Dependence of the peptide amide III vibration on the phi dihedral angle.

Peptide and protein UV Raman spectra excited at ∼200 nm show strong enhancement of the peptide backbone amide I, II, and III and (C)CR-H vibrations.1-4 We recently demonstrated that these spectra could be analyzed to quantitatively determine secondary structure by modeling the measured spectra with a linear combination of basis spectra of the different secondary structure motifs.4 These Raman studies also demonstrated that the amide III vibration is the most sensitive to peptide conformation; for example, the amide III frequency shifts from 1235 cm-1 in the -sheet to ∼1300 cm-1 in the R-helix conformation.1-4 Since our recent investigations of the UV Raman spectra of peptides indicate that the vibrations of individual peptide bonds are in general uncoupled,5 it may be possible to approximate protein Raman spectra as the linear sum of the spectra of individual peptide bonds. To test this possibility we need to know the dependence of the individual peptide bond Raman spectra on local secondary structure, as well as on their side chain identity and geometry. Our recent observation, that the Raman basis spectra of oligopeptides have bandwidths significantly narrower than those observed for proteins, may simply result from a side-chain dependence of the amide Raman band frequencies or from a narrow distribution of peptide band geometry.6 As previously suggested this dependence could result from either an involvement of side-chain atom motion to the amide III vibrational normal coordinate or a side-chain dependence of the amide bond conformation.7 Some of the earliest studies of amide vibrations empirically demonstrated a strong amide III frequency dependence on the Ramachandran Ψ angle.8 In addition, the experimental results of Williams et al.9,10and theoretical calculations showed an amide III frequency dependence on the Ramachandran Φ angle. They calculated the amide I and III potential energy distributions and the vibrational frequencies for the -sheet and R-helix conformations of dipeptides. For these conformations, they found no systematic dependence of the frequencies and potential energy distribution on side-chain molecular weight. Instead, they found that the amide III frequency shifts resulted from side-chain conformational preferences for particular dihedral angles. Their calculations demonstrated a Φ angle dependence for the magnitude of the contribution of N-H bending and methine proton bending to the amide III vibrational potential energy distribution. To further investigate the Φ angular dependence of the amide III frequency, we have examined the UV resonance Raman spectra of model dipeptides with side chains which are known to have a span of Φ angle preferences. These Φ angle preferences were calculated from Serrano’s11 recent NMR study of the side-chain Φ angle preference in proteins. It should be pointed out that the Φ angle conformational preferences are mainly determined by the side-chain identity and are independent of whether the side chain occurs in a random coil or in an ordered region of the protein.11 We calculated the average side-chain Φ angle from Serrano’s tables by weighting the different Φ angles by their relative probability of occurrence. We measured the UV Raman spectra of a series of acetylated amino acid esters to prevent obfuscation by a spectral dependence on charged penultimate groups. Figure 1 shows the UV Raman spectra of the amide III bands of different side-chain derivatives Ac-X-OCH3, where X is ala, lys, leu, ile, and val. The amide III frequency in this sequence monotonically downshifts from 1322 to 1313 cm-1. These derivatives should show only a Φ dihedral angle conformational dependence because the amide is not linked to a second amide group, and the penultimate CH3 group is 3-fold symmetric with respect to rotation around the CH3-C(O) bond. As suggested by William et al.,9 the amide III frequency does not depend on the side-chain mass, but linearly depends on the side-chain preferences for particular Φ angles11-14 (Figure 2, diamonds). Ac-ala-OCH3, which has a preference for the least negative Φ angle (-78°), has the highest amide III frequency, while Ac-val-OCH3, which has a preference for the most negative Φ angle (-96°, and prefers the -structure11-14), shows the lowest amide III frequency. The other side-chain data fall between that of ala and val. To support these empirical results, which show a Φ angle dependence of the amide III frequency, and to verify that this dependence does not result from mixing in of side chain atomic motion into the amide III vibration, we applied Gaussian 9815 density functional theory (DFT) calculations at the B3YLP 6-31G* To whom correspondence should be addressed. (1) Dudik, J. M.; Johnson, C. R.; Asher, S. A. J. Phys. Chem. 1985, 89, 3805-3814. (2) Song, S.; Asher, S. A. J. Am. Chem. Soc. 1989, 111, 4295-4305. (3) Holtz, J. S. W.; Holtz, J. H.; Chi, Z.; Asher, S. A. Biophys. J. 1999, 76, 3227-3234. (4) Chi, Z.; Chen X. G.; Holtz J. S. W.; Asher, S. A. Biochemistry 1998, 37, 2854-2864. (5) Mix, G.; Schweitzer-Stenner, R.; Asher, S. A. J. Am. Chem. Soc. 2000, 122, 9028-9029. (6) Lednev, I. K.; Karnoup, A. S.; Sparrow, M. C.; Asher, S. A. J. Am. Chem. Soc. 1999, 121, 8074-8086. (7) Krimm, S.; Bandekar, J. AdV. Protein Chem. 1986, 38, 183-364. (8) Lord, R. C. Appl. Spectrosc. 1977, 31, 187-194. (9) Williams, R. W.; Weaver, J. L.; Lowrey, A. H. Biopolymers 1990, 30, 599-608. (10) Weaver, J. L.; Williams, R. W. Biopolymers 1990, 30, 593-598. (11) Serrano, L. J. Mol. Biol. 1995, 254, 322-333. (12) Blaber, M.; Zhang, X.-J.; Matthews, B. W. Science 1993, 260, 16371640. (13) Chou, P. Y.; Fasman, G. D. Biochemistry 1974, 13, 211-222. (14) Koehl, P.; Levitt, M. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 1252412529. Figure 1. 204 nm excited UV RR spectra of peptides (10 mM, pH 6). The instrumentation is described in ref 6. The samples were measured in a temperature-controlled free-surface flowing stream. An accumulation time of 25 min was used for each spectrum. Spectral resolution was 8 cm-1. 7433 J. Am. Chem. Soc. 2001, 123, 7433-7434