<%BANNER%>

Dependence of Head-to-Tail Cyclization on Primary Structures of Peptides in Collision-Induced Dissociation

Permanent Link: http://ufdc.ufl.edu/UFE0042539/00001

Material Information

Title: Dependence of Head-to-Tail Cyclization on Primary Structures of Peptides in Collision-Induced Dissociation
Physical Description: 1 online resource (148 p.)
Language: english
Creator: Chen, Xian
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: cyclization, fticr, irmpd, macrocycle, ms, oxazolone, peptide, sequencing
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Collision-induced dissociation (CID) of peptides to deduce their sequences is the key technology in identifying proteins by mass spectrometry. However, during activation the linear peptide structures can undergo a head-to-tail cyclization reaction, where the N- and C-terminus of a peptide ?b? fragment fuse into a macrocycle structure. When this macrocycle structure opens up at a different site than where it was originally formed, a scrambling of the sequence information will occur. There are few techniques that yield direct structural information on the minute quantities of gas-phase ions inside mass spectrometers, such as infrared spectroscopy, ion mobility, and hydrogen/deuterium exchange (HDX). IR spectroscopy confirms chemical structures based on diagnostic vibrations, however it is difficult to obtain relative abundances. Conversely, ion mobility can yield relative abundances of structures, but the structural interpretation is often ambiguous. HDX had so far not produced either information. In our research, we combined infrared multiple photon dissociation (IRMPD) spectroscopy and gas-phase HDX to structurally characterize and quantify macrocycle, as well as oxazolone, structures for a series of ?b? ions from selected peptide systems. For a series of glycine-based ?b? fragment ions, this approach shows a size-dependency, where smaller ?b? fragments exclusively adopt oxazolones, while larger fragments display a mixture of oxazolones and macrocycles. The results are consistent with the finding from HDX, where smaller ?b? fragments display a single HDX rate, whereas larger fragments show two distinct rates. Relative abundances of oxazolones and macrocycles are approximated from HDX kinetic fitting. Similar trends are found in the peptide Leucine-enkephalin (YGGFL). The correlation between peptide sequence and propensity for macrocycle formation for the b6 motif QWFGLM is investigated. The IRMPD spectrum for the b6 fragment of QWFGLMPG is nearly identical to that for protonated cyclo(QWFGLM), which confirms the exclusive presence of macrocycle structures for b6 (from QWFGLMPG). The incorporation of a proline in for instance QPFGLMPG is found to reduce the propensity of the formation of macrocycle structures in the corresponding b6 fragment. A systematic chemical protection study was performed to the QWFGLG system. IRMPD spectra of b6 ions generated from QWFGLGPG, Ac-QWFGLGPG, Ac-Q(N-ethyl)WFGLGPG, and Q(N-ethyl)WFGLGPG indicate that no cyclization from the glutamine side chain occurs.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Xian Chen.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Polfer, Nicolas Camille.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042539:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042539/00001

Material Information

Title: Dependence of Head-to-Tail Cyclization on Primary Structures of Peptides in Collision-Induced Dissociation
Physical Description: 1 online resource (148 p.)
Language: english
Creator: Chen, Xian
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: cyclization, fticr, irmpd, macrocycle, ms, oxazolone, peptide, sequencing
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Collision-induced dissociation (CID) of peptides to deduce their sequences is the key technology in identifying proteins by mass spectrometry. However, during activation the linear peptide structures can undergo a head-to-tail cyclization reaction, where the N- and C-terminus of a peptide ?b? fragment fuse into a macrocycle structure. When this macrocycle structure opens up at a different site than where it was originally formed, a scrambling of the sequence information will occur. There are few techniques that yield direct structural information on the minute quantities of gas-phase ions inside mass spectrometers, such as infrared spectroscopy, ion mobility, and hydrogen/deuterium exchange (HDX). IR spectroscopy confirms chemical structures based on diagnostic vibrations, however it is difficult to obtain relative abundances. Conversely, ion mobility can yield relative abundances of structures, but the structural interpretation is often ambiguous. HDX had so far not produced either information. In our research, we combined infrared multiple photon dissociation (IRMPD) spectroscopy and gas-phase HDX to structurally characterize and quantify macrocycle, as well as oxazolone, structures for a series of ?b? ions from selected peptide systems. For a series of glycine-based ?b? fragment ions, this approach shows a size-dependency, where smaller ?b? fragments exclusively adopt oxazolones, while larger fragments display a mixture of oxazolones and macrocycles. The results are consistent with the finding from HDX, where smaller ?b? fragments display a single HDX rate, whereas larger fragments show two distinct rates. Relative abundances of oxazolones and macrocycles are approximated from HDX kinetic fitting. Similar trends are found in the peptide Leucine-enkephalin (YGGFL). The correlation between peptide sequence and propensity for macrocycle formation for the b6 motif QWFGLM is investigated. The IRMPD spectrum for the b6 fragment of QWFGLMPG is nearly identical to that for protonated cyclo(QWFGLM), which confirms the exclusive presence of macrocycle structures for b6 (from QWFGLMPG). The incorporation of a proline in for instance QPFGLMPG is found to reduce the propensity of the formation of macrocycle structures in the corresponding b6 fragment. A systematic chemical protection study was performed to the QWFGLG system. IRMPD spectra of b6 ions generated from QWFGLGPG, Ac-QWFGLGPG, Ac-Q(N-ethyl)WFGLGPG, and Q(N-ethyl)WFGLGPG indicate that no cyclization from the glutamine side chain occurs.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Xian Chen.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Polfer, Nicolas Camille.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2012-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042539:00001


This item has the following downloads:


Full Text

PAGE 1

1 DEPENDENCE OF HEAD TO TAIL CYCLIZATION ON PRIMARY STRUCTURES O F PEPTIDES IN COLLISIO N INDUCED DISSOCIATION By XIAN CHEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT O F THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Xian Chen

PAGE 3

3 To my parents and Yilin

PAGE 4

4 ACKNOWLEDGMENTS Throughout my graduate study at the University of Florida, I have benefited a lot from the help from a number of people. First and foremost, I would like to express my deep gratitu de to my advisor, Dr. Nicolas Polfer, for his guidance, support, and patience over the past three years. He encouraged and supported me to present our work at the national conferences. I also thank him for giving me the opportunity to travel to the FOM institute in The Netherlands to carry out the IRMPD experiments and to the University of Amsterdam for peptide synthesis. It was a great experience in my life. I want to ackno wledge my committee members, Drs. David Powell, Richard Yost, Kirk Schanze, and David Barber for helpful suggestions. given for our research and my practice talks for conferences and seminars. I would like to thank all the people with whom I collaborated because this dissertation will not be finished without their kind help. I thank Dr. David Powell for generously allowing us to use his FTICR mass spectrometer, with which we performed all the gas phase H/D exch ange experiments. He is always very patient for all my questions. Dr s. Jos Oomens and Jeff Steill at FOM institute are greatly acknowledged for their help on the IRMPD experiments and helpful discussions. I want to thank Dr. Britta Redlich for arranging th e guesthouse for me while I was working at the FOM institute. I also want to thank Dr. Jan van Maarseveen for welcoming me to carry some peptide synthesis experiments in his lab Jochem Rutt ers is acknowledged for teaching me how to synthesize cyclic pepti des in solution and answering my questions My third project will not be accomplished without their help. I would like to acknowledge Mr. Alfred Chuang from the ICBR at University of F lorida. He helped us synthesize and

PAGE 5

5 characterize some peptides and gener ously shared his experience on peptide synthesis with me. My colleagues in the Polfer group are acknowledged for the discussion and the suggestions on my research. They also helped me preparing my every presentation. I would like to thank Yu Long for doing the calculations to generate the theoretical spectra for the first project. Some former Eyler group members provided a lot of help to me. I thank Dr. Julia Rummel for teaching me how to use the FTICR instrument, generously sharing her experience, and bein g patient for my questions. I also want to thank Dr. Sarah Stefan and Dr. Michelle Sweeney for offering help and suggestions. Finally, I want to thank my family. I thank my parents Shanfeng Chen and Suyun Liao, for the ir love and support through these yea rs They are always very supportive for the decisions I made, including studying abroad for a PhD degree. I also want to acknowledge my husband, Yilin Meng for being a wonderful friend and companion. It is his love, patience, support and encourage ment mak es this dissertation possible.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 9 LI ST OF FIGURES ................................ ................................ ................................ ........ 10 ABSTRACT ................................ ................................ ................................ ................... 15 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 17 Peptide C hemistry ................................ ................................ ................................ .. 17 Amino Acids ................................ ................................ ................................ ..... 17 Peptide Synthesis ................................ ................................ ............................. 18 Fmoc solid phase peptide synthesis ................................ .......................... 18 Synthesis of macrocyclic peptides ................................ ............................. 20 Protein Identification ................................ ................................ ............................... 21 Edman Degradation ................................ ................................ ......................... 22 Protein Identification with Mass Spectrometry ................................ .................. 23 Peptide Fragmentation ................................ ................................ ............................ 24 Mass Spectrometry ................................ ................................ .......................... 24 Electrospray ionization ................................ ................................ ............... 25 Tandem mass spectrometry ................................ ................................ ....... 27 Fragmentation Methods ................................ ................................ .................... 28 Mobile Proton Model ................................ ................................ ........................ 29 Structures of b Fragment Ions ................................ ................................ ................ 30 Objective of This Research ................................ ................................ ..................... 32 2 INSTRUMENTATION AND METHODS ................................ ................................ .. 43 Fourier Transform Ion Cyclotron Resonance Mass Spectrometry .......................... 43 Principles ................................ ................................ ................................ .......... 43 Ion cyclotron motion ................................ ................................ ................... 43 Trapping motion ................................ ................................ ......................... 46 Magnetron motion ................................ ................................ ...................... 47 Apparatus ................................ ................................ ................................ ......... 48 Magnet ................................ ................................ ................................ ....... 48 Analyzer cell ................................ ................................ ............................... 49 Ultra high vacuum system ................................ ................................ .......... 49 Data system ................................ ................................ ............................... 50 Excitation and Detection ................................ ................................ ................... 50 Mass Accuracy ................................ ................................ ................................ 51 Mass Resolution ................................ ................................ ............................... 52

PAGE 7

7 Space Charge Effects ................................ ................................ ...................... 54 Infrared Multiple Photon Dissociation Spectroscopy ................................ ............... 55 Gas Phase Hydrogen/Deuterium Exchange ................................ ........................... 58 3 INVESTIGATION OF THE INFLUENCE OF PEPTIDE B FRAGMENT IONS ON THE FORMATION OF MACROCYCLE STRUCTURES THROUGH HEAD TO TAIL CYCLIZATION IN THE CASE OF OLIGOGLYCINES ................................ .... 69 Background ................................ ................................ ................................ ............. 69 Experimental ................................ ................................ ................................ ........... 70 Sample Preparation ................................ ................................ .......................... 70 Mass Spectrometry and Hydrogen/Deuterium Exchange ................................ 71 Mass Spect rometry and Infrared Photodissociation Spectroscopy ................... 72 Results and Discussion ................................ ................................ ........................... 73 Scrambling in Isotope Tagged Peptide ................................ ............................ 73 Infrared Spectroscopy Results ................................ ................................ ......... 74 b 2 ion ................................ ................................ ................................ .......... 74 b 5 ion ................................ ................................ ................................ .......... 76 b 8 ion ................................ ................................ ................................ .......... 77 Summary of IRMPD results ................................ ................................ ........ 78 s ................................ ....... 78 Kinetic Fitting of HDX Data ................................ ................................ ............... 79 Chemical Basis for Differences in HDX Rates ................................ .................. 81 Summary ................................ ................................ ................................ ................ 82 4 STRUCTURAL ANALYSIS FOR LEU ENKEPHALIN B2 B4 WITH INFRARED MULTIPLE PHOTON DISSOCIATION SPECTROSCOPY AND GAS PHASE HYDROGEN/DEUTERIUM EXCHANGE ................................ ................................ 96 Background ................................ ................................ ................................ ............. 96 Experimental ................................ ................................ ................................ ........... 97 Sample Preparation ................................ ................................ .......................... 97 Mass Spectrometry and H ydrogen/ D euterium E xchange (HDX) ...................... 97 Mass Spectrometry and Infrared Photodissociation Spectroscopy ................... 98 Computations ................................ ................................ ................................ ... 99 Results and Discussion ................................ ................................ ......................... 100 Infrared Spectroscopy Results ................................ ................................ ....... 100 Hydrogen Deuterium Exchange (HDX) Experiments ................................ ..... 103 Kinetic Fitting of HDX Data ................................ ................................ ............. 104 Comparison with Oligoglycines Study ................................ ............................ 105 Summary ................................ ................................ ................................ .............. 107 5 DEPENDENCE OF HEAD TO TAIL CYCLIZATION ON PRIMARY STRUCTURE OF PEPTIDES I N COLLISION INDUCED DISSOCIATION .......... 115 Background ................................ ................................ ................................ ........... 115 Experimental ................................ ................................ ................................ ......... 116

PAGE 8

8 Materials ................................ ................................ ................................ ......... 116 Preparation of Peptides ................................ ................................ .................. 117 Fmoc synthesis of linear peptides ................................ ............................ 117 Synthesis of head to tail cyclic peptides ................................ .................. 117 Purification of synthesized head to tail cyclic peptides ............................ 118 Mass Spectrometry ................................ ................................ ........................ 118 Results and Discussion ................................ ................................ ......................... 119 Commercial Cyclo(QWFGLM) as Direct Reference ................................ ....... 119 IRMPD of cyclo(QWFGLM) and analogs ................................ ................. 119 HDX of cyclo(QWFGLM) and analogs ................................ ..................... 120 The Effect of Prol ine on Head to Tail Cyclization ................................ ........... 122 Charaterization of synthetic peptides ................................ ....................... 122 IRMPD results ................................ ................................ .......................... 122 The Effect of Glutamine Side Chain on the Head to Tail Cyclization ............. 123 Summary ................................ ................................ ................................ .............. 125 6 CONCLUSIONS AND FUTURE DIRECTIONS ................................ .................... 134 LIST OF REFERENCES ................................ ................................ ............................. 138 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 148

PAGE 9

9 LIST OF TABLES Tabl e page 1 1 Effect of yield per coupling step on final product yields. 3 ................................ .... 39 1 2 Sources for MS based protein identificati on tools. 73 ................................ ........... 39 2 1 Proton affinities of common amino acids. Numbers are adapted from Ref. 129. ................................ ................................ ................................ .............. 67 3 1 Mass to charge ratios of a ll b ions of interested. ................................ ................ 85 3 2 Photofragments of b 2 protonated cyclo(Gly Gly) and b 5 ................................ ... 85 3 3 Kinetic fitting results for the ln[d 0 / d n ] plots vs. H/D exchange time for the oligoglycine fragments b 2 b 8 ................................ ................................ .............. 94 4 1 Energies for the lowest energy conformers of each chemical structure of b 2 The electronic energy for each c onformer at the MP2/6 31G+(d,p) level was corrected for the zero point energy (ZPE) derived at the B3LYP/6 31G+(d,p) level. ................................ ................................ ................................ ................. 109 4 2 exchanging structures for the Leu enkephalin fragments b 2 b 4 ................................ ........... 114 4 3 Kinetic fitting results for the ln[d 0 / d n ] plots vs. H/D exchange time for the Leu enkephalin fragments b 2 b 4 ................................ ................................ ...... 114 5 1 Mass to charge ratios ( m/z ) of the ions of interest in this project. ..................... 126

PAGE 10

10 LIST OF FIGURES Figure page 1 1 The chemical structure of an alpha amino acid. (A) An alpha amino acid which has one amine group, one carboxylic group and one distinct functional group (R) attached to one alpha carbon atom; (B) Fischer projection formulas for an L amino acid; (C) Fischer projection formulas for a D amino acid; (D) Stereo representations for an L amino acid; and (E) Stereo representations for a D amino acid. ................................ ................................ ............................. 34 1 2 A dipeptide is formed through a condensation reaction. Peptide bond is bracketed in red, and the N and C termini are indicated in blue and purple, respectively. ................................ ................................ ................................ ........ 35 1 3 The twenty amino acids found in proteins. Their names, three letter and one letter abbreviations are given below each s tructure. ................................ ........... 36 1 4 The general procedure of SPPS. ................................ ................................ ........ 37 1 5 Mechanism of Edman degradation. ................................ ................................ .... 38 1 6 The general procedure of MS based protein analysis. Figure is adapted from Ref 10 with permission. ................................ ................................ ...................... 38 1 7 Scheme of ESI source. Reproduced with permission from John Wiley and Sons, Inc. from Cech, N.B.; Enke, C.G. Mass Spectrom. Rev. 2001 20 364. ... 40 1 8 Comparison of tandem mass spectrometry in space and in time. Figure is modified from Ref 20. ................................ ................................ ......................... 40 1 9 Nomenclature of common peptide fragment ion types for a protonated pentapeptide. ................................ ................................ ................................ ...... 41 1 10 Structures of b ions that have been proposed. ................................ ................... 41 1 11 Sequence scrambling of a b 5 ion generated from a hexapeptide. Figure modified with Harrison et al. J. Am. Chem. Soc. 2006 128 10364. .................. 42 2 1 Ion cyclotron motion. A spatially uniform magnetic field has a direction that is perpendicular to and going into the plane of the paper and an ion is moving in the plane of the paper. The moving path of the ion is a circle resulting from the magnetic Lorentz force. Positive and negative ions travel in opposite directions. ................................ ................................ ................................ ........... 61 2 2 Ion path resulting from the combination of cyclotron, tapping, and magnetron motions. Figure reproduced from Ref. 80 with permission. ................................ 61

PAGE 11

11 2 3 Schematic diagram of Bruker APEXII 4.7T FTICR instrument. .......................... 62 2 4 FTICR MS performance as a function of magnetic field strength B 0 Figure reproduced from Ref. 80 with permission. ................................ .......................... 62 2 5 Schematic presentation of a cubic (left) and a cylindrical (right) FTICR MS analyzer cell. Both types of cell have six plates, and each pair functions fo r excitation, trapping, and detection. ................................ ................................ ..... 63 2 6 Schematic showing cross section of cylindrical ICR cell. Ions are excited to a larger radius (shown in blue) by applying an rf potential to the excitation plates. The motion of the ion packet is then detected on the detection plates in the form of an image current. ................................ ................................ .......... 63 2 7 Schematic showing how time domain data is converted to a frequency domain spectrum, followed by conversion to a mass spectrum. Figure taken from Ref 91 with permission of The Royal Society of Chemistry. http://dx.doi.org/10.1039/ b403880k ................................ ................................ ... 64 2 8 Demonstration of the effect of dataset size on resolution. The three mass spectra correspond to different acquisition dataset sizes for the same sample. Figure reproduced from Ref 91 with permission of The Royal Society of Ch emistry. http://dx.doi.org/10.1039/ b403880k ................................ ............. 65 2 9 The dependence of resolution on both measured m/z and the lowest m/z cutoff. Figure reproduced from Ref 91 with permission of The Royal Society of Chemistry. http://dx.doi.org/10.1039/ b403880k ................................ ............. 65 2 10 Schematic presentation of IRMPD mechanism. ................................ ................. 66 2 11 Schematic representation of the laboratory constructed FTI CR instrument equipped with an ESI source coupled to FELIX. Figure taken from Ref 104 with permission ................................ ................................ ................................ .. 66 2 12 Gas phase H/D exchange mechanisms proposed by Beauchamp and co workers. Figure is adapted from Ref 127 with permission. ................................ 68 3 1 HDX m ass spectrum with HDX products labeled. ................................ ............... 85 3 2 Inserts from the nozzle skimmer CID spectrum of octaglycine labeled with a 13 C Gly as the N terminal glycine residue, showing the b isotope distributions for (A) b 4 (B) b 5 (C) b 6 and (D) b 7 The 13 C labeled b n peaks denote b ions that incorporate the 13 C Gly label, whereas 12 C only b n peaks are entirely composed of 12 C Gly resi dues. ................................ ................................ ........... 86 3 3 The cartoon mechanism rationalizes the appearance of the 12 C only b 7 peak. .. 86

PAGE 12

12 3 4 IRMPD spectrum of the b 2 fragment generated from Gly Gly Gly, compared to computed spectra for (A) diketopiperazine structure protonated on a carbony l O, (B) oxazolone structure protonated on the oxazolone ring N, and (C) oxazolone structure protonated on the N terminus. ................................ ...... 87 3 5 IRMPD spectrum of protonated cyclo(Gly Gly), compared to calculated spectra for (A) diketopiperazine with proton pointing to CH 2 group and (B) diketopiperazine ......................... 88 3 6 Overlaid IRMPD spectra of protonated cyclo(Gly Gly) and b 2 from triglycine. .... 89 3 7 Possible structures for b 5 G 8 considered in theoretical study. (A) macrocycle structure protonated on backbone carbonyl, (B) oxazolone structure protonated on N terminus, and (C) oxazolone structure protonated on oxazolone ring N. ................................ ................................ ................................ 89 3 8 IRMPD spectrum of b 5 generated from octaglycine compared to the two lowest energy conformers for the various chemical structures: (A) macrocycle structure protonated on backbone carbonyl, (B) oxazolone structure protonated on N terminus, (C) oxazolone structure protonated on oxazolone ring N. The chem ical diagnostic bands and the relative energies to the lowest conformer are indicated. ................................ ................................ .... 90 3 9 Overlaid mid IRMPD spectra of b 5 G 8 b 5 G 5 and b 8 G 8 The chemically diagnostic modes are indicated. ................................ ................................ ......... 91 3 10 Representative H/D exchange (10 8 Torr CH 3 OD) mass spectra for (A) b 2 generated from tri glycine, (B) b 3 generated from octa glycine, (C) b 4 generated from penta glycine, (D) b 5 generated from octa glycine, (E) b 6 generated from octa glycine, (F) b 7 generated from octa glycine, and (G) b 8 generated from octa glycine. ................................ ................................ .............. 92 3 11 Kinetic fitting of the HDX results for glycine based b fragment ions. (A) b 2 generat ed from triglycine, (B) b 3 generated from octaglycine, (C) b 4 generated from pentaglycine, (D) b 5 generated from octaglycine, (E) b 6 generated from octaglycine, (F) b 7 generated from octaglycine, and (G) b 8 generated from octaglycine. (H) Relative abundanc exchanging structure as a function of b n fragment size. ................................ ......................... 93 3 12 The relative abundance of slow exchanging stru cture as a function of b fragment size. ................................ ................................ ................................ ..... 95 4 1 Chemical structure of Leu enkephalin. ................................ ............................. 109 4 2 Experimental IRMPD spectrum of b 2 compared with theoretical spectra by computational studies. (A) IRMPD spectrum of the b 2 f ragment generated from protonated Leu enkephalin, (B) diketopiperazine structure protonated on a carbonyl O, (C) oxazolone structure protonated on the on the N

