Plasmodium falciparum infected erythrocytes induce hepcidin (HAMP) mRNA synthesis by peripheral blood mononuclear cells

Many factors contribute to malarial anaemia, including haemolysis, increased erythrophagocytosis and suppressed erythropoiesis (Lamikanra et al, 2007). Reduced production of red blood cells can be caused by a restriction in iron supply to the bone marrow. Hepcidin blocks iron transport through ferroportin, and inhibits both dietary iron uptake and recycling of red cell iron through macrophages (Nemeth et al, 2004a; Andrews, 2008). Thus, increased hepcidin levels would provide a rationale to explain some aspects of malarial anaemia, and recently, raised hepcidin levels have been observed in the urine and serum of naturally and experimentally Plasmodium infected humans (Howard et al, 2007; de Mast et al, 2009a,b). During the course of a blood stage Plasmodium infection, white blood cells are exposed to both infected red cells and the products of infected cell lysis as a result of parasite multiplication. We hypothesized that the leucocyte host response to parasitized cells might provide a source for hepcidin during malaria, analogous to hepcidin synthesis by leucocytes after exposure to bacterial lipopolysaccharride (LPS) (Peyssonnaux et al, 2006). LPS strongly induces cytokines, including interleukin-6 (IL6), which then stimulate hepcidin production (Nemeth et al, 2004b; Theurl et al, 2008). To test this idea, we co-cultured P. falciparum infected red blood cells (iRBC) with peripheral blood mononuclear cells (PBMC) from uninfected donors at various red cell to white cell ratios, with uninfected red cells (uRBC) used as controls. The laboratory P. falciparum clone A4 and its sub-clone, 2B2, were cultured as previously described (Kyes et al, 2007). Changes in HAMP (hepcidin) and IL6 mRNA relative to GAPDH were measured by real-time quantitative reverse transcription polymerase chain reaction (qRT-PCR). Increasing iRBC numbers caused a parallel increase in HAMP mRNA synthesis by PBMC after 3 h of co-culture, while uninfected red cells did not induce HAMP (Fig 1A). However, unexpectedly, iRBC did not strongly induce IL6 mRNA (Fig 1B). We further explored the mechanism of HAMP induction by iRBC by comparison with classical Toll-like receptor (TLR) agonists. We measured HAMP and IL6 mRNA induction in several donor PBMC in response to P. falciparum parasitized RBC or to the TLR agonists LPS and flagellin. HAMP mRNA was induced by iRBC to a highly significant extent (P = 0·0039, Willcoxon matched pairs test) compared to uRBC, whereas significant induction was not observed by uRBC compared to untreated cells (Fig 2A). HAMP mRNA levels induced by iRBC were comparable to those induced by LPS or flagellin whereas IL6 mRNA induction by iRBC was low or not apparent, but was very strong by LPS and flagellin (Fig 2B). The discrepancy between IL6 and HAMP induction by iRBC versus TLR agonists was more obvious when comparing the ratio of HAMP to IL6 induction (Fig 2C). For iRBC the ratio was 4·26 whereas for LPS and flagellin it was 0·061 and 0·162, respectively. Although iRBC also induced other cytokines characteristic of TLR ligation (TNF, IL8 and IL1B) the mRNA levels of these cytokines were also substantially less than those induced by purified TLR agonists (Fig 2D). Figure 2D also reinforces the observation that, while IL6 shows the greatest induction by TLR agonists, the fold-change of HAMP induced by iRBC is greater than the fold-change of the pro-inflammatory cytokines. The data are consistent with iRBC inducing TLR signalling and pro-inflammatory cytokine synthesis to a relatively low extent but inducing hepcidin via a mechanism that is at least partly independent of IL6. During inflammation and bacterial infections, both hepatocytes and haematopoietic cells can make hepcidin in response to IL6 (Nemeth et al, 2004b; Theurl et al, 2008). However in a recent report of urinary hepcidin in individuals naturally infected with Plasmodium, although IL6 and other cytokines correlated with hepcidin levels, reverse stepwise multivariate linear regression analysis showed IL6 levels were independent of hepcidin levels (de Mast et al, 2009a). This observation suggests other factors aside from IL6 could be responsible for hepcidin synthesis in the context of malaria, and is consistent with the data presented here. Hepcidin induction by P. falciparum parasitized erythrocytes. 