Hemozoin Induces Lung Inflammation and Correlates ... - ATS Journals

Hemozoin Induces Lung Inflammation and Correlates with Malaria-Associated Acute Respiratory Distress Syndrome Katrien Deroost1, Ariane Tyberghein1, Natacha Lays1, Sam Noppen2, Evelin Schwarzer4, Els Vanstreels2, Mina Komuta3, Mauro Prato4,5, Jing-Wen Lin6, Ana Pamplona7, Chris J. Janse6, Paolo Arese4, Tania Roskams3, Dirk Daelemans2, Ghislain Opdenakker1, and Philippe E. Van den Steen1 1

Laboratory of Immunobiology, and 2Laboratory of Virology and Chemotherapy, Department of Microbiology and Immunology, Rega Institute for Medical Research, and 3Translational Cell and Tissue Research, KU Leuven–University of Leuven, Leuven, Belgium; 4Department of Genetics, Biology, and Biochemistry, and 5Department of Neuroscience, University of Turin, Turin, Italy; 6Leiden Malaria Research Group, Department of Parasitology, Leiden University Medical Center, Leiden, The Netherlands; and 7Unidade de Imunologia Molecular, Instituto de Medicina Molecular, Faculdade de Medicina, Universidade de Lisboa, Lisbon, Portugal

Malaria-associated acute respiratory distress syndrome (MA-ARDS) is a deadly complication of malaria, and its pathophysiology is insufficiently understood. Both in humans and in murine models, MA-ARDS is characterized by marked pulmonary inflammation. We investigated the role of hemozoin in MA-ARDS in C57Bl/6 mice infected with Plasmodium berghei NK65, P. berghei ANKA, and P. chabaudi AS. By quantifying hemozoin in the lungs and measuring the disease parameters of MA-ARDS, we demonstrated a highly significant correlation between pulmonary hemozoin concentrations, lung weights, and alveolar edema. Histological analysis of the lungs demonstrated that hemozoin is localized in phagocytes and infected erythrocytes, and only occasionally in granulocytes. Species-specific differences in hemozoin production, as measured among individual schizonts, were associated with variations in pulmonary pathogenicity. Furthermore, both pulmonary hemozoin and lung pathology were correlated with the number of infiltrating inflammatory cells, an increased pulmonary expression of cytokines, chemokines, and enzymes, and concentrations of alveolar vascular endothelial growth factor. The causal relationship between hemozoin and inflammation was investigated by injecting P. falciparum–derived hemozoin intravenously into malaria-free mice. Hemozoin potently induced the pulmonary expression of proinflammatory chemokines (interferon-g inducible protein–10/CXC-chemokine ligand (CXCL)10, monocyte chemotactic protein–1/CC-chemokine ligand 2, and keratinocyte-derived chemokine/CXCL1), cytokines (IL-1b, IL-6, IL-10, TNF, and transforming growth factor–b), and other inflammatory mediators (inducible nitric oxide synthase, heme oxygenase–1, nicotinamide adenine dinucleotide phosphate– oxidase–2, and intercellular adhesion molecule–1). Thus, hemozoin correlates with MA-ARDS and induces pulmonary inflammation.

(Received in original form November 7, 2012 and in final form December 30, 2012) This study was supported by the Agency for Innovation by Science and Technology (Agentschap voor Innovatie door Wetenschap en Technologie [IWT]), by Geconcerteerde Onderzoeks Acties grants GOA 2,012/017 and GOA 2,013/ 014 of the Research Fund of KU Leuven–University of Leuven, by the Fund for Scientific Research in Flanders (Fonds voor Wetenschappelijk Onderzoek [FWO]– Vlaanderen), and by EviMalar (P.A. and E.S.). K.D. is funded by a Ph.D. grant of the IWT, P.V.d.S. is a research professor at KU Leuven–University of Leuven, and J.-W.L. is supported by the China Scholarship Council–Leiden University Joint Program. Correspondence and requests for reprints should be addressed to Philippe E. Van den Steen, Ph.D., Laboratory of Immunobiology, Department of Microbiology and Immunology, Rega Institute for Medical Research, KU Leuven–University of Leuven, Leuven 3000, Belgium. E-mail: [email protected] This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org Am J Respir Cell Mol Biol Vol 48, Iss. 5, pp 589–600, May 2013 Copyright ª 2013 by the American Thoracic Society Originally Published in Press as DOI: 10.1165/rcmb.2012-0450OC on January 17, 2013 Internet address: www.atsjournals.org

CLINICAL RELEVANCE Malaria-associated acute respiratory distress syndrome is an often lethal complication of malaria. This study, using a murine model with significant similarity to human pathology, shows that hemozoin is associated with several disease parameters, and that it induces inflammation in the lungs.

Keywords: malaria; hemozoin; ARDS; pulmonary inflammation; Plasmodium berghei NK65

In 2010, malaria caused an estimated 655,000 deaths, and was the fifth highest cause of death in low-income countries, accounting for 5.2% of all deaths (World Malaria Report 2011, World Health Organization). Most of these deaths were attributable to life-threatening complications such as cerebral malaria, severe malarial anemia, respiratory problems, and pregnancy-associated malaria, for which no successful adjuvant therapy is available (1). Acute lung injury (ALI) and its more severe form, acute respiratory distress syndrome (ARDS), are major life-threatening complications of noncardiogenic and nonmetabolic etiologies (2). ALI and ARDS mostly develop secondary to one or more predisposing conditions such as sepsis, severe burns, trauma, and infections of different etiologies, and are often the result of a systemic inflammatory response with systemic leukocyte activation. Malaria-associated (MA) ALI/ARDS attributable to Plasmodium falciparum occurs as a single complication, or is accompanied by additional disturbances, leading to multiorgan dysfunction (3, 4). Furthermore, it is the most frequent complication in P. knowlesi infections (5, 6), and may even develop during or after antimalarial treatment, as is increasingly reported in P. vivax malaria (7, 8). MA-ARDS is characterized by acute pulmonary inflammation and increased capillary endothelial and alveolar epithelial permeability, leading to interstitial and alveolar edema and hyalinemembrane formation, resulting in ventilation–perfusion mismatch and impairment of gas exchange (3, 4). The pathogenesis and pathophysiology of MA-ARDS are incompletely understood. By screening various combinations of murine and parasite strains, we previously demonstrated that C57BL/6J mice infected with P. berghei NK65 (PbNK65) succumbed to MA-ARDS (9). Lungs from infected mice had a brown-grayish discoloration because of hemorrhages and hemozoin (Hz) deposition, in addition to significantly increased lung weights and massive edema. A pronounced inflammatory response with leukocyte infiltration was observed in the affected lungs, and this strikingly resembled the histopathological findings in postmortem studies of patients with MA-ARDS (10, 11).

