SAG2A protein from Toxoplasma gondii interacts with both innate and adaptive immune compartments of infected hosts

Abstract

Background: Toxoplasma gondii is an intracellular parasite that causes relevant clinical disease in humans and animals. Several studies have been performed in order to understand the interactions between proteins of the parasite and host cells. SAG2A is a 22 kDa protein that is mainly found in the surface of tachyzoites. In the present work, our aim was to correlate the predicted three-dimensional structure of this protein with the immune system of infected hosts.

Methods: To accomplish our goals, we performed in silico analysis of the amino acid sequence of SAG2A, correlating the predictions with in vitro stimulation of antigen presenting cells and serological assays.

Results: Structure modeling predicts that SAG2A protein possesses an unfolded C-terminal end, which varies its conformation within distinct strain types of T. gondii. This structure within the protein shelters a known B-cell immunodominant epitope, which presents low identity with its closest phyllogenetically related protein, an orthologue predicted in Neospora caninum. In agreement with the in silico observations, sera of known T. gondii infected mice and goats recognized recombinant SAG2A, whereas no serological cross-reactivity was observed with samples from N. caninum animals. Additionally, the C-terminal end of the protein was able to down-modulate pro-inflammatory responses of activated macrophages and dendritic cells.

Conclusions: Altogether, we demonstrate herein that recombinant SAG2A protein from T. gondii is immunologically relevant in the host-parasite interface and may be targeted in therapeutic and diagnostic procedures designed against the infection.

Figure 1

Structural modeling of surface SAG2A protein and comparison with related Toxoplasma gondii antigens SAG1 and BSR4. Structural modeling of surface SAG2A protein and comparison with related Toxoplasma gondii antigens SAG1 and BSR4. Three-dimensional models of SAG-related surface proteins. The predicted model of the single domain of SAG2A, with its unfolded C-terminal end (red highlight). This feature is not observed in the structure of the external (D1) and internal domains (D2) of the related tachyzoite surface proteins SAG1, and in correlated bradyzoite antigen BSR4.

Figure 2

Anchorage site, carbon structure and charge distribution of SAG2A protein. (A) SAG2A anchorage in Toxoplasma gondii’s surface surface by glycosyl-phosphatidylinositol (GPI) was predicted to be in a leucin at position 169 of its amino acid sequence, located at the C-terminal end of SAG2A. (B) The modeled carbon structure of SAG2A evidences a disordered amino acid sequence, absent in the SAG1 and BSR4 proteins. (C) The C-terminal end of SAG2A presents a relevant hydrophobic portion, with distinct polar amino acids at positions 134 and 137.

Figure 3

Orthologue SAG2A from Neospora caninum presents distinct conformation of the C-terminal end and does not share immunodominant B cell-epitope with T. gondii. (A) The consensus tree of SAG2 and SAG1 protein sequences demonstrates higher proximity of SAG2A with its predicted orthologue from Neospora caninum (NcSAG2A). (B) Modeling of NcSAG2A suggests that overall structure of the orthologues is similar, although its loop is composed with beta-sheets instead of the largely disordered structure of T. gondii’s protein. Additionally, the B cell-epitope sequence found in SAG2A (purple highlight) is not present in its orthologue. (C) Sequence alignments produced by ClustalW with SAG2A orthologues demonstrate the percentage of identity and similarity between these proteins, found to be low in the loop sequence and practically nonexistent in the B cell-epitope region of the amino acid sequence. The unfolded C-terminal end of T. gondii SAG2A is highlighted in blue, which includes an immunodominant epitope NDGSSA highlighted in red. (D) Reactivity of IgG antibodies from naïve mice and mice experimentally infected with N. caninum or T. gondii against recombinant SAG2A protein of T. gondii. Sera reactivity was expressed as ELISA index (EI). (E) Recognition profile of recombinant SAG2A by serum samples from experimentally infected mice and naturally infected goats with T. gondii and N. caninum in Western Blot.

Figure 4

C-terminal end of SAG2A protein interacts with innate immune response. (A) Three-dimensional model of recombinant SAG2A (rSAG2A) and the truncated form of the protein (rSAG2A^∆135). Structural analysis shows that depletion of the C-terminal end did not affect the overall predicted three-dimensional structure or distribution of positive and negative charges along the protein. (B) Bone marrow-derived macrophages (BMMs) and (C) dendritic cells (BMDCs) were treated with different concentrations of rSAG2A and rSAG2A^∆135 for 48 and 24 h, prior to determination of nitrite (B) and IL-12p40 levels (C), respectively. As controls, cells were left untreated (RPMI) or exposed to LPS (1 μg/ml) in similar time spam. Results are presented as mean ±SEM. Dashed lines indicate mean values obtained for untreated BMMs. * Statistical significance (p < 0.05) in relation to untreated controls.

Figure 5

Strain-dependent SAG2A polymorphisms in the C-terminal end alter protein conformation and regulatory features. (A) The consensus tree of SAG2A from the three clonal strains (I, II, III) of Toxoplasma gondii and the orthologue expressed in Neospora caninum. Alignment of the immunodominant epitope region shows that type II SAG2A displays a single additional glycine (position 142) within its IUP domain sequence. (B) Modeling of SAG2A from type I/III strains with the epitope region highlighted within the C-terminal end. The addition of glycine is responsible for significant changes in the SAG2A structure in type II strains, creating a predicted coil. (C) Bone marrow-derived macrophages (BMMs) and (D) Bone marrow-derived dendritic cells (BMDCs) were previously exposed to different parasite:cell ratios (multiplicity of infection – MOI) with RH and Me49 strain tachyzoites, in the presence of A4D12 monoclonal antibody or irrelevant IgG for 24 h. BMMs activation were promoted by LPS (10 ng/ml) + IFN-γ (100 ng/ml), 48 h prior to determination of nitrite levels. BMDC activation was promoted by LPS (1 μg/ml), 24 h prior to determination of IL-12p40 levels. Results are presented as mean ±SEM. Dashed lines indicate mean values for experimental controls following description. * Statistical significance (p < 0.05) related to antibody treatments.