27–31]. However, it is worth noting that recent studies have also indicated that chloroquine inhibits hemozoin formation in S. mansoni, reducing parasite burden as well as egg deposition in mice [32], suggesting that hemozoin inhibition may also be a viable strategy for treatment of schistosomiasis. Recently hemozoin has also been garnering interest because of its possible role in the immunological

1290

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

response of the host to malarial infection and for its potential diagnostic utility. Readers are directed to a recent review on these applied and medical aspects [33]. With the elucidation of the chemical composition and structure of hemozoin, the major outstanding question has been the mechanism of its formation. In a review on hemozoin published five years ago it was pointed out that the formation of hemozoin in vivo was poorly understood and that the details of its formation remained to be determined [34]. The last three years has seen major strides, both in describing the biological environment in which hemozoin formation occurs and in elucidating the molecular details of the process. This will be the main focus of this review. 2. Abiotic syntheses of b-hematin The synthetic counterpart of hemozoin, b-hematin, can be prepared in several ways. Fundamentally, however, these can be divided into two alternative procedures: 2H2 O–Fe(III)PPIX ! (Fe(III)PPIX)2 + 2H2 O

ð1Þ

or 2Cl–Fe(III)PPIX + B ! (Fe(III)PPIX)2 + BHþ Cl

ð2Þ

where B is the organic base 2,6-lutidine. Process (1) is sometimes referred to as the acid-catalyzed method [35] and proceeds in aqueous media under acidic conditions, often at elevated temperatures. This was the method originally used by Slater et al. who incubated a solution of hemin (Cl–Fe(III)PPIX) dissolved in 0.1 M NaOH (which likely produces HO–Fe(III)PPIX) in 4.5 M acetic acid at 70 °C for 18 h [6]. Later it was shown that the reaction proceeds at 60 °C in 30 min in 4.5 M acetate at pH 4.5 [22]. This process has recently been used in a 96-well plate assay for measuring inhibition of b-hematin formation [36]. The reaction also proceeds at lower temperature and has been used in similar assays performed at 37 °C, typically over 16 h [37–39]. Acids other than acetic acid, including propionic and benzoic acid promote bhematin formation as well [6,40]. Variations on this aqueous method have also been reported in which detergents or alcohols are used to promote b-hematin formation [41–43]. In the presence of alcohols there is evidence that carboxylic acids are not essential and sulfuric or hydrochloric acid can be used to acidify the medium [41]. Process (2) occurs under strictly anhydrous conditions in methanol at room temperature. It was originally developed by Bohle and Helms [44] and has subsequently been used mainly by Bohle and co-workers to study spectroscopic, structural and physical characteristics of b-hematin [7,8,35,45–47], but has also been used to prepare b-hematin to investigate its immunological effects [48]. Over the years there have been some disputes about the nature of products obtained under various conditions. Suggestions that method (1) produces Fe(III)PPIX–acetate complex [49] have been disproved by a combination of ele-

mental analysis and X-ray diffraction studies [50]. Similarly, suggestions that b-hematin is only formed during the drying process used for infrared spectroscopy [51] is discounted by X-ray diffraction and infrared spectra showing that it is present in the undried product [52]. Nonetheless, it has been demonstrated by synchrotron X-ray diffraction that crystalline phases other than b-hematin can be present in the products and that reaction at high temperature produces much smaller and less regular crystals compared to the anhydrous method (Fig. 2) [35]. Some of these non-crystalline or poorly crystalline phases probably account for observation of a strange product that has been referred to as B-hematin, which seems to be chemically identical to b-hematin, but lacks the degree of crystallinity of the latter [41]. An intriguing possibility is that this substance may include non-centrosymmetric dimers of Fe(III)PPIX which have been discussed as potentially responsible for limiting crystal growth of b-hematin by Buller et al. [53]. Different hydration states of b-hematin are also possible as water can be absorbed up to a limit of about 14% by mass by dry b-hematin. X-ray diffraction indicates improved crystallinity in the hydrated crystal relative to the dehydrated product. Such disputes highlight the importance of using not only infrared, but also X-ray diffraction and preferably scanning electron microscopy (SEM) or transmission electron microscopy (TEM) in the characterization of these products. A recent publication has suggested, based on EXAFS data, that hemozoin exhibits a Fe(III)–O bond length ˚ longer than b-hematin [54], a 1.5% difference 0.029 A ˚ , respectively). This difference is smaller (1.878 vs. 1.849 A than the statistical error in the Fe(III)–O bond length in ˚ ), which the authors attribute to disorhemozoin (0.098 A der in this part of the structure in hemozoin. Conversely, they reported increased disorder in the fourth shell of atoms around the Fe center in b-hematin. It is difficult to assess the importance of these apparent differences as it is unclear what sort of inter-sample differences are likely to

Fig. 2. Comparison between b-hematin crystals produced by (a) the acidcatalyzed thermal method and (b) the anhydrous method as reported by Bohle and co-workers [35]. Reproduced from: D.S. Bohle, A.D. Kosar, P.W. Stephens, Phase homogeneity and crystal morphology of the malaria pigment b-hematin, Acta Cryst. D 58 (2002) 1752–1756; with permission of the International Union of Crystallography, .

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

be seen either in b-hematin or hemozoin, or indeed what the effect would be of different methods of preparing bhematin, as only process (2) was used in this study. However, it is to be noted that another recent publication reported the Fe–O bond length from solution of the X˚, ray powder diffraction pattern of b-hematin to be 1.85 A while hemozoin from R. prolixus was in this case shorter ˚ ) [55]. As described above, changes in the hydration (1.82 A state of b-hematin are known to have subtle effects on both the structure and disorder of the crystal lattice [46] and may also play a role in the observed differences. It would therefore seem premature to conclude that these subtle discrepancies represent any systematic distinction in structure between b-hematin and hemozoin. 3. Role of proteins and lipids in hemozoin formation in vivo When Slater et al. found that b-hematin contains bonds between the propionate group of one Fe(III)PPIX and the Fe(III) center of another, they assumed that this substance is a polymer consisting of infinite chains of the form (Fe(III)PPIX)n [6]. In 1992 Slater and Cerami showed that parasite extracts promote b-hematin formation and that this apparent catalytic effect is seemingly abolished by boiling. Furthermore, the process seems to occur optimally at about pH 5.5. This led them to conclude that an enzyme, dubbed heme polymerase is responsible for forming hemozoin in the parasite [21]. It is noteworthy, however, that the bell-shaped pH dependence of b-hematin formation can be expected on the basis of the fact that the product requires one of the porphyrin propionates to be protonated and the other deprotonated and thus will form optimally within a certain pH range [52]. Given the pKa of carboxylic acids, an optimum around 5 is not unexpected. Later Dorn et al. confirmed that parasite extracts promote b-hematin formation, but disputed the claim that heating destroys this capacity and also showed that the effect is resistant to proteases. Rather, they found that hemozoin present in the parasites promotes b-hematin formation in an apparent autocatalytic process [23]. This was almost immediately disputed by Bendrat et al. who suggested that lipids are in fact responsible for b-hematin formation in these extracts and that even when preformed b-hematin is used as autocatalyst, it is in fact contaminating lipids that are responsible for the catalytic effect [56]. It has subsequently been unequivocally demonstrated that hemozoin/b-hematin does indeed promote b-hematin formation [57] and the process has even been used as the basis for high throughput screening assays for hemozoin inhibitors [58,59]. Indeed, this is exactly what is expected for a crystal growth process and growth onto pre-existing crystals with addition of Fe(III)PPIX has now been directly demonstrated by SEM [60]. In addition, it has also been demonstrated that lipids do indeed promote b-hematin formation. Dorn et al. showed that lipids extracted with acetonitrile from trophozoites and even extracts of uninfected red cells support b-hematin formation [57]. The role

