Citation
Mass Spectrometric Studies of Asparagine Synthetase and Its Role in the Drug-Resistant Form of Acute Lymphoblastic Leukemia

Material Information

Title:
Mass Spectrometric Studies of Asparagine Synthetase and Its Role in the Drug-Resistant Form of Acute Lymphoblastic Leukemia
Creator:
ABBATIELLO, SUSAN EUGENIA LINDYBERG
Copyright Date:
2008

Subjects

Subjects / Keywords:
Amino acids ( jstor )
Cell lines ( jstor )
Digestion ( jstor )
Fluorescence ( jstor )
Gels ( jstor )
Ions ( jstor )
Mass spectroscopy ( jstor )
Nuclear proteins ( jstor )
Proteomics ( jstor )
Reagents ( jstor )
City of Gainesville ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Susan Eugenia Lindyberg Abbatiello. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
11/30/2006
Resource Identifier:
496613321 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

MASS SPECTROMETRIC STUDIES OF ASPARAGINE SYNTHETASE AND ITS ROLE IN THE DRUG-RESISTANT FORM OF ACUTE LYMPHOBLASTIC LEUKEMIA By SUSAN EUGENIA LIN DYBERG ABBATIELLO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Susan Eugenia Lindyberg Abbatiello

PAGE 3

This dissertation is dedicated to the memory of Virginia Schwab

PAGE 4

iv ACKNOWLEDGMENTS Completion of this degree to my satisfac tion would simply not have been possible if it weren’t for the support, guidance, and in spiration of the followi ng people. First, I would like to thank my advisor, Dr. John R. Eyler. His gentle support and wealth of advice and experience truly helped to keep me focused on my goals. My gratitude goes to Dr. Nigel G. J. Richards, who went a bove and beyond the role of a committee member to provide support for my research, and w hose high standards kept me striving to do better. I’d like to thank my committee memb ers, Dr. Nancy Denslow, Dr. David Powell, and Dr. Richard Yost, three individuals who have offered advice and guidance in times when I greatly needed it. My family has always supported my crazy ideas, including returning to graduate school. My parents deserve a w ealth of thanks for their ne ver ending support, for their sacrifices to get me to this point, and for ne ver telling me I couldn’t do it. Special thanks to my brother, who told me I was stubborn enoug h to finish my degree, and to my sister, who labored through graduate schoo l at the same time. And I wish to thank Al and Mary Grace Abbatiello for their support and faith in my abilities. Graduate school brings you together with people you might never meet, even if they lived only 8 miles away from you for year s. Lani Cardasis has been a great friend and support through all of the challenging times, and she has been my finest arch rival. Lisa Regalla has provided motivation and much help with scientific techniques, and has

PAGE 5

v been a supportive and dependable friend. Cory Toyota and Jemy Gutierrez of the Richards lab have made “doing th e science” fun and tolerable. I fervently thank them all. The bulk of my best data was accomplished due to the efforts of Thomas P. Conrads. I wish to thank him for taking a chance on me, for his seemingly endless support, and for allowing my repeated visits to his lab, which have resulted in success beyond what I had hoped. Finally, I would like to ac knowledge the support of my husband, Russ Abbatiello. It is not easy to return to school, but it is equally difficult for the spouse of the graduate student. His support, advice, patience, and sacr ifice have been the primary reasons I have made it to this point.

PAGE 6

vi TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................xi LIST OF FIGURES.........................................................................................................xiii LIST OF OBJECTS........................................................................................................xvii ABSTRACT...................................................................................................................xvii i CHAPTER 1 INTRODUCTION........................................................................................................1 Acute Lymphoblastic Leukemia...................................................................................1 Symptoms and Diagnosis......................................................................................1 Treatment of ALL..................................................................................................1 The Role of Asparagine Synthe tase in Drug-resistant ALL.........................................2 The MOLT-4 Cell Line.........................................................................................3 Studies with Patients..............................................................................................4 Detection Methods for Asparagine Synthetase.....................................................5 Objectives of Research.................................................................................................7 Characterization of the Primary St ructure of Recombinant Human AS...............7 Development of a Method for Quantitation of hAS in Cell Lines........................7 Evaluation of Proteomics Me thods for Detection of AS.......................................8 Investigation of Changes in Global Pr otein Expression as a Function of LAsparaginase Treatment.....................................................................................9 2 TECHNOLOGIES SUPPORTING STUDIES OF THE PROTEOME.....................10 Introduction.................................................................................................................10 Separation Methods for the Studies of Proteins in Complex Mixtures......................11 Two-Dimensional Gel Electrophoresis...............................................................12 Liquid Chromatography......................................................................................15 Sample Derivatization.........................................................................................17 MS Detection and Identification of Proteins..............................................................19 Ionization.............................................................................................................19 Electrospray ionization.................................................................................19

PAGE 7

vii Matrix-assisted laser desorption/ionization..................................................21 Mass Analyzers...................................................................................................22 Quadrupole ion trap......................................................................................22 Triple quadrupole.........................................................................................24 Time-of-flight...............................................................................................25 Fourier transform ion cyclotron resonance..................................................25 Hybrid mass analyzers.................................................................................27 Mass Detectors....................................................................................................28 Application of Technology.........................................................................................29 3 CHARACTERIZATION OF RECOMBINANT HUMAN ASPARAGINE SYNTHETASE...........................................................................................................30 Introduction.................................................................................................................30 Experimental Procedure..............................................................................................31 Expression and Purification of rhAS...................................................................31 Simple, In-Solution Digestion of Purified rhAS.................................................33 Denaturation, Reduction, Al kylation and Digestion...........................................33 Investigation of Enzyme to Substrate Ratio........................................................34 SDS-PAGE Separation and In-gel Digestion......................................................34 Desalt of Digest Using Single-Step Elution........................................................35 Desalt of Digest Using Multi -Step Gradient Elution..........................................36 ESI-FTICR MS Analysis of Protein Digests.......................................................36 MALDI-TOF MS Analysis of Protein Digests...................................................37 Results and Discussion...............................................................................................38 Effect of Denaturation, Reduction and Alkylation of Cysteines.........................42 In-Gel Digestion of rhAS....................................................................................44 Investigation of Off-Line Desalt Conditions.......................................................46 Analysis of rhAS Digests by Altern ative Modes of Ionization and Mass Analyzers.........................................................................................................48 Conclusions.................................................................................................................51 4 METHOD DEVELOPMENT FOR THE QU ANTITATION OF ASPARAGINE SYNTHETASE USING MASS SPECTROMETRY.................................................52 Introduction.................................................................................................................52 Materials and Methods...............................................................................................54 Evaluation of Recombinant Human As paragine Synthetase for Robust Peptide Standards.............................................................................................54 Heavy-isotope Peptide Preparation and Addition...............................................55 Generation of the Response Curve for Each Peptide..........................................56 LC-ESI-MS/MS analysis.....................................................................................56 MS Data Analysis................................................................................................57 Cell Lysis and Desalt...........................................................................................58 Preparation of ALL Patient Samples...................................................................60 SDS-PAGE Separation........................................................................................60 Western Blot Analysis.........................................................................................61

PAGE 8

viii In-gel Digestion and Heavy-Isotope Peptide Addition.......................................62 Extraction and Analysis of Digested Peptides.....................................................63 Results and Discussion...............................................................................................64 Selection of Peptides for Use as Heavy-Isotope Standards.................................65 Generation of the Response Curve for Standard Peptides of rhAS by LC/MS/MS.......................................................................................................66 Investigation of Protein Extrac ts from Cancer Cell Lines...................................74 Analysis of additional MOLT4 S and R protein samples...........................78 Investigation of peptid e internal standards...................................................79 Analysis of ALL Patient Samples.......................................................................81 Conclusions.................................................................................................................86 5 INVESTIGATION OF TWO DIMENSIONAL GEL ELECTROPHORESIS FOR DETECTION OF ASPARAGINE SYNTH ETASE IN MOLT-4 CELL LINES......87 Introduction.................................................................................................................87 Materials and Methods...............................................................................................90 2D-DIGE Experiment..........................................................................................90 Preparation of cell lysate samples for 2D-DIGE..........................................91 CyDye labeling procedure............................................................................91 Isoelectric focusing of 100 g combined CyDye labeled sample on pH 311 IPG strip.............................................................................................92 Isoelectric focusing of CyDye labe led samples with 5 different pH gradients..................................................................................................92 SDS-PAGE second dimension.....................................................................93 Modification of method to targ et asparagine synthetase..............................93 Visual Staining and Protein Identification...................................................94 MALDI-Q-TOF MS analysis of in-gel digests............................................95 LC/MS/MS analysis of in-gel digests..........................................................95 Peptide and protein identification................................................................96 Results and Discussion...............................................................................................97 2D-DIGE of MOLT-4 Samples Using Broad pH Range....................................97 Analysis of MOLT-4 Protei ns Over pH Range 5.5-6.7.....................................100 Analysis of MOLT-4 Proteins Usin g Several Narrow pH Range IPG Strips.....................................................................................................103 Conclusions...............................................................................................................105 6 TARGETED DETECTION AND R ELATIVE QUANTITATION OF ASPARAGINE SYNTHETASE IN MO LT-4 CELL LINES USING ISOTOPECODED AFFINITY TAGS......................................................................................106 Introduction...............................................................................................................106 Materials and Methods.............................................................................................109 Preliminary Peptide Derivatization Experiments..............................................109 ESI-FTICR MS Analysis of Biotinylated Peptides...........................................110 ICAT Derivatization of Purified rhAS..............................................................110

PAGE 9

ix ICAT Labeling of MOLT-4 S Cytosolic Proteins with 1% and 0.1% rhAS Spike..............................................................................................................112 ICAT Labeling of MOLT-4 S and R Cell Lines...............................................113 LC/MS/MS Analysis of ICAT Labeled rhAS............................................113 Results and Discussion.............................................................................................114 Analysis of Biotinylated Peptid es in Mock ICAT Experiment.........................114 Detection of ICAT Labeled Pe ptides From Purified rhAS...............................120 Spike in study of 1% rhAS into MOLT-4 S cytosolic protein samples.....123 Spike in study of 0.1% rhAS into MOLT-4 S cytosolic protein samples..126 Results of ICAT analysis of MOLT-4 S and R cytosolic protein samples127 Changes in protein expression be tween MOLT-4 S and R cell lines.........128 Conclusions...............................................................................................................128 7 IDENTIFICATION OF CHANGES IN MOLT-4 S AND R CELL LINES BY ICAT AND LABEL-FREE PROTEOMIC ANALYSES........................................131 Introduction...............................................................................................................131 Experimental Procedure............................................................................................133 Cell Culture.......................................................................................................133 Cell Lysis for Total Protein Recovery...............................................................134 Cell Lysis for Nuclear Protein Recovery...........................................................134 Protein Desalt....................................................................................................135 Cleavable ICAT Labeling of Total Protein Samples.........................................136 Cleavable ICAT Labeling of Nuclear Protein Samples....................................136 Trypsin Digestion..............................................................................................137 Avidin Affinity Separation of cICAT Labeled Peptides...................................137 Strong Cation Exchange Fractionati on of cICAT-Labeled Peptides................138 Strong Cation Exchange Fractionati on of Label-Free MOLT-4 Nuclear Protein Digests...............................................................................................139 Reversed-Phase Liquid Chromatography ESI MS/MS for cICAT Samples.....139 Reversed-Phase Liquid Chromatography ESI MS/MS for Label-Free Nuclear Protein Digests...............................................................................................140 Peptide Identification and Quantitation.............................................................141 Western Blot Analysis.......................................................................................142 Results and Discussion.............................................................................................142 cICAT Analysis of MOLT-4 S and R Total Protein Fractions.........................143 cICAT Analysis of MOLT-4 S a nd R Nuclear Protein Fractions.....................150 Label-Free Analysis of MOLT-4 S and R Nuclear Protein Fractions...............154 Conclusions...............................................................................................................160 8 CONCLUDING REMARKS....................................................................................162 Utility of MS in Proteomics Studies..................................................................162 Characterization of rhAS by Mass Spectrometry..............................................164 Successful Method Development for th e Quantitation of AS Protein in Cancer Cell Lines and Leukemia Patient Samples........................................164

PAGE 10

x Evaluation of Proteomics Me thods for Detection of AS...................................166 Exploration of Protein Expression Be tween MOLT-4 S and R Cell Lines as a Function of L-Asparaginase Challenge.........................................................167 APPENDIX A TWO-DIMENSIONAL DIFFERE NTIAL GEL ELECTROPHORESIS FLUORESCENCE AND SILVER -STAINED GEL IMAGES...............................169 B PROTEINS EXHIBITING CHANGE IN EXPRESSION IN MOLT-4 CELL LINES AS A FUNCTION OF LASPARAGINASE CHALLENGE.....................180 C RESULTS OF ICAT INVESTIGATI ON OF MOLT-4 S AND R PROTEINS FROM TOTAL CELL LYSATE..............................................................................195 D RESULTS OF ICAT INVESTIGATI ON OF MOLT-4 S AND R PROTEINS FROM NUCLEAR PROTEIN FRACTION............................................................196 E PROTEINS IDENTIFIED BY LAB EL FREE PROTEOMIC INVESTIGATION OF MOLT-4 S AND R NUCLEAR PROTEINS.....................................................197 LIST OF REFERENCES.................................................................................................198 BIOGRAPHICAL SKETCH...........................................................................................208

PAGE 11

xi LIST OF TABLES Table page 4-1 Figures of Merit for Peptides...................................................................................68 4-2 Quantitation of AS from 7 Cancer Cell Lines..........................................................74 4-3 Fragment Ion Ratios for Natural Abundance and Heavy Isotope Peptide WINATDPSAR........................................................................................................80 4-4 Amounts of AS present in MOLT-4 S and R Cytosolic Protein Samples Detected by MS........................................................................................................81 4-5 Sample Code, Cell Count and Total Protein Recovery from ALL Patient Samples....................................................................................................................82 4-6 Results of AS Quantitation Using the WINATDPSAR Peptide..............................83 4-7 Results of AS Quantitation Using the ETFEDSNLIPK Peptide..............................83 4-8 Fragment Ion Ratios for WINATDPSA R Peptide in ALL Patient Samples............84 4-9 Fragment Ion Ratios for ETFEDSNLIPK Peptide in ALL Patient Samples...........85 5-1 Proteins Identified from 2D-DIGE, pH Range 5-5-6-7 (Figure 5-5)-....................102 5-2 Identification of Proteins Detect ed in 2D-DIGE Gel Spots by LC/MS/MS...........104 6-1 Peptides Used for Derivatization Ex periments with Biotinylation Reagent..........109 6-2 Cysteine-containing Peptides Found in rhAS........................................................114 6-3 Theoretical and Observed m/z Signals for Unlabeled and PEO-Biotin Labeled Peptides..................................................................................................................118 6-4 Theoretical ICAT Peptides from Trypsin Digested rhAS......................................121 6-5 Ratios of ICAT Labeled Peptid es from Spike-In Experiments..............................126 7-1 Proteins Identified in MOLT-4 S Nuclear Protein Faction Only...........................159

PAGE 12

xii 7-2 Proteins Identified in MOLT-4 R Nuclear Protein Faction Only...........................159 B-1 List of Proteins Identified and Quantitated in MOLT-4 S and R Cytosolic Protein Samples using ICAT..................................................................................180

PAGE 13

xiii LIST OF FIGURES Figure page 1-1 Schematic of the reaction catalyzed by L-asparaginase.............................................2 1-2 Schematic representing the reaction cat alyzed by asparagine synthetase. ...............3 3-1 Amino acid sequence of rhAS. ...............................................................................39 3-2 SDS-PAGE of expression and purification of rhAS. .............................................39 3-3 ESI-FTICR mass spectrum of trypsin digested rhAS. ...........................................41 3-4 Enlarged region of ESI-FTICR mass spectrum shown in Figure 3.3. ....................41 3-5 Comparison of sequence coverage of rh AS based on E:S ratio and alkylation. ....43 3-6 ESI-FTICR mass spectrum of an optimized rhAS digest. .....................................43 3-7 SDS-PAGE of rhAS and BSA samples for in-gel digestion....................................45 3-8 ESI-FTICR mass spectra for off-line step gradient separation of rhAS solution digest........................................................................................................................4 7 3-9 Sequence coverage of rhAS by ESI-FTI CR MS with and without off-line step gradient separation. ................................................................................................48 3-10 Comparison of rhAS sequence cove rage as detected by ESI-FTICR and MALDI-TOF MS. ..................................................................................................49 3-11 ESI-FTICR mass spectrum of rhAS digest obtained with over 260 K resolution. 50 4-1 ESI-FTICR MS spectrum of trypsin digested rhAS. ..............................................66 4-2 Full scan MS/MS spectra of pep tides WINATDPSAR a nd WINATD*PSAR. ....69 4-3 Full scan MS/MS spectra of peptides ETFEDSNLIPK and ET*FEDSNLIPK. ....70 4-4 Ion chromatograms for 1:1 ratio of light and heavy isotope peptides. ...................71 4-5 Ion chromatograms for a 1:1 ratio of light and heavy isotope peptides. ................72

PAGE 14

xiv 4-6 Graphical representation of the ra tio of MS response of light and heavy WINATDPSAR peptide versus ra tio of peptide amount. ......................................73 4-7 Graphical representation of the ra tio of MS response of light and heavy ETFEDSNLIPK peptide versus ratio of peptide amount. ......................................73 4-8 SDS-PAGE gel of total pr otein lysate from 7 differ ent cancer cell lines. .............74 4-9 Graphical representation of MS quan titation of AS in 7 different cancer cell lines.......................................................................................................................... 75 4-10 Western blot of SDS-PAGE gel si milar to that shown in Figure 4.8. ....................76 4-11 Comparison of Western blot and MS quant itation data for the detection of AS in cancer cell lines. .....................................................................................................77 4-12 Comparison of Western blot and MS quantitation data. ........................................78 4-13 SDS-PAGE and Western Blot of MOLT-4 S and R samples. ...............................79 4-14 SDS-PAGE and Western Blot of ALL Patient Samples. .......................................82 5-1 CyDye structures with exc itation and emission wavelengths..................................89 5-2 Schematic representation of the work flow of the 2D DIGE experiment. .............89 5-3 SDS-PAGE analysis of CyDye labele d MOLT-4 S and R protein samples. .........98 5-4 2D-DIGE merged fluorescence imag e of the MOLT-4 S and R cytosolic proteins. ..................................................................................................................99 5-5 2D-DIGE fluorescence image of 150 g MOLT-4 S and R cytosolic proteins, pH range 5.5-6.7. ..................................................................................................101 6-1 Schematic representation of the ICAT reagent and its f our main functional components. ..........................................................................................................106 6-2 Schematic representation of th e workflow of the ICAT method. ........................108 6-3 EZ-Link® Iodoacetyl-PEO2-Biotin {(+)-biotinyl-iodoacetamidyl-3,6dioxaoctanediamine}. ...........................................................................................115 6-4 RP-HPLC chromatograms of the PEObiotin reagent, peptide 1 alone, and peptide 1 with PEO-biotin. ...................................................................................116 6-5 Chromatograms of peptides 2, 3, 4, and 5 before and after PEO-biotin derivatization. .......................................................................................................117 6-6 ESI-FTICR MS spectrum of pep tide 1, sequence WGDMAAAYAK. ................118

PAGE 15

xv 6-7 ESI-FTICR MS spectrum of peptides 2, 3, 4, and 5 after derivatization with PEO-biotin reagent. ..............................................................................................119 6-8 Representative total ion chromatogram (T IC) of rhAS peptides labeled with light and heavy ICAT. ..................................................................................................120 6-9 MS spectrum showing the comparison of peptide I252-K271 with the light and heavy ICAT labels. ...............................................................................................121 6-10 MS spectrum and MS/MS spectrum of rh AS peptide F33-R48 detected in 1% rhAS spike sample. ...............................................................................................124 6-11 MS spectrum and MS/MS spectrum of rh AS peptide I252-K271 detected in 1% rhAS spike sample. ...............................................................................................125 7-1 Schematic representation of the cICAT workflow. ..............................................144 7-2 SCX chromatogram and corresponding histogram of peptides and proteins identified from the MOLT-4 total protein cICAT experiment. ............................146 7-3 Pie charts illustrating the types and lo cation of proteins identified from the MOLT-4 S and R cICAT experiment. .................................................................147 7-4 Histogram of the number of unique prot eins identified versus the number of peptides detected in the MOLT-4 total protein cICAT experiment.......................150 7-5 Pie charts illustrating the types and lo cation of proteins identified from the MOLT-4 S and R nuclear protein cICAT experiment. ........................................151 7-6 Histogram of the number of unique prot eins identified versus the number of peptides detected in the MOLT-4 nuc lear protein cICAT experiment..................153 7-7 Venn diagram of the overlap in unique proteins identified by ICAT analyses of the total and nuclear protein fract ions of the MOLT-4 cell line. .........................153 7-8 Strong cation exchange chromatogram s of MOLT-4 S and R nuclear protein digests. ..................................................................................................................156 7-9 Pie charts illustrating the types and lo cation of proteins identified from the MOLT-4 S and R nuclear protei n label-free experiment. ....................................157 7-10 Venn diagram comparing the proteins id entified in the MOLT-4 S and R nuclear protein fractions by cICAT and Labe l Free proteomic investigations...................160 A-1 Overlaid fluorescence image of 2D SDS-PAGE with pH gradient 3-11, non linear.......................................................................................................................17 0 A-2 Silver stained 2D SDS-PAGE with pH gradient 3-11, non linear.........................171

PAGE 16

xvi A-3 Overlaid fluorescence image of 2D SDS-PAGE with pH gradient 3-5.6. ...........172 A-4 Silver stained 2D SDS-P AGE with pH gradient 3-5.6...........................................173 A-5 Overlaid fluorescence image of SD S-PAGE with pH gradient 5.3-6.7. ..............174 A-6 Silver stained 2D SDS-PAGE with pH gradient 5.3-6.7........................................175 A-7 Overlaid fluorescence image of SDS-PAGE with pH gradient 6.2-7.5.................176 A-8 Silver stained 2D SDS-PAGE with pH gradient 6.2-7.5........................................177 A-9 Overlaid fluorescence image of SD S-PAGE with pH gradient 7-11. ..................178 A-10 Silver stained 2D SDS-PAGE with pH gradient 7-11. .........................................179

PAGE 17

xvii LIST OF OBJECTS Object page 1 ICAT Investigation of MOLT-4 S and R Proteins from Total Cell Lysate...........195 2 ICAT Investigation of MOLT-4 S and R Proteins from Nuclear Fraction............196 3 Label-Free Investigation of MOLT-4 S a nd R Proteins from Nuclear Fraction....197

PAGE 18

xviii Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MASS SPECTROMETRIC STUDIES OF ASPARAGINE SYNTHETASE AND ITS ROLE IN THE DRUG-RESISTANT FORM OF ACUTE LYMPHOBLASTIC LEUKEMIA By Susan Eugenia Lindyberg Abbatiello May 2006 Chair: John R. Eyler Major Department: Chemistry Development of drug-resistant acute lymphoblastic leukemia (ALL) is often accompanied by an increase in expression of asparagine synthetase (AS) by leukemic cells. Detection of this enzyme has historic ally been accomplished using either mRNA or antibody-based assays. Preliminary charac terization of recombinant human AS (rhAS) resulted in the identification of a number of peptides suitable for detection by mass spectrometry (MS) methods. Herein is desc ribed the development and evaluation of a method for direct detection a nd quantification of asparagine synthetase in leukemia cell lines using stable isotope labeled standard peptides and mass spectrometry. Quantification of AS was determined in seven different cancer cell lines and ALL patient samples using this LC/MS/MS-bas ed method, and results agreed well with Western blotting analyses. The cell lines we re differentiated into 4 different groups based on the amount of AS protein detect ed. The linear range of the method was established between 100 amol and 200 fmol , while the lower limit of detection was

PAGE 19

xix identified at 30 amol. Analysis of four cl inical samples from patients diagnosed with ALL resulted in the positive detection and quantitation of AS, providing a new method for direct measurement of the protein in complex samples. Additional experiments evaluating the cha nges in protein expr ession between drugsensitive and drug-resistant cancer cell lines were conducted to determine what other proteins were strongly affected by the e xposure to the chemotherapeutic enzyme Lasparaginase. The identification of such pr oteins might provide al ternative therapeutic targets for treatment.

PAGE 20

1 CHAPTER 1 INTRODUCTION Acute Lymphoblastic Leukemia Acute lymphoblastic leukemia (ALL) is a cancer of the blood, afflicting 30-40 people per million annually in the United Stat es, with the predominance of ALL cases occurring in children under the age of 10. Survival rate base d on initial diagnosis is quite high; however after treatment 10-20% of child ren will develop drug-re sistance. In 2005, there were 3970 new cases of ALL, with 1490 estimated deaths from the disease.1 Symptoms and Diagnosis Blood-forming stem cells are continually produced in the bone marrow and differentiate to form the red blood cells (e rythrocytes), white blood cells (leukocytes), and platelets (thrombocytes). ALL affects the development process of the leukocytes from the immature blast form of the cell, producing elevated levels of non-functional white blood cells, or lymphoblasts, in the blood. This condition result s in a crowding of the normally functioning erythrocytes, leukocyt es and thrombocytes in the blood, which can be diagnosed by symptoms including anem ia, bruising, and propensity for infection. Diagnosis of ALL usually relies on analysis of a blood or bone marrow sample, revealing elevated numbers of immature lymphoblasts. Treatment of ALL Treatment of ALL consists of a combinati on of chemotherapeutic drugs designed to kill the leukemia cells and prevent them from propagating. In many cases, the enzyme Lasparaginase is administered during induction therapy (treatment to reduce the number of

PAGE 21

2 cancer cells). The function of this enzyme is to convert the amino acid asparagine to aspartate (Figure 1.1), thus de pleting the bodyÂ’s supply of as paragine, a necessary amino acid for the synthesis of a large number of proteins within the cell. As a result, leukemic cells are unable to acquire a sufficient supply of asparagine to maintain their proliferative state and rapidly die. In the case of recurr ence, the leukemia cells become resistant to treatment with L-asparaginase, and prognosis is very poor. Figure 1.1. Schematic of the reaction cataly zed by L-asparaginase. L-Asparaginase converts the amino acid aspara gine to aspartate. The Role of Asparagine Synthetase in Drug-resistant ALL Leukemic lymphoblasts normally express very low, if any, amounts of the enzyme asparagine synthetase (AS). AS is a housekeeping enzyme that produces the bodyÂ’s supply of asparagine from as partate, adenosine triphospha te (ATP), and a source of ammonia, usually glutamine (Figure 1.2). A number of studies have explored the incidence of L-asparaginase-r esistance in leukemia cells to determine its cause and mechanism, and have uncovered a correlation between L-asparaginase resistance and an increase in the cellular expression of AS.2-6

PAGE 22

3 Figure 1.2. Schematic representi ng the reaction catalyzed by as paragine synthetase (AS). AS converts aspartate to asparagine in the presen ce of ATP and a source of ammonia, usually glutamine. The MOLT-4 Cell Line Medical research studies involving human patients or human samples are often difficult to implement because of a limite d sampling population and invasiveness of the sample collection. Interp retation of results is al so complicated by countless uncontrollable variables, including age, sex, di et, medical history, etc. These difficulties often lead researchers to cell line models th at can be grown in cu lture under carefully controlled conditions such that the response to certain stimuli can be measured using a variety of methods. The MOLT4 cell line is an excellent in vitro model for the study of alterations in the phenotype of leukemia cells in response to changing stimuli. The cells were originally isolated in 1971 from a 19-yea r-old male patient dia gnosed with recurrent ALL.7 The cells were immortalized so that th ey would not exhibit signs of age as a function of replication, while st ill maintaining original cell genotype, and are widely used for the study of ALL.8 Initial studies revealed that th e MOLT-4 cell line could be grown in the presence of increasing amounts of L-asparaginase (from 0.00001 – 1 Unit/mL) to create a resistant line.2, 5 This resistant line, called MOLT-4 R, demonstrated elevated levels of asparagine synthetase.5 Further studies ha ve explored the cellular response to L-

PAGE 23

4 asparaginase and have resulted in discovery of the unfolded protein response and the amino acid response pathways that appear to strongly influence the expression of AS in the resistant cells.6, 9, 10 Studies with Patients A number of studies have been cond ucted with blood or bone marrow samples from leukemia patients at initial diagnosis, in itiation of induction th erapy, and throughout the course of treatment, to bett er understand the function of AS in vivo and to monitor the patientsÂ’ response to the treatment.11-21 Initial results support ed the hypothesis developed from the in vitro studies that L-asparagi nase resistance was coupled with an increase in cellular expression of AS.11 Further investigations revealed that a difference in the cell lineage of the patient populati on strongly influenced the sens itivity to drug treatment.17, 21 The TEL-AML1-positive form of ALL results from the fusion of the TEL (translocation-Ets-leukemia) and AML1 (acute-myeloid-leukemia-1) genes.22, 23 This gene fusion is found in 1 out of 100 new borns, but only 1 out of every 100 TEL-AML1positive children will develop leukemia. The connection is unclear between this gene fusion and the development of leukemia, howev er it does result in a different lineage of leukemic cells, B cells, which have an altere d resistance to L-asparaginase and usually predict a favorable outcome.14, 21-23 In addition, this cell lineage has also been reported to express lower levels of AS than the TEL-AM L1-negative patients, supporting the results from in vitro studies.21 A very recent study by I. M. Appel et al. investigated Lasparaginase resistance in patient sample s and reported no significant changes in the expression of AS between TEL-AML1-pos itive and TEL-AML1-negative samples.24 While the disagreement between in vitro studies with leukemia ce ll lines and studies with

PAGE 24

5 patient samples is reasonable, the lack of unanimity in even the population of patient sample studies is troubling. Detection Methods for Asparagine Synthetase There are a variety of methods for the de tection of AS in cell lines and human samples. Investigation of cell lines is slight ly more direct, because all of the cells are leukemic, and there is little chance of cont amination with non-leukemi c cells in culture. In the case of human samples, the cells must first be isolated from the sample matrix, either blood or bone marrow. Once the cells ar e isolated, they are examined to determine what percentage are leukemic blast cells. Th e blasts can then be separated from non-blast cells through the use of immunomagnetic beads, as described by G. J. Kaspers et al .25 However, in cases of delayed initial diagnosis or relapse, the leuke mic blast cell count is quite high, a characteristic upon which diagnosis is based. Some current methods of AS detection rely on the selectiv e identification of messenger RNA (mRNA) that is contained in the cell, indi cating transcription of the specific gene, or DNA encoded for AS.26 Transcription of the DNA into mRNA, however does not directly correlate with prot ein translation and expression in the cell.27 Nonetheless, analysis of mRNA from cellular samples provides a good basis for comparison of the cellÂ’s response to a stim ulus. One such method is the real-time quantitative polymerase chain reaction (RQ-PCR) assay developed by T. Irino et al .16 This method has been demonstrated on both in vitro samples as well as those isolated from patients. The method has been determin ed to be quantitative a nd linear with a large dynamic range. In addition, it is easily au tomated and requires small amounts of samples, making it suitable for clinical studies. To be clear, results from this assay are quantitative in that the amount of AS mRNA de tected is expressed as a ratio of mRNA

PAGE 25

6 from AS and a control protein, GAPDH (g lyceraldehyde-3-phosphate dehydrogenase), which is expressed in high levels in most cel ls. Therefore, the quantitative value obtained from the RQ-PCR assay for AS does not dete rmine the amount of AS mRNA present in the cell, but instead a relative quantity th at can be compared to other samples. Additionally, the RQ-PCR assay does not de tect protein, and can not monitor posttranscriptional regu lation of AS. Methods that are used for the detection of AS protein from cell samples are limited to antibody-based assays and e xploit the highly se lective interaction of an antibody with its target protein. These methods of immunostaining28, 29 and Western blotting are considered very specific, and are very sensit ive, detecting ng to pg levels of protein.30 However, the observed signal is not directly from AS, but instead from a secondary antibody with a chemiluminescent (in the case of Western blotting) or fluorescent reagent attached, which is specific for the primary, -AS antibody. The detection of a signal using antibody-based assays is usually compar ed to either the molecular weight (as in Western blotting), or other methods of sample identification or loca lization. In addition, the procedures involved in Western blo tting and immunostaining of cells are laborintensive and cumbersome, and not suitable for routine clinical testing.16 Finally, activity assays are used to detect the function of the targ et enzyme, in this case, the production of asparagine.31-33 While these methods are specific for the presence of asparagine, they do not simultaneously measur e the amount of AS present. Therefore, AS activity results should be compared to ot her assays described above to ensure the source of asparagine is indeed from the functional enzyme.