590

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL 48 2013

During the digestion of hemoglobin during parasite maturation, toxic heme is released inside the food vacuole (12). Before the heme can damage the parasite, it is detoxified through biocrystallization into Hz. These crystals, surrounded by host- and parasite-derived lipids and proteins, are released into the circulation at schizont rupture, and are rapidly removed by phagocytosis. The number of pigment-containing leukocytes in the peripheral blood correlates with disease severity in P. falciparum–infected patients (13, 14), and several inflammatory and immunomodulating effects of Hz were found in vitro, as reviewed elsewhere (15–17). Injections of b-hematin or synthetic Hz induced inflammation in air pouches and in the liver (18). Furthermore, we previously showed that mice infected with lethal P. berghei parasites had significantly more Hz than mice infected with the nonlethal P. chabaudi AS (PcAS) parasite (19). Because Hz has inflammatory properties and a strong pulmonary inflammatory response is induced in MA-ARDS, we investigated the role of Hz in a murine model of MA-ARDS. The aims of this study were (1) to analyze pulmonary Hz concentrations with different parasite species, and to correlate Hz with the disease parameters of MA-ARDS; (2) to compare Hz production and degradation in the host between the different parasite species; (3) to determine correlations between lung Hz, local inflammatory mediators (including cytokines, chemokines, and enzymes), and pathology; (4) to demonstrate the role of Hz in lung inflammation by the intravenous injection of P. falciparum–derived Hz (PfHz) in malaria-free mice; and (5) to investigate the effects of Hz on pulmonary cytokine and chemokine expression. Our results indicate that Hz induces pulmonary inflammation, and is associated with MA-ARDS in mice.

MATERIALS AND METHODS Additional information about methods can be found in the online supplement.

Infection and Dissection of Mice and Determination of Hz C57BL/6J mice were infected with 104 PcAS, PbNK65, or P. berghei ANKA (PbANKA). Where indicated, antimalarial treatment involved chloroquine and pyrimethamine. At the indicated times, mice were killed and their left lungs were pinched off. Subsequently, bronchoalveolar lavage (BAL) was performed on the right lungs, after which the mice were systemically perfused. The Hz content in lavaged and perfused lungs was determined by chemoluminescence, as previously described (19). All experiments were approved by the local ethics committee (License LA121251, Belgium).

Quantification of Hz in Individual Schizonts The ex vivo cultivation of P. berghei or PcAS schizonts was performed as described previously (20, 21). Schizont-enriched cell suspensions were incubated with carboxyfluorescein diacetate succinimidyl ester (CFSE) and 3,3-dioctadecyloxycarbocyanine perchlorate (DiO), fixed with 0.37% formaldehyde, smeared on glass slides, and mounted with ProLong Gold antifade reagent containing 49,6-diamidino-2phenylindole (DAPI). Hz in individual schizonts was quantified by densitometry upon imaging by transmitted light and confocal fluorescence microscopy. Alternatively, schizonts were imaged by reflection contrast microscopy (RCM).

Histological Analysis Paraffin-embedded sections were stained with hematoxylin and eosin or by immunohistochemistry with monoclonal anti-mouse F4/80 IgG2b. Visualization was performed via reaction with 3,39-diaminobenzidine, which produces a brown color in the presence of peroxide.

Quantitative Reverse Transcription-Polymerase Chain Reaction and Determination of Vascular Endothelial Growth Factor Protein After mechanical homogenization of the left lungs, total RNA was extracted and quantified, cDNA was synthesized, and quantitative PCR was performed on 25 ng and 12.5 ng cDNA with primer and probe sets from Applied Biosystems (Foster City, CA) or Integrated DNA Technologies (Leuven, Belgium). Data were normalized to 18S ribosomal RNA concentrations. Alveolar vascular endothelial growth factor-A (VEGF-A) concentrations were determined in bronchoalveolar lavage (BAL) fluids, using an ELISA kit according to the manufacturer’s instructions (1/2 dilution; R&D Systems Euope Ltd, Abingdon, UK).

Flow Cytometry At the indicated time points after infection, mice were bled by heart puncture, BAL fluid from both lungs was collected, and mice were systemically perfused. A portion of lung tissue was mechanically homogenized and passed through a 70-mm nylon cell strainer. The resulting cell suspensions were centrifuged, and the pellets were resuspended in 40% Percoll solution (GE Healthcare, Uppsala, Sweden) and layered on a 72% Percoll solution. After centrifugation, leukocytes at the interphase between both Percoll layers were collected. Subsequently, the remaining erythrocytes were lysed and the leukocytes were washed, enumerated with a Bürker-type hemacytometer (Schrek, Goldbach, Germany), and used for flow cytometric analysis.

Intravenous Injection of Pf Hz PfHz was prepared from cultured P. falciparum–infected erythrocytes (iRBCs) and treated with DNase. Contamination with endotoxin (LPS) was excluded via the E-Toxate assay (Sigma, St. Louis, MO). Different amounts of PfHz (100–900 nmol) in 200 ml of PBS or PBS alone were injected intravenously. After 6 hours, the mice were killed and dissected as already described. Right lungs were used for PfHz quantification, and left lungs were used for quantitative RT-PCR analysis.

Statistical Analysis GraphPad Prism software (La Jolla, CA) was used to calculate P values for the differences between two groups (Mann-Whitney U-test), nonparametric Spearman correlation coefficients, and values for areas under the curve (AUCs). Multiple linear regression analysis was performed with IBM SPSS software (IBM Corp., Armonk, NY). P , 0.05 was considered statistically significant.

RESULTS Pulmonary Hz Concentrations Are Correlated with Disease Parameters of MA-ARDS in Mice

Previously, we found more Hz in organs, including lungs, of P. berghei–infected C57BL/6J mice than in those of PcAS-infected mice (19). In this study, we investigated whether Hz plays a role in the pathogenesis of MA-ARDS. Pulmonary Hz concentrations increased progressively during infection with PbANKA, PbNK65, and PcAS. At the same time after infection and at similar peripheral parasitemia levels, the amount of Hz was comparable in lungs from PbANKA-infected and PbNK65infected mice, whereas lungs from PcAS-infected mice contained significantly less Hz (Figure 1A). On Day 10, when all PbANKA-infected mice had already succumbed to cerebral pathology, the Hz concentrations in PbNK65-infected lungs further increased. Lung weights (Figure 1B) and the protein content of BAL fluids (Figure 1C) were also significantly higher in PbANKAinfected and PbNK65-infected lungs, compared with PcAS-infected lungs, in line with previous findings (9). We found a highly significant correlation between pulmonary Hz concentrations, lung weights, and alveolar edema (Figures 1D–1F). Importantly, this