1291

of lipids was further explored by Fitch and co-workers who showed that 70% of the b-hematin promoting activity of parasite extracts could be recovered in chloroform, with no activity remaining in the residue. These authors confirmed that uninfected red blood cell ghost extracts can promote b-hematin formation, but the ghosts themselves seem to contain no such capacity. It was reasonably suggested that non-lipid components of red blood cell membranes render them incapable of supporting b-hematin formation. Various unsaturated fatty acids, mono- and dioleoylglycerol and a series of detergents were also shown to promote the process, while the lipids trioleoylglycerol and cholesterol among others and saturated fatty acids palmitic and stearic acid were reported to be inactive. In this study, dried lipids were suspended in buffered reaction medium to which hematin (HO–Fe(III)PPIX) was introduced. In some cases reactions proceeded at hitherto unprecedented rates, with half-lives approaching several hours at 37 °C [61]. We now know that hemozoin is a crystalline dimer of Fe(III)PPIX with a well defined crystal habit (i.e. shape) which can be described as lath-like [8,53,62]. These crystals are typically close to a micron in length and several 100 nm wide making them almost macroscopic objects. While an enzymatic process was a perfectly reasonable possibility for a polymerization process as originally conceived by Slater et al. [6,21], it is difficult if not impossible to conceive how an enzyme could bring about crystal formation. In the former process, the enzyme would catalyze the formation of the links between the monomers, with the resultant polymer eventually precipitating to produce the macroscopic structure. In the latter process, the product is the macroscopic crystal which is hundreds of times larger than a protein. This comment does not, however, rule out proteins from the process of hemozoin formation, but rather changes the focus of the role of such a protein. Instead of being a catalyst, the protein would have to act as a nucleation site for crystal growth. This is common in biomineralization processes [63]. In 1996 Sullivan et al. showed that histidine rich protein II (HRP II) can promote b-hematin formation, and suggested that it is responsible for hemozoin formation in the malaria parasite [64]. Later Ziegler et al. showed that a dendrimer containing the peptide repeat unit of HRP II (His-His-AlaHis-His-Ala-Ala-Asp) also supports b-hematin formation and suggested that this sequence represents a biomineralizing template [65]. The consequence of these many views on the biosynthesis of hemozoin formation in malaria parasites is that there has been no consensus on the mechanism of hemozoin formation in vivo. It is fairly clear, however, that autocatalysis can only play a role once the process had been initiated and that initiation could be by lipids, proteins (possibly HRP II), or both. Pandey et al. noted that neither the lipid supported process, nor HRP II were fast enough to account for hemozoin formation in vivo and proposed that both are involved in a two-step process [66].

1292

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

4. Thermodynamic and kinetic considerations Before describing the latest findings on hemozoin formation, it is worth digressing to discuss the thermodynamics of b-hematin formation as well as kinetic studies that have been conducted under abiotic conditions as these are informative in understanding the biological process. The thermodynamics of b-hematin remain unexplored. However, this does not mean that inferences cannot be drawn from available data. For example, less soluble solids are thermodynamically more stable than their more soluble counterparts and amorphous solids are less stable (and more soluble) than crystalline materials [63]. On the other hand, amorphous solids are precipitated faster than crystalline phases and often convert to the most stable solid in a stepwise fashion. Thus, for example, in the formation of the bone mineral hydroxyapatite (Ca10(PO4)6(OH)2) in vitro the process occurs through rapid precipitation of amorphous calcium phosphate, followed by slow conversion first to less soluble crystalline octacalcium phosphate (Ca8(HPO4)2(PO4)4  H2O) at higher pH or dicalcium phosphate dihydrate (Ca2(HPO4)2  2H2O) at low pH and then to the least soluble hydroxyapatite. This is a generally observed sequential process seen in the formation of crystalline solids referred to as the ‘‘Ostwald rule” [63]. The solubility product of b-hematin and Fe(III)PPIX are unknown. However, available evidence suggests that bhematin is highly insoluble [6], while Fe(III)PPIX (here probably H2O–Fe(III)PPIX) is slightly soluble at pH 5 [22]. Certainly, X-ray diffraction data show that the former is crystalline [6–8,50,52], while the latter appears to be amorphous [52]. Thus a schematic phase diagram can be drawn for b-hematin and Fe(III)PPIX in an aqueous milieu (Fig. 3). At sufficiently low pH, protons will compete with Fe(III) for the bound propionate group, destabilizing bhematin. At pH values much above the pKa of the second propionic acid group of Fe(III)PPIX, b-hematin is also unstable as deprotonation will remove the hydrogen atoms

Fig. 3. A schematic representation of the possible phase diagram for hemozoin/b-hematin (Hz) and Fe(III)PPIX (H). Here Fe(III)PPIX is probably H2O–Fe(III)PPIX.

required for hydrogen bonding between the dimers in the crystal. Thus b-hematin is stable only within a certain acidic pH range. Fe(III)PPIX similarly precipitates when one propionate is unprotonated and the other is protonated giving a neutral complex. Thus, this will have similar pH dependence to b-hematin. However, the pKa of the first propionic acid group will be higher than that of b-hematin because binding to Fe(III) can be expected to depress the pKa in b-hematin. Similarly, the pKa of the second propionic acid group is likely to be lower than that of b-hematin because hydrogen bonding will stabilize the protonated form in the latter case. This means that neutral solid H2O–Fe(III)PPIX is probably thermodynamically metastable with respect to b-hematin formation throughout the pH range of its existence. The importance of this point is that it ought to be possible to produce b-hematin directly from solution under appropriate conditions without precipitating Fe(III)PPIX. As indicated above, there is no direct evidence for significant proportions of Fe(III)PPIX (either solid or in solution) being present in malaria parasites and it therefore seems likely that hemozoin is indeed directly nucleated from solution. The first attempt to follow the time course of b-hematin formation used Mo¨ssbauer spectroscopy to monitor the process. Discreet samples were collected during formation of the product at 60 °C in 4.5 M sodium acetate at pH 4.5. The results seemed to indicate a zero-order process continuing at a constant rate until completion [67]. However, follow up studies using infrared spectroscopy showed that the reaction in fact follows sigmoidal kinetics under the same conditions [52]. The relatively large errors inherent in attempting to determine relative quantities of starting material and product by Mo¨ssbauer spectroscopy probably obscured the true profile of the reaction. Although the use of the 1210 cm1 infrared peak to monitor the kinetics has been criticized on the grounds that bhematin exhibits the same peak, the results of the infrared study were consistent with X-ray diffraction measurements and with changes observed by SEM [52]. Subsequent measurements using differential solubilization in 5% pyridine solution have also recently produced identical kinetics for this system [40] and there was no evidence of B-hematin formation in either study. Sigmoidal kinetics are typical of processes involving nucleation and growth phases and the data were modeled using the Avrami equation, an equation used to describe such processes. Under these conditions, Fe(III)PPIX precipitates at the beginning of the process. Little apparent change occurs for some time and then once sufficient nucleation has taken place growth of b-hematin begins to dominate and rapid conversion occurs. Stirring rate was found to influence the geometry of growth from nucleation sites and the process was found to depend strongly on acetate concentration, temperature and pH. The starting material was found to be amorphous or at most contain nano-scale crystallites. It was proposed that formation of b-hematin involves rapid precipitation of amorphous Fe(III)PPIX, followed by slow conversion to