PAGE 13

13 terminus, and (D) oxazolone structure protonated on the oxazolone ring N. The scaling factor is 0.965. Corresponding structures are presented on the right. The site of proton attachment (red arrow) and oxazolone rings (black arrow) are indicated. ................................ ................................ ......................... 110 4 3 Overlaid mid IRMPD spectra of b 2 b 3 and b 4 The spectrum of b 4 is adapted from previous publication of Polfer at el. 39 The chemically diagnostic modes are indicated. ................................ ................................ ................................ .... 111 4 4 H/D exchange (10 8 Torr CH 3 OD) mass spectra for (A) b 2 (B) b 3 and (C) b 4 for different exchange times. ................................ ................................ ............ 112 4 5 Structure of b 2 of Try Gly of an oxazolone structure protonated at the N terminus. Exchangeable hydrogens are shown in red. ................................ ..... 112 4 6 Kinetic fitting of the HDX results for (A) b 2 (B) b 3 and (C) b 4 .......................... 113 5 1 Overlaid IRMPD spectra of protonated cyclo( QWFGLM ) and b 6 from QWFGLMPG and Ac QWFGLMPG ................................ ................................ 126 5 2 Natural logarithm of relative d 0 depletion, ln[d 0 n ], as a function of time for (A) protonated cyclo(Gln Trp Phe Gly Leu Met), and (B) b 6 generated from protonated linear QWFGLMPG. ................................ ................................ ....... 127 5 3 Mass spectra of Ac QWFGLMPG b 6 ion after exchanging with CH 3 OD for (A) 0s and (B) 90s. No exchange is seen after 90s of exchange, indicating that such reaction is extremely slow. ................................ ................................ ....... 127 5 4 Schematic presentation of an oxazolone exchanges with CH 3 OD. .................. 128 5 5 LC/MS spectrum of linear Q(Trt)PFGLM after purification. ............................... 128 5 6 LC/MS spectrum of aliquot taken from cyclization reaction solution. The top figure is the HPLC spectrum of the aliquot. The middle figure is the MS spectrum of the synthetic cyc lo(QPFGLM) with a retention time of 6.14 min. The bottom figure is the MS spectrum of the synthetic cyclo(QPFGLM) 2 with a retention time of 6.95 min. ................................ ................................ ............. 129 5 7 IRMPD spectra of (A) QWFGLMPG b 6 (A) QPFGLMPG b 6 and (A) QWPFGLMPG b 7 In the inserts, in the intensities are magnified by 5x in the range of 1750 1930 cm 1 ................................ ................................ ................. 130 5 8 Overlaid IRMPD spectra of cyclo(QPFGLM) and QPFGLMPG b 6 The oxazolone C=O stretch band is pointed by the red arrow. ................................ 131 5 9 The structures of the b 6 ions from four peptides of QWFGLG system. The possible attack from the N terminus is labeled in blue, and the one from the glutamine side chain is labeled in orange. ................................ ........................ 132

PAGE 14

14 5 10 IRMPD spectra of b 6 ions generated from (A) QWFGLGPG, (B) Ac QWFGLGPG, (C) Q(N ethyl)WFGLGPG, and (D) Ac Q(N ethyl)WFGLGPG. The range for oxazolone bands is highlighted in pink. ......... 133

PAGE 15

15 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DEPENDENCE OF HEAD TO TAIL CYCLIZ ATION ON PRIMARY STR UCTURES OF PEPTIDES IN COLLISIO N INDUCED DISSOCIATION By Xian Chen December 2010 Chair: Nicolas C. Polfer Major: Chemistry Collision induced dissociation (CID) of peptides to deduce their sequences is the key technology in identifyin g proteins by mass spectrometry. However, during activation the linear peptide structures can undergo a head to tail cyclization reaction, where the N and C macrocycle structure. When this macrocycle structur e opens up at a different site than where it was originally formed, a scrambling of the sequence information will occur. There are few techniques that yield direct structural information on the minute quantities of gas phase ions inside mass spectrometers such as infrared spectroscopy, ion mobility, and hydrogen/deuterium exchange (HDX). IR spectroscopy confirms chemical structures based on diagnostic vibrations, however it is difficult to obtain relative abundances. Conversely, ion mobility can yield rel ative abundances of structures, but the structural interpretation is often ambiguous. HDX ha d so far not produced either information. In our research, we combined infrared multiple photon dissociation (IRMPD) spectroscopy and gas phase HDX to structurally characterize and quantify macrocycle as well as oxazolone selected peptide systems.

PAGE 16

16 For a series of glycine dependency, where smaller b fragments exclusively adopt oxazolones while larger fragments display a mixture of oxazolones and macrocycle s. The results are consistent with the finding from HDX, where smaller b fragments display a single HDX rate, whereas larger fragments show two distinct rates. Relativ e abundances of oxazolones and macrocycle s are approximated from HDX kinetic fitting. Similar trends are found in the peptide Leucine enkephalin (YGGFL). T he correlation between peptide sequence and propensity for macrocycle formation for the b 6 motif QWF GLM is investigated The IR MPD spectrum for the b 6 fragment of QWFGLMPG is nearly identical to that for protonated cyclo(QWFGLM), which confirms the exclusive presence of macrocycle structures for b 6 (from QWFGLMPG). The incorporation of a proline in for i nstance Q P FGLMPG is found to reduce the propensity of the formation of macrocycle structures in the corresponding b 6 fragment. A systematic chemical protection study was performed to the QWFGLG system. IRMPD spectra of b 6 ions generated from QWFGLG PG, Ac Q WFGLG PG Ac Q(N ethyl)WFGLG PG and Q(N ethyl)WFGLG PG indicate that no cyclization from the glutamine side chain occurs.

PAGE 17

17 CHAPTER 1 INTRODUCTION Peptide Chemistry The four main groups of molecules that play crucial role in cellular function are proteins, n ucleic acids, lipids, and carbohydrates. A protein is a molecule that has more than 50 amino acid residues, while a peptide is composed of less than 50 residues. Amino Acids Amino acids are the building blocks of all peptides and proteins. The structure of a typical amino acid is illustrated in Figure 1 1A Bonded to the alpha carbon are one amino group, one carboxylic group, one hydrogen, and one distinct functional group R on the side chain. It is the side chain group R that differs for each amino acid. A n alpha amino acid is formed when all these three functional group are attached to one alpha carbon atom. Except for the glycine, which has a hydrogen atom at the side chain, the four groups attached to the alpha carbon atom are different, the carbon atom then allows for steroisomerism. As a result, amino acids are chiral. There are two enatiomers : L and D amino acids b ased on the position of the amino group and hydrogen. Amino acids having amino group locating at the left hand side of the chiral center a tom have L configuration, and D amino acid are those with a hydrogen atom on the left hand side, as illustrated in Figure 1 1 B E. All amino acids found in proteins have the L configuration. The amine group from one amino acid can react with the carboxyli c group from another amino acid, eliminating a water molecule and forming a molecule where the two amino acid residues link via a covalent amide bond, which is commonly called the peptide bond in peptides and proteins. Peptides are formed through a seri e s of such

PAGE 18

18 condensation reactions. Figure 1 2 shows how a dipeptide is formed through the loss of a water molecule There are twenty amino acids commonly found in human proteins, and the vast majority of proteins in nature are composed of these twenty amino a cids (with the exception of some fungi) The structures and abbreviations of these common amino acids are shown in Figure 1 3. Commonly amino acids are classified into four categories based on the polarity of the side chains: nonpolar or hydrophobic amino acids, neutral (uncharged) but polar amino acids, acidic amino acids, and basic amino acids. 1 Peptide Synthesis As mentioned above, peptides are formed through the polymerization of amino acids. However, during synthesis it is possible that internal reaction will occur and result in unintended products. For example, to synthesize a dipeptide of Ala Gly, simply mixing the two amino aci ds Ala and Gly together will yield not just the target peptide Ala Gly, but also by products Ala Ala and Gly Gly. In order to direct synthesis, it is important to have protecting groups. Nowadays, solid phase peptide synthesis is the most popular method. F moc solid phase peptide synthesis Solid phase peptide synthesis (SPPS) was developed by Robert Bruce Merrifield, 2 whose contribution to the solid phase chemical synthesis was recognized by the 1984 Nobel Prize in Chemistry. Peptides are synthesized from the C terminus, or carboxy l group side. The general procedures of SPPS are shown in Figure 1 4. The ass embly of a peptide chain involves washing coupling The C terminal amino acid residue of the desired peptide is attached to an insoluble

PAGE 19

19 porous solid bead (resin) via its carboxyl group. Functional groups on amino acid side chains are blocked by permanent protecting groups that cannot be removed under the conditions during assemb y The temporary protecting group on the alpha amino group The amino group is temporarily protected while its carboxyl group is activated, along with coupling reagents. The two amino acids are then linked via an amide bond resulting in a dipeptide that is protect ed on the N terminus. Notably, the N terminal protected amino acids and coupling reagents are all soluble, whereas the peptide is immobilized on the resin, thus, after each coupling step excess reagents and solubl e by products can be removed by washing Following the washing step, the protecting group on the amino group of the dipeptide is then removed, prior to addition of a third amino acid residue. This process is repeated until the desired peptide sequence is a ttained The deprotection of the final N terminus protecting group is done prior to the final step, chain protecting groups are removed. Typically the resins and side chain protecting g roups are chosen so that they can be removed under the same conditions. The major limitation in SPPS is the final yield. To obtain pure and high yield product requires extremely high yield in every step. The effect of chemical efficiency of each step on t he yield of the final product is shown in Table 1 1. The by products, resulting from incomple te coupling, often have very similar properties as the target peptide, hence making the purification difficult. 3

PAGE 20

20 The two major SPPS methods involve Fmoc and Boc chemistr ies and differ from each other by the N termin al protecting groups and the resins. The Boc method name is derived fr om the protecting group on the amine tert butoxycarbonyl (Boc), the removal of which necessitates trifluoroacetic acid (TFA). The resin used in the Boc method is h oxymethylphenylace tamidomethyl polystyrene (PAM) resin, and to release the peptide requires the use of hydrogen fluoride (HF). The Boc method can give very high yield and thus is powerful in synthesis of larger peptides and proteins. 4 However, the need to use toxic HF requires special equipment and caution. Fmoc chemistry is a more popular method because it is safe and easy to operate. Similarly to the Boc method, the Fmoc method is named after its amino protecting group 9 fluorenylmethyloxycarbony (Fmoc). The deprot ecting reagent used is 20~50% piperidine in N,N dimethylforma mide (DMF). Coupling is achieved by addition of Fmoc protected amino acids and coupling reagent s diisopropylethylamine (DIPEA) and 2 (1H Benzotriazole 1 yl) 1,1,3,3 tetramethyluronium hexafluoro phos phate (HBTU). Cleavage of the peptide from the resin and removal of the protecting groups on side chains is done with 95% TFA. Synthesis of macrocyclic peptides Macrocyclic peptides are cyclic polypeptides who se amino and carboxyl termini are linked together via a peptide bond in a circular chain. C yclic peptides often play important roles in biological processes, such as for instance gramicidin S, which is an antibiotic agent This explains the interest for synthetic approaches in making cyclic pepti des. Cyclization of linear peptides with side chain protected in the solution phase is the widely used method. Synthesis of intramolecular head to tail cyclic peptides is normally performed in dilut e conditions (10 3 10 4 M ) 5 to minimize the competing

PAGE 21

21 polymerization of linear peptides. Given the dilute solution conditions and compet ing reactions, cyclization reactions are typically slow and give low yield. Protein Identification Each protein is unique from others based on its distinctive sequence of amino acids in the polypeptide chain Note that by convention the amino acid seque nce of a peptide or protein is read from the N terminus of the polypeptide chain to the C terminal end. 1 The primary structure of the protein also gives rise to higher order structure, which account for its distinctive functions in biological system. T h e amino acid sequence is based on the genetic information encoded by DNA Given the human genome project, and other genome sequencing p rojects, a wealth of information is now available on expected protein sequences, and this information is stored in DNA/protein databases. Nonetheless, information on the genetic make up of cells does not directly correlate with biological function. Instead the primary actors in cells are proteins, the expression of which is steered by highly in complex cellular regulatory networks. 6 The central theme in proteomics is to identify the role of each protein in these processes. For that purpose, it is essential that proteins can be identified from minute and highly heterogeneous biological samples. Since the identity can be established from its sequence, analytical techniq ues are required that can confirm the primary structures of proteins. Twenty years ago, Edman degradation an automated, stepwise chemical degradation, was used to digest and identify the amino acid sequence of proteins. The method has been largely replace d by mass spectrometry Since the development of two soft ionization methods electrospray ionization (ESI) 7 and matrix assisted laser desorption/ionization (MALDI) 8,9 in the late 1980s, mass spectrometry (MS) has become

PAGE 22

22 a key approach in the field of proteomics. These methods solved the difficult problem of generating ions from large, nonvolatile analytes such as proteins and peptides without inducing analyte fragmentation. C ompared to Edman degradation, MS is much more sensitive, can analyze peptides in seconds, does not require proteins or peptides to be purified, and has no problem in identifying blocked or modified proteins. 10 Tandem MS (MS/MS) involves mass isolating a peptide of interest and subject ing it to fragm entation through collision s The mass to charge ratios ( m/z ) of the fragment ions are then employed to derive information on the amino acid sequence of the peptide Edman Degradation Edman degradation, developed by Peer Edman, was the key technique in prot ein and peptide sequencing In this method, amino acids are chemically cleaved in a stepwise way from the amino terminus of the proteins, followed by identification. The mechanism of Edman degradat ion is illustrated in Figure 1 5 In Edman degradation, phenylisothiocyanate (PITC) reacts with the alpha amino group at the N terminus of a peptide, forming a phenylthiocarbamyl (PTC) adduct. Under anhydrous acidic conditions the N terminal amino acid residue is then cleaved from the peptide chai n, resulting in a heterocyclic derivative through the attack of the sulfur of the PTC adduct on the carbonyl component of the first peptide bond. The cleaved amino acid derivative is separated from the residual peptide by extraction with an organic solvent and then gets identified by ultraviolet (UV) absorbance spectroscopy. The remaining peptide chain is then subjected to further cycles of coupling of PITC, cleavage and identification. 11 Edman degradation is capable of reasonable sensitivity (i.e., 10 100 picomoles ) However, its major limitation is the need of a free amino group to couple PITC H ence

PAGE 23

23 peptides cannot be identified if they are acetylated. In addition to needing purified samples, one of the main drawbacks of Edman degradation is the time require d for analysis, which can stretch from hours to days for a single peptide/protein Protein Identification with Mass Spectrometry Mass spectrometry (MS) is currently the most popular method for peptide/protein identification due to its higher sensitiv ivity (i.e., attomol, 10 15 mol) and that the identification of a single peptide can be done in the timeframe of seconds. Furthermore, compared to Edman degradation, MS does not require purificat ion of peptides or proteins, and can identify modified proteins (e.g., N terminal acetylated proteins). The general procedure for MS based proteom ics is illustrated in Figure 1 6 A protein is first prepared from a biological sample, by sodium dodecyl sul fate polyacrylamide gel electrophoresis ( SDS PAGE ) The gel bands correspond to individual proteins and can be cut out using a scalpel. The protein is digested into peptides using an enzyme (e.g. trypsin). The extracted peptides are then ionized by ESI or MALDI for mass analysis In order to reduce sample complexity, the peptides may be further separated by high performance liquid chromatography (HPLC) prior to ionization. Hyphenated HPLC separation and ESI can in fact be carried out on line with mass spect rometric analysis. In the mass spectrometer, the mass of the ionized peptide of interest is first measured. The peptide is then isolated and subjected to fragmentation, where the ionized peptide is broke n into smaller pieces (i.e., fragments). The mass spe ctra of these fragment ions contain sequence specific information. Lastly, the peptide sequencing data that are obtained form the mass spectra are searched against protein databases using one of a number of database searching programs, 10 which are listed in Table 1 2.

PAGE 24

24 The method described above is also called the as the peptide sequence information from an enzymatic digestion is used to identify a whole protein One difficulty with this method is that the complete protein sequence can only be obtained when the peptide coverage is also complete ; often, this is not possible, as some peptides have low ionization efficiencies 12,13 An alternative approach for protein identification involves ionizing and fragmenting the entire protein in the m ass spectrometer, and is hence referred to as the method 14 18 Given the highly complex mass spectra in this approach, this requires the use of high resolution and high mass accuracy mass spectrometers, s uch as Fourier transformed ion cyclotron resonance (FTICR) mass spectrometer s.Note that the theory of FTICR MS will be discussed in Chapter 2. protein identification by mass spectrometry, as it is compatible with all mass analyzers. Peptide Fragmentation information of a peptide, the peptide has to unde rgo fragmentation, where it is broken up into smaller pieces. Peptide fragmentation in the mass spectrometer is the crucial step in peptide sequencing. In this section, the basic concept and theory of mass spectrometry techniques will be introduced, and te chniques for peptide fragmentation will be discussed. Mass Spectrometry Among all the analytical techniques, mass spectrometry is the most widely applicable. This technique can be used to get the elemental composition of samples of matter, to confirm the s tructures of inorganic, organic and biological molecules, to do

PAGE 25

25 both qualitative and quantitative analysis of complex mixtures, and to get isotopic ratios of atoms in samples. 19 A mass spectrometer has three major parts: an ion source, a mass analyzer and a detector. In a typical analysis with a mass spectrometer, analytes have to be first transformed into gaseous i ons, which is done at the ion source. The generated analyte gas phase ions are then transferred through a series of ion optics to the mass analyzer, where the mass to charge ratios ( m/z ) are measured. A detector is used to count the ions from the mass anal yzer and to measure their abundance. Electrospray ionization A number of ionization techniques have been developed and they are adopted based on different research needs. These techniques can be classified as hard and soft ionization methods. With har d ionization method, samples are not just get ionized, but also dissociated during the ionization process. Soft methods produce little or no fragmentation during the ionization process. One of the main examples as a hard ionization technique is e lectron i onization (EI) which is particularly suitable in the analysis of organic compounds. In this method, electrons are emitted from a heated filament by thermionic emission The electron beam is intersected with gaseous molecules of the sample of interest. In positive ion mode, the incident electron ejects an electron from the molecule generating an ion in the process The number of ions I produced per unit time can be calculated using the equation 20 : I=NpiV (1 1) where N is a constant proportionality coefficient, p is the pressure, i is the electron current, and V is the volume.

PAGE 26

26 In terms of peptide/protein analysis, EI is not a useful technique. EI cau ses extensive dissociation of amino acids. Further, the vapor pressure of peptides/proteins is much too low to bring sufficient densities into the gas phase. In fact, heating the sample to achieve higher vapor pressures results in dissociation of the protein, as opposed to their tr ansfer into the vapor phase. The ionization techniques that have revolutionized biological mass spectrometry of macromolecules are e lectrospray ionization (ESI) 7 and matrix assisted laser desorption/ionization (MALDI) 8,9 Both of these are soft ionization techniques, as they are capable of transferring large molecules in the gas phase while leaving them intact. This dissertation will focus on ESI since it is the only ioniza tion technique employed throughout the research. ESI is an atmospheric pressure ionization method While the phenomenon of electrospray has been known for more than a century, Fenn and co workers first coupled electrospray to a mass analyzer for the purpos e of ionizing large molecules in Since then, the technique has been employed to routinely ionize a wide range of molecules, including lipids, carbohydrates, polymers, peptides, proteins, protein complexes and even entire viruses The capabilit y of ESI for transferring very large complexes into the gas phase in unparalleled, and no upper mass limit has been determined so far. ESI allows very high sensitivity to be reached and is easy to couple to separation techniques, such as high performance l iquid chromatography (HPLC) or capillary electrophoresis. 20 In ESI, gaseous ionized samples are produced by applying a high electric field to solution phase molecules contained in droplets. Figu re 1 7 illustrates the ESI process. A high voltage drop (2 5 kV) is applied between the capillary and the metal plate, which

PAGE 27

27 is near the entrance to the mass spectrometer. The electric field induces a charge accumulation at the liquid surface located at t he end of the capillary, which will break to form highly charged droplets. 20 generate the droplets, and the onset voltage depends on the surface tension of the solvent. At the end of the capillary, when the onset voltage is reached, the pressure is higher than the surface tension, then the shape of the flow will change to a Taylor cone and small droplets are released. These droplets contain solvent molecules, sample molecules, and charges. As they drift toward the mass spectrometer entrance, solvent evaporation causes the charged droplets to undergo Rayleigh explosions since the surface tension of the droplet can no longer withstand the Coulombic r epulsion. This solvent evaporation is aided by a heated nebulizing gas (normally inert gas, e.g. N 2 ). 21,22 The process of Rayleigh explosions is repeated until the analyte ion is brought into the gas phase. There is still contention on the exact mechanism for the final step, when the ion is brought into the vapor phase. It is generally accepted that smaller ions may be capable to escape the droplet from the surface, while larger ions require a boiling off of all solv ent molecules. 23,24 Tandem mass spectrometry The charged analyte produced with ESI enters the mass spectrometer through the inlet, and its mass to charge ratio is then measured in the mass analyzer. To solve the pro blem of coupling an atmospheric pressure ionization source to a mass analyzer which must be operated in vacuum conditions (i.e., <10 5 mbar) the inlet has to restrict air flow into the mass spectrometer, and multiple pumping stages are required For prot eomics measurements mass spectrometer can be used either to measure simply the molecular mass of a polypeptide, or to determine additional structural features

PAGE 28

28 including the amino acid sequence or the site of attachment and type of post translational modif ications. 6 For the first purpose, a single stage mass spectrometer is used. In the second case, tandem mass spectrometry has to be utilized. Tandem mass spectrometry (MS/MS) is a method that involves at least two stages of mass analysis. In such an experiment, after determining the mass of a precursor ion, the ion is isolated and then subjected to fragmentation. The analysis of the fragment ions is essential in struct ural elucidation There are two types of instrument s that can carry out MS/MS measurements : tandem mass spectrometer s in space or in time A tandem mass spectrometer in space has two distinct mass analyzers coupling together, while the one in time is done via performing a sequence of events inside one ion storage device. The comparison of a product ion scan performed by a space based and a time based i nstrument is shown in Figure 1 8 In our research, we use Fourier transform ion cyclotron resonance mass s pectrometry (FTICR MS). Its mass analyzer is the ICR cell, which is an ion trapping device and hence, the tandem mass spectrometry experiment is performed in time. Fragmentation Methods To get the sequence of a peptide, one has to break the peptide into fragment ions. Generally, this is done for protonated peptide cations. The nomenclature 25 for peptid e fragment products is shown in Figure 1 9 Cleavage of a chemical bond along the backbone will result in either a b or c ions if the charge is retained on the N terminal side, or x y and z ions if the charge is on the C terminal side. The subscript in dicates the number of amino acid residues present in the ion counting from either the N or C terminus For the pentapeptide in Figure 1 10 when the peptide bond that links the third and the fourth residue dissociates, either a b 3 or y 2 ion will be produ ced.

