Journal of Inorganic Biochemistry 104 (2010) 1119–1124 Contents lists available at ScienceDirect Journal of Inorganic Biochemistry j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j i n o r g b i o On the kinetics of formation of hemozoin, the malaria pigment Robert F. Pasternack a,⁎, Ben Munda a, Abigail Bickford b, Esther J. Gibbs b, Luigi Monsù Scolaro c a b c Department of Chemistry & Biochemistry, Swarthmore College Swarthmore, PA 19081, USA Department of Chemistry, Goucher College, Towson, MD 21204, USA Dipartimento di Chimica Inorganica, Chimica Analitica e Chimica Fisica, C.I.R.C.M.S.B., Università di Messina, Italy a r t i c l e i n f o Article history: Received 24 February 2010 Received in revised form 30 June 2010 Accepted 30 June 2010 Available online 29 July 2010 a b s t r a c t We report on the kinetics of formation of hemozoin as a function of hemin concentration and reaction medium. Evidence is presented for the critical role played by interfacial regions in the efficient conversion of hemin to the malaria pigment. © 2010 Elsevier Inc. All rights reserved. Keywords: Malaria Hemozoin Chloroquine Heme Hemin 1. Introduction Despite years of intensive study malaria remains one of mankind's most dreaded scourges with annual infections and deaths counted in the hundreds of millions [1]. A critical factor in the expression of the disease is a pathway by which the Plasmodium parasite protects itself from the presence of extra-protein heme. As the parasite feeds on the protein portion of hemoglobin, large amounts of toxic heme are released. The heme is rapidly oxidized to hemin (iron in the + 3 oxidation state) which, in a series of steps, is converted to and stored as an insoluble, non-toxic, largely unreactive assembly called hemozoin (“malaria pigment”) [2,3]. It is generally believed that quinine and similar drugs owe their efficacy to interference with hemozoin formation [4], but some considerable uncertainty remains as to the molecular details of this inhibition [5]. Although studied for nearly three centuries, the composition and structure of hemozoin were elucidated only within the last twenty years [6,7]; and the mechanism of its formation remains an open question. The basic unit of the assembly is a reciprocal dimer in which a propionate group of one hemin occupies an axial position on the iron of a neighboring hemin, and vice versa [6]. The iron(III) centers have been shown to be high-spin [8]. Whereas hemin is known to aggregate extensively in aqueous solution, the “normal” dimer involves either stacking interactions or the formation of a μ-oxo linkage between iron centers [9–11]. Some controversy exists as to ⁎ Corresponding author. Department of Chemistry and Biochemistry, Swarthmore College, 500 College Avenue, Swarthmore, PA 19081, USA. Tel.: + 1 610 3288559; fax: + 1 610 3287355. E-mail address: rpaster1@swarthmore.edu (R.F. Pasternack). 0162-0134/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2010.07.001 which of these forms is prevalent under the conditions of interest here (pH ~ 5, vide infra); recent evidence indicates that the dimer structure is highly dependent on pH and solvent composition [9,10,12,13]. What is unequivocal is that under these pH conditions, hemin is sparingly soluble in aqueous media and precipitates as α-hematin aggregates. This material is readily converted back to a soluble form by modifying the pH or by treatment with any of a number of reagents including DMSO, bicarbonate or pyridine. In contrast, the insoluble aggregate formed by Plasmodium based on a reciprocal dimer is resistant to attack by most reagents (including those listed above) and to changes in pH over an extended range. This so-called β-hematin aggregate or hemozoin involves hydrogen bonding between adjacent dimer units for which the non-liganded propionate groupings play a key role [7]. It is in this form that Plasmodium stores released heme [14]. The formation of hemozoin occurs in a food vacuole produced by the parasite for protease “digestion” of hemoglobin [15]. The optimum pH for this protein degradation activity – the pH at which hemozoin is formed – is about five, and the half-life for the conversion of free hemin into hemozoin has been estimated as in the order of minutes [16]. To simulate in vivo conditions, most laboratory studies on the kinetics of hemozoin formation are carried out in the presence of acetate or citrate buffer to maintain the physiologically relevant pH. In one such protocol [17], hemin in DMSO was added to an aqueous solution containing acetate and a water-soluble alcohol. After 16 h the reaction mixture was tested for the presence of hemozoin. A correlation was found between the yield of the malaria pigment and the lipophilicity of the dissolved alcohol. A second protocol to study hemozoin formation, developed by Egan and coworkers [4,18], uses a closed system reactor in which a large amount of hemin (75 mg) is placed in about 25 mL of 4.5 M 1120 R.F. Pasternack et al. / Journal of Inorganic Biochemistry 104 (2010) 1119–1124 acetate buffer and maintained at 60 °C (the reaction is extremely slow at physiological temperature). Under these conditions hemin rapidly aggregates and precipitates in the α-hematin form. Samples are withdrawn at intervals over the course of many hours and are cooled on ice, dried for 48 h and eventually studied via infrared spectroscopy to determine the extent of conversion to β-hematin. This protocol has been critiqued by its originators who pointed out the non-correspondence of the conditions (especially temperature and method of hemin delivery) to those found in vivo [16]. In still another protocol developed by this group a small sample of a hemin stock in aqueous acetone is injected at the interface of a citrate containing aqueous/ organic mixture [16]. Using this approach, efficient conversion to hemozoin was observed in the presence of several lipids and two alcohols (1-pentanol and 1-octanol). We describe below an alternative, complementary method for studying the kinetics of hemozoin formation, which we believe has several advantages; it is more flexible than previous protocols and lends itself to systematic kinetic studies. We compare the results obtained using this new protocol to those already reported, and discuss factors that appear vital for rapid, efficient conversion of hemin into hemozoin. 2. Experimental 2.1. Materials and methods Fluka and Frontier Scientific hemin were used for these experiments. Chloroquine (ChlQ) diphosphate was a MP Biomedicals product, and L-α-Dioleoylphosphatidyl-ethanolamine (DOPE) (N99% purity) was supplied by Avanti Polar Lipids, Inc. (Alabama). Other more standard chemicals were obtained from Fisher Scientific and Aldrich Chemicals and met ACS standards for high purity. Reaction vials (8 mL) fitted with teflon caps were used as reaction vessels. These were maintained at a constant temperature using an Isotemp 202 S bath supplied by Fisher Scientific. The vortex devices employed were a Thermolyne MaxiMix II and a Scientific Industry Vortex-2 Genie. Centrifugations were performed on a Sorvall Instruments RC5C fitted with a SS-34 head. Spectroscopic measurements were conducted on a JASCO V-560 spectrophotometer. of the lipid is added to this stock. This mixture is monophasic for all but one of the co-solvents we have tested and report on below (heptane). In the heptane case, the organic mixture and acetate solutions were added separately to the reaction vials. In the standard protocol, 250 μL of Stock D is added to each of a selected number of screw-top reaction vials (usually 15–20) that are then equilibrated to 37 °C. (iii) To initiate the reaction, 750 μL of the thermally equilibrated hemin (Stock C) is added to each vial. The reaction mixture is inverted twice and immediately returned to the constant temperature bath. In an earlier study XRD, FTIR and Resonance Raman spectroscopies were used to show that hemin produces hemozoin in such media [16,21]. Under these conditions, depending on the co-solvent (or presence of a lipid), the hemin