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

the final product by dissolution and re-precipitation with acetate aiding in solubilizing the initial precipitate. This is reminiscent of hydroxyapatite formation and it was proposed that hemozoin formation is indeed a form of biomineralization [52]. This interpretation of how the process proceeds has recently been supported by Huy et al. who have studied b-hematin formation in the presence of methanol, ethanol, n-propanol and n-butanol [43], showing that these alcohols induce b-hematin formation with increasing activity in that order. This is also the order in which they both solubilize and monomerize Fe(III)PPIX as well as the order in which they reduce the surface tension of water. The latter is important for nucleation as a decrease in interfacial tension reduces the energy barrier for critical nucleus formation. A recent study on b-hematin formation in the presence of benzoic acids and substituted benzoic acids under conditions otherwise the same as those used in the study performed in acetate medium has revealed that these acids are far more active in supporting b-hematin formation than acetic acid [40]. For example, b-hematin forms within 2 h in the presence of 0.05 M benzoic acid at pH 4.5 and 60 °C, compared with a similar reaction time under the same conditions in the presence of 3.5 M acetate [52]. In the case of benzoic acids, however, there is no evidence that the acid is able to solubilize the initial precipitate, the process is not influenced by stirring speed, except when stirring is very vigorous (which appears to break up the nuclei and causes a deviation from Avrami behavior) and studies with substituted benzoic acids indicate increased efficacy when electron withdrawing groups are attached to the aromatic ring. The last observation was interpreted as evidence that p-stacking of the aromatic acid over the Fe(III)PPIX helps to break up Fe(III)PPIX interactions and promote rearrangement to b-hematin. Suggestions that the acids also help to disrupt hydrogen bonds seems to be discounted by the observations of Huy et al. on alcohols, as the more hydrophilic methanol would be expected to disrupt hydrogen bonding more than n-butanol, but is in fact least active in promoting b-hematin formation [43]. Very few studies on b-hematin formation in the presence of biological materials have made an attempt to obtain time courses for the process. This is a pity, because such data are far more informative than overall yields of product at fixed times which are often used to compare effectiveness of different substances in producing b-hematin. Nonetheless, from the available evidence at least autocatalytic b-hematin formation [59] has been shown to exhibit sigmoidal kinetics. The study by Fitch et al. on lipid promoted b-hematin formation does not contain enough points at short times to be sure whether sigmoidal kinetics are obeyed [61]. Comparable data appears to be lacking for HRP II. Interestingly, the recent study on hemozoin formation in live single cells reported by Gligorijevic et al. using spinning disk confocal microscopy has shown an overall sigmoidal curve for the growth of hemozoin in the parasite

1293

[16]. However, it seems doubtful that this is reflective of the mechanism of hemozoin formation itself, as similar sigmoidal growth is observed for the food vacuole volume as well [68]. Thus, this sigmoidal profile is probably representative of the parasite maturation process in this case. The most notable feature of the kinetics of b-hematin formation in studies conducted to date is that the factor which appears to dominate the process is the ability to solubilize or disrupt the initial precipitate of Fe(III)PPIX. This comes about because all of these studies involve introducing a fixed quantity of dissolved hemin or hematin to an acidic aqueous environment which brings about immediate precipitation. In the malaria parasite and probably in other hemozoin forming organisms, heme is released from hemoglobin in a continuous manner as the protein is digested. In P. falciparum neither Mo¨ssbauer spectra nor TEM provide any indication of observable precipitation of Fe(III)PPIX in the food vacuole in any form other than hemozoin. Given that most studies in the presence of biological materials have also been conducted by adding a bolus of Fe(III)PPIX to acidic aqueous solution at the beginning of the process, the conclusions reached about their rates, roles in hemozoin formation in vivo and overall capacity to promote hemozoin formation must be treated with extreme caution. 5. Recent biological evidence In 2001, Papalexis et al. showed by immunofluorescence labeling that HRP II is indeed present in the food vacuole of P. falciparum. However, a striking feature of this study is that a considerable proportion of the HRP II is located outside the food vacuole either in the red cell or even parasite cytoplasm [69]. While this could simply represent the route of trafficking of HRP to the food vacuole it would seem rather an inefficient distribution of HRP if its primary role is to promote hemozoin formation. Furthermore, investigations on the extent of b-hematin formation in a 24 h period suggested that HRP II could only convert about a third of the Fe(III)PPIX to hemozoin at concentrations at which it is likely to occur. Later it was shown that a P. falciparum clone lacking the genes for both HRP II and HRP III forms hemozoin normally. Indeed the hemozoin crystals from this clone are indistinguishable from other parasites [60]. This strongly indicates that HRP is not in fact involved in hemozoin formation, although it has been pointed out that other histidine rich proteins occur in the parasite and could compensate for the loss of HRP II and III, so that HRP cannot be entirely ruled out on these grounds alone [69]. However, HRP homologues have not been identified in P. vivax or in the rodent malaria parasite P. berghei, both of which produce hemozoin [70]. There also appear to be no reports of HRP in other hemozoin forming organisms. In sum, these observations appear to rule out a role for HRP in hemozoin formation. By contrast to the declining evidence for involvement of HRP, evidence for the role of lipids has been growing. Pal-