Deroost, Tyberghein, Lays, et al.: Inflammatory Role of Hemozoin in MA-ARDS

591

Figure 1. The disease parameters of malaria-associated acute respiratory distress syndrome (MA-ARDS) are correlated with pulmonary hemozoin (Hz) levels. C57BL/6J mice were infected with Plasmodium berghei NK65 (PbNK65, shaded circles), P. berghei ANKA (PbANKA, open circles), or P. chabaudi AS (PcAS, solid circles), and killed at different time points after infection. PbANKA-infected mice die around Day 8, and thus no data were available on Day 10. (A) Hz concentrations in perfused right lungs. (B) Lung weights of nonperfused left lungs. (C) Protein concentrations in bronchoalveolar lavage (BAL) fluid. (D–F) Spearman correlations between pulmonary Hz amounts, lung weights, and alveolar edema. (G) Spearman r and P values and adjusted R2 and P values between different groups. Data are derived from two separate experiments. Each dot represents the result from an individual mouse. Horizontal bars between individual data points represent group medians, and horizontal lines below asterisks in panels A–C indicate statistical differences between groups, adjusted for peripheral parasitemia, sex, day after infection, and parasite strain. Asterisks above individual datasets indicate statistical differences with the uninfected control group. *P , 0.05. **P , 0.01. ***P , 0.001.

relationship remained significant in a multiple regression model after adjusting for sex, day after infection, peripheral parasitemia, and the parasite strain used (Figure 1G), indicating that pulmonary Hz accumulation and pathology are tightly associated in C57BL/6J mice. Differences in Pulmonary Hz Content Are Not Attributable to Hz Degradation

To investigate whether strain-specific differences in pulmonary Hz concentrations were attributable to differential Hz degradation, we blocked parasite replication and Hz synthesis. Mice were infected with PbNK65 or PcAS and treated for 6 days with a combination of chloroquine diphosphate and pyrimethamine, starting 7 to 8 days after infection. At different time points, Hz

concentrations were quantified in perfused lungs, livers, and spleens, because these organs contain most of the Hz (. 95%) (19). The total Hz content in both PbNK65-infected and PcAS-infected mice did not significantly change after Day 30 (Figure 2), indicating that after an initial increase during the treatment by which parasites were cleared from the circulation, the total quantity of Hz remained stable. Hz concentrations decreased in lungs and increased in livers and spleens, suggesting a redistribution between organs. These data are in full agreement with previous studies by Levesque and colleagues (22) with P. chabaudi adami and P. yoelii 17X, and by Frita and colleagues (23) with PbNK65, and they indicate that the higher amounts of pulmonary Hz with PbNK65 were not attributable to differences in Hz degradation.

592


Figure 2. Hz is not degraded after therapeutic parasite clearance. (A–H) C57BL/6J mice were infected with PbNK65 (A, C, E, and G) or PcAS (B, D, F, and H), and antimalarial treatment was administered for 6 days, starting 7–8 days after infection, to clear parasites from the circulation. Mice were killed at different time points after infection (7–8, 20, 30, 60, and 120 days). The amount of Hz was determined in lungs (A and B), livers (C and D), and spleens (E and F). The organ-specific Hz content was calculated by multiplying the Hz quantity/mg tissue with the organ weight, and was divided by the peripheral parasitemia on the day of antimalarial treatment to account for variations in the number of Hz-producing parasites. All measured Hz values were above the background value of the organs, as determined for organs from uninfected control mice. (G and H) The total amount of Hz was calculated by summing the Hz contents of the lungs, liver, and spleen for each mouse separately. Data are derived from two separate experiments. Each dot represents the result from an individual mouse. Horizontal bars between individual data points represent group medians, and horizontal lines below asterisks indicate statistical differences between groups. *P , 0.05. **P , 0.01. ***P , 0.001.

P. berghei Schizonts Contain Significantly More Hz than Do PcAS Schizonts

To determine whether differences in Hz production may form the basis of the observed differences in Hz load among the organs of mice infected with different Plasmodium species, Hz was quantified in schizonts (i.e., the erythrocytic stage during which the Hz content is maximal). Because mature-stage iRBCs sequester in the microvasculature and only a few schizonts can be observed in the peripheral circulation (Figure 3A), schizont-

enriched cultures were obtained by ex vivo cultivation of peripheral blood from malaria-infected mice. Because the rupture of PcAS schizonts occurs at night, mice were housed on a reversed day–night cycle, and the optimal time frame to obtain blood for the ex vivo culture was determined (see Figure E1A in the online supplement). Figure 3B shows schizonts from the different parasites obtained from schizont-enriched cultures and stained with fluorescently labeled dyes to visualize the nuclei (Figure 3B, blue) and the cytoplasm and membranes (Figure 3B, green) of merozoites. Because schizonts from the ex vivo culture


593

Figure 3. P. berghei schizonts contain significantly more Hz than do PcAS schizonts. (A) Giemsa-stained schizonts from tail-vein blood of malariainfected mice. (B) Schizonts from schizont-enriched cultures after ex vivo cultivation at 378 C. Upper row: Images were taken by confocal microscopy after staining the cytoplasm and lipids (green) and the nuclei (blue). Lower row: Bright field images of the same schizonts. (C) Hz crystals in schizonts were observed under reflection contrast microscopy. Upper row: Images taken with polarized light. Lower row: Bright field images of the same schizonts. (D) Relative Hz quantity/schizont was obtained via the combined confocal and transmitted light microscopy technique. (E) Relative Hz quantity/schizont was obtained with reflection contrast microscopy under polarized light. In the lower graphs of D and E, data are stratified according to the number of nuclei per schizont. Data are derived from two separate experiments. Each dot represents the value from an individual schizont. PbNK65, shaded circles; PbANKA, open circles; PcAS, solid circles. Horizontal bars between individual data points represent group medians, and horizontal lines below asterisks indicate statistical differences between groups. *P , 0.05. **P , 0.01. ***P , 0.001. ****P , 0.0001. Additional data about the optimal time point chosen to obtain blood for the ex vivo culture of PcAS parasites, and how we calculated the relative Hz quantity/schizont, are shown in Figure E1 of the online supplement. #, numbers of.

