Ligand binding properties of two Brugia malayi fatty acid and retinol (FAR) binding proteins and their vaccine efficacies against challenge infection in gerbils

Abstract

Parasitic nematodes produce an unusual class of fatty acid and retinol (FAR)-binding proteins that may scavenge host fatty acids and retinoids. Two FARs from Brugia malayi (Bm-FAR-1 and Bm-FAR-2) were expressed as recombinant proteins, and their ligand binding, structural characteristics, and immunogenicities examined. Circular dichroism showed that rBm-FAR-1 and rBm-FAR-2 are similarly rich in α-helix structure. Unexpectedly, however, their lipid binding activities were found to be readily differentiated. Both FARs bound retinol and cis-parinaric acid similarly, but, while rBm-FAR-1 induced a dramatic increase in fluorescence emission and blue shift in peak emission by the fluorophore-tagged fatty acid (dansyl-undecanoic acid), rBm-FAR-2 did not. Recombinant forms of the related proteins from Onchocerca volvulus, rOv-FAR-1 and rOv-FAR-2, were found to be similarly distinguishable. This is the first FAR-2 protein from parasitic nematodes that is being characterized. The relative protein abundance of Bm-FAR-1 was higher than Bm-FAR-2 in the lysates of different developmental stages of B. malayi. Both FAR proteins were targets of strong IgG1, IgG3 and IgE antibody in infected individuals and individuals who were classified as endemic normal or putatively immune. In a B. malayi infection model in gerbils, immunization with rBm-FAR-1 and rBm-FAR-2 formulated in a water-in-oil-emulsion (®Montanide-720) or alum elicited high titers of antigen-specific IgG, but only gerbils immunized with rBm-FAR-1 formulated with the former produced a statistically significant reduction in adult worms (68%) following challenge with B. malayi infective larvae. These results suggest that FAR proteins may play important roles in the survival of filarial nematodes in the host, and represent potential candidates for vaccine development against lymphatic filariasis and related filarial infections.

Fig 1. Sequence and structural analysis of Bm-FAR-1 and Bm-FAR-2.

(A) Phylogenetic tree of Bm-FAR-1 and Bm-FAR-2 and their homologues in other nematodes, with bootstrap values (B) Structure based sequence comparison of Bm-FAR-1 and Bm-FAR-2, as well as Ov-FAR-1 and Ov-FAR-2, with Na-FAR-1 and Ce-FAR-7. The sequences are aligned with Clustal Omega and the secondary structural features are illustrated with the coordinates of Na-FAR-1 and Ce-FAR-7 using ESPript [52]. The different secondary structure elements shown are alpha helices (α), 3₁₀-helices (η), beta strands (arrows), and beta turns (TT). Identical residues are shown in solid red, and conserved residues are in red. Black box indicates the conserved casein kinase II phosphorylation site. The N-glycosylation site at 53-NFS for Bm-FAR-2 and Ov-FAR-2 is underlined. The amino acid sequence identity of orthologues between B. malayi and O. volvulus is shown at the end of sequence alignment. (C) Predicted tertiary structures of Bm-FAR-1 and Bm-FAR-2 showing retinol-binding pocket (P1) and fatty acid binding pocket (P2). The palmitate molecules from the Na-FAR-1 structure are modeled into the pockets as magenta spheres. Top panel shows transparent surfaces of the models and the retinol-binding pocket P1 is formed by helices α2 and α6, which are well conserved. The fatty acid binding cavity P2 is formed by helices α4 and α5, which are poorly conserved. The bottom panel of the solid surface plot shows that the retinol-binding pockets (P1) are similarly located and oriented in both models, whereas the fatty acid binding pockets (P2) are smaller and less accessible in Bm-FAR-2.

Fig 2. Bubble plot showing the normalized spectral abundance levels of BmFAR-1 and BmFAR-2 across the mature microfilaria (MF), immature microfilariae (UTMF), L3 larvae, in-vitro derived L4 larvae (L4), adult female (AF) and male (AM) worms based on proteomic analysis.

The size of the bubble is proportional to the abundance of Bm-FAR-1.

Fig 3. SDS-PAGE of yeast expressed rBm-FAR-1 and E. coli expressed rBm-FAR-2 and their binding to alum (Rehydrogel).

