Structure and function of an aromatic ensemble that restricts the dynamics of the hydrophobic core of a designed helix-loop-helix dimer

GTD-43, a polypeptide with 43 amino acid residues, has been designed to fold into a hairpin helix-loop-helix motif that can dimerize to form a four-helix bundle. The mean residue ellipticity is -21000 deg cm(2) dmol(-1), and the magnitude is independent of pH in the interval of 2.4-4.6. The chemical shift dispersion of the amide proton and methyl regions of the H-1 NMR spectrum is only slightly less than that of Interleukin-4, a naturally occurring four-helix bundle, and much larger than those reported for molten globule structures. The temperature dependence of the H-1 NMR spectrum shows that GTD-43 is in slow exchange on thr NMR time scale and that the temperature interval for thermal denaturation is narrow. These spectroscopic probes show that GTD-43 exhibits remarkable structural characteristics for a de novo designed four-helix bundle, and the structural basis for the observed properties has now been determined. GTD-43 folds into a symmetric dimer, and the secondary structure as well as the formation of the hairpin helix-loop-helix motif has been established from NMR spectroscopy, GTD-43 has been designed to form complementary hydrophobic interfaces between the helices, and an aromatic ensemble consisting of Phe10, Trp13, and Phe34 has been incorporated in the interior of The hydrophobic core to restrict the dynamics and give rise to a well-defined tertiary structure. The interaction between the residues in the aromatic ensemble has been established from the observation of a number of long-range NOEs. The formation of a well-defined tertiary structure in the region of the Phe-Trp-Phe residues has been demonstrated by slow amide proton exchange rates and Trp fluorescence. This is the first report of a designed hydrophobic core in a four-helix bundle based on aromatic residues, and the structural property of GTD-43 suggests that the design principle is of general applicability in the de novo design of proteins.