Autoantibodies inhibit Plasmodium falciparum growth and are associated with protection from clinical malaria

Abstract

Many infections, including malaria, are associated with an increase in autoantibodies (AAbs). Prior studies have reported an association between genetic markers of susceptibility to autoimmune disease and resistance to malaria, but the underlying mechanisms are unclear. Here, we performed a longitudinal study of children and adults (n = 602) in Mali and found that high levels of plasma AAbs before the malaria season independently predicted a reduced risk of clinical malaria in children during the ensuing malaria season. Baseline AAb seroprevalence increased with age and asymptomatic Plasmodium falciparum infection. We found that AAbs purified from the plasma of protected individuals inhibit the growth of blood-stage parasites and bind P. falciparum proteins that mediate parasite invasion. Protected individuals had higher plasma immunoglobulin G (IgG) reactivity against 33 of the 123 antigens assessed in an autoantigen microarray. This study provides evidence in support of the hypothesis that a propensity toward autoimmunity offers a survival advantage against malaria.

Keywords: Plasmodium; autoantibodies; autoimmunity; cross-reactivity; infection; malaria.

Figure 1.. High ANA reactivity is independently associated with a reduced risk of clinical malaria.

(A) Plasma samples of participants in the Mali cohort (n=602) were tested for ANAs and stratified into negative (ANA⁻), medium (ANA⁺), and high (ANA⁺⁺) reactivity based on immunofluorescence intensity. (B) ANA status of Malian participants by age (3–11 mos, n=28; 12–23 mos, n=30; 2–5 yr, n=123; 6–12 yr, n=382; 13–17 yr, n=15; and 18–25 yr, n=24), as well as SLE-diagnosed U.S. adults (n=50) and healthy U.S. adults (n=20). Column widths are proportional to sample sizes. (C) ANA status of Mali participants by the presence (n=287) or absence (n=315) of asymptomatic Pf infection before the malaria season, as determined by PCR. (D) ANA status of Mali participants with low (n=199), medium (n=198) or high (n=205) plasma IgG reactivity to Pf GB4 blood-stage lysate. Individuals were stratified into three equal-sized tertiles. (E) Kaplan-Meier plot stratified by ANA status for time to first clinical malaria episode among study participants aged 3 mos to 12 yr (n=563). P value determined by log-rank test. (F) Forest plot showing multivariate analysis of time to first clinical malaria episode using the Cox proportional-hazards regression model among study participants aged 3 mos to 12 yr (n=563), with age, sex, ethnicity, anemia (defined as Hb<12 g/dL (female) or Hb<13.5 g/dL (male)), HbAS, asymptomatic Pf or S. haematobium infection prior to malaria season and anti-Pf IgG titers as covariates. Error bars represent 95% confidence intervals, p values were determined using the Wald statistic. See also supplemental Table 2.

Figure 2.. AAbs isolated from ANA⁺⁺ Malians inhibit Pf growth in vitro.

(A) Schematic illustrating AAb isolation strategy. HEp-2 cell lysate-conjugated resin was used to affinity-purify AAbs from plasma. Validation and quality control of AAb purifications shown in Fig. S1. (B) ANA⁺⁺ and ANA⁻ plasma (pools of n=6, ages 8–12 yr) (dilution 1:40) and IgG AAb fractions obtained from ANA⁺⁺ or ANA⁻ plasma pools (1:40) were used for ANA testing. (C) In vitro growth inhibitory activity of total IgG and paired IgG AAbs purified malaria-exposed ANA⁺⁺ individuals (ages 9–12 years, n=16). All IgG samples were tested at 0.4 mg/mL (~1:5 dilution of IgG AAbs). Positive control polyclonal rabbit anti-PfAMA1-C2 . Statistical analysis: unpaired nonparametric Mann-Whitney test. Cumulative data of four independent experiments are shown. See also Fig. S2.

Figure 3.. Purified Malian AAbs bind to whole Pf-iRBCs and target key invasion proteins.

(A-B) Pf-3D7 trophozoite and schizont stages were incubated with ANA⁺⁺ and ANA⁻plasma (pools of n=6, ages 8–12 yr) at 1:500 (A), as well as with the corresponding AAb fractions (1:300) or malaria-naïve U.S. plasma (pool of n=20, mean age 56±SD 9.5, 1:500) and detected with anti-human IgG (green) and nuclear stain DAPI (blue). (B). A-B size bars: 10 μm. See also Fig. S3. (C-D) Pf-3D7 trophozoite (top panel) and schizont (lower panel) stages were co-stained with ANA⁺⁺ IgG AAbs, detected with anti-human IgG (pools of n=6, ages 8–12 yr) (green) and anti-Band3 antibody (magenta) (C), or anti-KAHRP antibody (magenta) (D) and nuclear stain DAPI (blue). C-D size bars: 5 μm. A-D: images are representative of three experimental replicates using three separate plasma pools for AAb purification. (E) Pf parasites co-stained with a serum pool obtained from ANA⁺⁺ SLE-diagnosed U.S. adults (n=5; 1:500) (green), anti-Band3 antibody (magenta) and nuclear stain DAPI (blue). Size bar: 5 μm. (F) Heatmap visualizes reactivity to 668 Pf proteins. Tested samples included total IgG from malaria-naïve healthy U.S. individuals (“healthy”, n=5), total IgG from malaria-naïve SLE-diagnosed ANA⁺⁺ U.S. individuals (“SLE”, n=5), total IgG purified from malaria-exposed ANA⁺⁺ individuals and corresponding IgG AAbs (ages 9–12 years; n=10), all tested at 3.3 ng/well. Pf proteins were ordered by average reactivity to Malian AAbs. 95^th percentile (representing 35 proteins) is highlighted and selected identified proteins are shown in the insert. See also Table S3.

Figure 4.. Healthy ANA⁺⁺ Malians have AAbs against a broad spectrum of autoantigens.

(A) 462 healthy Malians were profiled for IgG AAb responses to 123 autoantigens. Heat map depicts variance-stabilized Ab-scores, which were centered across rows. Within ANA⁻ (n=213), ANA⁺ (n=131) or ANA⁺⁺ (n=118) groups, samples were sorted by increasing age from left to right. Autoantigens are sorted by decreasing median reactivity in the ANA⁺⁺ group. Scale bar values reflect the range of Ab scores. (B) Volcano plot illustrating differences in IgG reactivity between ANA⁻ and ANA⁺⁺ groups. Significance (p-values determined using Empirical Bayes Statistics) versus log fold-change is plotted. 33 autoantigens with differential reactivity between ANA⁻ and ANA⁺⁺ groups (FC>0.5; p<0.05) were identified. (C) List displays 33 autoantigens with significant differential reactivity and FC>0.5, p value<1×10^–12. (D) Frequency of staining patterns observed in ANA⁺⁺ individuals (n=71) in the indirect immunofluorescence assay utilizing HEp-2 cell substrates. See also Table S4.