Ancylostoma ceylanicum excretory-secretory protein 2 adopts a netrin-like fold and defines a novel family of nematode proteins

Abstract

Hookworms are human parasites that have devastating effects on global health, particularly in underdeveloped countries. Ancylostoma ceylanicum infects humans and animals, making it a useful model organism to study disease pathogenesis. A. ceylanicum excretory-secretory protein 2 (AceES-2), a highly immunoreactive molecule secreted by adult worms at the site of intestinal attachment, is partially protective when administered as a mucosal vaccine against hookworm anemia. The crystal structure of AceES-2 determined at 1.75 Å resolution shows that it adopts a netrin-like fold similar to that found in tissue inhibitors of matrix metalloproteases (TIMPs) and in complement factors C3 and C5. However, recombinant AceES-2 does not significantly inhibit the 10 most abundant human matrix metalloproteases or complement-mediated cell lysis. The presence of a highly acidic surface on AceES-2 suggests that it may function as a cytokine decoy receptor. Several small nematode proteins that have been annotated as TIMPs or netrin-domain-containing proteins display sequence homology in structurally important regions of AceES-2's netrin-like fold. Together, our results suggest that AceES-2 defines a novel family of nematode netrin-like proteins, which may function to modulate the host immune response to hookworm and other parasites.

Fig. 1

Overall structure of AceES-2. (a) AceES-2 adopts an NTR-domain fold with a five-stranded β-sheet (blue) capped by long loops (grey). N- and C-terminal α-helices (orange) lie on one face of the protein. Intramolecular disulfide bonds (Cys3-Cys62 and Cys77-Cys84) are shown in yellow, and the nonnative N-terminal portion of recombinant AceES-2 (residues -3 to 0) is shown in pink. The position of Glu1, the N-terminal residue of endogenous AceES-2, is labeled “N”. (b) Schematic of the secondary structure created with TopDraw. (c) Solvent accessible surface representation colored by electrostatic potential, calculated with APBS assuming physiological ionic strength. The units in the color bar are k_BT e_c^-1, where T = 310 K, k_B is the Boltzmann constant, and e_c is the charge of an electron. The view is rotated approximately 155° along the y-axis relative to Fig. 1a. Crystals of rAceES-2 grew by hanging drop vapor diffusion at 4°C. rAceES-2 at 24 mg/ml in 10 mM HEPES 7.5, 0.2 M KCl, 0-5% sucrose was mixed with an equal volume of 0.1 M HEPES pH 6.8 and 60% MPD. Lead derivative crystals were obtained by crystallizing rAceES-2 under the same conditions plus 1 mM Pb(NO₃)₂. Crystals were frozen in liquid nitrogen. Lead derivative crystals belonged to space group C222₁ with three molecules per asymmetric unit, while native crystals belonged to space group P4₃ with seven molecules per asymmetric unit. The structure was determined by single anomalous diffraction (SAD) using a lead derivative crystal. Data were collected at 100 K at the experimentally determined peak anomalous diffraction wavelength of lead (0.9496 Å). Data processing was performed using HKL2000. Heavy atom sites were located and experimental phases were calculated using SHELX and HKL2MAP. Buccaneer in the CCP4 suite was used for automated model building. In order to take advantage of the higher resolution native data the initial model was placed in the native crystal by molecular replacement with Phaser. Six molecules were located in the asymmetric unit of the native crystal, and the model was improved with rounds of model building, refinement and water placement using Coot and REFMAC5. Initial refinement statistics indicated that the native data were partially twinned. Amplitude-based twin refinement detected 28% merohedral twinning and revealed a seventh molecule in the asymmetric unit in the electron density. Final refinement included optimization of twin and TLS parameters to model anisotropic displacements within the protein, yielding R_work and R_free of 14.6% and 17.5% respectively. 90% of all residues are in the most favored regions of the Ramachandran plot, and three residues (Ala-31 in subunits B, D and E) are in disallowed regions. See Table S1 for data collection and refinement statistics.

Fig. 2

The structure of AceES-2 resembles the NTR domains of TIMPs and the C345C domain of complement C3. (a) Overall structure of AceES-2 (blue). The view is rotated 180° along the y-axis relative to Fig. 1a. (b) The NTR domain of TIMP-1 (green) in complex with MMP-14 (grey, PDB 3MA2 chains C, D). (c) The NTR domain of TIMP-3 (red) in complex with ADAM-17/TACE (grey, PDB 3CKI). (d) The C345C domain of cobra venom factor, a complement C3b analogue, in complex with factor B (PDB 3HRZ). The C345C domain (chain C, residues 1469-1620) is shown in pink; the rest of the structure is shown in grey. Disulfide-bonded cysteines within each NTR domain are shown in yellow. Figures 1a, 1c and 2 were created with Pymol.

Fig. 3

AceES-2 defines a family of nematode single domain NTR/TIMP proteins. The signal sequences have been omitted. The AceES-2 secondary structure is shown at the top and the four AceES-2 cysteines involved in disulfide bonds are labeled according to which disulfide bond they belong to. Strictly conserved residues are highlighted in red, substantially conserved residues in yellow. The sequence accession numbers are: AceES-2 Q6R7N7; A. duodenale TIMP ABP88131.1; A. ceylanicum TIMP CB176262; A. caninum TIMP-1 Q963I8; A. caninum TIMP-2 ACB13195.1; H. glycines hypothetical esophageal gland cell secretory protein 12 (SP12) Q9NDF1; C. elegans TIMP Q21265; C. elegans NTR Q21267.

Fig. 4

MMP activity and complement-mediated cell lysis inhibition assays with rAceES-2. (a) A colorimetric kit was used to test the ability of rAceES-2 to inhibit cleavage of a thiopeptide substrate by human MMPs 1, 2, 3, 7, 8, 9, 10, 12, 13 and 14 (Enzo Life Sciences). The cleavage product reacts with 5,5’-dithiobis(2-nitrobenzoic acid) to yield 2-nitro-5-thiobenzoic acid, which can be detected by absorbance at 412 nm. Final MMP concentrations in the reactions were 17-126 nM according the manufacturer's protocol. AceES-2 with either N- or C-terminal histidine tags (NHis and CHis, respectively) do not significantly inhibit any of MMPs tested. Shown is one representative experiment using 2.0 μM rAceES-2, which corresponds to a rAceES-2:MMP molar ratio of between 15:1 and 115:1, depending on the concentration of MMP used. Similar results where obtained using up to 20 μM rAceES-2. The general peptide inhibitor NNGH was used as an inhibitory control. (b) The ability of rAceES-2 to complement pathways was tested using standard erythrocyte lysis assays (see Supplementary Data for detailed methods). rAceES-2 does not significantly inhibit complement-mediated lysis of sheep erythrocytes in the classical pathway relative to a buffer control. Bovine serum albumen (BSA) at 50 μM inhibits this pathway non-specifically. (c) rAceES-2 does not significantly inhibit complement-mediated lysis of rabbit erythrocytes in the alternative pathway.