1294

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

acpac et al. demonstrated the presence of lipid bodies containing triacylglycerols in P. falciparum [71] and showed that the enzyme responsible for synthesis of triacylglycerols is essential for intraerythrocytic proliferation of the parasite [72]. Jackson et al. used Nile Red labeling to demonstrate that neutral lipid bodies are associated with the food vacuole, are several hundred nanometers in diameter and appear to consist largely of diacyl- and triacylglycerols. The model lipids mono- and dimyristoyl- and mono- and dioleoylglycerol were found to convert about 60–70% of a 150 lM suspension of Fe(III)PPIX at pH 5 and 37 °C in 24 h, considerably more efficient than HRP II [73]. In 2005, Coppens and Vielemeyer reviewed the literature on neutral lipid bodies in Apicomplexa, a relatively unexplored field [74]. In this review they published a previously unreported TEM of a malaria parasite stained with malachite green, a substance used to fix neutral lipids. This startling image shows that hemozoin is completely embedded within a neutral lipid body (Fig. 4a), a feature never observed before because processing for TEM dissolves such lipids and washes them out of the sample. Subsequently Pisciotta et al. have published a detailed study on these neutral lipid bodies and their ability to promote bhematin formation [75]. This study included further images of hemozoin within what have now been dubbed lipid nanospheres (Fig. 4b). Furthermore, these bodies were isolated by sucrose cushion centrifugation and the lipids characterized. Thin layer chromatography showed that the major constituents of these nanospheres are monoacylglycerols, with diminished quantities of di- and triacylglycerols. Hydrolysis of the lipids followed by gas–liquid chromatography–mass spectrometry showed that the predominant fatty acid components of the lipids are stearic and palmitic acid suggesting that the nanospheres consist mainly of monopalmitoyl- and monostearoylglycerol. Gel electrophoresis with Coomassie Brilliant Blue staining detected

no evidence of proteins associated with the lipid nanospheres and Western blotting showed no evidence of the presence of HRP II or of the food vacuole membrane protein PfCRT. P. falciparum aldolase could also not be detected by Western blotting. These findings argue against involvement of proteins in hemozoin formation and appear to finally rule out HRP II. However, proteins cannot be entirely excluded as it remains possible that an unknown protein associated with the surface of the lipid nanosphere and present at very low levels could be involved in nucleating hemozoin. Nonetheless, monopalmitoylglycerol (MPG) in particular was found to be extremely efficient in promoting b-hematin formation, with formation of the product within 10 min. Furthermore, the reaction was found to be unaffected by globin, a key finding since the food vacuole is likely to contain high concentrations of this protein in the process of being digested. Finally, the lipids also protect Fe(III)PPIX from degradation by H2O2 which can be expected to be formed as consequence of the oxidation of heme (Fe(II)PPIX) to Fe(III)PPIX in the food vacuole. In parallel with these findings in P. falciparum, the close association between hemozoin formation and lipids has also been demonstrated in the unrelated organisms S. mansoni and R. prolixus [55]. Oliveira et al. showed that hemozoin formation in S. mansoni occurs within lipid droplets in the worm’s gut (Fig. 5a). These bodies are roughly spherical and contain no evidence of a bilayer at their surface, the hallmark of a lipid droplet. This study, together with a subsequent one [76], have shown that some droplets contain only small quantities of hemozoin, while others are almost completely filled with crystals. However, where small quantities of crystals are present these are invariably located near the inner surface of the lipid droplet, making contact with the aqueous interface (Fig. 5b). Indeed, TEMs seem to indicate that the hemozoin grows from the interface into the lipid droplet (Fig. 5b) [76]. In the case of R. prolixus,

Fig. 4. The association between hemozoin formation and lipids in P. falciparum. (a) Formation of hemozoin within a lipid body (marked by arrows) is evident when neutral lipids are fixed with malachite green (from Coppens and Vielemeyer [74]). Reprinted from: I. Coppens, O. Vielemeyer, Insights into unique physiological features of neutral lipids in Apicomplexa: from storage to potential mediation in parasite metabolic activities, Int. J. Parasitol. 35, 597–615. ÓAustralian Society for Parasitology (2005), with permission from Elsevier. (b) Hemozoin formation within lipid nanospheres has recently been demonstrated by Pisciotta et al. [75]. These are readily seen with malachite green fixation. Reprinted from: J.M. Pisciotta, I. Coppens, A.K. Tripathi, P.F. Scholl, J. Shuman, S. Bajad, V. Shulaev, D.J. Sullivan, Biochemical Journal 402 (2007) 197–204. ÓThe Biochemical Society. Scale bars = 0.5 (a) and 1.0 lm (b), respectively.

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

1295

Fig. 5. (a) In S. mansoni hemozoin (marked c) can clearly been seen to form in lipid droplets, objects bounded by a single lipid layer (inset). (b) These crystals are clearly associated with surface of the droplet and appear to grow inwards. The droplets in this image seem to show consecutive stages of growth, numbered 1–5. (c) Hemozoin (marked c) formation in R. prolixus is also associated with lipids. In this case the crystals appear to form within vesicles bounded by a bilayer (inset). Images in (a) and (c) from Oliveira et al. [55] and in (b) from Soares et al. [76]. Scale bars in insets: 25 nm. (a) and (c) Reprinted from: M.F. Oliveira, S. Kycia, A. Gonzales, A.D. Kosar, D.S. Bohle, E. Hempelmann, D. Menezes, M. Vannier-Santos, P.L. Oliveira, S.T. Ferreira, Structural and morphological characterization of hemozoin produced by Schistosoma mansoni and Rhodnius prolixus, FEBS Lett. 579, 6010– 6016. ÓFederation of European Biochemical Societies (2005), with permission from Elsevier. (b) Reprinted from: J.B.R.C. Soares, F.A. Lara, P.R.B.B. Cunha, G.C. Atella, C.M. Maya-Monteiro, J.C.P. d’Avila, D. Menezes, M. Vannier-Santos, P.L. Oliveira, T.J. Egan, M.F. Oliveira, Extracellular lipid droplets promote hemozoin crystallization in the gut of the blood fluke Schistosoma mansoni, FEBS Lett. 581, 1742–1750. ÓFederation of European Biochemical Societies (2007), with permission from Elsevier.

hemozoin formation occurs within vesicles bounded by a bilayer membrane (Fig. 5c). The crystals make contact with the bilayer, but seemingly grow into the aqueous environment [55]. This indicates that the process may be somewhat different to that in either the malaria parasite or schistosomal worm. Lipid droplets from S. mansoni support b-hematin formation and the process can be fitted to the Avrami equation [76]. The reaction is relatively slow, but this can probably be ascribed to the points raised earlier, namely that Fe(III)PPIX is introduced into an acidic environment where it precipitates instead of being continuously released and delivered to the lipid droplets as might be expected to occur in vivo. Taken together, these studies provide incontrovertible evidence that hemozoin formation occurs in association with lipids in three unrelated organisms. In P. falciparum and S. mansoni the crystals undoubtedly form within the lipid and there is no evidence of protein involvement in the case of the malaria parasite. However, the rates of bhematin formation in the presence of S. mansoni regurgitates or even lipids extracted from these regurgitates [76] as well as the data on b-hematin formation in the presence of MPG and a lipid blend containing mono-acylglycerols suggest that proteins may act as chaperones to deliver Fe(III)PPIX to these lipid bodies [75]. 6. Mechanism of hemozoin formation Three recent papers shed considerable light on the mechanism of hemozoin formation in lipid bodies. The observation that hemozoin formation in S. mansoni lipid droplets occurs at the interface of the lipid and aqueous phases strongly suggests that the interface is central to the process. As these droplets are less than a micron in diameter direct investigation of the interface is extremely difficult. However, a model system which closely mimics this environment has been investigated. Octanol is routinely used to mimic