PAGE 29

29 There are several methods to fragment peptides, including electron capture dissociation (ECD), 26 29 blackbody infrared radiative dissociation (BIRD), 30 surface induced dissociation (SID), 31 electron transfer dissociation (ETD), 32 collision induced dissociation (CI D), 33,34 and infrared multiple photon dissociation (IRMPD) 35 The most widely used ion activation and dissociation process is collision induced dissociation (CID), where the mass selected protonated peptide collides with neutral backgrou nd gas molecules (typically He, N 2 and Ar) in a collision cell or ion trap. Generally, the masses of the CID product ions are expected to reflect the amino acid sequence of the peptide, and therefore be valuable for sequencing. 36,37 In low energy CID, bond breakage mostly occurs at the peptide (i.e., amide) bonds, which are often the lowest energy pathways, leading to b and y ions. In CID, the ion is vibrationally excited. Vibrational excitation o f ions can also be achieved by absorption of infrared photons from an infrared laser. In i nfrared multiple photon dissociation (IRMPD) similar fragmentation is observed as in CID. In Chapter 2, the technique of IRMPD spectroscopy will be introduced, which is a method to record the infrared spectra of ions. This technique is based on the premise that the ion absorbs a number of photon (i.e., tens to hundreds) when a vibrational mode of the molecule is resonant with the laser wavelength. 38,39 Mobi le Proton Model In the fragmentation of protonated peptides, the proton in fact plays an important role in the dissociation mechanism. T Wysocki and co workers, 40 and Gaskell and collaborators 41 states that the proton migrate s along the backbone to induce charge site initiated dissociation In particular, the attachment of the proton to the amide NH is thought to weaken the amide bond, and

PAGE 30

30 hence promote cleavage at that site. Basic amino acid side chains (i.e., Arg, Lys, His) sequester the proton more tightly, and for such peptides higher activation energies are required. Theoretical research has made important contributi on to the development of this theory. To evaluate the competing reaction pathways that take place, Paizs and competition 42 Computations based on this model provide information about the competing fragmentation mechanisms after the liberation of the mobile protons on to the peptide backbone In term of the different reaction mechanisms, experimental approaches are required to validate the chemical structures that are formed. 43 Structures of b Fragment Ions The chemical structures of b ions had been the subject of a debate in the mass spectrometry community. b 1 ions are very instable in mass spectrome ter, so it is normally not observed in MS spectra, whereas larger b ions are often abundant in the CID spectra. It was initially thought that b ions form acylium structures 44 as shown in Figure 1 10 (A). However, acylium structures are not stable and are expected to spontaneously lose CO. In order to rationalize stable b ions, Harrison and co workers suggested that oxazolone structure s are formed via nucleophilic attack from a backbone carbonyl (structure shown in Figure 1 10B ). 45 50 The first direct evidence for oxazolone structure s was obtained by IRMPD spectroscopy by Polfer et al. 39 In their study the IR spectrum of b 4 ions generated from Leu enkephalin (YGGFL) showed that the oxazolone diagnostic C=O stretch bands were observed, whereas the diagnostic acylium C=O stretch mode was not seen

PAGE 31

31 A third possibl e b fragment structure was first proposed by Wesdemiotis and co workers, involving a cyclization via the N terminus, in a so called diketopiperazine structure, as shown in Figure 1 10C. For smaller b 2 ions, the evidence so far has shown that diketopiperazi ne structures are rarely formed. 51 55 For larger b ions, however, there is increasing evidence that such head to tail macrocycle structures are possible. The presence of such macrocycle structures is particularly wo rrisome, since their presence can explain some of the sequence permutations that are observed in CID. 56 58 With respect to peptide sequencing via MS/MS, it is possible to add substantial internal energy and the frag ment ions of peptides can undergo rearrangement processes, leading to the formation of a macrocycle struc ture, as shown in Figure 1 10D Oxazolone structures can isomerize to the macrocycle structure. When the macrocycle structure is present, reopening of this macrocycle structure will lead to the scrambling of peptide sequence. The mechanism is proposed by Harrison and Paizs as illustrated in Figure 1 11 for a b 5 ion made from protonated hexapeptide R 1 R 2 R 3 R 4 R 5 R 6 where R i is the name of an amino acid. The hea d to tail cyclization reaction results in a macrocyclic b 5 ion, and the reopening leads to oxazolones with original and permutated sequences. Further fragmenta t ion of b 5 with original sequence will generate direct sequence ions, which contain informati on of the initial sequence. However, fragmentation of permutated oxazolones will lead to non direct sequence ions, in which the original primary sequence information is lost. The phenomenon is problematic for MS based peptide sequencing because the algorit hm s of current software for peptide sequencing do not take this into consideration, which might lead to false identification

PAGE 32

32 The phenomena of sequence scrambling have been observed in several peptide systems with MS/MS method. CID is performed to generat e fragments from ions of interest and the masses of fragment ions are then collected and analyzed. Any internal neutral loss from an ion may suggest that a macrocycle structure is formed prior to the fragmentation, and may also gives information about frag mentation pathways. 59 62 Experimental methods for investigat ing b ion structures include IRMPD, isotope labeling, ion mobility spectrometry (IMS) and gas phase hydrogen/deuterium exchange (HDX). Since it was shown that IRMPD spectroscopy can provide direct evidence for chemical moieties, 63 it has become a popular approach. 39,51,53 55,64,65 IMS is also a very useful technique that can distinguish isomers by their different collision cross sections in the drift tube. Polfer et al. reported a work done by ion mobility spectroscopy (IMS) that showed further evidence of the presence of cyclic b 4 and a 4 ions of protonated YGGFL. 66 Gaskell and coworkers published a paper showing further evidence of cyclic b 5 ion of protonated YAGFL NH 2 by IMS 67 Gas phase HDX has been employed to structurally characterize b ions. 54,55,65,68 72 The pre mise of this approach rests on the fact that isomers of b ions are expected to exchange with the deuterating reagent (e.g. CH 3 OD) at different rates. In other words, if multiple chemical structures are present, more than one HDX kinetic rate should be obse rved. The mechanism and theory of gas phase HDX will be discussed in detail in Chapter 2. Objective of This Research This thesis describes a systematic study of the effect of primary structure on the formation of oxazolone / macrocycle b structure s under CID conditions All experiments were conducted on FTICR instrumentation, where the peptide of interest was generated

PAGE 33

33 by ESI. IRMPD experiment s w ere performed at the Free Electro n Laser for Infrared eXperiments (FELIX) facility (Nieuwegein, The Netherlands) wi th the tunable free electron laser HDX experiments were carried out at the University of Florida, by introducing CH 3 OD vapor into ICR cell to react with isolated ions of interest. Computations were conducted at the High Performance Computing (HPC) center at the University of Florida. In Chapter 2, the main experimental techniques employed here are introduced, namely FTICR mass spectrometry, IRMPD spectroscopy and HDX. In Chapter 3, a series of b ions generated by CID from oligoglycines were used to invest igate the size effect on the formation of the macrocycle structure. Chapter 4 discusses a similar study for the peptide Leu enkephalin (YGGFL) In Chapter 5, synthetic chemistry approaches are employed to create reference compounds for IRMPD spectroscopy a nd to investigate the effect of site mutations in the primary structure on the prevalence of oxazolone / macrocycle structures. Lastly, in Chapter 6, future directions are discussed that could lead to further insights in the dissociation chemistry of protona ted peptides.

PAGE 34

34 Figure 1 1. The chemical s tructure of an alpha amino acid. (A) An alpha amino acid which has one amine group, one carboxylic group and one distinct functional group (R) attached to one alpha carbon atom ; (B) Fischer projection formulas for an L amino acid; (C) Fischer projection formulas for a D amino acid; (D) Stereo representations for an L amino acid; and (E) Stereo representations for a D amino acid.

PAGE 35

35 Figure 1 2. A di peptide is forme d through a condensation reaction Peptide bon d is bracketed in red and the N and C termini are indicated in blue and purple, respectively.

PAGE 36

36 Figure 1 3. The twenty amino acids found in protein s Their names, three letter and one letter abbreviations are given below each structure.

PAGE 37

37 Figure 1 4. T he general procedure of SPPS.

PAGE 38

38 Figure 1 5 M echanism of Edman degradation. Figure 1 6 The general procedure of MS based protein analysis Figure is adapted from Ref 10 with permission.

PAGE 39

39 Table 1 1. Effect of yield per coupling step on final produ ct yields. 3 No. of coupling steps Yield per coupling step (%) 100 99 95 90 10 Overall yield 100 90 60 35 20 100 81 36 12 30 40 50 100 74 21 4 100 67 13 1 100 61 8 <1 Table 1 2. Sources for MS based protein identification tools 73 Application Sponsor Uniform resource locator (URL) MassSearch Eidgenossische Technische Hochschule http://cbrg.inf.ethz.ch PeptideSearch European Molecular Biology Laboratory http://www.mann.emblheidelberg.de ExPAS y Swiss Institute of Bioinformatics http://www.expasy.ch/tools Mascot Matrix Science http://www.matrixscience.com PepFrag, ProFound Rockefeller University http://prowl.rockefeller.edu MOWSE Human Genome Research Center http://www.seqnet.dl.ac.u k MS Tag, MS Fit, MS Seq University of California http://prospector.ucsf.edu COMET Institute for Systems Biology (COMET) http://www.systembiology.org SEQUEST University of Washington htt p://thompson.mbt.washington.edu/ sequest

PAGE 40

40 Figure 1 7. Sch eme of ESI source. Reproduced with permission from John Wiley and Sons, Inc. from Cech, N.B.; Enke, C.G. Mass Spectrom. Rev. 2001 20 364. Figure 1 8 Comparison of tandem mass spectrometry in space and in time. Figure is modified from Ref 2 0

PAGE 41

41 Figu re 1 9 Nomenclature of common peptide fragment ion types for a protonated pentapeptide Figure 1 10. Structures of b ions that have been proposed.

PAGE 42

42 Figure 1 11 Sequence scrambling of a b 5 ion generated from a hexapeptide Figure modified wit h Harrison et al. J. Am. Chem. Soc. 2006 128 10364.

PAGE 43

43 CHAPTER 2 INSTRUMENTATION AND METHODS Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Fourier transform ion cyclotron resonance mass spectromet ry (FTICR MS) is a metry technique FTICR MS has the highest resolving power and mass accuracy, and is thus very suitable for analysis of complex systems, e.g. petroleum 74 a nd proteins 75 79 There are a few excellent reviews on FTIC R MS available. 80 82 M ode rn FTICR MS derives from the invention of the cyclotron by Lawrence in the 1930s, when Law rence and co workers built the first cyclotron accel erator and used it to investigate atomic fundamental properties. 83 Two decades later, Simmer and co workers incorporated the theory of ICR to a mass spe ctrometer called the omegatron 84 In 1974, Comisarow and Marshall first described the concept of combining the Fourier transform with ICR mass spectrometry, and built the first FTICR MS ins trument. 85,86 Fourier transformation a llows one to measure several ions in the cell at the same time, instead of only measuri ng one ion at a time. Since that time, interests in FTICR MS have greatly increased and so has the number of FTICR MS instruments. Nowadays, commercial FTICR MS instruments are offered by several companies on the market. Principles Ion cyclotron motion T he mass analyzer in FTICR MS is the ICR cell, on which a magnetic field and an electric field is applied. The magnetic field is uniform, unidirectional, and homogeneous. The motion of ions in the ICR cell is determined by the combination of the magnetic an d electric fields. When an ion moves in a magnetic field (B) and an electric field (E), it is

PAGE 44

44 subject to the Lorentz force (F), as described in Equation 2 1, where m is the mass of the ion, a is the acceleration, q is the charge that the ion carries, v is the ion velocity, and F E and F M refer to the electric and magnetic components of the Lorentz force. (2 1) The ion cyclotron motion arises from the interaction of an ion with the magnetic field. When a m agnetic field has a direction that is perpendicular to the direction of the ion velocity, the direction of magnetic component, F M is perpendicular to the plane determined by v and B If there are no collisions, the velocity of the ion will remain constant and hence the ion will be forced by the magnetic field to move in a circular path with a radius of r as shown in Figure 2 1. If we set the z axis along the direction of the magnetic field B, then an x y plane can be defined perpendicular to z. The ion velocity on the x y plane is denoted as v xy and hence the angular acceleration, dv xy /dt can be expressed as v 2 xy /r When the electric field is missing, and we only consider the magnitude, Equation 2 1 then becomes: (2 2) Equation 2 2 can then be rearranged to yield the cyclotron motion radius: (2 3) Since the kinetic energy KE=mv 2 /2 v can be written as then Equation 2 3 can also be expressed as: (2 4)

PAGE 45

45 This equation shows that the radius of an ion in the ICR cell depends on the kinetic energy. Thus, in order to be detected, the ion will need to be excited so that the cyclotron radius is increased to a significant fraction of the dimensions of the analyzer cell. 82 The angular velocity (in rad/s) is defined as: (2 5) Substituting Equation 2 5 in Equation 2 3 then becomes: (2 6) Rearranging this equation gives the cyclotron frequency c : (2 7) The cyclotron frequency can also be defined in Hz as: (2 8) In Equation 2 8, m/z is the mass to charge ratio of an ion. For example, a singly charged ion with a mass of 151 at 4.7T magnetic field has a frequency of 484.9 kHz: Equation 2 8 shows that the ICR frequency depends on the m/z and the magnetic field strength, while it is independent of the velocity and hence the kinetic energy. In essence, all ions with the same mass to charge ratio have the same cyclotron frequency. There are two more motions observed in ICR, the trapping and magnetron motions, which will be introduced in the next two sections. Of these three ty pes of

PAGE 46

46 motion, the ion cyclotron frequency has the highest magnitude and is the only one to be exploited in terms of mass analysis. Trapping motion When an ion travels along the z axis, parallel to the magnetic field, there is no force from the magnetic fi eld, and hence the ion motion along the z axis is unconstrained. To prevent the ions from escaping from the cell along the z axis, a trapping voltage is applied to the end caps or plates of the cell, and this electric potential causes another type of ion m otion the trapping motion. Typically, a three dimensional axial quadrupolar electrostatic trapping potential is used, and it has a form as: (2 9) where V trap is th e applied trapping voltage, r is the radial position of the ion in the x y plane and equals to a is the trap size, and and are constants that are determined by the trap shape. 80 87 Ion motion at the z axis can then be solved using Equation 2 9: (2 10) For an ion that oscillates along the z axis sinusoidally with time, its position can be expressed as (2 11 ) T he trapping frequency of the ion can then be described as : (2 12)

PAGE 47

47 Magnetron motion A c ombination of electric and magnetic fields produces a three dimensional trap which allows ions to be stored in the ICR cell for extensive intervals of time (i.e., seconds, minutes, and even hours). The magnetic and electric fields generate cyclotron and trapping motions independently, and the two motions do not couple. However, the combination of magnetic and electric fields together create another ion motion, called the magnetron motion. First, the trapping potential in Equation 2 9 creates a radial for ce, F radial expressed by: (2 13) The radial electric field that operate s on the ion generates an outward electric force opposite to the inward Lorentz magnetic force from the applied magnetic field. Combining Equation 2 13 and 2 5 gives (2 14) Solving Equation 2 14 we can then get two results for : (2 15 ) (2 16) in which (2 17)

PAGE 48

48 + is called the reduced or observed cyclotron frequency, which is reduced by the magnetron frequency, A calibra tion can relate the observed cyclotron frequency to the unperturbed cyclotron frequency, c = Bq/m. Combination of the three ion motions in the cell the cyclotron, trapping, and magnetron motions makes ions behave in a way illustrated in Figure 2 2. As mentioned before, trapping and magnetron frequencies are much lower than the cyclotron frequency, particularly for low DC trapping voltages. The cyclotron frequency is most useful in terms of its simple relationship to the m/z of the ions 80 Apparatus There are several types of FTICR instruments available, and they all have four specific components: a magnet, an analyzer cell, an ultra high vacuum, and a data system, which will be introduced in this s ection. A schematic diagram of the commercial FTICR mass spectrometer (Bruker Daltonics, Billerica, MA) equipped with a 4.7 T superconducting magnet that is employed in this research is shown in Figure 2 3. Magnet A magnet is required to create a strong m agnetic field. The performance of an FTICR mass spectrometer improves as the magnetic field strength increases, as shown in Figure 2 4. Some of these characteristics scale linearly, whereas others increase as a function of the square of the magnetic field strength. The magnet can either be a permanent magnet, an electromagnet, or a superconducting magnet. Permanent magnets have low field strengths (< 1 T), which limit the performance of the instrument and are hence not widely used. Electromagnets have highe r field strengths (~1 3T), which is good enough for ions of low mass to charge ratio. Superconducting magnets

PAGE 49

49 have been used since they can provide much stronger fields, generally 3~12T, with the largest commercial magnets now reaching 14.5 T. 88 However, there are two main drawbacks about superconducting magnets First, the cost of a superconducting magnet is high, as well as the maintenance for cryogens. Second, the sizes of these magnets are large, thus imposing space restrictions. Analyzer cell For biomolecular mass spectrometry, ions are made externally, usu ally by ESI or MALDI, and then injected into the cell. The analyzer cell in FTICR MS, also called the ICR cell, is surrounded by a magnet as shown in Figure 2 3. Ions are stored, mass selected, activated, excited, and detected in the ICR cell. The advantag e of an ion trap is that mass isolation and activation events can be carried out in time as opposed to in space and are hence only limited by ion signal. Historically, two common types of cell designs were employed, having cubic or cylindrical shape, as shown in Figure 2 5. The cubic cell was the first design, but has been largely phased out in favor of the cylindrical design. In the most basic design, ICR cells consist of three pairs of electrodes. The purpose of the trapping plates is to store ions insi de the cell via a static DC potential. A dipolar excitation signal, of the same frequency as the cyclotron frequency, is applied to the excitation plates to excite ions to larger radii. The detection plates detect the image current from ions excited to lar ger radii. These concepts are explained in more detail in 2.1.3. Ultra high vacuum system An ultra high vacuum system is critical for all FTICR instruments. While all mass spectrometers require vacuum to analyze and detect ion, FTICR MS is more sensitive

PAGE 50

50 to background pressure, due to the long distances that the ions travel. To prevent collisions of ions with other molecules background pressures of 10 9 10 10 Torr are required for optimal performance. To obtain such low pressures in combination with highe r pressure ionization sources, a series of pumping stages are employed. In an ESI source, mechanical pumps are used for the first (rough) pumping stageand then turbo molecular pumps are utilized for lower pressures. FTICR instruments are often equipped wit h pulsed valves or leak valves to allow introduction of a volatile neutral compound into the high vacuum region. This neutral molecule can then be employed in ion molecule reactions with the trapped ions, or can serve as a collision partner for dissociatio n. In this research, a Varian leak valve, equipped with a sapphire crystal, is used to introduce CH 3 OD vapor into the ICR cell to do exchange with the ions of interest that are trapped in the cell. Dissociation experiments were performed by momentarily int roducing a collision gas (e.g. Xe) into the cell via a pulsed valve, and applying an excitation pulse to the excitation plates. Data system Another main component of all FTICR instruments is a complex data system for acquisition and processing the data. With the data system, the signals produced by excited ions on the detection plates are collected and transformed into more straightforward information. The data system includes a frequency synthesizer, delay pulse generator, broadband radio frequency (r.f. ) amplifier and preamplifier, a fast transient digitizer and a computer system, similar to those used for FT NMR systems. Excitation and Detection When ions are transferred into the ICR cell, they are subject to motions discussed in Section 2.1.1. Initial ly, the radii of the cyclotron orbits are too small to be detected due

PAGE 51

51 to the fact that their kinetic energy in the xy plane is low (<1eV). 82 In order to be detected, the ions must be excited to a detectable radius. This is achieved by using a radio frequency (rf) p ulse to the two excitation plates at the resonant cyclotron frequency of the ions, as shown in Figure 2 6. Following excitation, ions of the same m/z move as a coherent packet undergoing the cyclotron motion. As the ion packet moves in between the detect ion plates, it induces a current in the detection circuitry, the image current. 82,89 Since the cyclotron motion of the ions is periodic, the generated image current is sinusoidal and of the same frequency than the cyclotron frequency of the ions Multiple m/z sweep. The image current on the detection plates then corresponds to a superposition of cyclotron frequencies. This time domain signal is known as a transient. In order t o applied, 90 which then yields the frequency domain spectrum, as shown in Figure 2 7. Since the cyclotron frequency is related to m/z the frequency domain spectrum can then be converted to a mass spectrum. Mass Accuracy FTICR MS has the highest mass accuracy and resolution of any mass spectrometric technique. Mass accuracy is often expressed in ppm (parts per million), and is an indicator of how well the measured m/z definition of mass accuracy is shown in Equation 2 18, by comparing the observ ed m/z (2 18)

PAGE 52

52 Other types of commercial mass spectrometers typically have mass accuracies of the orders of tens of ppm, whereas FTICR instruments usually give mass accuracies of 1 ppm or lower. 91 Mass Resolution Mass resolution is a measurement of how well adjacent peaks in a mas s spectrum can be separated. It is very important for complex mixtures or multiply charged ions, as in these cases the s of ions are closely spaced. For example, a fullerene derivative that contains nitrogen can be identified with a 9.4T FTICR mass spe ctrometer, as 12 C 59 14 N + ( m/z 722.002525) and 12 C 5 8 1 3 N 2 + ( m/z 722.006161), but cannot be distinuguished with a lower resolution mass analyzer. 91,92 Equation 2 19 defines the mass resolution: (2 19) In Equation 2 1 9, m is the m/z of the peak of interest and is either the difference in m/z between two adjacent peaks or the width of the peak. The resolution of many commercial time of flight mass spectrometers may achieve resolutions of 10 4 while FTICR mass spectrom eters can reach resolutions >10 5 The only mass analyzer that can in some cases approach the mass resolution of FTICR mass spectrometers are orbitraps 93 Orbitraps measure frequencies of ions orbiting a central electrode by non destructive image currents; the technique is hence very similar to FTICR MS in terms of data acquisition and processing, even if the orbitrap is not suitable for actual trapping of the ions for extended periods of time. Prior to the data acquisition, mass spectra are recorded as a finite number of data points in the case of FTICR MS, which determines the number of data points used in

PAGE 53

53 time domain spectrum. The time for data acquisition depends on the sampling rate (or sampling frequency), and the number of data points: ( 2 20 ) where T acq is the acquisition time in seconds, N is the datase t size, and S is the sampling frequency (determined by the low m/z cutoff) in Hz. According to the Nyquist theorem, the sampling frequency must equal to at least twice the highest frequency (i.e., the low m/z limit) of interest. 90 Thus, the maximum resolution for a given dataset size is determined by: ( 2 21 ) in which R is the mass resolution, is the cycl otron frequency, and T acq is the duration of acquisition. From Equation 2 21, it can be found that the resolution of FTICR MS is proportional to the magnetic field strength, the dataset size, but is inversely proportional to m/z and the sampling rate. For a given FTICR instrument, to achieve better resolution, one can increase the dataset size or use a higher value for the lowest m/z cutoff. The effects of dataset size and lowest m/z cutoff on the resolution are illustrated in Figures 2 8 and 2 9, respecti vely. It can be seen from Figure 2 8 that the mass spectrum with larger dataset sizes (and hence longer acquisition times) are necessary for higher resolution. Figure 2 9 demonstrates that resolution increases when the value of the lowest m/z cutoff is inc reased. The mass resolution also increases inversely with m/z in the mass spectrum. One trade off with longer transients is that the signal to noise ratio is decreased. In Figure 2 8, the longer transients 128k and 512k show a clear decay of the signal,

