A theoretical and matrix-isolation FT-IR study on amino acids, N-acetylated amino acids and H-bonded complexes between amino acids and DNA bases.

Publication date: 2011-03-04

Author:

Boeckx, Bram
Maes, Guido

Keywords:

ab initio, amino acids, FT-IR, DNA bases

Abstract:

1. Amino acids The knowledge of the conformational behavior of amino acids is fundamental to understand the conformational flexibility of proteins which is important in the biological relationship between structure and functionality. Therefore, a detailed theoretical and matrix-isolation FT-IR investigation of the conformational landscape was performed for phenylalanine,isoleucine, lysine and asparagine, as well as for N-acetylcysteine, N-acetylproline, N-acetylalanine and N-acetylglycine. In a first stage of the study, non acetylated amino acids were investigated. For phenylalanine five conformations with an abundance larger than 5% were predicted. Two of them contain the amino acid backbone type I characterized by an OH…N intramolecular H-bond, two of them have the amino acid backbone type II characterized by a C=O…H2NH-bond and one has type III with a NH2…O(H) H-bond. The H-bond involved vibrational modes were used to identify these three different types of backbone in the matrix-isolation FT-IR spectrum. An experimental rotamerization constant of 3.9 between the conformations with andwithout an intramolecular OH…N H-bond could be estimated, which appeared to be in good accordance with the MP2 theoretical value. The investigation of the conformational landscape of isoleucine resulted in five conformations with an abundance larger than 5%. Among these, the amino acid backbones type I, II and III were present. All these conformations could be identified in an Ar matrix. The incorporation of the entropy term was necessary to correctly predict the observed conformations. A rotamerization constant of 7.1 could be estimated. The accordance with the theoretically predicted MP2 valueappeared to be good, whereas the accordance with the DFT value was poor.Small differences in the aliphatic side chain of lysine result in a large flexibility. The observed energy differences between the different conformations are largely due to the additional intramolecular H-bond formed at the NH2 group in the side chain. The lowest energy conformations contain an additional H-bond between the amino acid backbone and the NH2group of the side chain. This energetic order of the conformations is completely changed after incorporation of the entropy term. This implies that the stability is strongly dependent on the sublimation temperature used. The abundant conformations of lysine can be classified by the type of amino acid backbone and the eventual additional H-bond into four groups, i.e. type II,type I, type I with an additional NH…Nsc H-bond and type IV with an additional OH…Nsc H-bond. These four groups were all predicted to be detectable in the matrix, as their abundances are all larger than 5% at the sublimation temperature of 380 K. In the experimental matrix-isolation FT-IR spectrum these four groups could be identified using the shifts of the H-bondinvolved modes. The theoretical investigation of the conformational landscape of asparagine resulted in four(DFT) or six (MP2) conformations which should be abundant in the experimental matrix spectrum. Among these six, four have a different kind of amino acid backbone and additional H-bonds, i.e. type I and two additional C=Osc…HNbb, C=Obb…HNsc H-bonds, type II with an additional N…HNsc H-bond, type I with one additional NH…Osc H-bond and type II and IIIwithout additional H-bond stability. Due to the large shifts of the H-bond involved modes, all these particular types of conformations were observed in the FT-IR matrix spectra. In a second stage of the research, N-acetylated amino acids were investigated with the aim to evaluate the influence of the peptide bond on the conformational behaviour of amino acids. The conformational analysis of N-acetylcysteine resulted in two conformations with an abundance larger than 5%. These conformations and one additional conformation with an intramolecular H-bond were observed in the matrix spectrum. The abundance of this H-bond containingconformation was estimated to be 3%, which was in good accordance with the MP2 predicted value. The DFT method gave no satisfying stability results for NAC, since thedispersion energy is poorly described by this method. For N-acetylproline, the most abundant predicted conformation (47.66%) has anintramolecular H-bond OH…O=Cac, and five other conformations with an abundance larger than 5% were predicted. In the matrix-isolation the H-bond involved modes as well as the non H-bond involved modes were clearly observed. An experimental rotamerization constantof 0.50 could be estimated, in reasonable accordance with the predicted value. The investigation of the conformational behaviour of N-acetylalanine resulted in two conformations with an abundance larger than 5% without any intramolecular H-bond. The most stable conformation with an intramolecular H-bond had a predicted abundance of only 