On the temperature dependence of amide I intensities of peptides in solution

doi:10.1016/j.vibspec.2008.07.004

Vibrational Spectroscopy

Volume 50, Issue 1, 26 May 2009, Pages 2-9

https://doi.org/10.1016/j.vibspec.2008.07.004 Get rights and content

Abstract

Temperature dependence of the amide I/I′ spectral intensities is investigated for N-methylacetamide (NMA) as a model compound for the peptide bond in D₂O and three organic solvents with different polarities (dimethyl sulfoxide (DMSO), acetonitrile and 1,4-dioxane). The total amide I/I′ intensity (dipole strength) systematically decreases in less polar solvents as well as with the increasing temperature. Decreased solvent polarity results in the narrowing of the amide I bandwidths, while increasing temperature predominantly reduces the peak absorbance, with only a small effect on the spectral width. In D₂O, the NMA amide I′ dipole strength decreases by 1.7 × 10⁻⁴ Debye²/deg, in DMSO, acetonitrile and 1,4-dioxane by 1.0 × 10⁻⁴ Debye²/deg. The amide I/I′ intensity variations in the non-protic solvents rule out hydrogen bonding as the sole source of these effects. The experimental NMA amide I dipole strengths in the organic solvents are accurately described by a simple theory based on the Onsager reaction field with temperature-dependent solvent dielectric constant, refractive index and the solute molecular cavity, which can be approximated using NMA density. Experimental results are compared to density functional theory (DFT) BPW91/cc-pVDZ/Onsager calculations. The computations significantly overestimate the absolute experimental amide I intensities, but comparison of the relative values underscores the importance of the temperature-dependent molecular cavity dimension (density) as well as the frequency-dependent response of the reaction field (index of refraction) for describing the amide I spectral intensities in polar solvents. Correlations between temperature-dependent amide I frequencies and intensities, and their possible utility for analyses of the temperature-dependent peptide and protein infrared spectroscopy (IR) spectra are discussed.

Introduction

Infrared spectroscopy (IR) is frequently used for studies of peptide and protein structures and dynamics of the structural transitions during protein folding or function [1]. Several advantages of the IR, such as higher resolution compared to the electronic spectroscopies, and fast intrinsic time-scale compared to the NMR, have made IR spectroscopy a technique of choice, for example, for numerous dynamic studies of peptide and protein folding [2], [3], [4], [5]. On the other hand, one of the disadvantages for investigations of protein structural properties is that most commonly used denaturants absorb strongly in the IR, interfering with the structurally sensitive protein signals. As a consequence, temperature is most often used to induce protein unfolding in the IR spectroscopic studies. The temperature-dependent IR spectra can, however, be complicated by the temperature-dependent changes unrelated to the peptide or protein structural properties. Such temperature effects must be well understood for the IR spectroscopy to provide accurate and reliable information about peptide and protein structure.

The majority of peptide and protein IR studies focus on the amide I band (predominantly amide C double bond O bond stretching, often measured in D₂O as amide I′ for N-deuterated peptides) since its conformational sensitivity has been best established [1], [6], [7]. Most frequently, amide I/I′ frequencies are the basis for structural assignments and the amide I/I′ frequency shifts for the interpretation of structural changes. In the recent study, we have investigated the temperature dependence of the peptide amide I/I′ vibrational frequencies in solution [8]. We found that an α-helical oligopeptide, random coil poly-l-lysine and a simple amide N-methylacetamide (NMA) exhibit the same amide I′ frequency shifts with temperature in D₂O. To understand the origin of these effects, we measured the amide I frequencies of NMA as a function of temperature in organic solvents of different polarities: dimethyl sulfoxide (DMSO), acetonitrile and 1,4-dioxane. Similar temperature dependence of the amide I vibrational frequencies in all these solvents demonstrated that the dominant cause for the spectral shifts is not hydrogen bonding (to the amide C double bond O), which is the most common interpretation. The frequency shifts in the organic solvents can be quantitatively explained by Buckingham theory [9] based on a simple, continuum model for the solvent, which takes into account the temperature dependence of the dielectric constant, index of refraction and density.

