Development and Application of Direct Dynamics Simulations for Reactivity of Biological Molecules – DynBioReact

The method employed is molecular dynamics allowing chemical reactions, named chemical dynamics. We have used different semi-empirical Hamiltonians allowing the study of relatively large molecules by direct dynamics of reactive processes. We have first tested and validated the different methods by comparison with more precise ones (which are computationally more demanding) and experiments. We have then considered two models of collision induced dissociation: explicit collision and internal energy activation. Results of the two models can provide information on characteristic products and mechanisms. Finally, we could point out how internal energy fragmentation can be related to statistical theories and used to obtain in a simple way quantitative information on reactivity for complex molecules (like e.g. peptides). Since we need a huge number of trajectories, we have developed an automatic analysis method based on graph theory which can provide, once the simulations are done, all the quantities needed on the whole set of trajectories.

It is now possible to obtain a mass spectrum of a given species directly from simulations. At this end, we have developed an approach based on trajectories and their automatic analysis. Given the initial structure (normally the most stable isomer), we run trajectories and determine products and isomers. New isomers can be used as new initial structure and the procedure iterated up to convergence. This approach can be used for fragmentation in theoretical mass spectrometry but also for ion-molecule reactions. Several new national and international collaborations were thus established.

-

We have published 18 articles on international journals and 1 book chapter and we have about 4 articles in preparation. Furthermore, a book on theoretical mass spectrometry was issued in 2018. Finally, we wrote an integrated software to analyze an ensemble of trajectories and we have modified a code to calculate unimolecular dissociation rate constant for the peptides. They will be finalized in 2019 and released to the community.

Overview. Chemical dynamics simulations will be performed to model: (1) collision induced dissociation (CID) experiments on peptides, carbohydrates, and steroids; and (2) peptide synthesis in interstellar media. The goal of the CID simulations is to get new insights about the mechanisms involved in the CID process and to obtain theoretical predictive MS/MS spectra. The goal of peptide synthesis simulations is to determine if this process is feasible in interstellar media via molecule-molecule and/or ion-molecule reactions, and with or without the presence of a catalyst. The simulations will be performed using direct dynamics, in which the methodology of chemical dynamics is directly interfaced with electronic structure theory.

Intellectual Merit. The proposed project is based upon a research collaboration with the Hase research group (USA). In previous studies, we have used direct dynamics simulations to study energy transfer and fragmentation dynamics in the collision-induced dissociation (CID) of protonated urea and [Ca(urea)]2+ complex. New developments and findings in the field of chemical reaction dynamics will benefit from a strong USA-France scientific cooperation through the NSF-ANR international program. In the research proposed here we will study:

1. Theoretical modeling of tandem mass spectrometry experiments of the fragmentation dynamics of peptides, carbohydrates, and steroids;

2. Bimolecular reactivity leading to peptide synthesis.

The significance of the first topic is illustrated by the study of peptides. Despite the large number of mass spectrometric studies dedicated to peptide dissociation, many fragmentation pathways are still not well understood and classified. Studies have shown that 40-70% of high signal/noise MS/MS spectra cannot be matched to predicted protein spectra (or are misidentified) by widely used search engines. One of the problems is post-fragmentation sequence scrambling. Another issue is that peptides do not necessarily fragment as expected from simple models. This likely occurs when amino acids undergo unexpected side chain fragmentations. Direct dynamics CID simulations can address both of these problems. Theoretical MS/MS spectra for peptides and other biological molecules will be obtained by combining the short-time non-statistical fragmentation dynamics of the CID simulations with long-time statistical RRKM calculations, which utilize the CID energy transfer distributions obtained from the simulations.

For the second topic, direct dynamics simulations will be used to address questions concerning the presence of biomolecules in interstellar media, and dynamical effects that may govern peptide synthesis in the astrophysical environment. The goal is to understand the reaction dynamics of peptide synthesis either in the gas phase or on graphite surfaces (to mimic meteorite surfaces), as has been suggested for astrochemical conditions. We will focus our attention on the formation of amino acids and di-peptides through ion-molecule and/or molecule-molecule reactions.

To facilitate the synergy between the two groups, frequent exchanges will be done and post-doctoral associates will switch between USA and France each year.