PAGE 26

7 Currently, no method exists to directly measure the presence and amount of AS present in cellular samples. Furthermor e, it is not known what amounts of AS are necessary to cause the L-aspa raginase resistant phenotype. The established methods for AS detection are not suitable for quantitation of protein expr ession in cancer cells. For these reasons, development of a single robust method for the identification and quantitation of AS expression in cancer cells is in order. Objectives of Research The factors relating to drug resistance in ALL patients are still not fully understood, and in the absence of a method for direct qua ntitation of AS expressed in leukemia cells, the functional role of this protein remains el usive. Therefore, the following goals were pursued to focus on aspects that will better ch aracterize the role of AS in the occurrence of drug-resistant leukemia. Characterization of the Primary St ructure of Recombinant Human AS Expression and purification of recombinan t human AS (rhAS) has been developed in the lab of Nigel Richards, Department of Chemistry, University of Florida (Gainesville, FL).34 The amino acid sequence of the pr otein has only been reported based on isolation of human cDNA that encodes the enzyme.35 Therefore, with large amounts of purified enzyme available, the primary st ructure was verified by proteolytic cleavage and mass spectrometric analysis of the resulti ng peptides, and is discussed in Chapter 3. Development of a Method for Quantitation of hAS in Cell Lines Prior to the current work, there was no me thod available for direct detection and quantitation of AS in comple x mixtures. A technique for the quantitation of AS, using stable-isotope labeled peptid es as internal standards w ith LC/MS/MS detection, was developed. The method was optimized to enable low limits of detection (10-18 mol) of

PAGE 27

8 AS, and analysis of several cancer cell line s and clinical samples. These efforts are described in Chapter 4. Evaluation of Proteomics Met hods for Detection of AS A number of proteomics-based methods are available for the analysis of changes in protein expression between tw o cell states. Two-dimensional differential imaging gel electrophoresis separates a complex mixture of proteins, based on isoelectric point and molecular weight, resulting in a visual map of most proteins expresse d in a particular cell system. Differential fluorescent labeling of th ese proteins allows combination of proteins from two different cell states, and comparis on of changes between the two samples on the same 2D gel. This method was explored to evaluate its utility in the separation of AS from other more abundant proteins in the sample for relative quantitation and identification. The second technique that was evaluated wa s the isotope-coded affinity tag (ICAT) method, which uses sample derivatization, a ffinity chromatography and isotope-coded tags to identify and relatively quantify pr oteins from two different samples using MS detection.36 These methods were evaluated for the detection of changes in AS expression in MOLT-4 drug-sensitive (S) and drug-resistan t (R) cell lines. In the case where the method was not suitable for detection of AS, minor modifications were made to enhance the sensitivity of each method for the target. Additionally, changes in protein expression across the entire cell lysate were examined to explore further effect s of L-asparaginase drug challenge. These methods are described in chapters 5 and 6.

PAGE 28

9 Investigation of Changes in Global Prot ein Expression as a Function of LAsparaginase Treatment In order to better understand the cellular implications of L-asparaginase, MOLT-4 S and R cell lines were examined to determin e what changes in ot her protein expression, besides AS, could be observed. The ICAT method was used for a large-scale examination of changes in protein expr ession as a function of exposure to Lasparaginase. Particular atte ntion was paid to the nuclear protein fraction, in which both ICAT and label-free proteomic analyses were conducted to evaluate the overall identification of proteins. Chapter 7 a ddresses the capabilities of each method for proteomic investigation and ev aluates the detected changes in protein expression that may be a result of the L-asparaginase challenge.

PAGE 29

10 CHAPTER 2 TECHNOLOGIES SUPPORTING ST UDIES OF THE PROTEOME Introduction The word “proteome” was coined by Marc Wilkins in 1994 to succinctly describe the entire complement of proteins expre ssed in a given cell or tissue under certain conditions.37 Wilkins, a post-doctoral associate from Australia first used the word at a conference in Italy, and the name stuck. Th e study of the proteome, called proteomics, has gained increasing momentum over the past two decades, especially after completion of the Human Genome Project in 2000. Gene s are essentially cellular blueprints for proteins, which carry out the functions with in a cell. Once the genetic code for an organism has been determined, the array of proteins it may produce can be deduced from the DNA. Unfortunately, this information does not provide details on the form or function of the proteins, nor does it indicat e possible mutations or post-translational modifications. In addition, there has yet to be established a correlation between the amount of RNA, which translates the geneti c code into functional proteins, and the amount of corresponding protein present in a cell.38 Exploration of a cell or organism’s prot eome may likely provide information to better understand the role of prot eins in the response of the cell to certain stimuli. For example, if a cell normally grow s on a source of glucose, wh en the glucose is replaced with another carbohydrate, such as galactose, the physical response of the cell (change in size, rate of reproduction, or death) is linked to the biochemical processes going on within it, and changes in protein expression pr ovide an explanation of that response.

PAGE 30

11 Studies of the proteome have been carried out for many years, but have only been formally recognized as “proteomics” in the pa st two decades. The first studies of total protein expression from cells were carried out by Klose and O’Fa rrell in 1975 using twodimensional electrophoresis.38, 39 This method provided hi gh resolution separation of complex mixtures of proteins obtained from sources such as rabbit embryos, hepatoma cells, and from the nematode, C. elegans .39 Two-dimensional elect rophoresis is still widely used and has become a corner st one of many proteomic investigations. Technologies that provide better sensitivity, resolution, and offer the benefit of protein identification are continually being developed to further th e information that can be gained from proteomic studies. The purpose of this chapter is to describe the technol ogies that are used in proteomic investigations, including the de velopment and progression of separations science, mass spectrometry, and the tools used for targeted protein detection as well as global protein identification and quantitati on. Mass spectrometry (MS) is the main method of detection in the studies discussed in this dissertatio n, and the predominant methods used for facilitating MS detection will receive particular attention in this chapter. Separation Methods for the Studies of Proteins in Complex Mixtures The study of the entire complement of prot eins expressed in a given cell has been challenging, specifically due to the large number of proteins , on the order of thousands, that are present in cell samples. The range in concentrations of the different proteins within the sample also causes difficulty in fi nding a suitable method that will be able to detect the low abundance proteins as well as the proteins that make up the majority of the sample. For example, in the human plasma proteome, the concentrations of proteins

PAGE 31

12 within the sample span 10 orders of ma gnitude, where albumin is the most abundant protein at 35-50 mg/mL.40 At the other end of the spectrum, interleukin 6 is normally present at 0-5 pg/mL in the blood and is rout inely detected at health facilities as a measure of infection or inflammation. When te sted separately, the de tection of these two proteins is possible, demonstrating the capability of the current technologies.40 However, detection of these two protein molecules within the same analytical investigation is more difficult because the high abundance proteins of ten obscure or drown out detection of the less abundant protein molecules. When MS is used as the method of detection, special attention must be paid to the composition of the sample, especially with regards to contaminants such as salts, detergents, and overly abundant analytes. Sample preparation prior to MS analysis is extremely importa nt, with additional methods of separation allowing for increased selectivity for pr otein detection and identification. Two-Dimensional Gel Electrophoresis Early investigation of samples with co mplex protein mixtures was accomplished using two-dimensional gel electrophoresis (2D-GE).38, 39 This method first applies isoelectric focusing (IEF) to separate soluble pr oteins based on their isoelectric point (pI). Current technologies achieve this by absorbing the sample into dehydrated strips composed of multi-ampholyte buffers that form an immobilized pH gr adient (usually pH 3-11). When a voltage is applied across the strip, the positively charged proteins move toward the cathode, while the negatively char ged proteins move toward the anode. The charge on the proteins decreases as they pass thro ugh the pH gradient, where they finally arrive at the pH corresponding to their pI, and the net charge on the proteins is zero. The focused proteins within the IEF strip can then be applied to a second dimension for additional resolution. The second dimension separation step is usually sodium dodecyl

PAGE 32

13 sulfate-polyacrylamide gel el ectrophoresis (SDS-PAGE). This method is useful for separating proteins based on apparent molecula r weight. In this separation method, the proteins are introduced to the detergent SDS, which denatures and coats the proteins with a large number of negative ch arges. The proteins are th en applied to the top of a polyacrylamide gel where a voltage potentia l is applied and the negatively charged proteins migrate through the porous gel towards the anode at the bottom, moving at a rate that is inversely proportional to their molecular weight, where the smaller proteins travel farther through the gel than th e large proteins. After staini ng with Coomassie or silver stain, the result is a map of well-defined prot ein spots that can be classified by their migration point on the gel and excised for further analysis. One benefit of this method is the large se parating capacity of the gel, where often hundreds to thousands of proteins can be individually resolved.39, 41, 42 In addition, the pH range of the first dimension step can be reduced to allow better resolution of certain proteins over smaller changes in gradient , with calculated separation of up to 10,000 spots.41, 43 Another benefit to this method is th e ability to directly compare the 2D-GE maps from similar samples,39 providing a basis for quantita tive proteomics. Finally, the ability to excise the gels fo r Edman degradation or identific ation by in-gel digestion and mass spectrometric analysis of the peptides ha s driven this method to become one of the most widely used techniques for proteomic studies. The drawbacks to 2D-GE are unfortunate ly plentiful. The method is labor intensive, with the total analysis time from sa mple isolation to visual ization of the protein spots lasting upwards of 72 hours, often longer. The large range in protein concentration in total cell lysate samples is a limitation to the method also. Even though 2D-GE has

PAGE 33

14 demonstrated resolving power for a large quant ity of proteins, results have demonstrated its ability to preferentially is olate higher abundance proteins.41 In addition, proteins that fall into the categories of very acidic, very basic, or membrane proteins are usually not resolved.41, 44 There has been debate regarding th e reproducibility of 2D-GE, with a limitation that protein identification can not be based solely on migration point on the gel.45, 46 The use of an additional method for protein identificati on is necessary, and useful in determining if the spot on the gel contains more than one protein. Improvements in 2D-GE technology have cause d the method to still be one of the most widely used methods of sample se paration for proteomic investigations. Developments of sample de rivatization methods, such as fluorescence labeling, have facilitated relative quantitation of proteins fr om two to three sample sources that can be run on the same gel, removing the questi on of reproducibility and experimental variability between two samples. Tw o-dimensional differential imaging gel electrophoresis (2D-DIGE) has been used in th e past few years explicitly for monitoring changes in protein expression betw een two or three protein samples.47, 48 This method of protein derivatization labels pr oteins from up to three differe nt cell sources with cyanine dyes, highly fluorescent dyes with excitation and emissi on wavelengths that do not overlap with one another. Once the labeled proteins are combined and separated by 2DGE, the fluorescence images for each dye gr oup can be overlaid, and the changes in intensity for each protein spot can be determ ine and quantified. One of the shortcomings of this method is the inability to differentiat e different proteins that co-migrate to the same spot on the gel. However, the overall ability to quantify rela tive changes in protein expression between 2 or 3 samples is a valu able attribute, and is being used as a

PAGE 34

15 contemporary proteomic technique. Chapter 5 of this dissertation describes in more detail the applicati on of 2D-DIGE to leukemia cell lines. Liquid Chromatography Reversed-phase high performance liquid chromatography (RP-HPLC) is a widely applied method for the separation of peptides and proteins prior to MS analysis. RPHPLC readily separates the peptides and proteins base d on their hydrophobicity while concurrently removing salts and very hydrophili c components of the sample matrix. The eluent from the RP-HPLC separation can be directly coupled to the inlet of a mass spectrometer and the analytes are analyzed following sepa ration, called on-line LC/MS.49, 50 The separation capacity of the RP-HPLC is governed by the type of resin used, the size of the column, and the length and of the gr adient. In order to couple the HPLC flow directly to a mass spectrometer inlet, the flow must either be maintained at a low flow rate (> 250 L/min), or the flow can be split prior to coupling with the MS. If a split is used, then a portion of the flow may be directed to an additi onal detector, such as UV or fluorescence. However, with the improvement s being made with low-flow HPLC pumps and micro-capillary columns, the amount of sample being an alyzed is usually not enough to be detected by UV absorbance. The flow rate and size of RP-HPLC columns have been greatly reduced over the years, with current methods using flow rates of 500 nL/min or less, and column diameters between 50-100 m.44, 51, 52 Further development has c oupled the column with the electrospray ionization (ESI) tip, resulting in an increase in sensitivity for the detection of sub-picomolar peptides and proteins by on-line LC/MS.53, 54

PAGE 35

16 As with any method, additional modes of separation increase the selectivity and resolution, while allowing an increased sample size to be analyzed versus a onedimensional method.55 A second dimension of liquid sepa ration prior to MS analysis that has had significant impact on proteomics is strong cation exchange (SCX) fractionation prior to RP separation. This type of 2D-L C has been applied to complex mixtures of peptides after enzymatic digestion of proteins isolated from a cell or tissue, and can be performed off line or on-line with reversed-pha se LC (RPLC)/MS. The principle of the method is to first use SCX to separate the peptides based on charge, then separate the fractionated peptides by RPLC.56 This greatly reduces the sample complexity, affording the detection of hundreds to t housands of proteins (based on peptide identification) per complex sample. When the two types of chromatography are coupled on-line, it is referred to as multi-dimensional prot ein identification technology (MudPIT).56, 57 When using MudPIT, the SCX and RP resins are pack ed into the same column, integrated with an ESI tip for direct infusion into the MS. The SCX resin is added upstream from the RP resin, and allows a greater loading capacity fo r the sample. The peptides are eluted in a salt step gradient from the SCX resin, then separated by the RP resi n for MS detection. The separation method, with increasing salt bolus es, is repeated until the sample has been depleted from the SCX resin. One of the advant ages of this method is that it is unbiased, and is able to detect protein varieties that are more difficult to detect by 2D-GE, such as low abundance and membrane proteins, and prot eins with extremely acidic or basic pIs.44, 56, 57 Other methods of liquid chromatogra phy have been applied to proteomic investigations, but are perhap s a little less common than thos e described above. Affinity

PAGE 36

17 chromatography, thiol-exchange chromat ography and immobilized metal affinity chromatography (IMAC) have all been used to exploit some key feature of certain proteins of interest.58-61 Within the realm of this research, online RPLC/MS was the most common method of separation, using an inte grated micro-capilla ry column (ID = 75 m) and integrated ESI tip. Off-line SCX fractionation was applied to cell lysate samples (discussed in Chapter 7), and greatly incr eased the number of proteins identified as compared to RPLC/MS only. Sample Derivatization In an effort to aid in peptide and prot ein detection or to facilitate relative quantitation of proteins betw een two or more samples, of ten a derivatizat ion step is included to help differentiate the source of each analyte being detected (ie, from a control or experimental cell state). This type of sample handling is seen in the 2D-DIGE method, described above. An alternate labeling strategy that is very similar to 2D-DIGE, but does not require gel electrophoresis, is stable isotope labeling of protein samples. In these experiments, a control sample with natural is otopic abundances would be compared to an experimental sample that has heavy-isotope enrichment of certain atoms. This derivatization methodolo gy is conducive to MS detection because the same protein or peptide from different samples would have the same sequence, and would differ only by the mass addition of the stable isotopes. Ch emically, the two isotopomers are identical, except for the molecular mass, so the differen tially labeled peptides would co-elute for simultaneous detection in the MS. Comparis on of the ion intensities from each peptide ion can be accomplished using mass chromatograms, and the peak area for each peptide can be quantitatively compared. Some heavy is otopes that have been successfully used

PAGE 37

18 are 2H, 13C, 15N, 18O and 34S. The use of deuterium in labeli ng of proteins or peptides is one of the least expensive me thods of stable-isotope inco rporation, but it has one major drawback. The addition of 4 or more de uterium atoms to a peptide may alter its chromatographic profile from the same pep tide with hydrogen, making the quantitation of each peptide difficult.62 Alternative methods of stable isotope labeling of peptides include enzymatic digestion in 18O water, dimethyl labeling of lysine and -amino groups,63 and D3-methylation of aspartic acid, glutamic acid and the C-terminus of peptides,64 to name a few. These methods are mean t to be simple, cost efficient means of sample differentiation so that relative quantitation can be ca rried out by MS detection. Addition of stable isotopes to a protein sample can be carried out in vivo , by growing cells in the presence of stable isotop e-enriched media. This method is referred to as SILAC (stable isotope labeling with ami no acids in cell culture), and several stable isotope-enriched amino acids are added to media lacking those same amino acids.65 The cells are able to incorporate the stable is otope labeled amino acid, thus changing the mass of the protein in which that amino acid occurs . This too, has been a successful means of stable isotope labeling an d is widely applied to in vivo studies. A final method of sample derivatization disc ussed here is the is otope-coded affinity tag (ICAT) method. This method is discussed in greater detail in chapters 6 and 7 of this dissertation. Briefly, protein samples are derivatized with an ICAT reagent that has 3fold functionality: first, the reagent has an iodoacetyl reac tive group that targets reduced cysteine residues; second, a linking region that incorporates stable isotope atoms to cause a mass difference between the “light” and “heavy” reagent, and third, a biotin moiety that facilitates enrichment of the sample in ICAT-derivatized peptides, post proteolysis.36, 66

PAGE 38

19 This method is unique in that it incorporates differential sa mple labeling and a separation step that reduces the overall sample complexity after proteolysis. The ICAT strategy has proven suitable for the detection of low a bundance proteins, and for relative quantitation of changes in protein expr ession by MS detection. Sample derivatization techniques are c onstantly evolving, and depending on what question is being pursued within a particul ar biological system, there is likely an acceptable experimental design fo r answering that question. MS Detection and Identification of Proteins The predominant method of detection for proteomic studies is MS. This method lends itself well to the detec tion, identification, and quantitati on of proteins and peptides, especially after suitable sample separation. The topics discussed herein will address the methods of ionization, mass selection, ion fragmentation and ion detection used in proteomics. Ionization MS detection of proteins or peptides can only occur when these molecules are present in the gas phase as ions. Most bi omolecules are charged in solution, due to the variety of functional groups pr esent, but transfer of these fragile molecules into the gas phase can be difficult without sample degrad ation. Two main me thods of ionization, developed in the 1980Â’s, helped to launch th e investigation of bi omolecules using mass spectrometry. Electrospray ionization The first method discussed here is electro spray ionization (ESI), which was applied to biomolecules in 1989 by Fenn and coworkers.67 Ionization of liquid phase biomolecules is accomplished by passing the liq uid through a small capillary with a high

PAGE 39

20 voltage applied to the tip, imposing an electr ic field between it a nd a counter-electrode within close proximity.68 The result is the formation of charged droplets that emerge from the ESI tip, which through evaporati on and coulombic repulsion shrink and split into increasingly smaller dropl ets until they are gas phase ions. The point at which the ions are desorbed from the droplet is calle d the Rayleigh limit, when the charge repulsion exceeds the surface tens ion of the droplet.56, 69 Despite the repulsi on and “explosion” of ions into the gas phase, this method is less destructive than are al ternative methods of ionization, such as electron ionization (EI), chem ical ionization (CI) or fast atom bombardment (FAB), and well suited to the field of proteomics. Optimization of ESI conditions for maximum sensitivity has resulted in a dramatically decreased flow rate of sample to the ion source, from 4-200 L/min down to 10-500 nL/min,70, 71 with a concurrent decrease in chromatography column size.51 It was established that an ESI tip with a small diameter orifice produces a st able electrospray cu rrent, and thus strong generation of ions.72 The use of ESI usually results in the generation of multiply-charged ions, which facilitates further analysis by a pplication of fragmentation methods to the ions, resulting in specific sequence informa tion. The multiply charged ions are also a benefit for certain mass analyzers, such as i on traps, that have a limited mass to charge (m/z) range for ion detection. Using ESI, ions may be generated in positive or negative mode, as determined by the polarity of the applied potential between the ESI tip and the counter electrode. Most proteins and peptides, when analyzed by on-lin e LC/MS, are separated in HPLC solvents containing acid. The low pH of the solv ents produces a positive charge on the Nterminus and basic amino acid side-chains su ch as lysine and ar ginine. Negative ion

PAGE 40

21 mode is often used for DNA and oligomer studi es, but has been app lied to the detection of peptides and proteins contai ning phosphate or sulfate groups.73, 74 The research carried out in the subsequent chapters us ed ESI in positive mode, only. Finally, ESI is easily coupled to on-line LC /MS analysis of biomolecules. The low flow rates of current LC methods concentrate the analytes as they elute from the column in small volumes, causing an increase in signal from the mass detector. This enhancement, coupled with MS techniques to improve sensitivity and selectivity have resulted in detection of peptides present at the attomole level in complex sample mixtures.75 Matrix-assisted laser desorption/ionization An additional method of sample ionization that has been broadly applied to proteomic studies is matrix-assisted lase r desorption/ionizatio n (MALDI). This ionization technique is also well suited fo r biomolecules and is considered a “soft” ionization method, due in part to generation of few fragment ions. In MALDI, the sample is co-crystallized with an organic mo lecule, usually an acid that functions as a matrix for the analyte. The crystallized mi xture is irradiated with a laser under vacuum and the energy is absorbed by the matrix molecules, which then transfer energy to the analyte molecules forming ions with little sample decomposition.76 One of the major differences between ESI and MALDI is th at MALDI usually generates only singly charged ions. The generation of such ions, especially when anal yzing large molecules such as proteins, requires mass analyzers that are capable of resolving and detecting high m/z ions. Usually MALDI is coupled to timeof-flight (TOF) mass an alyzers, but recent developments in MS technology have resulted in successful joining of the MALDI source with ion trap mass spectrometers for analysis of peptides.77, 78

PAGE 41

22 Due to the nature of MALDI, it is not readily coupled with on-line LC/MS analysis. The majority of the experiments conducted fo r this body of research used ESI for the predominant method of ionization. Mass Analyzers A wide variety of mass analyzers are available for chemical, physical, and biochemical research; however, four mass an alyzers have been central to proteomic investigations: the quad rupole ion trap (QIT), triple qua drupole, time-of-flight (TOF), and Fourier transform ion cyclotron resona nce (FTICR) analyzers. Each of these analyzers has a different mode of mass sele ction, contributing to its use for specific proteomics applications. Selection of a singl e mass analyzer for a particular application must take into account the needs for mass accuracy, resolution, tandem MS capabilities and sensitivity. This section will address the basic principles of operation for each analyzer used in the studies desc ribed in chapters 3 through 7. Quadrupole ion trap The QIT is currently available in two forms in the field: a 3-dimensional ion trap and a 2-dimensional, or linear ion trap (LIT). The LIT is an improved version of the 3D trap, but the basic principles of operation are si milar. The 3D ion trap is composed of a hyperbolic ring electrode with two end-cap el ectrodes through which ions are injected into the trap and ejected upon de stabilization. Ions are trappe d within the center volume of the chamber by application of a radio-freque ncy (RF) applied to the ring electrode and a potential applied to the end cap electr odes (or they are set to ground), and by the introduction of helium, which dampens the ion motion through low-energy collisions to bring ions to the center of the trap.79, 80 The ions can be sequentially scanned out of the trap by changing the amplitude of the RF voltage applied to the ring electrode. Ramping

PAGE 42

23 the RF from low to high voltage causes a destabilization of ions within the trap, proportional to their m/z, and they ar e ejected through the end cap electrode.81 The ions are ejected sequentially so that the abundance for a particular m/z can be independently recognized by the detector, a feature of extreme importa nce for quantitation by MS. Another key feature of ion trap MS is the capability for ion fragmentation. Trapped ions can be selectively isolated by stored waveform inverse Fourier transform (SWIFT), which uses a series of waveforms supplied by the endcap electrodes to remove all but the selected m/z ion window from the trap. The is olated ion can then be either ejected for a selected ion monitoring (SIM) scan, or it can be fragmented by resonant excitation through the endcap electrodes. Fragment ions are then ejected sequentially and detected. Even though this type of mass analyzer is widely used in peptide and protein studies, including LC/MS analysis of complex mixtures, it has a number of drawbacks. The first is the limited ion capacity. When th e trap is overcrowded by ions, a significant deviation in m/z is observed, due to the “space charge” effect.82 These shortcomings have been improved upon in the LIT. Instead of a 3D trap, the LIT is constructed similar to a quadrupole, with 4 parallel rods with hyperbolic surfaces that are segmented into three sections, each held at di screte DC levels, which facilitate trapping of the ions.83 The increased ion trapping capacity greatly minimizes space charge effects and increases the sensitivity of the mass an alyzer. Also, radial ejection of the ions through a slit in opposite rods allows for dual detection, which increases sensitivity. With the improvements in scan rates and electronics, the duty cycle of the LIT is much faster than that of the 3D ion trap, resulting in a la rger number of ion scans per analysis.

