Hemozoin (malarial pigment) directly promotes apoptosis of erythroid precursors

Abstract

Severe malarial anemia is the most common syndrome of severe malaria in endemic areas. The pathophysiology of chronic malaria is characterised by a striking degree of abnormal development of erythroid precursors (dyserythropoiesis) and an inadequate erythropoietic response in spite of elevated levels of erythropoietin. The cause of dyserythropoiesis is unclear although it has been suggested that bone-marrow macrophages release cytokines, chemokines or lipo-peroxides after exposure to hemozoin, a crystalloid form of undigested heme moieties from malarial infected erythrocytes, and so inhibit erythropoiesis. However, we have previously shown that hemozoin may directly inhibit erythroid development in vitro and the levels of hemozoin in plasma from patients with malarial anemia and hemozoin within the bone marrow was associated with reduced reticulocyte response. We hypothesized that macrophages may reduce, not enhance, the inhibitory effect of hemozoin on erythropoiesis. In an in vitro model of erythropoiesis, we now show that inhibition of erythroid cell development by hemozoin isolated from P. falciparum is characterised by delayed expression of the erythroid markers and increased apoptosis of progenitor cells. Crucially, macrophages appear to protect erythroid cells from hemozoin, consistent with a direct contribution of hemozoin to the depression of reticulocyte output from the bone marrow in children with malarial anemia. Moreover, hemozoin isolated from P. falciparum in vitro inhibits erythroid development independently of inflammatory mediators by inducing apoptotic pathways that not only involve activation of caspase 8 and cleavage of caspase 3 but also loss of mitochondrial potential. Taken together these data are consistent with a direct effect of hemozoin in inducing apoptosis in developing erythroid cells in malarial anemia. Accumulation of hemozoin in the bone marrow could therefore result in inadequate reticulocytosis in children that have adequate levels of circulating erythropoietin.

Figure 1. Erythroid Expansion in vitro.

A, Cytospins of cells stained with 1% dianisidine/10% Giemsa are shown to illustrate the morphological changes within this culture as erythroblasts differentiate from larger basophilic cells on day 7 to smaller hemoglobin positive cells on day 10 and with pyknotic nuclei on day 14. The brown stain is indicative of hemoglobinization. B, The generation of erythroid precursors was monitored by labelling cells with fluorescent antibodies specific to the cell surface markers CD71 (transferrin receptor) and CD235a (glycophorin A) and analysed by flow cytometry. C, The changes in FSC and SSC of gated cells are shown in B to indicate the reduced size of erythroid precursors. Representative plots and cytospins viewed at ×60 magnification from 3 or more independent experiments are shown.

Figure 2. Hemozoin inhibits cell expansion and differentiation of erythroid progenitors.

A, An example of the inhibition of progression to basophilic CD71⁺CD235⁺ precursors on day 7 is shown and B, the proportion of cells incubated with 6µg/ml hemozoin (H) that express CD71⁺CD235⁺ on day 7 is compared to media controls (M). * p = 0.030. C, The absolute number of viable cells on day 7 in control cultures with media or with hemozoin. D, An example of reduction in CD71^loCD235⁺ precursors on day 14 and E, the proportion of cells incubated with hemozoin that express CD71^loCD235⁺ on day 14 is compared to media controls. **p = 0.006. F, The absolute number of viable cells on day 14 following culture with hemozoin or control cultures. *p = 0.012. Dot plots show the percentage of live cells expressing markers for each quadrant. Bar graphs show the average of 3 or more independent experiments with SEMs where for each experiment values for CD71⁺CD235⁺ or CD71^loCD235⁺ have been normalized to cultures grown in media alone.

Figure 3. Assessment of reactive oxygen species (ROS) in erythroid cells following incubation with hemozoin.

A, Relative levels of ROS were determined using the fluorescent dye 2, 7-dichlorofluoroscein diacetate. The mean fluorescence intensity relative to media controls at each time point is shown. B, Viable erythroid cells in the same cultures with controls (MEDIA) or with 6µg/ml hemozoin (HZ). C, The proportion of late erythroblasts (CD71^loCD235a⁺) and D, the total number of CD235a⁺ erythroid cells in culture on day 14 when treated with the anti-oxidant vitamin E. Cells were incubated with 6µg/ml hemozoin on day 0 or with the same dose of hemozoin after pre incubation with 30µM vitamin E and normalized to media controls in the absence or presence of vitamin E respectively. The averages of 3 independent experiments with SEMs are shown. A * p = 0.021, B *** p = 0.0003 C * p = 0.0278 and D * p = 0.0236.

