CHEMICAL BIOLOGY
Hemozoin: A Paradigm for Biominerals in Disease
Melissa D. Carter, Anh N. Hoang and David W. Wright, Vanderbilt University, Nashville, Tennessee
doi: 10.1002/9780470048672.wecb649
Biomineralization is the formation of organic-inorganic composites by organisms. Originally evolved as a protective mechanism, this complex process has also become a recognized contributor to several disease states, which range from kidney stone disease (nephrolithiasis) to parasitic diseases like malaria. The characteristic three-step process for the formation of biominerals is defined by the supramolecular preorganization of a nucleating template, the interfacial molecular recognition of crystal nuclei and the cellular processing of resultant aggregates. Hemozoin formed in the heme detoxification pathway used by the malarial parasite Plasmodium falciparum represents a paradigm for pathogenic biominerals. Current research indicates that a supramolecular lipid template organizes heme released previously during hemoglobin catabolism. Nucleation and growth of the heme aggregate serves to protect the parasite from the toxic effects of free heme. Given the mechanisms of biomineralization, it is not surprising to discover that century-old antimalarial compounds function by disrupting key interactions between the heme substrate and template. Subsequently, the heme-aggregate is released into the host vasculature and deposits in patients' brains, spleens, and livers where it disrupts host innate immune response. The underlying basis of this immunomodulating activity seems to result from hemozoin mediated lipid peroxidation. Understanding the relationships between hemozoin formation and its pathogenic activity with the host immune response represents a significant challenge to chemical biology.
Biomineralization is the biologic formation of organic-inorganic composites generally organized on a nucleating template of organic material. During bone and teeth formation, hydroxyapatite, which is a calcium phosphate derivative enriched with carbonate, mineralizes on a template of collagen fibrils and other proteins (1). Likewise, calcium oxalate monohydrate crystals mineralize on a biologic matrix of proteins in both plants and humans. Although these crystals function as tissue support in plants (2), they cause painful inflammation in humans (3). In fact, biomineralization is associated with several human diseases. Nephrolithiasis is the mineralization of calcium oxalate monohydrate in the urinary tract of patients as kidney stones. In gout or metabolic arthritis, monosodium urate crystals accumulate in the articular joints. A patient’s innate immune response or the on-rush of monocytes and neutrophils at the sites of crystal sedimentation causes the characteristic tissue inflammation and pain associated with these diseases. Biomineralization can also contribute to the pathology of parasitic diseases like malaria and schistosomiasis. When the malarial protozoa Plasmodium falciparum deposits heme-derived aggregates in host vasculature, these aggregates travel through the bloodstream and collect in the brain, spleen, and liver. When they are phagocytosed by innate immune cells such as monocytes and neutrophils, the cells’ ability to produce reactive oxygen and nitrogen species is impaired. This immunomodulation is typical of the pathogenesis caused by many biomineral-associated diseases.
Heme Homeostasis
Over 40% of the world’s population is at risk from malaria. The disease causes severe illness in over 500 million people and results in over 1.7 million deaths each year (4). Transmission is most prevalent in the world’s poorest countries, predominantly sub-Saharan Africa, and accounts for 40% of public healthcare costs. In addition, a 2007 World Health Organization report estimates a 1.3% decline in annual economic growth for countries with high rates of malarial infection (4). Compounded over time, this drop contributes significantly to GDP disparities between those countries where malarial infection is endemic and those where it is not.
Challenges of heme homeostasis
Over 100 Plasmodium species contribute to the spread of malaria, but only four of these (P. falciparum, P. vivax, P. ovale, and P. malariae) account for human infection, the deadliest being P. falciparum. The malaria life cycle exists first in a mosquito, and then it passes to a human host. An infected female Anopholes mosquito is the host of the parasite’s sporogonic life cycle. Mature P. falciparum sporozoites reach the salivary glands of the mosquito, and the parasite is transmitted to a human host when the mosquito feeds. During this blood meal, sporozoites are released into the bloodstream where they penetrate hepatic cells and mature into schizonts. The liver cells rupture after approximately two weeks, discharging merozoites into the bloodstream whereupon they infect red blood cells (RBCs). Every 48 to 78 hours, mature merozoites rupture from spent RBCs and either they differentiate into gametocytes or they infect more RBCs. This blood stage is responsible for the clinical manifestation of the disease (5).
P. falciparum ingest and degrade up to 80% of host erythrocyte hemoglobin (Hb) (5) to provide the parasite with essential amino acids for growth and maturation. During high parasitemia (20%), up to 100 g of the 750 g of circulating host Hb can be catabolized by the parasite (5). Hb is broken down in the parasite’s acidic (pH 4.5-5.2) digestive food vacuole (DV) by a suite of proteinases that includes four aspartic acid proteinases (plasmepsins PfPM1, PfPM2, PfHAP, and PfPM4), three cysteine proteinases (falcipains PfFP2, PfFP2', and PfFP3), and a metalloproteinase (falcilysin) (Fig. 1) (6). This ordered catabolic process is initiated by PfPM1 and PfPM2, which cleave between residues α33Phe and α34Leu in the hinge region of Hb. Recent quadruple knockout studies show individual plasmepsin redundancy and suggest that although each aspartic proteinase may not be essential to the intraerythrocytic stage, they could play unique roles outside Hb digestion. On plasmepsin cleavage, the protein unfolds, which releases free heme and exposes other peptide bonds to the cysteine proteinases. Knockout studies suggest gene redundancy between PfFP2 and PfFP2', which shows no effect on parasite development with changes in the falcipain present. In the complete absence of cysteine proteinases, however, undigested Hb accumulates in the parasite’s DV and causes lysis. After falcipain cleavage, subsequent protein fragments are broken down by falcilysin. The remaining peptide fragments are transported from the DV to the cytoplasm where they are degraded into the final amino acids needed by the parasite (6).