Two μg of recombinant proteins were loaded in each lane (red arrow). The same amount of recombinant proteins was absorbed with Rehydrogel and the absorbed supernatant was loaded to adjacent well.

Fig 4. Circular dichroism spectra of rBm-FAR-1, rBm-FAR-2 and rOv-FAR-2 in 20 mM Tris, 20 mM NaCl, pH 6.8 TBS, 25 ^oC.

Fig 5. Lipid binding properties of rBm-FAR-1 and rBm-FAR-2.

Fluorescence emission spectra were recorded, and spectrofluorimetry conditions set, as described in Materials and Methods. (A) Retinol in ethanol was added to 2 ml PBS, pH 7.2 (to create an approximately 4 μM solution of retinol) alone, or 2 ml PBS containing rBm-FAR-1 or rBm-FAR-2. Both proteins clearly bind retinol. (B) When rBm-FAR-1 and rBm-FAR-2 were added to 2 ml of an approximately 1 μM solution of DAUDA in PBS, Bm-FAR-1 elicited a dramatic increase and blue shift in DAUDA fluorescence emission, whereas Bm-FAR-2 did so to a minimal extent. (C) Addition of oleic acid to pre-formed Bm-FAR-1:DAUDA complexes results in competitive displacement of DAUDA from the protein binding site, indicating that there exists a binding site with preference for natural fatty acids. (D) Both proteins bind cis-parinaric acid (cPnA), albeit yielding different amplitudes of change in cPnA emission, but this nevertheless indicates that both proteins bind fatty acids. Similar experiments were carried out with the orthologues of these proteins from the river blindness parasite O. volvulus (rOv-FAR-1 and rOv-FAR-2) and the results were directly comparable to results with the Bm-FARs (see S2 Fig).

Fig 6. IgG1, IgG3 and IgE antibody responses against Bm-FARs and Ov-FARs in human populations living in filariae endemic regions.

Anti-rBm-FAR-1 and rBm-FAR-2 IgG1 (A) and IgG3 (B) responses in children that were infected (INF) or uninfected (EN) who lived in Cook Islands, an endemic region for lymphatic filariasis. Anti-rOv-FAR-1 and rOv-FAR-2 IgG1 (C), IgG3 (D) and IgE (E) responses in putatively immune (PI) and infected (INF) individuals who lived in onchocerciasis highly endemic area in Cameroon. The ODs obtained with a pool of 10 control normal human sera samples at 1:100 (IgG1 and IgG3) or 1:50 (IgE) dilution were used to calculate the cutoff for each assay based on mean OD ± 3X SD (dotted line): 0.1 for anti-rBm-FAR-1 and anti-rBm-FAR-2 IgG1; 0.05 for anti-rBm-FAR-1 anti-rBm-FAR-2 IgG3; 0.135 for anti-rOv-FAR-1 and anti-rOv-FAR-2 IgG1; 0.08 for anti-rOv-FAR-1 and anti-rOv-FAR-2 IgG3; and 0.1 for anti-rOv-FAR-1 and anti-rOv-FAR-2 IgE. The cutoff was used to calculate the number of responders, individual having IgG1, IgG3 or IgE OD values more than mean + 3X SD of normal human sera. Analysis was done using the Mann-Whitney test.

Fig 7. Protective immunity induced by immunization of gerbils with rBm-FAR-1 and rBm-FAR-2 formulated with Montanide-720 or alum.

(A) Total worm count from gerbils (n = 20 per group, which are presented as a combination of two separate experiments) immunized with rBm-FAR-1 formulated with alum or Montanide-720 or adjuvant controls 42 days after being challenged with 100 B. malayi L3. (B) Titers of anti-Bm-FAR-1 IgG in sera of gerbils immunized rBm-FAR-1 formulated with alum or Montanide-720. (C) Total worm count from gerbils (n = 10 per group performed only once) immunized with rBm-FAR-2 formulated with alum or Montanide-720 or adjuvant controls. (D) Titers of anti-Bm-FAR-2 IgG in sera taken from gerbils immunized with rBm-FAR-2 formulated with alum or Montanide-720. The median in A and C is marked by a line. Geometric mean and 95% CI are marked in B and D.