lipids as a system for measuring lipophilicity of drugs and has recently been shown to form an organized layer at the aqueous interface [77]. When hematin dissolved in 0.1 M NaOH is introduced to the interface between octanol and an aqueous solution buffered with citrate at pH 4.8 (a recent estimate of the food vacuole pH [78]) the Fe(III)PPIX spreads very rapidly at the interface and the base is almost immediately neutralized. Within a period of half an hour the Fe(III)PPIX is converted to b-hematin [79] at 37 °C. The process also occurs with a pentanol–water system. Infrared spectra of the products are identical to those of hemozoin extracted from P. falciparum and both SEM and X-ray diffraction definitively identify the product as b-hematin. The crystals are similar in size to hemozoin and show a high degree of regularity. The process does not proceed in either the aqueous or organic mediums alone. Furthermore, a resonance Raman microprobe revealed that the process occurs at or close to the interface (within 1 lm). In order to mimic the real biological process more closely, the organic layer was replaced with the model lipid monomyristoylglycerol (MMG). The MMG was dissolved in 1:10 acetone:methanol (v/v) and layered onto the aqueous substrate. As both acetone and methanol are completely miscible with water it is likely that they rapidly diffuse away leaving a film of pure lipid at the surface (the aqueous medium then contains a maximum of only 0.18% acetone and 1.8% methanol by volume, assuming no evaporation of these volatile solvents). Application of a solution of Fe(III)PPIX in 0.1 M NaOH to this interface resulted in rapid formation of b-hematin in an apparent first-order process with a half-life of 5.3 min. This is by far the fastest process of b-hematin formation that has been observed under physiologically realistic conditions. The process was shown to be extremely efficient in the presence of a series of other lipids. This study is notable in demonstrating that b-hematin formation is extremely rapid in a system mimicking the biological environments recently uncovered in P. falciparum and S. mansoni as the actual sites of hemozoin

1296

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

formation. The major difference between this study and other studies involving lipids is that the Fe(III)PPIX is delivered in solution directly to the lipid–water interface which appears to be critical to its formation. A prerequisite to understanding the molecular mechanism of the process is to determine how Fe(III)PPIX behaves in aqueous solution. For 35 years it has been believed that it spontaneously forms a l-oxo dimer in aqueous solution. This is based on infrared evidence from the solid precipitated by addition of either solid NaOH or 1 M NaOH to an aqueous solution [80], together with spectroscopic evidence of dimerization in aqueous solution [81]. A later study showing that the characteristic spectrum of the l-oxo dimer is only obtained when an organic base such as pyridine is included in a strongly alkaline solution of Fe(III)PPIX has been largely ignored [82]. In fact recent evidence confirms dimerization based on both spectrophotometric titration and diffusion measurements, but the visible spectra obtained when hematin is dissolved in aqueous solution over the pH range 6–9.7 are markedly different from the l-oxo dimer and the 1H NMR spectrum which spans almost 80 ppm closely resembles that of Cl–Fe(III)PPIX dimethyl ester in chloroform [83], and which is characteristic of a five-coordinate high spin complex, but is utterly different from the l-oxo dimer induced by 10% v/v pyridine in 0.1 M NaOH which spans only the usual 10 ppm owing to strong antiferromagnetic coupling [84]. Magnetic susceptibility measurements obtained by the Evans method further support this [84], although an error was made in using the equation for a continuous wave instrument with a perpendicular field in the original publication. However, recalculation using the appropriate equation for a modern NMR instrument gives a magnetic susceptibility for the loxo dimer of 1.05 BM, in excellent agreement with the expected value of 1.1 BM at 298 K [82], while that for Fe(III)PPIX in aqueous solution (4.3 BM) is in reasonable agreement with that observed for the dimer of [H2O– Fe(III)octaethylporphyrin]ClO4  2H2O (4.8 BM at 298 K) which exhibits a mixed S = 3/2, S = 5/2 spin state as well as p-stacking between the porphyrin rings and not a l-oxo bridge, a reasonable model for the Fe(III)PPIX dimer. The fact that the observed magnetic susceptibility is not that expected for an isolated S = 5/2 high-spin d5 complex (5.99 BM) supports its similarity to H2O– Fe(III)octaethylporphyrin and indicates antiferromagnetic coupling between the iron centers. In the case of a l-oxo dimer this is very strong, but it is much weaker in the pstacked dimer. Indeed the somewhat lower magnetic susceptibility of the proposed H2O–Fe(III)PPIX p-stacked dimer relative to H2O–Fe(III)octaethylporphyrin probably arises from stronger interaction between the Fe(III) centers in accordance with the smaller lateral shift between the porphyrins predicted by molecular dynamics simulations (Fig. 6a). Further molecular dynamics simulations of the interactions of such a p-stacked dimer of H2O–Fe(III)PPIX in vacuum indicate that the porphyrins rapidly rearrange to

Fig. 6. (a) The structure of the H2O–Fe(III)PPIX dimer in aqueous solution proposed in a recent study [84], reprinted from: K.A. de Villiers, C.H. Kaschula, T.J. Egan, H.M. Marques, Speciation and structure of ferriprotoporphyrin IX in aqueous solution: spectroscopic and diffusion measurements demonstrate dimerization, but not l-oxo dimer formation, J. Biol. Inorg. Chem. 12, 101–117, Fig. 8C. ÓThe Society of Biological Inorganic Chemistry (2007), with kind permission from Springer Science and Business Media. (b) A hemozoin precursor dimer formed by interaction of two H2O–Fe(III)PPIX molecules in vacuum and modeled by molecular dynamics simulation. (c) A graphical representation of the conversion of the hemozoin precursor dimer to the hemozoin dimer. (d) The hemozoin precursor dimer is not stable in the presence of water molecules because the propionate groups are drawn away from the Fe(III) center and interact with solvent molecules. (b–d) From Egan et al. [79]. Reprinted from: T.J. Egan, J.Y.-J. Chen, K.A. de Villiers, T.E. Mabotha, K.J. Naidoo, K.K. Ncokazi, S.J. Langford, D. McNaughton, S. Pandiancherri, B.R. Wood, Haemozoin (b-haematin) biomineralization occurs by self-assembly near the lipid/water interface, FEBS Lett. 580, 5105– 5110. ÓFederation of European Biochemical Societies (2006), with permission from Elsevier.

form a b-hematin precursor [79]. This involves an increase in the lateral shift between the porphyrins and electrostatic interaction of the negatively charged propionate of one H2O–Fe(III)PPIX with the positively charged Fe(III) center of the other (Fig. 6b). Such interaction might also be expected in the low dielectric environment of the lipid body. Subsequent loss of the axial water molecules from the porphyrin Fe(III) centers and concomitant formation of propionate to iron bonds results in formation of the bhematin dimer (Fig. 6c). In the absence of competing hydrogen bonding from the solvent, these dimers can then hydrogen bond to each other to assemble the crystal [79].