The French military physician Charles Louis Alphonse Laveran was credited with the 1880 discovery of the malaria protozoa Oscilliaria malariae, which was later termed Plasmodium. In malaria necropsies, Laveran noted the presence of dark pigmented bodies in the bloodstream and in the brain, spleen, and liver (7). This pigmented body was later characterized as a unique biomineral composed of a dimeric ferriprotoporphyrin IX (Fe(in)PPIX) aggregate, which is known commonly as hemozoin (HZ).
Although the nutritional needs of P. falciparum are met by Hb catabolism, the parasite is endangered by the release of heme. During intraerythrocytic phases, free heme can accumulate at concentrations that reach 400 mmol/L in the DV. Subsequently, the heme can cause disruption of cellular function by the inhibition of enzymes, the generation of oxidative free radicals, and the peroxidation of membranes (5), which leads ultimately to parasite death. Unlike higher organisms, P. falciparum lacks a heme-oxygenase pathway for the cellular disposal of heme. Rather, the protozoa achieves effective detoxification by removing the heme from solution through biomineralization mechanisms, which results in crystalline HZ (8).
Figure 1. Ordered pathway of hemoglobin catabolism. Host Hb cleavage is initiated by PfPM1 and PfPM2. Hb unfolds, releases heme and exposes other peptide bonds to the falcipains. Resultant protein fragments are degraded even more by falcilysin and transported to the cytosol where they are broken down into needed amino acids. The heme released in the first step aggregates via biomineralization to form HZ.
Biomineralization
Over 570 million years, the complex process of organizing inorganic molecules, which generally contain calcium, iron, or silicon, into crystal aggregates evolved as a protective mechanism for organisms. Fossil records show that biomineralization surfaced first in the form of organism scales and skeletons during the neoproterozoic era (9), and now, biominerals are recognized commonly in a range of functional structures including shells (10), vertebrate teeth, and bone (11, 12), diatom silicates (13) and unicellular silver deposits (14). The formation of these materials is attributed to two processes: 1) biologically-induced mineralization, which is the deposition of minerals via adventitious precipitation (15) and 2) biologically-controlled mineralization, which is the regulated formation of minerals that have a specific function and structure (16). Generally, induced minerals are heterogenous in character, whereas minerals that result from cellularly controlled processes exhibit uniform composition and morphology (1).
Chemistry of biomineral formation
The processes of crystal nucleation and growth are driven by the basic laws of thermodynamics in which a greater free energy must exist in the original solution phase than the resultant crystalline phase (17). However, it must be noted that the free energy distribution at the crystal surface differs from these phases (3). Because molecules on the crystal surface are not bound as strongly as molecules in the preliminary bulk solution, their free energy contribution to the system is greater. This energy difference between molecules at the surface and in solution is known as the interfacial free energy. Acting on the destabilization of crystal nuclei, the interfacial free energy can cause either nucleus dissolution or growth of the nucleus to a large enough size that its stability prevails over the affects of surface free energy, and a crystal is formed. The nucleation pathways that lead to the most stable crystal phase are the foundations of the Ostwald-Lussac law of phases, which explains that nucleation occurs in a series of pathways with increasing stability before reaching the final crystalline state (3). The propagation and continued growth of a stable crystal is attributable to a uniform nucleation template from which a new phase is formed from an old phase that has become higher in free energy (3, 18).
In the context of crystal nucleation and growth, the formation of biominerals can be cast as a three-step process that involves the supramolecular preorganization of a template, the interfacial molecular recognition of crystal nuclei, and the cellular processing of resultant aggregates (19). The first-order assembly of organic matrices such as protein and lipid networks provides a foundation for the second-order assembly of inorganic species. Typically, these frameworks have functionalized surfaces that behave as templates for inorganic nucleation and are governed by electrostatic, structural, and stereochemical complementarity at the organic-inorganic interface. Without cellular intervention to control the flux of metal ions, crystal nuclei would continue to grow along these scaffolds, proceeding to their bulk state. Clearly, such unconstrained growth represents a danger to cellular integrity. The final stage of biomineral construction is cellular processing, which is often the distinguishing step between native biomineral morphology and that of its synthetic analogs. The intracellular or extracellular environment in which a crystal grows ultimately influences its crystallographic structure and morphology (19) and ensures the function of laden cells.