In this report we focus on the amide I/I′ intensities and their changes with temperature. Overall, spectral intensities have received much less attention than the vibrational frequencies, even though the intensities are equally important in determining the appearance of the experimental amide I spectral contours. While in the gas phase the IR absorption intensities are temperature independent, at least in the harmonic approximation [10], the amide I intensity, along with its frequency, is highly sensitive to the solvent environment. For example, the amide I dipole strength of NMA has been shown to more than double between acetonitrile and aqueous solution [11]. This effect could be, again, expected to arise primarily from hydrogen bonding, as suggested by studies on NMA association in CH₃Cl [12] and of NMA and isotopically labeled peptides in D₂O at cryogenic temperatures [13]. However, bulk solvent electrostatics may play an important, or even dominant role, as in the case of the vibrational frequency shifts [8], [11], and as also suggested by theoretical simulations of the NMA spectra with explicit and implicit solvent models [11]. Since the dielectric properties of the solvents are temperature dependent, so will be the solvent-induced effects on the spectral intensities [14].

It is perhaps well appreciated that the spectral intensities inherently depend on temperature, but the IR studies of peptide or protein thermal denaturation seldom account for any intrinsic temperature dependence of the spectra. In order to reliably extract the structural information, it is necessary to separate the temperature dependence of the spectral changes arising from structural transitions from those due to the solvent and other structure unrelated effects. To date, and to our best knowledge, no systematic study of the amide infrared intensities as a function of solvent and temperature has been reported. We again use NMA as a model for the amide bond to eliminate all possible effects from the conformational equilibria, the Buckingham theory [9] and density functional theory (DFT) calculations to quantitatively model the temperature-dependent amide I spectral intensities under different solvent conditions.

Section snippets

Experimental

All solvents used in this study were of analytical or better grade. Acetonitrile and 1,4-dioxane were used without further purification. DMSO was found to contain a significant amount of water and therefore was dehydrated over activated molecular sieves; the dryness was evidenced by a complete absence of the water infrared (IR) absorption bands. NMA was purchased from Sigma–Aldrich and D₂O was from Cambridge Isotope Laboratories.

To obtain accurate absolute intensities, repeated IR spectral

Results and discussion

Experimental NMA amide I′ spectra in D₂O and amide I in DMSO, acetonitrile and 1,4-dioxane are shown in Fig. 1. For all studied solvents, nearly symmetric absorption bands with slight inhomogeneous broadening toward the lower frequencies are observed. Symmetric profiles imply monomeric NMA, as can be expected at relatively low concentrations [15]. Although the inhomogeneous broadening could be caused by intermolecular association, this would result in concentration and, most likely,

Conclusion

We have investigated the intrinsic temperature dependence of the amide I IR intensities in solution, using NMA as a model system. The NMA amide I intensity depends significantly on the solvent as well as on temperature. Interestingly, the solvent polarity seems to affect predominantly the bandwidths, while the increase in temperature is manifested by decrease in the peak absorption, but without significant broadening. A simple theory based on the Onsager reaction field was used to

Acknowledgements

This work was supported by the Faculty Grant-in-Aid and Basic Research Grant programs of the University of Wyoming. The computations were made possible by the National Science Foundation under the following NSF programs: Partnerships for Advanced Computational Infrastructure, Distributed Terascale Facility (DTF) and Terascale Extensions: Enhancements to the Extensible Terascale Facility. Loren Ackels was supported in part by the University of Wyoming EPSCoR undergraduate research fellowship.

References (42)

R.B. Dyer
Curr. Opin. Struct. Biol.
(2007)
K.H. Illinger et al.
J. Mol. Spectrosc.
(1962)
L.M. Kuznetsova et al.
J. Mol. Struct.
(1996)
Y. Luo et al.
Chem. Phys. Lett.
(1997)
G. Keresztury et al.
J. Mol. Struct.
(1997)
A. Barth et al.
Q. Rev. Biophys.
(2002)
C.M. Phillips et al.
Proc. Natl. Acad. Sci. U.S.A.
(1995)
R.B. Dyer et al.
Acc. Chem. Res.
(1998)
H.S. Chung et al.
Proc. Natl. Acad. Sci. U.S.A.
(2007)
P.I. Haris
Infrared analysis of peptides and proteins: principles and applications

E. Goormaghtigh et al.

K.E. Amunson et al.

J. Phys. Chem. B

(2007)

A.D. Buckingham

Proc. R. Soc., Lond. A

(1958)

J. Kubelka et al.

J. Phys. Chem. A

(2001)

E.S. Manas et al.

