Gammaherpesvirus Co-infection with Malaria Suppresses Anti-parasitic Humoral Immunity

Abstract

Immunity to non-cerebral severe malaria is estimated to occur within 1-2 infections in areas of endemic transmission for Plasmodium falciparum. Yet, nearly 20% of infected children die annually as a result of severe malaria. Multiple risk factors are postulated to exacerbate malarial disease, one being co-infections with other pathogens. Children living in Sub-Saharan Africa are seropositive for Epstein Barr Virus (EBV) by the age of 6 months. This timing overlaps with the waning of protective maternal antibodies and susceptibility to primary Plasmodium infection. However, the impact of acute EBV infection on the generation of anti-malarial immunity is unknown. Using well established mouse models of infection, we show here that acute, but not latent murine gammaherpesvirus 68 (MHV68) infection suppresses the anti-malarial humoral response to a secondary malaria infection. Importantly, this resulted in the transformation of a non-lethal P. yoelii XNL infection into a lethal one; an outcome that is correlated with a defect in the maintenance of germinal center B cells and T follicular helper (Tfh) cells in the spleen. Furthermore, we have identified the MHV68 M2 protein as an important virus encoded protein that can: (i) suppress anti-MHV68 humoral responses during acute MHV68 infection; and (ii) plays a critical role in the observed suppression of anti-malarial humoral responses in the setting of co-infection. Notably, co-infection with an M2-null mutant MHV68 eliminates lethality of P. yoelii XNL. Collectively, our data demonstrates that an acute gammaherpesvirus infection can negatively impact the development of an anti-malarial immune response. This suggests that acute infection with EBV should be investigated as a risk factor for non-cerebral severe malaria in young children living in areas endemic for Plasmodium transmission.

Fig 1. MHV68 co-infection with the non-lethal P. yoelii XNL in C57BL/6 results in lethal malarial disease and suppressed Plasmodium specific IgG response.

(A) Timeline of infection. 6–8 week old C57BL/6 mice were infected with 1000 PFU of MHV68 on day -7 followed by infection with 10⁵pRBCs of non-lethal P. yoelii XNL or P. chabaudi AS. Infections consisted of 5 experimental groups: MHV68 + Plasmodium, Plasmodium, MHV68 or mock infected. Each experimental group consisted of n = 5 and was repeated twice. Animals were sacrificed at days 8, 12, 16 and 23 post P. yoelii XNL infection or day 7, 11, 15 and 23 post P. chabaudi AS infection for collection of spleen, lung and blood. (B) Survival analysis of animals co-infected with MHV68 and P. yoelii XNL or P. chabaudi AS. Total IgG and IgM levels in serum in (C) P. yoelii XNL (Day 23 IgG—P. yoelii vs co-infected: p<0.05 Mann Whitney U-test) or (D) P. chabaudi AS co-infection model (Day 11 IgG—P. chabaudi vs co-infected: p<0.05 Mann Whitney U-test). Parasite specific IgG levels in serum during (E) P. yoelii XNL (day 23 post infection, P. yoelii vs co-infected: p<0.05 Mann Whitney U-test) or (F) P. chabaudi AS (day 11 post infection, P. chabaudi vs co-infected: p<0.05 Mann Whitney U-test) co-infection.

Fig 2. P. yoelii XNL requires Plasmodium specific IgG response to clear primary peak of parasitemia.

(A) Percent parasitemia in the periphery during P. yoelii XNL (p<0.05; area under curve, Mann Whitney U-test) or P. chabaudi AS co-infection models (p>0.05; area under the curve, Mann Whitney U-test). (B) Anemia during P. yoelii XNL (p>0.05; area over curve, Mann Whitney U-test, P. yoelii vs. co-infected) or P. chabaudi AS co-infection (p>0.05; area over curve, Mann Whitney U-test, P. chabaudi vs. co-infected). (C) Percent parasitemia in periphery during infection of single P. yoelii XNL or P. chabaudi AS in C57BL/6 or μMT (B cell-deficient) mice. (D) Anemia during infection of single P. yoelii XNL or P. chabaudi AS in C57BL/6 or μMT mice.

Fig 3. MHV68 suppresses splenic B cell responses during co-infection with Plasmodium.