Search for hemozoin's bionucleating template
Investigations into the mechanisms of HZ formation have centered on protein or lipid-rich nucleating templates. Early hypotheses of a catalytic heme-polymerase in trophozoite lysates were abandoned because of failed attempts at identification and purification of the enzymatic activity. An alternative to such a heme polymerase was proposed by Sullivan (20) from their investigations of a family of histidine-rich proteins (HRP) isolated from the DVs of P. falciparum. HRPII is a 30 kDa protein with 76% of its composition being His or Ala residues. HRPIII is 27 kDa and is 56% His or Ala residues. Both proteins have repeats of the tripeptide Ala-His-His, 51 repeats in HRPII, and 28 repeats in HRPIII. This Ala-His-His recurring domain provides an archetypal biomineralization scaffold for the nucleation and propagation of free heme into HZ. When HRPII or HRPIII were incubated with heme in aqueous acetate solution, the protein templates bound 17 molecules of heme, mediated the formation of HZ and was inhibited by the antimalarial drug chloroquine.
This HRP nucleating domain was explored even more by Ziegler et al. (21) who used template design principles to construct a dendrimer-based template model system. These bionucleating templates (BNT I and II) were composed of the nucleating domain of HRP II of P. falciparum coupled to a tetralysine dendrimer core. The templates could bind to near stochiometric amounts of Fe(III)PPIX, although substrate specificity experiments suggested that substrate recognition was dependent on the porphyrin moiety rather than specific metal recognition. Moreover, HRP substrate recognition was not shown to be mediated by histidine axial ligation to a metal ion, but rather it was attributed to π-stacking and electrostatic interactions. Chloroquine inhibition of the bionucleating templates was comparable with HRP II and III impaired formation of HZ when treated with the antimalarial (21).
These templates, like HRPII, were shown to promote HZ formation at a parasite DV relevant pH 4.0-4.5. In fact, HRPII is active from pH 4.5 to pH 6.0. HZ formation slows at pH values below 2.0 and above 5.0 (22). Although an increase in pH does cause an increase in binding of heme to HRPII, the HZ produced actually decreases leading, instead, to the formation of μ-oxo-heme dimers (22). Therefore, the bis-histidyl heme binding observed on other proteins like histidine-rich glycoprotein (HRG) at physiologic pH 7.0 differs from the HRPII nucleation of HZ in the parasite’s acidic DV (23), although a pH 7 HRPII model was attempted by Schneider and Marietta in 2005 (24). At this pH, binding is consistent with a nucleating template that serves as an organizing function rather than a tight heme-binding function.
Subsequent experiments revealed, however, that the HRP’s were not the likely template for HZ formation. Double deletion mutants of HRP II and HRP III did not prevent the formation of HZ in the DV (20), which suggests the existence of an alternative template. In addition, HRP II is not located solely in the DV, but rather it is secreted into the serum of victims at high concentrations. Histology labeling experiments show that HRPs in the DV are simply captured during the endocytosis of host hemoglobin and not specifically targeted to the DV (20). Such surreptitious colocalization would also suggest the existence of a non-HRP template. In light of these experimental results, the search for HZ’s biomineralization template turned toward other possibilities.
Lipids are a possible template for HZ biomineralization. As a bionucleating template, lipids can increase the solubility of Fe(In)PPIX, localize high concentrations of Fe(IlI)PPIX with iron-hydrophobic headgroup interactions, and provide a layer of free heme intercalation. The propionate groups of free heme within the lipid layer draw into their positively charged Fe(III) centers, which weakens any hydrogen bonds with water and enables the hydrogen bonds of Fe(III)PPIX dimers. Fitch et al. (25) showed that HZ formation could be mediated at pH 5.0 in the presence of fatty acids like arachidonic acid and glycerols of oleic acid as well as detergents like polyoxyethylene sorbitan monooleate (TWEEN 80). Increasing support for a lipid template led to the membrane sacrifice theory by Hemplemann et al. (25), which suggested that inner membranes of RBC transport vesicles in the parasite cytosome were degraded by free heme, releasing lipids that increase heme solubility and aggregation. In the sacrifice of the inner membrane, the outer membrane is preserved to prevent additional oxidative damage to the parasite (26). In vitro lipid-initiated HZ crystallization was reported by Egan et al. (27) using a range of lipids that includes myristoyl, oleoyl, and palmitoyl glycerides; phosphocholines; and cholesterol to initiate HZ formation along the lipid-water interface. Simulations of HZ formation in these studies indicated that the hydrophobic Fe(III)PPIX dimer was more stable in the lipid layer to foster hydrogen bonds between the protonated propionic acid groups and thus, HZ crystal assembly (27). Transmission electron microscopy images of HZ localized in lipid nanospheres within an infected trophozoite stage RBC provided increased support of a lipid-mediated HZ bionucleating process. The extraction of these fatty acyl glycerides (including monostearic, monopalmitic, dipalmitic, dioleic, and dilinoleic glycerols) and their incubation with substrate hemin, developed a competent in vitro template for HZ formation (28).
Target of antimalarials
Interruption of the parasite heme biomineralization pathway is a logical target of antimalarial drug development. As antimalarials amass in the parasite’s DV, HZ formation is inhibited, and the parasite becomes flooded in toxic heme. Beginning with the hypothesis that neutral lipid droplets indeed serve as biomineralization templates for the formation of HZ, a variety of limiting cases exist in which inhibitors may disrupt the aggregation of HZ (Fig. 2). An inhibitor may bind the heme substrate in such a manner that the heme-inhibitor complex cannot be recognized by the template. Alternately, a drug could interact with the template, blocking the heme binding site. Finally, a HZ aggregation inhibitor might trap the heme bound to the template, which prevents formation of the dimeric unit or nucleation of the extended crystal. These possible modalities, which are derived from the paradigm of biomineral formation, can be used to understand the mechanism of action of some antimalarial compounds.