either rapidly precipitates (as α-hematin) and/or is selectively taken up in the co-solvent that has formed a separate phase upon the addition of the aqueous hemin stock. The co-solvent represents about 2% of the total reaction volume, and the recorded pH* of the reaction mixture was 5.3. (iv) At selected times, reaction vials are removed from the bath and plunged into ice. As quickly as possible, 5 mL of a solvent system containing 50% acetone/aqueous pyridine (5%) buffered with 26 mM HEPES is added. The resulting mixture is vortexed for 60 s to ensure that the pyridine has encountered all the hemin regardless of its state. At this point, all hemin that is not part of a hemozoin assembly is converted to soluble hemin (py)2. This assay solution was shown – through application of X-ray diffraction and infrared spectroscopy – not to react with hemozoin but to readily dissolve hemin and α-hematin and convert them to the low-spin dipyridine hemin complex [21]. (v) The mixture is centrifuged for 30 min at 5000 rpm to separate the insoluble hemozoin which settles to the bottom of the vial. The supernatant is then tested spectrophotometrically (λmax at 406 nm) for the presence of hemin as the dipyridine adduct. The data is plotted (Figs. 1–5) as the percentage of hemin converted to hemozoin (% conversion) vs. time. 2.2. Protocol A new protocol used for studying the kinetics of hemozoin formation involves several steps: (i) Prepare a concentrated solution (~ 10–20 mM) of hemin in DMSO (Stock A). Hemin is very soluble in this solvent and exists as a monomer [19]. To keep the DMSO concentration below 0.5% in the final reaction mixture, it proves convenient to prepare Stock B in aqueous 1 mM phosphate buffer (final pH ~ 10) by a dilution of Stock A by 20 fold. The concentration of Stock B is determined spectrophotometrically using a sample diluted by a factor of one hundred with a 0.1 M NaOH solution. The molar absorptivity of hemin in this medium has been reported as ε = 5.84 × 104 M− 1 cm− 1 [20]. Finally, a large volume (~20 mL) of a hemin reaction solution (Stock C) in 1 mM phosphate buffer (pH ~ 10) is prepared from Stock B such that hemin in the reaction vials will have the desired final concentration (50 μM for most of these experiments). Stock C, which is monophasic, is equilibrated at 37o in a constant temperature bath. When diluted by a factor of one-third (vide infra), DMSO is less than 0.5% by volume in the reaction mixture. (ii) Another stock solution (Stock D) is prepared which typically contains 9 mL methanol, 1 mL of co-solvent, and 2.5 mL of 1 M acetate buffer, pH ~ 5. For lipid experiments, a weighed quantity Fig. 1. Kinetic profiles for the conversion of hemin into hemozoin for several co-solvents listed in Table 1. Although the co-solvents represent only about 2% of the reaction volume, they can have a profound impact on the kinetics of conversion. The curves shown represent the best fit of the data obtained by applying Eq. (1). For 1-hexanol, k = 0.14 min− 1, n = 1.3, ks = 0.0054 min− 1; while for CCl4, k = 0.012 min− 1 and n = 0.20. The formation of hemozoin in a heptane/methanol/aqueous medium represents less than 10% conversion over the time window studied (~200 min). R.F. Pasternack et al. / Journal of Inorganic Biochemistry 104 (2010) 1119–1124 Fig. 2. Comparison of the kinetics of hemozoin production in the presence of several straight chain alcohols. Under the conditions of the experiments (9:1 methanol to cosolvent) both 1-hexanol and 1-octanol form a separate phase when aqueous hemin is added while 1-pentanol does not. 2.3. Kinetic analysis The kinetic data obtained for a number of the systems studied here are biphasic (cf. Fig. 1, 1-hexanol) having a major absorbance change that is quite rapid (comparable to the half-life estimated for in vivo formation of hemozoin), and a smaller, slower effect whose half-life is measured in hours. The data can be fit as two parallel processes in which the primary change is represented by a “stretched exponential” (Avrami equation [22]) and the slower process (for which less precise data is available) by a simple exponential. The fit to the data for 1hexanol in Fig. 1, as well as several other co-solvents, was obtained using the working equation:   n + 1 y = ð1−A∞ = Ao Þ100−y1 exp −ðkt Þ = n + 1 −y2 expð−ks t Þ ð1Þ where “y” is the total % Conversion of hemin to hemozoin at the time, t, as calculated from absorbance measurements at 406 nm for the dipyridine hemin complex, and “y1” and “y2” are the contributions made to the conversion via the more rapid and slower processes. A∞ and Ao represent the final and initial absorbances at 406, respectively. After 24 h, the absorbance at 406 nm approaches zero for the more active solvent systems investigated. Fig. 3. Study of the impact of an aqueous phase on the conversion of hemin to hemozoin in the presence of 1-octanol as co-solvent. 1121 Fig. 4. Effect of added lipid (DOPE) on the kinetics of hemozoin formation. The methanol/chloroform/aqueous system is among the least active in producing hemozoin, in contrast to its reactivity with lipid added. In a kinetic model developed earlier for autocatalytic processes [23,24], a key step involves the rate-determining formation of a “seed” or “nucleus” catalyzed by a (fractal) surface. In the limiting case in which the “seed” size is one (that is, the rate-determining step involves a solute species undergoing a transformation rather than the formation of a multi-molecular seed), the resulting integrated rate law has a “stretched exponential” functional form as in the first term of Eq. (1). We use this semi-empirical approach to compare and contrast reaction systems differing in co-solvent, added lipid and/or added chloroquine. For the results shown in Fig. 1, we obtain: for 1-hexanol as co-solvent, k = 0.14 min− 1, n = 1.3, and ks = 0.0054 min− 1; for carbon tetrachloride as co-solvent, k = 0.012 min− 1 and n = 0.20. To facilitate comparisons of kinetic results obtained under a variety of conditions, we introduce estimates of the initial half-life (t1/2′) for the stretched exponential function (t1/2′ = {[0.693 (n + 1)]^1/(n + 1)}/k), and for the slow (simple exponential) process (t1/2 s = 0.693/ks). These calculations are based on the approximation that the two kinetic processes are sufficiently different in rate that they can be effectively decoupled. For the results shown in Fig. 1, t1/2′ = 8.7 min and t1/2 s = 2.1 h for 1-hexanol; t1/2′ = 71 min for CCl4. Fig. 5. Influence of added chloroquine (ChlQ) on the formation of hemozoin with 1octanol as co-solvent. The ChlQ has a marked inhibitory effect, which scales with its concentration. 1122 R.F. Pasternack et al. / Journal of Inorganic Biochemistry 104 (2010) 1119–1124 3. Results Table 2 Summary of kinetic resultsa. 3.1. Survey of co-solvents System Experiments were conducted using the protocol described above that differed from one another in the identity of the co-solvent — which comprises about 2% by volume of the final reaction mixture. Reagents were selected having various functional groups and lipophilicities, and the results were used as a method of screening the several organic solvents for their impact on hemozoin production. The criterion that was applied was the percentage conversion after 100 min of reaction time. A summary of the results is shown in Table 1. Fuller kinetic studies were conducted for those co-solvents that are particularly active. One of the active alcohols (1-octanol) was selected as co-solvent to probe the impact on the kinetics of hemin and DMSO concentration. Three experiments were conducted: [hemin]= 50 μM, DMSO ~ 0.3% by volume; [hemin]= 50 μM, DMSO ~ 5% and [hemin]= 400 μM, DMSO ~5%. The kinetic profiles obtained for these experiments are nearly indistinguishable with k = 0.16 min− 1, n = 0.83; ks = 0.0066 min− 1; k = 0.16 min− 1, n = 0.93, ks = 0.0060 min− 1; and k = 0.17 min− 1, n = 0.79, ks = 0.0082 min− 1, respectively (Table 2). Kinetic profiles for three straight chain alcohols, 1-octanol, 1hexanol and 1-pentanol, are shown in Fig. 