The combination of molecular dynamics with crystallography for elucidatingprotein-ligand interactions: a case study involving peanut lectin complexes with T-antigen and lactose
Jv. Pratap et al., The combination of molecular dynamics with crystallography for elucidatingprotein-ligand interactions: a case study involving peanut lectin complexes with T-antigen and lactose, ACT CRYST D, 57, 2001, pp. 1584-1594
Peanut lectin binds T-antigen [Gal beta (1-3)GalNAc] with an order of magni
tude higher affinity than it binds the disaccharide lactose. The crystal st
ructures of the two complexes indicate that the higher affinity for T-antig
en is generated by two water bridges involving the acetamido group. Fresh c
alorimetric measurements on the two complexes have been carried out in the
temperature range 280-313 K. Four sets of nanosecond molecular-dynamics (MD
) simulations, two at 293 K and the other two at 313 K, were performed on e
ach of the two complexes. At each temperature, two somewhat different proto
cols were used to hydrate the complex in the two runs. Two MD runs under sl
ightly different conditions for each complex served to assess the reliabili
ty of the approach for exploring protein-ligand interactions. Enthalpies ba
sed on static calculations and on MD simulations favour complexation involv
ing T-antigen. The simulations also brought to light ensembles of direct an
d water-mediated protein-sugar interactions in both the cases. These ensemb
les provide a qualitative explanation for the temperature dependence of the
thermodynamic parameters of peanut lectin-T-antigen interaction and for th
e results of one of the two mutational studies on the lectin. They also sup
port the earlier conclusion that the increased affinity of peanut lectin fo
r T-antigen compared with that for lactose is primarily caused by additiona
l water bridges involving the acetamido group. The calculations provide a r
ationale for the observed sugar-binding affinity of one of the two availabl
e mutants. Detailed examination of the calculations point to the need for e
xercising caution in interpreting results of MD simulations: while long sim
ulations are not possible owing to computational reasons, it is desirable t
o carry out several short simulations with somewhat different initial condi
tions.