Structure of the trehalose-6-phosphate phosphatase from Brugia malayi reveals key design principles for anthelmintic drugs

Abstract

Parasitic nematodes are responsible for devastating illnesses that plague many of the world's poorest populations indigenous to the tropical areas of developing nations. Among these diseases is lymphatic filariasis, a major cause of permanent and long-term disability. Proteins essential to nematodes that do not have mammalian counterparts represent targets for therapeutic inhibitor discovery. One promising target is trehalose-6-phosphate phosphatase (T6PP) from Brugia malayi. In the model nematode Caenorhabditis elegans, T6PP is essential for survival due to the toxic effect(s) of the accumulation of trehalose 6-phosphate. T6PP has also been shown to be essential in Mycobacterium tuberculosis. We determined the X-ray crystal structure of T6PP from B. malayi. The protein structure revealed a stabilizing N-terminal MIT-like domain and a catalytic C-terminal C2B-type HAD phosphatase fold. Structure-guided mutagenesis, combined with kinetic analyses using a designed competitive inhibitor, trehalose 6-sulfate, identified five residues important for binding and catalysis. This structure-function analysis along with computational mapping provided the basis for the proposed model of the T6PP-trehalose 6-phosphate complex. The model indicates a substrate-binding mode wherein shape complementarity and van der Waals interactions drive recognition. The mode of binding is in sharp contrast to the homolog sucrose-6-phosphate phosphatase where extensive hydrogen-bond interactions are made to the substrate. Together these results suggest that high-affinity inhibitors will be bi-dentate, taking advantage of substrate-like binding to the phosphoryl-binding pocket while simultaneously utilizing non-native binding to the trehalose pocket. The conservation of the key residues that enforce the shape of the substrate pocket in T6PP enzymes suggest that development of broad-range anthelmintic and antibacterial therapeutics employing this platform may be possible.

Figure 1. Schematic showing the two-step synthesis of trehalose.

Trehalose is made in a two-step process catalyzed by trehalose-6-phosphate synthase (TPS) and trehalose-6-phosphate phosphatase (T6PP).

Figure 2. Structure of B. malayi T6PP and ortholog.

Ribbon diagram of the X-ray crystal structure of T6PP from Brugia malayi with selected helices or strands labeled (PDB ID 4OFZ) with the MIT-like domain (green), connector region (purple) the catalytic core Rossmann-fold HAD domain (blue) and the HAD cap domain (gold) colored differentially (A). A single magnesium ion (magenta sphere) marks the active site. The T. acidophilum T6PP-like enzyme lacks the MIT-like domain found in B. malayi (B). An overlay of these enzymes reveals a slightly more closed cap orientation in the T. acidophylum enzyme, but a nearly identical conformation of the C1 loop (rmsd = 1.04 Å mainchain atoms) (C and above). This molecular figure and all others, unless otherwise noted, were generated with UCSF Chimera v1.8.

Figure 3. Comparison of the N-terminal MIT-like domain of T6PP with other MIT domains and its interaction with the HAD domain.

The PDBeFold Server was used to identify structures with similar folds to that of the MIT-like domain. An overlay is depicted (A) between the MIT-domains from B. malayi T6PP (green) and the AAA ATPases Vps4 from Sulfolobus solfataricus (blue; PDB ID 2V6Y) and Sulfolobus acidocaldarius (pink; PDB ID 2W2U). The interaction between the α1–α3 interface of the MIT-like domain and the C1-loop of the HAD core domain is highlighted (B). Thermal stability analysis of B. malayi Δ59-T6PP reveals a strong interaction between the MIT and HAD domains. Thermal melt analysis by CD reveals one transition (C), suggesting the two domains are co-dependent.

Figure 4. Putative substrate interacting residues in B. malayi analyzed by mutagenesis and kinetics.

Residues in the cap and core regions analyzed by mutagenesis are depicted as sticks and colored according to kinetic parameters: no effect on kinetic parameters (grey); catalytically inactive (blue); decreases in k_cat (green); significant changes in k_cat and K_I for a T6S substrate analog (red).

Figure 5. Proposed model of trehalose 6-phosphate in the active site of T6PP.

The FTMap server was used to identify hot spots where protein-substrate interactions may occur. Analysis of the T6PP enzyme from T. acidophilium (1U02) (A), and B. malayi (B) reveal hot spots near the interface of the cap and core domains. These hot spots are cradled by the structurally conserved C1-Loop. T6P was placed manually into the active site of T6PP by coordinating the Mg²⁺ cation with the phosphate group (C). The residues identified as important via mutagenesis and kinetics are labeled and can be seen in proximity to the trehalose moiety.

Figure 6. Comparison of ligand binding between sucrose-6-phosphate phosphatase and trehalose-6-phosphate phosphatase.

An analysis of the crystal structure of the S6PP-sucrose 6-phosphate (S6P) complex (PDB: 1U2T) and the model of the T6PP-trehalose 6-phosphate (T6P) complex revealed that the stereochemistry of the glycosidic bond might affect specificity. The cap can be found in a different position in the S6PP-S6P complex than either the crystal structure or cap-closed model of T6PP, affecting the size and shape of the active site cavity. An extensive hydrogen-bonding network exists between S6PP and S6P (black dashed lines), utilizing residues from both the cap domain (orange) and the core domain (blue) (A). Positioning of the cap in S6P may be impacted by the α(1→2)β glycosidic bond of sucrose 6-phosphate versus the α(1→1)α glycosidic bond in trehalose 6-phosphate. An overlay of T6P from our model and S6P reveals a putative clash between S6P and residues 254–265 (light grey loop) in T6PP (B) potentially explaining the lack of turnover or binding of S6P.

Figure 7. Two binding pockets in the T6PP enzyme.

Analysis of the hypothesized binding model for trehalose 6-phosphate and enzyme kinetics suggests that inhibitors should interact with two pockets in order to maximize interactions. Trehalose 6-phosphate (A) and trehalose 6-sulfate (B) presumably bind in the phosphoryl-binding and sugar-binding pockets while glucose 6-phosphate (C) and trehalose (D) interact in only one pocket.