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

In an aqueous environment, where competing hydrogen bonding occurs, the simulation indicates that the b-hematin precursor rapidly rearranges to the type of p-stacked dimer discussed earlier, with the propionate groups interacting with solvent molecules, rather than the Fe(III) center (Fig. 6d). These simulations suggest that the lipid environment promotes interaction of the propionate group with the Fe(III) center, assists in driving water off the axial position (as suggested by Pisciotta et al. [75]) and promotes

1297

hydrogen bonding required for hemozoin crystal formation. While the molecular dynamics simulations provide an indication of how the hemozoin dimer forms in a lipid body and suggests the role of such an environment in promoting hydrogen bonding between the dimers required for assembly of the crystal, it does not explain how the nucleation process itself occurs. A recent study using grazing incidence X-ray diffraction (GIXD), X-ray reflectivity (XR) and spec-

Fig. 7. (a) The characteristic morphology of hemozoin crystals with the faces labeled according to the {h, k, l} indices according to Buller et al. [53]. (b) Evidence from GIXD showing that b-hematin crystals floating on water preferentially lie on their {1 0 0} faces. Here qh and qv are the X-ray scattering vector components oriented horizontal and vertical to the water surface respectively. Crystal planes that lie parallel to the water surface make no contribution to the diffraction. The presence of a very small {1 0 0} indicates that only a minor fraction of crystals do not lie on their {1 0 0} face [62]. (c) Evidence from specular X-ray diffraction h/2h scans that b-hematin crystals nucleated from chloroform using a DBPC–OTS–Si wafer monolayer are preferentially oriented with their {1 0 0} faces parallel to the DBPC layer. The scan senses only crystallographic planes parallel to the OTS–Si wafer. The very prominent {1 0 0} peak and absence of other normally prominent peaks in the diffraction pattern (e.g. {0 2 0} and {0 3 1}) indicates the preferred orientation [62]. (d) A schematic representation of epitaxial growth at the surface of a phospholipid. Here the carboxylic acid groups which emerge obliquely through the {1 0 0} face of the crystal are shown interacting with the head groups of the lipid molecules (represented by spheres). (a) Reprinted ¨ . Almarsson, L. Leiserowitz, Quinoline binding site on malaria pigment crystal: a rational pathway for with permission from: R. Buller, M.L. Peterson, O antimalaria drug design. Cryst. Growth Des. 2 (2002) 553–562. ÓThe American Chemical Society (2002). (b, c) reprinted with permission from: I. Solomonov, M. Osipova, Y. Feldman, C. Baehtz, K. Kjaer, I.K. Robinson, G.T. Webster, D. McNaughton, B.R. Wood, I. Weissbuch, L. Leiserowitz, J. Am. Chem. Soc. 129 (2007) 2615–2627. ÓThe American Chemical Society (2007).

1298

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

ular X-ray diffraction scans has begun to shed some light on this issue [62]. Although the authors of this study were unable to produce b-hematin from Cl–Fe(III)PPIX and 2,6-lutidine at the surface of water, even using a monolayer of dipalmitoylphosphatidyl-choline or -ethanolamine, they were able to demonstrate by GIXD and XR that preformed b-hematin crystals float preferentially on their {1 0 0} face (Fig. 7a and b). When a surface layer of the lipid dibehenoyl-L-a-phosphatidylcholine (DBPC) was transferred from a water surface to an octadecyltrichlorosilane–silicon (OTS–Si) wafer, it was able to nucleate b-hematin crystals from chloroform with a measure of preferred orientation. Specular X-ray diffraction scans demonstrated a minor preference for the {1 0 0} face to lie parallel to the surface plane of the DBPC–OTS–Si wafer (Fig. 7c). These studies hint that the surface of the lipid body may well play a role in orienting the crystals. Crystal growth occurs fastest along the c-axis of the crystal [53] and the authors point to evidence from TEM of malaria parasites suggesting that crystals appear to be aligned parallel to c. This indicates that nucleation occurs via the {1 0 0} or {0 1 0} faces (Fig. 7d) and may involve epitaxial nucleation by the lipid head groups. While these observations are intriguing, the details remain to be investigated. It is likely that application of GIXD and XR to lipid–water systems known to promote b-hematin formation will soon permit elucidation of the process of hemozoin nucleation.

ospheres from P. falciparum [75]. Much more likely is a role for proteins in transporting Fe(III)PPIX to the lipid bodies, possibly as chaperones as has recently been suggested [75,79]. The site of oxidation of heme (Fe(II)PPIX) to Fe(III)PPIX is also not known. Indeed little if anything is currently really known about this step in the process, which is an area in need of investigation. Finally, it is worth pointing out that with notable exceptions [85–87] the emphasis on Fe(III)PPIX–drug interactions has been in the aqueous medium and most studies on inhibition of b-hematin formation have also concentrated on the aqueous environment. A rethink is now needed, with a new emphasis on Fe(III)PPIX in non-aqueous environments that mimic lipids and indeed within the lipid environment itself. Such studies could enhance our chances of finding new antimalarials of this type.

7. Conclusions

[1] G.M. Lancisi, De Noxiis Paludum Effluvius Eorumque Remediis, J.M. Salvioni, Rome, 1717. [2] R. Ross, Br. Med. J. ii (1897) 1736–1788. [3] W. Brown, J. Exp. Med. 13 (1911) 290–299. [4] C.D. Fitch, P. Kanjananggulpan, J. Biol. Chem. 262 (1987) 15552– 15555. [5] A. Hamsik, Z. Physiol. Chem. 190 (1936) 199–215. [6] A.F.G. Slater, W.J. Swiggard, B.R. Orton, W.D. Flitter, D.E. Goldberg, A. Cerami, G.B. Henderson, Proc. Natl. Acad. Sci. USA 88 (1991) 325–329. [7] D.S. Bohle, R.E. Dinnebier, S.K. Madsen, P.W. Stephens, J. Biol. Chem. 272 (1997) 713–716. [8] S. Pagola, P.W. Stephens, D.S. Bohle, A.D. Kosar, S.K. Madsen, Nature 404 (2000) 307–310. [9] S.H. Vincent, Semin. Hematol. 26 (1989) 105–113. [10] A.C. Chou, C.D. Fitch, J. Clin. Invest. 68 (1981) 672–677. [11] M.F. Oliveira, B. Timm, E.A. Machado, K. Miranda, M. Attias, J.R. Silva, M. Dansa-Petretski, M.A. de Oliveira, W. de Souza, N.M. Pinhal, J.J.F. Sousa, N.V. Vugman, P.L. Oliveira, FEBS Lett. 512 (2002) 139–144. [12] H.M. Baker, B.F. Anderson, E.N. Baker, Proc. Natl. Acad. Sci. USA 100 (2003) 3579–3583. [13] F. Zhang, M. Krugliak, H. Ginsburg, Mol. Biochem. Parasitol. 99 (1999) 129–141. [14] P. Loria, S. Miller, M. Foley, L. Tilley, Biochem. J. 339 (1999) 363– 370. [15] T.J. Egan, J.M. Combrinck, J. Egan, G.R. Hearne, H.M. Marques, S. Ntenteni, B.T. Sewell, P.J. Smith, D. Taylor, D.A. van Schalkwyk, J.C. Walden, Biochem. J. 365 (2002) 343–347. [16] B. Gligorijevic, R. McAllister, J.S. Urbach, P.D. Roepe, Biochemistry 45 (2006) 12400–12410. [17] M.F. Oliveira, J.R. Silva, M. Dansa-Petretski, W. de Souza, U. Lins, C.M.S. Braga, H. Masuda, P.L. Oliveira, Nature 400 (1999) 517–518. [18] M. Oliveira, J.C.P. d’Avila, C.R. Torres, P.L. Oliveira, A.J. Tempone, F.D. Rumjanek, C.M.S. Braga, J.R. Silva, M. Dansa-