demonstrated a morphology and number of merozoites comparable to those of schizonts observed in the peripheral circulation, the maturation process of ex vivo cultured parasites appeared to be normal. When sufficient schizonts were present in the cultures, the amount of Hz was quantified in more than 100 individual schizonts with segregated merozoites by a combination of confocal and transmitted light microscopy. A z-stack was created throughout the cell, and Hz was quantified in the complete parasite by measuring Hz at different depths of the z-stack (Figure E1B). The area under the curve (AUC) was calculated to estimate the relative Hz quantity in the entire schizonts. The results demonstrate that P. berghei schizonts (both PbNK65 and PbANKA) contained significantly more Hz than did PcAS schizonts (Figure 3D). Furthermore, a significant difference was found in the amount of Hz/cell between schizonts of the two P. berghei strains. To investigate whether these differences were attributable to variations in schizont maturation, the Hz quantification data were stratified according to the number of nuclei per cell. Schizonts with 5 to 8 nuclei were the most prevalent, and were detected in cultures from all three parasite strains. Smaller schizonts with less than five nuclei were

mostly observed in PcAS cultures, and schizonts with more than eight nuclei were typical of P. berghei cultures. When schizonts with similar numbers of nuclei were compared, P. berghei schizonts still contained significantly more Hz than did PcAS schizonts. More Hz was also detected in PbNK65 schizonts than in PbANKA schizonts, but only in the schizonts with 5 to 8 nuclei. This occurred because in this group, more PbNK65 schizonts contained eight nuclei compared with PbANKA schizonts (58% of PbNK65 schizonts, compared with 22% of PbANKA schizonts). When larger schizonts were compared, similar quantities were found for the two parasite strains. To confirm the differences in Hz content between schizonts from different parasites, we quantified the Hz content in schizonts from the same cultures, using RCM. Because of the birefringent nature of Hz (24) and a polarization block consisting of two crossed polarizers, the entire Hz crystal could be visualized as shown in Figure 3C. Densitometry was applied to quantify the crystal on polarized light images of individual cells, and the relative densitometric value/cell was used to compare the data between the different parasite strains. This technique confirmed the significantly higher amounts of Hz found in P. berghei

594


To investigate the cellular location of pulmonary Hz, paraffin sections from perfused lungs were stained with hematoxylin and eosin. Most of the Hz was observed in lungs from PbNK65-infected mice 10 days after infection, corroborating the Hz quantification data (Figure 1A). Intravascular marginating leukocytes (Figure 4A, arrows) were found with all three parasite strains, but the amount of inflammatory cells containing Hz was highest in lungs from P. berghei–infected mice. Cytoadherent iRBCs (Figure 4A, arrowheads) were frequently observed in PbANKA-infected and PbNK65-infected lungs, and only rarely in lungs from PcASinfected mice. Most iRBCs were trophozoites, which were recognizable because of the Hz dispersed throughout the cell, but fully segmented schizonts with a central Hz crystal were also frequently found. Furthermore, small vessels in the interstitium of P. berghei–infected lungs were congested with both uninfected and Hz-containing iRBCs, whereas this pattern was not observed in lungs from control or PcAS-infected mice. To investigate the nature of Hz-containing cells, immunohistochemistry was performed for the monocyte/macrophage-specific antigen F4/80. Hz (Figure 4B, dark brown crystals) was mostly situated inside F4/801 cells (Figure 4B, brown staining) and inside cytoadherent iRBCs, and only occasionally inside granulocytes.

LT-a (Figure 5 and Table E2). In a multiple regression model, the relationship between Hz/mg lung tissue and IL-6, IL-10, MCP-1, and MHC-II remained statistically significant and was independent of sex, peripheral parasitemia, parasite strain, and the number of days after infection. The relationship with IL-4 and Hmox1 was also statistically significant, but was dependent on the level of peripheral parasitemia. In addition, the association between Hz/mg tissue and alveolar edema was independent of peripheral parasitemia, sex, parasite strain, and number of days after infection, and was dependent on IL-1b, IL-4, IL-6, LT-a, IP-10, and MHC-II. A similar relationship was found between Hz/mg lung tissue and lung weights, and this relationship was dependent on TNF, LT-a, and VEGF. VEGF is important in the pathogenesis of PbANKA-induced MA-ALI in DBA/2 mice (25). Although VEGF is a strong inducer of vascular permeability, and high concentrations of VEGF are produced in the lungs (26), a down-regulated transcription was evident in lung homogenates of P. berghei–infected C57BL/6J mice (Table E1). Because increased vascular and epithelial permeability with interstitial and alveolar edema is a prominent feature of our MA-ARDS model, and because large numbers of phagocytes accumulating in the lungs may be a source of VEGF protein, alveolar VEGF protein concentrations were determined in BAL fluids. A significant increase in BAL VEGF concentrations was observed with both P. berghei strains, but not with PcAS (Figure 5I). Furthermore, a strong correlation existed between VEGF protein concentrations, pulmonary Hz concentrations, and the degree of alveolar edema (Figures 5I–5J). These observations indicate that pulmonary Hz levels, inflammatory cytokine and chemokine expression, the release of VEGF in the alveoli, and pulmonary edema are tightly associated with each other.

Hz, Inflammation, and Pathology Are Tightly Associated in MA-ARDS

Inflammatory Cell Recruitment to the Lungs Is Associated with Hz

Previously we found that the transcription of a number of cytokines and chemokines was more strongly induced in the lungs of mice with PbNK65-induced MA-ARDS than in mice infected with PcAS (9). Here, we extend this analysis to a broader range of inflammatory mediators, and also include a comparison with PbANKA-infected mice. The pulmonary mRNA expression of various chemokines (interferon-g inducible protein–10 [IP-10]/ CXC-chemokine ligand 10 [CXCL10], monocyte chemotactic protein–1 [MCP-1]/CC-chemokine ligand 2 [CCL2], and keratinocyte-derived chemokine [KC/CXCL1]), cytokines (IL1b, IL-4, IL-10, TNF, and transforming growth factor–b [TGFb]), and other inflammatory mediators (inducible nitric oxide synthase [iNOS], nicotinamide adenine dinucleotide phosphate oxidase–2 [NOX2], and heme oxygenase–1 [Hmox1]) was significantly increased, whereas the mRNAs of major histocompatibility complex class II (MHC-II) and vascular endothelial growth factor (VEGF) were significantly decreased in C57BL/6J mice infected with P. berghei parasites (which cause MA-ALI/ ARDS), compared with C57BL/6J mice infected with PcAS. The transcription of IL-6, IL-12, lymphotoxin-a (LT-a), intercellular adhesion molecule–1 (ICAM-1), matrix metalloproteinase-9/ gelatinase B (MMP-9), perforin, and Fas ligand (FasL) was not significantly different between strains, although the transcription was significantly different from that in uninfected control mice (Table E1). The association between Hz, lung pathology, and the induction of these mediators was determined by Spearman correlation analysis. A positive correlation with both the amount of Hz/mg lung tissue and lung pathology was found for the mRNA expression of IL-6, IL-10, TNF, MCP-1, IP-10, and Hmox1, whereas a negative correlation was found for MHC-II, VEGF, and