J. Am. Chem. Soc.

(2000)

W.B. Person

J. Chem. Phys.

(1958)

R. Schweitzer-Stenner et al.

J. Phys. Chem.

(1998)

P.R. Bevington et al.

Data Reduction and Error Analysis for the Physical Sciences

(1992)

L. Onsager

J. Am. Chem. Soc.

(1936)

C.G. Zhan et al.

J. Chem. Phys.

(1998)

B. Linder et al.

J. Chem. Phys.

(1967)

Cited by (34)

Normal mode calculation and infrared spectroscopy of proteins in water solution: Relationship between amide I transition dipole strength and secondary structure
2021, International Journal of Biological Macromolecules
Dipole Strength (DS) of the amides has gained a renewed interest in chemical physics since it provides an important tool to disclose the on-site vibrational energy distributions. Apart from earlier experimental efforts on polypeptides, little is still known about DS in complex proteins. We accurately measured the Fourier Transform Infrared absorption spectra of nine proteins in water solution obtaining their Molar Extinction Coefficient in the amide I and II spectral region. Our results show that the amide I DS value depends on the protein secondary structure, being that of the α-rich and unstructured proteins lower by a factor of 2 than that of the β-rich proteins. The average DS for amino acids in α and β secondary structures confirms this finding. Normal Mode calculation and Molecular Dynamics were performed and used as tools for data analysis and interpretation. The present outcomes corroborate the hypothesis that antiparallel β-sheet environment is more prone to delocalize the on-site CO stretching vibration through coupling mechanisms between carbonyl groups, whereas α-helix structures are energetically less stable to permit vibrational mode delocalization.
Temperature dependence of peptide carboxylic group IR spectra in the amide I′ region at neutral and acidic pH
2015, Vibrational Spectroscopy
Citation Excerpt :
This, however, can be corrected by pathlength calibration [15]. More significant is the sensitivity to temperature of both the frequency [16–21] and intensity [21] of the amide I′ bands. The frequency shift in particular may negatively affect both single-frequency and multivariate analyses of the amide I′ band changes and their interpretations in terms of structural transitions.
Analysis of the temperature-dependent amide I′ bands of peptides and proteins can be complicated by their overlap with other IR bands, particularly those of carboxylic groups of amino acid side-chains and the C-terminus. Previously, we reported IR spectra of charged carboxylic side-chains in Asp and Glu amino acids, and C-terminal groups of several amino acids and dipeptides at neutral pH. To complement these results, here we investigate the IR absorptions of Asp and Glu side-chains in capped amino acids (AcAspNMe and AcGluNMe), at both neutral and acidic pH. Spectra of protonated (acidic pH) C-terminal group absorptions are also investigated, using three dipeptides (AlaGly GlyAla and GlyGly) as model compounds. Sets of temperature-dependent experimental IR spectra were analyzed using pseudo-Voigt lineshape profiles. We find that the temperature-dependent behavior of the IR bands of deprotonated (neutral pH) side-chains in AcAspNMe and AcGluNMe dipeptides are generally similar to those reported previously for Asp and Glu. Protonated carboxylic group (acidic pH) IR bands behave uniformly with respect to temperature, showing very similar magnitude frequency shifts and intensity changes. Implications for analyses of amide I′ bands of peptides and proteins are discussed.
Temperature dependence of C-terminal carboxylic group IR absorptions in the amide I′ region
2015, Spectrochimica Acta - Part A: Molecular and Biomolecular Spectroscopy
Citation Excerpt :
However, this also introduces additional complications associated with the analysis of the temperature dependent amide I′ spectra (prime denotes N-deuterated amides, as measured in D2O based solutions). Namely, the amide I′ exhibits a significant intrinsic frequency shift [26–30] as well as a loss in intensity [27,31] as the temperature is raised. The accurate intensity measurement is further hampered by the thermal expansion of the IR sample cell and resulting changes in the optical pathlength [32], although this can be corrected.