PAGE 43

24 The QIT mass spectrometers are the least complicated, least expensive and most robust mass analyzers that provide sensitive on-line LC/MS detection, MSn fragmentation and are well suited to proteomics investigations . With development of the newer LIT, the shortcomings of the 3D QIT are being circ umvented, resulting in the LIT becoming the workhorse of proteomics laboratories. Triple quadrupole The triple quadrupole (TQ) mass analyzer, as the name implies, is a series of three quadrupoles that can be operated in a number of different modes for parent ion detection and structural determination of mixtures of analytes. The design of the TQ couples an ion source to a quadrupole mass filter, an RF-only quadrupole collision chamber, and another quadrupole mass filter.84 The first and third quadr upoles may be operated in mass selective mode, while the middle quadrup ole serves as a collision cell for ion fragmentation. Once generated, ions can be analyzed in several distinct modes.85 First, a single dimension mass spectrum may be produc ed by scanning the first quadrupole and passing the ions through quadrupoles 2 and 3 in RF mode. Second, a product-ion scan can be generated by selecting a certain m/z in the first quadru pole, fragmenting it in the second, and scanning the third. A precursor ion scan is accomplis hed by scanning the first quadrupole, fragmenting in the sec ond, and allowing only a fixed m/z to pass through the third. Finally, in a neutral loss scan, the first and th ird quadrupoles are scanned in a constant m/z offset. This mode is useful in studies of phosphorylation of peptides, in which the CID causes loss of th e phosphate group prior to peptide backbone fragmentation. The TQ mass analyzer is well designed for quantitation experiments. The selectivity demonstrated by isol ation of parent (precursor) and specific daughter (product)

PAGE 44

25 ions through scanning the firs t and third quadrupoles, respec tively, results in unparalleled mass selection and identification of characteri zed analytes in complex mixtures. While the capability of the 3D QIT to attempt thes e same modes of operati on may fall short, the 2D QIT has demonstrated similar selectiv ity and sensitivity, and when a TQ is unavailable, the 2D QIT may be a suitable re placement for quantita tive studies by MS. Time-of-flight The TOF mass analyzer has historically been coupled to MALDI sources. The TOF analyzer has demonstrated superior re solution capabilities, making measurements of a large range of m/z possible. In the case of MALDI-generat ed ions, a group of ions are produced and accelerated to a fixed kinetic en ergy by an electric potential. The ions travel through a field free region where they separate based on mass. The relationship between ion velocity and m/z of the ions is inversely proportional, resulting in lighter ions arriving at the detector first. The reso lution of the analyzer can be increased when it is operated in reflectron mode, in which the distribution in an ionÂ’s velocity, normally resulting in a broadening of the ion packet, is reduced by ap plication of an ion mirror in the flight path of the ions. Use of the ion mirror corrects for some of the broadening caused by the ionsÂ’ initial velo cities, and better resolution of the peaks is accomplished.86 MALDI-TOF analysis is rou tinely conducted on in-gel dige sts of single proteins or simple protein mixtures, as we ll as intact protein samples. MS/MS fragmentation using TOF instruments has only recently been accomplished through development of hybrid and coupled TOF-TOF instruments.87 Fourier transform ion cyclotron resonance The FTICR mass analyzer is well known for its ultra-high resolving power and high mass accuracy. These characteristics, inhe rent to its operational principles of

PAGE 45

26 detecting ions based on measurement of fre quency, make it the most accurate method of mass analysis available today.88 Often, accurate mass measurements of analyte mixtures is sufficient to identify the composition of the analytes without further dimensions of MS analysis. FTICR MS is also an ion trappi ng method, and isolates ions in the ICR cell between two axial electrodes (t rapping plates) for resonant ex citation by application of an RF signal. Ions are trapped radially by the presence of a high magnetic field. In the case of this research, magnets with 4.7 and 7 Tesla field strength were use d. The resolution of this method is proportional to the magnetic fi eld strength, and there has been movement for the use of higher strength magnets for proteomic and other studies.89 One significant attribute to the trapping mechanism of FTICR is that it is conducive to fragmentation by electron capture dissociation (ECD). EC D is a fragmentation method employing low energy electrons to dissociate peptide and protei n molecules. It is best used for sitespecific determination of labile post-transla tional modifications, su ch as phosphorylation and glycosylation, present on protein molecules. It is perhaps most commonly used in the investigation of “top dow n” proteomics, in which intact protein molecules are characterized and identified directly in the gas phase by MS analysis.90 A number of drawbacks to this analyzer pr ohibit its widespread use, as compared to the QIT. First, maintenance of a supe rconducting magnet is expensive and requires scheduled care. Second, the ultra-high v acuum region required by the detection cell necessitates a number of turbomolecular pumps and differential pumping regions, especially when incorporating an ion source at atmospheric pressure, such as ESI. Finally, MS/MS fragmentation employing collision gas is s lightly more time involved than CID in a QIT, in part because of the duty cycle required to co llect the ions in the

PAGE 46

27 cell, fragmentation with a collision gas, pumping down of the ce ll region, and detection of the ions. However, mass analysis using FTICR MS results in mass spectra with ultrahigh resolution, a tradeoff which may be worth the expe nse, depending on the specific application. Hybrid mass analyzers With the ever changing needs of the prot eomics researcher, mass analyzers have also been changing and deve loping to produce superior se nsitivity, resolution and mass accuracy. In order to accomplish this goa l, hybrid mass spectrometers have been developed, piecing several diffe rent mass analyzers togeth er for optimum detection parameters. Two of these hybrids are discussed below. The quadrupole time-of-flight (q-TOF) mass spectrometer couples the triplequadrupole mass filter with the TOF mass filter in order to combine the mass selection and fragmentation capabilities of the triple quadrupole with the high resolution of the TOF analyzer. In addition to resolution, the TOF analyzer provides additional sensitivity and an increased m/z of detection, as compared to a quadrupole or QIT.91 One drawback is that the duty cycle is considerably longer than that of a contemporary LIT, where the qTOF cycle for 1 parent ion scan and 2 daughter ion scans can take up to 7 seconds, while the LIT can complete 1 parent ion scan and 7 daughter ion scans in about one second. Once again, the application guide s the selection of the mass spectrometer, and scan time might not be as important as resolution, and in this case the q-TOF w ould be appropriate. The combination of the LIT with the FTI CR mass analyzer has revolutionized the use of FTICR mass spectrometry in the field of proteomics.92 The LIT has been coupled to the front end of the FTICR, affording an additional dimension of MS prior to the ultrahigh resolution detection. The LIT can ope rate as a mass analyzer and detector

PAGE 47

28 independently, but when ions collected in th e LIT are transferred to the FTICR cell, the accuracy of mass measurement and resolution greatly increase the quality of the data.93 Also, fragmentation of parent ions takes place in the LIT, removing the problem of gas load from CID fragmentation in the ICR cell. In chapter 7 of this dissertation, the LITFTICR (called the LTQ-FT, from Thermo Finni gan, San Jose, CA) is operated such that a single parent ion scan is accumulated in the ICR cell to generate accurate mass measurement of the parent ions, and 7 MS/M S events are carried out in the LIT to fragment the 7 most abundant parent ions for sequence identification. This combination of high throughput analysis, 2 ppm mass accu racy and resolving power of 500,000 make the LIT-FTICR a valuable instrument for proteomics analyses.92 Mass Detectors An important feature to consider in sel ection of mass detector s is whether or not quantitation of ion signal is important to th e experiment. In most proteomic studies, including those discussed in this dissertation, relative quant itation of analyte ions is secondary only to identification of the pe ptide samples. Quantitative analysis experiments must use a detector that will be able to produce a signal output equal in intensity to the number of ions being generated. For example, the 2D and 3D QITs sold by Thermo Finnigan use a conversion dynode a nd electron multiplier for ion detection. The dynode is maintained at -15 kV so that ions ejected from the trap strike the surface of the detector, generating secondary particles. The secondary particles are emitted toward the electron multiplier cathode, which generates an electron current for each ion scanned out of the trap. The gain of the electron multiplier is on the order of 105, with the final current proportional to the ion cu rrent impinging the conversion dynode.92 This method of ion detection is well suited for QITs.

PAGE 48

29 As an alternative the multichannel plate (M CP) detector is described as a 2D array of channels that can detect individual ions ar riving within a certain period of time. The gain is not as high as with an electron multip lier, but ions within a range of m/z values can be detected simultaneously, resulting in increased sensitivity. The disadvantage to using this detector is that only one arriving ion can be count ed per time period, resulting in an under-estimation of ion intensity if tw o or more ions strike the MCP within one counting period.92 This detector would not be ideal for use in quant itation experiments. Application of Technology As alluded to in the sections above, it is important to first ev aluate the biological system at hand and decide what question is to be answered. If mass accuracy and resolution are important, then TOF and FTI CR mass analyzers would be key instruments to use. If high throughput, MS/MS frag mentation, and sequen ce identification of complex samples mixtures is the focus, th en the QIT mass analyzers might be better suited to the experimental parameters. As ma ny graduate researchers find, it is often not possible to choose the ideal in strument for the selected ap plication, due to cost or availability. In that case, experimentati on through trial and erro r is often the best approach to generate the desired results, whet her it is increased sens itivity, mass accuracy or resolution. It is often th ese experimental and instrumental modifications that lead to development of new configurations of mass spectrometers, and if nothing else, a better understanding of the mass spectrometer itself.

PAGE 49

30 CHAPTER 3 CHARACTERIZATION OF RECOMBINANT HUMAN ASPARAGINE SYNTHETASE Introduction Asparagine synthetase is proving to have a significan t role in understanding the mechanism of L-asparaginase resistance in leukemia cells. A variety of methods have been developed to specifically de tect the presence of the enzyme,28, 29 the presence of the mRNA which translates the enzyme,16 and the synthetase activity in the cell.31-33 The known amino acid sequence of AS has been deduced following the isolation of human cDNA that encodes the enzyme.35, 94 This information does not take into account any post-translational modifications of the enzyme that might be necessary for its function, such as glycosylation, phosphorylation, or oxidation. Characterization of the primary structure of AS would not only verify the amino acid sequence from the cDNA data, but would also provide a reference for direct analys is of the protein. Such data could then be used for designing an assay for the direct de tection and quantitation of AS in complex sample matrices, allowing for analysis of ce ll lines or human samples. A new method of expression and purification of recombinant human AS (rhAS) from Sf9 insect cells was developed in the Richards lab at the University of Florida.34 With the availability of large amounts of rhAS, experiments were carried out to verify the amino acid sequence of the enzyme and to investigate the potential fo r post-translational m odifications using a combination of techniques including enzyma tic digestion, step-gradient separation, and electrospray ionization Four ier transform ion cyclotron resonance mass spectrometry

PAGE 50

31 (ESI-FTICR MS). The conditions for digestion were adjusted to explore the effect of enzyme to substrate ratio, denaturation of the protein, reduction and derivatization of cysteine (Cys) residues, and solution and in -gel digestion. Additional methods of mass spectrometric analyses, together with offline step gradient separation, on-line HPLC separation and tandem-MS detection were i nvestigated to optimize sequence coverage, and to provide data for future proteomics experiments. Experimental Procedure Expression and Purification of rhAS Expression of rhAS was carried out in Sf9 insect cells that were transfected with a virus encoding for the DNA sequence of the human enzyme with the addition of a Cterminal region containing a c-myc and a multi-histidine tag, as described.34 Development of the baculovirus vector a nd expression system was accomplished by M. Ciustea.95 The viral stock used to transfect the Sf9 insect cells (Invitr ogen) was previously prepared with a viral tite r concentration of 2 x 108 pfu (plaque forming units)/mL. Sf9 cells were grown in serum-free me dium to a cell density of 3 x 106 cells/mL and 95% viability. The 1 L cell culture was transfected with the viral stock at a multiplicity of infection (MOI) of unity. Th e cell density and viability were monitored every 12 hours, beginning 24 hours post infection. Sixty hour s post infection, cell viability was approximately 50%, and cells were harves ted by centrifugation. The cell pellet was resuspended in lysis buffer (50 mM EPPS, pH 8, 300 mM NaCl, 10 mM imidazole, 1% Triton X-100 and 0.5 mM DTT) and incubated on ice for 1 hour. The cells were lysed using three 15 second intervals of sonicat ion, and the cell debr is were removed by centrifugation at 19,000 x g. The supernatant, containing th e soluble protein fraction,

PAGE 51

32 was filtered through a 0.45 m membrane to clarify th e solution and remove any insoluble materials. Selective isolation of C-terminally tagged rhAS was accomplished by applying the filtrate to a Ni-NTA (n ickel nitrilo-triacetic acid) column, an immobilized metal affinity column charged with Ni2+ ions, which selectively binds histidine-rich regions of prot eins. The column was previous ly equilibrated with lysis buffer. Molecules with non-specific interacti ons between the filtered lysate and the resin were eluted using a wash buffer with an inte rmediate concentration of imidazole (50 mM EPPS, pH 8, 300 mM NaCl, 20 mM imidazole, and 0.5 mM DTT). In order to recover the C-terminally His-tagged protein, elution buffer containing highe r concentrations of imidazole (50 mM EPPS, pH 8, 300 mM NaCl , 250 mM imidazole, and 0.5 mM DTT) was applied to the column, a nd eluate was collected in 0.5 mL fractions. Each of the elution fractions was analyzed by absorbance at 280 nm and for specific enzymatic activity of AS. Active eluti on fractions with significant UV absorbance were pooled and subjected to precipitation usi ng a solution of 70% ammonium sulfate. The precipitate was pelleted by centrifugation and resuspended in storage buffer (50 mM EPPS, pH 8, 5 mM DTT). The sample was then dialyzed fo r 16 hours against the same buffer at 4ºC to remove any extraneous salts or contaminants from previous steps. Protein concentration was determined using the Bradford Assay (Pierce), consisting of a Coomassie dye reagent that binds to lysine, ar ginine, and aromatic residues.96 A calibration curve was prepared using bovine serum albumin (B SA) as the control, and absorbance measurements of dilutions of the final rhAS protein product were recorded and used to calculate concentration, which was 4 mg/mL. Fractions sampled from the lysis and purification steps of the pr ocedure were analyzed by SDS-PAGE to determine the

PAGE 52

33 approximate molecular weight of the isolated protein, and to determine the success of the purification procedure. Simple, In-Solution Digestion of Purified rhAS A 100 g aliquot of purified rhAS was diluted to a concentration of 1 mg/mL in 50 mM NH4HCO3, pH 8. Trypsin (Promega) was added at an enzyme to substrate (E:S) ratio of 1:20 (5 g trypsin), and the mixture was digest ed overnight (18 hours) at ambient temperature. The digest was quenched by addition of 5 L acetic acid. Denaturation, Reduction, Alkylation and Digestion Alkylation of the cysteine residues was investigated to determine the effect on peptide recovery. Using a 30,000 molecular weight cutoff membrane filter, 300 g of rhAS were exchanged from storage buffer to 6 M urea by rinsing four times with 200 L water, then recovering the protein solution in 25 L and combining with 75 L of 8 M urea, 50 mM EPPS, pH 7.8, 20 mM CaCl2, to denature the protein sample. A stock solution of dithiothreito l (DTT) was prepared in water to a concentration of 0.1 M, and 22.5 L of the stock was added to the denature d protein to reduce the disulfide bonds. Cysteine concentration wa s calculated to be 225 M, so the selected concentration of DTT was in excess at 18 mM. The sample was incubated for 40 minutes at 37ºC. Iodoacetamide and iodoacetic acid were prepar ed in separate solutions of 0.5 M. The denatured, reduced protein sample was split in half (61.25 L each), and 5 L of iodoacetamide was added to one tube, 5 L of iodoacetic acid was added to the other, for final alkylant concentration of 40 mM. The reactions were incubated in the dark at ambient temperature for 1 hour, a nd quenched by the addition of 10 L of 144 mM Lcysteine methyl ester to a final concentrati on of 18 mM. Both solutions were incubated

PAGE 53

34 for 15 minutes at ambient temperatur e in the dark, then diluted to 250 L with digestion buffer (50 mM Tris, 100 mM NaCl, pH 8) to re duce the urea concentration to less than 2 M. Trypsin was added at an E:S ratio of 1:10 (1.5 g trypsin added). The solution was digested overnight (18 hours) at ambient temperature and quenched by addition of acetic acid to a final concen tration of 5%. Investigation of Enzyme to Substrate Ratio Three 100 g aliquots of rhAS were digested with different E:S ratios of trypsin to determine the best ratio for digestion. Tr ypsin stock (1 mg/mL) was added in either g (1:20), 10 g (1:10), or 20 g (1:5) amounts, to all samples, which were diluted to 1 mg/mL with digestion buffer (50 mM Tris, 100 mM NaCl, pH 8) and incubated for 18 hours at ambient temperature. The sa mples were quenched by addition of 5 L acetic acid. SDS-PAGE Separation and In-gel Digestion Aliquots of purified rhAS (0.5, 1.0, 2.5, and 5 g) were resolved on a 12% acrylamide SDS-PAGE gel. Bovine serum al bumin (BSA), a well-characterized protein of the same approximate molecular mass, was analyzed as a control. After electrophoresis, the gel was stained with Coomassie blue stain for one hour, and destained with a solution of methanol/wat er/acetic acid (5:4:1) until the background of the gel was clear. The darkened bands around the 66 kDa marker were excised and placed in separate 0.6 mL microcentrifuge tubes. The in-gel digestion procedure is based on the procedure described by A. Schevenko et al .97 The gel bands were washed repeatedly with 200 L of 50% acetonitrile (v/v), 50 mM NH4HCO3, pH 8.3, to remove SDS and Coomassie stain. Gel slices were dehydrated with 100% acetonitrile, the wash

PAGE 54

35 solvent was removed and the gel pieces were allowed to dry in an exhaust hood. Samples were rehydrated and reduced by addition of 100 L 45 mM DTT in 250 mM NH4HCO3, pH 8 to each piece. Samples were incubate d for 30 minutes at 55º C, then liquid was removed and discarded. Alkylation was accomplished by incubation with 100 L of 100 mM iodoacetamide in 250 mM NH4HCO3, pH 8 in the dark at ambient temperature. Alkylation solution was removed and discarded, and gel pieces were washed twice with 100 L 50% acetonitrile, 50 mM NH4HCO3, pH 8.3. Following dehydration with 100 L 100% acetonitrile, 20 L of trypsin (12 ng/ L) was added to each tube, and the samples were incubated on ice while the trypsin solution was absorbed into the gel. After 45 minutes, excess trypsin solution was removed and replaced with 20 L 50 mM NH4HCO3, pH 8.3, then incubated at 37ºC for 18 hours. The digest was quenched by addition of 1 L acetic acid. Desalt of Digest Usin g Single-Step Elution For samples with less than 5 g of digested protein, C18 ZipTips (Millipore) were used for desalting prior to ESI-MS. The resin in the tip was wet with 20 L 50% acetonitrile, then equilibra ted three times with 20 L of 0.1% formic acid in water (v/v). The sample was applied to the resin bed by aspirating and dispensing 4-5 times. Salts and contaminants were removed by washing the resin with 10 L 0.1% formic acid in water (v/v), 4-5 times. Elution of the pep tides was achieved with a single solution of 50% acetonitrile (v/v), 0.1% formic acid in water (v/v). The solution was passed over the resin of the tip in 10 L increments to accumulate a total volume of 30 L. PepClean C18 Spin Columns (Pierce Biotechnologies, In c.) were used for samples containing more than 5 g of digested protein, up to 30 g. The procedure using the spin columns was

PAGE 55

36 very similar to that used for ZipTips, ex cept that a centrifuge was used to drive the sample and solutions through the resin bed in to a 1.5 mL microcentrifuge tube. Briefly, the resin was wet with 200 L 50% acetonitrile, then equilibrated with 200 L 5% acetonitrile (v/v), 0.5% formic acid (v/v) twice. The sample was diluted in a 1:3 ratio with 20% acetonitrile (v/v), 2% formic acid (v/v), so that th e final solution contained 5% acetonitrile, 0.5% formic acid in 150 L or less. The sample was applied to the column, centrifuged for 1 minute at 1,500 x g, and the el uate was reapplied twice. The column was washed three times with 200 L 5% acetonitrile, 0.5% formic acid, and the eluate was discarded. Elution of the peptides was accomplished by adding 70% acetonitrile, 0.5% formic acid in 20 L volumes twice. The total el ution volume was then reapplied to the column and centrifuged for maximum recovery of peptides. Desalt of Digest Using Mu lti-Step Gradient Elution Six elution buffers with increasing concentr ations of acetonitrile were prepared in order to produce a step-gradient that would preferentially elute peptides from the C18 resin based on hydrophobicity. The same desa lt procedure as described above for the ZipTips and PepClean Spin Columns was follo wed, except for the elution step. Instead of a single elution solution, six solu tions containing 10, 20, 30, 40, 50, and 70% acetonitrile (v/v) in 0.1% formic acid (v /v) were applied to the resin in 20 L volumes twice, in order of increasing percentage of acetonitrile. Elution fractions for each acetonitrile percentage were collected into separate microcentrifuge tubes. ESI-FTICR MS Analysis of Protein Digests Desalted peptide samples of rhAS were an alyzed by electrospray ionization-Fourier transform ion cyclotron resonance (ESI-FTI CR) mass spectrometry using a Bruker Apex

PAGE 56

37 47e FTICR MS with magnetic field strength of 4.7 Tesla, and equipped with a Modular ICR Data Acquisition System (MIDAS).98 Samples were directly infused into the ionization source region in 50% acetonitrile, 0.1% formic acid at a flow rate of 300 nL/min. The source was an Analytica of Bran ford electrospray ionization source, with a heated metal capillary and a hexapole tr apping region. Ions were produced by the application of a high voltage (+ 2.5 kV) to a metal union in contact with the sample solution. The charged solution was pumped to wards a heated metal capillary, maintained around + 140 V and 120°C. Ions were trapped in the hexapole region between a skimmer (+ 5 V) and a trapping electrode, usually ~ +8V, for approximately 0.5 to 1 second. The voltage on the trapping electrode was droppe d to -12 V for 1.5 msec, and ions were pulsed towards the ICR cell where they were trapped and excited by application of a chirp waveform. Detection was accomplished by collecting the induced current for 128 K data point transients prior to Fourier tr ansformation with accumulation of 100 scans. The resultant summed spectrum was processed using the Hanning method of apodization. MALDI-TOF MS Analysis of Protein Digests Desalted peptide samples to be analyzed by matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectro metry were prepared in a 1:10 ratio with -cyano-4-hydroxycinnaminic acid, and spotted in 1 L volumes on a MALDI target. The approximate concentrati on for each peptide, prior to dilution with matrix, was 1 x 10-6 M, assuming complete digestion and 100% recovery from the desalt step. MALDI-TOF MS analyses were conduc ted on a Bruker II Reflex, retrofitted with delayed extraction, operated in Reflectron mode.

PAGE 57

38 Results and Discussion The amino acid sequence of rhAS is show n in Figure 3.1. It is a 66.4 kDa protein with no known post-translationa l modifications. Recombinant human AS contains a consensus sequence for N-linked glycosylation, 542 Asn-Ala-Thr 544, but there is no evidence suggesting the site is occupie d. Recombinant protein expression in E. coli systems prohibits the possibili ty of the post-translationa l event of glycosylation. Therefore, the Sf9 insect cell expression system was selected for its ability to carry out such post-translational modifications.99 Expression and purification of the prot ein using immobilized metal affinity chromatography (IMAC) resulted in a rather pure sample of rhAS, as determined by SDS-PAGE analysis (Figure 3.2). The IMAC step selectively binds histidine-rich regions of proteins through the interaction of the histidines with immobilized nickel ions. Recombinant human AS was designed with a C-terminal region containing 6 histidine residues, specifically to assi st in the purification process.34, 95, 100 Free imidazole, similar to the functional group of histidine, is used at low concentrations in a wash step to disrupt any non-specific interactions between the column and other proteins with lower binding affinity. Once the nickel-bound protein is wash ed, a higher concentration of imidazole is used to release the protein, which is p ooled into a single fraction based on the UV absorbance and enzymatic activity of the elution fractions.34 Trypsin was chosen as the proteolytic en zyme for digestion of rhAS. Trypsin cleaves on the C-terminal side of basic lysi ne (K) and arginine (R) residues, producing peptides with positively charged side-chains in acidic solutions. Generation of such peptides greatly facilitates detection by MS because most peptides are already positively charged in the liquid sample.

PAGE 58

39 1 CGIWALFGSD DCLSVQCLSA MKIAHRGPDA FRFENVNGYT 40 41 NCCFGFHRLA VVDPLFGMQP IRVKKYPYLW LCYNGEIYNH 80 81 KKMQQHFEFE YQTKVDGEII LHLYDKGGIE QTICMLDGVF 120 121 AFVLLDTANK KVFLGRDTYG VRPLFKAMTE DGFLAVCSEA 160 161 KGLVTLKHSA TPFLKVEPFL PGHYEVLDLK PNGKVASVEM 200 201 VKYHHCRDEP LHALYDNVEK LFPGFEIETV KNNLRILFNN 240 241 AVKKRLMTDR RIGCLLSGGL DSSLVAATLL KQLKEAQVQY 280 281 PLQTFAIGME DSPDLLAARK VADHIGSEHY EVLFNSEEGI 320 321 QALDEVIFSL ETYDITTVRA SVGMYLISKY IRKNTDSVVI 360 361 FSGEGSDELT QGYIYFHKAP SPEKAEEESE RLLRELYLFD 400 401 VLRADRTTAA HGLELRVPFL DHRFSSYYLS LPPEMRIPKN 440 441 GIEKHLLRET FEDSNLIPKE ILWRPKEAFS DGITSVKNSW 480 481 FKILQEYVEH QVDDAMMANA AQKFPFNTPK TKEGYYYRQV 520 521 FERHYPGRAD WLSHYWMPKW INATDPSART LTHYKSAVKA 560 561 EQKLISEEDL LEHHHHHH 578 Figure 3.1 Amino acid sequence of rhAS. The c-myc and multi-His tag are at the Cterminus, residues 561-572 a nd 573-578, respectively. Figure 3.2. SDS-PAGE of expr ession and purification of rhAS. Lane 1: molecular weight markers; Lane 2: so luble protein from the cell lysate; Lane 3: total cell lysate after filtration; Lane 4: flow through from the IMAC column; Lane 5: wash from the IMAC column; Lane 6: blank; Lane 7: pooled IMAC elution fractions containing rhAS; Lane 8: rhAS after dialysis; Lane 9: molecular weight markers; Lane 10: blank, a nd Lane 11: total cell lysate. Fourier-transform ion cyclotron re sonance mass spectrometry (FTICR MS) provides the highest resolution and mass accuracy available for peptide studies.88 For the analysis of single protein dige sts, it isnÂ’t necessary to sepa rate the peptides prior to

PAGE 59

40 analysis because the resolution of the peaks is such that an accura te mass of each peptide ion can be calculated, and most peak s are well separated from each other.88, 101 Initial studies of rhAS by solution digestion and ESI-FTICR MS detection resulted in 62% sequence coverage, based on the molecular weig ht of the predicted peptides. Figures 3.3 and 3.4 are typical ESI-FTICR ma ss spectra from a sample of digested rhAS. Figure 3.3 demonstrates the full m/z range that was detected by the MS, and Figure 3.4 is an enlarged view of a small m/z window fr om the spectrum shown in Figure 3.3. The monoisotopic masses were manually calculated fro m the peaks detected in these spectra, based on their isotope spacing to determine the charge state, and the m/z for the monoisotopic peak. The peptide masses dete rmined from the spectra were compared against the theoretical masses of trypsin digested rhAS using the PAWS program (Proteometrics LLC). A number of different dige stion conditions were evaluated to determine which would provide the highest sequence covera ge of the protein. The experimental parameters adjusted were: denaturation of the protein followed by reduction and alkylation of Cys residues with either iodo acetic acid or iodoacetamide, enzyme to substrate (E:S) ratio, and in-gel digestion. Each of these steps, with the exception of adjusting the E:S ratio, adds considerable tim e to the sample preparation and digestion procedure. In each case, the digested protei n samples were desalted into a single elution fraction of 50% acetonitrile and 0.1% formic acid, and directly an alyzed by ESI-FTICR MS.

PAGE 60

41 5007009001100 m/z 0.0e+00 4.0e+06 8.0e+06 1.2e+07 1.6e+07 a.i. 5007009001100 m/z 5007009001100 m/z 5007009001100 m/z 0.0e+00 4.0e+06 8.0e+06 1.2e+07 1.6e+07 a.i. 0.0e+00 4.0e+06 8.0e+06 1.2e+07 1.6e+07 a.i. Figure 3.3. ESI-FTICR mass spectrum of trypsin digested rhAS. The data represent 100 accumulated scans with a sample flow rate of 300 nL/min. The m/z region underlined in red is enlarged and shown in Figure 3.4. Figure 3.4. Enlarged region of ESI-FTICR mass spectrum shown in Figure 3.3. The isotope distribution for each peptide is clearly visible, and the peaks are assigned with their corresponding peptide matches. Some peaks were identified as having sodium ions atta ched, indicating the desalt step might have been inefficient.

PAGE 61

42 Effect of Denaturation, Reduction and Alkylation of Cysteines Recombinant human AS contains 10 Cys re sidues in its sequence, including one at the N-terminus. When digested with trypsin, the cysteines are distri buted over 7 different peptides. Cysteines are labile amino acid residues that can form disulfide bonds with other cysteines (to form cystines) or free su lfhydryl groups in solu tion. These disulfide bonds may occur within one protein molecule , or between two separate molecules, creating dimers or larger multimers.102 Many researchers use deri vatization reagents that will permanently block the reactivity of cy steines, such as iodoacetic acid or iodoacetamide. The iodoacetyl group from these chemicals easily reacts with a sulfhydryl group, forming a carboxymet hyl or carboxyamido group on the Cys, respectively. The first experiments analyzing rhAS dige sted in solution without denaturants or derivatizing reagents resulted in detection of 6 out of 10 cysteines in 5 of the 7 predicted peptides. The largest peptide, 1Cys-Lys22, contains 3 cysteines, and the other peptide that was undetected was 107Gly-Lys130. The large size of these peptides could be a reason they were not detected in the initial experiments.103 Additionally, the low isoelectric point (pI) for each peptide may have prohibited adequate ionizati on in positive mode ESI. Identical amounts of rhAS were denatured to unfold the protein molecule, reduced with DTT, which breaks apart any disulfide bonds th at might be present, and alkylated with either iodoacetic acid or iodoacetamide prior to trypsin digestion. After desalting, the MS data revealed the addition of either the carboxymethyl or carboxyamido group to the Cys-containing peptides that were detected. However, the derivatization of cysteines with either reagent did not appear to ha ve a large effect on recovery of the Cys-

PAGE 62

43 containing peptides. A graphical repres entation comparing the amino acid sequence coverage is shown in Figure 3.5. Figure 3.5. Comparison of sequenc e coverage of rhAS based on E:S ratio and alkylation. Cysteine residues are depict ed as blue lines. Regions shown in red represent peptides verified by mass. Regions in white were not detected. For all conditions, the N-terminal peptide was not detected. Figure 3.6. ESI-FTICR mass spectrum of an op timized rhAS digest. The protein was digested with trypsin in a 1:20 ratio for 16 hours at ambient temperature in solution. This spectrum represents 71% sequence coverage of rhAS.