PAGE 54

54 w hereas this is not apparent yet in the 32k transient. This decay is due to a de phasing of ions inside of the ion packet as a result of Coulombic repulsion and collisions with background molecules. The latter effect demonstrates the importance of an ultra high vacuum in FTICR MS. Space Charge Effects Ions are stored in the analyzer cell for detection in the FTICR mass spectrometer, generating space charge. The ion space charge in the cell will affect the observed frequencies, and hence result in inaccurate measurements of masses. The frequency shifts were shown to correlate with the changes in space charge. 94 Jeffries and co workers developed the theory of space charge induced frequency shifts in Penning cells with different geometries, and mass calibration can be approximated by the followin g equations. 95 They derived the flowing expression for the frequencies of the natural modes of single ion motion in a cubic cell: (2 22 ) where + is the angular frequency, B is the magnetic field, V T is the trapping voltage, is the charge density, m is the mass of the ion, and c is the cyclotron frequency defined in Equation 2 7. An approximated mass calibration can then be derived at the point m=0, V T =0, and =0, and Equation 2 22 then becomes (2 23 ) Thus, the frequency differency between two ions are (2 24 )

PAGE 55

55 When B is known, by measuring the value of eff eff the m/z of an unknown ion can be determined relative to a reference ion. Infrared Multiple Photon Dissociation Spectroscopy In Chap ter 1, it has been mentioned that several techniques are used for fragmenting ions in mass spectrometers, and collision induced dissociation ( CID ) is the most widely employed method. In ion trap mass spectrometry, such as FTICR MS, i nfrared m ultiple p hoton d issociation ( IRMPD) has bee n used as an alternative to CID, with the advantage that it does not require the introduction of a collision gas into the ion trap Beauchamp and co workers first demonstrated that trapped ions in an ICR cell could be dissociate d with CO 2 lasers in the late 1970s, 35,96 98 The irradiation by the IR light results in fragmentation by stepwise vibrational excitation of the molecules. Infrared photons are less energetic, therefore absorption of multiple IR photons are needed to induce dissociation in a molecule. The mechanism of IRMPD is represented in Figure 2 10. The absorption of photons by trapped ions in IRMPD is a sequential process. An infrared photon with a frequency that is in r esonanc e with the frequency of a vibrational mode (vi), is absorbed by a n ion. This energy is rapidly dissipated to all other vibrational modes by intramolecular vibrational redistribution (IVR) (
PAGE 56

56 observed as a change in mass The IRMPD yield reflects the efficiency of IR absorption at a particular wavelength. By employing a tunable laser and measuring the IRMPD yield as a function of wavelength, the IRMPD spectrum of an ion can be recorded. 103,104 Direct infrared spectroscopy measurements on trapped ions in mass spectrometers are not possible, due to the ultra low ion densities. IRMPD spectroscopy hence offers an indirect approach to measure infrared spectra of gaseous ions. Such an appro causes a change in the species that can be observed, which in this case is a change in the m/z of the ion. Since multiple photons are required for dissociation, moderately powe rful lasers are necessary, and several types of lasers that have been successfully implemented in IRMPD spectroscopy studies. Historically, line tunable CO 2 lasers were first employed. Such lasers cover the region from 9.2 to 10.8 While there are several reports on IRMPD spectroscopy with CO 2 lasers, 105 107 their main limitation is the limited wavelength range. Free electron lasers (FELs) are another type of laser used for IRMPD spectrosc opy studies. Several facilities have coupled FELs with FTICR MS for IRMPD research, including the Free Electron Laser for Infrared experiments (FELIX) at the FOM Institute for Plasma Physics Rijnhuizen in The Netherlands, the Centre Laser Infrarough Orsay (CLIO) in Orsay, France, and the FEL at the Science University of Tokyo (SUT) in Tokyo, Japan. FELs provide high power over a wide wavelength range (5 250 for FELIX), and are hence ideally suited for IRMPD experiments. Nonetheless, their cost is substantial and access to these lasers is limited. The IRMPD

PAGE 57

57 experiments in this dissertation were done at the FELIX facility. Figure 2 11 shows the schematic rep resentation of FEIXL coupled to laboratory constructed FTICR mass spectrometer. There are two beamlines at FELIX FEL 1 and FEL 2. The free electrons are accelerated to either 15 to 25 or 25 to 45 MeV by one or two r adio frequency linear accelerators. Usi ng the first a ccelerator allows FEL 1 to access wavelengths from 25 to two accelerators are used in conjunction with each other, FEL 2 can access wavelengths from 5 The operation of a free electron laser rests on the principle that when a rela tivistic electron enters a magnetic structure, called an undulator it can spontaneously emit radiation. This radiation can be tuned by adjusting the kinetic energy, since the laser wavelength in the labframe is substantially Doppler shifted from the mm to the m range. For a particular kinetic energy setting, the wavelength can be tuned by a factor of 2 3 by adjusting the magnetic field in the undulator via positioning of samariumcobalt permanent magnets FELIX is a pulsed laser composed of micro and macro pulses that reflect the injection of electron bunches into the laser cavity The micropulses are spaced 1 ns apart and have a duration of 1ps to multiple ps Macropulses generally have a duration of 5 7 repetition rate of 5 Hz or 10 Hz More recent ly, our group has implemented a benchtop optical parametric oscillator (OPO) laser to carry out IRMPD experiments on mass selected ions in an ICR cell. This laser was shown to be able to photodissociate metal bound carbohydrate 108 and amino acid complexes. 109 In terms of IRMPD, the collision less environment of an ICR cell is certainly advantageous; as the absorbed energy is not lost due to collisions with

PAGE 58

58 background molecules. Moreover, the irradiation time can be extended to multiple seconds, thus increasing the IRMPD yield. All of the IRMPD results described in this thesis employ the free electron laser FELIX in c onjunction with the laboratory constructed FTICR instrument in Figure 2 11. Gas Phase Hydrogen/Deuterium Exchange Hydrogen/ d euterium exchange (HDX) is a powerful technique to investigate structures and conf o rmations of peptides and proteins. Hydrogen atoms bonded to oxygen, nitrogen, and sulfur can exchange with deuterium when molecules are exposed to an environment of deuterated molecules. Typically, H/D exchange reactions are performed on proteins in solution, where a protein is incubated in a D 2 O buffer, and the degree of subsequent deuterium incorporation determined by nuclear magnetic resonance (NMR) or mass spectrometry 110 113 The isotopic exchange of these labile hydrogens is an important tool to determine pro tein structures, and to study dynamic processes such as protein folding 114 117 Exchanges at the side chain are typically too fast to be kinetically resolved but hydrogens at the backbone amide positions exchange a t rates that can be measured. The lability of amide hydrogens also offer most insights into the secondary and tertiary structure of proteins. In gas phase HDX, reaction rate constants are much lower than those in solutions due to the very diluted condition s 118 122 so it has been proposed as an alternative method for structural analysis of amino acids. 121,123 125 Mass spectrometers are very suitable for performing gas phase HD X experiments because the deuterium incorporation can be monitored directly from the mass increase of the peptide In addition, with ion trapping techniques, such as ion cyclotron resonance (ICR) and

PAGE 59

59 quadrupole ion traps, ions can be trapped for extended t ime periods and thus allowing for observation of the exchange processes, so information about gas phase H/D exchange kinetics can be obtained 121,126 Systematic studies on model peptides had shown that the exchange rate in gas phase HDX depends on the proton affinity (PA) and/or gas phase basicity (GB) of the peptide molecule and the deuterating reagent. 127,128 The PA is defined as the enthalpy of reaction 2 25 whereas the G B stands for the free energy. 129 B + H + BH + (2 25 ) Several groups have contributed to the evaluation of the gas phase proton affinities for amino acids and peptides with either experimental or theore tical methods. 130 134 The gas phase proton affinities of typical deuterating reagents including D 2 O, CD 3 O D CD 3 C OO D and ND 3 are 166.5, 181.9, 190.2, and 204.0 kcal/mol, respectively. Deuterating reagents with hig her proton affinities display faster exchange. The evaluated proton affinities of amino acids are listed in Table 2 1. Proton affinities of peptides depend on the proton affinity of every single amino acid residue, as well as its chain length. For example, the proton affinities of Gly 2 Gly 3 Gly 4 and Gly 5 are 210.0, 213.0 218.1, and 218.4 kcal/mol, respectively. 129,133 There are several aspects that will affect the HDX: the difference in PAs between the analyte and the deuterating reagent, the size of the analyte molecule, and the number of deuterons on the deuterating reagent. When the PA is the main factor, if the deuterating reagent has a proton affinity be promoted. Of all the common deuterating chemicals listed before, ND 3 gives the fastest exchange rates when exchanging with peptides or amino acids.

PAGE 60

60 T he understanding to the mechanism of gas phase HDX remains inadequate In fact, Beauchamp and co worker s proposed a number of mechanisms that are active, depending on the deuterating molecule For low basicity reagents, such as CD 3 OD or D 2 forms two hydrogen bonds: one with th e protonated site on the N terminus, and one the N terminal proton to the deuterating molecule, and movement of a deuteron to a carbonyl oxygen. In effect, the proton of the p eptide ends up on the deuterating molecule, while a deuteron ends up on the peptide. For exchanges to occur, it was proposed that the energy gained by forming hydrogen bonds must be larger than the difference in the proton affinities of the exchange sites and the energy lost by opening the internally solvated structure. 128 More basic deuterating agents, such as ND 3 are 3 is capable of solvating the proton as an intermediate. In this thesis, deuterated methanol (CH 3 OD) is employed as mechanism is active.

PAGE 61

61 Figure 2 1. Ion cyclotron motion. A spatially uniform magnetic field has a direction that is perpendicu lar to and going into the plane of the paper and an ion is moving in the plane of the paper. The moving path of the ion is a circle resulting from the magnetic Lorentz force. Positive and negative ions travel in opposite directions. Figure 2 2. Ion p ath resulting from the combination of cyclotron, tapping, and magnetron motions. Figure reproduced from Ref. 8 0 with permission.

PAGE 62

62 Figure 2 3. Schematic diagram of Bruker APEXII 4.7T FTICR instrument. Figure 2 4. FTICR MS performance as a functio n of magnetic field strength B 0 Figure reproduced from Ref. 8 0 with permission.

PAGE 63

63 Figure 2 5. Schematic presentation of a cubic (left) and a cylindrical (right) FTICR MS analyzer cell. Both types of cell have six plates, and each pair functions for ex citation, trapping, and detection. Figure 2 6. Schematic showing cross section of cylindrical ICR cell. Ions are excited to a larger radius (shown in blue) by applying an rf potential to the excitation plates. The motion of the ion packet is then detect ed on the detection plates in the form of an image current.

PAGE 64

64 Figure 2 7. Schematic showing how time domain data is converted to a frequency domain spectrum, followed by conversion to a mass sp ectrum. Figure taken from Ref 91 with permission of The Royal Society of Chemistry http://dx.doi.org/10.1039/ b403880k

PAGE 65

65 Figure 2 8. Demonstration of the effect of dataset size on resolution. The three mass spectra correspond to different acquisition dataset sizes for the same sampl e. Figure reproduced from Ref 91 with permission of The Royal Society of Chemistry http://dx.doi.org/10.1039/ b403880k Figure 2 9. The dependence of resolution on both measured m/z and the lowest m/z cutoff. Figure reproduced from Ref 9 1 with permission of The Royal Society of Chemist ry http://dx.doi.org/10.1039/ b403880k

PAGE 66

66 Figure 2 10. Schematic presentation of IRMPD mechanism. Figure 2 11. Schematic representation of the laboratory constructed FTICR instrument equipped with an ESI source coupled to FELIX. Figure taken from Ref 1 04 with permission

PAGE 67

67 Table 2 1. Proton affinities of common amino acids. Numbers are adapted from Ref. 129 Amino Acid Proton Affinity (kcal/mol) Amino Acid Proton Affinity (kcal/mol) Gly 210.5 Asn 220.6 Cys 214.0 Tyr 220.9 Ala 214.2 Met 221.1 Ser 215 .2 Gln 222.0 Asp 216.4 Pro 222.1 Val 216.5 Glu 223.4 Leu 217.4 Trp 223.9 Ile 218.6 Lys 235.6 Thr 219.5 His 231.5 Phe 219.9 Arg 244.8

PAGE 68

68 Figure 2 12. Gas phase H/D exchange mechanisms proposed by Beauchamp and co workers. Figure is adapted from Ref 1 2 7 with permission.

PAGE 69

69 CHAPTER 3 INVESTIGATION OF THE INFLUENCE OF PEPTIDE B FRAGMENT IONS ON T HE FORMATION OF MACROCY CLE STRUCTURES THROU GH HEAD TO TAIL CYCLIZATION IN THE C ASE OF OLIGOGLYCINES Background It has been discussed in Chapter 1 that b ions gene rated by collision induced dissociation will undergo head to tail cyclization and form macrocyclic b ion isomers, which will then lead to sequence scrambling due the reopening of the ring structures at different sites. So far, to fully investigate the driv ing forces of the formation of macrocycle structures, systematic studies are required. In this chapter, the size effect of b ions will be studied. As mentioned before, infrared photodissociation spectroscopy can provide direct evidence for structures of b ions, and one example is that the IRMPD spectrum of b 4 fragment of the pentapeptide Leu enkephalin (Tyr Gly Gly Phe Leu) shows prominent bands for both oxazolone and macrocycle structures. 39 However, IRMPD spe ctroscopy cannot give information about relative abundances because the intensity of the band does not directly relate to the concentration of the structures. Gas phase hydrogen/deuterium exchange (HDX) has been employed to study the ion structures in the gas phase. workers employed HDX to investigate a series of b fragments made from oligoglycines, where they considered several water loss reaction channels. 135 Wysocki and Somogyi have used HDX to determine that b ions often exhibit b imodal distributions, which can be rationalized by the presence of Reproduced in part with permission from Chen, X.; Yu, L.; Steill, J.D.; Oomens, J.; Polfer, N.C. Effect of Peptide Fragment Size on the Propensity of Cyclization in Collision Induced Dissociation: Oligoglycine b 2 b 8 J. Am. Chem. Soc. 2009, 131, 18272 18282. Copyright 2009 American Chemical Society.

PAGE 70

70 two different chemical structures. 71,136 Paizs and Somogyi have used H/D exchange and DFT calcula tions for tryp t ic digest peptide b 2 fragments to show that exclusively oxazolone structures are formed. 70 Since the inadequate understanding of the HDX mechanism, the interpretation of HDX data is complicated, and hence making direct structural assignment difficult. S table isotope labeling techniques have been used by many gr oups for the mechanistic elucidation of gas phase reactions and fragmentations over the years. 137 142 Influence of size on scrambling has been studied by Van Stipdonk and co workers using tan dem mass spectrometry on a serie s of peptides, ranging from tetrapeptides to decapeptides. Minimal scrambling was observed in b 3 ion, while significant nondirect ions were found in all other ions. 142 However, since the peptides they used had different residues, there might be influence from the size chain of some amino acid. In addition, with j ust MS n method, they were not able to characterize the structures. Finally, the quantification of the isomeric structures was not obtained. In this project we combine isotope labeling, infrared spectroscopy, gas phase HDX, and computational approaches to structurally analyze a series of b product ions generated by CID of oligoglycine, from b 2 to b 8 Glycine is the simplest amino acid and no side chain nucleophiles are present, and hence the competition between oxazolone and macrocycle formation can be s tudied as a function of size of the fragment generated. Experimental Sample P reparation The triglycine (Sigma Aldrich, St. Louis, MO), pentaglycine (Sigma Aldrich, St. Louis, MO) cyclo (Gly Gly) (Bachem, Torrance, CA) and deuterating reagent CH 3 OD

PAGE 71

71 (Sigma Aldrich, St. Louis, MO) were employed without further purification. Octaglycine was prepared by solid phase synthesis and was isotopically labeled by incorporating a 13 C Gly as the N terminal residue (ICBR, University of Florida), and the peptide was puri fied by high performance liquid chromatography (HPLC). All peptides were used at 20 solution 49:49:2 (vol:vol:vol) water/methanol/acetic acid solutions The acetic acid was added to aid the protonation. Mass Spectrometry and Hydrogen/Deuterium Exchange The hydrogen/deuterium exchange (HDX) experiments were carried out at the Universit y of Florida using a commercially available Fourier transform ion cyclotron resonance (FT ICR) mass spectrometer (4.7 T actively shielded APEX II, Bruker Daltonics, Billerica, MA) in Dr. David ions were activa sk adjusting the voltage drop between the metal plated glass capillary and the first transfer to the ICR cell. The exception to this was the b 2 fragment, which was generated in the ICR cell by sustained off resonance irradiation collision induced dissociation (SORI CID) from protonated triglycine by pulsing in Xe gas, as low m/z (<130) ions could not be transferred from the hexa pole to the ICR cell successfully. The mass to charge ratios of all the ions of interest and the peptides used are listed in Table 3 1. The most abundant isotope peak of the b n product ion of interest was mass isolated and subjected to gas phase hydrogen/ deuterium exchange (HDX) with CH 3 OD, which was leaked into the vacuum chamber at a constant pressure of 10 8 Torr through a Varian leak valve. Mass spectra with different exchange times were recorded, and the ab undances for the undeuterated (d 0 ), singly de uterated (d 1 ), doubly deuterated (d 2 ),

PAGE 72

72 etc. peaks in the resulting distributions were determined as Campbell showed in their HDX kinetic study 119,128 (Figure 3 1) The normalized abundance of d 0 is defined as ( 3 1 ) in which I refers to the relative abundance of each product, and i to the number of deuterium exchanges. For each fragment, the natural logarithm of [d 0 ] was plotted against the exchang e time, 65,68 and then linear fits were performed to determine the reaction kinetics and abundances of the two structures. Since the exchange reagent is conside red to be in great excess of the analyte, the H/D exchange reaction can be assumed as a pseudo first order reaction. The slope obtained represent s the exchange rate constant The larger the slope the faster the exchange is For fragments with two chemical structures, two diff erent slopes should be obtained Mass Spectrometry and Infrared Photodissociation Spectroscopy The infrared photodissociation experiments were conducted at the FOM institute FELIX 143 coupled to a laboratory built FT ICR mass spectrometer, described in detail in previous publicat ions. 144,145 The b 2 b 5 and b 8 CID products dissoci ation, in a similar scheme as described in section 3.2.2 using protonated triglycine, pentaglycine, and octaglycine, respectively. The notation b 2 G3 for instance relates to the b 2 ion generated from triglycine. The fragment ion was accumulated in a hexap ole, prior to transfer to the ICR cell. As a control experiment for b 2 proton attached cyclo (Gly Gly) was produced by ESI in a separate experim ent to measure its IR spectrum. Following mass isolation, the ion of interest was irradiated with the tunable ou

PAGE 73

73 train of micropulses at a GHz repetition rate. The pulse energy per macropulse is cropulses were employed to induce efficient photodissociation. Infrared multiple photon dissociation (IR MPD) 146 spectro scopy works on the premise that when the laser frequency is in resonance with a fundamental vibration in the molecule, many photons ( i.e., tens to hundreds) are absorbed, leading to photodissociation. This manifests itself in the mass spectrum by a depleti on of the precursor ion and appearance of photofragments. The photofragments of b 2 protonated cyclo(Gly Gly) and b 5 observed in the photodissocation are listed in Table 3 2. The IR photodissociation spectrum was obtained by plotting the IR MPD yield as a function of the wavelength using the relation shown below. Given the many CID product ions generated from octaglycine, the photodissociation products for b 5 G 8 and b 8 G 8 were n ot detectable, and hence the IR MPD depletion spectra are shown here. ( 3 2 ) In the photodissociation experiments presented here, the wavelength was 1 Note that in the latter experiments, mirrors with a dichroic coating were employed in the laser cavity, to specifical ly allow reflection (and hence lasing) at the third harmonic, but not at the fundamental wavelength. Results and Discussion Scrambling in Isotope Tagged Peptide In order to test for scrambling in oligoglycine peptides, the octaglycine peptide was isotopica lly labeled at the N terminal position with a 13 C glycine residue. The mass

PAGE 74

74 spectra of CID products b 4 b 7 of the protonated precursor ion ( m / z 476) are shown in Figure 3 2. For all these b ions, in addition to the b fragment ions that have this residue inc orporated 13 C labeled b n b ions with masses that relates to the loss of the 13 C Gly residue, 1 2 C only b n are also observed in the mass spectra The possible explanation for the case of b 7 are illustrated in Figure 3 3 The loss of N terminal 13 C Gly fro m b 7 can be rationalized by a rearrangement process, where a macrocylic b 8 is first formed and then the reopening leading to an oxazolone with 13 C Gly on the C terminal s ide of the molecule. Upon amide bond cleavage between the seventh and eighth residues, 13 C Gly is then lost as a neutral fragment, yielding 1 2 C only b 7 The relative abundance of these 12 C only b ions increases as one goes down the series from b 7 to b 5 This is not surprising, as the loss of 13 C Gly becomes statistically more likely for sm aller fragments. All of these results are consistent with the hypothesis that cyclization in these peptides occurs, followed by sequence scrambling and the appearance of nondirect b ion Infrared Spectroscopy Results b 2 ion The IRMPD spectrum of b 2 ion ( m / z 115) made from protonated triglycine with CID was recorded. To identify the structure of the ion, this spectrum is compared to theoretical spectra generated through a computational study. All the theoretical spectra in this chapter were calculated by Lon g Yu from the Polfer group. For this b 2 G 3 fragment, a number of chemical structures have to be considered, including an oxazolone structure protonated on the N terminus ( oxazolone N prot ), an oxazolone structure protonated on the oxazolone ring N ( oxazolo ne ox prot ), and a cyclic diketopiperazine structure protonated on a carbonyl O ( diketopiperazine O prot ). Figure 3 4 presents a comparison of the experimental spectrum recorded in the mid IR range

PAGE 75

75 2000 cm 1 ) with the calculated spectra of the lowest energy conformer for each chemical structure, along with their st ructures and relative energies. Based on the pr ominent band centered at 1960 cm 1 oxazolone ox prot (Figure 3 3 B) can be identified unambiguously. Conversely, the oxazolone N prot structur e (Figure 3 3 C) gives a poorer match, suggesting that this structure either is not present at all or is at a much reduced abundance relative to oxazolone ox prot In fact, the predicted energy gap (> 23 kJ mol 1 ) between both of these structures is conside rable, supporting the claim that merely oxazolone ox prot is populated at room temperature. For the diketopiperazine O prot structure, there is no match between the experimental spectrum and the calculated spectrum despite the fact this structure is margi nally lower in energy than oxazolone ox prot Similar results have also been observed by others for related b 2 fragments Ala Ala (from Ala Ala Ala) by Oomens and Van Stipdonk and co workers, 51 Ala Gly (from Ala Gly Gly) by Wysocki and co workers, 64 and results by Paizs and Maitre and co workers. 53 The only exception so far to this rule has been the very recent study of b 2 of His Ala by Wysocki and co workers, where the evidence from IR spectroscopy and HDX suggested a mixture of oxazolone and diketopiperazine structures. 147 To further confirm that b 2 from Gly Gly Gly does not adopt a diketopiperazine structure, the IRMPD spectrum of p rotonated cyclo(Gly Gly), cyclo(Gly Gly)H + was recorded with FELIX. As shown in Figure 3 5 theory predicts the experimental spectrum qualita tively well, even if the scaling factor (0.98) is not optimal for all of the vibrations considered. Subtle differences in the interaction of the proton with the side chain CH (Figure 3 5A 3 5 B), can be