Figure 2. Modes of hemozoin inhibition. On a neutral lipid droplet template (T), heme can aggregate to form the biomineral HZ. Antimalarials may inhibit this aggregation by binding heme substrate, interacting with the lipid template or trapping heme bound to the template. All actions serve to prevent the formation of HZ.
Well-known quinoline-based drugs like chloroquine and amodiaquine are thought to target this HZ biomineralization pathway. These drugs have been the standards in treatment of malaria, but growing parasite resistance threatens their use. Quinoline-based drugs include the well-known 4-aminoquinoline derivatives chloroquine, amodiaquine, halofantrine, quinine, and bisquinoline (29). These drugs are thought to cap monomeric heme or bind the μ-oxo-dimer of oxidized heme to prevent HZ formation. The π-π interactions at the {010} heme face control resultant adduct formation. Computational models of HZ and quinoline interactions indicated that drug adsorption occurred on the {001} and {011} crystal faces (Fig. 3) (30). These highly symmetrical crystals also showed tapering at each drug-bound {001} or {011} ledge, which suggested weakening of the quinoline inhibition along the crystal’s c-axis which resulted in thinner crystal cross-sections (30).
Other inhibitors of HZ formation are thought to act similarly to the quinoline family. Binding of the antifungal azole-based drugs (clotrimazole, ketoconazole, and miconazole) to heme is thought to damage parasite cell membranes and cause death (29). The isonitrile terpenes (diisocyanoadociane and axisonitrile-2), isolated from marine sponges and the synthetically derived methylene blue analogs (azures A, B, and C; thion; celestine blue; and phenosaphranin) most likely prohibit HZ formation by binding monomeric heme. The xanthone family of antimalarials, such as the hydroxyxanthones and the bis-(N,N-diethylamino)-ethoxy xanthones, were found also to bind at the HZ crystal faces {001} and {011}. Specifically, the drugs’ terminal amino groups bind the carboxyl group exposed at each face, which inhibits the nucleation and growth of the crystal (30). The crystal engineering prospect of future drug design is an interesting one that beckons additional exploration and promises applications in a variety of biomineral-associated diseases.
Figure 3. Hemozoin-chloroquine binding model. Chloroquine is thought to bind to the {001} and {011} faces of heme as well as π-π stacking at the {010} face. To prevent HZ formation completely, chloroquine would have to bind both faces of each heme monomer or crystal extension would be blocked in only one direction as shown.
Prior to the definitive X-ray powder diffraction characterization of HZ, the true structure of the aggregate remained elusive because of its limited solubility. HZ was soluble only in NaOH, which completely degraded the structure and prevented any attempt to determine the intramolecular atomic interactions. The insolubility of the product rendered useless many “standard” experimental methods for the characterization of bioinorganic systems. This aspect of the biomineral’s identification undoubtedly contributed to Ridley’s description of HZ as “a black insoluble mass of material that can be soul-destroying to work with” (31).
Physical characterization
Studies by Slater et al. (31) provided an initial “fingerprint” of HZ’s chemical structure by applying Fourier-transform IR (FT-IR) spectroscopy to intact HZ crystals. The IR spectrum of HZ revealed intense absorbance patterns at 1664 and 1211 cm-1, which indicate the axial propionate C = O and C-O stretching, respectively (Fig. 4a). These peaks were absent from the spectra of the synthetic substrates hemin chloride and hematin. These data suggested a direct coordination between the carboxylate of one heme monomer and the iron center of another, which rules out a direct iron-carbon bond that would have characteristic stretching near 1900 cm-1. Imagining a long carboxyl-propogated “polymer,” this structure accounted for the aggregate’s insolubility.
Using field emission inlens scanning electron microscopy, gross physical analysis of isolated HZ crystals obtained from mammalian Plasmodium species show a uniform morphology that resembles smooth-sided bricks arranged at right angles (33). Unlike the smooth, flat sides of the aggregates, the crystals’ tapering, square ends indicate a face of continuous nucleation and growth. The average dimensions of these crystals were 100 nm x 100 nm x 300-500 nm. Even strains of P. falciparum that lack HRPII or HRPIII were shown to produce uniform HZ crystals (33). This observed consistency in morphology is caused by the final step of in vivo biomineralization, which is cellular processing. In this stage, cells organize HZ crystals to best fit the confines of their storage vacuole as illustrated in an in situ HZ RBC electron micrograph image (Fig. 4c) (23). This differs from the seemingly less-ordered in vitro production of HZ’s synthetic analog, which is known as P-hematin (BH), in a reaction vessel, which fosters continued crystal extension without cellular restraint (Fig. 4d).