Recent evidence from electron microscopy of three different hemozoin forming organisms, the malaria parasite P. falciparum, the helminth worm S. mansoni and the insect R. prolixus provide incontrovertible evidence that hemozoin formation occurs in close association with lipids [55,74–76]. Indeed, in the case of the first two organisms it occurs within lipid bodies. In vitro growth of b-hematin under biomimetic conditions indicate that this crystalline solid forms exceptionally efficiently in a lipid environment [75,79], especially when Fe(III)PPIX is directly introduced at or near the interface between the lipid and aqueous solution [79]. The rate at which this occurs suggests that the lipid/water interface itself is probably sufficient to bring about hemozoin formation in vivo. Indeed, the lipid bodies may be produced by the organism specifically for the purpose of heme detoxification through hemozoin formation. Modeling studies suggest that the low dielectric constant of the lipid environment promotes self-assembly of the hemozoin crystal. Despite these advances in understanding hemozoin formation, the nucleation of hemozoin crystals still remains poorly understood, notwithstanding recent evidence hinting at epitaxial nucleation by the lipid surface [62]. Understanding this nucleation process is now a priority in this field. The question of whether a protein may play a role in nucleation remains an open one. However, if any such protein is involved it must act at exceptionally low concentrations as no evidence of proteins was found in lipid nan-

Acknowledgements The author acknowledges the financial support of the National Research Foundation under Grant No. 2069079, the Medical Research Council of South Africa, the South African Malaria Initiative (SAMI) and the University of Cape Town. References

T.J. Egan / Journal of Inorganic Biochemistry 102 (2008) 1288–1299

[19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32]

[33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48]

[49] [50] [51] [52] [53] [54]

Petretski, M.A. Oliveira, W. de Souza, S.T. Ferreira, Mol. Biochem. Parasitol. 111 (2000) 217–221. M.M. Chen, L. Shi, D.J. Sullivan, Mol. Biochem. Parasitol. 113 (2001) 1–8. J.M. Pisciotta, E.L. Ponder, B. Fried, D.J. Sullivan, Int. J. Parasitol. 35 (2005) 1037–1042. A.F.G. Slater, A. Cerami, Nature 355 (1992) 167–169. T.J. Egan, D.C. Ross, P.A. Adams, FEBS Lett. 352 (1994) 54–57. A. Dorn, R. Stoffel, H. Matile, A. Bubendorf, R.G. Ridley, Nature 374 (1995) 269–271. M. Mungthin, P.G. Bray, R.G. Ridley, S.A. Ward, Antimicrob. Agents Chemother. 42 (1998) 2973–2977. D.J. Sullivan, I.Y. Gluzman, D.G. Russell, D.E. Goldberg, Proc. Natl. Acad. Sci. USA 93 (1996) 11865–11870. T.J. Egan, J. Inorg. Biochem. 100 (2006) 916–926. J. Ziegler, R. Linck, D.W. Wright, Curr. Med. Chem. 8 (2001) 171– 189. T.J. Egan, Drug Design Revs. 1 (2004) 93–110, Online. B.L. Tekwani, L.A. Walker, Comb. Chem. High Througput Scr. 8 (2005) 61–77. P.G. Bray, S.A. Ward, P.M. O’Neill, Curr. Top. Microbiol. Immunol. 295 (2005) 3–38. S. Kumar, M. Guha, V. Choubey, P. Maity, U. Bandyopadhyay, Life Sci. 80 (2007) 813–828. M.F. Oliveira, J.C. d’Avila, A.J. Tempone, J. Soares, F.D. Rumjanek, A. Ferreira-Pereira, S.T. Ferreira, P.L. Oliveira, J. Infect. Dis. 190 (2004) 843–852. T. Ha¨nscheid, T.J. Egan, M.P. Grobusch, Lancet Infect. Dis. 7 (2007) 675–685. T.J. Egan, J. Inorg. Biochem. 91 (2002) 19–26. D.S. Bohle, A.D. Kosar, P.W. Stephens, Acta Cryst. D 58 (2002) 1752–1756. K.K. Ncokazi, T.J. Egan, Anal. Biochem. 338 (2005) 306–319. N. Basilico, E. Pagani, D. Monti, P. Olliaro, D. Taramelli, J. Antimicrob. Chemother. 42 (1998) 55–60. R. Baelmans, E. Deharo, G. Bourdy, V. Munoz, C. Quenevo, M. Sauvain, H. Ginsburg, J. Ethnopharm. 73 (2000) 271–275. S. Parapini, N. Basilico, E. Pasini, T.J. Egan, P. Olliaro, D. Taramelli, D. Monti, Exp. Parasitol. 96 (2000) 249–256. T.J. Egan, M.G. Tshivhase, Dalton Trans. (2006) 5024–5032. G. Blauer, M. Akkawi, Arch. Biochem. Biophys. 398 (2002) 7–11. N.T. Huy, D.T. Uyen, A. Maeda, D.T.X. Trang, T. Oida, S. Harada, K. Kamei, Antimicrob. Agents Chemother. 51 (2007) 350–353. N.T. Huy, A. Maeda, D.T. Uyen, D.T.X. Trang, M. Sasai, T. Shiono, T. Oida, S. Harada, K. Kamei, Acta Trop. 101 (2007) 130–138. D.S. Bohle, J.B. Helms, Biochem. Biophys. Res. Commun. 193 (1993) 504–508. D.S. Bohle, P. Debrunner, P.A. Jordan, S.K. Madsen, C.E. Schulz, J. Am. Chem. Soc. 120 (1998) 8255–8256. D.S. Bohle, A.D. Kosar, P.W. Stephens, Can. J. Chem. 81 (2003) 1285–1291. A. Sienkiewicz, J. Krzystek, B. Vileno, G. Chatain, A.J. Kosar, D.S. Bohle, L. Forro´, J. Am. Chem. Soc. 128 (2006) 4534–4535. C.K. Carney, A.C. Schrimpe, K. Halfpenny, S. Reese, C.M. Miller, M. Broncel, S.L. Sewell, J.E. Schaff, D. Ravinder, M.D. Carter, D.W. Wright, J. Biol. Inorg. Chem. 11 (2006) 917–929. A.V. Pandey, B.L. Tekwani, FEBS Lett. 393 (1996) 189–192. T.J. Egan, E. Hempelmann, W.W. Mavuso, J. Inorg. Biochem. 73 (1999) 101–107. G. Blauer, M. Akkawi, Biochem. J. 346 (2000) 249–250. T.J. Egan, W.W. Mavuso, K.K. Ncokazi, Biochemistry 40 (2001) 204–213. ¨ . Almarsson, L. Leiserowitz, Cryst. R. Buller, M.L. Peterson, O Growth Des. 2 (2002) 553–562. M. Walczak, K. Lawniczak-Janlonska, A. Sienkiewicz, I.N. Demchenko, E. Piskorska, G. Chatain, D.S. Bohle, Nucl. Instr. Meth., Phys. Res. B 238 (2005) 32–38.