2. Both the octanol and hexanol systems exhibit phase separation upon addition of the aqueous hemin solution (Step (iii) of the Protocol) and the hemin is concentrated in the now separated co-solvent volume (although some color remains in the major aqueous/methanol phase). However, 1-pentanol under these conditions shows no such phase separation. Note how different the 1-pentanol kinetic profile is from the other alcohols. Values of t1/2′ are 110 min for 1-pentanol and only 7–9 min for the other two alcohols. If the experimental conditions are modified somewhat (Step ii) so that the organic mixture is now 6:4 methanol to 1-pentanol, an organic phase does separate on aqueous hemin addition, and the 1-pentanol profile is now very similar to the other alcohols. Under these conditions of 6:4 methanol to alcohol, t1/2′ for 1-octanol and 1-pentanol are 12 and 10 min, respectively. The importance of interface regions for rapid kinetics is further illustrated by an experiment in which Stock D (methanol with 1octanol) was prepared so that it did not contain acetate buffer. The Step (iii) addition was thus divided into two parts. First, 50 μL of an aqueous solution of 1 M acetate buffer was added to 750 μL hemin Stock C in each room temperature reaction vial, leading to rapid precipitation of α-hematin at pH ~ 5. After several minutes, these mixtures were centrifuged and the near-colorless supernatant was discarded. Next, some 200 μL of the pre-equilibrated organic mixture, 9:1 methanol/1-octanol, was added to the solid material remaining in Table 1 Comparison of organic co-solvents. a Co-solvent % Conversion in 100 min logPa Acetone Aniline Carbon tetrachloride Chloroform Cyclohexanol Cyclopentanone Dichloromethane Heptane 1-Hexanol Nitromethane 1-Octanol 1-Pentanol b5 b5 ~ 30 b5 b5 b5 b5 b5 ~ 85 b5 ~ 80 ~ 20 −0.157 0.936 2.861 1.755 1.341 0.198 1.187 4.475 1.938 −0.195 3.001 1.407 logP are partition values obtained by Advanced Chemistry Development. The larger the value, the more lipophilic is the solvent. i) 50 μM hemin 9:1 methanol/CCl4 9:1 methanol/1-hexanol (Figs. 1 and 2) 9:1 methanol/1-octanol (Fig. 2) 9:1 methanol/1-octanol (5% DMSO) 6:4 methanol/1-octanol 9:1 methanol/1-pentanol (Fig. 2) 6:4 methanol/1-pentanol 9:1 methanol/chloroform (DOPE added) (Fig. 4) 9:1 methanol/1-octanol (0.1 mM chloroquine, Fig. 5) ii) 400 μM hemin 9:1 methanol/1-octanol (5% DMSO) n k (min− 1) t1/2′ (min) t1/2 s (h) 0.20 1.3 0.012 0.14 71 8.7 – 2.1 0.83 0.16 7.1 1.8 0.93 0.16 7.3 1.9 3.2 11 0.11 0.011 12 110 1.6 – 1.4 0.33 0.13 0.30 10 3.1 0.83 1.3 9.1 0.022 55 2.2 0.79 0.17 6.6 1.4 a All systems are 0.3–0.5% DMSO unless otherwise specified. the vial, and the mixture was vortexed and maintained at 37 °C. The usual detection procedure was then followed — vials were removed periodically, the reaction was quenched on ice, and buffered pyridine solution was added, followed by vortexing, centrifugation and spectroscopic determination of conversion to hemozoin. The results are shown in Fig. 3; the conversion to hemozoin in the absence of an aqueous phase proceeds very slowly. The data in the absence of an aqueous phase is fit as a first-order reaction with a rate constant of 0.0088 min− 1. 3.2. Experiments with a lipid (DOPE) It is widely agreed that lipids play a critical role in the formation of hemozoin in vivo [25–28]. Lipid structures have been observed within the Plasmodium food vacuole in which hemozoin is completely embedded (so-called “lipid nanospheres”). These lipid droplets putatively catalyze the formation of hemozoin by concentrating the extra-protein hemin and excluding water to facilitate the formation of the reciprocal dimer. In the current study we considered the influence of DOPE on hemozoin production by modifying the present protocol so that solid lipid was dissolved in Stock D (0.50 mg/mL). When aqueous hemin is added (Step iii), the mixture becomes extremely turbid as DOPE separates out of solution. The lipid has a remarkable effect when CHCl3 was used as the co-solvent (Fig. 4) giving results (k = 0.30 min− 1; n = 0.33, ks = 0.0088 min− 1) that are rapid compared even to 1-hexanol or 1octanol as co-solvent; in the absence of the lipid, CHCl3 produces practically no hemozoin (Table 1 and Fig. 4). By way of contrast, the addition of DOPE to the methanol/1-octanol system has little effect on the rate of hemozoin formation. 