Solution structure and backbone dynamics of the human alpha 3-chain type VI collagen C-terminal Kunitz domain

The solution structure and backbone dynamics of the 58-residue C-terminal Kunitz domain fragment [alpha 3(VI)] of human alpha 3-chain type VI collagen has been studied by two-dimensional H-1-H-1 and H-1-N-15 nuclear magnetic resonance spectroscopy at 303 K. The solution structure is represented by an ensemble of 20 structures calculated with X-PLOR using 612 distance and 47 dihedral angle restraints. The distance restraints were obtained by a complete relaxation matrix analysis using MARDIGRAS. The root mean squared (rms) deviation is 0.91 Angstrom for the backbone atoms of the residues Thr2(8)-Gly12(18), Arg15(21)-Tyr35(41), and Gly40(46)-Pro57(63). The central beta-sheet [residues Ile18(24)-Tyr35(41)] and the C-terminal alpha-helix [residues Gln48(54)-Cys55(61)] are better defined with a backbone rms deviation of 0.46 Angstrom. The solution structure of alpha 3(VI) is virtually identical to the crystal structure of alpha 3(VI) and to the solution structure of bovine pancreatic trypsin inhibitor (BPTI). The N-15 spin-lattice and spin-spin relaxation rates and the H-1-N-15 heteronuclear nuclear Overhauser enhancement (NOE) were analyzed using both the ''model-free'' formalism [Lipari, G., & Szabo, A. (1982) J. Am. Chem. Soc. 104, 4546-4559 and 4559-4570] and the reduced spectral density mapping procedure [Farrow, N. A., Szabo, A., Torchia, D. A., & Kay, L. E. (1995) J. Biomol. NMR 6, 153-162]. The results obtained from the ''model-free'' analysis include an overall correlation time tau(c), of 3.00 ns and backbone order parameters S-2 in the range from 0.28 to 0.93. The necessity of including an exchange term in the analysis of the relaxation data from 14 residues indicated that these residues are involved in motions on the micro-to millisecond time scale. The majority of the 14 residues are located in the vicinity of the Cys14(20)-Cys38(44) disulfide bond, suggesting the presence of a disulfide bond isomerization similar to the one observed in BPTI [Otting, G., Liepinsh, E., & Wuthrich, K. (1993) Biochemistry 32, 3571-3582]. It is suggested that this disulfide bond isomerization is the main reason for the surprisingly small effect on trypsin inhibition observed when Thr13(19) of alpha 3(VI) is substituted with Pro.