Figure 4. Macrophages do not contribute to erythroid inhibition induced by hemozoin.

A, Cytokine levels in supernatants from erythroid cultures in the absence or presence of a high dose of hemozoin (25µg/ml) are shown for days 7 and 14. Empty bars represent levels detected in control cultures with media alone for each time point. The average of 3 or more independent experiments is shown with SEMs. B, Induction of MCP-1 by hemozoin is compared between erythroid cultures depleted of CD14⁺ macrophages on day 0 (−MØ) and cultures in which macrophages are present (+MØ). A representative experiment of 2 is shown with error bars as standard deviation of measurements in triplicate. C, The maturation of erythroblasts on day 14 to CD71^loCD235a⁺ precursors in cultures depleted of CD14⁺ on day 0 (−MØ) is reduced compared with the same cultures that have not had macrophages removed (+MØ). The effect of macrophage loss on erythroid development was taken into account by normalizing the yield of erythroblasts with hemozoin to the yield obtained in macrophage-depleted control cultures, where media but no hemozoin was added. The average of 2 independent experiments is shown where error bars are SEMs. **p = 0.003. D, O-dianisidine staining of cytospin preparations to show changes in morphology and hemoglobinization (brown staining) of cells cultured without macrophages and cultures with macrophages that are found associated with larger clusters of hemozoin. Images are representative of 3 or more experiments and were taken at ×40 magnification.

Figure 5. Macrophage content required to reduce inhibition by hemozoin.

Increasing proportions of isogenic CD14⁺ cells (MØ) isolated from the same culture were added on day 0 together with hemozoin. A, Absolute numbers of erythroblasts on day 14 and B, normalization of the same cell counts to media controls with each concentration of CD14⁺ cells to allow assessment of rescue by CD14⁺ cells. A representative experiment of 3 is shown where error bars are SDs of measurements in triplicate. ** p = 0.007.

Figure 6. Apoptosis of erythroid cells is enhanced by hemozoin.

A, Basophilic erythroblasts from day 6 cultures were incubated with 20µg/ml anti-CD95 (CD95) or hemozoin at 6µg/ml (Hz) and 24 hours later assessed for i, the relative proportion of CD71⁺ erythroblasts ii, exposure of phosphatidylserine on non necrotic 7AAD⁻ cells and iii, activation of caspase 8 (FLICA LETD⁺) in Annexin V⁺ cells; iv, a proportion of cells from the same experimental groups were also assessed for permeability of mitochondria using the fluorescent dye JC-1. Fluorescence from JC-1 aggregates in mitochondria (JC1-A) and from monomers in the cytosol (JC1-M) is used to calculate the ratio of JC1-A to JC1-M to allow for comparison between experiments. Loss of membrane potential is indicated by values less than 1 shown in bold in the lower right quadrant of each dot plot. The proportion of cells with JC1-A^hi and JC1-A^lo fluorescence is also shown to the left. Controls are isotype controls. B, The average of 3 or more independent experiments as described in A following incubation with media, anti-CD95, or hemozoin (Hz) demonstrating fold changes in apoptotic markers. Error bars are of SEMs and p values are from analyses using the student's t test.

Figure 7. Activation of Caspases in erythroid cells.

A, Induction of cleaved caspase 3 and cytochrome C in basophilic erythroblasts purified from day 6 cultures 24 hours after incubation with media (controls), anti-CD95 or 6µg/ml hemozoin. B, Western blots of lysates taken from purified basophilic erythroblasts incubated with hemozoin for 4 hours demonstrate increased levels of cleaved caspase 8 and activation of caspase 3 with slight induction of cleaved Bid (tBid). C, Detection of cleaved caspase 9 after 4 and 24 hours incubation with hemozoin. Each band was normalised to that for alpha tubulin. Fold changes in normalized levels of pro-apoptotic proteins relative to those in lysates from media controls are shown below each band.