3.3. Chloroquine experiments Finally, some preliminary experiments were conducted in which chloroquine was incubated with hemin before addition to a methanol/ 1-octanol/acetate buffer solution. The results are summarized in Fig. 5. The droplet of 1-octanol which forms on aqueous hemin addition is intensely colored as in the absence of chloroquine. Yet, the inhibition of hemozoin formation is virtually complete at a chloroquine concentration of 1 mM, but is only partial at [ChlQ]= 0.1 mM. Under these latter conditions, k = 0.022 min− 1, n = 9.1, ks = 0.0052 min− 1. These effects due to ChlQ are not the result of pH changes; pH* = 5.3 for all three reaction mixtures. R.F. Pasternack et al. / Journal of Inorganic Biochemistry 104 (2010) 1119–1124 4. Discussion Results described here parallel observations reported for the formation of hemozoin in the intact biological system. At a pH of about 5 at 37 °C, the half-life for the primary pathway to hemozoin formation is of the order of a few minutes, lipids greatly enhance the rate of the process and chloroquine interferes in a concentration-dependent manner. Although conversion of hemin/α-hematin into a hemozoin assembly is fostered by a lipophilic environment, the dependence is not straightforward. Both the most and least lipophilic co-solvents studied produce very little hemozoin over the time frame of interest (Table 1). Our results point to an important role played by interface regions between aqueous and non-aqueous phases. The slow kinetics observed in a non-aqueous methanol/1-octanol environment (Fig. 3) is a case in point. The reaction profile in this medium is fit as first-order with a halflife of nearly one and a half hours. For several other systems we observe such slow (usually minor) processes, which we interpret as arising from conversion to hemozoin within a single phase rather than at the interface. Recall that in the method of Huy et al. [17] hemin, delivered in DMSO to an aqueous phase containing a water-soluble alcohol, produces hemozoin (as determined after 16 h) with an efficiency of conversion that correlates with the lipophilicity of the alcohol. Preferential solvation of hemin may be an important factor in this procedure. (Such preferential solvation of water-soluble metalloporphyrins by organic solvents has been previously reported [29]). One can envision a de facto separation in which hemin is partially screened from water by the alcohol; the more lipophilic the alcohol, the better is the screening. In a slightly modified Huy procedure, when non-water-soluble, longer chain alcohols are employed, the conversion is very much more rapid. Further insight into the importance of interface regions emerges from a comparison of results obtained here to ones reported by Egan et al. [16]. The Egan procedure involves the layering of an organic solvent on an aqueous phase (cf. Introduction). For 1-octanol, results obtained with our protocol are remarkably similar to those reported by Egan et al. Using a first-order rate law, they obtained a value for k of 0.13 min− 1, while we obtain 0.16 min− 1 (Fig. 2). However, in the solvent separated method of Egan et al., 1-pentanol was found to be very effective at producing hemozoin from aqueous acetone-containing hemin, giving results quite similar to 1-octanol. In contrast, we find using our protocol that 1-pentanol as co-solvent (regardless of whether we combine it with methanol or acetone) leads to a kinetic profile which involves an extended incubation period; quite different from that obtained for 1octanol (Fig. 2). This difference in behavior, we suggest, is the consequence of 9:1 methanol:1-pentanol and acetone:1-pentanol media not separating into