Studies of structural changes in peptides and proteins using IR spectroscopy often rely on subtle changes in the amide I′ band as a function of temperature. However, these changes can be obscured by the overlap with other absorptions, namely the side-chain and terminal carboxylic groups. The former were the subject of our previous report (Anderson et al., 2014). In this paper we investigate the IR spectra of the asymmetric stretch of α-carboxylic groups for amino acids representing all major types (Gly, Ala, Val, Leu, Ser, Thr, Asp, Glu, Lys, Asn, His, Trp, Pro) as well as the C-terminal groups of three dipeptides (Gly–Gly, Gly–Ala, Ala–Gly) in D₂O at neutral pH. Experimental temperature dependent IR spectra were analyzed by fitting of both symmetric and asymmetric pseudo-Voigt functions. Qualitatively the spectra exhibit shifts to higher frequency, loss in intensity and narrowing with increased temperature, similar to that observed previously for the side-chain carboxylic groups of Asp. The observed dependence of the band parameters (frequency, intensity, width and shape) on temperature is in all cases linear: simple linear regression is therefore used to describe the spectral changes. The spectral parameters vary between individual amino acids and show systematic differences between the free amino acids and dipeptides, particularly in the absolute peak frequencies, but the temperature variations are comparable. The relative variations between the dipeptide spectral parameters are most sensitive to the C-terminal amino acid, and follow the trends observed in the free amino acid spectra. General rules for modeling the α-carboxylic IR absorption bands in peptides and proteins as the function of temperature are proposed.
The study of secondary effects in vibrational and hydrogen bonding properties of 2- and 3-ethynylpyridine and ethynylbenzene by IR spectroscopy
2014, Spectrochimica Acta - Part A: Molecular and Biomolecular Spectroscopy
Citation Excerpt :
Overall behavior is quite expected, i.e. the integrated intensities of the dominant bands decreases while the frequencies increases (0.02–0.03 cm−1/°C) with the increasing temperature. It can be accounted for by the Onsager reaction field theory with temperature dependent solvent refractive index and dielectric constant and the solute molecular cavity [27]. The temperature increase from 25 °C to 115 °C affects the ph OH stretching band (Fig. S4).
Weak hydrogen bonds formed by 2- and 3-ethynylpyridine and ethynylbenzene with trimethylphosphate and phenol were characterized by IR spectroscopy and DFT calculations (B3LYP/6-311++G(d, p)). The structure and stability of ethynylpyridines and ethynylbenzene in the gas phase and in the complexes with trimethylphosphate and phenol are discussed in terms of geometry and electronic charge redistribution. Anharmonic effects are taken into account when calculating vibrational wavenumbers of these systems what lead to partial improvement of agreement with experiment. The changes in the electronic charge distribution are behind the frequency shifts of the CC stretching in opposite direction depending on the role the ethyne molecule has in a hydrogen bonded complex ( $Δ \tilde{ν} = + 9$ cm⁻¹ in trimethylphosphate complexes, $Δ \tilde{ν} = - 3$ cm⁻¹ in phenol complexes). The association constants were determined by keeping the concentrations of proton donors approximately constant and low enough to avoid self-association and the proton acceptors were present in excess. The values obtained for the association constants and enthalpy changes in C₂Cl₄ (for trimethylphosphate complexes K ≈ 0.5–1.0 mol⁻¹ dm³ and $- Δ_{r} H^{⦵} \approx 6 – 8 kJ {mol}^{- 1}$ , for phenol complexes K ≈ 20–40 mol⁻¹ dm³ $- Δ_{r} H^{⦵} \approx 17 – 22 kJ {mol}^{- 1})$ are in good agreement with literature data.
Chasing weakly-bound biological water in aqueous environment near the peptide backbone by ultrafast 2D infrared spectroscopy
2024, Communications Chemistry
Refining protein amide I spectrum simulations with simple yet effective electrostatic models for local wavenumbers and dipole derivative magnitudes
2023, Physical Chemistry Chemical Physics

View all citing articles on Scopus

View full text

On the temperature dependence of amide I intensities of peptides in solution

Abstract

Introduction

Section snippets

Experimental

Results and discussion

Conclusion

Acknowledgements

Curr. Opin. Struct. Biol.

J. Mol. Spectrosc.

J. Mol. Struct.

Chem. Phys. Lett.

J. Mol. Struct.

Q. Rev. Biophys.

Proc. Natl. Acad. Sci. U.S.A.

Acc. Chem. Res.

Proc. Natl. Acad. Sci. U.S.A.

Infrared analysis of peptides and proteins: principles and applications

J. Phys. Chem. B

Proc. R. Soc., Lond. A

J. Phys. Chem. A

J. Am. Chem. Soc.

J. Chem. Phys.

J. Phys. Chem.

Data Reduction and Error Analysis for the Physical Sciences

J. Am. Chem. Soc.

J. Chem. Phys.

J. Chem. Phys.