Inflammatory cells, including Hz-containing phagocytes, are abundantly present in the lungs of mice with MA-ARDS, as shown by histological analysis of lung tissue (Figure 4). Therefore, the relationship between Hz and leukocytes was analyzed by flow cytometry. Leukocytes were isolated from the lungs and stained with cell-specific markers to identify different cell subsets. Strong correlations were found between pulmonary Hz amounts (Figures 6A– 6H) or protein concentrations in BAL fluid (Figure 6I) and the absolute number of monocytes/macrophages (F4/801CD11b1), inflammatory monocytes (F4/801CD11b1Gr-11), neutrophils (F4/802CD11b1Gr-11), CXCR31CD41 T cells, and CXCR31 CD81 T cells in the lungs. Total cell numbers and CD81 cells correlated with pulmonary Hz but not with the amount of edema, whereas no correlation was found between Hz and CD41 cells.

schizonts compared with PcAS schizonts, and no differences were detected between schizonts of the PbANKA and PbNK65 strains (Figure 3E). Together, these results indicate that parasites inducing MA-ALI/ARDS (P. berghei) produce greater amounts of Hz than parasites that do not induce this pathology (PcAS). Pulmonary Hz Is Located in Cytoadherent iRBCs and Infiltrating Monocytes/Macrophages

Injection of Naive Mice with Pf Hz Mimics the Inflammatory Profile Observed during MA-ARDS

To confirm the causal relationship between Hz and pulmonary inflammation, PfHz was injected intravenously into malariafree C57BL/6J mice at doses between 100 and 900 nmol/mouse. These doses are in a range similar to the total amounts of Hz (z 250 nmol) found in PbNK65-infected mice 10 days after infection (19). Mice were killed 6 hours after injection, and the distribution of PfHz was investigated in perfused lungs, livers, and spleens. The amount of PfHz retrieved in the different organs varied between experiments and between PfHz batches. Most of the PfHz (80–90%) was trapped in the liver, whereas only around 10% was found in the spleen (Figure 7A). Less than 5% of the PfHz was detected in the lungs, which is comparable to the Hz distribution


595

Figure 4. Hz in the lungs is located inside cytoadherent infected erythrocytes (iRBCs) and monocytes/macrophages. (A) Paraffin sections were prepared from lungs of uninfected mice (Con) and from mice infected with PbANKA (Day 8), PbNK65 (Day 10), or PcAS (Day 10), and stained with hematoxylin and eosin. (B) Sections were stained with anti-F4/80 monoclonal antibodies (brown) to identify monocytes/macrophages. Hz is visible as dark-brown crystals. Representative images are shown (original magnification, 340; scale bars, 10 mm). Insets show magnifications of Hz-containing monocytes/ macrophages (arrows), or iRBCs containing Hz (arrowheads) (scale bars, 10 mm).

in infected mice (Figure 2). Furthermore, circulating phagocytes had taken up PfHz (Figure 7B). BAL VEGF concentrations were not increased after PfHz injection, indicating that PfHz does not directly induce VEGF in the lungs at the time point studied (Figure 7C). To investigate the effects of PfHz in the lungs, inflammation was investigated at the transcription level. Several chemokines (IP-10, MCP-1,

and KC), cytokines (IL-1b, IL-6, IL-10, TNF, and TGF-b), and other inflammatory mediators (iNOS, Hmox1, NOX2, and ICAM-1) were significantly induced (a greater than twofold increase) in the lungs by PfHz, whereas the transcription of other inflammatory mediators (LT-a, MHC-II, FasL, perforin, MMP-9, and VEGF) was not significantly altered or was undetectable (IL-4 and IL-12) at the time

596


Figure 5. Hz is associated with pulmonary inflammation. (A–H) Mice were infected with PbANKA (open circles), PbNK65 (shaded circles), or PcAS (solid circles), and pulmonary mRNA expression levels were determined at different time points after infection (Days 6, 8, and 10) by quantitative RT-PCR, and normalized to 18S ribosomal RNA concentrations. Spearman correlations were determined between pulmonary mRNA concentrations of TNF (A and B), monocyte chemotactic protein–1 (MCP-1) (C and D), IL-10 (E and F), or vascular endothelial growth factor (VEGF) (G and H) and the amount of Hz/mg lung tissue (A, C, E, and G) or the protein content in BAL fluid (B, D, F, and H). (I and J) Spearman correlations between BAL VEGF concentrations and the amount of Hz/mg lung tissue (I) or BAL protein content (J). Data are derived from two separate experiments. Spearman r and P values are indicated for each graph. Additional expression and correlation data are provided in Tables E1 and E2 of the online supplement.

point studied. This was particularly clear when stratifying the data according to pulmonary PfHz concentrations (Figures 7C–7I and Table E3). These results demonstrate that PfHz injection induces the pulmonary expression of several cytokines and chemokines that are associated with pulmonary Hz concentrations and alveolar edema in mice with MA-ARDS.

DISCUSSION By documenting strong associative data between Hz, inflammation, and pathology, and by demonstrating that PfHz, when injected intravenously into malaria-free mice, induces pulmonary inflammation, we demonstrated that Hz is a prominent inflammatory factor in the pathogenesis of MA-ARDS in


597

Figure 6. Hz is associated with inflammatory cell numbers in the lungs. (A–I) C57BL/6J mice were infected with PbNK65 (shaded circles) or PcAS (solid circles), and 10 days after infection, pulmonary leukocyte numbers were analyzed by flow cytometry. Spearman correlations were determined between total cells (A), monocytes/macrophages (F4/801CD11b1 cells; B), inflammatory monocytes (F4/801CD11b1Gr-11 cells; C), neutrophils (F4/802CD11b1Gr-11 cells; D), CD41 cells (E), CXCR31CD41 T cells (F), CD81 cells (G), or CXCR31CD81 T cells (H) and the total amount of Hz in the lungs (A–H) or the protein content in BAL fluid (I). Data are derived from two separate experiments. Absolute cell numbers are shown. Spearman r and P values are indicated for each comparison.

mice. An analogous pathogenic mechanism operates in sepsisassociated ARDS, in which bacteria in the circulation produce bioactive molecules such as peptidoglycan and LPS, which can be sufficient to cause pulmonary inflammation and ARDS, even if these bacteria do not specifically accumulate in the lungs (27). Parasite sequestration is a central feature in P. falciparum– induced complications, particularly in cerebral pathology (28). The role of inflammatory cells in cerebral malaria, although crucial in the PbANKA murine model, is still heavily debated, in part because only limited inflammatory infiltrates are present in the brains of patients with cerebral malaria (29). The presence of abundant inflammatory cells in the lungs of patients with MA-ARDS, in addition to sequestering parasites, indicates that inflammation may play a more prominent role in the pathogenesis of MA-ARDS than in cerebral malaria (3, 8). Furthermore, MA-ARDS is often induced after the initiation of antimalarial treatment, indicating that excessive inflammation induced by massive killing of iRBCs in the pulmonary microvasculature may be responsible for the induction of MA-ARDS.