The introduction of high-resolution synchrotron X-ray radiation enabled the X-ray powder diffraction of both native HZ and BH. Their structure was determined using simulated annealing techniques on the diffraction data (34). This method included the use of the Le Bail algorithm to compare integrated peak intensities, the Rietveld refinement of the structure’s molecular bonding geometry and Fourier difference calculations of atomic positions. The resultant diffraction pattern confirmed a triclinic unit cell with a space group assignment of P-1, which demonstrates an inversion of symmetry between the unit cells of the Fe(III)PPIX dimer (Fig. 4b). This pattern led to the definitive structure of a five-coordinate Fe(III)PPIX dimer bound by reciprocal monodentate carboxylate interactions with the propionic side chains of each PPIX. Hydrogen bonding between these heme dimers accounted for the aggregate’s stable, extended crystalline network. Importantly, this work established that HZ and BH were crystallographically, chemically and spectroscopically identical (34).
Figure 4. Physical characterization of hemozoin. (a) The axial propionate linkages between heme units in HZ are seen by FT-IR fingerprints of C=O and C-O stretching at 1664 and 1211 cm-1, respectively. (b) Characteristic 2:1 peaks are seen at 7°: 21° and 24° 2θ for (i) native and (ii) synthetic HZ. Absent from these aggregates is the 23° 2θ peak observed in the diffraction pattern of (iii) substrate hemin chloride. (c) Electron micrograph of HZ in P. falciparum infected RBC where (i) is the host RBC, (ii) is the parasite, and (iii) is the DV. Image reproduced with kind permission of Springer Science and Business Media (23). (d) SEM image of uniform BH crystals.
Identification of native lipid components of hemozoin
In vivo native HZ transmission electron microscopy images from trophozoite-infected RBCs depicted the localization of HZ crystals within neutral lipid nanospheres (28). Trophozoite DVs were isolated using Percoll/sucrose bottom separation techniques, and the lipid content of these isolates was extracted by Bligh-Dyer (chloroform/methanol) lipid extraction. Methyl ester fatty acid characterization by gas chromatography-MS (GC-MS) and lithium (Li+) adduct electrospray ionization mass spectrometry of trophozoite fractions revealed a suite of neutral lipids adhered to the crystal’s surface. When these lipid extracts (a ratio of 4 monostearic-: 2 monopalmitic-: 1 dipalmitic-: 1 dioleic-: and 1 dilinoleic glycerol) were incubated in the presence of substrate hemin, HZ formation was observed (28). This competent nucleating template of fatty acyl glycerides is consistent with the hypothesis of a lipid scaffold for heme aggregation in P. falciparum.
Also extracted from native HZ were polar hydroxylated fatty acids derived presumably from cellular arachidonic and linoleic acids. Native HZ was purified from infected RBCs by a series of centrifugation steps followed by organic extraction of its lipid coat. Analysis of the lipid coat revealed the presence of hydroxylated polyunsaturated fatty acids. These polar lipids were separated using reverse phase-, normal phase-, and chiral phase high performance liquid chromatography and subsequently examined by GC-MS. Native HZ lipid coat GC-MS analysis revealed the presence of 15-, 12-, 11-, 9-, 8-, and 5- hydroxyeicosatetraenoic acids (HETEs) as well as 13- and 9- hydroxyoctadecadienoic acids (HODEs) (35).
Native HZ produced in the parasite DV most likely encounters cellular debris such as arachidonic and linoleic fatty acids or other lipids when released from a bursting RBC. Redox cycling of surface exposed heme units within HZ can lead to the initiation of lipid peroxidation (LPO). Abstraction of a bisallylic hydrogen atom from a fatty acid such as arachidonic acid results in an unpaired electron on the methylene carbon (36). This unpaired electron promotes the rearrangement of the double bonds adjacent to the methylene group that produces an alkyl radical (L●). In the presence of oxygen, alkyl radicals react to form peroxyl radicals (LOO●) that can abstract a hydrogen atom that yields a lipid peroxide (LOOH) and can propagate more reactions. Reduction of the peroxyl radical would result in the production of racemic mixtures of hydroxylated fatty acids like the HETEs and HODEs described previously in the HZ lipid coat. Secondary oxidation and chain β-cleavage results in reactive 4-hydroxy-2-nonenal (HNE) (Fig. 5).
Lipid peroxidation was demonstrated in vitro by reacting arachidonic acid with native HZ purified of all cellular lipid and protein content (37). Resultant HETE products were extracted from the reaction supernatant and identified using reverse phase high performance liquid chromatography (37). Recent studies of BH reactions with arachidonic acid revealed an identical reaction profile to that of HZ with all six positional HETE isomers identified by ultra high pressure reverse phase and chiral phase liquid chromatography tandem MS/MS (chiral-RP-LC-MS/MS) (38). LC-MS analysis has enabled the separation and identification of these cellular metabolites at femtomole levels of detection. The presence of these oxidation products in the lipid coat of native HZ and their formation upon arachidonic acid incubation with purified HZ and BH strongly supports the HZ-mediated LPO of cellular fatty acids. Additionally, the known immunomodulatory activity of these lipid peroxidation products is intriguing given the reported ability of HZ to disrupt the function of innate immune cells.
Figure 5. Hemozoin-mediated lipid peroxidation. Redox cycling of iron in complexes like HZ can initiate lipid peroxidation of fatty acids like arachidonic acid (LH). Abstraction of the bisallylic hydrogen leaves an unpaired electron on the methylene carbon that can rearrange to form a reactive alkyl radical (L●). Oxidation of this radical leads to a peroxyl radical (LOO●), and on reduction, forms a lipid peroxide (LOOH) that can undergo additional reactions to yield a variety of secondary oxidation products.