PAGE 63

44 Optimized conditions for rhAS solution dige st resulted in 83% sequence coverage. The ESI-FTICR mass spectrum shown in Figure 3.6 represents a sample of rhAS that was digested in a 1:20 E:S ratio with trypsin, in solution (50 mM Tris, pH 7.5, 100 mM NaCl) for 16 hours at ambient temperature. A porti on of the sample was desalted with a C18 ZipTip and eluted into one fraction of 50% acetonitrile, 0.1% formic acid (v/v) to an approximate concentration of 10 M. This preparation and analysis resulted in 71% sequence coverage. The E:S ratio was also examined to determine the best conditions for optimal sequence coverage. The addition of more tr ypsin (enzyme) resulted in more efficient digestion of rhAS (substrate), providing a larger number of peptides with lower molecular weight. A longer digestion pe riod usually provides optimal conditions for thorough digestion of the enzyme, and few misc leavages (missed sites where the trypsin should normally cleave). Alternatively, a smaller E:S ratio might actually promote miscleavages, resulting in large peptides with internal lysine or ar ginine residues. The results shown here (Figure 3.5) demonstrate that better peptide se quence coverage was achieved (71%) with the smallest E:S ratio investigated, 1:20. In-Gel Digestion of rhAS Several different quantities of rhAS and BSA were resolved by SDS-PAGE analysis in preparation for in-gel diges tion (Figure 3.7). The gel was stained with Coomassie blue to visualize th e protein bands for excision. The darkened rhAS and BSA protein bands were excised and washed to remove SDS and Coomassie blue dye. Coomassie blue will bind to the protein at lysine, arginine and aromatic amino acids, and protein presen t at a few hundred nanograms can usually be

PAGE 64

45 visualized.96 It is essential to remove the major ity of the dye, otherw ise trypsin is unable to cleave at lysines and argi nines that are coupled to dye molecules. The E:S ratio was not adjusted for this series of experiment s. Instead, the trypsin was added to the destained, dehydrated gel pieces at a concentration of 12 ng/mL. The volume absorbed by the gel was estimated to be about 25 L, resulting in 300 ng trypsin. This produced an E:S ratio between 1:3 and 1:16, depending on th e amount of rhAS in the gel. The E:S ratio for in-gel digests is us ually a little higher than for in-solution digests because the enzymeÂ’s activity and movement may be re stricted in the polyacrylamide gel. Figure 3.7 SDS-PAGE of rhAS and BSA sample s for in-gel digestion. Lane 1: molecular weight markers; Lanes 2, 7, a nd 10: blank; Lane 3: rhAS, 1 g; Lane 4: rhAS, 5 g; Lane 5: rhAS 0.5 g; Lane 6: rhAS, 2.5 g; Lane 8: BSA, 1 g; Lane 9: BSA, 5 g. The MS results obtained from in-gel dige stion were comparable to those obtained from in-solution digestion of rhAS, although there was slightly less sequence coverage.

PAGE 65

46 The digestion conditions that were determin ed in these experiments were used as a foundation for the quantitation method described in Chapter 4 of this thesis. The one adjustment that was made, to increase the sp eed and ease of sample preparation, was to omit the reduction and alkylation step in the in -gel digestion procedur e. Recovery of the Cys-containing peptides after in-gel digestion was poor, a nd the reduction and alkylation did not seem to improve the results (see Figure 3. 5), therefore it was omitted from the procedure. In addition, prior to loading th e protein samples onto the SDS-PAGE gel, the samples were boiled for 10 minutes in a de naturing buffer containing SDS and DTT, and it is assumed the proteins were sufficientl y reduced and unfolded to allow for maximum cleavage by trypsin. Investigation of Off-Line Desalt Conditions The above digest samples were all desa lted using either C18 ZipTips or C18 PepClean spin columns, with the peptides eluted into single fractions. Separation of peptides based on their hydrophobicity by re versed-phase HPLC is a commonly used method to decrease sample complexit y, minimize ion suppression, and increase sensitivity in the MS detector, when coupled on-line with ESI-MS.51, 104, 105 Without a suitable HPLC available for coupling to the ESI-FTICR MS in our lab, several attempts at off-line separation were pursued. The mo st successful off-line separation technique was an acetonitrile step gradient elution of th e peptides from either the ZipTips or C18 spin columns. ESI-FTICR mass spectra for th e different elution fr actions are shown in Figure 3.8.

PAGE 66

47 Figure 3.8. ESI-FTICR mass spectra for off-lin e step gradient separation of rhAS solution digest. The spectra shown in Figure 3.8 demonstrate the capability of off-line step gradient separation to sufficiently resolve peptides in a single protein digest. There appears to be a trend in the m/z of peptides as a function of acetonitrile percentage. Closer inspection of the mass spectra revealed that not only were the ions of higher m/z in the higher percentages of acetonitrile, but they were of higher mass, as well (multiply-charged states were maintained even at high acetonitrile cont ent). This suggests that perhaps peptide mass and hydrophobicity are linked, or that there is an increa sed likelihood that larger peptides are more hydrophobic simply because they are comprised of more amino acids

PAGE 67

48 and have a higher chance of incorpora ting more hydrophobic residues than smaller peptides. In addition, this separation step improved the sequence coverage of rhAS by 20% (Figure 3.9), resulting in 91% coverage. It is therefore assumed that sample fractionation, whether off-line or on-line prior to MS detection increases the likelihood of complete sample ionization and detection in the mass analyzer.106 1 CGIWALFGSDDCLSVQCLSAMK IAHRGPDAFRFENVNGYT 40 41 NCCFGFHRLAVVDPLFGMQPIR VKK YPYLWLCYNGEIYNH 80 81 K KMQQHFEFEYQTK VDGEIILHLYDKGGIEQTICMLDGVF 120 121 AFVLLDTANKKVFLGR DTYGVRPLFKAMTEDGFLAVCSEA 160 161 KGLVTLKHSATPFLKVEPFLPGHYEVLDLKPNGKVASVEM 200 201 VKYHHCRDEPLHALYDNVEKLFPGFEIETVKNNLRILFNN 240 241 AVKK RLMTDRRIGCLLSGGLDSSLVAATLLK QLK EAQVQY 280 281 PLQTFAIGMEDSPDLLAARKVADHIGSEHYEVLFNSEEGI 320 321 QALDEVIFSLETYDITTVRASVGMYLISKYIRKNTDSVVI 360 361 FSGEGSDELTQGYIYFHKAPSPEKAEEESERLLRELYLFD 400 401 VLRADRTTAAHGLELRVPFLDHRFSSYYLSLPPEMR IPKN 440 441 GIEKHLLR ETFEDSNLIPKEILWRPKEAFSDGITSVKNSW 480 481 FKILQEYVEHQVDDAMMANAAQKFPFNTPK TK EGYYYR QV 520 521 FER HYPGRADWLSHYWMPKWINATDPSARTLTHYK SAVKA 560 561 EQK LISEEDLLEHHHHHH 578 Figure 3.9. Sequence coverage of rhAS by ES I-FTICR MS with and without off-line step gradient separation. Resi dues in red were detected using the simple desalt step into one fraction. Residues in gr een were detected only after off-line peptide separation using the step gradient. Residues in blue have not been conclusively detected. Analysis of rhAS Digests by Alternativ e Modes of Ionization and Mass Analyzers In an effort to determine if the N-term inal peptide would be detected using an alternative method of ionization, a sample of rhAS digested in solution with trypsin was desalted using the step gradient method and each fraction was analyzed by ESI-FTICR MS and MALDI-TOF MS. The sequence cove rage obtained for each instrument was compared, shown in Figure 3.10.

PAGE 68

49 Figure 3.10. Comparison of rhAS sequence c overage as detected by ESI-FTICR and MALDI-TOF MS. The areas shown in red represent peptides that were detected by MS. The areas in grey were not detected. The Venn diagram illustrates that majority of the sequence coverage was detected by both methods (66%), with both ESI-F TICR and MALDI-TOF contributing significant additional coverage. The results demonstrate that ESI-FTICR MS analysis of the samples resulted in a higher percentage of sequence verification ve rsus MALDI-TOF MS. The explanation for this is based on the resolving powers of each of the mass analyzers. MALDI-TOF MS, when operated in reflectron mode, is cap able of 1200-3000 resolution for peptides.86 ESI-FTICR MS is, however, an ultra-high reso lution technique, resu lting in resolution between 50,000 – 100,000, on average.56 Therefore, the resolution and mass accuracy capabilities of the FTICR mass an alyzer allow identification of a number of peptides that were not resolved well enough in the MALDI-TOF mass spectrum. In order to demonstrate the resolution cap abilities of the FTICR mass analyzer, a sample of the rhAS solution digest was an alyzed by ESI-FTICR MS at the National High Magnetic Field Laboratory at Fl orida State University. The FTICR used had a magnetic field strength of 9.4 Telsa, and has been previously described.107, 108 Two peptides in the

PAGE 69

50 rhAS amino acid sequence ha ve the same nominal mass of 1067 Da: 340A-K349, with a mass of 1067.568 Da, and 407T-R416, with a mass of 1067.572 Da. These two peptides have a difference in exact mass of only 4 mD a. In the doubly charged state, this would correspond to a m/z difference of 0.002 m/z un its between the two ions. The resolution required to adequately separate these peaks is equal to m/ m, were m is the mass of the peptide and m is the difference between the apex of each of the peaks to be resolved, and is calculated to be 267,400. The mass sp ectrum is shown in Figure 3.11. The peaks were resolved well enough to identify the appear ance of two peaks at th at m/z. The error in measurement of these ions was 0.9 ppm for peptide 407T-R416, and the error for peptide 340A-K349 was 3 ppm. These data truly represent the high resolution and mass accuracy possible with FTICR mass analyzers. Unfortunately, using even this method of mass analysis the elusive N-termin al peptide was not detected. Figure 3.11. ESI-FTICR mass spectrum of rhAS digest obtained with over 260 K resolution. The peptides shown above with the same nominal mass were resolved with mass accuracies of 3 ppm (340A-K349) and 0.9 ppm (409TR416).

PAGE 70

51 Conclusions The results of primary structure character ization of rhAS have resulted in 91% sequence verification, as determined by accu rate mass ESI-FTICR ma ss analysis of a solution digested sample separated using a si mple step gradient of acetonitrile. Sample derivatization did not appear to significantly affect the recovery of Cys-containing peptides. In-gel digestion also did not a ppear to increase the number of detected peptides. Detection by MALDI-TOF MS added an additional 9% of sequence coverage to the peptides detected by ESI-FTICR MS. If MALDI wa s used as the method of ionization with a mass analyzer capable of higher resolution, then it is likely it would provide a better improvement in sequence c overage. Finally, the use of a 9.4 Tesla FTICR mass spectrometer allowe d resolution of two peptides that would normally not be separated in a direct infusion mi xture of digested peptides. These results have provided experimental parameters as well as information on peptides that are robustly and routinely iden tified by a number of different methods of mass analysis. Additional experiments, discussed in Chapter 4, demonstrated the use of this characterization of rhAS peptides in th e development of a quantitative assay relying on MS detection.

PAGE 71

52 CHAPTER 4 METHOD DEVELOPMENT FOR THE QU ANTITATION OF ASPARAGINE SYNTHETASE USING MASS SPECTROMETRY Introduction Quantitative measurement of biochemical markers for disease is the basis for design of effective drug regimens. Detecti on and quantitation of a signature protein during the course of a disease and its treatment may provide researchers and clinicians a sensitive way to modify patient drug therapy, as well as to monitor the progression of the illness. For example, in acute lymphoblastic leukemia (ALL), drug-resistance is often accompanied by an increase in expression of asparagine synthetase (AS) by leukemic cells, according to a number of in vitro studies.1-5 This increase in AS is induced by either deprivation of asparagine from the cell medium, or by introduction of the chemotherapeutic agent L-asparaginase.1, 2 Results of these studies indicate that expression of AS by leukemia cells is suffici ent to induce resistance to L-asparaginase. This conclusion is valid because L-asparagina se depletes extracellula r asparagine, and the leukemic cells are able to produce the necessary levels of asparagine for survival. In addition, in vitro studies with U937 cells demonstrated that L-asparaginase sensitive cells can be made resistant by repeated exposure to increasing levels of sub-lethal doses of the drug.4 Interestingly, results from in vivo studies involving human leukemia patients have been less consistent, leading to much discussion as to the role of AS in L-asparaginase resistance and prognosis.6-11

PAGE 72

53 A number of methods have been developed for the detection and quantitation of AS in both in vitro and in vivo studies. Most recently, a re al-time quantitative polymerase chain reaction (RQ-PCR) method was desc ribed as an effective means for the quantitation of AS expressi on in both leukemia cell lines and samples from human patients.12 This method relies on isolation of mRNA from the sample, and through amplification of the AS mRNA and a control mRNA fo r glyceraldehyde-3-phosphate dehydrogenase (GAPDH), the resultant quantit ative value is a normalized expression value of the AS signal divided by that of the GAPDH signal. While this provides an adequate normalized value for AS expre ssion within a sample, it does not measure expression of the AS protein, but instead the amount of AS mRNA that is present. In addition, the RQ-PCR assay does not provide absolute quantitation of how much AS mRNA has been expressed, but instead a rela tive value, based on the internal control protein. To date, there has yet to be es tablished a strong corr elation between mRNA levels and protein expression.13 The presence of mRNA for AS does not definitively prove that the protein is presen t, so it is important to monitor either the protein or its activity in the sample to conclude that th e elevation of mRNA is significant to the biology of the system. Alternative methods for detection of AS in cells rely on immunostaining and Western blotting (antibody-based assays), and RNA microarrays. While antibody-based experiments can be highly selective in the detection of AS, the observed signal is not from the protein itself, but from a seconda ry antibody that is specific for the primary antibody, which is specific for AS. Detection of a signal does not necessarily prove the presence of AS, but instead a molecule to which the primary (or secondary) antibody can

PAGE 73

54 bind. While most antibodies are highly select ive for single antigens, alternate methods of sample identification are beneficial. With the disparity in reports of AS expre ssion in cell lines and patient samples, it is important to understand not only the limitations of the methods chosen, but also the critical differences between in vitro and in vivo studies. Not only are in vivo studies inherently more difficult to control, but a number of other factors might be responsible for the L-asparaginase resistant phenotype, including age, genetic subtypes and cell lineage.14, 15 In an effort to provide a method to inves tigate and establish a biologically relevant level of AS that produces the L-asparaginase resistant phenotype, the remainder of this chapter describes development and evaluati on of a mass spectrometry-based technique for the direct detection and qua ntitation of AS from complex mixtures, such as total cell lysates from cancer cell lines and periphe ral blood samples from human leukemia patients. Quantitation based on the addition of a heavy-isotope pep tide standard allows relative determination of the amount of protei n (in moles or grams) isolated from the sample, data that have yet to be provided by any other method, and may present suitable evidence to better understand th e role of AS in L-asparaginase resistance. Materials and Methods Evaluation of Recombinant Human Asparagine Synthetase for Robust Peptide Standards Purified recombinant human AS (100 µg) was diluted in digestion buffer (50 mM NH4HCO3, pH 8.4), and trypsin (Sigma) added such that the AS:trypsin ratio was 1:20 (w/w) (total volume 120 µL). Prot eolysis was carried out at 30 C for 18 h before the reaction was terminated by flash freezing in liquid N2. Desalting of the tryptic digest

PAGE 74

55 prior to MS detection was accomplished us ing C18 PepClean spin columns (Pierce Biotechnology, Inc.), and the peptide mixtur e was eluted in one fraction using 50% aqueous acetonitrile containing 0.1% formic aci d. Samples were analyzed on a 4.7 Tesla Bruker Apex 47e Fourier transform ion cy clotron resonance mass spectrometer (FTICR MS), using an electrospray ionization (ES I) source (0.3 µL/min) equipped with a MIDAS data acquisition system.16 Peptide masses were calculated based on the mono-isotopic peak and charge state of each ion cluster a nd compared to those expected for tryptic peptides derived from recombinant, C-terminally tagged, human AS.17 Peptides providing reproducible ionization and few labile amino acids were selected for use as heavy-isotope internal standards. Heavy-isotope Peptide Prep aration and Addition Heavy-isotope amino acids were purch ased from Cambridge Isotopes, Inc (Andover, MA). Fluorenylmethoxycarbonyl (Fmoc)-derivatized L-phenylalanine containing six 13C (99%) isotopes was incorporated into the following peptide: ET* F EDSNLIPK. Fmoc-derivatized L-proline containing five 13C (98%) and one 15N (98%) isotopes was incorporated into the following peptide: WINATD* P SAR. Heavyisotope and natural abundance is otope peptides (light-isotope peptides) were synthesized by Mr. Alfred Chung at the ICBR Protein Core , University of Florida, Gainesville, Florida. Synthesized peptides were pur ified by HPLC and the masses verified by MALDI-TOF MS and ESI-FTICR MS. Lyophilized peptides were prepared in ~1 mg/mL solutions in water and the concentratio ns verified by amino acid analysis. This type of analysis involves acid hydrolysis of the peptide followed by amino acid derivatization with phenylisothiocya nate to produce phenylthiocarbamyl amino acid derivatives. The derivatized amino acids fr om the sample are resolved by HPLC and

PAGE 75

56 compared to a standard for quantitation. Con centration of the peptide solution is based on amino acid recovery. Generation of the Response Curve for Each Peptide All peptides were serially dilu ted from stock solutions to 125 M, separately in water. The heavy-isotope peptide solutions were combined in a 1:1 molar ratio to a concentration of 1.25 M each (5 L each 125 M peptide stock and 490 L water, for a total of 500 L). This was repeated for the light-i sotope peptides. Combined peptide stock solutions were diluted to 500 nM, then 25 nM in water. Twel ve solutions with a fixed concentration of heavy-isotope peptides at 12.5 nM and varying concentrations of light-isotope peptides between 12.5 pM and 125 nM were prepared to generate a standard curve for LC/MS/MS analysis. All solutions were prepared with a total volume of 50 L: 25 L of 25 nM heavy-isotope peptide stock was combined with 500 nM, 25 nM, 2.5 nM or 0.25 nM stock solutions of light-isotope pe ptides and water. Th ree standard curves were prepared on separate days and analy zed by LC-ESI-MS/MS as described below. Three microliters of each sample were anal yzed, providing 37.5 fmol of heavy-isotope peptide and between 37.5 amol and 375 fmol light-isotope peptide on-column. LC-ESI-MS/MS analysis Samples were separated by on-line micro RPLC ( RPLC) ESI MS/MS. The end of a length of fused silica capillary, 75 m ID x 360 m OD (Polymicro Technologies, Phoenix, AZ) was pulled to a fine tip (5-7 m) using a butane torch. The capillary was slurry packed with Jupiter C18 reversed phase resin, with 5 m bead diameter and 200 Ã… pore size (Phenomenex, Torrance, CA) to a bed length of 10 cm, as previously described.18 Solvent flow was supplied by an Agile nt 1100 capillary LC system (Agilent

PAGE 76

57 Technologies, Palo Alto, CA). Samples were loaded using an Agilent HPLC autosampler, maintained at 4ºC. After loading the sample, the column was washed with 2% B for 30 minutes at a flow rate of 0.5 L/min, then the flow rate was decreased to 0.250 L/min before initiation of the gradient. The gradient was as follows: 2-40% B over 40 minutes, 40-98% B over 30 minutes, where mobile phase A was 0.1% formic acid in water (v/v) and mobile phase B was 0.1% formic aci d in acetonitrile (v/v). High voltage contact for electrospray ioniza tion was provided through a metal union connecting the microcapillary column to the LC pump. The Thermo LTQ mass spectrometer method was created for full-scan MS then full-scan MS/MS of 5 most intense ions (data dependent mode), except during the retention times of the target peptides. During method times 23-28 minut es, the LTQ method was programmed for full-scan MS, then full-scan MS/MS (150 – 900 m/z) of 566.12 m/z and 569.12 m/z, sequentially, the light and heavy doubly-ch arged ions of pept ide WINATDPSAR. During method times 28-33 minutes, the LTQ method was programmed for full-scan MS, then full-scan MS/MS (150 – 1100 m/z) of 647.00 m/z and 650.00 m/z, sequentially, the light and heavy doubly-charge d ions of ETFEDSNLIPK. The retention time windows were based on an initial injec tion of the heavy-isotope peptides to determine where they eluted. The parent ion isolat ion width in full-scan MS/MS mode was set for 2.5 m/z with collision energy at 35%, and q = 0.250. MS Data Analysis A reconstructed ion chromatogram (RIC) for each peptide was generated using Xcaliber software (Thermo Electron, San Jose, CA). The three most abundant and consistent fragment ions for each parent peptide ion were chosen based on MS/MS

PAGE 77

58 fragmentation data of each pur ified peptide. Reconstructe d ion chromatograms (RICs) for each peptide were generated by plotting the sum of the three selected fragment ions versus time using an MS/MS filter for the pa rent ion, resulting in a peak area for each light and heavy-isotope pep tide, and all peak areas were recorded manually in a spreadsheet file. Additional RICs were ge nerated for each separate fragment ion, and those peak areas were also recorded. The ra tio of the MS peak area for each light and heavy peptide pair was plotted versus the ratio of the light a nd heavy peptide amount analyzed. The data were fit using a line f unction, and regression anal ysis of the data was carried out to determine the error in the slope of the line and the y-intercept. Cell Lysis and Desalt The human acute lymphoblastic leukem ia cell line MOLT-4 (ATCC CRL 1582) was propagated in RPMI-1640 medium supplem ented with 10% (v/v) fetal bovine serum (FBS), 10 mL/L ABAM ( 100 U/mL penicillin, 100 g/mL streptomycin, 0.25 g/mL amphotericin B) (GIBCO, Gaithersburg, MD) and 30 g/mL gentamycin (Sigma, St. Louis, MO) as previously described.2 All suspension cultures were maintained at 37 °C in a 5% CO2 incubator (Nuaire, Plymouth MN ). Twenty-four hours before all experiments, cells were collected by centrif ugation for 5 min at 288 x g, rinsed once with phosphate buffered saline PBS (0.15 M sodium , chloride, 10 mM sodium phosphate, pH 7.4), and resuspended at a dens ity of approximately 5 x 105 cells/mL in fresh medium. MOLT-4 parental (MOLT-4 S) cells were maintained in RPMI-1640 medium without any asparaginase (ASNase). MOLT-4 resistan t (MOLT-4 R) cells were maintained in RPMI-1640 medium containing 1 Unit/mL ASNase (Merck, West Point, PA). MOLT-4 S and MOLT-4 R cells were collected by cen trifugation for 5 minutes at 288x g, rinsed

PAGE 78

59 twice with PBS, and re-suspended in 1 mL lysis buffer (10 mM HEPES, pH 7.9, 10 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 20%glylcerol, 1 mM DTT, 1 mM PMSF, Protease Inhibitor Cocktail(Roche)) by ge ntly pipetting up and down several times. Samples were transferred to 2 mL microcen trifuge tubes and incubated on ice for 15 min. Triton-X 100 was added to a final concentration of 1% (100 L of 10% solution), and samples were vortexed for 5 seconds. Centrif ugation at 500 x g for 5 min at 4 °C provided the cytosolic protein fraction in the supernatant. These protein samples were provided by Nan Su and Dr. Michael Kilberg, Department of Biochemi stry, University of Florida (Gainesville, FL). Additional MOLT-4 S and R cells were ly sed as described, except the original lysis buffer contained 1% Triton X-100, and the so luble protein fraction collected after centrifugation was considered the to tal soluble protein fraction. Total protein extracts from the following cell lines were provided by Dr. Mi Zhou and Dr. Stephen Hunger, University of Florid a, College of Medicine (Gainesville, FL): K562, Jurkat, Nalm6, REH, RCH-ACV, MOLT-4 S and R. Protein samples were precipitated using 20% trichl oroacetic acid and cold acetone to remove salts and detergents. Proteins were re-dissolved in 50 mM NH4HCO3, pH 8.3. Protein concentration was determined using th e bicinchoninic (BCA) assay (Pierce Biotechnologies, Inc, Rockford, IL). Samples were diluted 1:5 (5 L sample, 25 L water) with water and combined with the BCA working reagent in a 1:1 ratio (10 L each sample and working reagent). A standard cu rve was prepared as described using BSA. The color was allowed to develop at 37ºC for 30 minutes. Absorbance was monitored at 562 nm using a ND-1000 nanodrop diode ar ray spectrophotometer (NanoDrop

PAGE 79

60 Technologies, Wilmington, DE). Sample c oncentration was calculated based on the calibration curve of BSA. Preparation of ALL Patient Samples Four samples of peripheral blood from pa tients diagnosed with ALL were provided by Dr. Alan Wayne, clinical director of the Pediatric Oncology Branch at the National Cancer Institute, Bethesda, MD. Thes e samples were designated M079880, M079837, M080788, and M082246, and were previously treat ed to recover the mononuclear cells using a ficoll-hypaque gradient, with the re covered blast cells counted and frozen for storage. In preparation for cell lysis, th e frozen cell samples were warmed in a 37°C water bath until almost completely thawed, th en washed twice in cold PBS. The PBS was removed using a pipette and replaced with 500 L lysis buffer (20 mM NH4HCO3, pH 8.3, 0.5% sodium dodecyl-sulfate (SDS), 5 mM tris-carboxyethyl phosphine (TCEP), 5 mM sodium orthovanadate (NaVO4), 10 mM sodium fluorid e (NaF), 1 mM EDTA, 25 mM glycerophoshate, and Protease Inhibitor Co cktail (Sigma-Aldri ch)) and incubated on ice for 15 minutes. The cell samples were lysed by 5 rounds of sonication, 15 seconds each at 12% amplitude, until homogenous. The samples were boiled for 10 minutes, and a small aliquot (10 L) was removed for protein concentration determination, as described previously. SDS-PAGE Separation Fifty micrograms of each sample were combined with SDS-reducing sample loading buffer and boiled for 10 minutes. After vortexing and centrifugation, samples were loaded in 20 L volumes onto NuPage 4-12% Bis-Tr is SDS-PAGE gels (Invitrogen, Carlsbad, CA). Gels were run at 60-100V pe r gel for over an hour, or until the dye-front

PAGE 80

61 reached the bottom of the gel. After electrophoresis, the gels were washed 5 minutes in de-ionized water, three times. Gels were stained with Simply Blue (Invitrogen) Coomassie stain until the protein lanes were visible. The gels were then washed in water to remove the background stain. Western Blot Analysis In the case of western blotting analysis, two identical gels were prepared for SDSPAGE. After SDS-PAGE separation, the gel was applied to a PVDF (polyvinyl difluoride) membrane (Invitroge n) and electrotransferred for 1 hour at 80 V using a mini Protean 3 Trans-Blot electrophor etic transfer apparatus (Bio -Rad, Hercules, CA). Nonspecific binding was blocked by incubating th e membrane in 5% non-fat dried milk in TBST (25 mM Tris-HCl, pH 8.0, 125 mM NaCl, 0.1% Tween 20) for 1 hour. The membrane was probed for one hour with -AS monoclonal anti body (University of Florida Hybridoma Core, Gainesville, FL) at a dilution of 1:100 in 5% non-fat dried milk in TBST. The membrane was rinsed 3 times for 5 minutes each in TBST prior to incubation with the secondary, horse radish peroxidase-conjugated goat-mouse antibody (Pierce, Rockford, IL) at a dilution of 1: 7500. The membrane was rinsed again 3 times for 5 minutes each in TBST prior to exposur e with enhanced chemiluminescence reagent (SuperSignal West Pico Substrate, Pierce, Rockford, IL). The peroxidase-conjugated antibody was visualized with x-ray film (C L-XPosure film, Pierce Biotechnologies, Inc) after treatment with SuperSignal West Pico chemiluminescent substrate (Pierce Biotechnologies, Inc), according to manufactur erÂ’s instructions. The x-ray film was photographed and the image was analyzed by UNSCAN-IT (Silk Scientific Corporation) to determine the pixel intens ity of each of the bands.