PAGE 76

76 distinguished. T he overlaid spectra of p rotonated cyclo(Gly Gly) and b 2 G 3 are shown in Figure 3 6. It is clear that they adopt different structures. This result indicates that no diketopiperazine structure is present in b 2 G 3 b 5 ion As for the study of b 2 t he exper imen tal mid 1 ) spectra for b 5 ( m / z 286) generated from octaglycine, b 5 G 8 was recorded and then compared to the theoretical spectra in order to identify the structures. The candidate structures considered in this case are shown in Figure 3 7. The experimental spectra are compared with the two lowest energy conformers for the various chemical structures as shown in Figure 3 8. The band at 1430cm 1 which is the chemical diagnostic band for macrocycle structures, suggests that there is a mac rocycle structure present in the b 5 ion population. This is also confirmed by the computational results (Figure 3 8), which predict the CO H + bending mode to o ccur in this region. In the macrocycle structure, the proton is partially shared between two carb onyl sites This results in a flat anharmonic potential, which is consistent with the 1 band (not shown) In fact, shared proton modes are often observed as broad features in IR spectra. For instance, in IR measurements by Johnson and co workers on proton bound dimers, the shared proton stretching band was found to be reasonably broad, even for these cold argon tagged complexes formed in a supersonic expansion. 148,149 For higher temperature proton bound complexes, the IR MPD spectral features typically become very broad. 150 153 The broadening of OH stretches as a consequence of strong H bonding has even been observed in several neutral amino acid and peptide systems. 154,155

PAGE 77

77 O xazolone N prot can be identified, based on the band at 1825 cm 1 assigned to C O stretch mode associated with the oxazolone ring moiety, which is in agreement with calculated spec tra for this chemical structure. The corresponding oxazolone C O stretch of oxazolone ox prot is not observed, and this may be du e to our limited scan to 1940 cm 1 Proton attachment on a backbone carbonyl oxygen would also be possible in principle for an oxazolone structure; however, such structures are not likely to be substantially populated at room temperature, given their lower proton affinity and, hence, relatively hi gh energetic penalty (>30 kJ mol 1 ) determined from previous studies. 39 b 8 ion IRMPD spectra of b 5 G 8 and b 8 G 8 are overlaid in Figure 3 9 Clear differences between the spectra can be seen, and t hese serve as useful guidance in the interpretation of the results. The spectrum for b 8 is clearly the simplest, lacking some of the spectral features present for b 5 For instanc e, the prominent band at 1775 cm 1 for b 5 is not observed for b 8 The weaker f eature at 1825 cm 1 is observed for b 5 but not for b 8 Recently, Maitre and co workers showed IR MPD evidence that b 5 from G 5 R exclusively gives rise to a macrocycle structure, based on the absence of oxazolone bands. 5 8 The positive identification of the macrocycle rests upon unambiguous assignment of modes associated to the proton attachment site on a backbone carbonyl, since this structure is chemically analogous to a peptide backbone and, therefore, lacks other diag nostic chemical moieties. In our comparison of b 5 and b 8 only b 8 is compatible with the exclusive presence of a macrocycle structure, as no oxazolone band is observed. Moreover, the presence of the band at 1430 cm 1 indicates the presence of the macrocycl e structure, as this mode has previously been suggested to be due to the

PAGE 78

78 CO H + (i.e., (C )O H + ) bending mode of a cyclic peptide structure. 39 Summary of IRMPD results From the IRMPD results for b 2 b 5 and b 8 i t can be seen that there are a number of differences between these b ions. b 2 exclusively adopt s oxazolone ox prot structures. Conversely, b 5 appears to give a mixture of oxazolone and macrocycle structures. The absence of oxazolone band in IR MPD spectrum of b 8 indicates the exclusive presence of macrocycle structures. HDX is employed as a method to quantify the relative amounts of oxazolone and macrocycle structures formed in CID, as a function of the b fragmen t size. Note that the fragment ions were made by nozzle skimmer dissociation and that these CID products were accumulated in a hexapole prior to transfer to the ICR cell, as opposed to generating the fragments in the ICR cell by SORI CID (with the exceptio n of b 2 G 3 as explained in the Experimental Section and Calculations). This approach presents a number of advantages over in cell CID. Relatively large and constant number densities of CID product ions can be generated in this manner. Moreover the pressu re of the deuterating agent can be held at a constant pressure, without the need to pulse in a collision gas for SORI CID. Finally, this approach is expected to yield thermalized fragment ions more readily, given the higher pressure environment in the stor age hexapole. As the ion temperature is likely to affect the HDX kinetics, it is important to control the ion temperature in the interest of reproducibility. The mass spectral distributions for different HDX times for the series b 2 b 8 are shown in Figure 3 10 In previous HDX studies, bimodal mass distribution in the mass spectra were observed for analytes with two or more isomeric structures, and this can

PAGE 79

79 be explained by different exchange rates of the isomers. 71,136 In all these b ions, no obvious bimodal distributions are seen here, however, a kinetic fitting of the HDX data may reveal the presence of more than one kinetic rate Kinetic F it ting of HDX D ata As discussed in the Experimental section, HDX can be considered as a pseudo first order reaction due to the fact that the amount of deuterating reagent is in great excess than the analyte inside the ICR cell. To determine pseudo first orde r HDX kinetics, the natural logarithm of the relative depletio n of the undeuterated peak, ln[d 0 n ], is plotted as a function of the HDX time. The complete series of HDX measurements from b 2 to b 8 is summarized in Figure 3 11. A single kinetic rate ( k = 0 .51 s 1 ) is observed in the case of b 2 whereas two distinct kinetic rates are required to fit the data for the larger b fragments ( b 4 b 7 ), with the notable exception of b 8 The reason for the appearance of two exchange rates may be the presence of two is omeric structures, which can exchange with CH 3 OD at different rates that can be distinguished with FTICR mass spectrometer. These two structures can then be called exchanging structures. By the assumption that exchanging popul ation is fully depleted at longer HDX times, the slower exchange rate can be exchanging structure. Hence the faster rate is the combination of exchanging structur e can then be determined. A least squares linear regression fit is employed to determine both the pseudo first order rate constant and the intercept. More conveniently, the depletion of d 0 can al so be represented as remaining d 0 (%) on a natural logarithm s exchanging linear exchanging structure at

PAGE 80

80 time zero. In the case of b 5 the intercept equates to a relative abundance of 23% for the be distinguished, as the difference in rate constant is almost an order of magnitude: k slow = 0.019 s 1 vs k fast = 0.17 s 1 k slow ). Table 3 4 summaries the kinetic fitting results for b 2 b 8 The exclusive presence of oxazolone ox prot structures for b 2 as confirmed by the IR MPD results, correlates well with the presence of one r ate of HDX exchange, which happens t 5 is shown to have a mixture of oxazolone and macrocycle structures, based on the IR MPD results, and this is confirmed by the and exchanging. Finally for b 8 the IR MPD results are consistent with the exclusive presence of a macrocycle structure, while the HDX results for b 8 exhibit merely one rate of re consistent with the view exchanging ions correspond to the macrocycle structure, whereas the exchanging structure corresponds to the oxazolone structure. exchanging structures from b 2 to b 7 are su mmarized in Figure 3 12 based on the kinetic fitting procedure, described above. The exchanging structure reaches 30% for b 7 which exchanging structure still accounts for the majority of the ion population. Note that the error bars in are determined from the standard deviation of the exchanging structure. The gradual increase of the relative amount of the macrocycle structure from b 4 to

PAGE 81

81 b 7 sho ws that macrocycle formation becomes more favorable for larger structures. A sudden oxazolone ) structure for b 8 is observed A recent study by Harrison showed that b 9 ions of Tyr(Ala) 9 (Ala) 4 Tyr(Ala) 5 and (Ala) 8 TyrAla are e xclusively macrocycle 59 As the size gets larger, the macrocycle structures become more favorable, and this probably due to the fact that more flexible backbone makes it easier for the N terminus t o make the attack. It would be interesting to see what would be observed for larger b n fragments. Preliminary experiments to generate larger b n product ions ( n > 8) from deca and dodecaglycine were not successful, instead resulting in dehydrated b ions, su ch as b 9 H 2 O and b 10 H 2 O (not shown). Further experiments are required to determine under what conditions such fragment ions can be produced. Chemical Basis for Differences in HDX Rates Due to the obvious structural difference between oxazolone and macrocy cle b fragment structures, it is expected that their proton affinities (PAs) should be different and hence their HDX rates would differ. Previous studies have shown that oxazolones have higher PAs than macrocycle structures, 39,42,156 due to the presence of more basic proton attachment sites (i.e., N terminus and oxazolone ring N, vs backbone carbonyl). In detailed studies by Beauchamp and co workers 127,128 and Lebrilla and co workers, 157,158 it has been shown that a smaller difference in proton affinity between the deuterating reagent an d the peptide results in faster HDX rates. The results for oxazolone and macrocycle b 5 seem to show the opposite effect, as the less basic macrocycle structure exchanges more slowly than the more basic oxazolone structure. These trends have also been obser ved by Wysocki and co workers in their combined IR MPD and HDX study on HA b 2 fragments. 147 Clearly, differences in PAs between reagent and analyte ions are not the only parameter affecting HDX rates. Another

PAGE 82

82 important parameter is the geometry of the analyte molecule, which enables (or inhibits) the gas phase HDX mechanism to occur. In the case of low basicity CH 3 OD, HDX is 122,128 This mechanism wor ks on the premise that the deuterating molecule simultaneously hydrogen bonds to the proton and a basic site on the peptide. The proton/deuteron transfer then takes place in a concerted mechanism, where the proton is transferred to the deuterating agent, w hile the deuteron is transferred to the basic hydrogen bonded site on the peptide. In light of oxazolone structure more readily assumes the correct geometry for such a concer ted mechanism, whereas a macrocycle structure is more prone to ring torsion strain, thus raising the transition state energy for this process. Summary In this chapter, the effect of chain length on peptide fragment structure formation is investigated using a range of gas phase techniques. Using stable isotopic labeling, sequence scrambling is confirmed in the oligoglycine system. The data from IRMPD and HDX indicate that smaller b fragments ( b 2 b 3 ) exclusively form oxazolone structures, whereas midsized fr agments ( b 4 b 7 ) display a mixture of oxazolone and macrocycle structures. Mixtures of both structures are in fact expected, given that a low energy isomerization pathway is available to interconvert oxazolone and cyclic b fragment structures. 56,159 This pathway is, however, apparently not available to smaller b fragments. In the case of b 8 the IR MPD and HDX results indicate an exclusive presence of macrocycle structures. The occurrence of larger macrocycle structures in CID i s particularly unsettling in

PAGE 83

83 structures are more likely to open up at a different site than where they were originally put together. In fact, recent studies by Harrison 59 and Van Stipdonk 142 seem to confirm that larger b fragments give rise to considerable sequence scrambling. Complementary information from a range of gas phase structural techniques was required to establish the qualitative and quantitative trends in the dissocia tion chemistry presented here. In particular, IR photodissociation spectroscopy allows identification of the chemical species that are generated, whereas H/D exchange enables quantification of the relative amounts that are made. Previous studies had shown that oxazolone structures could be readily identified based on the diagnostic oxazolone ring C O stretch. 39,51,52,160 Identification of the macrocycle can be established based o n the + bending mode (1430 cm 1 ), even if this mode appears in a potentially more congested regi on of the IR spectrum. While IR MPD spectroscopy gives valuable information on the chemical structures of reaction products, such as the chemical moieties formed 39,160,161 and the site of proton attachment, 162 164 the IR MPD yield cannot typically be related to relative abundances of structures in a mixture. It is demonstrated here that the relative abundances of oxazolone and macrocycle can be inferred from the kinetic fitting of HDX da ta. The complementary information from these IR spectroscopy and HDX results allows identification of the oxazolone and macrocycle for b 5 fragment structures The results from HDX kinetic analysis show that b 2 and b 8 display single exchange rate, character Combing the results from IRMPD that b 2 is exclusively oxazolone and that b 8 is exclusively macrocycle it is then clear tha t oxazolone structures display HDX rate

PAGE 84

84 constants that are an order of magnitude higher than those of macrocycle structures. In exchanging structure, or macrocycle can be trend of the formation of macrocycle structures is then established as a function of b fragment ion size.

PAGE 85

85 Table 3 1. Mass to charge ratios of all b ions of interested. b ions Precursor peptide Mass to charge ratio ( m/z ) b 2 (Gly) 3 115 b 3 (Gly) 8 173 b 4 (Gly) 5 229 b 5 (Gly) 5 (Gly) 8 287 b 6 (Gly) 8 344 b 7 (Gly) 8 401 b 8 (Gly) 8 458 Figure 3 1 HDX mass spectrum with HDX products labeled. Table 3 2. Photofragments of b 2 protonated cyclo(Gly Gly) and b 5 b ions Photofragments b 2 G 3 m/z 87 (a 2 ) Cyclo(Gly Gly)H + m/z 87 (a 2 ) and m / z 59 ( a 2 H 2 O) b 5 G 5 m/z 125 and m/z 154 ( b 3 H 2 O )

PAGE 86

86 Figure 3 2 Inserts from the nozzle skimmer CID spectru m of octaglycine labeled with a 13 C Gly as the N terminal glycine residue, showing the b isotope distri butions for (A) b 4 (B) b 5 (C) b 6 and (D) b 7 The 13 C label ed b n peaks denote b ions that incorporate the 13 C Gly label, whereas 12 C only b n peaks are entirely composed of 12 C Gly residues. Figure 3 3. The cartoon mechanism rationalizes the appearance of the 12 C only b 7 peak.

PAGE 87

87 Figure 3 4. IR MPD spectrum of the b 2 fragment generated from Gly Gly Gly, compared to computed spectra for (A) diketopiperazine structure protonated on a carbonyl O, (B) oxazolone structure protonated on the oxazolone ring N, and (C) oxazolone structure protonated on the N terminus.

PAGE 88

88 Figure 3 5. IR MPD spectrum of protonated cyclo(Gly Gly), compared to calculated spectra for (A) diketopiperazine with proton pointing to C H 2 group and (B) dik etopiperazine with proton pointing

PAGE 89

89 Figure 3 6. Overlaid IR MPD spectr a of protonated cyclo(Gly Gly) and b 2 from triglycine Figure 3 7. Possible structures for b 5 G 8 considered in theoretic al study. (A) macrocycle structure protonated on backbone carbonyl, (B) oxazolone structure protonated on N terminus, and (C) oxazolone structure protonated on oxazolone ring N. A B C

PAGE 90

90 Figure 3 8 IR MPD spectrum of b 5 generated from octaglycine compared to the two lowest energy conformers for the various chemical structures: (A) macrocycle structure protonated on backbone carbonyl, ( B ) oxazolone structure protonated on N terminus, ( C ) oxazolone structure protonated on oxazolone ring N. The chemical diagnost ic bands and the relative energies to the lowest conformer are indicated.

PAGE 91

91 Figure 3 9 Overlaid mid IRMPD spectra of b 5 G 8 b 5 G 5 and b 8 G 8 The chemically diagnostic modes are indicated.

PAGE 92

92 Figure 3 10 Representative H/D exchange (10 8 Torr CH 3 OD) m ass spectra for (A) b 2 generated from tri glycine, (B) b 3 generated from octa glycine, (C) b 4 generated from penta glycine, (D) b 5 generated from octa glycine, (E) b 6 generated from octa glycine, (F) b 7 generated from octa glycine, and (G) b 8 generated fro m octa glycine.

PAGE 93

93 Figure 3 11 Kinetic fitting of the HDX results for glycine based b fragment ions. (A) b 2 generated from triglycine, (B) b 3 generated from octaglycine, (C) b 4 generated from pentaglycine, (D) b 5 generated from octaglycine, (E) b 6 gener ated from octaglycine, (F) b 7 generated from octaglycine, and (G) b 8 generated from exchanging structure as a function of b n fragment size.

PAGE 94

94 Table 3 3 Kinetic fitting results for the ln[d 0 / d n ] plots vs. H/D exchange time for the oligoglycine fragments b 2 b 8 b n b 2 b 3 b 4 b 5 F ast exchanging structure slope 0.513 0.122 0.350 0.185 Error in slope 0.018 0.014 0.042 0.032 intercept 0.177 9.53 E 4 0.0609 0.109 Error in intercept 0.070 0.0427 0.0784 0 .106 R 0.997 0.987 0.986 0.958 Standard deviation about regression 0.108 0.0499 0.0938 0.124 Slow exchanging structure slope 0.00795 0.0189 Error in slope 6E 4 0.0056 intercept 2.38 1.45 Error in intercept 0.016 0.09 R 0.99 4 0.958 Standard deviation about regression 0.0134 0.0400 b n b 6 b 7 b 8 F ast exchanging structure slope 0.266 0.212 Error in slope 0.042 0.014 intercept 0.0667 0.0506 Error in intercept 0.0785 0.0426 R 0.976 0.991 Standard deviati on about regression 0.0938 0.0589 Slow exchanging structure slope 0.00403 0.986 0.0142 Error in slope 3 E 4 0.002 5 E 4 intercept 1.14 1.12 0.0409 Error in intercept 0.01 0.05 0.0145 R 0.992 0.986 0.99 7 Standard deviation about regressi on 0.00524 0.0400 0.0266

PAGE 95

95 Figure 3 12 The relative abundance of slow exchanging structure as a function of b fragment size.

PAGE 96

96 CHAPTER 4 STRUCTURAL ANALYSIS FOR LEU ENKEPHALIN B2 B4 WITH INFRARED MULTIPLE PHOTON DISSOCIATION SPECTROSCOPY AND GAS PHASE H YDROGEN/DEUTERIUM EX CHANGE Background In the last chapter, the dependence of macrocycle structure formation on the chain length of b fragment ions has been studied, and the system used is oligoglycines where all the amino acid residues are the same. In th is chapter, another peptide, Leu enkephalin (Try Gly Gly Phe Leu), will be investigated. Leu enkephalin is a pentapeptide that has been investigated in numerous studies The first direct evidence for a mixture of oxazolone and macrocycle structures comes f rom infrared multiple photon dissociation (IRMPD) spectroscopy by Polfer et al. on b 4 generated from YGGFL. 39 Oxazolone and macrocycle structures are identified based on diagnostic vibrations, involving the oxa zolone ring C=O stretch and macrocycle CO H + bending. Rearrangement processes were also seen for an a type fragment in Leu enkephalin by Vachet and Glish. 165 Structure of Try Gly Gly Phe Leu is as shown in Figure 4 1. The combined approach of IRMPD spectroscopy and gas phase HDX is hence a suitable solution to this problem, as shown in the previous chapter. We apply this complementary methodology here to study the b 2 b 4 CID products of Leu enkephalin. Both techniques are found to be highly complementary, as they cancel out ea weaknesses. While IRMPD spectroscopy confirms the chemical structures that are formed, HDX allows their quantification. Reproduced in part with permission f rom Chen, X.; Steill, J.D.; Oomens, J.; Polfer, N.C. Oxazolone vs. Macrocycle Structures for Leu Enkephalin b 2 b 4 : Insights from Infrared Multiple Photon Dissociation Spectroscopy and Gas Phase Hydrogen/Deuterium Exchange, J. Am. Soc. Mass Spectrom. 2010, 21, 1313 1321 Copyright 2010 A merican Society for Mass Spectrometry

PAGE 97

97 Experimental Sample P reparation The pentapeptide Leu enkephalin (Tyr Gly Gly Phe Leu) (Sigma Aldrich, St. Louis, MO) and deu terating reagent CH 3 OD (Sigma Aldrich, St. Louis, MO) were employed without further purification. Leu enkephalin was used as 20 solution in 50 : 50 (vol:vol) water/methanol with 2% acetic acid to aid protonation. Mass Spectrometry and H ydrogen/ D euterium E xchange (HDX) The hydrogen/deuterium exchange (HDX) experiments were carried out at the University of Florida using a commerci ally available 4.7 Tesla actively shielded Bruker Bioapex II Fourier transform ion cyclotron resonance (FT ICR) mass spectrometer equipped with an Apollo API 100 source (Bruker Daltonics, Billerica, MA) at Dr. David as reported in Chapter 3 The singly protonated Leu enkephalin ( m/z 556) in the ESI source region to generate CID product ions The produced ions were then accumulated in the hexapole for 3s, pr ior to transfer to the ICR cell. A frequency sweep was performed in the FT ICR experimental sequence to eject other ions and to mass isolate each desired b product ion (b 2 b 4 ). The mass to charge ratios of b 2 b 3 and b 4 are 221, 278, and 423, respectivel y. The monoisotopic peak of each species was then subjected to gas phase hydrogen/deuterium exchange (HDX) with CH 3 OD in the ICR cell for different time periods. CH 3 OD was leaked into the vacuum chamber using a Varian leak valve to attain a constant pressu re of 1x10 8 Torr. Note that CH 3 OD was degassed using several freeze thaw cycles before introduction into the mass spectrometer.

PAGE 98

98 Mass spectra with different HDX times were recorded. The data are represented here by plotting the natural logarithm of the ra tio of d 0 divided by the sum of all ions, ln[d 0 n ], as a function of the HDX time, as described previously. The depletion of d 0 is also represented as a percentage, which is plotted on a natural logarithm scale. Because the exchange reagent is considered to be in great excess of the analyte, the H/D exchange reaction can be approximated as a first order reaction in the analyte concentration. Linear fits were performed to determine the reaction kinetics. Mass Spectrometry and Infrared Photodissociation Spectroscopy The infrared photodissociation exper iments were performed at the FOM institute built FT ICR mass spectrometer described in detail previously was employed for the mass spectrometry measurements. The fragment ion s (b 2 b 3 4 is taken from a previously published study 39 The fragment ion was accumulated in the hexapole, and then transferred to t he ICR cell. The ion of interest was mass selected and irradiated with the tunable output from the free electron laser. The IR photodissociation spectrum was recorded by monitoring the infrared multiple photon dissociation (IRMPD) yield as a function of t he wavelength (here 1 ). This yield is represented with the equation (3 2). The yield is further normalized linearly with FELIX laser power at each wavelength step. The main photodissociation product of b 3 ( m/z 278) was found to be m/z 221 (b 2 ) as well as a minor fragment at m/z 193 (a 2 ). For b 2 ( m/z 221), the m/z 193 (a 2 ) photofragment was most abundant.