distinct phases when aqueous hemin is added; the liquids are miscible under these conditions. We repeated this study using other ratios of methanol to 1-pentanol, 1:9; 4:6; 6:4. In all these three cases, 1-pentanol forms a discrete phase when aqueous hemin is added, and in all three cases hemozoin is produced at rates comparable to that observed with 1-octanol, t1/2′ = 10 and 12 min, respectively in a 6:4 mixture. Taken together, these results provide direct evidence for the importance of an interface region for rapid hemozoin formation. However, notice should be taken that the existence of an interface is not sufficient to insure efficient conversion of hemin into hemozoin. Heptane forms an interface with aqueous media but it is among the least effective co-solvents at producing hemozoin. The role of neutral lipids in producing hemozoin is illustrated by the impact of added DOPE to a methanol/CHCl3 solvent system. The methanol/CHCl3 system is among the least efficient of those studied at producing hemozoin, but when the lipid is added the rate of conversion to hemozoin is comparable to or even faster than that observed for methanol/1-hexanol or methanol/1-octanol (cf. Table 2). We believe that structured interface regions serve as reaction sites that are critical for rapid hemozoin formation. A recent study [30] using Transmission Electron Microscopy has shown that 1123 emulsions form near a lipid/aqueous interface and that hemozoin is produced preferentially at the surface of the emulsion droplets. Even in the presence of such interface reaction sites, chloroquine is capable of inhibiting the formation of hemozoin. It is apparent in the methanol/1-octanol system where direct visual detection is possible, that hemin is still concentrated in the lipophilic phase in the presence of chloroquine but virtually no reaction is observed at 1 mM chloroquine. A lower concentration of chloroquine effectively blocks formation of hemozoin for a period of time, but then an efficient conversion occurs (Fig. 5). The profile suggests a mechanism in which chloroquine interferes with the formation of microcrystallites which serve as templates for hemozoin crystallization. Once such microcrystallites form, hemozoin crystals are rapidly produced. Various step-wise chemical models have been proposed to account for the stretched exponential kinetic profile employed in this study. However, in most of these a concentration dependence is predicted [31], but not observed in the present system. These models have in common an autocatalytic process in which each molecule of product serves as a catalytic site. In an approach previously suggested by us for DNA-bound porphyrin assembly [23,24], the catalyst is a (fractal) surface rather than individual product molecules. The resulting integrated rate law shows no concentration dependence for the case in which the rate-determining step involves the transformation of a solute species. We propose that the key step in the mechanism of hemozoin formation occurs at the aqueous/non-aqueous interface where hemin exists as a dimer, and that the interface helps promote the conversion of the standard dimer to the reciprocal dimer. Once formed, other dimers attack the reciprocal dimer, are transformed, and add to it (via hydrogen bonding) to form the hemozoin crystal. Hemin exists as a monomer in organic layers and α-hematin assemblies in aqueous phase (pH ~ 5) — and both of these forms are slow to convert to reciprocal dimer assemblies. Thus, one way to affect the kinetics of the process is by modifying the interface region rather than changing the hemin concentration. It has been reported that the rate of hemozoin production in stirred solutions depends on the rate of stirring [18] — such stirring may well have an impact on the properties of the interface region. That the existence and properties of the interface have such a profound effect on the kinetics of hemozoin production leads to a consideration of an alternative mechanism for chloroquine activity. Previous workers have