The timeline and experimental set up was identical to that shown in Fig 1A. (A) Absolute numbers of splenic GC B cell populations (B220+ GL7+ CD95+) during P. yoelii XNL and P. chabaudi AS co-infection models with representative gating strategy (Day 12 post P. yoelii or Day 15 post P. chabaudi; Plasmodium vs. co-infected, p<0.05, Mann Whitney U-test). (B) Absolute numbers of splenic plasma cell populations (CD3- B220int CD138+) during P. yoelii XNL AND P. chabaudi AS co-infection models with representative gating strategy (Day 12 post P. yoelii or Day 11 post P. chabaudi; Plasmodium vs. co-infected, p<0.05, Mann Whitney U-test). (C) Spleen section for mock infected, MHV68 infected, P. yoelii XNL infected and MHV68 and P. yoelii XNL co-infected animals at day 8 post infection with P. yoelii XNL (or day 15 post-infection with MHV68). Green: B220-FITC (B cells), Blue: GL7-AF660 (Germinal center B cells) and Red: CD3-AF568 (T cells).

Fig 4. MHV68 and Plasmodium co-infection results in defective splenic T follicular helper (Tfh) response.

The timeline and experimental set up was identical to that shown in Fig 1A. (A) Representative flow plots for gating strategies used to define the global Tfh population (CD4+ PD-1+ CXCR5+), germinal center Tfh (CD4+ GL7+ CXCR5+) and activated/antigen specific Tfh (CD4+ CD44+ PD-1+ CXCR5+). (B) Absolute values for all three Tfh subsets are plotted for the P. yoelii XNL (Day 23, all Tfh subsets, P. yoelii vs. co-infected, p<0.05 Mann Whitney U-test) or (C) P. chabaudi co-infection models at multiple time points (Day 23, all Tfh subsets, P. chabaudi vs. co-infected, p<0.05 Mann Whitney U-test).

Fig 5. Acute, but not latent, MHV68 infection results in suppressed humoral response.

(A) Timeline of infection. C57BL/6 mice were infected with 1000 PFU of MHV68 IN at day -60, -30, -15 or -7 and challenged with 10⁵ pRBCs on day 0. Absolute number of (B) splenic GC B cell (B220+ GL7+ CD95+) and plasma cell (CD3- B220int CD138+) populations at day 16 post P. yoelii XNL infection (For GC and PC: Day -7 and Day -15 co-infected vs. P. yoelii, Kruskal Wallis p<0.05; Dunn’s pairwise comparison test p<0.05/ Day -30 co-infected vs. P. yoelii, Kruskal Wallis p<0.05; Dunn’s pairwise comparison test p>0.05). (C) MHV68 and P. yoelii XNL specific IgG responses at day 16 post P. yoelii XNL infection (Day -7 and Day -15 co-infected vs. P. yoelii, Kruskal Wallis p<0.05; Dunn’s pairwise comparison test p<0.05/ Day -30 co-infected vs. P. yoelii, Kruskal Wallis p<0.05; Dunn’s pairwise comparison test p>0.05). (D) Global Tfh population (CD4+ PD-1+ CXCR5+), germinal center Tfh (CD4+ GL7+ CXCR5+) and activated/antigen specific Tfh (CD4+ CD44+ PD-1+ CXCR5+) in the spleen at day 16 post P. yoelii XNL infection.

Fig 6. The MHV68 M2 gene product is necessary for virus mediated humoral suppression and lethality during Plasmodium co-infection.

(A) MHV68 specific IgG titers from serum of animals infected with the MR (M2.Marker Rescue) or M2.Stop (ST, M2-null) viruses. Serum was collected and analyzed on days 7, 14 and 21 post infection with either virus (n = 10/ virus) (Day 21, MR vs. M2.Stop, Kruskal Wallis p<0.05; Dunn’s pairwise comparison test p<0.05). (B) P. yoelii XNL specific IgG response during P. yoelii XNL co-infection with either the M2.MR or M2.Stop virus. Serum was collected at day 20 post infection with P. yoelii XNL (WT + P. yoelii co-infected vs. P. yoelii, Kruskal Wallis p<0.05; Dunn’s pairwise comparison test p<0.05/ WT + P. yoelii co-infected vs. M2.Stop + P. yoelii, Kruskal Wallis p<0.05; Dunn’s pairwise comparison test p>0.05). (C) Survival curve during P. yoelii XNL co-infection with either the M2.MR or M2.Stop virus. Note: data representing P. yoelii XNL + MHV68 co-infection is the identical data set to that in Fig 1B. It was added in panel C for comparative purposes. (D) % parasitemia in the periphery during P. yoelii XNL, P. yoelii XNL +MR and P. yoelii XNL + M2.Stop infection.