Basis of Hemozoin Immunomodulation
Once thought to be a simple “inert” detoxification biomineral, an increasing appreciation exists of HZ’s reactivity and subsequent perturbation of the host immune response. HZ reactivity in the modulation of innate and adaptive immunity has been attributed to a range of effects that include toll-like receptor 9 (TLR9) activation, cytokine production, LPO, and dendritic cell development. Recent studies found that the P. falciparum DNA that remained on native HZ extracts was the true source of TLR9 activation, and once native HZ was treated with nuclease, no activation was observed (39). HZ disruption of dendritic cell function was shown to cause a decline in T cell activation that led to a weakened adaptive immune response. The role of HZ as an immunoreactive aggregate may impact the high rates of patient secondary infection as well as the decline in vaccine efficacy.
Functionally, phagocytosis of HZ has been reported to impair macrophage oxidative burst, to downregulate iNOS activation, and to perturb cytokine profiles in infected patients. It has also been shown to correlate with increased levels of immunomodulatory LPO products such as prostaglandin E2 (PGE2), HNE, and HETEs (40) in monocytes. The biologic activity of these compounds is generally derived from either of two mechanisms. In the first mode of action, the reactive intermediates or products may form adducts to DNA or proteins. Thus, a variety of chain-terminating reactions of peroxyl, alkoxyl, or epoxyperoxyl radicals can result in oxidative cross-links to DNA or proteins. Furthermore, the electrophilic alkenals, such as HNE, readily form Schiff-base adducts to lysine residues and Michael addition adducts to histidine and cysteine residues to perturb the function of many proteins (41). In the second mode of action, the LPO products may act as alternate ligands to several different proteins and receptors to initiate an ultimately pathogenic signaling cascade. The hydroxylated fatty acids 9- and 13-HODE, as well as 15-HETE have all been found to be activators of the important nuclear receptor protein PPAR-γ, which is involved in key cellular regulatory and differentiation functions in monocytes (42). 15-HETE has also been found to stimulate RBC adherence to capillary endothelia, to enhance vascular permeability and edema, and to induce chemotaxis and chemokinesis, although the precise pathways remain unclear. From these examples, it is easy to imagine how such promiscuous reactivity often manifests itself in the pathogenesis of disease states. These findings suggest that HZ’s ability to rectify cellular responses may actually be caused by formation of primary and secondary LPO products (40).
To study this hypothesis, BH was incubated with ghost RBC membrane lipids, which are similar to those that the native biomineral would be exposed to upon RBC rupture (40). RAW macrophage-like cells were treated with the supernatant of this reaction, and reactive oxygen and nitrogen species (ROS and RNS) produced via the NADPH oxidase and iNOS pathways were inhibited. The levels of inhibition paralleled the effects of individual LPO products such as 15-HETE or HNE at pathologically reported concentrations. Additionally, treatment of cells with either BH or unreacted ghost supernatant did not cause a decrease in ROS and RNS production, which indicates that the products of HZ-mediated LPO were responsible for the observed disruption of macrophage function, not the dimeric heme component of the aggregate itself (40). Mechanisms for such malarial host-pathogen interactions and, more broadly, biomineral-to-cell relationships are primarily undefined, which creates a significant treatment barrier. Unraveling the formation and role of HZ in the pathogenesis of P. falciparum infection may provide additional insight into the prevention and treatment of a variety of diseases that result from pathogenic biominerals.
Summary
Biomineralization results in an expansive array of complex materials. These natural biominerals often represent unique crystal forms that extend over several size domains that are synthesized under ambient conditions. Increasingly, it is understood that biomineralization processes play important roles in the pathologies of several diseases. In malaria, the parasite forms the HZ biomineral in response to the heme released during hemoglobin catabolism. HZ serves an important detoxification role that allows the organism to maintain homeostasis during its intraerythrocytic phases. Despite the fact that HZ represents a validated drug target for P. falciparum, many fundamental questions remain concerning its formation; questions such as how or if the parasite assembles neutral lipid droplets specifically, how the heme is transported and deposited in these lipid domains, and whether new strategies exist that could be used to design drugs to disrupt this process. Tackling the rational drug design problem from a crystal engineering perspective offers an interesting direction for drug discovery, whether designing crystal-specific antibodies or fabricating selectively binding compounds. Consequently, studies on the in vivo formation of HZ afford an opportunity for chemists with a variety of interests (e.g., organic, inorganic, biologic, supramolecular, materials) to make significant contributions in attacking this neglected and devastating disease.
A second, emerging area of research is the pathophysiologic responses between biominerals and immune cells. A commonality between the diseases that maintain a pathogenic biomineral is the inflammatory response. Be it an auto immune reaction, as in the case of gout, or down regulation of the innate immune system as in malaria, it is clear that the interface between biominerals and cells is important in mitigating these responses. In the case of HZ, growing evidence suggests that the biomineral reacts with cellular fatty acids to produce a suite of reactive oxidized eicosanoids. The immunoreactivity of these oxidation products is likely a significant contributor to the inflammation and discomfort experienced by patients. Developing therapies for these effects and for the cause of infection could provide a dual approach in the treatment of many pathogenic biominerals and proffers the potential for novel initiatives in disease prevention and treatment.