PAGE 81

62 In the case of the ALL patient samples, the antibody concentration and incubation times were increased to improve detection of AS: the -AS monoclonal antibody was prepared at a dilution of 1:100 and the membra ne was incubated overnight (12-14 hours). The secondary antibody was diluted 1:5000 and in cubated with the membrane for 1 hour. After treatment with chemilu minescent substrate, the x-ray film was exposed to the membrane for 5 minutes. Longer exposure times did not appear to affe ct signal intensity. In order to verify equal sample lo ading, the membrane was probed with -Actin at a dilution of 1:10,000 for 1 hour and re-expos ed to the secondary antibody (1:20,000 dilution) for 1 hour. Chemiluminescent signal wa s detected by film exposure to the blot for 1-2 seconds. In-gel Digestion and Heavy-Isotope Peptide Addition Protein bands were excised using a scalpel and placed in separate 600 L microcentrifuge tubes. Location of the ar eas of excision was based upon the molecular weight markers and a sample of purified recombinant human asparagine synthetase (rhAS) run in a separate lane, and was nor mally near the 64 kDa molecular weight marker. The rhAS sequence contains an extended C-terminal region and normally migrates to a higher position on the SDS-PAGE gels than endogenous human asparagine synthetase. Therefore a gene rous region of the gel was excised between the 49 kDa and just above the 62 kDa molecula r weight markers (SeeBlue Plus2 markers, Invitrogen). Gel bands were destained in two volumes of 500 L of 25 mM NH4HCO3, pH 8.3, 50% acetonitrile. The excised gel bands were c hopped into smaller pieces using a scalpel. Gel pieces were dehydrated by vortexing for 10 minutes in 100% ace tonitrile. Bulk acetonitrile was removed and samples were lyophilized for 20 minutes in a vacuum

PAGE 82

63 centrifuge. Trypsin was prepared to a final concentration of 20 ng/ L in 25 mM NH4HCO3, pH 8.3 (Promega) and added in 40-50 L volumes to the dehydrated gel pieces, then incubated on ice for 1 hour until the gel pieces were fully rehydrated. Excess trypsin solution was removed and replaced with 40 L of combined heavy-isotope peptide standards (2.5 nM of WINATD *P SAR and ET *F EDSNLIPK) diluted in 25 mM NH4HCO3, pH 8.3, resulting in 100 fmol of each h eavy-isotope peptide. Samples were incubated overnight at 37ºC. Alternativel y, the heavy-isotope peptide standards were spiked in at higher or lower amounts (50 fmol or 150 fmol, each). Peptide solutions were prepared fresh for each experiment, starting from the 500 nM combined peptide stock, and diluted serially to 2.5 nM with 25 mM NH4HCO3, pH 8.3. In the case of the human ALL patient sa mples, whenever possible, the SDS-PAGE gel was aligned with a corres ponding western blot to target the excision of the AS band, which was not visible by Coomassie stain, a nd to minimize excess gel in the sample. Extraction and Analysis of Digested Peptides Trypsin digestion was quenched by addition of 50 L of 70% acetonitrile, 5% formic acid. Gel samples were sonicated for 10 minutes, and the solution removed to fresh tubes. Extraction was repeated twice more. Peptide extract was lyophilized by vacuum centrifugation. Samples were rehydrated with 20 L 0.1% TFA and vortexed for 15 minutes. Samples were desalted using C18 ZipTips (Millipore). After initial desalt, the unbound portion of samples was re-subjected to fresh C18 Zi pTips. The eluates were combined and lyophilized, then re-dissolved in 15 L of 0.1% TFA in preparation for RPLC-ESI-MS/MS analysis (as described a bove). Seven microliters of each sample

PAGE 83

64 were analyzed twice by RPLC-ESI-MS/MS. Data interpretation was carried out as described above. Results and Discussion The use of heavy-isotope internal standa rds has long been recognized as a suitable method for quantitation by MS detection.19-22 Over the past decade, isotope dilution mass spectrometry (IDMS) has been explored for the quantitation of peptides and proteins from complex sample matrices.23-28 In the majority of these experiments, a peptide sequence is selected from the target protein of interest, and a heavy-isotope version of the peptide, through enrichment of 13C or 15N, is synthesized. The heavyisotope peptide is spiked into the sample in a known amount and digested with trypsin. The resultant mixture is then analyzed by LC/MS/MS and the parent ion mass of the natural-abundance peptide and the heavy-isotope standard ar e specifically targeted for fragmentation. In some cases, either due to the comple xity of the sample or the very low abundance of the target protein, the sample s are enriched prior to digestion by SDSPAGE,26 size-exclusion chromatography27 or antibody-mediated isolation.28 While sample recovery during these steps may be diminished,28 the reduction in overall sample complexity usually results in enhanced analyte detection. Most IDMS studies employ a triple qu adrupole mass spectrometer (TQMS) for fragmentation and detection of the target parent and fragment ions. The TQMS is capable of selected reaction monitoring (S RM) and multiple reaction monitoring (MRM), methods by which a selected parent ion is ex clusively isolated for fragmentation in the second quadrupole, and only select ed fragment ions are passe d into the third quadrupole

PAGE 84

65 for mass analysis.29 The TQMS is known for its selectiv ity and sensitivity, related to the efficiency of transferring ions from one quadrupole to the next and its ability to efficiently contain ions dur ing the fragmentation step.29, 30 Recently, a linear ion trap (LIT) mass spect rometer was introduced that exhibited increased ion storage capacity and altered de tector configuration to allow for improved ion detection.31 This mass spectrometer was used for quantitative LC-MS/MS analysis of AS from the cellular lysate of cancer cell lines. The resu lts described in this chapter demonstrate the reproducibility, selectivity a nd sensitivity of the measurements of the LIT MS for quantitative analyses of peptides using the IDMS methodology. Selection of Peptides for Us e as Heavy-Isotope Standards Purified rhAS was digested with trypsin and analyzed by ESI-FTICR MS for selection of suitable peptides for use as intern al standards. The desalted protein digest was directly analyzed without chromatographic separation of peptides in order to create a sample mixture of some complexity (Figur e 4.1). Additionally, a sample of MOLT-4 R cytosolic proteins was analyzed by SD S-PAGE and the region around the 62 kDa molecular weight marker was excised, in-gel digested, and analyzed by LC-ESIMS/MS. The data were searched against the human protein database using SEQUEST,32 and AS was identified by a number of peptid es, three of which corresponded to stronglyionizing peptides in the digest sample analyzed by ESI-FTICR MS. The peptides selected for use as heavy-isotope standards were 540WINATDPSAR549 and 467ETFEDSNLIPK477, based on the amino acid sequence of human AS (accession number P08243). These peptides were chosen because of their adequate size (between 10 and 15 amino acids), their ability to ionize and be detected by MS in a somewhat complex

PAGE 85

66 mixture, and the absence of labile amino aci ds such as methionine and cysteine. The additional peptide shown in Figure 4.1, E395R403 was not pursued for further study because of a potential site of degradation in its sequence. The heavy-isotope amino acids proline and phenylalanine were selected based on the incorp oration of six heavy-isotope atoms (13C and 15N) per amino acid, and the availability of an N-terminal Fmoc group for protection during pep tide synthesis. Figure 4.1. ESI-FTICR MS spectrum of trypsin digested rhAS. Spectrum was obtained by direct infusion of digested rhAS at 300 nL/min with 128K data collection, and is an accumulation of 100 scans. Pe ptides labeled in red were chosen as strongly ionizing, non-labile peptides that were composed of fewer than 15 amino acids and were also identified in LC-ESI-MS/MS analysis of MOLT4 R samples. Peptide E395-R403 was not used as a standard based on a potential site of degrada tion in its sequence. Generation of the Response Curve for St andard Peptides of rhAS by LC/MS/MS A standard curve was prepared for each of the peptides selected for use as a heavyisotope standard. Fragmentation conditions were optimized to provide consistent fragmentation patterns of each of the heavy and light isotope peptide pairs. Three of the

PAGE 86

67 most abundant fragment ions from the MS/M S spectrum for each peptide were selected for quantitation using mass chromatograms constructed from the MS/MS fragment data. The fragment ions chosen for the WINATDPSAR peptide pair were y4, y6 and y8 (Figure 4.2). The fragment ions chosen for the ETFEDSNLIPK peptide were y2, b8-H2O, and b9H2O (Figure 4.3). Twelve solutions containing 12.5 nM each of WINATD*PSAR and ET*FEDSNLIPK peptides and varying concentra tions of the natural-abundance form of each peptide (12.5 pM to 125 nM) were analyzed by LC/MS/MS to determine the lower limit of detection, lower limit of quantitation, and the linear response range of each peptide. Standard curves were prepared in triplicate from three independent sample dilutions and analyzed by LC-ESI-MS/MS. Full-scan MS/MS of each target peptide ion pair during their elution times provided 15-25 data points across each peak for generation of a RIC based on three major fr agment ions (Figures 4.4-4.5). The ratio of the peak area in the mass chromatograms for each light and heavy peptide pair was plotted against the ratio of light vs heavy peptide amount anal yzed for each run (Figures 4.6-4.7), using a linear fit. Regression analysis was carried out to determine the error in the slope of the curves and the y-intercepts. In order to de termine the lower limit of detection (LLOD) for each peptide, peak area values for th e natural abundance isotope forms of each peptide were recorded from 12 blank analyses. The average (Sbl) and standard deviation (sbl) of the blank measurements were calcula ted and incorporated into equation 1. SLLOD = Sbl + 3(sbl) (1) where SLLOD is the signal required to satisfy the LLOD. The constant 3 in equation 1 represents a confidence level of detecti on of 89% or greater, and is the value

PAGE 87

68 recommended in literature.33, 34 This value is then incorpor ated into equation 2, in which the value for m was experimentally determined from the calibration curve for each peptide (data not shown). The slope of each calibration curve, m, represents the sensitivity of the detection method and CLLOD represents the concentration of the LLOD. CLLOD = (SLLOD – Sbl)/m (2) The LLOD for ETFEDSNLIPK was found to be 30 amol, while the LLOD for WINATDPSAR was found to be 100 amol. Values for the lower limit of quantitation (CLLOQ) were calculated as ten times the standard deviation of the blank measurement, sbl (equation 3). CLLOQ = (10 x sbl)/m (3) The lower limit of quantitation for peptide ETFEDSNLIPK was determined to be 100 amol, while for WINATDPSAR is was 350 amol. These figures of merit are summarized in Table 4.1. It must be empha sized that these figures of merit were established with purified synthetic peptides dissolved in water. The addition of a complex sample matrix may alter the lower li mits of detection and quantitation with the possibility of contaminant analytes. However, it is important to note that the linear ion trap mass detector is capable of three orde rs of magnitude linearity, and sub-femtomole limits of detection in this study. Table 4.1. Figures of Merit for Peptides. WINATDPSAR ETFEDSNLIPK Linear Range 350 amol – 400 fmol 100 amol – 200 fmol Limit of Detection 100 amol 30 amol Lower Limit of Quantitation 350 amol 100 amol

PAGE 88

69 Figure 4.2. Full scan MS/MS spectra of peptides WINATDPSAR (top panel) and WINATD*PSAR (bottom panel). Fragment ions matching the theoretical fragment ions for the proposed peptide sequence are shown in red. Note the 6 m/z unit shift in many of the fragment ions, reflecting the position of the heavy-isotope labeled proline (*P). Each MS/MS spectrum represents 37.5 fmol of the peptide.

PAGE 89

70 Figure 4.3. Full scan MS/MS spectra of peptides ETFEDSNLIPK (top panel) and ET*FEDSNLIPK (bottom panel). Fragment ions matching the theoretical fragment ions for the proposed peptide sequence are shown in red. Note the 6 m/z unit shift in many of the fragment ions, reflecting the position of the heavy-isotope labeled phenylalanine (*F) . Each MS/MS spectrum represents 37.5 fmol of the peptide.

PAGE 90

71 Figure 4.4. Ion chromatograms for 1:1 ratio of light and heavy isotope peptides. Panel A is the base peak chromatogram. Panel B is a mass chromatogram for the 3 fragment ions of peptid e WINATDPSAR, resulting from fragmentation of m/z = 566.12. Panel C is a mass chromatogram for the 3 fragment ions of WINATD*PSAR, resulting from frag mentation of m/z = 569.12. Both peptides were present at approximately 37.5 fmol. Use of mass chromatograms enhances selectivity and clearly indicates re tention times of peptides. The MS method was programme d to only fragment the light and heavy peptide ions during their elution time (23-28 minutes). The absence of ion signal in other areas of chromatogram s B and C is due to a lack of MS/MS data for those parent ions.

PAGE 91

72 Figure 4.5. Ion chromatograms for a 1:1 ratio of light and heavy isot ope peptides. Panel A is the base peak chromatogram. Panel B is a mass chromatogram for the 3 fragment ions of peptide ETFEDSNLIPK , resulting from fragmentation of m/z = 647.0. Panel C is a mass chromatogram for the 3 fragment ions of ETFEDSNLI*PK, resulting from fr agmentation of m/z = 650.0. Both peptides were present at approxima tely 37.5 fmol. The MS method was programmed to only fragment the light and heavy peptide ions during its elution time (23-28 minutes). The abse nce of ion signal in other areas of chromatograms B and C is due to a lack of MS/MS data for those parent ions.

PAGE 92

73 Figure 4.6. Graphical representati on of the ratio of MS res ponse of light and heavy (L:H) WINATDPSAR peptide versus ratio of pe ptide amount (L:H). The data are an average of 3 separate dilutions, and the error bars represent the standard deviation. The line equation and errors in slope and y-intercept are shown. Figure 4.7. Graphical representati on of the ratio of MS res ponse of light and heavy (L:H) ETFEDSNLIPK peptide versus ratio of peptide amount (L:H). The data are an average of 3 separate dilutions, and the error bars represent the standard deviation. The line equation and errors in slope and y-intercept are shown.

PAGE 93

74 Investigation of Protein Extracts from Cancer Cell Lines The LC-ESI-MS/MS method was applied to several protein samples isolated from cancer cell lines. Fifty micrograms of total protein extract from MOLT-4 S, MOLT-4 R, K562, REH, Jurkat, Nalm6, and RCH-ACV were resolved by SDS-PAGE, and compared to a standard of purified rhAS (Figure 4.8). Figure 4.8. SDS-PAGE gel of total protein ly sate from 7 differen t cancer cell lines, before (panel A) and after (panel B) excision of the band where AS migrates. Lane 1: purified rhAS, 1 g; Lane 2: blank; Lane 3: 50 g MOLT-4 S; Lane 4: 50 g MOLT-4 R, Lane 5: 50 g K562; Lane 6: 50 g Jurkat; Lane 7: 50 g RCH-ACV; Lane 8: 50 g Nalm6; Lane 9: 50 g REH. Analysis of the gel slices using the de scribed method resulted in detection and quantitation of AS, as summa rized in Table 4.2. The uncer tainty in these values represents a 95% confidence interval. Table 4.2. Quantitation of AS from 7 Cancer Cell Lines Gel Lane Cell Line Moles Detected Grams of AS in Sample % of AS in Sample 3 MOLT-4 S 1.1 ± 0.7 fmol 150 ± 100 pg 0.0003 ± 0.0002% 4 MOLT-4 R 170 ± 100 fmol 20 ± 10 ng 0.05 ± 0.03% 5 K562 40 ± 15 fmol 5 ± 2 ng 0.010 ± 0.004% 6 Jurkat 6 ± 3 fmol 800 ± 400 pg 0.0020 ± 0.0006% 7 RCH-ACV 4 ± 3 fmol 600 ± 400 pg 0.0012 ± 0.0006% 8 Nalm6 900 ± 400 amol 130 ± 60 pg 0.0003 ± 0.0001% 9 REH 4.3 ± 3.2 fmol 599 ± 440 pg 0.0012 ± 0.0008%

PAGE 94

75 The values shown in Table 4.2 demonstrate th at each of the cell lines analyzed was shown to have AS present within the limits of detection and quantitation of the method (30 and 100 amol, respectively). Figure 4.9 graphically illustrate s the moles of AS present in each cell line, based on analysis of 50 g of total protein for each sample. Error bars represent 95% confidence intervals. Figure 4.9. Graphical representa tion of MS quantitation of AS in 7 different cancer cell lines. The detected amount of AS is plotted on the y-axis, with error bars representing a 95% confid ence interval. The inset provides a closer look at the cell lines containing lower levels of AS. These data from these seven cell lines shown in Table 4.2 and Figure 4.9 were compared using the StudentÂ’s t-test to dete rmine if each cell line contained significantly different levels of AS. The results indicate that all but the REH cell line can be divided into 4 distint groups: MOLT-4 R, K562, Jurk at/RCH, and MOLT-4S/Nalm6 (listed in order of decreasing AS measured). The e rror in replicate measurements of the REH sample caused it to significantly overlap with the measurements from the Jurkat/RCH and

PAGE 95

76 the MOLT-4 S/Nalm6 groups, which were found to be significantly different from each other. Western blot analysis was carried out on a similarly prepared ge l as that shown in Figure 4.8, with the exception th at the amount of purified rhAS loaded as a control was 5 ng instead of 1 g (Figure 4.10). The Western blot was probed with the antibody for AS, and the darkened bands correspond to the pres ence of the protein. In Figure 4.10, Lane 1, the rhAS sample is shifted upward from the endogenous hAS detected in the cell line samples because it contains a C-terminal c-myc and multi-His tag, causing its molecular weight to be about 2 kD higher th an that of the endogenous protein.17 Figure 4.10. Western blot of SD S-PAGE gel similar to that shown in Figure 4.8. Lane 1: rhAS, 5 ng; Lane 2: blan k; Lane 3: MOLT-4 S, 50 g (total protein load); Lane 4: MOLT-4 R, 50 g; Lane 5: K562, 50 g; Lane 6: Jurkat, 50 g; Lane 7: RCH-ACV, 50 g; Lane 8: Nalm6, 50 g; Lane 9: REH, 50 g. The membrane was probed with mouse-AS primary antibody at a dilution of 1:100, then with goat -mouse secondary antibody, coupled to horseradish peroxidase. After introduction to EC L reagents, the chemiluminescent signal was detected by exposure to x-ray film for 30 seconds. When the results of the MS quantitation assay are compared to the Western blot data, the overall trends app ear to agree. MS data we re normalized to the sample

PAGE 96

77 containing the largest amount of AS, MOLT-4 R. Western blot data (two data sets) were also normalized to the MOLT-4 R value and th e results are plotted in Figure 4.11. Error bars represent the standard deviation in the measurement of the Western blot intensities, and the relative standard deviation of the normalized MS data. The MOLT-4 R and K562 samples were detected as having the larg est amount of AS by both MS and Western methods. In order to observ e the trends in the samples containing lower amounts of AS, the data were plotted without the MO LT-4 R and K562 data (Figure 4.12). Figure 4.11. Comparison of Wester n blot and MS quantitation data for the detection of AS in cancer cell lines. Both Western blot and MS data were normalized to the MOLT-4 R sample data, which was de termined to have the largest amount of AS by both methods. Error bars in the Western blot data represent the standard deviation of two measurements for each cell line. Error bars in the MS data represent the rela tive standard deviation of the normalized data.

PAGE 97

78 Figure 4.12. Comparison of Wester n blot and MS quantitation data with the omission of MOLT-4 R and K562 samples so that the lower level samples can be visualized. From Figure 4.11 and 4.12, it seems that the MS data have a similar trend to that of the Western blot data, when normalized to th e MOLT-4 R sample. While the data do not agree exactly, in most cases the error bars from each method overlap within each cell line. These data indicate th at the MS method of quantitati on is at least as good as data obtained by Western blotting. In addition, th e MS method provides an estimate of how much AS is present in each sample, whic h is not possible with Western blotting. Analysis of additional MO LT-4 S and R protein samples A separate set of MOLT-4 S and R protei ns were analyzed by Western blotting analysis and the quantitative MS method. The results of the SDS-PAGE gel and Western blot are shown in Figures 4.13 A and B.

PAGE 98

79 Figure 4.13. SDS-PAGE and West ern Blot of MOLT-4 S and R samples. Panel A: SDSPAGE of 50 g each MOLT-4 S (lanes 1-3) and MOLT-4 R (lanes 4-6) proteins, stained with Coomassie blue . Lanes 7, 8, and 9 contain 10, 28, and 100 ng of purified rhAS, respectively. Pa nel B: Western blot for the detection of AS. Lane assignments are the same as for the gel (panel A). No AS is detected in the MOLT-4 S samples by Western blot analysis. The Western blot analysis of the samples was unable to detect AS in the MOLT-4 S sample, although AS was detected in the MOLT-4 R and purified rhAS samples. The purified rhAS samples were loaded in such a low amount that they were below the level of detection for Coomassi e stain (Figure 4.13 A). Investigation of peptide internal standards During the analysis of the MOLT-4 S a nd R samples, there appeared to be significant discrepancy in values of AS calcula ted from one peptide st andard to the other, where the WINATDPSAR peptide was consistently 5-fold higher in the cell line samples than the ETFEDSLIPK peptide. In or der to determine what was causing this disagreement, the data were re-analyzed and instead of constructing one RIC for the three fragment ions of each peptide, a separate RIC was constructed for each fragment ion for each peptide and its corresponding internal st andard. Additionally, the data from the standard curves shown in Figures 4.6 and 4.7 we re re-analyzed in this same manner. The peak areas of the three fragment ions for each peptide were normalized to the ion with the

PAGE 99

80 smallest peak area for each data point: the y4, y6, and y8 ions of WINATDPSAR were all divided by the y8 peak area for each point; the y2, b8-H2O, and b9-H2O ions were divided by the b8-H2O peak area for each point. The re sults for the ETFEDSNLIPK peptide showed a y2:b8-H2O: b9-H2O ratio of 1.4:1:1.7, which was consistent with only 5% RSD for both the natural abundance and heavy-isot ope peptide, as long as the data were gathered in the linear range established for the method (Table 4.1). The WINATDPSAR peptideÂ’s y4:y6:y8 ratio was found to be 2.2:1:3.85 fo r the natural abundance and heavyisotope peptides analyzed in the standard cu rve. However, in the samples analyzed from cell lines, the ratio of the na tural abundance peptide, arising from endogenous AS in the sample was found to be significantly different from the established fragment ratio, and the y6 ion was consistently the fragment ion with the largest peak area, instead of the smallest (Table 4.3). This ch ange in fragmentation pattern of the peptide is most likely caused by co-isolation of an ion with similar m/z as the doubly-charged parent ion of the natural abundance peptide, 566.12 m/z. The likel y cause of the elevation in the peak area of fragment ion y6 is a contaminant ion that produces a fragment ion at the same m/z. There is no elevation in the y6 fragment i on of the heavy-isotope peptide, so it is probably not directly related to fragmentation of the peptide itself. The quantitative data generated from the WINATDPSAR peptide for the cell lines was therefore not used in calculating the amount of AS present in the samples. Table 4.3. Fragment Ion Ratios for Natural Abundance and Heavy Isotope Peptide WINATDPSAR Fragment ions (*= heavy isotope standard ions) y4 y6 y8 y4* y6* y8* Standard curve determination of fragment ion ratios 2.1 1 3.85 2.1 1 3.85 MOLT-4 S, n =9 0.065 1 0.10 1.9 1 3.85 MOLT-4 R n = 9 0.64 1 1.16 1.98 1 3.98

PAGE 100

81 The amount of AS was then calcula ted using only the data from the ETFEDSNLIPK peptide, which retained its fr agment ion ratio throughout all analyses. The data are summarized in Table 4.4. Table 4.4. Amounts of AS present in MO LT-4 S and R Cytosolic Protein Samples Detected by MS Sample Moles Detected Grams in Sample % of AS in Sample MOLT-4 S, n = 9 1.0 ± 0.2 fmol 140 pg ± 30 pg 0.00010 ± 0.00006% MOLT-4 R, n = 9 90 ± 20 fmol 12 ± 3 ng 0.024 ± 0.007% The values shown in Table 4.4 for the MOLT-4 S and R proteins differ slightly from the values shown in Table 4.2. Nearly a ll of these values fall within the calculated 95% confidence interval. The composition of th e samples was also different in that the samples reported in Table 4.2 were from a to tal protein fraction, and the samples reported in Table 4.4 were a cytosolic protein fraction. The cellular location of AS has not been reported. During the protein fractionation proc ess the nuclear pellets are removed from the soluble cytosolic proteins. It is possible that some cytosolic proteins might have been lost during this process. Alternatively, a popul ation of AS may be pr esent in the nucleus of the MOLT-4 cells, and removal of these pr oteins might decrease the overall quantity of AS recovered. Analysis of ALL Patient Samples The composition of the patient samples an alyzed in this study was nearly pure (100%) blast cells isolated from peripheral blood. An alternative source of sampling from an ALL patient is the bone marrow, wh ich is considerably more invasive and painful than a blood draw. The sample iden tification code, cell c ount and total protein recovery after lysis are shown in Table 4.5.

PAGE 101

82 Table 4.5. Sample Code, Cell Count and To tal Protein Recovery from ALL Patient Samples Sample Code Cell Count (cells/vial) Recovered Protein M079880 24.4 million 2.9 mg M079837 49.5 million 3.0 mg M080788 35.6 million 2.2 mg M082246 46.6 million 2.6 mg Two gels were prepared with the ALL pa tient samples, and with total protein fractions of MOLT-4 S and MOLT-4 R as controls , as well as a sample of rhAS. Fifty micrograms of each sample were loaded on the gel, except rhAS, in which only 5 ng were loaded. After electrophoresis, one gel was st ained to visualize the protein bands, while the other was electroblotted to a PVDF membrane for Western analysis. Once the Western blot results were obtained, the gel wa s overlaid on the blot film to specifically target the regions of the gel that had a co rresponding signal in the Western blot. The SDS-PAGE gel and Western blot are shown in Figure 4.14. Figure 4.14. SDS-PAGE and West ern Blot of ALL Patient Samples. Panel A is a Coomassie stained SDS-PAGE gel, while Panel B is the corresponding Western blot, probed for the presence of AS. Lane assignments for each are identical: Lane 1: M082246; Lane 2: M079837; Lane 3: M080788; Lane 4: M079880, Lane 5: blank; Lane 6: MOLT4 S; Lane 7: molecular weight markers; Lane 8: MOLT-4 R; Lane 9: blank; Lane 10: 5 ng rhAS.

PAGE 102

83 Based on results of the Western Blot show n in Figure 4.14, the levels of AS present in the patient samples are quite low, and vary among the samples. Additional bands appeared at lower molecular weight in the Western blot, indicating perhaps a smaller or truncated form of AS is present in the sa mples and reacts with the antibodies. These lower molecular weight bands were al so excised and analyzed using the RPLC/MS/MS method for quantitation of AS. Five additional samples for each patient were prepared on 4 separate SDS-PAGE gels, one set of 5 re plicates per gel with MOLT-4 S and R and 5 ng rhAS control lanes and the ~62 kDa ba nd was excised for in-gel digestion and quantitative analysis. Results are summarized in Tables 4.6 and 4.7. Table 4.6. Results of AS Quantita tion Using the WINATDPSAR Peptide Gel Lane Sample Code Moles AS Detected Moles AS in Sample Copies of AS per Cell 1 M082246 300 ± 100 amol 700 ± 200 amol 450 ± 120 2 M079837 1.0 ± 0.3 fmol 2.2 ± 0.6 fmol 1200 ± 300 3 M080788 260 ± 70 amol 600 ± 200 amol 400 ± 100 4 M079880 170 ± 50 amol 370 ± 120 amol 500 ± 200 6 MOLT-4 S 620 ± 200 amol 1.3 ± 0.4 fmol Not determined 8 MOLT-4 R 10 ± 2 fmol 20 ± 5 fmol Not determined Table 4.7. Results of AS Quantitati on Using the ETFEDSNLIPK Peptide Gel Lane Sample Code Moles AS Detected Moles AS in Sample Copies of AS per Cell 1 M082246 25 ± 8 amol 54 ± 17 amol 40 ± 10 2 M079837 170 ± 40 amol 360 ± 80 amol 200 ± 40 3 M080788 60 ± 20 amol 130 ± 50 amol 100 ± 30 4 M079880 30 ± 20 amol 60 ± 30 amol 100 ± 50 6 MOLT-4 S 240 ± 50 amol 500 ± 100 amol Not determined 8 MOLT-4 R 6 ± 2 fmol 12 ± 4 fmol Not determined The calculation for “Copies of AS per Cell” is as follows: where A is Avogadro’s number.

PAGE 103

84 There are several aspects of this method that must be addressed in order to properly interpret the data. First, tw o peptides from AS were used for quantitation, but the data from each peptide provided different quantitative results, with the WINATDPSAR peptide presenting about a 5-fold higher esti mation of protein. These results are similar to what was found upon initial analyses of the MOLT-4 S and R cell lines, however, in the ALL patient samples, the fragmentation ion peak ratios for the WINATDPSAR fragment ions are more consistent with that of the heavy-isotope control (Tables 4.8 and 4.9), and appear to be less of a contributing fact or to the discrepancy in calculation of AS from each heavy-isotope peptide standard. Th ese data do not seem to be affected by a contaminant ion as shown previously (Tab le 4.3). Nevertheless the difference in calculation of AS between th e two standard peptides is something that should be addressed. It is perhaps an effect of cleavage, where the ETFEDSNLIPK peptide may not be produced at the same rate as the WINATDPSAR peptide from in-gel digestion with trypsin. Degradation of the peptide post-digestion may be ruled out because the heavy-isotope standard w ould likely undergo the same degradation under the same conditions in the sample. The ETFEDSNLIPK peptide has less variation in fragment ion measurement from the ALL patient samples, a nd may be the more reliable standard used for quantitation. Table 4.8. Fragment Ion Ratios for WINAT DPSAR Peptide in ALL Patient Samples Sample y4 y6 y8 y4* y6* y8* M079837 3.6 ± 1.0 1 3.3 ± 1.5 2.1 ± 0.1 1 3.8 ± 0.1 M080788 3 ± 2 1 3 ± 2 2.6 ± 1.0 1 4.8 ± 2.0 M082246 4 ± 3 1 1.2 ± 0.7 2.2 ± 0.1 1 3.9 ± 0.1 M079880 5 ± 5 1 1.2 ± 0.8 2.0 ± 0.1 1 3.9 ± 0.2

PAGE 104

85 Table 4.9. Fragment Ion Ratios for ETFEDSNLIPK Peptide in ALL Patient Samples Sample y2 b8 – H2O b9 H2O y2* b8 – H2O* b9 H2O * M079837 1.3 ± 0.4 1 2.1 ± 0.8 1.4 ± 0.1 1 1.9 ± 0.4 M080788 1.4 ± 0.5 1 2.0 ± 1.4 1.4 ± 0.1 1 1.9 ± 0.1 M082246 7.3 ± 6.8 1 4.3 ± 6.1 1.4 ± 0.1 1 1.8 ± 0.1 M079880 1.1 ± 0.6 1 1.6 ± 1.2 1.4 ± 0.1 1 1.9 ± 0.1 The standard deviation in the fragment ion ratios shown in Tables 4.8 and 4.9 is significantly larger (in most cases) for the na tural abundance peptides than it is for the heavy-isotope peptides. Also, the ratios are clos er to those of the heavy-isotope values in samples in which more AS was detected. This may indicate that the fragment ions used for detection and quantitation of AS are at or below the lower limit of the linear range in these complex samples. Therefore, mon itoring the individual c ontributions of the fragment ions to the peak area in the RIC will provide useful criteria to determine if the samples analyzed contain AS within the linear range and lower limit of detection of the method. The question of sample recovery from the gel is something that also should be addressed. While the efficiency of extracti on of AS from the gel pieces is unknown, we may be able to assume that the extraction effi ciency is comparable for similar samples. The calculation of “copies of AS per cel l” as shown above assumes 100% percent efficiency in extraction for all samples. The exact number of AS molecules per cell may deviate significantly from these calculated values, but we can certainly determine the amount of AS detected in each analysis. More importantly, this method provides data from which an absolute quantit y of AS may be calculated. The results from these analyses are significan t in that AS has been detected in ALL patient samples using a mass spectrometry-ba sed method. The resu lts from the Western

PAGE 105

86 blot analyses imply that the amount of AS present in the patient samples is much less than that present in the MOLT-4 S cell line, however the response of the signal from Western blot analyses is not linear over large ch anges in protein amount.35 The results from the MS quantitative method also demonstrate the ever increasing sensitivity and selectivity of MS-based methods for de tection of targeted analyte ions. Conclusions A method for the direct dete ction and quantitation of AS in cell lines has been developed using LC/MS/MS detection. The quantitative MS assay has demonstrated strong correlation with Western blotting resu lts while providing an estimation of the amount of AS isolated from the sample. In the evaluation of the data generated for quantitation, it was found that the fragmentation patterns of both the heavy-isotope and natural abundance peptides s hould not only be identical in the specific fragment ions generated (with the exception of additional mass of the incorporated heavy-atoms), but also the intensities of each of the ions should maintain a constant ratio relative to one another, at least within the linear range of the mass spectrometer. The lower limit of detection for this method was established at 37.5 amol, with a lower limit of quantitation of 375 amol. These lower limits demonstrate th e increasing sensitivit y and selectivity of LC/MS/MS-based methods, providing a suitable resource for detec tion and quantitation of low level proteins in complex mixtures. This method was applied to the analysis of ALL patient samples obtained from peripheral blood, and provided a reasonable estimate of the amount of AS recovered from each sample, and showed agreement with the Western blot data. Further development of this method will focus on increasing the sensitivity and reproducibility, and stream lining the procedure for automation.

