Structural and functional difference of pheromone binding proteins in discriminating chemicals in the gypsy moth, Lymantria dispar

Abstract

Pheromone-binding proteins (PBPs) of the gypsy moth, Lymantria dispar L., play an important role in olfaction. Here structures of PBPs were first built by Homology Modeling, and each model of PBPs had seven α-helices and a large hydrophobic cavity including 25 residues for PBP1 and 30 residues for PBP2. Three potential semiochemicals were first screened by CDOCKER program based on the PBP models and chemical database. These chemicals were Palmitic acid n-butyl ester (Pal), Bis(3,4-epoxycyclohexylmethyl) adipate (Bis), L-trans-epoxysuccinyl-isoleucyl-proline methyl ester propylamide (CA-074). The analysis of chemicals docking the proteins showed one hydrogen bond was established between the residues Lys94 and (+)-Disparlure ((+)-D), and л-л interactions were present between Phe36 of PBP1 and (+)-D. The Lys94 of PBP1 formed two and three hydrogen bonds with Bis and CA-074, respectively. There was no residue of PBP2 interacting with these four chemicals except Bis forming one hydrogen bond with Lys121. After simulating the conformational changes of LdisPBPs at pH7.3 and 5.5 by constant pH molecular dynamics simulation in implicit solvent, the N-terminal sequences of PBPs was unfolded, only having five α-helices, and PBP2 had larger binding pocket at 7.3 than PBP1. To investigate the changes of α-helices at different pH, far-UV and near-UV circular dichroism showed PBPs consist of α-helices, and the tertiary structures of PBP1 and PBP2 were influenced at pH7.3 and 5.5. The fluorescence binding assay indicated that PBP1 and PBP2 have similarly binding affinity to (+)-D at pH 5.5 and 7.3, respectively. At pH 5.5, the dissociation constant of the complex between PBP1 and 2-decyl-1-oxaspiro [2.2] pentane (OXP1) was 0.68 ± 0.01 μM, for (+)-D was 5.32 ± 0.11 μM, while PBP2 with OXP1 and (+)-D were 1.88 ± 0.02 μM and 5.54 ± 0.04 μM, respectively. Three chemicals screened had higher affinity to PBP1 than (+)-D except Pal at pH5.5, and had lower affinity than (+)-D at pH7.3. To PBP2, these chemicals had lower affinity than the sex pheromone except Bis at pH 5.5 and pH 7.3. Only PBP1 had higher affinity with Sal than the sex pheromone at pH 5.5. Therefore, the structures of PBP1 and PBP2 had different changes at pH5.5 and 7.3, showing different affinity to chemicals. This study helps understanding the role of PBPs as well as in developing more efficient chemicals for pest control.

Keywords: Lymantria dispar; bioinformatics; discrimination; pheromone-binding protein; semiochemicals.

Figure 1

Sequence alignment of LdisPBP1, LdisPBP2, BmorPBP and ApolPBP. Identical residues were highlighted in white letters with a red background, residues with similar physico-chemical properties were shown in red letters. Aligned segments were framed in blue. The six conserved cysteines were marked with 1, 2 and 3, the same number indicating two cysteines connected by a disulfide bond.

Figure 2

Superimposed structures of ApolPBP(purple) and LdisPBP1 (A, green) and PBP2 (B, green). The NMR structure of the template ApolPBP was analyzed at pH 6.3. The models of LdisPBPs presented seven α-helices.

Figure 3

Molecular docking of LdisPBPs. The residues that were determined to be important for ligand binding were represented as stick models. Hydrogen bonds were shown as dashed lines in green. (A) (+)-D formed the hydrogen bond with residues Lys94 of LdisPBP1, and there was л-л interaction with residues Phe36. (B) Bis (purple) and CA-074 (blue) respectively formed two and three hydrogen bonds with residues Lys94 of LdisPBP1, and Pal (green) without forming any interaction. (C) (+)-D bound the binding pocket of LdisPBP2, without the hydrogen bonds forming. (D) Only Bis (purple) formed one hydrogen bond with residues Lys121 of LdisPBP2.

Figure 4

Molecular dynamics of PBPs. The models were shown in red, and the conformational changes were shown at pH 7.3 (purple) and 5.5 (green), seven α-helices were labeled. (A) The first α-helix of PBP1 unfolded at pH 7.3 and 5.5. (B) The first α-helix of PBP2 also unfolded completely, and the second α-helix of PBP2 extended from the third, increasing the binding cavity.

Figure 5

Purification of PBP1 (A) and PBP2 (B) by gel filtration chromatography after eluting from Ni resin. Proteins from indicated fractions (F2 to Waste) were separated by 14% SDS-PAGE gel and stained with Coomassie blue.

Figure 6

MALDI-TOF analysis of PBP1 (A) and PBP2 (B) following digestion with trypsin. The ratios of m/z were shown on the top of peaks.

Figure 7

Far-UV CD spectra (A, B) of LdisPBP1 and LdisPBP2 were measured at pH 7.3 (black trace) and pH 5.5 (red trace). The behaviours of PBP1 and PBP2 at pH 7.3 and 5.5 were opposite. Near-UV CD spectra (C, D) of LdisPBP1 and LdisPBP2 were measured at pH 7.3 (black trace) and pH 5.5 (red trace). Disruption of hydrogen bonds that keep the C-terminus covering the binding pocket may account for the changes in the environment of aromatic residues (Phe, Tyr and Trp) at low pH and consequently the decrease in amplitude.

Figure 8

Binding curves and relative Scatchard plots of the fluorescence probe (1-NPN) binding to recombinant PBP1 and PBP2 of gypsy moth. A and C report the curves relative to LdisPBP1 at pH 7.3 and 5.5, B and D those relative to LdisPBP2 at pH 7.3 and 5.5.

Figure 9

Competitive binding curves of some ligands to LdisPBPs. Solutions containing LdisPBP1or LdisPBP2 at pH 5.5 or 7.3 and 1-NPN, both at 2 μM concentration, were titrated with increasing amounts of competing ligands. The fluorescence intensity of LdisPBP1 (A) and PBP2 (C) binding chemicals at pH 5.5 were increased at their low concentration, the competitive effect occurred at high concentration. The competitive effect was obvious for LdisPBP1(B) and PBP2 (D) at pH 7.3. The structures of the compounds are shown, and the abbreviations of the compounds are reported after their names.