References
1. Mann S. Biomineralization: principles and concepts in bioinorganic materials chemistry. In: Oxford Chemistry Masters. Compton RG, Davies SG, Evans J, eds. 2001. Oxford University Press, Oxford.
2. Nakata PA. Advances in our understanding of calcium oxalate crystal formation and function in plants. Plant Sci. 2003; 164:901-909.
3. De Yoreo JJ, Vekilov PG. Principles of crystal nucleation and growth. In: Reviews in Mineralogy and Geochemistry. Dove PM, De Yoreo JJ, Weiner S, eds. 2003. The Mineralogical Society of America, Washington, DC. pp. 57-90.
4. Malaria. Fact Sheet N°94. World Health Organization. 2007.
5. Ziegler J, Linck R, Wright DW. Heme aggregation inhibitors: antimalarial drugs targeting an essential biomineralization process. Curr. Med. Chem. 2001; 8:171-189.
6. Bonilla JA, Moura PA, Bonilla TD, Yowell CA, Fidock DA, Dame JB. Effects on growth, hemoglobin metabolism and paralogous gene expression resulting from disruption of genes encoding the digestive vacuole plasmepsins of Plasmodium falciparum. Int. J. Parasitol. 2007; 37:317-327.
7. Haas L. Neurological stamp. Charles Louis Alphonse Laveran (1845-1922). J. Neurol. Neurosurg. Psych. 1999; 67:520.
8. Scheibel L, Sherman I. In: Malaria: Principles and Practice of Malariology. Wernsdorfer WH, McGregor I, eds. 1988. Churchill-Livingstone, Edinburgh.
9. Knoll AH. Biomineralization and evolutionary history. In: Reviews in Mineralogy and Geochemistry. Dove PM, De Yoreo JJ, Weiner S, eds. 2003. The Mineralogical Society of America, Washington DC. pp. 329-350.
10. Rodriguez-Navarro AB, Cabraldmelo C, Batista N, Morimoto N, Alvarez-Lloret P, Ortega-Huertas M, Fuenzalida VM, Arias JI, Wiff JP, Arias JL. Microstructure and crystallographic-texture of giant barnacle (Austromegabalanus psittacus) shell. J. Struct. Biol. 2006; 156:355-362.
11. He G, Ramachandran A, Dahl T, George S, Schultz D, Cookson D, Veis A, George A. Phosphorylation of phosphophoryn is crucial for its function as a mediator of biomineralization. J. Biol. Chem. 2005; 280:33109-33114.
12. Bozec L, de Groot J, Odlyha M, Nicholls B, Nesbitt S, Flanagan A, Horton M. Atomic force microscopy of collagen structure in bone and dentine revealed by osteoclastic resorption. Ultramicroscopy. 2005; 105:79-89.
13. Perry CC. Silification: the processes by which organisms capture and mineralize silica. In: Reviews in Mineralogy and Geochemistry. Dove PM., De Yoreo JJ, Weiner S, eds. 2003. The Mineralogical Society of America, Washington DC. pp. 304-323.
14. Nie S, Emory Sr. Probing single molecules and single nanoparticles by surface-enhanced raman scattering. Science 1997; 275:1102-1106.
15. Lowenstam HA. Minerals formed by organisms. Science 1981; 211:1126-1131.
16. Mann S. Molecular recognition in biomineralization. Nature 1988; 332:119-124.
17. Gibbs J. On the equilibrium of heterogenous substances. Trans. Connect. Acad. Sci. 1878; 16:343-524.
18. Hohenberg PC, Halperin BI. Theory of dynamic critical phenomena. Rev. Mod. Phys. 1977; 49:435-475.
19. Mann S. Molecular tectonics in biomineralization and biomimetic materials chemistry. Nature 1993; 365:499-505.
20. Sullivan DJ. Theories on malarial pigment formation and quinoline action. Int J Parasitol. 2002; 32:1645-1653.
21. Ziegler J, Change R, Wright D. Multiple-antigenic peptides of histidine-rich protein II of Plasmodium falciparum: den- drimeric biomineralization template. J. Am. Chem. Soc. 1999; 121:2395-2400.
22. Egan T, Mavuso WW, Ncokazi KK. The mechanism of beta-hematin formation in acetate solution. Parallels between hemozoin formation and biomineralization processes. Biochemistry 2001; 40:204-213.
23. Carney CK, Harry SR, Sewell SL, Wright DW. Detoxification biominerals. In: Biomineralization I: Crystallization and Self-Organization Process. Naka, Kensuke, eds. 2007. Springer Verlag, Berlin. pp. 155-185.
24. Schneider EL, Marietta MA. Heme binding to the histidine-rich protein II from Plasmodium falciparum. Biochemistry 2005; 44:979-986.
25. Fitch C, Cai G-z, Chen Y-F, Shoemaker JD. Involvement of lipids in ferriprotoporphyrin IX polymerization in malaria. Biochim. Biophys. Acta. 1999; 1454:31-37.
26. Hemplemann E, Motta C, Hughes R, Ward S, Bray P. Plasmodium falciparum: sacrificing membrane to grow crystals? Trends Parasitol. 2003; 19:23-26.