5 × 106 PBMC were co-cultured for 3 h with parasite-infected erythrocytes (iRBC) or uninfected erythrocytes (uRBC) at various PBMC:erythrocyte ratios, or left untreated. RNA was extracted from pooled adherent and pelleted non-adherent cell lysates using the RNeasy Mini Kit with QIAshredder homogenisation (Qiagen, Crawley, West Sussex, UK); RNA was reverse transcribed using the High Capacity RNA-to-cDNA kit, and HAMP (A) and IL6 (B) expression was quantified relative to GAPDH by qRT-PCR with TaqMan Gene Expression assays and Master Mix (all Applied Biosystems, Foster City, CA, USA). Fold-changes relative to levels in untreated cells from the same donor were calculated using the 2ΔΔCT method. Further details of all methods are available from the authors on request. Values shown are means of three experiments using PBMC from three different donors; error bars show standard error of the mean (SEM). Data was processed and analysed using Microsoft Excel and graphpad prism software (GraphPad Software, Inc., La Jolla, Ca). Hepcidin induction by parasitized erythrocytes differs from hepcidin induction by TLR agonists. Donor (n = 9) PBMC were co-cultured for 3 h either with RBC parasitized with the A4 clone or the 2B2 subclone of P. falciparum at red cell ratios of 20:1 (solid triangles)or 10:1 (open triangles), or with the purified TLR agonists E. coli K12 LPS (1 μg/ml) or Salmonella typhimurium flagellin (0·5 μg/ml) (Invivogen, San Diego, CA, USA). Mean changes in HAMP (A), IL6 (B) mRNA expression relative to untreated PBMC. (C) Ratio of HAMP to IL6 fold-changes relative to untreated PBMC (=1). (D) Induction of HAMP and cytokines by uRBC, iRBC, LPS or flagellin, relative to untreated PBMC (=1). Bars represent the mean, error bars represent the SEM in C, D. In endemic areas, malaria parasitaemia is a continuous variable with much of the population showing low or undetectable parasitaemia during asymptomatic infection. In the case of episodes of clinical disease, although varying with age and endemicity, clinical thresholds may lie roughly between 5000 and 10 000 parasites/μl. Depending on the haematocrit of the patient this translates to between 0·1% and 0·4% of red cells infected, but parasitaemia will rise much higher during disease episodes. In vivo, the ratio of red cells to leucocytes is approximately 600:1 and in our experiments, the ratio of red cells to PBMC was 20:1 at a parasitaemia of c. 80%. If PBMC are the only source of parasite-induced hepcidin, then our data suggest significant hepcidin induction would only occur during clinical infection. We have only measured HAMP mRNA induction by haematopoietic cells in the presence of iRBC. We have not excluded hepatocytes as an important source of hepcidin during malarial infection, either during the liver-stage of infection or, perhaps more likely, during blood-stage malarial fever. Suppressed erythropoiesis contributes to malarial anaemia, as new red cells are not produced to replace those lost by haemolysis and ingestion by macrophages. Normally, hepcidin levels are inhibited during periods of strong bone marrow activity, increasing iron release from stores and absorption from the diet (Andrews, 2008). As a result, more iron is available to supply erythropoiesis. In contrast, raised hepcidin levels during malaria infection would reverse this flow of iron and further starve the erythron when its need for iron is urgent. Work by de Mast et al (2009a,b) and Howard et al (2007) showed that increased hepcidin is indeed present during the blood stage of malaria infections in vivo, and here we show that P. falciparum parasitized eryrthocytes induce hepcidin expression from co-cultured leucocytes. If hepcidin does contribute to malarial anaemia, further understanding of the molecular mechanism of Plasmodium-induced hepcidin may enable interventions to reduce hepcidin expression during infection. This could provide a new approach to modify the anaemia associated with malaria. We thank the MRC, Wellcome Trust and the Beit Memorial Fellowship for Medical Research for support.