The murine MA-ARDS model used in this study bears strong resemblance to the human disease, because both iRBC sequestration and inflammation are pathogenic features observed in the lungs. The same leukocyte subtypes accumulate in the lungs of patients with MA-ARDS and in our murine model, and consist mainly of macrophages (including Hz-containing macrophages), lymphocytes, and a limited number of neutrophils (4, 9, 10). Moreover, this study indicates that Hz, inflammation, and pulmonary pathology are tightly associated with each other, independent of the parasite strain used. In a previous study, we found that mice infected with lethal P. berghei parasites contain significantly higher amounts of Hz in different organs, including the lungs, at similar peripheral parasitemia levels, compared with mice infected with nonlethal PcAS (19). The current data indicate that this is mainly attributable to differences in Hz production by the parasites, and not to differences in degradation rates by the host. Although the amount of Hz decreased in the lungs after therapeutic parasite clearance, no Hz degradation was found during a 4-month

598


Figure 7. Plasmodium falciparum–derived Hz (PfHz) induces the expression of inflammatory mediators in the lungs. Malaria-free mice were injected intravenously with different amounts of PfHz, and dissected 6 hours later. (A) PfHz distribution in different organs was measured according to the chemoluminescence method. (B) Cytospins were prepared from peripheral blood leukocytes, and stained with hemacolor (Merck, Darmstadt, Germany). Internalized PfHz (arrows) in macrophages was observed in the peripheral circulation (original magnification, 3100; scale bar, 10 mm). (C) VEGF concentrations in the BAL fluid of PfHz-injected mice. Data were stratified according to the total amount of PfHz measured in the lungs. (D–I) Pulmonary mRNA expression levels of interferon-g inducible protein–10 (IP-10) (D), monocyte chemotactic protein-1 (MCP-1) (E), keratinocytederived chemokine (KC) (F), IL-1b (G), IL-6 (H), or TNF (I), as determined by quantitative RT-PCR and normalized to 18S ribosomal RNA concentrations. Data were stratified according to the total amount of PfHz measured in the lungs. Additional expression data are provided in Table E3. Horizontal bars between individual data points indicate group medians. Horizontal lines below asterisks indicate statistical differences between groups. Asterisks above individual datasets indicate statistical differences with the vehicle-injected control group (PBS). *P , 0.05. **P , 0.01. ***P , 0.001.

observation period, which is consistent with previous reports (22, 23). Interestingly, the genetic ablation of plasmepsin-4, a protease of the food vacuole, resulted in decreased Hz production (30) and in the attenuation of parasite growth and pathology (31). The current data demonstrate that Hz, pulmonary edema, and inflammation are tightly associated with each other during MA-ARDS in mice. Furthermore, the injection of PfHz into malaria-free mice stimulated the pulmonary expression of chemokines and cytokines in the absence of a malaria infection, confirming the important role of Hz in pulmonary inflammation. MCP-1 and IP-10 can recruit monocytes/macrophages and

activated T cells to the lungs, where they may become a source of local cytokine production. Cytotoxic T cells participate in the pathogenesis, of MA-ARDS because the depletion of this cell type protected mice from pulmonary pathology (9). Neutrophils were also present in the lungs of mice with MA-ARDS, although their numbers were lower than those of mononuclear cells. The injection of PfHz induced the pulmonary expression of the neutrophil-attracting chemokine KC, but also at a lower concentration compared with chemokines recruiting mononuclear cells. Furthermore, Hz is present in monocytes/macrophages in lungs of patients with MA-ARDS (4, 10, 11), IL-1b and TNF are induced by Hz in human monocytes isolated from


the peripheral blood (32), and IL-1b and TNF are expressed in inflammatory cells collected from the BAL fluid of patients with nonmalarial septic ARDS (33), substantiating the relevance of our findings for human pathology. The PfHz used for the intravenous injections in this study was purified from the supernatant of P. falciparum cultures and treated with DNase to remove any potentially contaminating DNA. This ensures that it closely mimics the state of Hz when it is released into the circulation. In particular, the crystal size and the presence of associated bioactive molecules such as hydroperoxy and hydroxyl fatty acids, 4-hydroxynonenal (4-HNE), and fibrinogen are key determinants of the immunological effects of Hz (34–36). How Hz triggers inflammation remains incompletely understood. Hz activates the nucleotide-binding oligomerization domain (NOD)-like receptor pyrin domain containing-3 (NLRP3) inflammasome, which causes the release of IL-1b and which is involved in neutrophil recruitment (37, 38). Furthermore, the interaction between Hz-bound fibrinogen and Toll-like receptor–4 (TLR4) on monocytes immediately elicits an extremely high oxidative burst, together with the secretion of MCP-1 and TNF (36). The possible involvement of TLR9 in Hz signaling is controversial (39–41), and its role in malaria may be confused by the insufficiently defined genetic background of TLR9 knockout mice (42). Once inside the phagocytes, Hz interacts with polyunsaturated fatty acids present in membranes, resulting in lipoperoxidation (34, 43, 44), and lipoperoxidation products such as 4-HNE or hydroxyleicosatetraenoic acids (HETEs) may exert opposite effects in monocytes. They can inhibit important monocyte functions such as phagocytosis, antigen presentation, and the differentiation of macrophages into dendritic cells, as reviewed elsewhere (17, 35), but they can also induce the expression of proinflammatory cytokines and chemokines (32, 45–47). In addition to its effects in immune cells, the high amounts of 4-HNE produced after Hz phagocytosis will have consequences for the adjacent cells in the alveoli. HNE is highly reactive and is able to cross the cell membrane, modifying the proteins and DNA of neighboring cells. Together, these data suggest that Hz may reside at the basis of an inflammatory cascade that causes damage to the endothelial and alveolar barrier. Furthermore, our data indicate a highly significant correlation between lung Hz, alveolar edema, and VEGF protein concentrations, thereby confirming the essential pathogenic role of VEGF as observed in the PbANKA-induced MA-ALI model (25). Nevertheless, we found that VEGF mRNA concentrations were decreased in lung homogenates, and that PfHz injection did not induce alveolar VEGF protein concentrations, suggesting that Hz does not directly trigger VEGF production or release in the alveoli. Pulmonary VEGF is predominantly produced by epithelial cells and diffuses through the alveolar–capillary membrane, where it binds to VEGF receptors on the vascular endothelium (26). Epithelial damage may thus explain the decreased VEGF transcripts that we found in the lungs. Similar decreased VEGF transcripts have been found in other ALI/ARDS models in which epithelial damage is evident (48). This implies that the increased VEGF protein concentrations probably arise from the increased cleavage of VEGF from the extracellular matrix by proteases, or from infiltrating phagocytes, which are able to release large amounts of VEGF (48, 49), rather than from de novo synthesis in the lungs. TNF and IL-6 were also induced in the lungs by Hz, and are also known to increase vascular permeability. Moreover, TNF reduces the pulmonary expression of the amiloride-sensitive epithelial sodium channel (ENaC) in the alveoli, thereby altering alveolar fluid clearance (26). Alveolar ENaC expression is decreased during pulmonary pathology in mice infected with the P. berghei strain K173, and this may have contributed to the observed interstitial edema (50).