27. Egan TJ, Chen JY-J, de Villiers KA, Mabotha TE, Naidoo KJ, Ncokazi KK, Langford SJ, McNaughton D, Pandiancherri S, Wood BR. Haemozoin (beta-haematin) biomineralization occurs by self-assembly near the lipid/water interface. FEBS Lett. 2006; 580:5105-5110.
28. Pisciotta JM, Coppens I, Tripathi AK, Scholl PF, Shuman J, Bajad S, Shulaev V, Sullivan D Jr. The role of neutral lipid nanospheres in Plasmodium falciparumhaem crystallization. Biochem. J. 2007; 402:197-204.
29. Kuman S, Guha M, Choubey V, Maity P, Bandyopadhyay U. Antimalarial drugs inhibiting hemozoin (beta-hematin) formation: a mechanistic update. Life Sci. 2007; 80:813-828.
30. Solomonov I, Osipova M, Feldman Y, Baehtz C, Kjaer K, Robinson IK, Webster GT, McNaughton D, Wood BR, Weissbuch I, Leiserowitz L. Crystal nucleation, growth and morphology of the synthetic malaria pigment beta-hematin and the effect thereone by quinoline additives: the malaria pigment as a target of various antimalarial drugs. J. Am. Chem. Soc. 2007; 129:2615-2627.
31. Ridley R. Haemozoin formation in malaria parasites: is there a haem polymerase? Trends Microbiol. 1996; 4:253.
32. Slater A, Swiggard W, Orton B, Flitter W, Goldberg D, Cerami A, Henderson G. An iron-carboxylate bond links the heme units of malaria pigment. Proc. Natl. Acad. Sci. U.S.A. 1991; 88:325-329.
33. Noland G, Briones N, Sullivan D Jr. The shape and size of hemozoin crystals distinguishes diverse Plasmodium species. Mol. Biochem. Parasitol. 2003; 130:91-99.
34. Pagola S, Stephens P, Bohle D, Kosar A, Madsen S. The structure of malaria pigment beta-hematin. Nature 2000; 404:307-310.
35. Schwarzer E, Kuhn H, Valente E, Arese P. Malaria-parasitized erythrocytes and hemozoin nonenzymatically generate large amounts of hydroxy fatty acids that inhibit monocyte functions. Blood 2003; 101:722-728.
36. Cheng Z, Li Y. What is responsible for the initiating chemistry of iron-mediated lipid peroxidation: an update. Chem. Rev. 2007; 107:748-766.
37. Green M, Xiao L, Lal A. Formation of hydroxyeicosatetraenoic acids from hemozoin-catalyzed oxidation of arachidonic acid. Mol. Biochem. Parasitol. 1996; 83:183-188.
38. Carter MD, Harry SR, Wright DW. Identification of hydroxyeicosatetraenoic acid components of schistosomal hemozoin. Biochem. Biophys. Res. Commun. 2007; 363:867-872.
39. Parroche P, Lauw FN, Goutagny N, Latz E, Monks BG, Visintin A, Halmen KA, Lamphier M, Olivier M, Bartholomeu DC, Gazzinelli RT, Golenbock DT. Malaria hemozoin is immunologically inert but radically enhances innate responses by presenting malaria DNA to Toll-like receptor 9. Proc. Natl. Acad. Sci. U.S.A. 2007; 104:1919-1924.
40. Carney CK, Schrimpe AC, Halfpenny K, Harry SR, Miller CM, Broncel M, Sewell SL, Schaff JE, Deol R, Carter MD, Wright DW. The basis of the immunomodulatory activity of malaria pigment (hemozoin). J. Biol. Inorg. Chem. 2006; 11:917-929.
41. Fenaille F, Guy PA. Study of protein modification by 4-hydroxy-2-nonenal and other short chain aldehydes analyzed by electrospray ionization tandem mass spectrometry. J. Am. Soc. Mass Spectrom. 2003; 14:215-226.
42. Skorokhod OA, Alessio M, Mordmuller B, Arese P, Schwarzer E. Hemozoin (malarial pigment) inhibits differentiation and maturation of human monocyte-derived dendritic cells: a peroxisome proliferator-activated receptor-gamma-mediated effect. J. Immunol. 2004; 173:4066-4074.
Further Reading
Carney CK, Pasierb L, Wright DW. Heme detoxification in malaria: a target rich environment. In: Medicinal Inorganic Chemistry. Sessler JL, Doctrow SR, McMurry TJ, Lippard SJ, eds. 2005. American Chemical Society, Washington, DC. pp. 263-280.
Lippard SJ, Berg JM. Principles of Bioinorganic Chemistry. 1994. University Science Books, Mill Valley, CA.
Marks F, Furstenberger G. Prostaglandins, Leukotrienes and Other Eicosanoids. 1999. Wiley-VCH, Weinheim.
Sherman IW. Malaria: Parasite Biology, Pathogenesis, and Protection. 1998. ASM Press, Washington, DC.
See Also
Lipids, Chemical Diversity of
Hemes in Biology
Metal Complexes, Assembly of
Redox Regulation and Signaling: Reactive Oxygen Species (ROS)
Interactions of Biomaterials with Cells and Tissues