PAGE 106

87 CHAPTER 5 INVESTIGATION OF TWO DIMENSIONAL GEL ELECTROPHORESIS FOR DETECTION OF ASPARAGINE SYNT HETASE IN MOLT-4 CELL LINES Introduction A number of proteomics methods have b een developed to monitor the relative changes in protein expression between two or more samples.42, 124-126 Historically, twodimensional gel electrophoresis (2D-GE) has accomplished this task by the separation of proteins from a complex sample, such as a cellular lysate, first by isoelectric focusing, and then by gel electrophoresis. The resultant gel can be st ained, and the visible protein spots are compared from one sample to the next, to identify any changes between the two samples.127 The use of such a 2D method is appe aling because of its proven resolving power, often separating thousands of protei n spots in a single 2D-GE experiment.128 There are a certain number of drawbacks th at make this method less desirable. First, the entire procedure takes several days to accomplish for one sample set. Second, the two separate samples to be compared are analyzed on different gels, evoking the question of reproducibility. Third, 2D-GE is capable of separating proteins, based on their pI and molecular weight, but it is unabl e to provide identifica tion for the separated protein spots. Fourth, this method may not be suitable for very large or very small proteins, or proteins th at are not very soluble.36 Finally, the comparison of the stained protein spots on the two gels is often done manually, not only intr oducing the opportunity for error, but also making the comparison process time-consuming. While the method has been used for a number of years for th e comparison of proteins from two or more

PAGE 107

88 complex samples, it clearly has its drawback s. Development of automated methods and enhanced sample preparation has continued to improve the reproduc ibility of 2D-GE for proteomic analyses.129 One example of enhanced sample preparation is discussed below. Two dimensional differential imaging ge l electrophoresis (2DDIGE) is a method used to separate the proteins in a comple x cellular mixture, and relatively quantify the changes in protein expression between tw o samples using fluorescence detection.47, 48 Prior to separation, two protein samples from different cell states, su ch as a control and diseased state, are separate ly labeled with fluorescent dye s, Cy3 and Cy5, respectively (Figures 5.1 and 5.2). The samples are then co mbined and subjected to normal 2D-GE. The Cy3 and Cy5 dyes have distinct a nd non-overlapping exc itation and emission wavelengths, allowing for select ive detection of the different ially labeled proteins from the same gel using a fluorescence detector. The detector first excites and detects the fluorescence from one label, then excites and detects the fluorescence from another label. Once the signals have been acquired, they are overlaid, and a resultant image is displayed. In the case where there are equa l amounts of a specific Cy3 and Cy5 labeled protein, the visible color is show n as yellow. If there is mo re of a Cy3 labeled protein, it appears green, and more of a Cy5 labeled protei n appears red. The inte nsities of the spots are measured using specific software, and th e relative change in protein amount for each spot is calculated. Protein spots exhibiti ng large changes in expr ession between the two samples can be excised, in-gel digested and analyzed by LC/MS/MS for peptide identification.

PAGE 108

89 Figure 5.1. CyDye structures with excitati on and emission wavelengths. Each of the dyes is functionalized with an N-hydr oxysuccinimide reactive group, which targets primary amines, such as lysine residues. Cell State 1 Control Cell State 2 ExperimentalLabel with Cy5 CombineSamples2D Electrophoresis Label with Cy3 Lysecells and extract protein Excise spots of interest Digest MS analysisFluorescence ImagingMolecular Weight pH Cell State 1 Control Cell State 1 Control Cell State 2 Experimental Cell State 2 ExperimentalLabel with Cy5 CombineSamples2D Electrophoresis Label with Cy3 Lysecells and extract protein Excise spots of interest Digest MS analysisFluorescence ImagingMolecular Weight pH Excise spots of interest Digest MS analysisFluorescence ImagingMolecular Weight pH Figure 5.2. Schematic representation of the work flow of the 2D DIGE experiment. Cells from two different states are lyse d, and the protein extracts are labeled with Cy3 (control) and Cy5 (experiment al) fluorescent dyes. The samples are combined and the proteins are separate d by 2D electrophoresis, then the gel is analyzed by fluorescence detection. Spot s of interest are excised, digested, and the samples are identified by LC/MS analysis. The cyanine dyes used in the 2D-DIGE experiment are functionalized with an Nhydroxysuccinimide (NHS) ester, which reacts with primary amines. The dyes are added

PAGE 109

90 in a limiting amount to the protein samples, such that only about 2-3% of the total primary amine groups, lysine residues, are labeled.48 The molar extinction coefficients of Cy3 and Cy5 are 150,000 and 250,000 M-1cm-1, respectively.130 This method was investigated for two purposes. The first was to determine what changes in protein expression between the MO LT-4 S and R cell lines could be detected. The second purpose was to examine how well this method was suited to detecting the established changes in expressi on of AS between the MOLT-4 samples. It is important to note that in the following sect ion describing the experimental parameters used for the 2DDIGE experiments, there was little deviat ion from the manuf acturerÂ’s protocols (Amersham Biosciences). Materials and Methods The 2D Clean-Up kit, 2D Quant kit, imm obilized pH gradient (IPG) strips and CyDyes for the 2D-DIGE experiment were pur chased from Amersham Biosciences. All experiments, with the exception of LC/MS/MS analysis of protein spots from the series of 5 2D-DIGE gels, were carri ed out at the ICBR Proteomi cs Core Facility at the University of Florida wi th Marjorie Chow. 2D-DIGE Experiment Cytosolic protein fractions from MOLT4 S and R cell lysates were provided by Nan Su of the Kilberg Lab at the University of Florida, Department of Biochemistry. Samples were received in lysis buffer (20 mM HEPES, 10 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 20% glycerol, 1% Triton X-100, 1 mM DTT, 1 mM PMSF, and 1x Protease Inhibitor Cocktail (Sigma)).

PAGE 110

91 Preparation of cell lysate samples for 2D-DIGE Protein samples were precipitated, using the 2D Clean-Up kit, according to the manufacturerÂ’s instruct ions. Approximately 250 L of each sample (~ 1000 g, each) was precipitated according to the protocol (u sing trichloroacetic acid-based agents and acetone). The protein pellets were re-dissolved in 125 L DIGE lysis buffer (8 M urea, 2 M thiourea, 4% CHAPS, 0.1% SDS, pH 8.5). The samples were spun in an ultracentrifuge at 100K x g for 30 minutes to remove any precipitate or insoluble material. After ultracentrifugation, the supern atants were dialyzed overnight against fresh DIGE lysis buffer (membrane with 1000 Da molecular weight cut-off). Protein concentration for each sample was determined using a Plus One 2D Quant kit, according to manufacturerÂ’s instructions. The kit procedure invo lves precipitation of a portion of the samples, resolubilization of the protein in a solution containing copper ions, which will bind with the amide bac kbone structure, then incubation with a colorimetric reagent that r eacts with free copper ions in solution. Intensity of the developed color is measured at 480 nm, and th e relationship of the absorbance to protein concentration is inverse. A calibration curv e was prepared with BSA, and the protein concentrations of the experimental samples were determined from the curve. CyDye labeling procedure Initially, this experiment was conducted with 100 g of each MOLT-4 S and R protein samples. The CyDye reagents were pr eviously prepared in a stock solution of 1 nmol/ L in N,N-dimethylformamide (DMF). A 400 pmol/ L working solution for each dye was prepared by combination of 1 L of the 1 nmol/ L stock with 1.5 L DMF. The MOLT-4 S protein sample was labeled with 2 L of the Cy3 dye, and the MOLT-4 R

PAGE 111

92 protein sample was labeled with 2 L of the Cy5 dye. The samples were incubated on ice for 30 minutes in the dark. Ex cess dye was reacted by addition of 1 L 10 mM lysine solution, and the samples were incubated on ice for 10 minutes in the dark. In a second experiment, aliquots of 250 g of each the MOLT-4 S and R cytosolic proteins samples were placed in separate tubes and labeled as above, except using 5 L of each of the 400 pmol/ L working reagents, and 5 L of the 10 mM lysine to quench. To verify the success of the labeling pro cedure, the MOLT-4 S and R samples were analyzed by 1D-SDS-PAGE, 5 g per lane, then scanned using a Typhoon 8800 fluorescence image detector (Amersham Biosciences). Isoelectric focusing of 100 g combined CyDye labeled sample on pH 3-11 IPG strip The MOLT-4 S and R CyDye labeled samp les were added together in equal amounts (45 g each), then combined with 40 L 1 M DTT, 2 L of the appropriate IPG buffer stock for a pH 3-11 non-linear gradient, 2 L Orange G dye (a tracking dye), and brought to 400 L volume with DIGE lysis buffer. The sample was incubated overnight with the dehydrated IPG strip (18 cm), cove red in oil to prevent evaporation of the sample solution and ensure rehydration of the strip. Once hydrated, the strip was placed in the IPGphor (Amersham), and the protein sample was focused for 21 hours, according to the recommended voltage settings. Isoelectric focusing of CyDye labeled samples with 5 different pH gradients In a separate experiment, the MOLT-4 S and R CyDye labeled samples were combined in equal amounts (45 g each), then aliquotted into 5 equal volumes in preparation for IEF using IPG strips with di fferent pH gradients. Each sample was combined with 40 L 1 M DTT, 2 L of the appropriate IPG buffer stock (determined by

PAGE 112

93 the pH gradient on the IPG strip to be used), 2 L Orange G dye (a tracking dye), and brought to 500 L volume with DIGE lysis buffer. Each of the 5 samples was introduced to individual IPG strips (20 cm) and allowed to rehydrate for 16 hours. The pH gradient strips used were pH 3-11 non linear, pH 3-5.6 linear, pH 5.3-6.6 linea r, pH 6.2-7.5 linear, and pH 7-11 linear. Each of the strips was focused separately, according to the recommended voltage settings. SDS-PAGE second dimension Strips subjected to IEF were removed from the IPGphor apparatus and either frozen at -20ºC, or immediately prepared for SDSPAGE. Strips were separately incubated, with shaking, in 15 mL Equilibration Buff er I containing DTT for 30 minutes, then moved to Equilibration Buffer II (15 mL), c ontaining iodoacetamide, and incubated for 30 minutes, in the dark with shaking. The fi rst pH 3-11 IPG strip was loaded on the top of an 18 cm wide SDS-PAGE gel. The rema ining 5 IPG strips of differing pH gradients were loaded on the top of 5 separate 22 cm wide SDS-PAGE gels. Molecular weight markers were loaded on one edge, and the IPG st rips were oriented so that the acidic end was closest to the markers. The samples were subjected to SDS-PAGE according to manufacturer’s instructions for each gel si ze. After electrophoresis, the gels were scanned for Cy3 and Cy5 fluorescence using the Typhoon 8800 fluorescence image detector. Modification of method to target asparagine synthetase After observing the fluorescent signals from the first pH 3-11 non-linear gradient, and noticing few obvious changes in the cyto solic proteome, the above experimental scheme was repeated using 3x the amount of MOLT-4 S and R proteins, and the sample

PAGE 113

94 was applied to a pH 5.5-6.7 IPG strip to better resolve the proteins in that pH range and target AS, which has a ca lculated pI of 6.4. Visual Staining and Protein Identification The gels were stained with silver stain, according to the manufacturer’s instructions. The silver stai ned protein spots were compared with the obvious changes observed from the fluorescence signals of the Cy Dyes, and protein spots that were clearly down-regulated (appeared green in fluorescence image) or up-regulated (appeared red in fluorescence image) were excised. Silver stai ned spots were destaine d in separate 1.5 mL microcentrifuge tubes using a 1:1 solution pr epared from 30 mM potassium ferricyanide and 100 mM sodium thiosulfate. Once the gel pieces were clear, they were rinsed three times with 200 L de-ionized water, then washed twice in 200 L 100 mM ammonium bicarbonate, pH 8. Acetonitrile was added to dehydrate the gel spots. A solution of trypsin was prepared to a final concen tration of 20 ng/mL in 25 mM ammonium bicarbonate. Twenty-five microliters of the trypsin solution was added to each gel spot and the gel was allowed to rehydrate for 45 mi nutes, on ice. The excess trypsin solution was removed and discarded, and replaced with 25 L 25 mM ammonium bicarbonate. Samples were incubated at 37ºC for 18 hours, and the digestion was quenched by addition of 1 L acetic acid. The solution was tran sferred to new tubes, and the gel pieces were extracted three times with 20 L 50% acetonitrile, 5% formic acid. The extraction solution was added to the digested peptides and dried by vacuum. Digests were dissolved by sonication with 10 L of 3% acetonitrile (v/v), 0.1% TFA (v/v) in preparation for MS analysis.

PAGE 114

95 MALDI-Q-TOF MS analysis of in-gel digests Half of each sample (5 L) was desalted using C18 ZipTips, and the peptides were eluted with a solution of 50% acetonitrile, 0.1% formic aci d in a final volume of ~1 L, as previously described in Chapter 3. A stock solution of -cyano-4-hydroxycinnaminic acid was diluted 1:5 with 50% acetonitrile (v/v), 0.1% TFA (v/v). The acid was combined 1:1 with the desalted sample and 1 L was spotted onto the MALDI target. Samples were analyzed on the QSTAR instru ment (Applied Biosystems). MS data acquisition was carried out at th e ICBR Protein Core Facility with Dr. Stanley Stevens. LC/MS/MS analysis of in-gel digests The remaining 5 L of each sample was analyzed by LC/MS/MS using the LCQ Classic ion trap mass spectrometer coupled to an Eldex MicroPro HPLC pump. A preliminary desalt was carried out on a C 18 CapTrap (LC Packings) for 5 minutes with 3% acetonitrile (v/v), 0.1% ace tic acid (v/v) (mobile phase A), with the eluent going to waste. Peptide separation was carried out with a 15 cm PepMap (NewObjectives, Inc) column packed with C18 resin. The gradient profile was as follows: 3% B for 5 minutes, 3% to 90% B for 30 minutes, a wash in 90% B for 5 minutes, and re-equilibration at 3% B for 10 minutes. The MS method was set up for a full MS scan, followed by 3 datadependent full MS/MS scans of the 3 most intense parent ions. The gel spots excised from the series of 5 gels with narrow pH ranges were also digested and analyzed by LC/MS/MS, but were analyzed on a linear ion trap mass spectrometer (LTQ, Thermo). After digesti on and peptide extracti on, the dried samples were redissolved in 15 L 0.1% TFA (v/v). Samples were separated by on-line micro RPLC ( RPLC) ESI MS/MS. The end of a length of fused s ilica capillary, 75 m ID x

PAGE 115

96 360 m OD (Polymicro Technologies, Phoenix, AZ) was pulled to a fine tip (5-7 m) using a butane torch. The capillary was sl urry packed with Magic AQ C18 reversed phase resin, with 5 m bead diameter and 200 Å pore si ze (Michrom Bioresources) to a bed length of 10 cm, as previously described.131 Solvent flow was supplied by an Agilent 1100 capillary LC system (Agilent Technologies , Palo Alto, CA). Samples were loaded in 7 L volumes using an Agilent HPLC autosampler, maintained at 4ºC. After loading the sample, the column was washed with 2% B for 30 minutes at a flow rate of 0.5 L/min, then the flow rate was decreased to 0.250 L/min before initiation of the gradient. The gradient was as follows: 2-40% B over 40 minutes, 40-98% B over 30 minutes, where mobile phase A was 0.1% formic acid in water (v/v) and mobile phase B was 0.1% formic acid in acetonitrile (v/v). High voltage contact for electrospray ionization was provided through a metal union connecting the microcap illary column to the LC pump. The Thermo LTQ mass spectro meter method was created for a full-scan MS and full-scan MS/MS of the 5 most in tense ions (data dependent mode). Peptide and protein identification Peptide and protein identification was acco mplished by generation of separate dta files for each MS/MS spectrum. The dta file is a special format of MS/MS data that lists the calculated molecular weight of the parent ion selected for frag mentation, the charge state that was observed, and a list of the fr agment ions. Using a database searching program such as SEQUEST, the dta files can be compared with theoretical MS/MS data for any number of proteins found in publicly available databases, such as that found at ftp://ftp.ncbi.nlm.nih.gov/blast/db (last accessed April 4, 2006).122, 132 The integrity of the peptide and protein matches is based upon a nu mber of criteria. First, a value is given

PAGE 116

97 to how well the intensity and number of fragme nt ions match the theoretical data, called the cross correlation value, or “xcorr”. Th e higher the number, the more likely it is a positive match, for example, an xcorr value of 1.5 does not support a strong match. An xcorr value of 2.5 or higher in dicates a very strong match of the data. One additional figure of judgment for the likelihood of a peptid e ID and match with a protein is the delta correlation ( CN) value. This number reflects the difference between the normalized xcorr values of the first two peptide matches for a particular MS/MS spectrum. In this case, a value greater than 0.1 is considered good, with increasing values indicating that the first peptide ID is probably the correct one. These data togeth er are summarized in a list of peptides, and the proteins in whic h those peptide sequences appear. Often, a peptide sequence is not unique to any one prot ein, or is found to be conserved throughout a family of proteins. This value is also pres ented in the data, allowi ng the user to decide if the peptide sequence is a di stinguishing one in the protein sequence. The dta files were searched against the hu man proteome database ( http://www.ebi.ac.uk/integr8/QuickSear ch.do?pageContext=201&action=doOrgSearch& geneName=&organismName=sapiens , last accessed April 4, 2006). Results and Discussion 2D-DIGE of MOLT-4 Samples Using Broad pH Range Approximately 50 g each of the MOLT-4 S and R cytosolic protein fractions were derivatized with Cy3 and Cy5 dyes, respectively. Verificat ion of the labeling step was accomplished by analysis of a small portion of the protein samples by 1D-SDS-PAGE and fluorescence imaging of the gel (Figure 5.3).

PAGE 117

98 160 105 75 50 35 30 25 15 10 250 kDaA B 160 105 75 50 35 30 25 15 10 250 160 105 75 50 35 30 25 15 10 250 kDaA B Figure 5.3. SDS-PAGE analysis of CyDye la beled MOLT-4 S and R protein samples. Panel A represents the merged fluores cence image for the Cy3 (left lanes, green), and Cy5 (right lanes, red) signals. Panel B is the same gel stained with colloidal blue. Each of the lanes contains 5 g MOLT-4 S protein (5 left lanes in each gel) or 5 g MOLT-4 R protein (5 right lanes in each gel). The appearance of fluorescence signals in the 1D SDS-PAGE step supported successful sample labeling and the CyDye labeled MOLT-4 S and R protein samples were combined and analyzed by 2D-GE. After electophoresis in the second dimension, the gel was scanned, using the fluorescence imag ing detector, and the fluorescent images from the Cy3 and Cy5 signals were overlaid (Figure 5.4). The resultant image displays three key features: protein spots appeari ng green are more abundant in the MOLT-4 S (Cy3) sample, protein spots appearing red are more abundant in the MOLT-4 R (Cy5) sample, and protein spots appearing yellow are present in about the same amount in each sample.

PAGE 118

99 160 105 75 50 35 30 25 15 10 250 kDa 3 11 6 4578910 pH 160 105 75 50 35 30 25 15 10 250 kDa 3 11 6 4578910 160 105 75 50 35 30 25 15 10 250 160 105 75 50 35 30 25 15 10 250 kDa 3 11 6 4578910 3 11 6 4578910 pH Figure 5.4. 2D-DIGE merged fluorescence im age of the MOLT-4 S and R cytosolic proteins. Spots appearing as green (Cy3) are more abundant in the MOLT-4 S sample, spots appearing as red (Cy5) are more abundant in the MOLT-4 R sample, and spots appearing as yellow are present in both samples in very similar amounts. The 2D gel was stained with silver stain to allow for visualization of the protein spots for excision, in-gel digestion and MS anal yses. Use of the broad pH range IEF strip resulted in reasonable separation of a large number of proteins. However, the primary interest in using this method was to determin e if the change in expression of AS between the MOLT-4 S and R cell lines was detectable. The pI of AS is calculated to be 6.4, and the molecular weight is approximately 64 kDa. When focusing in on this region of the

PAGE 119

100 2D-DIGE gel, there were few obvious spots that exhibited the re d Cy5 signal, which would indicate increased expr ession of AS in the MOLT-4 R cell line. A small number of gel spots were excised, de stained, in-gel digested, and analyzed by both MALDI-MS and LC/MS/MS to determine the identification of the proteins. Unfortunately, practically all of the results were inconclusive, with keratin and trypsin iden tified as the primary proteins in each spot. Keratin is a common pr otein contaminant that originates from skin cells and is ubiquitous in the lab environment.133-135 Trypsin is the protease used for digestion, and due to autolysis or chymotr ypsinogen activity, often contributes a number of peptides to the sample.134, 136 Analysis of MOLT-4 Protei ns Over pH Range 5.5-6.7 A second 2D-DIGE experiment, utilizi ng triple the amount of protein (150 g from each of the MOLT-4 S and R samples), was ca rried out using a narrower pH range IEF strip, intending to focus around the pI of AS, an d increase resolution of the protein spots in that range. The merged fluorescent imag e is shown in Figure 5.5. A number of red spots surrounding the pH 6.4 and 64 kDa were cl early visible on this second gel. After visualization with silver stain, the spots i ndicated in Figure 5.5 we re excised, destained, in-gel digested and analyzed by MS for protein identification.

PAGE 120

101 Figure 5.5. 2D-DIGE fluorescence image of 150 g MOLT-4 S and R cytosolic proteins, pH range 5.5-6.7. The spots shown with white circles are in the approximate pH and molecular weight ranges where AS may migrate. These spots were excised and subjected to in-gel dige stion and protein identification by MS analyses. Spots labeled A and B were the only two that were conclusively identified after digestion and MS analyses. Of the nine red spots that were sel ected for in-gel digestion and protein identification, five were conclu sively identified, based on th e MS data, and are shown in Table 5.1. Protein identification was based on the detection of two or more unique peptides with a xcorr value of 2.1 or higher fo r the 2+ charge state. The remaining gel spots were not conclusively identified due to large signals from both trypsin and keratin. The trypsin peptides arose from autolysis of the protease used for digestion, while keratin A B C D E

PAGE 121

102 is a common contaminant that may be present at a larger abundance than the protein or proteins excised in the gel s pot, and may have suppressed an y signal from a less abundant analyte. Table 5.1. Proteins Identified from 2D-DIGE, pH Range 5.5-6.7 (Figure 5.5). Gel Spot Protein Accession Number Number of Peptides MW pI A Heptacellular carcinoma-associated antigen 59 Q9NZ63 2 33688 6.33 B Glutamine synthetase P15104 5 41933 6.42 C Ciliary dynein heavy chain 7 Q8WXX0 2 461143 5.70 D Phosphoglycerate dehydrogenase O43175 7 56519 6.31 E Tyrosyl-tRNA synthetase P54577 11 59012 6.62 Asparagine synthetase was not identified in any of th e red, “up-regulated” protein spots that were analyzed, but this can be e xplained in a number of ways. First, the 2DGE method assumes that AS will be soluble in the buffer conditions chosen, and maintain its solubility throughout each step. The proce ss of isoelectric focusing of the protein sample causes each protein to migrate to its pI, where the net charge on the molecule is neutral. It has been reported in some cases that proteins at thei r isoelectric point are unable to transfer into the second dimension becau se they have become insoluble. If this were the case with AS, then the protein would not be present on the SDS-PAGE gel, and would therefore be undetected.137 Second, the 2D-GE method has also been known to exclude highly acidic, basic, and hydrophobic proteins as we ll as low abundance proteins.36, 137, 138 While AS does not fall into the first three categories, it is lik ely that it is present at low abundance, in comparison to other proteins in the sample, and may not have been sufficiently resolved or detected. Finally, at the time of analysis, the am ount of AS present in the sample was unknown. While the fluorescence signals from AS in each the MOLT-4 S and R samples

PAGE 122

103 would allow detection of the protein at very low levels, if there was not enough protein to be visibly stained using silver stain, then identifi cation of the protein by in-gel digestion and MS detection would be very difficult. Sens itivity of silver staining of gels has been reported to the picogram level, which w ould correspond to hundreds of attomoles of protein.139, 140 Analysis of MOLT-4 Prot eins Using Several Narrow pH Range IPG Strips In a final effort to track down the lo cation of AS in the 2D electrophoresis experiments, a series of 5 2D gels with va rying pH ranges were used to separate the MOLT-4 S and R CyDye labeled proteins. As the pH range is decreased and the same amount of protein is loaded on the IPG strip, the proteins become better resolved during isoelectric focusing, and fewer interferences compete for protein detection. The four narrow pH ranges, pH 3-5.6, pH 5.3-6.6, pH 6.2-7.5, and pH 7-11 were compared to a single, broad-range pH 3-11 nonlinear gradient (Figures sh own in Appendix A). After fluorescence imaging, the protein spots that sh owed greatest differenc es (either appeared obviously green or red in the image) and were visualized by silver staining were excised for in-gel digestion and pr otein identification by LC/MS/ MS. Proteins were ranked according to the number of peptides identified when more than one protein was identified per spot. Peptide identifications were also filtered to provide onl y those with high xcorr values and sequences unique to single protei ns. Spots identified as red, or up-regulated, are designated with the letter R, while s pots appearing green, or down-regulated, are designated with the letter G, and the results are shown in Table 5.2.