PAGE 99

99 Computations Calculations were carried out at the High Performance Computing (HPC) Center at the University of Florida using the AMBER force field 166 and the Gaussian03 package. 167 Candidate structures for b 2 fragments were generated using an in house developed method, involving conformational searching by molecular mechanics and frequency calculations by density functional theory (DFT). The chemical structures ( oxazolone and diketopiperazine ) were initially built and optimized using semi empirical methods (AM1) in HyperChem (Hypercube Inc., Gainesville, FL). For oxazolones two proton attachment sites were considered, at the N term inus and oxazolone ring nitrogen. For the diketopiperazine structure, merely the backbone carbonyl oxygen was considered The chemical input structures for the molecular mechanics calculations required a parameterization procedure. The AM1 structures we re optimized using DFT (B3LYP/6 31 g*), followed by an electrostatic potential fitting with ab initio methods (HF/6 31 g*) to derive the atomic point charges. The geometry and ESP derived charges were imported into the AMBER suite of programs, where a rest rained electrostatic potential (RESP) fitting was performed. 168 Each chemical structure was parameterized separately in AMBER, foll owed by a conformational search using simulated annealing cycles. Two separate runs with starting temperatures at 300 and 500 K were carried out, resulting in 200 candidate structures per dynamics simulation. All output structures from AMBER were then opti mized again at the DFT level, initially using B3LYP/3 21. Further optimization was performed at the B3LYP/6 31G(d). Merely the 20 unique lowest energy conformers were then optimized at B3LYP/6

PAGE 100

100 31G+(d,p), followed by a single point MP2/6 31G+(d,p) calculati on. The MP2 electronic energy for each conformer was corrected for the zero point energy (ZPE) from B3LYP/6 31G+(d,p) of theory to yield the final ZPE corrected energies. Note that MP2 energies were considered to account for the dispersion interaction invo lving the tyrosine side chain. All energies are presented here are relative to the lowest energy conformer. The frequency spectra of the lowest energy structures at the B3LYP/6 31G+(d,p) level were scaled by 0.965. 169 Stick spectra w ere convoluted using a 20 cm 1 full width at half maximum (fwhm) Gaussian profile to allow easier comparison with the recorded IR photodissociation spectra. Results and Discussion Infrared Spectroscopy Results IRMPD spectrum of t he b 2 fragment ( m/z 221), CID from protonated Leu enkephalin, was recorded in the mid IR range (1300 1975 cm 1 ). To structurally analyze the ion, the exper imental spectrum is compared to theoretical spectra of a number of candidate chemical structures. For theoretical study, the chemical structures considered includ e an oxazolone structure protonated on the N terminus ( oxazolone N prot), an oxazolone structure protonated on the oxazolone ri ng N ( oxazolone ox prot), and a diketopiperazine structure proto nated on a carbonyl O ( diketopiperazine O prot). Figure 4 2 displays a comparison of the experimental results with the calculated spectra of the lowest energy conformer for each chemical structure, along with their structures and relative energies. As expl ained in the experimental section, the electronic energies (MP2) are corrected for the zero point energy (ZPE) at

PAGE 101

101 B3LYP/6 31g+(d,p). Both stick spectra and convoluted Gaussian profiles are shown for easier comparison. The results from theoretical study are shown in Table 4 1. It is clear that t here is no match between the experimental spectrum and the diagnostic modes of protonated diketopiperazine structure, despite the fact that this structure is lowest in energy. The prominent C=O (1810 cm 1 ) and C N (17 15 cm 1 ) stretches are clearly not observed. This result suggests that no diketopiperazine structure is present in b 2 of Try Gly, and this is similar to the result from Chapter 3 that b 2 made from triglycine adopts exclusively oxazolone structure Similar results have been observed by others for b 2 fragments in different systems, Ala Ala by Oomens and Van Stipdonk and co workers 51 and Ala Gly by Wysocki and co workers 52 and a number of b 2 fragments from tryptic digest peptides by and co workers 53 This indicates that the kinetic barrier to a nucleophilic attack from the N terminus is too high, and hence the oxazolone formation pathway is favored for kinetic reasons, as predicted by Paizs. 170 The broad band at around 1900 cm 1 clearly shows one main peak and a prominent shoulder, which is consistent with the presence of oxazolone N prot and ox prot structures. The peak at lower frequency is assigned to the oxazolone C=O stretch of the N prot structure, whereas the higher frequency shoulder matches the corresponding band of the oxazolone ox prot structure. The predicted energy gap, calculated at the MP2 (Mller Plesset) level (~ 11 kJ mol 1 ) is somewhat large to account for the presence of both structures, and hence this may be due to slight inaccuracies in the calculation. Nonetheless, the computations favor the N prot oxazolone which also seems to be validated by the significantly more intense band for

PAGE 102

102 this structure in the IRMPD spectrum. This is in strong contrast to the b 2 fragment involving primary structures Gly Gly, Ala Gly and Ala Ala, where the ox prot structur e mainly accounts for the structures that are observed. The structures indicate that the tyrosine side chain in b 2 Try G ly plays an important role in stabilizing the N prot oxazolone due to interaction between the proton and the tyrosine cloud. In order to account for these interactions more accurately, the energies are computed at the MP2 level, which includes dispersion interactions. The overlaid experimental mid IRMPD (1300 1975 cm 1 ) spectra for b 2 ( m / z 221), b 3 ( m/z 278) and b 4 ( m/z 425) are shown in Figure 4 3 for comparison Note that only b 2 and b 3 were recorded in this study, whereas b 4 originates from the study by Polfer et al. 39 It is clear that there are some differences between these spectra The chemical interpretations of the spectral bands are indicated and are based on comparisons with DFT calculated spectra for b 2 and b 4 The bands in the region 1780 1940 cm 1 are chemicall y diagnostic modes for oxazolone appear in a higher frequency region than the amide C=O stretch. It can be seen that the position of the oxazolone C=O stretch is progressively red shifted in larger b fragments, due to an increase in hydrogen bonding interactions. b 2 and b 4 exhibit multiple oxazolone bands due to different protonation sites while b 3 merely appears to show one band, and this suggest s that only one site of proton attachment is favored in b 3 For identi fication of macrocycle structures the macrocycle + bending mode is typically observed at ~1445 cm 1 39 + bending of a proton shared b etween two carbonyls. In the IR MPD spectrum of b 2 where such a + ben ding mode is not possiblethe band at 1445 cm 1 corresponds to CH 2 bending

PAGE 103

103 on the ring. Nonetheless, the proton bending mode lends brightness to the band at this position. While this band is found to be relatively weak in b 3 its relative intensity is much enhanced in b 4 This is consistent with the hypothesis that b 4 contains a higher proportion of macrocycle structures than b 3 In summary, the IR spectra provide strong evidence that b 2 is exclusively composed of oxazolone structures, whereas a mixture of oxazolone and macrocycle structures are observed for b 4 For b 3 the identification of oxazolone is unambiguous, whereas conclusion for macrocycle is difficult to make due to the weak band at 1450 cm 1 and possible overlap between CO H + bending and oxazolo ne ring CH 2 bending. The complementary technique of HDX will be employed to shed more information on the chemical structures that are present. Hydrogen Deuterium Exchange (HDX) Experiments The mass spectral distributions for different HDX times for the b 2 b 3 and b 4 are shown in Figure 4 4. There are 4 exchangeable hydrogens in b 2 as illustrated in Figure 4 5. It can be seen that only 3 exchanges occurred in b 2 For the oxazolone N prot and ox prot structures, this suggests that the nitrogen bound hydr ogens are exchanged, whereas the side chain tyrosine OH is not. Similar phenomena are also found for b 3 and b 4 which show one less exchange than the number of exchangeable hydrogens. These results are consistent with the picture that the tyrosine OH is mu ch less labile to HDX with CH 3 128,171 A full deuteration of the peptide fr NH amide sites on the backbone. Within this model, the deuteron does not exchange for

PAGE 104

104 the tyrosine OH, and hence HDX is primarily sensitive to the chemical structure of the peptide fra gment (i.e, sites of proton attachment). Similarly to previous HDX studies 71,136 bimodal distributions are observed for b 3 and b 4 On the other hand, b 2 merely d isplays one distribution. These results are consistent with the findings from IRMPD above, where b 2 is exclusively composed of oxazolones whereas b 4 is made up of a mixture of oxazolone and macrocycle To calculate the relative abundances of these structu res, kinetic fitting for these HDX data is required. Kinetic Fitting of HDX Data To determine pseudo first order HDX kinetics, the natural logarithm of the relative depletion of the undeuterated peak, ln[d 0 n ], is plotted against the HDX time for b 2 b 3 and b 4 as shown in Figure 4 6 A single kinetic rate ( k = 0.43 s 1 ) is observed for b 2 while two distinct kinetic rates are required to fit the data for b 3 and b 4 A least squares linear regression fit is employed to determine both the pseudo first order rate constant and the intercept. As introduced in the last chapter, exchanging population can be determined accurately exchanging population is fully depleted at longer times T he relative abundances of the exchanging structures can be approximated by exchanging reaction. The rates and abundances from the HDX kinetic fitting results are summar ized in Table 4 2, and Table 4 3 summarize the kinetic fitting results. In the case of b 3 the exchanging structure at the beginning of the experiment. Note that the large error bars in this case ( 34%) are

PAGE 105

105 due to the low ion abundance for b 3 exchanging structure accounts for the remainder (i.e., 94%). Note that the higher rate at shorter HDX times in Figure 4 6 stinguished, as the difference in rate constant is more than an order of magnitude: k slow = 0.019 s 1 vs k fast = 0.40 s 1 k slow 4 is cture. The exclusive presence of oxazolone structures for b 2 correlates well with a single 3 and b 4 suggest the presence of two distinct chemical structures. Given the unambiguous identifi cation of the oxazolone structure for b 3 but more tenuous identification of the macrocycle this oxazolone macrocycle 2 and b 3 corresponds to the oxazolone from b 3 (6%) to b 4 (31%). This trend is in agreement with th e increase in intensity of the 1440 cm 1 band from b 3 to b 4 exchanging structure corresponds to the macrocycle exchanging structure is the oxazolone The same trends were observ ed in the study of size effect in Chapter 3. Moreover, in a recent study by Wysocki and co workers on H is A la b 2 oxazolone and diketopiperazine structures were considered. 147 Comparison with Oligoglycines Study T here are some diff erences between the present study and the oligoglycine b fragment study. For oligoglycine b 2 b 8

PAGE 106

106 exchanging structures could be established Such a categorization is less straight forward here, as the magnit ude for k fast and k slow drop by a factor of ~10 from b 3 to b 4 (see Table 4 2 ). The slower kinetics for b 4 are possibly due to the bulky phenylalanine side chain, which affects the H/D exchange as a result of steric hindering. Conversely, for oligoglycine b fragments, no such side chain effects are expected. Fortunately, the for b 3 and b 4 thus the separation of b 3 and b 4 fragment ions can be done through the kinetic analysis. Another dif exchanging structure (i.e., macrocycle ) for b 4 which appears to be considerably enhanced in Leu enkephalin compared to pentaglycine ( 31% vs. 9%). This difference must be related to the su btle primary structure differences between Tyr Gly Gly Phe and G ly G ly G ly G ly which shows that the primary structure affects the propensity for to which will be dis cussed in Chapter 5. An important realization i n the comparison between the IR MPD and HDX results for Leu enkephalin is that the appearance of two distinct kinetics rates in the HDX experiments is only related to the presence of two considerably different chemical structures (i.e., oxazolone and macrocycle ), not to the presence of different protonation sites of the same chemical structure. For b 2 IR MPD measurements confirm the presence of N prot and ox prot oxazolones ; this, however, only results in a sing le kinetic HDX rate. On the other hand, the presence of a mixture of oxazolone and macrocycle in b 3 and b 4 does result in the observation of two distinct HDX rates. This suggests that in the structural characterization of these b fragments, HDX is not sens itive to the site of

PAGE 107

107 proton attachment as such, but rather the chemical structure. This observation is basicity deuterating reagants (such as CH 3 OD 128 ). As previously discussed, the transition state fo r HDX is likely to be higher for a macrocycle compared to an oxazolone structure, due to the ring strain in the macrocycle structure to allow this mechanism to take place. Recently Solouki and co workers employed the more basic deuter at ing reagent ND 3 to exchanging structures 72 This observation is intriguing, since the mechanism of HDX for higher basicity HDX reagents, such as ND 3 mecha nism ). This shows that a number of approaches are possible in the HDX characterization of b fragments, even if it remains to be seen whether the interpretation is always unambiguous. Summary In this study, we have applied IRMPD spectroscopy and ga s phase H/D exchange to the characterization of b 2 b 4 from Leu enkephalin. IR MPD was used to qualitatively identify structures, whereas HDX was employed to quantify the structures. For b 2 by comparing the measured IR MPD spectrum (1300 1975 cm 1 ) to theore tical spectra, the diketopiperazine structure was excluded based on the fact that the prominent C=O (1810 cm 1 ) and C N (1715 cm 1 ) stretches were not observed. Conversely, the characteristic oxazolone C=O stretch modes at ~1900 cm 1 allowed identificatio n of N prot and ox prot oxazolones The exclusive presence of one chemical structure (i.e., oxazolone ) correlates well with the presence one rate constant in the HDX measurements of b 2 which happens MPD spectra of b 2 b 3 and b 4 shows evidence for both macrocycle and oxazolone structures in b 3 and b 4

PAGE 108

108 based on vibrations at 1440 cm 1 (CO H + bending of macrocycle ) and >1770 cm 1 ( oxazolone C=O). This is in agreement with the presence of two distinct HDX rates and bimodal distribu tions in the corresponding HDX mass spectra. Using a recently the kinetic fitting analysis, as shown in Chapter 3 for oligoglycine b CID products, the and exchanging structures were inferred from kinetic fitting o f the HDX data. Similarly to the more extensive study on oligoglycine b 2 b 8 Leu enkephalin b 2 b 4 show an increase in the relative abundance of the macrocycle structure with fragment size. In fact, the relative abundance for macrocycle b 4 for the sequence Try Gly Gly Phe is considerably larger than for G ly G ly G ly G ly (31% vs. 9%). This shows that apart from the chain length, the primary structure also plays a key role in the relative propensity in forming oxazolone vs. macrocycle structure. While more comp lementary IRMPD/HDX studies are required to validate the hypothesis that macrocycle structures exhibit slower HDX kinetics compared to oxazolones these results show that HDX is a promising technique in quantifying both structures. In fact, HDX is much sen sitive relative to IR MPD in detecting low abundance structures, such as the macrocycle for b 3

PAGE 109

109 Figure 4 1. Chemical structure of Leu enkephalin. Table 4 1. Energies for the lowest energy conformers of each chemical structure of b 2 The electronic energy for each conformer at the MP2/6 31G+(d,p) level was corrected for the zero point energy (ZPE) derived at the B3LYP/6 31G+(d,p) level. Struct u re Electronic energy / Hartrees ZPE corrected energy / Hartrees Diketopiperazine O pro t 759.757356 759.520926 Oxazolone N prot 759.7521904 759.5155784 Oxazolone ox prot 759.7469593 759.5114213

PAGE 110

110 Figure 4 2. Experimental IRMPD spectrum of b 2 compared with theoretical spectra by computational studies. (A) IRMPD spectrum of the b 2 f ragment generated from protonated Leu enkephalin, (B) diketopiperazine structure protonated on a carbonyl O, (C) oxazolone structure protonated on the on the N terminus, and (D) oxazolone structure protonated on the oxazolone ring N. The scaling factor is 0.965. Corresponding structures are presented on the right. The site of proton attachment (red arrow) and oxazolone rings (black arrow) are indicated.

PAGE 111

111 Figure 4 3. Overlaid mid IR MPD spectra of b 2 b 3 and b 4 The spectrum of b 4 is adapted from previous p ublication of Polfer at el. 39 The chemically diagnostic modes are indicated.

PAGE 112

112 Figure 4 4. H/D exchange (10 8 Torr CH 3 OD) mass spectra for ( A) b 2 (B) b 3 and (C) b 4 for different exchange times. Figure 4 5 Structure of b 2 of Try Gly of an oxazolone structure protonated at the N terminus. Exchangeable hydrogens are shown in red.

PAGE 113

113 Figure 4 6 Kinetic fitting of the HDX results for (A) b 2 (B) b 3 and (C ) b 4

PAGE 114

114 Table 4 2. Exchange rates and relative abundan ces of fast and slow exchanging structures for the Leu enkephalin fragments b 2 b 4 Fast exchanging structure Slow exchanging structure b 2 Rate 0.43 0.04 Abundance (%) 100 b 3 Rate 0.40 0.03 0.019 0.012 Abundance (%) 93.6 34.5 6.4 34.5 b 4 Rate 0.034 0.002 0.0046 0.0002 Abundance (%) 69.4 3.9 30.6 3.9 Table 4 3 Kinetic fitting results for the ln[d 0 / d n ] plots vs. H/D exchange time for the Leu enkephalin fragments b 2 b 4 b n b 2 b 3 b 4 Fast exchanging structure slope 0.42 8 0.424 0.0393 Error in slope 0.0370 0.0451 0.00205 intercept 0.128 0.156 0.0486 Error in intercept 0.10 1 0.12 4 0.0346 R 0.99 3 0.98 9 0.99 6 Standard deviation about regression 0.138 0.16 9 0.0472 Slow exchanging structure slope 0.0189 0.00459 Error in s lope 0.0122 2.42E 4 intercept 2.75 1.18 Error in intercept 0.264 0.037 8 R 0.839 0.99 6 Standard deviation about regression 0.173 0.02 30

PAGE 115

115 CHAPTER 5 DEPENDENCE OF HEAD TO TAIL CYCLIZATION ON PRIMARY STRUCTURE OF PEPTIDES IN COLLISIO N INDUCED DISSOCIATION Background The influence of the backbone length on the formation of macrocycle structure has been shown in Chapters 3 and 4, based on the results for series of b ions from oligoglycines and from YGGFL.There are some differences in the relativ e abundances of oxazolone / macrocycle structures in both of those datasets, which are probably due to the different amino acid compositions of the ions. This suggests that the primary structure of the peptide also affects the formation of macrocycle structu res. The effect of certain amino acids on scrambling has been investigated by several groups. Van Stipdonk and co workers studied the influence of some amino acids with multiple stage tandem mass spectrometry experiments. It was found that with the presenc e of arginine, only direct sequence ions were observed. This was interpreted in the sense that arginine could inhibit the formation of macrocyclic b ions regardless of its sequence position. 61 Another study by the same group showed that the reopening of macrocycle structures is influenced by the amino acid side chains. For a series of permuted isomers with glutamine, b 5 + ions showed nearly identical MS n spectra, which suggested that macrocycle strutures tend to reopen at the postion of glutamine. 62 Histidine residue was investigated by Paizs, Harrison and co workers by tandem mass spectrometry and theoretical studies on the singly protonated peptides containing His residue. The results indicate that cyclization/reop ening is less active for b n ions containing His residue than for those with only aliphatic residues. 172 It has been seen in the previous IRMPD results that unambiguous results can be made for oxazolon e structures based on the chemically diagnostic bands in 1750 1950

PAGE 116

116 cm 1 However, since the macrocylic structures have chemically diagnostic bands located in a very congested region, it is very difficult to identify macrocycle structures from IRMPD spectra In Chapter 3, for the study of b 2 from triglycine, cyclo(Gly Gly) was utilized as a reference system to exclude the diketopiperizine structure for b 2 A direct comparison of IRMPD spectra of b fragments to IRMPD spectra of synthetic cyclic peptides is a useful and straightforward approach to identify macrocycle structures. In this chapter, a systematic study of the effect of specific residues on head to tail cyclization is presented for the b 6 sequence motif QWFGLM, where all amino acid residues are diffe rent, thus in complete contrast to the oligoglycine studies. In addition, this peptide contains no basic amino acids (i.e., arginine, lysine, or histidine) and hence no proton will be located at the side chain. The experimental methods employed include IR MPD and gas phase HDX. Commercially available cyclo( QWFGLM ) is a reference peptide in the study of the b 6 motif QWFGLM. For the investigation of the proline effect on the head to tail cyclization, a synthetically made peptide, cyclo(QPFGLM) is utilized. Ex perimental Materials All resins were purchased from Advanced ChemTech ( Louisville, KY). Fmoc protected amino acids and 2 (1H Benzotriazole 1 yl) 1,1,3,3 tetramethyluronium hexafluorophosphate ( HBTU) were obtained from AnaSpec ( Fremont, CA). Trifluoroace tic acid (TFA 99% ), N,N Diisopropylethylamine (DIPEA), triisopropylsilane (TIS), N,N dimethylforma mide (DMF) dichloromethane (DCM), piperidine (>99.5%) water, methanol, and formic acid were available from Sigma Aldrich (St. Louis, MO). Cyclo(QWFGLM) wa s purchased from Bachem ( Torrance, CA).

PAGE 117

117 Preparation of Peptides Fmoc synthesis of linear peptides Most linear peptides used in this project were prepared using solid phase synthesis techniques using 9 fluorenylmethoxycarbonyl (Fmoc) amino acid loaded Wang resin s 3 and Fmoc protected amino acids. After the final deprotection, peptides were cleaved from the resins were achieved ut ilizing a cleavage cocktail with 95:2.5:2.5 (vol:vol:vol) TFA:H 2 O:TIS. The synthesized peptides that were dissolved in the cleavage solution were precipitated by addition of diethyl ether. The products were collected by centrifugation and washed several ti mes with diethyl ether. The solids were then dried in air at ambient temperature. N terminal acetylated peptides were obtained by mixing 50 molar excess of acetic anhydride and DIPEA with peptides in DMF for 30 min prior to the last deprotection step. Afte r the acetylation, the peptides underwent deprotection, cleavage, and purification as other linear peptides. Q(Trt)PFGLM was made with Fmoc protected resin s, preloaded with 2 chlorotrityl The cleavage was done with 5% TFA and 2.5% TIS in DCM. The trityl p rotecting group on the side chain of Gln can be retained after the cleavage. Following cleavag e, the peptide was dissolved in acetonitrile and freeze dried. S ynthesis of head to tail cyclic peptides The synthesis of cyclic peptides were done in Synthetic O rganic Chemistry group of the University of Amsterdam with the help from Jochem Rutters and Dr. Jan M a arsev e en Linear peptide Q(Trt)PFGLM, made with solid phase synthesis, were dissolved in THF at concentrations of less than 10 3 mol/L, with 4.4 equivalen t of DIPEA. 2.2 equivalent of HATU and 3H 1,2,3 Triazolo[4,5 b]pyridin 3 ol, were added into the

PAGE 118

118 solution. The reaction solution was kept at room temperature while stirring for 24h. The aliquot from the reaction solution was characterized by LC/MS to make sure that the reaction is complete. Purification of synthesized head to tail cyclic peptides The THF was evaporated off from the reaction solution, and the remaining solid was then dissolved in ethyl acetate. 1M KHSO 4 aqueous solution was added to dissolv e the unreacted coupling reagents. The organic phase was collected and solid Na 2 SO 4 was added to remove the remaining water. The organic phase was lyophilized to obtain solid products. The trityl protecting group at the glutamine side chain was then remove d by addition of 95%TFA. Crude peptides was purified by reversed phase HPLC (RF HPLC) on a C 18 column using a gradient of 0 80 % B (Buffer A: water/0.05%TFA; Buffer B: 90% ace tonitrile/10% water/0.045% TFA) over 30 min. Mass S pectrometry All sample solutio ns were prepared by dissolving peptides in 49:29:2 (vol:vol:vol) MeOH:H 2 O:formic acid at concentrations of 100 M. Infrared multiple photon dissociation (IRMPD) experiments were performed with the same set up as described in Chapter 3, using the free electr on laser FELIX at FOM MS. The mass to charge ratios ( m/z ) of the ions of interested generated by are listed in Table 5 1. In the photodissociation experiments presented here, the wavele cm 1 Gas phase hydrogen/deuterium exchange experiments were performed on cyclo(QWFGLM)H + QWFGLMPG b 6 and Ac QWFGLMPG b 6 using the 4.7 T FTICR instrument at the University of Florida and the sam e procedure described in Chapter 3.

