H. Xie et al., Structure, biological activity and membrane partitioning of analogs of theisoprenylated a-factor mating peptide of Saccharomyces cerevisiae, J PEPT RES, 55(5), 2000, pp. 372-383
Previous biochemical investigations on the Saccharomyces cerevisiae a-facto
r indicated that this lipopeptide pheromone [YIIKGVFWDPAC(farnesyl)OMe] mig
ht adopt a type II beta-turn at positions 4 and 5 of the peptide sequence.
To test this hypothesis, we synthesized five analogs of a-factor, in which
residues at positions 4 and 5 were replaced with: L-Pro(4)(I); D-Pro(4)(II)
; L-Pro(4)-D-Ala(5)(III); D-Pro(4)-L-Ala(5)(IV); or Nle(4)(V). Analogs were
purified to > 99% homogeneity as evidenced by HPLC and TLC and were charac
terized by mass spectrometry and amino acid analysis. Using a growth arrest
assay the conformationally restricted a-factor analogs I and III were foun
d to be almost 50-fold more active than the diastereometric homologs II and
IV and were equally active to wild-type a-factor. Replacement of Lys(4) wi
th the isosteric Nle(4) almost abolished the activity of the pheromone. Thu
s, the incorporation of residues that promote a type II beta-turn compensat
ed for the loss of the favorable contribution of the Lys4 side chain to phe
romone activity. CD spectra on these peptides suggested that they were esse
ntially disordered in both TFE/H2O and in the presence of DMPC vesicles. Th
ere was no correlation between CD peak shape and biological activity. Using
fluorescence spectroscopy we measured the interaction of lipid vesicles wi
th these position 4 and 5 analogs as well as with three a-factor analogs wi
th a modified farnesyl group. The results indicated that modifications of b
oth the peptide sequence and the lipid moiety affect partitioning into lipi
d, and that no correlation existed between the propensity of a pheromone to
partition into the lipid and its biological activity.