Supporting information for Maglio et al. (2003) Proc. Natl. Acad. Sci. USA, 10.1073/pnas.0730771100


Supporting Text

Guanidine (Gdn) Denaturation. Samples of Dueferri 1 (DF1), L13A-DF1, and L13G-DF1 were prepared from stock peptide solutions in H2O/20% CH3CN (0.1% trifluoroacetic acid), diluted to the desired final concentrations with 10 mM phosphate buffer (pH 5.5) at room temperature (1-cm path length cell). Mean residue ellipticities were calculated by using the equation [θ] = θobsd/10lcn, in which θobsd is the ellipticity measured in millidegrees, l is the length of the cell in centimeters, c is the concentration in moles per liter, and n is the number of residues in the peptide.

The thermodynamic stability of each peptide was determined by analyzing Gdn denaturation curves at two different peptide concentrations. In DF1, L13A-DF1, and L13G-DF1 the midpoint of the unfolding transition is shifted to higher denaturant concentrations as the protein concentration is increased. These concentration-dependent shifts in denaturation curves can be used to determine both the stability and the oligomerization state of a protein (1).Indeed, the data for the three proteins conform well to a dimerization-induced two-state folding model (Scheme 1) in which a folded dimer, F2, is in equilibrium with unfolded monomers, M.

GIF ImageScheme I

The data were analyzed by the linear extrapolation method using the following equation, for a mononomer–dimer equilibrium in which folding and dimer formation are thermodynamically linked (2, 3).

θobs = [(θu° + au [Gdn]) – (θf° + bf [Gdn])] {[e–2(ΔGo– m[Gdn])/RT + 8Pte–(ΔGo – m[Gdn])/RT)]1/2e–(ΔGo – m [Gdn] )/RT]}/4Pt + (θf° + bf [Gdn]),

in which θu°, θf°, au, and bf are the intercepts and the slopes of the pre- and posttransition baselines, respectively, ΔG0 is the free energy of dimerization and folding (1 M standard state) extrapolated to 0 M Gdn, and m is the linear change in ΔG0 with respect to the concentration of Gdn. The pretransition baseline was particularly well defined for the most stable protein, DF1, and the parameters obtained from this protein were used for the other two proteins, the pretransition baselines of which were not as well defined. Similarly, the posttransition baseline of the least stable variant, Leu-13-Gly, was used to fit the other two proteins. These baseline parameters were determined by globally fitting the data at two concentrations for the appropriate protein. ΔG0 and m were treated as global parameters, which were allowed to vary for multiple concentrations of a single protein.

NMR Spectroscopy and Structure Calculations. The solution conditions for NMR experiments were 1.0 mM protein concentration in 90% H2O/10% DMSO d6, pH 4.0. All NMR spectra were carried out at 298 K on a Bruker Avance 600 spectrometer operating at a nominal frequency of 600.13 MHz. To obtain 1H resonance assignments, a double quantum filtered correlated spectroscopyspectrum, a clean-total correlation spectroscopy spectrum with a mixing time of 60 ms, and two NOESYspectra with mixing times of 120 and 150 ms were recorded by using standard pulse sequences and phase cycling.All spectra were acquired in the phase-sensitive mode with quadrature detection by using the time-proportional phase-incrementation technique (TPPI) (4) in the F1 dimension. Data size was typically 512 ´ 2,048 complex points in t1 and t2 time domains, which was zero-filled to obtain 2,048 ´ 4,096 data points in the spectrum. Solvent suppression was achieved with watergate sequence (5).

The1H chemical shifts (in ppm) of apo-DF1 are reported in Table 3. The most relevant NMR data for secondary structure determination, 3JαN coupling constants, medium-range NOEs, and α-proton chemical shift indices (CSI), are reported in Fig. 7. Most of the 3JαN values are < 6 Hz and are consistent with an α-helix. 3JαN values > 6 Hz were observed for residues located around the loop region. A high number of sequential and medium-range connectivities such as dNN(i,i+1), dNN(i,i+2), dαN(i,i+3), dαN(i,i+4), and dαβ(i,i+3), were identified in the NOESY spectra as expected for a helical structure. Upfield-shifted α-proton resonances, relative to the corresponding random coil values, are indicative of α-helical conformation (6): they were observed in two long sequential stretches of residues, 1–23 (except Glu-10) and 27–46 (see CSI in Fig. 7). Downfield chemical shifts, which are typically found in β-structures, loop, and turns (6), were observed in the designed loop region for residues Val-24 and Leu-26. The good correlation between all the experimental data allows us to define two stable helical segments for apo-DF1 in solution: helices A (residues 7–22) and B (residues 27–46). The region between residues Val-24and Leu-26 constituted the loop between the two helical regions, as designed.

The amide proton chemical shifts are sensitive to hydrogen bonding and solvent exposure (6–8). As expected from the helical structure of DF1, the amide proton chemical shifts vary periodically along the sequence (see Fig. 8). A repeating 3.6-residue pattern is observed for residues 7–22 and 27–46, which provides additional support for the assignment of the helices.

Fig. 9 shows the rms deviation per residue in the dimer structure for the best 40 calculated structures.

  1. Boice, J. A., Dieckmann, G. R., DeGrado, W. F. & Fairman, R. (1996) Biochemistry35, 14480–14485.
  2. Gittelman, M. S. & Matthews, C. R. (1990) Biochemistry29, 7011–7020.
  3. Ghirlanda, G., Lear, J. D., Lombardi, A. & DeGrado, W. F. (1998) J. Mol. Biol. 281, 379–391.
  4. Marion, D., Ikura, K., Tschudin, R. & Bax, A. (1989) J. Magn. Reson.85, 393–399.
  5. Piotto, M., Saudek, V. & Sklenar, V. (1992) J. Biomol. NMR2, 661–665.
  6. Wishart, D. S., Sykes, B. D. & Richards, F. M. (1992) Biochemistry31, 1647–1651.
  7. Wagner, G., Pardi, A. & Wütrich, K. (1983) Eur. J. Biochem.137, 445–454.
  8. Kuntz, I. D., Kosen, P. A. & Craig, E. C. (1991) J. Am. Chem. Soc.113, 1406–1408.
  9. Goodman, E. M. & Kim, P. S. (1991) Biochemistry30, 11615–11620.