computer modeling in molecular biology.pdf

46 D. 1 Osnuthorpe and r! K. C. Paulthe second generation of antagonists 100 to a 1000 times more active than the earliestanalogues and the present day antagonists a similar proportion more active than thesecond generation analogues.3.3.3 Structural Studies on LHRHThe early energy calculations on LHRH by Momany, published as two papers inJACS in 1976, supported a conformation for LHRH which was characterised by aTyr5-Gly6-Leu7-Arg* modified type 11' /3 turn [19, 201. Later calculations by othergroups also tended to support this conformation [21]. However, unequivocal NMRevidence for such a conformation of LHRH itself was not forthcoming though manygroups had worked on the problem. This however could be attributed to the solventin which NMR was performed (mainly water or D20) and also the conformationalvariety exhibited by the molecule. However, it was generally accepted that the lowenergy conformations of LHRH were characterised by the 5 -8 p-turn.In the design of antagonists therefore, it was thought essential to retain this conformationalfeature of LHRH as it was expected that a similar affinity to the receptorwould be conferred by the presence of the type 11' 5-8 8-turn. A common practicein designing peptide analogues is to design analogues with restricted conformationalspace. Cyclisation is one of the ways to achieve this restriction [22]. A set ofLHRH analogues, cyclised through the side chains of residues 5 and 8, was foundto be potent agonists [23]. The analogue: Ac-D-Phe'(C1)-D-Phe2(C1)-D-Trp3-Ser4-Glu5-D-Arg6-Leu7-Lys8-Prog-D-Ala'o-NH2 was found to be a highly potent LHRHantagonist. This antagonist was conformationally constrained to form a cyclic peptidebetween residues 5 and 8 in order to mimic the p-turn that was shown by energycalculations to be present in the native peptide.An exhaustive study of the conformational preferences of this analogue was carriedout using conformational search MD and minimisation techniques [24]. A conformationalsearch of the cyclic part of the analogue was carried out first, followedby minimisations of all the generated conformations that were consistent withcyclisation. Among the 22 conformations obtained after minimisation, only two hada &turn conformation at position 6-7, and these were =lo kcal/mol higher inenergy than the most stable conformation. In all other generated conformations thecentral two residues were in extended or y-turn conformations. A MD study of thisanalogue was performed next. The initial structure was based on the lowest energyconformation found for the cyclic part. Additional investigations of the conformationalpreferences of the N and C terminal of the peptide defined a plausible conformationfor 3 more residues. The others were set to an extended conformation. Afterabout 24 ps the peptide adopted a 8 sheet conformation, with a /3 turn betweenresidues 3-6. The evolution of the 8-turn 8 sheet conformation from an extended