2%. Using the H-bond involved modes these three conformations could be observed in an Ar matrix.The experimental abundances of these three conformations were estimated to be 71.8% for NAA1,7.9% for NAA2 and 20.2% for NAA3, which are in good accordance with the theoretical DFT and MP2 predictions. The most stable conformation of N-acetylglycine had a predicted abundance larger than 90%. Despite the small predicted abundance of two other conformations, these three conformations could be observed in the FT-IR spectrum. The experimental abundances of these three conformations were estimated to be 82.5% for NAG1, 11.8% for NAG2 and 5.7% for NAG3, which was in good agreement with the theoretical (DFT and MP2) predictions. For all these monomeric compounds, the mean frequency deviation between prediction andexperiment was about 10 cm-1, which is in accordance with literature data for other, smaller amino acids. This implies that the scaling factors used are suitable for the investigation of more complex and N-acetylated amino acids. The experimental obtained abundances were in good accordance with the obtained MP2 stabilities whereas the agreement with the DFT stabilities wassometimes poor. This demonstrates that the MP2 method is necessary to accurately predict the abundances of these molecules. The Gibbs free energy was not always in line with the electronic energy, due to the largeimpact of the (-T∆S°) term for H-bond containing systems. Therefore, the incorporation of this term is necessary for an accurate prediction of H-bond containing conformations. The H-bond properties such as the distances r(XH), r(XH…Y), ∆r(XH) and the shifts ν(XH…),γ(XH…) or the angle (XH…Y), were carefully investigated and appeared to follow the previously obtained correlations described in the literature. All these properties appeared to be good parameters to quantify the strength of the H-bond. This demonstrates that the spectral assignments performed for the H-bond sensitive vibrational modes are correct. The H-bonds in N-acetylated amino acids were found to be stronger compared to the non modified amino acids which is demonstrated by the stronger frequency shifts and the smaller H-bond distances. This work succeeded to reveal the most stable conformations of the monomeric compounds.This obtained knowledge can be important for further studies on larger protein systems. Due to the N-acetylation of the amino acid backbone the rotameric possibilities of the backboneconformations are limited. The FT-IR of the amino acids phenylalanine and isoleucine and the studied N-acetylated compounds are analyzed and the different conformations could be identified. In contrast to this, for the more complex amino acids the identification was limited to the different types of H-bonded structures for lysine and asparagine. For NAP, NAA and NAG an overtone or combination band of ν(C=O)ac is observed which can be used as a spectral tool to distinguish between the acetylated and non acetylated amino acid. 2. H-bonded complexes of amino acid . nucleic acid base A fundamental knowledge of the interaction between amino acids and DNA bases is importantsince nearly all DNA functions are dependent on the interactions with proteins. Therefore, some amino acid.DNA base complexes were investigated by a theoretical and matrix-isolation FT-IR study and their most stable configurations have been revealed. For the experimental study a new, dual mini-furnace was successfully developed, which allows to sublimate both complex partners at their optimal sublimation temperature. The H-bond interaction between glycine and 1-methyluracil was first theoretically investigated. The accuracy of the DFT(B3LYP)/6-31G geometries obtained was evaluated with single point energy calculations on the structures optimized at this and at the DFT(B3LYP)/6-31++G** level. This evaluation revealed that the DFT(B3LYP)/6-31G optimized geometries yield satisfying structures. The most stable complex had a H-bond interaction energy of -54.15 kJ.mol-1.In the experimental matrix spectrum, shifted modes due to H-bonding could be observedand the presence of three different complexes was suggested by the observation of three different ν(OH…) bands. For the most stable complex, the other H-bond involved vibrational modes have also been observed. The theoretical investigation of the H-bond complex phenylalanine.1-methyluracil resulted inthree stable configurations. The most stable one contained a C=OPHE…HN1MC and an OHPHE…N1MC closed H-bond system. The obtained H-bond interaction energy of the most stable complex was -68.99 kJ.mol. This stabilization energy is clearly larger compared to the stacking interaction energy of phenylalanine complexes, which demonstrates that H-bonds are more stabilizing than stacking interactions. The presence of the complex in the FT-IR spectrum could be demonstrated by the observation ofall the H-bond involved modes. The evaluation of several correlations between H-bond distances, frequency shifts, angles and interaction energies allow to conclude that the spectral assignments for the closed H-bond complexes are correct.