Effects of protein glycosylation on structural variation, thermal stability and ligand specificities of two Tachypleus plasma lectins

Two lectins, Tachypleus plasma lectin-1 (TPL-1) and Tachypleus plasma lectin-2 (TPL-2), derived from hemolymph of Tachypleus tridentatus (horseshoe crab) recognize peptidoglycan (PGN) in Gram-positive bacteria and lipopolysaccharide (LPS) in Gram-negative bacteria, respectively (1, 2). They serve as pattern recognition receptors (PRRs) and exert important roles in innate immunity of host. Previously, we utilized yeast Pichia pastoris to express glycosylated TPL-1 and TPL-2 (yTPL-1 and yTPL-2) with PGN-binding and LPS-trapping activity, but expression yields of both proteins were low (3). In this study, I tried to overexpress non-glycosylated TPL-1 and TPL2 (EcTPL-1 and EcTPL-2) using prokaryotic system and compared their ligand specificities with glycosylated equivalents. Non-glycosylated EcTPL-1 lost the PGN-binding activity, and its affinity to N-acetylmonosaccharide was weaker compared to that of yTPL-1. EcTPL-2 specifically interacted with L-rhamnose (L-Rha) and LPS from P. aeruginosa, but other O-antigen-containing LPS from E. coli, effective ligands for yTPL-2, could not bind to EcTPL-2. Both EcTPL-1 and yTPL-1 were dominated by β–sheet structure, whereas EcTPL-1 contained more random coil. The secondary structure of EcTPL-2 was also predominant in a β–sheet conformation differing from more unordered yTPL-2. As evidenced by analytic ultracentrifugation (AUC), EcTPL-1 formed a dimer and yTPL-1 were mixtures of monomer and dimer. Both EcTPL-2 and yTPL-2 migrated as a dimer under non-reducing SDS-PAGE analysis, and dimeric EcTPL-2 further associated to become an octamer. In the thermal unfolding profiles of differential scanning calorimetry (DSC), both non-glycosylated TPLs have Tm values a little higher than glycosylated counterparts, suggesting that in the present case glycosylation modification of protein could not contribute to its thermal stability. igand-induced conformational change and variation of association status of proteins were further examined by circular dichroism (CD), AUC. The addition of N-acetylmuramic acid (MurNAc), a common ligand for both EcTPL-1 and yTPL-1, rendered EcTPL-1 a tetramer and induced monomerization of yTPL-1. Moreover, monomerization of yTPL-1 upon addition of muramyl-dipeptide correlated to more random coil structure. P. aeruginosa LPS-treated EcTPL-2 remained as an octamer and gained more compact β–sheet conformation. By contrast, E. coli LPS-bound yTPL-2 exhibited significant oligomerization and minor conformational change. Finally, we utilized isothermal titration calorimetry (ITC) and dynamic light scattering (DLS) to investigate the physical property of LPS-TPL-2 complex and corresponding thermodynamic scenario. TPL-2 possessed disaggregating effect to disrupt the hydrophobic force among LPS micelle, and the expense of energy might be compensated by entropy-driven protein-ligand interaction and further stabilized by the multivalency of oligomeric TPL-2. Overall, these studies demonstrate that glycosylation modification of TPLs does influence its ligand specificities and causes structural variations even though it does not change thermostability.