Relating the infrared absorption of proteins to their structure
Title: Relating the infrared absorption of proteins to their structure
SNIC Project: SNIC 2020/5-371
Project Type: SNIC Medium Compute
Principal Investigator: Andreas Barth <>
Affiliation: Stockholms universitet
Duration: 2020-06-30 – 2021-07-01
Classification: 10603 10402 10302


The leading application of infrared spectroscopy in the life sciences is the analysis of protein structure. It exploits the conformational sensitivity of the amide I vibrations, which are collective vibrations of the amide groups of the protein backbone. However, their theoretical description is currently limited and therefore the interpretation of experimental results is guided by (sometimes conflicting) empirical correlations. Therefore, a reliable theoretical description is desirable and can lead to additional structural insight (e.g. Baronio, Baldassarre, & Barth, PCCP 2019). The large number of atoms in proteins prevents calculating their vibrational spectrum from density functional theory (DFT). Therefore, present calculations of the amide I spectrum use a different approach. They mostly consider each amide group as a local amide I oscillator, which has an intrinsic frequency and which couples with other amide groups. The intrinsic frequencies are affected by the local backbone conformation and elec­trostatic effects including hydrogen bonding. The former influences also the coupling between nearest neighbors. Long-range coupling is described by transition dipole coupling (TDC) between the transition dipole moments (TDMs) of close-by amide I oscillators. Once the intrinsic frequencies and the coupling constants are known, the spectrum of the de­localized amide I normal modes can be calculated. This ap­proach describes general trends well, but needs to be improved for detailed structural analysis. Therefore, we will use DFT calculations on small peptide systems (up to 21 amide groups) to study three aspects that are fundamental for the calculation of the amide I spectrum. a) An effect of conformation on the magnitude and direction of the TDM was found in our recent DFT calculations (Baronio & Barth, J. Phys. Chem. B 2020). Implementing secondary structure dependent TDM parameters into our amide I calculations might cause abrupt changes of these parameters upon relatively small changes in backbone structure. To avoid that, we will explore whether the different dihedral angles are the underlying reason for the different TDM parameters using DFT calculations of tetra­peptides (three complete amide groups) in different conformations. We will start with model structures of helices, parallel and antiparallel β-sheets for which geometry optimized structures are already available from our previous work. b) An effect of hydrogen bonding on the TDM parameters is also indicated by our previous work and we will use the same approach to study it, but using larger structures (10–21 amide groups), some of which are already geometry optimized. c) The nature of the coupling between nearest neighbors is unclear. Possible causes are electrostatic interactions between neighboring amide I vibrations or mechanical coupling, which is the influence of one amide group on its nearest neighbors via the displacement of the intervening atoms. To explore the origin of the coupling, we will calculate tetrapeptide spectra where we selectively switch off mechanical coupling or both, mechanical and electrostatic coupling.