599

Interestingly, Gillrie and colleagues demonstrated that merozoite proteins alter endothelial barrier function through a Src-family kinase–dependent mechanism (51). This may also occur in the lungs, and suggests that Hz may not be the only factor involved in the pathogenesis of MA-ARDS. Together, these data demonstrate that Hz, which exists at the interface between host and parasite, is a prominent inflammatory virulence factor in the pathogenesis of MA-ARDS. These experiments were performed in a murine model with high similarity to human MA-ARDS and with natural PfHz from a human malaria parasite. Therefore (and because several postmortem studies [10, 11] describe the presence of Hz in phagocytes and iRBCs in lungs of patients with MA-ARDS), we believe that the data presented in this study are particularly relevant for the understanding of this lethal complication. Author disclosures are available with the text of this article at www.atsjournals.org. Acknowledgments: The authors thank Professor Hubertine Heremans, Dr. Nele Berghmans, Paula Aertsen, and Lieve Ophalvens for their excellent help and technical skills.

References 1. Day N, Dondorp AM. The management of patients with severe malaria. Am J Trop Med Hyg 2007;77:29–35. 2. Matthay MA, Zemans RL. The acute respiratory distress syndrome: pathogenesis and treatment. Annu Rev Pathol 2011;6:147–163. 3. Mohan A, Sharma SK, Bollineni S. Acute lung injury and acute respiratory distress syndrome in malaria. J Vector Borne Dis 2008;45: 179–193. 4. Taylor WR, Hanson J, Turner GD, White NJ, Dondorp AM. Respiratory manifestations of malaria. Chest 2012;142:492–505. 5. Daneshvar C, Davis TM, Cox-Singh J, Rafa’ee MZ, Zakaria SK, Divis PC, Singh B. Clinical and laboratory features of human Plasmodium knowlesi infection. Clin Infect Dis 2009;49:852–860. 6. William T, Menon J, Rajahram G, Chan L, Ma G, Donaldson S, Khoo S, Frederick C, Jelip J, Anstey NM, et al. Severe Plasmodium knowlesi malaria in a tertiary care hospital, Sabah, Malaysia. Emerg Infect Dis 2011;17:1248–1255. 7. Tan LK, Yacoub S, Scott S, Bhagani S, Jacobs M. Acute lung injury and other serious complications of Plasmodium vivax malaria. Lancet Infect Dis 2008;8:449–454. 8. Anstey NM, Russell B, Yeo TW, Price RN. The pathophysiology of vivax malaria. Trends Parasitol 2009;25:220–227. 9. Van den Steen PE, Geurts N, Deroost K, Van Aelst I, Verhenne S, Heremans H, Van Damme J, Opdenakker G. Immunopathology and dexamethasone therapy in a new model for malaria-associated acute respiratory distress syndrome. Am J Respir Crit Care Med 2010;181: 957–968. 10. Valecha N, Pinto RG, Turner GD, Kumar A, Rodrigues S, Dubhashi NG, Rodrigues E, Banaulikar SS, Singh R, Dash AP, et al. Histopathology of fatal respiratory distress caused by Plasmodium vivax malaria. Am J Trop Med Hyg 2009;81:758–762. 11. Duarte MI, Corbett CE, Boulos M, Amato NV. Ultrastructure of the lung in Falciparum malaria. Am J Trop Med Hyg 1985;34:31–35. 12. Egan TJ. Haemozoin formation. Mol Biochem Parasitol 2008;157: 127–136. 13. Amodu OK, Adeyemo AA, Olumese PE, Gbadegesin RA. Intraleucocytic malaria pigment and clinical severity of malaria in children. Trans R Soc Trop Med Hyg 1998;92:54–56. 14. Nguyen PH, Day N, Pram TD, Ferguson DJ, White NJ. Intraleucocytic malaria pigment and prognosis in severe malaria. Trans R Soc Trop Med Hyg 1995;89:200–204. 15. Hanscheid T, Egan TJ, Grobusch MP. Haemozoin: from melatonin pigment to drug target, diagnostic tool, and immune modulator. Lancet Infect Dis 2007;7:675–685. 16. Shio MT, Kassa FA, Bellemare MJ, Olivier M. Innate inflammatory response to the malarial pigment hemozoin. Microbes Infect 2010;12: 889–899.