PAGE 123

104 Table 5.2. Identification of Proteins Det ected in 2D-DIGE Gel Spots by LC/MS/MS Gel pH gradient and spot Protein Name Accession Number Number of Peptides MW (kDa) pI pH 3-11, spot R1 Leucine zipper transcri ption factor-like 1 Hypothetical protein FLJ39100 Q9NQ56 Q8N1N4 11 2 34591 56819 5.37 5.79 pH 3-11, spot R2 D-3-phosphoglycerate dehydrogenase Protein Plunc precursor Anti-RhD monoclonal T125 kappa light chain precursor Lactotransferrin precursor Beta-tubulin cofactor E O43175 Q9NP55 Q5EFE6 P02788 Q15813 9 3 2 2 2 56519 26712 25697 78338 59346 6.31 5.42 8.69 8.56 6.32 pH 3-11, spot R3 Tyrosyl-tRNA synthetase, cytoplasmic 78 kDa glucose-regulated protein precursor Eukaryotic translation initiation factor 4E transporter P54577 P11021 Q9NRA8 3 1 1 59012 72332 108200 6.64 5.07 8.45 pH 3-11, spot G1 Endoplasmin precursor NPD007 Nucleolin P14625 Q9HBK7 P19338 1 1 1 92468 46650 76212 4.76 8.64 4.59 pH 3-11, spot G2 Annexin A1 Actin Bcl-2-associated transcription factor 1 P04083 P62736 Q9NYF8 2 1 1 38583 42008 106122 6.64 5.24 9.99 pH 3-5.6, spot G1 Hydroxymethylglutaryl-CoA synthase, Junction plakoglobin Histone H4 LAP3 Protein Q01581 P14923 Q6FGB8 Q6IAM6 4 3 2 2 57293 81498 11393 56136 5.22 5.95 11.36 8.03 pH 3-5.6, spot G2 Adenosine deaminase P00813 9 40633 5.63 pH 3-5.6, spot R1 Branched-chain-amino-acid aminotransferase, cytosolic P54687 3 42952 5.17 pH 3-5.6, spot R2 Heat shock cognate 71 kDa protein Leucine zipper transc ription factor-like1 Heat shock 70 kDa protein 1L P11142 Q9NQ56 P34931 3 3 2 70898 34591 70375 5.37 5.37 5.76 pH 5.3-6.6, spot G1 Annexin A1 P04083 10 38583 6.64 pH 5.3-6.6, spot G2 D-3-phosphoglycerate dehydrogenase O43175 2 56519 6.31 pH 5.3-6.6, spot R1 60S ribosomal protein L4 ATP synthase beta chain, mitochondrial precursor P36578 P06576 1 1 47697 56559 11.09 5.26 pH 5.3-6.6, spot R2 Synaptic vesicle membrane protein VAT-1 homolog Q99536 4 41920 5.88 pH 5.3-6.6, spot R3 Seryl-tRNA synthetase P49591 8 58646 6.06 pH 5.3-6.6, spot R4 D-3-phosphoglycerate dehydrogenase O43175 7 56519 6.31 pH 5.3-6.6, spot R5 78 kDa glucose-regulated protein precursor Heat shock-related 70 kDa protein 2 P11021 P54652 1 1 72332 70020 5.07 5.56 pH 5.3-6.6, spot R6 Heat shock-related 70 kDa protein 2 Collagen alpha 4(IV) chain precursor P54652 P53420 2 2 70020 164095 5.56 8.90

PAGE 124

105 Table 5.2. Continued. Gel pH gradient and spot Protein Name Accession Number Number of Peptides MW (kDa) pI pH 6.2-7.5, spot G1 Inosine-5'-monophosphate dehydrogenase 2 Filaggrin.OS P12268 Q5T583 12 2 55804 435169 6.44 9.24 pH 6.2-7.5, spot R1 D-3-phosphoglycerate dehydrogenase O43175 11 56519 6.31 pH 6.2-7.5, spot R2 Ash1 Hypothetical protein FLJ42667 Myosin-11 Q5T714 Q6ZVE3 P35749 1 1 1 332790 15400 227339 9.46 10.03 5.42 Conclusions The proteins identified by LC/MS/MS analysis from the 2D-DIGE gel spots are believed to be the most abundant proteins pres ent in those spots. Asparagine synthetase was not detected using this method. This may indicate that either AS should be considered a low abundance protein or that it co-migrates with another protein to the same pH and molecular weight, and that the other protein is more readily identified. There were no efforts made to tailor th e LC/MS/MS method fo r the detection of AS. The five most abundant peptide ions per full MS scan were selected for MS/MS fragmentation. The best conclu sion to support these data is that AS is not present at abundant enough levels for detection under the described conditions. This conclusion leads to the question: what ot her proteins are exhibiting sign ificant changes in expression and are not being detected by current proteomics methods? Unfortunately, there is no straightforward answer to this question. Instead, experiments should be designed to answer specific questions using a number of orthogonal techniques to ensure the most comprehensive set of data.

PAGE 125

106 CHAPTER 6 TARGETED DETECTION AN D RELATIVE QUANTITAT ION OF ASPARAGINE SYNTHETASE IN MOLT-4 CELL LINES USING ISOTOPE-CODED AFFINITY TAGS Introduction The Isotope-Coded Affinity Tag (ICAT) is a type of reagent used to derivatize complex protein samples from two different s ources so that when they are combined and analyzed by LC/MS, the relative abundan ce of most proteins can be compared.36 The ICAT reagent is made up of 4 functional co mponents (Figure 6.1): a thiol-reactive group, an isotope-coded tag, containing 9 x 12C or 9 x 13C, an acid cleavable region, and a biotin affinity group. IodoacetylReactive Group Protein Isotope-Coded Tag (C10H17N3O3) Heavy: 9 x 13C (236 amu) Light: 9 x 12C (227amu) Acid Cleavable Region Biotin Affinity Tag IodoacetylReactive Group Protein Isotope-Coded Tag (C10H17N3O3) Heavy: 9 x 13C (236 amu) Light: 9 x 12C (227amu) Acid Cleavable Region Biotin Affinity Tag Figure 6.1. Schematic representation of the ICAT reagent and its four main functional components. Figure adapted from ICAT kit protocol, Applied Biosystems.141 The premise behind the ICAT reagent is that most proteins contain cysteine residues, and isolation of cysteine-containi ng peptides after proteolysis will reduce the sample complexity to facilitate better detection during the LC/MS step. In order to

PAGE 126

107 accomplish this, two protein samples are separa tely derivatized with the light and heavy isoforms of the ICAT reagent, then combined (Figure 6.2). The sample is digested with trypsin, producing ICAT-labeled and unlabel ed peptides. The labeled peptides are selectively isolated by exploiting the high affini ty between a stationary avidin solid-phase and the biotin moiety of the ICAT reagent. Once the labeled peptides are released from the avidin column, the bulky biotin group is removed by acidifying the solution and cleaving at the acid cleavabl e linker region. The remaining portion of the tag still contains the isotope coded tag, differing by 9 Daltons, allo wing for mass discrimination of peptides based on their initial source by MS. Due to the capabilities of MS for peptide sequence determination by fragmentation, the peptide sequence can be determined. The peptide fragmentation can, in turn, be comp ared to known peptide sequences in protein databases, and protein identification may be ascertained. Additionally, in the resultant mass spectrum of a peptide that was found to be present in both samples, the peak areas of the light and heavy ICAT ve rsions of the peptide can be compared to determine the relative change in expr ession of the protein from which it arises. It is known that the MOLT-4 drug resistant (R) cell line expresse s a larger quantity of asparagine synthetase (AS) than the MOLT-4 drug sensi tive (S) cell line.4, 5 However, until recently, the only method for relative co mparison of the amount of AS in each sample has been antibody-based Western blotting analysis.16 This method is semiquantitative at best, and the response is not linear over several orde rs of magnitude, and often requires a standard curve of the protein to be analyzed simultaneously.30 Therefore, analysis of the MOLT-4 S and R cell lines using the ICAT method was considered to determine the relative change s in expression of AS.

PAGE 127

108 Sample 1 Sample 2Light ICAT reagentlabeled cysteines Heavy ICAT reagentlabeled cysteines Combine and Digest Sample Derivatization Peptide Separation LC/MS/MS Protein Identification and Quantitation 1. Cation Exchange 2. Avidin Affinity Separation 200400600800100012001400160018002000 m/z, amu 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4Intensity, counts 616.4335 715.5196 493.2715 510.3055 545.4114 278.1410 254.1372474.35311002.6726 828.6156 367.2201484.2707 1117.7141 597.3171 353.2354 271.1628 630.3690 456.3006 175.08381344.81581431.8240 315.1681 566.3395 871.5128 438.2988 325.2230 689.3729840.4588928.5526 1100.7093 y7y6y4y8y10y11y14y9y15y16b6y4y5MS/MS Identification 107610781080108210841086108810901092 m/z, amu 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 147Intensity, counts 1080.1789 1084.6926 1084.1922 1079.6802 1080.6810 1085.1 942MS Quantitation Light Heavy Sample 1 Sample 2Light ICAT reagentlabeled cysteines Heavy ICAT reagentlabeled cysteines Combine and Digest Sample Derivatization Peptide Separation LC/MS/MS Protein Identification and Quantitation 1. Cation Exchange 2. Avidin Affinity Separation 200400600800100012001400160018002000 m/z, amu 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4Intensity, counts 616.4335 715.5196 493.2715 510.3055 545.4114 278.1410 254.1372474.35311002.6726 828.6156 367.2201484.2707 1117.7141 597.3171 353.2354 271.1628 630.3690 456.3006 175.08381344.81581431.8240 315.1681 566.3395 871.5128 438.2988 325.2230 689.3729840.4588928.5526 1100.7093 y7y6y4y8y10y11y14y9y15y16b6y4y5MS/MS Identification 200400600800100012001400160018002000 m/z, amu 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4Intensity, counts 616.4335 715.5196 493.2715 510.3055 545.4114 278.1410 254.1372474.35311002.6726 828.6156 367.2201484.2707 1117.7141 597.3171 353.2354 271.1628 630.3690 456.3006 175.08381344.81581431.8240 315.1681 566.3395 871.5128 438.2988 325.2230 689.3729840.4588928.5526 1100.7093 y7y6y4y8y10y11y14y9y15y16b6y4y5 200400600800100012001400160018002000 m/z, amu 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4Intensity, counts 616.4335 715.5196 493.2715 510.3055 545.4114 278.1410 254.1372474.35311002.6726 828.6156 367.2201484.2707 1117.7141 597.3171 353.2354 271.1628 630.3690 456.3006 175.08381344.81581431.8240 315.1681 566.3395 871.5128 438.2988 325.2230 689.3729840.4588928.5526 1100.7093 y7y6y4y8y10y11y14y9y15y16b6y4y5MS/MS Identification 107610781080108210841086108810901092 m/z, amu 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 147Intensity, counts 1080.1789 1084.6926 1084.1922 1079.6802 1080.6810 1085.1 942MS Quantitation Light Heavy 107610781080108210841086108810901092 m/z, amu 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 147Intensity, counts 1080.1789 1084.6926 1084.1922 1079.6802 1080.6810 1085.1 942 107610781080108210841086108810901092 m/z, amu 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 147Intensity, counts 1080.1789 1084.6926 1084.1922 1079.6802 1080.6810 1085.1 942 107610781080108210841086108810901092 m/z, amu 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 147Intensity, counts 1080.1789 1084.6926 1084.1922 1079.6802 1080.6810 1085.1 942MS Quantitation Light Heavy Figure 6.2. Schematic representation of the wo rkflow of the ICAT method. After initial derivatization of two se parate samples with the light and heavy ICAT reagents, the samples are combined, dige sted, and the ICAT-labeled peptides are isolated by cation exchange, then avidin affinity chromatography. The labeled peptides are then analyzed by LC/MS/MS for peptide quantitation (by MS) and identification (by MS/MS). This figure was adapted from Applied Biosystems, Inc.142 The first goal of this work was to investigate the sample derivatization process using a series of model peptides with slight ly varied amino acid sequences to determine the specificity of the reacti on, and if it proceeded to completion. The second goal was to label purified recombinant human asparagine synthetase (rhAS) with the ICAT reagent and observe the number of cysteine-containing peptides that were identified. The third goal was to examine the limit of detection of AS in this method, using complex protein samples spiked with purified recombinant human AS (rhAS). Lastly, the ICAT protocol was followed using MOLT-4 S and R cytosolic pr otein fractions to determine if AS could be identified and quantitated from the two samples.

PAGE 128

109 Materials and Methods Preliminary Peptide Deri vatization Experiments A mock ICAT experiment with 5 decap eptides was conducted to determine the efficiency of the labeling procedure. Five peptides, each with 10 amino acid residues, were synthesized at the ICBR Protein Core Faci lity at the University of Florida. The sequences of the peptides are shown in Tabl e 6.1. An iodoacetylreactive biotinylating reagent similar in structure to the ICAT r eagent was purchased from Pierce and used in the labeling experiment. Table 6.1. Peptides Used for Derivatization Experiments with Biotinylation Reagent Peptide Sequence Mass Mass + PEO-Biotin Label 1 WGDMAAAYAK 1082.485 1082.485 2 WGDMACAYAK 1114.458 1528.648 3 WGVMACAYAK 1098.499 1512.689 4 WGAMACAYAK 1070.468 1484.658 5 WGKMACAYAK 1127.526 1541.716 EZ-Link® Iodoacetyl-PEO2-Biotin {(+)-biotinyl-iodoacetamidyl-3,6dioxaoctanediamine} (PEO-biotin) was disso lved in 1 mL 50 mM Tris-HCl, 5 mM EDTA, pH 8.3 to a final concentration of 2.0 mM. Peptides were weighed in 1 mg quantities and dissolved with 500 L of the PEO-biotin reagent solution. The samples were incubated for 90 minutes at 37°C in th e dark. Reactions were quenched by dilution with 250 L 0.1% TFA (v/v), in preparation for HPLC purification. HPLC purification of the biotinylated peptides was accomplished using a Rainin HPLC system with a Dynamax UV detector. Mobile phase A was 0.1% TFA (v/v) in water, mobile phase B was 90% acetonitrile (v/v), 0.1% TFA (v/v) in water. The gradient was as follows: 10-50% B over 15 minutes, 50-100% B over 15 seconds, wash at 100% B for 1 minute, then 5 second return to 10% B for a 9 minute re-equilibration.

PAGE 129

110 The flow rate was 1.0 mL/min, the column was reversed-phase C18, 4.6 mm x 250 mm (Vydac), and absorbance was monitored at 280 nm. Peaks absorbing strongly at 280 nm were collected and analyzed by ESI-FTICR MS to determine the success of the labeling reaction. Peptide control samples, exposed to the same buffers and incubation times, but not the labeling reagent, were also purified by HPLC and collected for MS analysis. ESI-FTICR MS Analysis of Biotinylated Peptides Samples collected from the HPLC purifi cation step after biotinylation were analyzed directly by ESI-FTICR MS to determ ine if the biotin reagent was successfully added to each of the peptides (except pep tide 1). Samples were loaded into a glass syringe and flowed through fused silica capillary (150 m ID) at a flow rate of 1 L/min. High voltage (~2.5 kV) was applied to a meta l union connecting the fu sed silica capillary transfer line to a fused silica ESI emitter wi th an externally tapered tip. The FTICR method was set up to collect 1 sec of ions in the external hexapole re gion prior to transfer to and detection in the ICR cell. MS an alysis was performed on a Bruker BioApex 47e FTICR mass spectrometer with a magnetic field strength of 4.7 T. The ESI source (Analytica of Branford) was modified with a he ated metal capillary, maintained at 120°C. Data collected consisted of 128K data point transients which were not apodized prior to Fourier transformation. ICAT Derivatization of Purified rhAS Two 100 g aliquots of rhAS were placed in separate tubes for precipitation. Trichloroacetic acid (20% w/v) was added in 300 L volumes to each tube of protein. Immediately, a cloudy precipitate formed. Solutions were incubated on ice for 30 minutes. Tubes were spun in a bench-top cen trifuge for 15 minutes at 14 K x g to pellet

PAGE 130

111 the protein. The supernatant wa s removed and replaced with 300 L cold acetone, allowing for little sample agitation. Tubes we re spun again for 5 minutes at 14 K x g, the supernatant was removed and pellets were allowed to air dry for 5 minutes. According to the instructions in cluded with the ICAT kit, 80 L of denaturing buffer (50 mM Tris, 0.1% SDS), was added to e ach of the two rhAS tubes to dissolve the protein pellets. Two microl iters of reducing reagent (50 mM tris-carboxyethylphosphine) was added to each sample and boiled for 10 minut es. One vial of light ICAT reagent and one vial of heavy ICAT reagent were each reconstituted with 20 L acetonitrile. Each of the reduced rhAS samples was transferred to one of the ICAT reag ent vials, and they were incubated at 37ºC for two hour s in the dark. An aliquot of 2 L was removed from each tube for SDS-PAGE analysis to confirm successful labeling. The contents of the light ICAT reagent vi al were transferred to the heavy ICAT reagent vial. A vial of trypsin (provi ded with the kit) was dissolved with 200 L deionized water, and added to the combined ICAT-labeled samples. The sample was incubated 16 hours at 37ºC. The ICAT-labeled peptide sample was dilu ted by the addition of 2 mL of “cation exchange buffer – load”. A cation exchange (CEX) cartridge, provided with the kit, was manually equilibrated with 2 mL of “cation exchange buffer – load”, at a flow rate of 2-3 drops per second using a 2 mL syringe. The samp le was loaded, in its entirety, at a rate of 1-2 drops per second. An additional 1 mL of the load buffer was used to wash the column. Peptides were eluted using 500 L of “cation exchange buffer – elute”, and collected in a single tube.

PAGE 131

112 The CEX purified peptide fraction was prepared for avidin chromatography separation using an avidin affin ity cartridge by the addition of 500 L of “affinity buffer – load”. The cartridge was conditioned w ith 2 mL “affinity buffer – elute”, then equilibrated with 2 mL “affinity buffer –loa d”. The peptide solution was loaded onto the avidin cartridge at a rate of ~1 drop/5 s econds. The cartridge was washed with an additional 500 L “affinity buffer – load”, then 1 mL “affinity buffer – wash1”. “Affinity buffer – wash 2” was used to rem ove the non-specifically bound peptides with a 1 mL volume. Finally, the last wash was acco mplished with 1 mL of deionized buffer. The ICAT-labeled peptides were el uted in a single fraction using 800 L “affinity buffer – elute”, and the last 750 L was collected in a tu be, and lyophilized. Cleavage of the biotin moiety from the labeled peptides was accomplished by dissolving the labeled peptides with a 90 L volume of a 95:5 mixture of “cleaving reagent A” and “cleaving reagent B”. The sample was incubated for 2 hours at 37ºC, then lyophilized. ICAT Labeling of MOLT-4 S Cytosolic Pr oteins with 1% and 0.1% rhAS Spike Purified rhAS was spiked into a 100 g aliquot of MOLT-4 S cytosolic protein sample at 1% (1 g) and 0.1% (100 ng) levels prior to sample precipitation. Each sample was prepared as described above, and labeled with heavy ICAT reagent. Two additional 100 g aliquots of MOLT-4 S cytosolic protei n were also precipitated and labeled separately with light ICAT reagent. Prio r to trypsin digestion, the MOLT-4 S + rhAS spike samples were each combined with a MOLT-4 S sample. The remaining ICAT procedure was carried out as previously described.

PAGE 132

113 ICAT Labeling of MOLT4 S and R Cell Lines The precipitation and labeling procedure was repeated us ing the cytosolic protein fraction from MOLT-4 S and R cell lines, provided by Dr. YuanXiang Pan, from the Kilberg Laboratory, Department of Biochemist ry, University of Fl orida. The protein concentration was determined by Bradford a ssay. The MOLT-4 S and R protein samples were precipitated, as previous ly described, in 100 ug aliquots . The samples were labeled with light (MOLT-4 S) and heavy (MOLT-4 R) ICAT reagents, and the procedure above was repeated. LC/MS/MS Analysis of ICAT Labeled rhAS Post-biotin cleavage, the lyophilized samples were resolubilized with 50 L mobile phase A (3% acetonitrile, 0.1% acetic acid). For the purified rhAS samples, a portion of the sample was diluted 10-fold with mobile phase A, and 5 L was analyzed by LC/MS/MS on a QSTAR MS. For the complex MOLT-4 S and R samples, and the MOLT-4 S + rhAS samples, 5 L (10%) of the original solution was analyzed. The sample was loaded onto a C18 CapTrap an d initially desalted for 5 minutes at a flow rate of 20 L/minute, with the eluent going to wast e. The peptides were then eluted onto a 75 m ID x 15 cm PepMap column packed with C18 resin using the following gradient: 0% to 40% mobile phase B (90% acetonitrile, 0.1% acetic acid) for 55 minutes, 40% to 90% for 5 minutes, 90% wash for 5 minutes, 90% to 0% for 5 minutes, and reequilibration at 0% for 60 mi nutes. MS data were collected from the initiation of the gradient for 65 minutes. Initially, the QS TAR MS was operated to provide a full MS scan of the parent ions, then an MS/MS sp ectrum for the two most abundant parent ions from the MS scan. After analysis of the MOLT-4 S and R sample, the cysteine-

PAGE 133

114 containing peptides from rhAS were entered into a “mass inclusion” li st to prioritize the ions, if detected in a full scan MS, for MS/MS fragmentation, and thus peptide identification (Table 6.2). Table 6.2. Cysteine-containing Peptides Found in rhAS. Peptide Sequence MW MW + light ICAT MW + heavy ICAT 1 CGIWALFGSDDCLSVQCLSAMK 22 2346.052 3027.44 3054.53 33 FENVNGYTNCCFGFHR 48 1906.788 2361.05 2370.08 66 YPYLWLCYNGEIYNHK 81 2074.961 2302.09 2311.12 107 GGIEQTICMLDGVFAFVLLDTANK 130 2554.281 2781.41 2790.44 147 AMTEDGFLAVCSEAK 161 1570.701 1797.83 1806.86 203 YHHCR 207 714.302 941.43 950.46 252 IGCLLSGGLDSSLVAATLLK 271 1930.081 2157.21 2166.24 The MS data were processed using ProI CAT software from Applied Biosystems, resulting in peptide and protein identification as well as relative qua ntitation of light and heavy ICAT labeled peptides. Results and Discussion Analysis of Biotinylated Peptid es in Mock ICAT Experiment Five peptides were synthesized with sp ecific amino acid sequences to determine how well the derivatization step of the ICAT experiment might work. The peptides each contained a C-terminal lysine, to mimic pep tides produced from dige stion with trypsin. Four out of the five peptides contained a cy steine residue, to react with the iodoacetylfunctional group on the PEO-biotin reagent, and all five peptides contained a methionine, included to test the specific ity of the sulfhydryl-specific iodoacetyl functional group of the reagent. Finally, each pe ptide contained a tryptophan re sidue so that detection by absorbance at 280 nm would be possible. The PEO-biotin reagent (Fi gure 6.3) was very similar in structure to the ICAT reagent, but is only availabl e in a “light” isotope, or na tural-abundance isotopic form.

PAGE 134

115 This chemical was considerably less expensiv e than the ICAT reagents, and suitable for this derivatiza tion study. HN O H NH H S O H N O O N H O I Figure 6.3. EZ-Link® Iodoacetyl-PEO2-Biotin {(+)-biotinyl-iodoacetamidyl-3,6dioxaoctanediamine}. The chemical structur e of this reagent is very similar to that of the ICAT reagents, and was used to establish the selectivity of labeling process. After the labeling procedure, each peptide was purified by RP-HPLC with UV detection. Control samples of each peptide th at were not subjected to derivatization were also analyzed by RP-HPLC, as well as a PEO-biotin reagent-only sample. Peaks absorbing strongly at 280 nm were colle cted and analyzed by ESI-FTICR MS to determine if the mass of the PEO-biotin reagen t was added, and to inve stigate if any other amino acids besides cysteine were derivatize d. Chromatograms for each of the peptides are shown in Figures 6.4 and 6.5. Each of the five peptides analyzed exhibite d a shift to a longer retention time after treatment with PEO-biotin, except for peptide 1 (sequence WGDMAAAYAK), which contained no cysteines, and pe ptide 3. It is uncl ear as to why peptide 3 did not shift, because the MS data provided evidence of the change in molecular weight of the peptide, due to addition of the PEO-biotin reagent.

PAGE 135

116 A b s o r b a n c e , 2 8 0 n m n o r m a l i z e d t o m o s t i n t e n s e p e a k Peptide 1 WDGMAAAYAK PEO-biotin Control Peptide 1 + PEO-biotin Time (min)A b s o r b a n c e , 2 8 0 n m n o r m a l i z e d t o m o s t i n t e n s e p e a k Peptide 1 WDGMAAAYAK PEO-biotin Control Peptide 1 + PEO-biotin Time (min) Figure 6.4. RP-HPLC chromatograms of the PE O-biotin reagent (top panel), peptide 1 alone (middle panel), and peptide 1 with PEO-biotin. The sequence of peptide 1 (WDGMAAAYAK) does not contain a cysteine residue, and is therefore unmodified by the PEO-biotin reagent. The PEO-biotin reagent also absorbed at 280 nm and eluted between 14-15 minutes.

PAGE 136

117 A b s o r b a n c e , 2 8 0 n m N o r m a l i z e d t o t h e m o s t i n t e n s e p e a kPeptide 2 WGDMACAYAK Peptide 3 WGVMACAYAK Time (min)Time (min)Peptide 2 + PEO-biotin Peptide 3 + PEO-biotin A b s o r b a n c e , 2 8 0 n m N o r m a l i z e d t o t h e m o s t i n t e n s e p e a kPeptide 2 WGDMACAYAK Peptide 3 WGVMACAYAK Time (min)Time (min)Peptide 2 + PEO-biotin Peptide 3 + PEO-biotin Peptide 4 WGAMACAYAK Peptide 5 WGKMACAYAK Peptide 4 + PEO-biotin Peptide 5 + PEO-biotin A b s o r b a n c e , 2 8 0 n m N o r m a l i z e d t o t h e m o s t i n t e n s e p e a kTime (min)Time (min) Peptide 4 WGAMACAYAK Peptide 5 WGKMACAYAK Peptide 4 + PEO-biotin Peptide 5 + PEO-biotin A b s o r b a n c e , 2 8 0 n m N o r m a l i z e d t o t h e m o s t i n t e n s e p e a kTime (min)Time (min) Figure 6.5. Chromatograms of peptides 2, 3, 4, and 5 before and after PEO-biotin derivatization. In most cases, the de rivatized peptide shifted to a longer retention time, indicating a potential ch ange to the molecule. The signal shown in the chromatogram for Peptid e 4 + PEO-biotin at the end of the gradient was due to a problem with the UV detector, but does not influence the preceding data in the chromatogram. MS analyses of each of the collected pe ptides are summarized in Table 6.3, and representative spectra of cont rol peptide 1 and the PEO-deriva tized peptides are shown in Figures 6.6 and 6.7.

PAGE 137

118 Table 6.3. Theoretical and Obse rved m/z Signals for Unlabeled and PEO-Biotin Labeled Peptides. Peptide Sequence Theoretical (M+2H)+2 m/z unlabeled Observed (M+2H)+2 m/z unlabeled Theoretical (M+2H)+2 m/z PEO-labeled Observed (M+2H)+2 m/z PEO-labeled 1 WGDMAAAYAK 542.251 542.277 542.251 542.277 2 WGDMACAYAK 558.237 558.251 765.332 765.378 3 WGVMACAYAK 550.258 550.287 757.353 757.397 4 WGAMACAYAK 536.242 536.258 743.365 743.381 5 WGKMACAYAK 564.771 564.787 771.866 771.911 (M+2H)+2 theo: 542.2506 (M+2H)+2 expt: 542.2772Loss of NH3( -17 mu) Ion intensity (arbitrary units)m/z 525 530535 540 545 550555560565 (M+2H)+2 theo: 542.2506 (M+2H)+2 expt: 542.2772Loss of NH3( -17 mu) Ion intensity (arbitrary units)m/z (M+2H)+2 theo: 542.2506 (M+2H)+2 expt: 542.2772Loss of NH3( -17 mu) Ion intensity (arbitrary units)m/z 525 530535 540 545 550555560565 Figure 6.6. ESI-FTICR MS spectrum of pe ptide 1, sequence WGDMAAAYAK. The peaks shown represent the 2+ charge state of the pe ptide. Theoretical and experimental m/z values are shown in Table 6.3.

PAGE 138

119 Ion intensity (arbitrary units)m/z 745 750 755 760 765 770 775 780 Ion intensity (arbitrary units)m/z 745 750 755 760 765 770 775 780 Figure 6.7. ESI-FTICR MS spectrum of peptides 2, 3, 4, and 5 after derivatization with PEO-biotin reagent. Peaks shown are the isotopic distributions for the 2+ charge state for each peptide. Theore tical and experimental m/z values are shown in Table 6.3. The data presented support the derivatization of four of the five peptides chosen for the derivatization experiments. The HPLC chromatograms demonstrate that the PEObiotin derivatized peptides exhibit a shift in retention time, and the MS data indicate a significant change in mass of each of the peptides, except peptide 1, which does not contain cysteine. In addition, peptide 1 cont ains the same amino acids as each of the other peptides, but is still not derivatized with the PEObiotin, indicating the reagent is relatively selective in the ami no acids with which it reacts. Thus, sample derivatization with the biotinylating ICAT r eagent seems feasible, and will most likely result in specific labeling of only the cysteine residues in rhAS.

