M. Tarnowska et al., A MOLECULAR MECHANICS STUDY OF THE EFFECT OF SUBSTITUTION IN POSITION-1 ON THE CONFORMATIONAL SPACE OF THE OXYTOCIN VASOPRESSIN RING/, Journal of computer-aided molecular design, 7(6), 1993, pp. 699-719
The effect of the substitution in position 1 on the low-energy conform
ations of the oxytocin/vasopressin 20-membered ring was investigated b
y means of molecular mechanics. Three representative substitutions wer
e considered: (beta'-mercapto-beta,beta-dimethyl)propionic acid (Dmp),
'-mercapto-beta,beta-cyclopentamethylene)propionic acid (Cpp), both f
orming strong antagonists, and (alpha,alpha-dimethyl-beta-mercapto)pro
pionic acid (alpha-Dmp), forming analogs of strongly reduced biologica
l activity, with the beta-mercaptopropionic (Mpa) residue taken as ref
erence. Both ECEPP/2 (rigid valence geometry) and AMBER (flexible vale
nce geometry) force fields were employed in the calculations. Three ba
sic types of backbone conformations were taken into account which are
distinguished by the type of beta-turn at residues 3 and 4: beta 1/bet
a III, beta II, and beta I'/beta III', all types containing one or two
intra-annular hydrogen bonds. The allowed (ring-closed) disulfide- br
idge conformations were searched by an algorithm formulated in terms o
f scanning the disulfide-bridge torsional angle C-beta-SS-C-beta, The
ECEPP/2 and AMBER energies of the obtained conformations were found to
be in reasonable agreement. Two of the low-energy conformers of the [
Mpa(1)]-compound agreed very well with the cyclic part of the two conf
ormers found in the crystal structure of [Mpa(1)]-oxytocin. An analysi
s of the effect of beta-substitution on relative energies showed that
the conformations with the N-C'-CH2-CH2 (psi(1)') and C'CH2-CH2-S (chi
(1)';) angles of the first residue around (-100 degrees, 60 degrees) a
nd (100 degrees, -60 degrees) are not affected; this in most cases imp
lies a left-handed disulfide bridge. In the case of alpha-substitution
the allowed values of psi(1)', are close to +/- 60 degrees. This requ
irement, being in contradiction to the one concerning beta-substitutio
n, could explain the very low biological activity of the alpha-substit
uted analogs. The conformational preferences of substituted compounds
can largely be explained by the analysis of local interactions within
the first residue. Based on the selection of the conformations which a
re low in energy for both the reference and beta-substituted compounds
, two distinct types of possible binding conformations were proposed,
the first one being similar to the crystal conformer with a left-hande
d disulfide bridge, the second one having a right-handed bridge, but a
geometry different from that of the crystal conformer with the right-
handed bridge. The first type of disulfide-bridge arrangement is equal
ly favorable for both beta I/beta III and beta II types of backbone st
ructure, while the second one is allowed only for the beta II type of
backbone. No conformation of the beta I'/beta III' type has a low enou
gh energy to be considered as a possible binding conformation for all
of the active compounds studied in this work.