600


17. Arese P, Schwarzer E. Malarial pigment (haemozoin): a very active “inert” substance. Ann Trop Med Parasitol 1997;91:501–516. 18. Jaramillo M, Plante I, Ouellet N, Vandal K, Tessier PA, Olivier M. Hemozoin-inducible proinflammatory events in vivo: potential role in malaria infection. J Immunol 2004;172:3101–3110. 19. Deroost K, Lays N, Noppen S, Martens E, Opdenakker G, Van den Steen PE. Improved methods for haemozoin quantification in tissues yield organ- and parasite-specific information in malaria-infected mice. Malar J 2012;11:166. 20. Janse CJ, Ramesar J, Waters AP. High-efficiency transfection and drug selection of genetically transformed blood stages of the rodent malaria parasite Plasmodium berghei. Nat Protoc 2006;1:346–356. 21. Spence PJ, Cunningham D, Jarra W, Lawton J, Langhorne J, Thompson J. Transformation of the rodent malaria parasite Plasmodium chabaudi. Nat Protoc 2011;6:553–561. 22. Levesque MA, Sullivan AD, Meshnick SR. Splenic and hepatic hemozoin in mice after malaria parasite clearance. J Parasitol 1999; 85:570–573. 23. Frita R, Carapau D, Mota MM, Hanscheid T. In vivo hemozoin kinetics after clearance of Plasmodium berghei infection in mice. Malar Res Treat 2012;2012:373086. 24. Lawrence C, Olson JA. Birefringent hemozoin identifies malaria. Am J Clin Pathol 1986;86:360–363. 25. Epiphanio S, Campos MG, Pamplona A, Carapau D, Pena AC, Ataide R, Monteiro CA, Felix N, Costa-Silva A, Marinho CR, et al. VEGF promotes malaria-associated acute lung injury in mice. PLoS Pathog 2010;6:e1000916. 26. Mura M, dos Santos CC, Stewart D, Liu M. Vascular endothelial growth factor and related molecules in acute lung injury. J Appl Physiol 2004; 97:1605–1617. 27. Strieter RM, Lynch JP III, Basha MA, Standiford TJ, Kasahara K, Kunkel SL. Host responses in mediating sepsis and adult respiratory distress syndrome. Semin Respir Infect 1990;5:233–247. 28. Taylor TE, Fu WJ, Carr RA, Whitten RO, Mueller JS, Fosiko NG, Lewallen S, Liomba NG, Molyneux ME. Differentiating the pathologies of cerebral malaria by postmortem parasite counts. Nat Med 2004;10:143–145. 29. Craig A, Grau G, Janse C, Kazura J, Milner D, Barnwell J, Turner G, Langhorne J. The role of animal models for research on severe malaria. PLoS Pathog 2012;8:e1002401. 30. Bonilla JA, Bonilla TD, Yowell CA, Fujioka H, Dame JB. Critical roles for the digestive vacuole plasmepsins of Plasmodium falciparum in vacuolar function. Mol Microbiol 2007;65:64–75. 31. Spaccapelo R, Janse CJ, Caterbi S, Franke-Fayard B, Bonilla JA, Syphard LM, Di Cristina M, Dottorini T, Savarino A, Cassone A, et al. Plasmepsin 4–deficient Plasmodium berghei are virulence attenuated and induce protective immunity against experimental malaria. Am J Pathol 2010; 176:205–217. 32. Giribaldi G, Prato M, Ulliers D, Gallo V, Schwarzer E, Akide-Ndunge OB, Valente E, Saviozzi S, Calogero RA, Arese P. Involvement of inflammatory chemokines in survival of human monocytes fed with malarial pigment. Infect Immun 2010;78:4912–4921. 33. Hashimoto S, Kobayashi A, Kooguchi K, Kitamura Y, Onodera H, Nakajima H. Upregulation of two death pathways of perforin/granzyme and FasL/Fas in septic acute respiratory distress syndrome. Am J Respir Crit Care Med 2000;161:237–243. 34. 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. 35. Schwarzer E, Skorokhod OA, Barrera V, Arese P. Hemozoin and the human monocyte: a brief review of their interactions. Parassitologia 2008;50:143–145.

36. Barrera V, Skorokhod OA, Baci D, Gremo G, Arese P, Schwarzer E. Host fibrinogen stably bound to hemozoin rapidly activates monocytes via TLR-4 and CD11b/CD18-integrin: a new paradigm of hemozoin action. Blood 2011;117:5674–5682. 37. Shio MT, Eisenbarth SC, Savaria M, Vinet AF, Bellemare MJ, Harder KW, Sutterwala FS, Bohle DS, Descoteaux A, Flavell RA, et al. Malarial hemozoin activates the NLRP3 inflammasome through Lyn and Syk kinases. PLoS Pathog 2009;5:e1000559. 38. Dostert C, Guarda G, Romero JF, Menu P, Gross O, Tardivel A, Suva ML, Stehle JC, Kopf M, Stamenkovic I, et al. Malarial hemozoin is a NALP3 inflammasome activating danger signal. PLoS ONE 2009;4:e6510. 39. Parroche P, Lauw FN, Goutagny N, Latz E, Monks BG, Visintin A, Halmen KA, Lamphier M, Olivier M, Bartholomeu DC, et al. Malaria hemozoin is immunologically inert but radically enhances innate responses by presenting malaria DNA to Toll-like receptor 9. Proc Natl Acad Sci USA 2007;104:1919–1924. 40. Coban C, Ishii KJ, Kawai T, Hemmi H, Sato S, Uematsu S, Yamamoto M, Takeuchi O, Itagaki S, Kumar N, et al. Toll-like receptor 9 mediates innate immune activation by the malaria pigment hemozoin. J Exp Med 2005;201:19–25. 41. Wu X, Gowda NM, Kumar S, Gowda DC. Protein–DNA complex is the exclusive malaria parasite component that activates dendritic cells and triggers innate immune responses. J Immunol 2010;184: 4338–4348. 42. Geurts N, Martens E, Verhenne S, Lays N, Thijs G, Magez S, Cauwe B, Li S, Heremans H, Opdenakker G, et al. Insufficiently defined genetic background confounds phenotypes in transgenic studies as exemplified by malaria infection in Tlr9 knockout mice. PLoS ONE 2011;6: e27131. 43. Schwarzer E, Muller O, Arese P, Siems WG, Grune T. Increased levels of 4-hydroxynonenal in human monocytes fed with malarial pigment hemozoin: a possible clue for hemozoin toxicity. FEBS Lett 1996;388: 119–122. 44. Schwarzer E, Turrini F, Ulliers D, Giribaldi G, Ginsburg H, Arese P. Impairment of macrophage functions after ingestion of Plasmodium falciparum–infected erythrocytes or isolated malarial pigment. J Exp Med 1992;176:1033–1041. 45. Prato M, Gallo V, Giribaldi G, Aldieri E, Arese P. Role of the NF-kappaB transcription pathway in the haemozoin- and 15-HETE–mediated activation of matrix metalloproteinase–9 in human adherent monocytes. Cell Microbiol 2010;12:1780–1791. 46. Dell’Agli M, Galli GV, Bulgari M, Basilico N, Romeo S, Bhattacharya D, Taramelli D, Bosisio E. Ellagitannins of the fruit rind of pomegranate (Punica granatum) antagonize in vitro the host inflammatory response mechanisms involved in the onset of malaria. Malar J 2010; 9:208. 47. Polimeni M, Valente E, Aldieri E, Khadjavi A, Giribaldi G, Prato M. Haemozoin induces early cytokine-mediated lysozyme release from human monocytes through p38. PLoS ONE 2012;7:e39497. 48. Ekekezie II, Thibeault DW, Rezaiekhaligh MH, Norberg M, Mabry S, Zhang X, Truog WE. Endostatin and vascular endothelial cell growth factor (VEGF) in piglet lungs: effect of inhaled nitric oxide and hyperoxia. Pediatr Res 2003;53:440–446. 49. Karmpaliotis D, Kosmidou I, Ingenito EP, Hong K, Malhotra A, Sunday ME, Haley KJ. Angiogenic growth factors in the pathophysiology of a murine model of acute lung injury. Am J Physiol Lung Cell Mol Physiol 2002;283:L585–L595. 50. Hee L, Dinudom A, Mitchell AJ, Grau GE, Cook DI, Hunt NH, Ball HJ. Reduced activity of the epithelial sodium channel in malariainduced pulmonary oedema in mice. Int J Parasitol 2011;41:81–88. 51. Gillrie MR, Krishnegowda G, Lee K, Buret AG, Robbins SM, Looareesuwan S, Gowda DC, Ho M. Src-family kinase dependent disruption of endothelial barrier function by Plasmodium falciparum merozoite proteins. Blood 2007;110:3426–3435.