PAGE 119

119 Again, the deuterating reagent used was CH 3 OD and the background pressure was kept constant at 10 8 Torr. Results and Discussion Commercial C yclo(QWFGLM) as Direct R eference IR MPD of c yclo(QWFGLM) and analogs The IR MPD s pectrum of protonated cyclo (QWFGLM ) is contrasted with those of b 6 from QWFGLMPG and Ac QWFGLMPG in Figure 5 1 The clear absence of oxazolone bands in the spectrum of b 6 from the linear peptide QWFGLMPG suggests that this ion does not adopt an oxazolone s tructure. To confirm that this ion is exclusively macrocyclic structure, its IRMPD spectrum is compared with that of commercial cyclo(QWFGLM). Both spectra are close to identical, clearly reproducing all of the main features. The only discrepancy is seen in the range from 1600 1650 cm 1 a range that includes NH 2 scissoring and C=O stretching modes. I t is conceivable that these minor differences are due to variations in subpopulations that are formed, although this is beyond the information that can be ob tained from this IRMPD spectrum Note that, t o promote isomerization from cyclic to oxazolone structures for the cyclic peptide, harsh source conditions were employed, where some of the protonated cyclo( QWFGLM ) precursor ions were fragmented by nozzle ski mmer CID. Nonetheless, the absence of bands at higher than 1700 cm 1 clearly indicates that there is no oxazolone structure present in the protonated cyclo(QWFGLM) in the gas phase even under these energetic conditions. For N terminally acetylated Ac QWF GLMPG, the N terminus is not a nucleophile, thus preclusing head to tail cyclization. In fact, the IRMPD spectrum for b 6 from Ac

PAGE 120

120 QWFGLMPG displays a prominent band at 1900 cm 1 which can be unambiguously assigned to the C=O stretch of a protonated oxazol one moiety This confirms that N terminal acetylation results in exclusive presence of an oxazolone structure for this b 6 ion. Compared to the other two spectra, the amide C=O stretch band is much broader, which is consistent with a more dynamic oxazolone structure. The band at 1440 cm 1 which had been assigned to the CO H + bending mode of macrocycle structures, and hence is used as the only diagnostic band for macrocycle structures in the mid IR range. Ac QWFGLMPG b 6 which is exclusively oxazolone also displays a band at this position, albeit at a lower intensity. This confirms that the 1440 cm 1 region is congested with other bands, and that identification of macrocycle structures is not unambiguous based on this feature.In principle, however, the highe r band intensities at 1440 cm 1 for QWFGLMPG b 6 and cyclo(QWFGLM)H + supports the case for the presence of the CO H + bending mode at 1440 cm 1 HDX of c yclo(QWFGLM) and analogs The three ions above were also subjected to gas phase hydrogen/deuterium exchan ge (HDX) with deuterated methanol (CH 3 OD) in the ICR cell of an FTICR mass spectrometer. As described in Chapter 3, the kinetic analysis of the data was done by plotting the relative abundance of the undeuterated peak, d 0 as a function of exchange time. T he kinetic plots of HDX for protonated cyclo(QWFGLM) and QWFGLMPG b 6 are shown in Figure 5 2 Two distinct rates of exchange can be resolved in both plots In addition, the kinetic rates are very close. Since both IRMPD spectra are nearly identical, it is not surprising that they display close values for the kinetic rates. A closer examination of the magnitudes of the rate constants for cyclo(QWFGLM)H + shows that

PAGE 121

121 exchanging structure for b 5 ( k 1 = 0.027 s 1 and k 2 = 0.009 s 1 vs k fast = 0.17 s 1 ). In fact, both of these rate constants are exchanging structure for b 5 (0.02 s 1 ). The absence of an oxazolone structure for cyclo( QWFGLM ) H + as confirmed by IR MPD spect exchanging structure in the HDX results. Similarly, both structures found in b 6 from QWFGLMPG are also exchanging structures, as k 1 = 0.037 s 1 and k 2 = 0.0069 s 1 The appearance of two ki netic rates indicates that HDX can separate the subpopulations of isomers which is compatible with the differences in the 1600 1650 cm 1 region of IRMPD spectra So far, it is not clear what isomers account for the subpopulations in macrocylic QWFGLM It is possible that different protonation sites have to be considered for this peptide, such as the tryptophan side chain for instance, or a number of backbone carbonyls. If the structures do not interconvert (i.e., no proton transfer), this may result in mul tiple exchange kinetics. Alternatively, cyclo( QWFGLM)H + is made up of different HDX for Ac QWFGLMPG b 6 shows an extremely slow exchange with CH 3 OD, as no exchange is observed even after 90 s(see Figure 5 3). This can be rationalized in the 128 Figure 5 4 illustrates the exchange progess with CH 3 OD. An oxazolone with a free N terminus exchanges with CH 3 OD through a concerted move ment of the oxazolone proton to the deuterating molecule, and movement of a deuteron to the N terminus. Both the N terminus and the oxazolone ring N have similar proton affinities, and hence there is no energetic penalty for substituting the proton from th e oxazolone ring for a deuteron on

PAGE 122

122 the N terminus, or vice versa. In Ac QWFGLMPG b 6 the N terminus is acetylated and hence the N terminus is no longer a basic site. Instead, the second most basic site in the molecule may be at a carbonyl oxygen, or the si de chain amine. Since these sites are much less basic than the N terminus, it is energetically unfavorable to substitute the proton from the oxazolone ring for a deuteron on a less basic site. The Effect of Proline on Head to T ail Cyclization Charaterizati on of synthetic peptides The LC/MS spectrum of purified linear Q(Trt)PFGLM is shown in Figure 5 5. As the HPLC spectrum shows, the product from the solid phase synthesis has high purity. This peptide was then subjected to head to tail cyclization in soluti on. The trityl protecting group on the side chain of glutamine is kept to prevent the cyclization from the side chain amine. Figure 5 6 shows the LC/MS spectrum of the aliquot from the cyclization reaction solution after 24h. The product with retention tim e of 6.15 min has the mass of 674, which is the mass of cyclo(QPFGLM). The product with retention of 6.95 min shows to intense peaks in the mass spectrum, 674 and 1347, and this indicates that cyclo(QPFGLM) 2 is also synthesized. Both of these two products were then purified and collected with HPLC. IRMPD results The IRMPD spectra of QWFGLMPG b 6 QPFGLMPG b 6 and QWPFGLMPG b 7 are compared in Figure 5 7, where the oxazolone region ( 1750 1930cm 1 ) is highlighted in pink. It is clear that the b 6 ion generated fr om QPFGLMPG displays a band at 1840cm 1 which is likely assigned to the C=O stretch mode of an oxazolone structure protonated at the N terminus. The spectrum of QWPFGLMPG b 7 ion, where the proline residue is inserted between the tryptophan and phenylalani ne, displays an even more intense

PAGE 123

123 oxazolone band. These results indicate that the presence of proline in these peptide play a role in favoring oxazolone structure formation. To confirm that the 1840 cm 1 band in fact corresponds to an oxazolone structure, the IRMPD spectrum of synthetically made cyclo(QPFGLM) is used as a direct reference. The overlaid IRMPD spectra of these two ions are shown in Figure 5 8. The spectrum of protonated cyclo(QPFGLM) shows no band in the oxazolone region, whereas the band fo r QPFGLMPG b 6 is apparent. This lends further credence to the claim that oxazolone structures are formed for QPFGLMPG b 6 and QWPFGLMPG b 7 As noted for QWFGLM systems, there are also differences in the range from 1600 1650 cm 1 where NH 2 scissoring and C= O stretching modes are located, although these differences are harder to explain from a structural point of view. The effect of the proline residue on the formation of macrocycle structure may be due to its chemical structure. Off all the natural amino ac ids, p roline is unique in that it contains a secondary amine involving a cyclic side chain. This feature makes the dihedral angle of proline particularly rigid (~60 o secondary structure of peptides at this position. This rigidity constrains the flexibility of the backbone, and is hence compatible with limiting nucleophilic attack from the N terminus. The Effect of Glutamine Side Chain on the Head to Tail Cyclization There are two amines in the structure of glutamine: one at the N terminus and the other at the side chain. In the gas phase, the head to tail cyclization is done through the nucleophilic attach from the N terminus. It is also possible that a competing nucleophilic attack can occur from the amine at the glutami ne side chain, forming different kind of macrocycle structure. To investigate whether the side chain of glutamine is involved in

PAGE 124

124 gas phase cyclization reactions during CID, a systematic study was performed using the QWFGLG motif. Four peptides were prepare d: QWFGLGPG, Ac QWFGLGPG, Q(N ethyl)WFGLGPG, and Ac Q(N ethyl)WFGLGPG. The structures the oxazolone structure of b 6 ions made from the four peptides are illustrated in Figure 5 9. For QWFGLGPG b 6 attack from both the N terminus and side chain are poss ible. For Ac QWFGLGPG b 6 only the attack from the glutamine side chain can occur, while for Q(N ethyl)WFGLGPG b 6 only the N terminus can attack. No cylization can happen for Ac Q(N ethyl)WFGLGPG b 6 since both amines are protected, and hence only oxaz olone structures can be generated. The IRMPD spectra of these four b 6 ions are shown in Figure 5 10, and the spectral range for oxazolone bands is again highlighted in pink. There is a very weak oxazolone band observed in the spectrum of b 6 from QWFGLGPG ~ 1800 cm 1 indicating that a small abundance of oxazolone structures are present. For Q(N ethyl)WFGLGPG b 6 no band is seen in this region suggesting that this ion is exclusively macrocycle Conversely, two bands are observed in the spectra of Ac QWFGLGP G b 6 and Ac Q(N ethyl)WFGLGPG b 6 at 1830 cm 1 and 1920 cm 1 assigned to oxazolone structures protonated at the oxazolone ring N and protonated at the N terminus respectively. It is apparent that only when the N terminus is blocked, do the oxazolone ban ds become prominent. These results indicate that in the gas phase, for peptides with glutamine at the N terminus, only the head to tail cyclization has to be considered, while cyclization from the glutamine side chain can be neglected.

PAGE 125

125 Summary In this cha pter, it has been shown that apart from the size effect, the primary structure has an influence on the formation of macrocycle structures. Synthetically made cyclic peptides can provide a direct comparison for IRMPD spectra and HDX, in order to identify ma crocycle structures. The IRMPD spectra of QWFGLMPG b 6 and cyclo(QWFGLM) are nearly identical, which indicates that QWFGLMPG b 6 exclusively adopts a macrocycle structure. This is confirmed by the gas phase HDX kinetic study, e xchanging rates that are close to each other. The influence of the proline residue on the head to tail cyclization is investigated. An oxazolone band is seen in the IRMPD spectra of both QPFGLMPG b 6 and QWPFGLMPG b 7 suggesting that the presence of proline reduce s the propensity for the formation of macrocycle structures. A systematic chemical protection study was performed on the QWFGLG motif, to investigate nucleophilic attacks from the glutamine side chain. Based on these results, it appears that such a side chain attack does not compete with the head to tail cyclization reaction

PAGE 126

126 Table 5 1. Mass to charge ratios ( m/z ) of the ions of interest in this project. Ions of interest Mass to charge ratio ( m/z ) Cyclo(QWFGLM)H + 763 Cyclo(QPFGLM)H + 674 QW FGLMPG b 6 763 QWFGLGPG b 6 689 QPFGLMPG b 6 674 QWPFGLMPG b 7 860 Ac QWFGLMPG b 6 805 Ac QWFGLGPG b 6 731 Q(ethyl)WFGLGPG b 6 717 Ac Q(ethyl)WFGLGPG b 6 759 Figure 5 1. Overlaid IR MPD spectra of protonated cyclo( QWFGLM ) and b 6 from QWFGLMPG and Ac Q WFGLMPG

PAGE 127

127 Figure 5 2 Natural logarithm of relative d 0 depletion, ln[ d 0 n ] as a function of time for (A) protonated cyclo(Gln Trp Phe Gly Leu Met), and (B) b 6 generated from protonated linear QWFGLMPG. Figure 5 3 Mass spectra of Ac QWFGLMPG b 6 ion after exchanging with CH 3 OD for (A) 0s and (B) 90s. No exchange is seen after 90s of exchange, indicating that such reaction is extremely slow.

PAGE 128

128 Figure 5 4 Schematic presentation of an oxazolone exchanges with CH 3 OD. Figure 5 5 LC/MS spectrum of linear Q(Trt)PFGLM after purification.

PAGE 129

129 Figure 5 6 LC/MS spectrum o f aliquot taken from cyclization reaction solution. The top figure is the HPLC spectrum of the aliquot. The middle figure is the MS spectrum of the synthetic cyclo(QPFGLM) with a retention time of 6.14 min. The bottom figure is the MS spectrum of the synth etic cyclo(QPFGLM) 2 with a retention time of 6.95 min.

PAGE 130

130 Figure 5 7 IRMPD spectra of (A) QWFGLMPG b 6 (A) QPFGLMPG b 6 and (A) QWPFGLMPG b 7 In the inserts, in the intensities are magnified by 5x in the range of 1750 1930 cm 1

PAGE 131

131 Figure 5 8 Overlaid IR MPD spectra of cyclo(QPFGLM) and QPFGLMPG b 6 The oxazolone C=O stretch band is pointed by the red arrow.

PAGE 132

132 Figure 5 9 The structures of the b 6 ions from four peptides of QWFGLG system. The possible attack from the N terminus is labeled in blue, and the one from the glutamine side chain is labeled in orange.

PAGE 133

133 Figure 5 10 IRMPD spectra of b 6 ions generated from (A) QWFGLGPG, (B) Ac QWFGLGPG, (C) Q(N ethyl)WFGLGPG, and (D) Ac Q(N ethyl)WFGLGPG The range for oxazolone bands is highlighted in pink.

PAGE 134

134 CHAPTER 6 CONCLUSIONS AND FUTU RE DIRECTIONS The results in this thesis have given insights into head to ions in collision induced dissociation (CID). This process rationalizes scrambling of the original amino acid sequence, and he nce it is important to understand the chemistry of b ions in the gas phase. It has been shown that combining infrared multiple photon dissociation (IRMPD) spectroscopy and gas phase hydrogen/deuterium exchange (HDX) is a powerful approach for structurally elucidating b ions. IRMPD spectra of b ions can be used to identify the structures based on the chemically diagnostic bands. In particular, oxazolone structures can be identified by the bands associated to the oxazolone C=O stretch modes in the range of 17 50 1950 cm 1 The diagnostic bands of macrocycle structures are locate d at around 1440 cm 1 which are assigned to CO H + bending mode. However, as this region is congested with other vibrational modes, unambiguous identification is often more challenging. A further limitation is that IRMPD spectroscopy cannot provide information about relative abundances in mixtures, since the intensity of the IRMPD band does not linearly relate to the abundance of one structure. On the other hand, the relative abundances o f oxazolone / macrocycle structures can be approximated with gas phase HDX. The exchange reaction in the ICR cell is assumed to be a first order reaction, and hence the kinetic plotting can be used to calculate the exchange rates and relative abundances of i somers. It has been found that the exchange rates of oxazolones are around 10 times faster than macrocycle structures. The first project on this dissertation is about the influence of the peptide backbone length on b ion structures. A series of b ions, b 2 b 8 made from oligoglycines were

PAGE 135

135 investigated by IRMPD spectroscopy and gas phase HDX. The glycine residue is the simplest amino acid, devoid of a nucleophilic side chain, and hence is an excellent candidate for the size effect study. It was found that the formation of macrocycle structures is favored for larger b fragments. Smaller b ions, b 2 and b 3 were identified as exclusively oxazolones whereas mid size b ions, b 4 b 7 were found to be a mixture of oxazolone and macrocycle structures. With the kinetic fitting of HDX data, the relative abundances of macrocycle structures were calculated to be 7%, 21%, 31% and 32%, respectively for b 4 b 7 Thus, oxazolone structures are still the majority of the population. The largest b ion, b 8 was found to exclusively adopt a macrocycle structure, as confirmed by the absence of an oxazolone band in the IRMPD spectrum and a single HDX exchange rate. CID product ions from Leu enkephalin (YGGFL) were studied with the same techniques as above. A similar dependency on the si ze of backbone was observed in this peptide. b 2 was found to adopt only oxazolone structures, as shown by the good match with the theoretical oxazolone spectra and a single HDX exchange rate. Both oxazolone and macrocycle structures were confirmed in b 3 an d b 4 ions, and the relative abundances were calculated to be 6% and 31%, respectively. The relative ratios of oxazolone / macrocycle for b 3 of GGG and YGG, and b 4 of GGGG and YGGF are different, which indicates that besides the size effect, the primary seque nce plays a role in the formation of macrocycle structures. Thus, a third project was conducted to study how the primary structure affects the formation of macrocycle structure for the sequence motif QWFGLM. In addition to IRMPD and HDX methods, synthetic cyclic peptides were used to provide a direct

PAGE 136

136 comparison, to verify macrocycle identification. By comparing IRMPD spectrum of QWFGLMPG b 6 ion with those of cyclo(QWFGLM) and Ac QWFGLMPG b 6 it was found that b 6 of QWFGLM was exclusively macrocycle The ef fect of proline on the head to tail cyclization was investigated by measuring IRMPD spectra of synthetic cyclo(QPFGLM), QPFGLMPG b 6 and QWPFGLMPG b 7 Results indicated that the presence of proline residue in the primary sequence could reduce the propensity of the formation of macrocycle structures, consistent with the limited flexibility of the backbone due to proline. Lastly, studies of b 6 ions of QWFGLG, Ac QWFGLG, Q(N ethyl)WFGLG, and Ac Q(N ethyl)WFGLG revealed that no cyclization from the amino gro up on the glutamine side chain takes place Several avenues for future work are proposed First, the size effect study has so far only been carried out for b ions up to b 8 .It is not clear yet whether larger b n fragments, where n>8, also exclusively adopt m acrocycle structures. While the propensity for head to tail cyclization increases with chain length for b 2 b 8 very large macrocycle are expected to be energetically disfavored due to entropic effects. In other words, the degrees of freedom of a macrocycl e structure is reduced vis vis a linear oxazolone structure. One might hence expect an upper limit for macrocycle structures, which remains to be confirmed. Much work remains to be done on the correlation between primary structure and propensity for mac rocycle formation. The presence of proline appears to reduce this tendency, possibly due to kinetic effects. Other residues (e.g. arginine, lysine) might offer competing nucleophilic attacks from their side chain groups. This was not confirmed for glutamin e, but may yet be confirmed for other amino acid residues.

PAGE 137

137 Lastly, although the scrambling phenomenon has been observed in a number of model peptides, there is no study showing the extent of scrambling in proteomics studies. Such an analysis is on going i n our laboratory, involving high resolution and high proteins) proteomics study.

PAGE 138

138 LIST OF REFERENCES (1) Garrett, R.; Grisham, C. M. Biochemistry ; 2nd ed.; Sa unders College Pub.: Fort Worth, 1999. (2) Merrifield, R. B. J. Am. Chem. Soc. 1963 85, 2149 2154. (3) Chan, W. C.; White, P. D. Fmoc solid phase peptide synthesis: a practical approach ; Oxford University Press: New York, 2000. (4) Kent, S. B. Annu Rev Bi ochem 1988 57 957 89. (5) Davies, J. S. J Pept Sci 2003 9 471 501. (6) Domon, B.; Aebersold, R. Science 2006 312 212 217. (7) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989 246 64 71. (8) Hillenkamp, F.; Karas, M.; Beavis, R. C.; Chait, B. T. Anal. Chem. 1991 63 A1193 A1202. (9) Karas, M.; Hillenkamp, F. Anal. Chem. 1988 60 2299 2301. (10) Steen, H.; Mann, M. Nature Rev. Mol. Cell Biol. 2004 5 699 711. (11) Niall, H. D. Methods Enzymol 1973 27 942 1010. (12) B erthod, A.; Crank, J. A.; Rundlett, K. L.; Armstrong, D. W. Rapid Comm Mass Spectrom 2009 23 3409 3422. (13) Standing, K. G. Curr. Opin. Struct. Biol. 2003 13 595 601. (14) Horn, D. M.; Zubarev, R. A.; McLafferty, F. W. Proc. Natl. Acad. Sci. USA 2000 97 10313 10317. (15) Zubarev, R. A.; Horn, D. M.; Fridriksson, E. K.; Kelleher, N. L.; Kruger, N. A.; Lewis, M. A.; Carpenter, B. K.; McLafferty, F. W. Anal. Chem. 2000 72 563 573. (16) Ge, Y.; Lawhorn, B. G.; ElNaggar, M.; Strauss, E.; Park, J. H.; Beg ley, T. P.; McLafferty, F. W. J. Am. Chem. Soc. 2002 124 672 678. (17) Ge, Y.; ElNaggar, M.; Sze, S. K.; Bin Oh, H.; Begley, T. P.; McLafferty, F. W.; Boshoff, H.; Barry, C. E. J. Am. Soc. Mass Spectrom. 2003 14 253 261. (18) Sze, S. K.; Ge, Y.; Oh, H. B.; McLafferty, F. W. Anal. Chem. 2003 75 1599 1603. (19) Skoog, D. A.; Holler, F. J.; Crouch, S. R. Principles of instrumental analysis ; 6th ed.; Thomson Brooks/Cole: Belmont, CA, 2007.

PAGE 139

139 (20) Hoffmann, E. d.; Stroobant, V. Mass spectrometry: principles an d applications ; 2nd ed.; Wiley: Chichester ; New York, 2001. (21) Rockwood, A. L.; Busman, M.; Udseth, H. R.; Smith, R. D. Rapid Comm Mass Spectrom 1991 5 582 585. (22) Cech, N. B.; Krone, J. R.; Enke, C. G. Anal. Chem. 2001 73 208 213. (23) Iribarne, J V.; Thomson, B. A. J. Chem. Phys. 1976 64 2287 2294. (24) de la Mora, J. F. Anal. Chim. Acta 2000 406 93 104. (25) Roepstorff, P.; Fohlmann, J. J. Biomed. Mass Spectrom. 1984 11 601. (26) Horn, D. M.; Ge, Y.; McLafferty, F. W. Anal. Chem. 2000 72 4778 4784. (27) Stensballe, A.; Jensen, O. N.; Olsen, J. V.; Haselmann, K. F.; Zubarev, R. A. Rapid Commun. Mass Spectrom. 2000 14 1793 1800. (28) Hakansson, K.; Cooper, H. J.; Emmett, M. R.; Costello, C. E.; Marshall, A. G.; Nilsson, C. L. Anal. Chem. 2001 73 4530 4536. (29) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem. Soc. 1998 120 3265 3266. (30) Schnier, P. D.; Price, W. D.; Strittmatter, E. F.; Williams, E. R. J. Am. Soc. Mass Spectrom. 1997 8 771 780. (31) Mabud, M. D. A.; D ekrey, M. J.; Cooks, R. G. Int. J. Mass Spectrom. 1985 67 285 294. (32) Coon, J. J.; Syka, J. E. P.; Schroeder, M. J.; Shabanovitz, J.; Hunt, D. F. Proc. Natl. Acad. Sci. USA 2004 101 9528 9533. (33) Wells, J. M.; McLuckey, S. A. Biol. Mass Spectrom. 2 005 402 148 185. (34) Sleno, L.; Volmer, D. A. J. Mass Spectrom. 2004 39 1091 1112. (35) Little, D. P.; Speir, J. P.; Senko, M. W.; Oconnor, P. B.; Mclafferty, F. W. Anal. Chem. 1994 66 2809 2815. (36) Biemann, K.; Scoble, H. A. Science 1987 237 992 998. (37) Chait, B. T.; Wang, R.; Beavis, R. C.; Kent, S. B. Science 1993 262 89 92. (38) Little, D. P.; Speir, J. P.; Senko, M. W.; O'Connor, P. B.; McLafferty, F. W. Anal. Chem. 1994 66 2809 2815.