1299

[55] M.F. Oliveira, S. Kycia, A. Gonzales, A.D. Kosar, D.S. Bohle, E. Hempelmann, D. Menezes, M. Vannier-Santos, P.L. Oliveira, S.T. Ferreira, FEBS Lett. 579 (2005) 6010–6016. [56] K. Bendrat, B.J. Berger, A. Cerami, Nature 378 (1995) 138–139. [57] A. Dorn, S.R. Vippagunta, H. Matile, A. Bubendorf, J.L. Vennerstrom, R.G. Ridley, Biochem. Pharmacol. 55 (1998) 737–747. [58] Y. Kurosawa, A. Dorn, M. Kitsuji-Shirane, H. Shimada, T. Satoh, H. Matile, W. Hofheinz, R. Masciadri, M. Kansy, R.G. Ridley, Antimicrob. Agents Chemother. 44 (2000) 2638–2644. [59] C.R. Chong, D.J. Sullivan, Biochem. Pharmacol. 66 (2003) 2201– 2212. [60] G.S. Noland, N. Briones, D.J. Sullivan, Mol. Biochem. Parasitol. 130 (2003) 91–99. [61] C.D. Fitch, G.-Z. Cai, Y.-F. Chen, J.D. Shoemaker, Biochim. Biophys. Acta 1454 (1999) 31–37. [62] I. Solomonov, M. Osipova, Y. Feldman, C. Baehtz, K. Kjaer, I.K. Robinson, G.T. Webster, D. McNaughton, B.R. Wood, I. Weissbuch, L. Leiserowitz, J. Am. Chem. Soc. 129 (2007) 2615–2627. [63] W. Kaim, B. Schwederski, in: Bioinorganic Chemistry: Inorganic Elements in the Chemistry of LifeAn introduction and guide, Wiley, Chichester, 1996, pp. 303–317. [64] D.J. Sullivan, I.Y. Gluzman, D.E. Goldberg, Science 271 (1996) 219– 222. [65] J. Ziegler, R.T. Chang, D.W. Wright, J. Am. Chem. Soc. 121 (1999) 2395–2400. [66] A.V. Pandey, V.K. Babbarwal, J.N. Okoyeh, R.M. Joshi, S.K. Puri, R.L. Singh, V.S. Chauhan, Biochem. Biophys. Res. Commun. 308 (2003) 736–743. [67] P.A. Adams, T.J. Egan, D.C. Ross, J. Silver, P.J. Marsh, Biochem. J. 318 (1996) 25–27. [68] B. Gligorijevic, T. Bennett, R. McAllister, J.S. Urbach, P.D. Roepe, Biochemistry 45 (2006) 12411–12423. [69] V. Papalexis, M.-A. Siomos, N. Campanale, X.-G. Guo, G. Kocak, M. Foley, L. Tilley, Mol. Biochem. Parasitol. 115 (2001) 77–86. [70] D.J. Sullivan, Int. J. Parasitol. 32 (2002) 1645–1653. [71] N.M.Q. Palacpac, Y. Hiramine, F. Mi-ichi, M. Torii, K. Kita, R. Hiramatsu, T. Horii, T. Mitamura, J. Cell Sci. 117 (2004) 1469–1480. [72] N.M.Q. Palacpac, Y. Hiramine, S. Seto, R. Hiramatsu, T. Horii, T. Mitamura, Biochem. Biophys. Res. Commun. 321 (2004) 1062–1068. [73] K.E. Jackson, N. Klonis, D.J.P. Ferguson, A. Adisa, C. Dogovski, L. Tilley, Mol. Microbiol. 54 (2004) 109–122. [74] I. Coppens, O. Vielemeyer, Int. J. Parasitol. 35 (2005) 597–615. [75] J.M. Pisciotta, I. Coppens, A.K. Tripathi, P.F. Scholl, J. Shuman, S. Bajad, V. Shulaev, D.J. Sullivan, Biochem. J. 402 (2007) 197–204. [76] J.B.R.C. Soares, F.A. Lara, P.R.B.B. Cunha, G.C. Atella, C.M. Maya-Monteiro, J.C.P. d’Avila, D. Menezes, M. Vannier-Santos, P.L. Oliveira, T.J. Egan, M.F. Oliveira, FEBS Lett. 581 (2007) 1742– 1750. [77] I. Benjamin, Chem. Phys. Lett. 393 (2004) 453–456. [78] R. Hayward, K.J. Saliba, K. Kirk, J. Cell Sci. 119 (2006) 1016–1025. [79] T.J. Egan, J.Y.-J. Chen, K.A. de Villiers, T.E. Mabotha, K.J. Naidoo, K.K. Ncokazi, S.J. Langford, D. McNaughton, S. Pandiancherri, B.R. Wood, FEBS Lett. 580 (2006) 5105–5110. [80] S.B. Brown, P. Jones, I.R. Lantzke, Nature 223 (1969) 960–961. [81] S.B. Brown, T.C. Dean, P. Jones, Biochem. J. 117 (1970) 733–739. [82] D.H. O’ Keeffe, C.H. Barlow, G.A. Smythe, W.H. Fuchsman, T.H. Moss, H.R. Lilienthal, W.S. Caughey, Bioinorg. Chem. 5 (1975) 125– 147. [83] D.L. Budd, G.N. Le Mar, K.C. Langry, K.M. Smith, R. NayyirMazhir, J. Am. Chem. Soc. 101 (1979) 6091–6096. [84] K.A. de Villiers, C.H. Kaschula, T.J. Egan, H.M. Marques, J. Biol. Inorg. Chem. 12 (2007) 101–117. [85] P.G. Bray, M. Mungthin, R.G. Ridley, S.A. Ward, Mol. Pharmacol. 54 (1998) 170–179. [86] P.G. Bray, O. Janneh, K. Raynes, M. Mungthin, H. Ginsburg, S.A. Ward, J. Cell Biol. 145 (1999) 363–376. [87] T.J. Egan, K.K. Ncokazi, J. Inorg. Biochem. 98 (2004) 144–152.