Systemic release of high mobility group box 1 (HMGB1) protein is associated with severe and fatal Plasmodium falciparum malaria

Abstract

Background: Severe falciparum malaria (SM) pathogenesis has been attributed, in part, to deleterious systemic host inflammatory responses to infection. High mobility group box 1 (HMGB1) protein is an important mediator of inflammation implicated in sepsis pathophysiology.

Methods: Plasma levels of HMGB1 were quantified in a cohort of febrile Ugandan children with Plasmodium falciparum infection, enrolled in a prospective observational case-controlled study, using a commercial enzyme-linked immunosorbent assay. The utility of HMGB1 to distinguish severe malaria (SM; n = 70) from uncomplicated malaria (UM; n = 33) patients and fatal (n = 21) versus non-fatal (n = 82) malaria, at presentation, was examined. Receiver operating characteristic curve analysis was used to assess the prognostic accuracy of HMGB1. The ability of P. falciparum-parasitized erythrocytes to induce HMGB1 from peripheral blood mononuclear cells was assessed in vitro. The effect of an anti-HMGB1 neutralizing antibody on disease outcome was assessed in the experimental Plasmodium berghei ANKA rodent parasite model of SM. Mortality and parasitaemia was assessed daily and compared to isotype antibody-treated controls.

Results: Elevated plasma HMGB1 levels at presentation were significantly associated with SM and a subsequent fatal outcome in paediatric patients with P. falciparum infection. In vitro, parasitized erythrocytes induced HMGB1 release from human peripheral blood mononuclear cells. Antibody-mediated neutralization of HMGB1 in the experimental murine model of severe malaria failed to reduce mortality.

Conclusion: These data suggest that elevated HMGB1 is an informative prognostic marker of disease severity in human SM, but do not support HMGB1 as a viable target for therapeutic intervention in experimental murine SM.

Figure 1

HMGB1 is released with Plasmodium falciparum infection and predictive of malarial disease severity and clinical outcome in Ugandan children. Plasma HMGB1 levels at clinical presentation are significantly higher in children with severe malaria (SM) compared to uncomplicated malaria (UM) (A) and in fatal cases when compared to non-fatal cases (B). Statistical analysis by Mann Whitney U test. * indicates a p value of <0.01. (C) Receiver operating characteristics (ROC) curves were generated for HMGB1 (solid line) and parasitaemia (dashed line) to assess the utility in predicting outcome. Area under the curve (AUC) is displayed with 95% confidence intervals in parentheses. p values were adjusted for multiple comparisons using Bonferroni-Holm’s correction (n = 2). * indicates a significant p value of <0.025 (D) HMGB1 is detectable in cultured media 24 hours following peripheral blood mononuclear cells (PBMC; 1.5 × 10⁶ per well) (isolated from healthy, malaria-naïve donors) stimulation with parasitized erythrocytes PEs (4.5 × 10⁶ per well) or LPS (100 ng/ml) but not after stimulation with uninfected erythrocytes (uE; 4.5 × 10⁶ per well) or media alone. HMGB1 was analysed by western blot using rabbit polyclonal anti-HMGB1 antibodies. (E) The kinetics of HMGB1 release from PBMCs (1.5 × 10⁶ per well) was assessed two, six and 24 hours after stimulated with PEs (4.5 × 10⁶ per well) or LPS (100 ng/ml).

Figure 2

HMGB1 is elevated in plasma of mice susceptible to developing severe and fatal malaria following Plasmodium berghei ANKA infection. Plasma expression levels of HMGB1 were measured by Western blot from plasma collected on day 5 post-infection with Plasmodium berghei-ANKA, a rodent adapted parasite strain (A-B) Band intensities for D0 (grey bar) and D5 (white bar) in each group were compared on the same blot by using the NIH Image J software. Statistically significant differences between D0 and D5 were assessed by the paired t-test in each group. * indicates a p value of <0.05 (p = 0.004). Histograms represent mean + SEM obtained from a pooled result from at least two independent Western blot analyses from two independent PbA infection studies (n = 15-20/group). Representative Western blots are shown in panel (B). Systemic levels of HMGB1 in plasma over the course of PbA infection (on day 1, 3, 5 and 6 post infection) were assessed by ELISA in ECM-susceptible C57BL/6 mice (C). HMGB1 levels are shown as the fold change compared to the geometric mean of baseline HMGB1 levels obtained from naïve, uninfected littermate C57BL/6 mice. Bars represent mean + SEM (n = 5/time point). Statistical significant difference between day 1 and day 5 levels were determined by Kruskal Wallis with Dunn’s multiple comparison test. * indicates a p value of <0.05.

Figure 3

HMGB1 is not necessary or sufficient for the development of severe malaria following Plasmodium berghei ANKA (PbA) infection in the experimental murine model. (A) Kaplan-Meier survival analysis after C57BL/6 mice were infected with 1 × 10⁶Plasmodium berghei-ANKA (PbA) infected erythrocytes. There was no significant difference in survival between mice receiving an HMGB1 neutralizing antibody (2G7) (black circle, solid line), or control IgG2b (grey diamond, solid line) (Log rank test, p = 0.9, median survival = eight days for both groups; n = 20/group) or PBS control (grey x, dashed line) (Log rank test, p = 0.06; n = 15). Data is pooled from two independent experiments. The kinetics of peripheral parasitaemia are shown up to day 9 post PbA infection for anti-HMGB1 2G7 treatment mice and isotype IgG control (B). Results expressed as a percentage of the parasitized erythrocytes (PE) and represent mean + SEM from one representative experiment (n = 10/group). Levels of TNF, IFN-γ, IL-10, IL-12 and MCP-1 were measured in serum collected from naïve, uninfected mice and PbA-infected mice on day 5 pi using the mouse cytometric bead array (C-G). Bars and whisker represent median and interquartile range (IQR). Statistical differences were assessed by Kruskal-Wallis one-way analysis of variance with Dunn’s multiple comparisons test. * indicates a p value of <0.05. Plasma levels of IFN-γ were confirmed by ELISA (H) and statistical difference between anti-HMGB1 2G7 and isotype IgG control was assessed by Mann–Whitney test (ns, p = 0.73). The line represents the median levels of all individual values in the scatter dot plot.