Computational Studies of Mechanisms of Peptide-Ion Fragmentation


Christopher R. Kinsinger


Proteomics deals with the full set of proteins encoded by a genome.  High-throughput proteomic studies currently use mass spectrometry to analyze thousands of proteins after digestion by trypsin.  Specifically, the recent developments of electrospray ionization and tandem mass spectrometry have made mass spectrometry the instrument of choice for proteomics.  Ideally, the fragmentation pattern of one protein can be used as a fingerprint for identifying that protein within a mixture of many proteins.  However, many mass-spectral peaks remain unidentified.  Additionally, information from peak intensities is not usually incorporated into current protein sequencing algorithms and databases.  These two shortcomings are due to limited understanding of both the variation between different types of instruments and the mechanisms of peptide fragmentation. 


Since ab initio methods were developed primarily for gas-phase calculations, quantum chemistry is well-suited for studying mechanisms of peptide dissociation in the gas phase.  The immediate question we are trying to answer is: “Why does fragmentation tend to occur on the N-terminal side or the C-terminal side of a given amino acid?”  We are considering three simple hypotheses to answer this question: (1) the peptide site with the highest proton affinity may correspond to the location of fragmentation (proton affinity hypothesis);  (2) the most thermodynamically stable set of products may be favored (thermodynamic hypothesis); (3) the site of fragmentation may be dictated by the lowest activation barrier (kinetic hypothesis).  We are currently exploring the first two of these hypotheses.  By using inexpensive levels of theory for geometry optimization and a correlated theory with a large basis set for calculating energies, we are able to reproduce some general trends of fragmentation.  We also compare our results to experimental spectra.  Preliminary results show that the proton affinities of different sites are not sufficiently different to influence fragmentation selectivity.  On the other hand, the thermodynamic hypothesis has led to some agreement between calculated and experimental results. Through computational chemistry and continued dialogue with experimentalists, we hope to develop fragmentation rules that accurately predict mass spectra.


Author:    Christopher R. Kinsinger, Karl K. Irikura, Division of Physical and Chemical Properties, CSTL

Room:     A157, Bldg 222

Mail:        8380

Phone:     x2526

Fax:         301-869-4020


Not a member of Sigma Xi