PAGE 140

140 (39) Polfer, N. C.; Oomens, J.; Suhai, S.; Paizs, B. J. Am. Chem. Soc. 2007 129 5887 5897. (40) Dongre, A. R.; Jones, J. L.; Somogyi, A.; Wysocki, V. H. J. Am. Chem. Soc. 1996 118 8365 8374. (41) Burlet, O.; Yang, C. Y.; Gaskell, S. J. J. Am. Soc. Mass Spectrom. 1992 3 337 344 (42) Paizs, B.; Suhai, S. Mass Spectrom. Rev. 2005 24 508 548. (43) Boyd, R.; Somogyi, A. J. Am. Soc. Mass Spectrom. 2010 21 1275 1278. (44) Biemann, K. Biomed. Environ. Mass Spectrom. 1988 16 99 111. (45) Yalcin, T.; Khouw, C.; Csizmadia, I. G.; Peterson, M. R.; Harrison, A. G. J. Am. Soc. Mass Spectrom. 1995 6 1165 1174. (46) Harrison, A. G. Mass Spectrom. Rev. 2009 28 640 654. (47) Yalcin, T.; Csizmadia, I. G.; Peterson, M. B.; Harrison, A. G. J. Am. Soc. Mass Spectrom. 1996 7 233 242. (48) Nold, M. J.; Wesdemiotis C.; Yalcin, T.; Harrison, A. G. Int. J. Mass Spectrom. Ion Proc. 1997 164 137 153. (49) Paizs, B.; Lendvay, G.; Vekey, K.; Suhai, S. Rapid Commun. Mass Spectrom. 1999 13 525 533. (50) Rodriquez, C. F.; Shoeib, T.; Chu, I. K.; Siu, K. W. M.; Hopkinson A. C. J. Phys. Chem. A 2000 104 5335 5342. (51) Oomens, J.; Young, S.; Molesworth, S.; van Stipdonk, M. J. J. Am. Soc. Mass Spectrom. 2009 20 334 339. (52) Yoon, S.; Chamot Rooke, J.; Perkins, B.; Hilderbrand, A. E.; Poutsma, J.; Wysocki, V. H. J. Am Chem. Soc. 2008 130 17644 17645. (53) Bythell, B.; Erlekam, U.; Paizs, B.; Maitre, P. Chem. Phys. Chem. 2009 10 883 885. (54) Chen, X.; Yu, L.; Steill, J. D.; Oomens, J.; Polfer, N. J. Am. Chem. Soc. 2009 131 18272 18282. (55) Chen, X ; Steill, J. D.; Oomens, J.; Po lfer, N. C. J. Am. Soc. Mass Spectrom. 2010 21 1313 1321. (56) Bleiholder, C.; Osburn, S.; Williams, T.; Suhai, S.; van Stipdonk, M. J.; Harrison, A. G.; Paizs, B. J. Am. Chem. Soc. 2008 130 17774 17789.

PAGE 141

141 (57) Harrison, A. G.; Young, A B.; Bleiholder, B.; Suhai, S.; Paizs, B. J. Am. Chem. Soc. 2006 128 10364 10365. (58) Erlekam, U.; Bythell, B.; Scuderi, D.; Van Stipdonk, M. J.; Paizs, B.; Maitre, P. J. Am. Chem. Soc. 2009 131 11503 11508. (59) Harrison, A. J. Am. Soc. Mass Spectro m. 2009 20, 2248 2253 (60) Harrison, A. G. J. Am. Soc. Mass Spectrom. 2008 19 1776 1780. (61) Molesworth, S. P.; Van Stipdonk, M. J. J. Am. Soc. Mass Spectrom. 2010 21 1322 1328. (62) Molesworth, S.; Osburn, S.; Van Stipdonk, M. J. Am. Soc. Mass Spec trom. 2010 21 1028 1036. (63) Polfer, N. C.; Oomens, J.; Suhai, S.; Paizs, B. J. Am. Chem. Soc. 2005 127 17154 17155. (64) Yoon, S. H.; Chamot Rooke, J.; Perkins, B. R.; Hilderbrand, A. E.; Poutsma, J. C.; Wysocki, V. H. J. Am. Chem. Soc 2008 130 17 644 17645 (65) Chen, X.; Yu, L.; Oomens, J.; Steill, J. D.; Powell, D. H.; Polfer, N. In Proceedings from the 57th ASMS Conference on Mass Spectrometry and Allied Topics Philadelphia, PA, 2009, p WOApm3:30. (66) Polfer, N. C.; Bohrer, B. C.; Plasencia, M. D.; Paizs, B.; Clemmer, D. E. J. Phys. Chem. A 2008 112 1286 1293. (67) Garcia, I.; Giles, K.; Bateman, R. H.; Gaskell, S. J. J. Am. Soc. Mass Spectrom. 2008 19 609 613. (68) Chen, X.; Powell, D. H.; Polfer, N. C. In Proceedings from the 56th American Society for Mass Spectrometry Conference Denver, CO, 2008, p MPMM 334. (69) Gucinski, A. C.; Somogyi, A.; Chamot Rooke, J.; Wysocki, V. H. J. Am. Soc. Mass Spectrom. 2010 21 1329 1338. (70) Bythell, B.; Somogyi, A.; Paizs, B. J. Am. Soc. Mass Spectrom. 2009 20 618 624. (71) Somogyi, A. J. Am. Soc. Mass Spectrom. 2008 19 1771 1775. (72) Fattahia, A.; Zekavata, B.; Solouki, T. J. Am. Soc. Mass Spectrom. 2009 21, 358 369 (73) Aebersold, R.; Goodlett, D. R. Chem. Rev. 2001 101 269 295.

PAGE 142

142 (74) Fernandez Lima, F. A.; Becker, C.; McKenna, A. M.; Rodgers, R. P.; Marshall, A. G.; Russell, D. H. Anal. Chem. 2009 81 9941 9947. (75) Stokes, A. A.; Clarke, D. J.; Weidt, S.; Langridge Smith, P.; Mackay, C. L. Int. J. Mass Spectrom. 2010 289 54 57. (76) Strohal m, M.; Novak, P.; Pompach, P.; Man, P.; Kavan, D.; Witt, M.; Dzubak, P.; Hajduch, M.; Havlicek, V. J. Mass Spectrom. 2009 44 1565 1570. (77) Tipton, J. D.; Carter, J. D.; Mathias, J. D.; Emmett, M. R.; Fanucci, G. E.; Marshall, A. G. Anal. Chem. 2009 81 7611 7617. (78) Li, L. J.; Masselon, C. D.; Anderson, G. A.; Pasa Tolic, L.; Lee, S. W.; Shen, Y. F.; Zhao, R.; Lipton, M. S.; Conrads, T. P.; Tolic, N.; Smith, R. D. Anal. Chem. 2001 73 3312 3322. (79) Patrie, S. M.; Robinson, D. E.; Meng, F. Y.; Du, Y .; Kelleher, N. L. Int. J. Mass Spectrom. 2004 234 175 184. (80) Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Mass Spectrom. Rev. 1998 17 1 35. (81) Marshall, A. G. Int. J. Mass Spectrom. 2000 200 331 356. (82) Amster, I. J. J. Mass Spectrom. 1996 31 1325 1337. (83) Lawrence, E. O.; Livingston, M. S. Phys. Rev. Lett. 1932 40 555. (84) Sommers, H.; Thomas, H. A.; Hipple, J. A. Phys. Rev. 1951 82 697 702 (85) Comisaro.Mb; Marshall, A. G. Chem. Phys. Lett. 1974 26 489 490. (86) Comisaro.M b; Marshall, A. G. Chem. Phys. Lett. 1974 25 282 283. (87) Dehmelt, H. Reviews of Modern Physics 1990 62 525 530. (88) Schaub, T. M.; Hendrickson, C. L.; Horning, S.; Quinn, J. P.; Senko, M. W.; Marshall, A. G. Anal. Chem. 2008 80 3985 3990. (89) Comi sarow, M. B. J. Chem. Phys. 1978 69 4097 4104. (90) Marshall, A. G.; Verdun, F. R. Fourier Transforms in NMR, Optical, and Mass Spectrometry: A User's Handbook ; First Edition ed.; Elsevier: Amsterdam, 1990. (91) Barrow, M. P.; Burkitt, W. I.; Derrick, P. J. Analyst 2005 130 18 28. (92) Feng, X. D.; Clipston, N.; Brown, T.; Cooper, H.; Reuther, U.; Hirsch, A.; Derrick, P. J.; Drewello, T. Rapid Comm Mass Spectrom 2000 14 368 370.

PAGE 143

143 (93) Hu, Q.; Noll, R. J.; Li, H.; Makarov, A.; Hardmanc, M.; Cooks, R. G. J. Mass Spectrom. 2005 40 430 443. (94) Ledford, E. B.; Rempel, D. L.; Gross, M. L. Anal. Chem. 1984 56 2744 2748. (95) Jeffries, J. B.; Barlow, S. E.; Dunn, G. H. Int. J. Mass Spectrom. 1983 54 169 187. (96) Woodin, R. L.; Bomse, D. S.; Beauchamp, J L. J. Am. Chem. Soc. 1978 100 3248 3250. (97) Bomse, D. S.; Woodin, R. L.; Beauchamp, J. L. J. Am. Chem. Soc. 1979 101 5503 5512. (98) Woodin, R. L.; Bomse, D. S.; Beauchamp, J. L. Chem. Phys. Lett. 1979 63 630 636. (99) Grant, E. R.; Schulz, P. A.; Sudbo, A. S.; Shen, Y. R.; Lee, Y. T. Phys. Rev. Lett. 1978 40 115 118. (100) Lehmann, K. K.; Scoles, G.; Pate, B. H. Annual Review of Physical Chemistry 1994 45 241 274. (101) Black, J. G.; Yablonovitch, E.; Bloembergen, N.; Mukamel, S. Phys. Rev. Let t. 1977 38 1131 1134. (102) Bagratashvili, V. N. Multiple photon infrared laser photophysics and photochemistry ; Harwood Academic Publishers: Chur [Switzerland] ; New York, NY, 1985. (103) Polfer, N. C.; Oomens, J. Mass Spectrom. Rev. 2009 28 468 494. (104) Eyler, J. R. Mass Spectrom. Rev. 2009 28 448 467. (105) Shin, S. K.; Beauchamp, J. L. J. Am. Chem. Soc. 1990 112 2057 2066. (106) Peiris, D. M.; Cheeseman, M. A.; Ramanathan, R.; Eyler, J. R. J. Phys. Chem. 1993 97 7839 7843. (107) Atrill, S. R .; Stace, A. Phys. Chem. Chem. Phys. 2000 2 823 830. (108) Cagmat, E. B.; Szczepanski, J.; Pearson, W. L.; Powell, D. H.; Eyler, J. R.; Polfer, N. C. Phys. Chem. Chem. Phys. 2010 12 3474 3479. (109) Mino Jr., W. K. S., J.; Pearson, W.L.; Powell,D.H.; D unbar, R.C.; Eyler, J.R.; Polfer, N.C. Int. J. Mass Spectrom. In press (110) Englander, S. W.; Sosnick, T. R.; Englander, J. J.; Mayne, L. Current Opinion in Structural Biology 1996 6 18 23.

PAGE 144

144 (111) Anderegg, R. J.; Wagner, D. S.; Stevenson, C. L.; Borchar dt, R. T. J. Am. Soc. Mass Spectrom. 1994 5 425 433. (112) Dharmasiri, K.; Smith, D. L. Anal. Chem. 1996 68 2340 2344. (113) Wagner, D. S.; Anderegg, R. J. Anal. Chem. 1994 66 706 711. (114) Englander, S. W. J. Am. Soc. Mass Spectrom. 2006 17 1481 14 89. (115) Engen, J. R.; Smith, D. L. Anal. Chem. 2001 73 256A 265A. (116) Busenlehner, L. S.; Armstrong, R. N. Archives of Biochemistry and Biophysics 2005 433 34 46. (117) Taylor, S. S.; Yang, J.; Wu, J.; Haste, N. M.; Radzio Andzelm, E.; Anand, G. Bio chim. Biophys. Acta, Proteins Proteomics 2004 1697 259 269. (118) Freitas, M. A.; Hendrickson, C. L.; Emmett, M. R.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 1998 9 1012 1019. (119) Campbell, S.; Rodgers, M. T.; Marzluff, E. M.; Beauchamp, J. L. J. Am. Chem. Soc. 1994 116 9765 9766. (120) Gard, E.; Green, M. K.; Bregar, J.; Lebrilla, C. B. J. Am. Soc. Mass Spectrom. 1994 5 623 631. (121) Green, M. K.; Lebrilla, C. B. Mass Spectrom. Rev. 1997 16 53 71. (122) Wyttenbach, T.; Bowers, M. T. J. Am. Soc. Mass Spectrom. 1999 10 9 14. (123) Chipuk, J. E.; Brodbelt, J. S. J. Am. Soc. Mass Spectrom. 2007 18 724 736. (124) Crestoni, M. E.; Fornarini, S. J. Mass Spectrom. 2003 38 854 861. (125) Freitas, M. A.; Marshall, A. G. J. Am. Soc. Mass Spectrom 2001 12 780 785. (126) Mo, J. J.; Hakansson, K. Anal. Chem. 2007 79 7893 7898. (127) Campbell, S.; Rodgers, M. T.; Marzluff, E. M.; Beauchamp, J. L. J. Am. Chem. Soc. 1994 116 9765 9766. (128) Campbell, S.; Rodgers, M. T.; Marzluff, E. M.; Beauchamp, J. L. J. Am. Chem. Soc. 1995 117 12840 12854. (129) Harrison, A. G. Mass Spectrom. Rev. 1997 16 201 217. (130) Bleiholder, C.; Suhai, S.; Paizs, B. J. Am. Soc. Mass Spectrom. 2006 17 1275 1281.

PAGE 145

145 (131) Wu, Z. C.; Fenselau, C. J. Am. Soc. Mass Spectrom 1992 3 863 866. (132) Wu, J. Y.; Lebrilla, C. B. J. Am. Chem. Soc. 1993 115 3270 3275. (133) Zhang, K.; Zimmerman, D. M.; Chungphillips, A.; Cassady, C. J. J. Am. Chem. Soc. 1993 115 10812 10822. (134) Gronert, S.; Simpson, D. C.; Conner, K. M. J. Am Soc. Mass Spectrom. 2009 20 2116 2123. (135) Reid, G. E.; Simpson, R. J.; O'Hair, R. A. J. Int. J. Mass Spectrom. 1999 191 209 230. (136) Herrmann, K. A.; Kuppannan, K.; Wysocki, V. H. Int. J. Mass Spectrom. 2006 249/250 93 105. (137) Garcia, I.; G iles, K.; Bateman, R. H.; Gaskell, S. J. J. Am. Soc. Mass Spectrom. 2008 19 1781 1787. (138) O'Hair, R. A. J.; Reid, G. E. Eur. Mass Spectrom. 1999 5 325 334. (139) Harrison, A. G. Int. J. Mass Spectrom. 2001 210 361 370. (140) Cooper, T.; Talaty, E. ; Grove, J.; Van Stipdonk, M. J.; Suhai, S.; Paizs, B. J. Am. Soc. Mass Spectrom. 2006 1654 1664. (141) Bythell, B.; Molesworth, S.; Osburn, S.; Cooper, T.; Paizs, B.; Van Stipdonk, M. J. J. Am. Soc. Mass Spectrom. 2008 19 1788 1798. (142) Molesworth, S .; Osburn, S.; van Stipdonk, M. J. J. Am. Soc. Mass Spectrom. 2009 20 2174 2181. (143) Oepts, D.; van der Meer, A. F. G.; van Amersfoort, P. W. Infrared phys. technol. 1995 36 297 308. (144) Valle, J. J.; Eyler, J. R.; Oomens, J.; Moore, D. T.; van der Meer, A. F. G.; von Helden, G.; Meijer, G.; Hendrickson, C. L.; Marshall, A. G.; Blakney, G. T. Rev. Sci. Instrum. 2005 76 23103 23109 (145) Polfer, N. C.; Oomens, J.; Moore, D. T.; von Helden, G.; Meijer, G.; Dunbar, R. C. J. Am. Chem. Soc. 2006 128 517 525. (146) Bagratashvili, V. N.; Letokov, V. S.; Makarov, A. A.; Ryabov, E. A. Multiple Photon Infrared Laser Photophysics and Photochemistry Harwood, Chur., 1985. (147) Perkins, B.; Chamot Rooke, J.; Yoon, S.; Gucinski, A.; Somogyi, A.; Wysocki, V. H J. Am. Chem. Soc. 2009 131 17528 17529.

PAGE 146

146 (148) Headrick, J. M.; Diken, E. G.; Walters, R. S.; Hammer, N. I.; Christie, R. A.; Cui, J.; Myshakin, E. M.; Duncan, M. A.; Johnson, M. A.; Jordan, K. D. Science 2005 308 1765 1769. (149) Roscioli, J. R.; McC unn, L. R.; Johnson, M. A. Science 2007 316 249 254. (150) Asmis, K. R.; Pivonka, N. L.; Santambrogio, G.; Brummer, M.; Kaposta, C.; Neumark, D. M.; Woste, L. Science 2003 299 1375 1377. (151) Fridgen, T. D.; McMahon, T. B.; MacAleese, L.; Lemaire, J.; Maitre, P. J. Phys. Chem. A 2004 108 9008 9010. (152) Moore, D. T.; Oomens, J.; van der Meer, A. F. G.; von Helden, G.; Meijer, G.; Valle, J. J.; Marshall, A. G.; Eyler, J. R. Chem. Phys. Chem. 2004 5 740 743. (153) Fridgen, T. D.; MacAleese, L.; Mait re, P.; McMahon, T. B.; Boissel, P.; Lemaire, J. Phys. Chem. Chem. Phys. 2005 7 2747 2755. (154) Chin, W.; Compagnon, I.; Dognon, J. P.; Canuel, C.; Piuzzi, F.; Dimicoli, I.; von Helden, G.; Meijer, G.; Mons, M. J. Am. Chem. Soc. 2005 127 1388 1389. (1 55) Inokuchi, Y.; Kobayashi, Y.; Ito, T.; Ebata, T. J. Phys. Chem. A 2007 11 3209 3215. (156) Bleiholder, C.; Osburn, S.; Williams, T. D.; Suhai, S.; Van Stipdonk, M.; Harrison, A. G.; Paizs, B. J. Am. Chem. Soc. 2008 130 17774 17789. (157) Green, M. K. ; Lebrilla, C. B. Mass Spectrom. Rev. 1997 16 53 71. (158) Gard, E.; Green, M. K.; Bregar, J.; Lebrilla, C. B. J. Am. Soc. Mass Spectrom. 1994 5 623 631. (159) Harrison, A. G.; Young, A. B.; Bleiholder, C.; Suhai, S.; Paizs, B. J. Am. Chem. Soc. 2006 1 28 10364 10365. (160) Polfer, N.; Paizs, B.; Snoek, L. C.; Compagnon, I.; Suhai, S.; Meijer, G.; von Helden, G.; Oomens, J. J. Am. Chem. Soc. 2005 127 8571 8579. (161) Frison, G.; van der Rest, G.; Turecek, F.; Besson, T.; Lemaire, J.; Maitre, P.; Chamo t Rooke, J. J. Am. Chem. Soc. 2008 130 14916 14917. (162) Lucas, B.; Gregoire, G.; Lemaire, J.; Maitre, P.; Glotin, F.; Schermann, J. P.; Desfrancois, C. Int. J. Mass Spectrom. 2005 243 97 105. (163) Lucas, B.; Gregoire, G.; Lemaire, J.; Maitre, P.; Or tega, J. M.; Rupenyan, A.; Reimann, B.; Schermann, J. P.; Desfrancois, C. Phys. Chem. Chem. Phys. 2004 6 2659 2663.

PAGE 147

147 ( 164) Simon, A.; MacAleese, L.; Maitre, P.; Lemaire, J.; McMahon, T. B. J. Am. Chem. Soc. 2007 129 2829 2840. (165) Vachet, R. W.; Bisho p, B. M.; Erickson, B. W.; Glish, G. L. J. Am. Chem. Soc. 1997 119 5481 5488. (166) Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham III, T. E.; Ross, W. S.; Simmerling, C. L.; Darden, T. A.; Merz, K., M.; Stanton, R. V.; Cheng, A. L.; Vincent, J. J.; Crowley, M.; Tsui, V.; Radmer, R.; Duan, Y.; Pitera, J.; Massova, I. G.; Seibel, G. L.; Singh, U. C.; Weiner, P. J.; Kollmann, P. A.; University of California: San Francisco. (167) Gaussian 03, Revision C.02, Frisch, M. J. T., G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O. ; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A; Gaussian, Inc.: Wallingford, CT, 2004. (168) Cieplak, P.; Cornell, W. D.; Bayly, C. I.; Kollmann, P. A. Computat. Chem. 1995 16 1357 1377. (169) Merrick, J. P.; Moran, D.; Rad om, L. J. Phys. Chem. A 2007 111 11683 11700. (170) Paizs, B.; Suhai, S. J. Am. Soc. Mass Spectrom. 2004 15 103 113. (171) Wyttenbach, T.; Bowers, M. T. J. Am. Soc. Mass Spectrom. 1999 10 9 14. (172) Bythell, B. J.; Knapp Mohammady, M.; Paizs, B.; H arrison, A. G. J. Am. Soc. Mass Spectrom. 2010 21 1352 1363.

PAGE 148

148 BIOGRAPHICAL SKETCH Xian Chen was born to Shanfeng Chen and Suyun Liao in Xiamen, a city on the southeast coast of China. She was the only child in the family. In 1999, she graduated from Xi amen No.1 Middle School after studying for six years. She then moved to Hefei, Anhui, to attend the University of Science and Technology of China (USTC), where she enrolled in the Department of Polymer Science and Engineering. She joined the group of Dr. G uangzhao Zhang and worked on her undergraduate research thesis entitled Charaterization of Surfactants using Quartz Crystal Microbalance (QCM). In July 2004, she received a Bachelor of Engineering degree. A month after graduation, she flew overseas to Gain esville, Florida, to begin her graduate studies in analytical chemistry in research group and got a masters degree with a thesis entitled Synthesis of Semiconductor Nano crystals using Selenium Dioxide in 2007. She then joined Dr. research group and began her work on investigating the influences on structures of peptide b fragment ions with infrared multiple photon dissociation, gas phase hydrogen/deuteri um exchange, and Fourier transform ion cyclotron resonance mass spectrometry. She received her Doctor of Philosophy from the University of Florida in December 2010