Dysregulation of the haem-haemopexin axis is associated with severe malaria in a case-control study of Ugandan children

Abstract

Background: Malaria is associated with haemolysis and the release of plasma haem. Plasma haem can cause endothelial injury and organ dysfunction, and is normally scavenged by haemopexin to limit toxicity. It was hypothesized that dysregulation of the haem-haemopexin pathway contributes to severe and fatal malaria infections.

Methods: Plasma levels of haemin (oxidized haem), haemopexin, haptoglobin, and haemoglobin were quantified in a case-control study of Ugandan children with Plasmodium falciparum malaria. Levels at presentation were compared in children with uncomplicated malaria (UM; n = 29), severe malarial anaemia (SMA; n = 27) or cerebral malaria (CM; n = 31), and evaluated for utility in predicting fatal (n = 19) vs non-fatal (n = 39) outcomes in severe disease. A causal role for haemopexin was assessed in a pre-clinical model of experimental cerebral malaria (ECM), following disruption of mouse haemopexin gene (hpx). Analysis was done using Kruskall Wallis tests, Mann-Whitney tests, log-rank tests for survival, and repeated measures ANOVA.

Results: In Ugandan children presenting with P. falciparum malaria, haemin levels were higher and haemopexin levels were lower in SMA and CM compared to children with UM (haemin, p < 0.01; haemopexin, p < 0.0001). Among all cases of severe malaria, elevated levels of haemin and cell-free haemoglobin at presentation were associated with subsequent mortality (p < 0.05). Compared to ECM-resistant BALB/c mice, susceptible C57BL/6 mice had lower circulating levels of haemopexin (p < 0.01), and targeted deletion of the haemopexin gene, hpx, resulted in increased mortality compared to their wild type littermates (p < 0.05).

Conclusions: These data indicate that plasma levels of haemin and haemopexin measured at presentation correlate with malaria severity and levels of haemin and cell-free haemoglobin predict outcome in paediatric severe malaria. Mechanistic studies in the ECM model support a causal role for the haem-haemopexin axis in ECM pathobiology.

Fig. 1

Alterations of the haem axis are associated with disease severity in children with malaria. Plasma samples were collected at presentation. Higher plasma levels of haemin, an oxidized form of haem, and lower levels of haemopexin and haptoglobin were associated with SMA and CM in Ugandan children compared to children with UM. Kruskal–Wallis Test followed by Dunn’s Multiple Comparison Test, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 2

Plasma levels of haemin and cell-free haemoglobin are associated with disease outcome in severe malaria. Children who died of SM, either SMA or CM, had higher plasma levels of haemin and cell-free haemoglobin at presentation compared to children who survived SM. Levels of haemopexin and haptoglobin were lower than those observed in children with UM but were not significantly different between those who survived SM and those who did not. The dotted lines indicate the median levels observed in children with UM. Mann–Whitney test, *p < 0.05

Fig. 3

Endogenous levels of haemopexin are associated with susceptibility to experimental cerebral malaria. a BALB/c mice are more resistant to ECM than C57BL/6 mice as demonstrated by improved survival (pooled data from two biological replicates; n = 15–17 per group; log rank test, **p < 0.01). Both strains of mice had similar levels of b parasitaemia throughout the course of infection (representative data from a single experiment; n = 7-8/group); c BALB/c mice had higher levels of plasma mouse (m) haemopexin compared to the more susceptible C57BL/6 mice when moribund (pooled data from two biological replicates; n = 12–13/group). mHpx Mann–Whitney test; **p < 0.01

Fig. 4

Haemopexin-deficient mice are more susceptible to experimental cerebral malaria. a Both haemopexin knockout (Hpx KO) and haemopexin heterozygous (Hpx HT) mice have decreased survival during ECM than their wild type (Hpx WT) littermates (pooled data from four biological replicates; n = 39–44/group; log rank test, *p < 0.05). All three genotypes had similar levels of b parasitaemia (representative data from a single experiment; n = 7–12/group). The Hpx KO animals had significantly higher levels of c plasma haemin on day 7 post infection, Mann–Whitney, **p < 0.01. The dotted line indicates upper limit of detection for the assay (8.4 uM). The plasma levels of d haemopexin, and e haptoglobin throughout the course of infection (n = 4/group); two-way ANOVA, d p = 0.0018, e p = 0.0002. Samples were collected on day 0 and day 7 post infection, then at cardiac puncture (CP) when mice were moribund