suggested that chloroquine acts by binding either to monomers or dimers of hemin or arises from a capping of hemozoin microcrystals [4,5]. However, it is possible that chloroquine and similar drugs show activity as a result of their impact on the aqueous/non-aqueous (lipid) interface so as to inhibit reaction. Further studies on this suggested inhibition mechanism are currently underway. Abbreviations ChlQ Chloroquine DOPE L-α-Dioleoylphosphatidyl-ethanolamine pH* measured pH in a mixed solvent medium t1/2',t1/2 s estimated first half-life for the stretched exponential kinetic process and half-life for the slower, first-order process HEPES 2-[4-(2-hydroxyethyl)-2,3,5,6-tetrahydropyrazin-1-yl] ethanesulfonate. Acknowledgement We are grateful to Dr. Giovanna De Luca for aiding in the preparation of this manuscript. References [1] World Health Organization, World Malaria Report, WHO, Geneva, 2008. [2] M.F. Oliveira, B. Timm, E.A. Machado, K. Miranda, M. Attias, J.R. Silva, M. DansaPetretski, 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. 1124 [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] R.F. Pasternack et al. / Journal of Inorganic Biochemistry 104 (2010) 1119–1124 C.D. Fitch, P. Kanjananggulpan, J. Biol. Chem. 262 (1987) 15552–15555. T.J. Egan, D.C. Ross, P.A. Adams, FEBS Lett. 352 (1994) 54–57. I. Weissbuch, L. Leiserowitz, Chem. Rev. 108 (2008) 4899–4914. A.F.G. Slater, W.J. Swiggard, B.R. Orten, W.D. Flitter, D.E. Goldberg, A. Cerami, G.B. Henderson, Proc. Natl Acad. Sci. USA 88 (1991) 325–329. S. Pagola, P.W. Stephens, D.S. Bohle, A.D. Kosar, S.K. Madsen, Nature 404 (2000) 307–310. B.R. Wood, S.J. Langford, B.M. Cooke, J. Lim, F.K. Glenister, M. Duriska, J.K. Unthank, D. McNaughton, J. Am. Chem. Soc. 126 (2004) 9233–9239. K.A. deVilliers, C.H. Kaschula, T.J. Egan, H.M. Marques, J. Biol. Inorg. Chem. 12 (2007) 101–117. C. Asher, K.A. deVilliers, T.J. Egan, Inorg. Chem. 48 (2009) 7994–8003. S.B. Brown, T.C. Dean, P. Jones, Biochem. J. 117 (1970) 733–739. L.B. Casabianca, D. An, J.K. Natarajan, J.N. Alumasa, P.D. Roepe, C. Wolf, A.C. de Dios, Inorg. Chem. 47 (2008) 6077–6081. L.B. Casabianca, J.B. Kallgren, J.K. Natarajan, J.N. Alumasa, P.D. Roepe, C. Wolf, A.C. de Dios, J. Inorg. Biochem. 103 (2009) 745–748. T.J. Egan, J.M. Combrinck, J. Egan, G.R. Hearne, H.H. Marques, S.N. Lenteni, B.T. Sewell, P.J. Smith, D. Taylor, D.A. van Schalkwyk, J.C. Walden, Biochem. J. 365 (2002) 343–347. D. Goldberg, A.F.G. Slater, A. Cerami, G.B. Henderson, Proc. Natl Acad. Sci. USA 87 (1990) 2931–2935. T.J. Egan, J.Y.-J. Chen, K.A. deVilliers, 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. [17] N.T. Huy, A. Maeda, D.T. Uyen, D.T.X. Trang, M. Sasai, T. Shino, T. Oida, S. harada, K. Kamei, Acta Trop. 101 (2007) 130–138. [18] T.J. Egan, W.M. Mavuso, K.K. Ncokazi, Biochemistry 40 (2001) 204–213. [19] G.S. Collier, J.M. Pratt, C.R. De Wet, C.F. Tshabalala, Biochem. J. 179 (1979) 281–289. [20] N. Shaklai, Y. Shviro, E. Rabizadeh, I. Kirschner-Zilber, Biochim. Biophys. Acta 821 (1985) 355–366. [21] K.K. Ncokazi, T.J. Egan, Anal. Biochem. 338 (2005) 306–319. [22] M. Avrami, J. Chem. Phys. 7 (1939) 1103–1112. [23] R.F. Pasternack, E.J. Gibbs, P.J. Collings, J.C. dePaula, L.C. Turzo, A. Terracina, J. Am. Chem. Soc. 120 (1998) 5873–5878. [24] R.F. Pasternack, E.J. Gibbs, D. Bruzewicz, D. Stewart, K.S. Engstrom, J. Am. Chem. Soc. 124 (2002) 3533–3539. [25] K.E. Jackson, N. Klonis, D.J.P. Ferguson, A. Adisa, C. Dogovski, L. Tilley, Mol. Microbiol. 54 (2004) 109–122. [26] K. Bendrat, B.J. Berger, A. Cerami, Nature 378 (1995) 138–139. [27] J.M. Pisciotta, I. Coppens, A.K. Tripathi, P.F. Scholl, J. Shuman, S. Bajad, V. Shulaev, D.J. Sullivan Jr., Biochem. J. 402 (2007) 197–204. [28] J.M. Pisciotta, D. Sullivan, Parasitol. Int. 57 (2008) 89–96. [29] R.F. Pasternack, E.G. Spiro, M. Teach, J. Inorg. Nucl. Chem. 36 (1974) 599–606. [30] A.N. Hoang, K.K. Ncokazi, K.A. de Villiers, D.W. Wright, T.J. Egan, Dalton Trans. 39 (2010) 1235–1244. [31] E.E. Finney, R.G. Finke, Chem. Mater. 21 (2009) 4692–4705.