PAGE 139

120 Detection of ICAT Labeled Peptides From Purified rhAS Purified rhAS samples labeled with light and heavy ICAT and combined in a 1:1 ratio were analyzed by LC/MS/MS on the QSTAR MS. A representative total-ionchromatogram (TIC) is shown in Figure 6.8. Of the seven possible cysteine-containing peptides that are generated by trypsin digestion of rhAS, the 5 peptides that were detected are listed in bold in Table 6.4. 51015202530354045505560 Time, min 0.0 2.0e4 4.0e4 6.0e4 8.0e4 1.0e5 1.2e5 1.4e5 1.6e5 1.8e5 2.0e5 2.1e5Intensity, cps 760.47 533.30 549.30 794.43 608.66 720.46 581.33 515.32 407.77 685.39 699.37 268.16 484.79 486.81 683.41 699.40 373.31 717.37 733.36 51015202530354045505560 Time, min 0.0 2.0e4 4.0e4 6.0e4 8.0e4 1.0e5 1.2e5 1.4e5 1.6e5 1.8e5 2.0e5 2.1e5 0.0 2.0e4 4.0e4 6.0e4 8.0e4 1.0e5 1.2e5 1.4e5 1.6e5 1.8e5 2.0e5 2.1e5Intensity, cps 760.47 533.30 549.30 794.43 608.66 720.46 581.33 515.32 407.77 685.39 699.37 268.16 484.79 486.81 683.41 699.40 373.31 717.37 733.36 Figure 6.8. Representative total ion chromatogr am (TIC) of rhAS peptides labeled with light and heavy ICAT. The ion intensity is shown on the y-axis, and retention time on the x-axis. The numbers shown above the peaks represent the m/z of the most intense parent ion during th e elution time of that peak. The detected peptides were iden tified by parent ion mass and MS/MS fragmentation. Identification of the peptid es was made automatically using ProICAT software with comparison of MS/MS spectra to the human protein database. Appearance of an oxidized form of peptide A147-K161 was not surprising. Methionine is a labile residue, easily susceptible to oxi dation due to sample handling.143

PAGE 140

121 Table 6.4. Theoretical ICAT Peptides from Trypsin Digested rhAS. Peptide Sequence MW MW + light ICAT MW + heavy ICAT 1 CGIWALFGSDDCLSVQCLSAMK 22 2346.0523027.44 3054.53 33 FENVNGYTNCCFGFHR 48 1906.7882361.05 2370.08 66 YPYLWLCYNGEIYNHK 81 2074.9612302.09 2311.12 107 GGIEQTICMLDGVFAFVLLDTANK 130 2554.2812781.41 2790.44 147 AMTEDGFLAVCSEAK 161 147 AMoxTEDGFLAVCSEAK 161 1570.701 1586.696 1797.83 1813.83 1806.86 1822.86 203 YHHCR 207 714.302941.43 950.46 252 IGCLLSGGLDSSLVAATLLK 271 1930.0812157.21 2166.24 Peptides detected by LC/MS/MS are shown in bold. In addition to peptide iden tification, relative quantitation was accomplished by the comparison of the MS peak heights for each li ght and heavy-ICAT peptide pair. In the case of the 1:1 mixture of rhAS labeled with light and heavy ICAT reagent, the calculated L:H ratio of the observed peptid es was around 1 (Figure 6.9). +TOF MS: Experiment 1, 44.961 to 45.478 min from 20040728 IDA Sue ICAT 10fold 1to1.wiff 107610781080108210841086108810901092 m/z, amu 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 147Intensity, counts 1080.1789 1084.6926 1084.1922 1079.6802 1080.6810 1085.1942 +TOF MS: Experiment 1, 44.961 to 45.478 min from 20040728 IDA Sue ICAT 10fold 1to1.wiff 107610781080108210841086108810901092 m/z, amu 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 147Intensity, counts 1080.1789 1084.6926 1084.1922 1079.6802 1080.6810 1085.1942 107610781080108210841086108810901092 m/z, amu 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 147Intensity, counts 1080.1789 1084.6926 1084.1922 1079.6802 1080.6810 1085.1942 Figure 6.9. MS spectrum showing the comparis on of peptide I252-K 271 with the light and heavy ICAT labels. Theoretical masses are shown in Table 6.3, and correspond well with the observed +2 char ge state. Theoretical monoisotopic peaks for the light and heavy labeled peptides are 1079.61 m/z and 1084.13, respectively.

PAGE 141

122 Two peptides were not identified in the sample, shown in Table 6.4. The reason these peptides were not dete cted is unknown. It may be due to incomplete digestion (more than 2 miscleavages), or that the actua l sequence is not the same as the predicted sequence. The N-terminal peptide, C1 – K22 has never been observed using any mass spectrometric methods (LC/MS/MS by qQq-TO F or ion trap MS, MALDI-TOF MS, or ESI-FTICR MS). However, the N-terminal cysteine residue is involved in the glutaminase active site of this enzyme, and if the N-terminus were missing or modified, the protein would not retain its glutaminase activity.144 The sample analyzed was found to retain its glutaminase activity (unpublishe d results, Jemy Gutierrez), so it was assumed that the N-terminus was intact. Additionall y, a sample of purified rhAS was submitted to the ICBR Protein Core Facility at the Un iversity of Florida for automated Edman degradation to verify the N-terminal sequence. The results supported the presence of the N-terminal cysteine residue, and confirme d the amino acid sequenc e through the first 13 residues (data not shown). The cysteine-containing peptides that were not observed by MS after ICAT sample processing may be poorly ionizing peptides. We can not exclude the possibility of ion suppression. The first peptide, C1-K22, ha s a calculated pI of 4.2, while the second peptide, G107-K130 has a calcula ted pI of 4.0. These isoele ctric points are a bit low, requiring the pH of the mobile phase to be sufficiently lower than pH 4.0 to cause the peptides to be positively charged. The acid used in the mobile phase for HPLC separation was acetic acid at a volume pe rcentage of 0.1%, resulting in a pH of approximately 3.1. While this pH is lower than 4.0, perhaps it was not low enough to adequately charge the peptide for ionization and MS detection. Additionally, both of

PAGE 142

123 these peptides are quite large, over 2.5 kDa with the addition of the ICAT labels. It is possible that detection of su ch peptides by MS, without us ing extreme measures for ion targeting and isolation, is nearly impossible under the selected conditions. Spike in study of 1% rhAS into MOLT-4 S cytosolic protein samples In order to determine the effect of increas ed sample complexity on the detection of rhAS using the ICAT method, the MOLT-4 S cy tosolic protein sample was spiked with purified rhAS to make up 1% and 0.1% of to tal protein. MOLT-4 S cells are believed to express extremely low amounts of AS, and of ten the presence of the protein is only faintly detected using western blotting.16 The sample of MOLT-4 S spiked with 1 g rhAS (1% spike) resulted in detection of onl y 2 of the 5 previously detected cysteinecontaining peptides found in rhAS. Peptides F33-R48 and I252-K271 were both detected by parent ion and MS/MS iden tification, and were also quantified. The ratio of H:L labeled peptides was calculated to have di fferent values, depending on which peptide was used (Figures 6.10 and 6.11). Peptide F33-R 48 had a H:L ratio of 3.5:1, while peptide I252-K271 had a H:L ratio of 8.2:1. While this method appears to work sufficiently in detecting rhAS in the presence of the comple x protein matrix of the MOLT-4 S cytosolic samples, it does not appear to prov ide reliable quantitative data.

PAGE 143

124 +TOF MS: Experiment 1, 44.998 to 45.450 min from 20041104 sue injection 2 again.wiffMax. 544.0 counts. 786.0787.0788.0789.0790.0791.0792.0793.0794.0795.0796.0797.0798.0799.0 0 10 20 30 40 50 60 70 80 90 100 110 120Intensity, counts 794.3949 794.0615 794.7260 +TOF MS: Experiment 1, 44.998 to 45.450 min from 20041104 sue injection 2 again.wiffMax. 544.0 counts. 786.0787.0788.0789.0790.0791.0792.0793.0794.0795.0796.0797.0798.0799.0 0 10 20 30 40 50 60 70 80 90 100 110 120Intensity, counts 794.3949 794.0615 794.7260 Max. 69.0 counts. 100200300400500600700800900100011001200130014001500160017001800 m/z, amu 0 5 10 15 20 25 30 35 40 45 50 55 60 65 69Intensity, counts 120.0828 516.2772 249.1262 277.1228 391.1684 312.1858 186.1171 663.3532 136.0736 459.2542 1002.5561 254.1303 928.9599 499.2408 340.1433728.3772 286.1176 794.4029 474.2534 752.3758985.4877 703.3712 b3 y4 y5 y6 b2 y2 y3 y7 y9 y8F E N V N G Y T N C C F G F H Rb2 b1y2y3y4y5y6y7y8y9Max. 69.0 counts. 100200300400500600700800900100011001200130014001500160017001800 m/z, amu 0 5 10 15 20 25 30 35 40 45 50 55 60 65 69Intensity, counts 120.0828 516.2772 249.1262 277.1228 391.1684 312.1858 186.1171 663.3532 136.0736 459.2542 1002.5561 254.1303 928.9599 499.2408 340.1433728.3772 286.1176 794.4029 474.2534 752.3758985.4877 703.3712 b3 y4 y5 y6 b2 y2 y3 y7 y9 y8F E N V N G Y T N C C F G F H Rb2 b1y2y3y4y5y6y7y8y9 Figure 6.10. MS spectrum (top panel) and MS /MS spectrum (bottom panel) of rhAS peptide F33-R48 detected in 1% rhAS spike sample. The calculated H:L ratio of the peptide is 3.5:1. The MS/MS sp ectrum shows the sequence coverage of the peptide.

PAGE 144

125 +TOF MS: Experiment 1, 85.195 to 86.085 min from 20041104 sue injection 2 again.wiffMax. 1030.5 counts. 719.0720.0721.0722.0723.0724.0725.0726.0727.0728.0729.0 m/z, amu 0 20 40 60 80 100 120 140 160 180 200 220 240Intensity, counts 723.4333 723.0996 723.7674 +TOF MS: Experiment 1, 85.195 to 86.085 min from 20041104 sue injection 2 again.wiffMax. 1030.5 counts. 719.0720.0721.0722.0723.0724.0725.0726.0727.0728.0729.0 m/z, amu 0 20 40 60 80 100 120 140 160 180 200 220 240Intensity, counts 723.4333 723.0996 723.7674 200400600800100012001400160018002000 m/z, amu 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4Intensity, counts 616.4335 715.5196 493.2715 510.3055 545.4114 278.1410 254.1372474.35311002.6726 828.6156 367.2201484.2707 1117.7141 597.3171 353.2354 271.1628 630.3690 456.3006 175.0838 1344.81581431.8240 315.1681 566.3395 871.5128 438.2988 325.2230 689.3729840.4588928.5526 1100.7093 I G C L L S G G L D S S L V A A T L L Ky7y6y4y5 y8y10y11y14y9 y15y16b6 y10 –NH3 ICAT ICAT y4 y5y6y7y8y9 y10y11y14y15y16 b6 200400600800100012001400160018002000 m/z, amu 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4Intensity, counts 616.4335 715.5196 493.2715 510.3055 545.4114 278.1410 254.1372474.35311002.6726 828.6156 367.2201484.2707 1117.7141 597.3171 353.2354 271.1628 630.3690 456.3006 175.0838 1344.81581431.8240 315.1681 566.3395 871.5128 438.2988 325.2230 689.3729840.4588928.5526 1100.7093 I G C L L S G G L D S S L V A A T L L Ky7y6y4y5 y8y10y11y14y9 y15y16b6 y10 –NH3 ICAT ICAT y4 y5y6y7y8y9 y10y11y14y15y16 b6 Figure 6.11. MS spectrum (top panel) and MS /MS spectrum (bottom panel) of rhAS peptide I252-K271 detected in 1% rh AS spike sample. The MS spectrum shows the +3 charge state of light an d heavy ICAT labeled peptide I252-K271 with a calculated H:L ratio of 8.2:1. The MS/MS spectrum shows the sequence coverage of the peptide.

PAGE 145

126 Spike in study of 0.1% rhAS into MOLT-4 S cytosolic protein samples Analysis of control MOLT-4 S cytosolic protein sample compared to MOLT-4 S cytosolic protein sample spiked with 0.1% ( 100 ng) rhAS also resulted in identification and quantitation of the same two peptides . The values determined for relative quantitation were 1.1 for peptide F33-R48 a nd 1.5 for peptide I252-K 271, and all spike-in ratios are summarized in Table 6.5. Table 6.5. Ratios of ICAT Labeled Peptid es from Spike-In Experiments. Peptide H:L ratio for 1% rhAS Spike H:L ratio for 0.1% rhAS Spike 33 FENVNGYTNCCFGFHR 48 3.5:1 1.1:1 252 IGCLLSGGLDSSLVAATLLK 271 8.2:1 1.5:1 The rhAS-spiked MOLT-4 S sample was always labeled with the heavy ICAT reagent. While the calculated H:L ratio for each of the peptides decreases with the decreasing rhAS spike percentage, the significance of each H:L ratio determined can not be established. It is uncl ear why peptide I252-K271 consis tently demonstrates a higher ratio. Whether it is caused by unequal labeli ng efficiency of the peptides by the ICAT reagents or differences in peptide generation from the digestion step, it is impossible to tease apart these steps and determine a singula r cause. These result s lead to questions regarding labeling efficiency of low-abundan ce proteins in complex mixtures and the dynamic range of the method for detecti ng changes greater than 10-fold. However, that the peptides were both iden tified and quantified with such a small spike-in amount (0.1%) is very promising. At the time of these anal yses, it was estimated that the percentage of AS in the MO LT-4 R samples was 0.2%, based on Western blotting results.145 Thus, ICAT analysis of MOLT-4 S and R samples could likely result in the detection and relati ve quantitation of AS.

PAGE 146

127 Results of ICAT analysis of MOLT4 S and R cytosolic protein samples After precipitation, 100 g samples of MOLT-4 S and R cytosolic proteins were labeled with the light and heavy ICAT reag ent, respectively, and processed according to the protocol. Initially, a 5 g aliquot of the avidin-separated peptides was analyzed by LC/MS/MS, and none of the AS peptides were identified. The sample was re-analyzed after modification of the MS method to include the m/z for each ICAT-labeled AS peptide that might be present. Normall y, only the two most abundant parent ions detected in a full MS scan are selected for MS/MS fragmentation. If any of the peptides of AS were present, but at lower intensity th an other ions in the scan event, they would not be selected for fragmentation, and sequence-specific iden tification would be impossible. After analysis with use of the peak list, only peptide F33-R48 from AS was identified. Unfortunately, only the heavy-labe led version of the peptide was detected, so relative quantitation was not possi ble. It is difficult to draw a conclusion regarding how much AS might be present in the MOLT-4 R samp le based on this result. In the spike-in studies, relative quantitation was successful, so it is understood that the AS peptides were detected in both the heavy (MOLT-4 S + spike) and light (MOLT-4 S only) labeled samples. If AS were present in 1% or le ss of the MOLT-4 R sample , then it still should have been detected in the MOLT-4 S sample (as it was in the spike-in studies), and quantitation would be possible. One explanation for the lack in quantitation is that the difference in AS between the MOLT-4 S and R sample was so large (greater than 10-20 fold) that the mass detector could not simulta neously detect the presence of both the light and heavy labeled peptides because of limits in dynamic range. This hypothesis is refuted by the quality of the MS/MS spectrum of the peptide F33-R48 identified in the

PAGE 147

128 MOLT-4 R sample. The signal to noise of the fragment ions was lower than that seen in the 1% spike-in study (Figures 6.10). An alternative expl anation is that the MOLT-4 R sample contained components that obscured detection of peptides that had been previously detected in the MOLT-4 S spike-in samples.146-148 The result of the ICAT analysis of MOLT-4 S and R cytosolic protein samples for AS was moderately successful: AS was identi fied as a protein present in the MOLT-4 R sample, but the difference in expression as compared to the MOLT-4 S sample was not determined. Changes in protein expression be tween MOLT-4 S and R cell lines The ICAT reagents and experimental desi gn were intended to measure changes in protein expression between two samples, and to provide identification of those proteins. While the principle goal in this chapter was to evaluate the ICAT method for detection of AS in the MOLT-4 samples, a considerable num ber of proteins was identified and their changes in expression as a f unction of L-asparaginase drug treatment were observed. These results are presented in Appendix B. Further analyses of the changes in protein expression are discussed in Chap ter 7 of this dissertation. Conclusions The ICAT method of sample derivatizati on and separation, th rough the use of the affinity tag, is very useful for compara tive analyses of complex protein samples by LC/MS/MS. The derivatization process is straightforward and greatly aids in the reduction of sample complexity after digesti on. It does not, however, seem to be the superlative method for identif ication and relative protein quantitation from complex mixtures. While we were able to identify a large number of proteins (>375), only 80 of these proteins were identified by more than 2 peptides (Appendix B). In addition, AS

PAGE 148

129 was only identified by the presence of one of its peptides, and relative quantitation between the MOLT-4 S and R samples was not accomplished. It is important to note that the protocol re leased with the ICAT reagents was strictly followed in these experiments. The workflow of the procedure calls for cation exchange separation of the peptides imme diately after trypsin digestion. The purpose for this step is to remove the peptides from undigested protein and the reagents used for the derivatization step. Af ter the cation exchange step, the pe ptides are eluted in a single fraction to proceed through to the avidin affi nity separation step. Cation exchange is a powerful separation technique that can resolve large numbers of peptides by their charge, further reducing the sample complexity of the mixture.149-151 In these analyses, the avidin is used as the final separation method prior to reversed-phase LC/MS/MS analysis. The order of these separations results in a larg e variety of ICAT-labeled peptides in one fraction, resulting in decreased protein identification by the LC/MS/MS step.152 The studies in Chapter 7 will explore multi-dimensional separation and fractionation of peptides after ICAT deriva tization, and large-scal e analyses of the MOLT-4 proteome to determine the changes in protein expression as a function of Lasparaginase challenge. Proteomics-based methods do not seem to be well suited for the direct detection of AS in complex sample mixtures. The methods are primarily designed to capture a large population of the proteins expressed by the cells, and work well for this purpose. Unfortunately, without adequate peptide separa tion, detection of a ll peptides by MS is highly unlikely, due to the nature of the detect or. MS detection reli es on ionization of the sample, and while the ionization process is not fully understood, it is assumed that

PAGE 149

130 intrinsic characteristics of different pep tides cause differences in their ionization efficiencies.153 While two peptides may be present in equal amounts, if ion ionizes more readily than the other, it will be detect ed more readily by the mass analyzer. The presence of AS in the MOLT-4 R sample is most likely considered to be “low abundance”, and will probably not be dire ctly detected using proteomics-based approaches.

PAGE 150

131 CHAPTER 7 IDENTIFICATION OF CHANGES IN MO LT-4 S AND R CELL LINES BY ICAT AND LABEL-FREE PROTEOMIC ANALYSES Introduction Preliminary investigations of the isot ope coded affinity tag (ICAT) method of differential labeling of the MOLT-4 sensitiv e (S) and resistant (R) protein samples resulted in detection of several hundred protei ns, with indication of relative changes in expression of these proteins as a function of L-asparaginase drug resi stance (discussed in chapter 6 of this dissertation). While result s from those experiments were informative, the method had been adjusted to prioritize th e detection of peptides from asparagine synthetase (AS), and had not been optimized for total proteomic analyses. Evaluations and improvements on the ICAT experimental methodologies have been conducted since its first publication by S. P. Gygi et al. in 1999.36 Investigations of labeling efficiency were carried out and found to be suitable unde r the conditions described in the labeling kit, resulting in full derivatization of cy steine residues in control samples with no evidence of non-specific labeli ng of additional amino acids.66, 154 In the printed materials released with the ICAT kit from Applied Bios ystems, the cation exchange step is carried out immediately after digestion of the sample w ith trypsin. The purpose of this step is to remove the reagents and denaturants used from the derivatization step and to remove any undigested protein, as well as tr ypsin from the sample, and all peptides are resolved into a single fraction. The avidin affinity step then isolates the biotinylated peptides from the unlabeled peptides in a single fraction, which is then subjected to acid treatment to cleave

PAGE 151

132 the biotin moiety from the ICAT labels, and the samples are analyzed by LC/MS/MS. This work flow of sample manipulation allows one to pull small por tions of the sample after each procedure to determine its success. The order of the cation exchange and avidin affinity steps do not suitably fractionate peptides or remove contaminants from the sample matrices to promote better sensitivity and detection of ICAT-derivatized peptides. In 2004, L.R. Yu et al. investigated the effect of removi ng the cleaved biotin moiety from the ICAT-derivatized peptides af ter avidin-affinity separation.154 The results indicated that a cation exchange step suitably rem oved the biotin moiety from the peptides, reducing the number of reagent contaminants and increasing the overall signal to noise ratio of the base peak chromatogram. Al so in 2004, K. A. Conrads and coworkers described an alteration to the work flow of th e ICAT procedure, resu lting in detection of over 2500 unique proteins.110 This alteration was a change in the order of the avidin affinity step, placing it prior to strong cati on exchange (SCX) frac tionation of the ICATlabeled peptides. The SCX fractions were separately analyzed by reversed-phase LC/MS/MS, resulting in greatly improved dete ction and identification of peptides. Recently, label-free proteomic i nvestigations have been explored to assess the need for sample derivatization or addition of isotope standards in quantitative studies of protein expression.155-158 This method of label-free proteomic quantitation has progressed from relying on an internal standard for sample normalization155, 156 to the direct comparison of separate LC/MS/MS anal yses of two samples, with the relative changes in peptide abundances observed through changes in mass spectral peak intensities and spectral counts (number of MS/MS spectra ob served for each peptide).157,

PAGE 152

133 158 Relative quantitation methods that do not subject the sample to excessive handling limit the problems associated with derivatiza tion and are more conducive to automation. The goal of this work was to inves tigate, using current proteomic-based technologies, the changes in protein expressi on that occur in MOLT-4 cells as a function of treatment with L-asparaginase (ASNase). There is already sufficient evidence that these leukemia cells exhibit an increase in expression of asparagine synthetase in vitro . However, it is unknown whether th is is a primary or secondary effect of the ASNase. While it makes sense that the cells respond to the deprivation of aspa ragine with a means to produce their own asparagine, the drug-resi stant phenotype of the cell may result from changes in a number of prot eins. These studies investig ated the changes in protein expression of the MOLT-4 S and R cells usi ng the ICAT method on bot h the total protein extract and the nuclear protein fraction. A significant focus was placed on the nuclear protein fraction because of the number of tran scription factors that are present in low abundance in the cell, which are usually focuse d in the nucleus. Isol ation of the nuclear proteins would allow an enrichment of th ese factors, which might not normally be identified when present with more abundant pr oteins in the total protein extract. In addition, a label-free experiment was conducte d with the nuclear pr otein fractions to determine how well the ICAT method compares to that of the label-free method, and to ascertain what proteins, if a ny, were not detected by the ICAT method due to the absence of cysteines in the protein sequence. Experimental Procedure Cell Culture The human acute lymphoblastic leukem ia cell line MOLT-4 (ATCC CRL 1582) was propagated in RPMI-1640 medium suppl emented with 10% (v/v) fetal bovine

PAGE 153

134 serum(FBS), 10mL/L ABAM(100U/mL peni cillin, 100ug/mL streptomycin, 0.25ug/mL amphotericin B)(GIBCO, Gaithersburg, MD) an d 30ug/mL gentamycin(Sigma, St. Louis, MO). MOLT-4 parental (MOLT-4 S) cells were maintained in RPMI-1640 medium without any asparaginase (ASNase), while MOLT-4 resistant (MOLT-4 R) cells were maintained in RPMI-1640 medium containi ng 1 U/mL ASNase (MERCK, West Point, PA). All suspension cultures were maintained at 37 ºC in a 5% CO2 incubator (Nuaire, Plymouth MN). Twenty-four hours before a ll experiments, cells were collected by centrifugation for 5 min at 288 x g, rinsed once with phosphate buffered saline (PBS) (0.15 M sodium, chloride, 10 mM sodium phos phate, pH 7.4), and resuspended at a density of approximately 5 X 105 cells/mL in fresh medium. These cells were grown and lysed for nuclear protein recovery by Na n Su in the lab of Dr. Michael Kilberg, Department of Biochemistry, University of Florida. Cell Lysis for Total Protein Recovery Cells were spun to form a pell et and washed twice with 500 L PBS. The supernatant was removed and cel ls were resuspended in 500 L lysis buffer (50 mM Tris-HCl, pH 7.2, 1% Triton X-100, 10 mM NaF, 1 mM EDTA. 1 mM Na3VO4, and 100 M phenylmethylsulphonylfluoride, PMSF). and sonicated twice for 30 seconds each with a 5 minute cooling period in-between sonication. MOLT-4 S and R cell lysates were centrifuged for 5 minutes at 13,000 rpm, and supernatant was collected into separate 0.5 mL centrifuge tubes. Cell Lysis for Nuclear Protein Recovery MOLT-4 S and R cells were washed twi ce in PBS, then PBS was removed and replaced with lysis buffer (20 mM HE PES, pH 7.9, 10 mM NaCl, 1.5 mM MgCl2, 0.2

PAGE 154

135 mM EDTA, 20% glycerol, 1 mM DTT, 1 mM PMSF, Protease Inhibitor Co cktail), and agitated to dissolve any clumps. The cells were incubated on ice for 15 minutes, then NP-40 was added to a final concentration of 0.1%. The cell samples were vortexed for 5 seconds, then immediately centrifuge d at 500 x g for 5 minutes at 4 oC. The supernatant was removed and conserved as they cytoplas mic extract. Two milliliters of nuclear extraction buffer was added to each sample (20 mM HEPES, pH 7.9, 400 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 25% glycerol, 1 mM D TT, 1 mM PMSF, Protease Inhibitor Cocktail), and sample tubes were rotated end over end for 1-2 hours at 4 oC. Samples were then centrifuged at 12,900 x g for 10 minutes at 4 oC. The supernatant, containing the soluble nuclear protein extract was collected in 2.0 mL tubes. Protein Desalt Total protein fractions a nd nuclear protein fractions from MOLT-4 S and R cell lysates were desalted using Excellulose Desalting Columns (Pierce Biotechnology, Inc, Rockford, IL, USA). For each set of experiments two columns were equilibrated with 25 mL of 50 mM ammonium bicarbonate (NH4HCO3), pH 8.3. Cell lysates were loaded onto the two columns, washed with 50 mM NH4HCO3 and eight 0.5 mL aliquots were collected for each sample. Fractions contai ning protein were pooled separately into MOLT-4 S and R protein fractions for each the total protein and nuclear protein experiments. Protein concentration for each sample was determined using the BCA assay (Pierce Biotechnologies, Inc, Rockford, IL). For total protein comp arison, aliquots of 1 mg of each sample of MOLT-4 S and R total protein fractions was removed and lyophilized separately in a v acuum centrifuge. For analysis of the nuclear protein fraction, 500 g of each sample were aliquotted and lyophilized. For analysis of label-

PAGE 155

136 free quantitation of the nu clear protein fraction, 200 g of each sample were aliquotted for digestion with trypsin. Cleavable ICAT Labeling of Total Protein Samples Lyophilized protein samples we re re-dissolved with 250 L 6 M guanidine hydrochloride in 50 mM NH4HCO3, pH 8.3. Tris-carboxyethyl phosphine (TCEP) was added to a final concentration of 10 mM, a nd the samples were boiled for 10 minutes. Cleavable ICAT (cICAT) reagents were purchas ed in bulk so that each vial contained enough reagent to completely labe l 1 mg of protein. Light (13C0) and heavy (13C9) cICAT reagents were dissolved with 40 L acetonitrile. The entire volume of MOLT-4 S sample (252.5 L) was added to the 13C0 vial, and the entire volum e of MOLT-4 R sample (252.5 L) was added to the 13C9 vial. The samples were incubated for 2 hours at 37ºC with agitation. After the derivatiz ation step, the MOLT-4 S and R samples were desalted separately into 50 mM NH4HCO3, pH 8.3, as previously described. Column eluate was collected in 0.5 mL fractions. The protein content of each fraction was determined by combining 10 L of each fraction with 300 L Coomassie Plus Protein Assay reagent (Pierce Biotechnology, Inc, Rockford, IL, USA), and color was allowed to develop. Fractions from each sample containing prot ein were pooled together, and finally the pooled samples of MOLT-4 S and R proteins were combined in a total volume of 4 mL. Cleavable ICAT Labeling of Nuclear Protein Samples Lyophilized samples of nuclear pr otein fractions, each containing 500 g protein, were redissolved in 125 L 6 M guanidine hydrochloride in 50 mM NH4HCO3, pH 8.3. TCEP was added to a final concentration of 10 mM, and the samples were boiled for 10 minutes. Five vials each of light and heavy cICAT reagent were pool ed separately after

PAGE 156

137 dissolving the reagent in each vial with 20 L 100% acetonitrile. The total volume MOLT-4 S nuclear protein was added to the lig ht cICAT vial, and th e total volume of MOLT-4 R nuclear protein was added to the he avy cICAT vial. Samples were incubated for 2 hours at 37 °C and treated as described above. Trypsin Digestion Two 20 g vials of modified trypsin (se quencing grade, Promega) were reconstituted using the combined MOLT-4 S and R total protein cICAT labeled sample and pooled into one vial, resulting in an enzyme to substrate ratio of 1:50 (w/w). For the cICAT labeled nuclear protein samples, one 20 g vial of modified trypsin was reconstituted with the entire combined sample volume. Finally, digestion of the labelfree nuclear protein samples was carried out separately. The MOLT-4 S and R nuclear fractions (200 g each) were each diluted to 350 L with 50 mM NH4HCO3, pH 8.3 and digested with 4 g modified trypsin, which was reconstituted with 50 mM NH4HCO3, pH 8.3. All digests proceeded at 37 °C ove rnight with shaking at 150 rpm. Avidin Affinity Separation of cICAT Labeled Peptides Digested samples were fractionated usi ng avidin affinity chromatography to separate the unlabeled peptides from the cICAT labeled peptides. Avidin columns were prepared from 2 glass pasteur pipettes bloc ked with glass wool and filled with 1.5 mL UltraLink Immobilized Monomeric Avidin (Pierce Biotechnology, Inc, Rockford, IL, USA). Each column was washed with 12 mL 2x PBS (0.2 M sodium phosphate, 0.3 M NaCl, pH 7.2), then blocked with 3 mL of 2 mM D-biotin in 2x PBS. After blocking, the columns were washed with 3 mL of 30% acetonitrile, 0.4% formic acid and reequilibrated with 15 mL 2x PBS to remove bio tin from the reversible binding sites. The

PAGE 157

138 cICAT digest was boiled for 10 minutes, th en cooled to room temperature. Phenylmethylsulphonylfluoride (PMSF) was a dded to a final concentration of 1 mM. Half of the sample (2 mL) was applied to each column and allowed to incubate for 1 hour at ambient temperature. Each column was then washed with 15 mL each of 2x PBS, 1x PBS, and 20% acetonitrile in 50 mM NH4HCO3, pH 8.3. Cleavable ICAT labeled peptides were eluted in 0.5 mL fractions us ing 30% acetonitrile, 0.4% formic acid, for a total of 8 fractions from each column. Pep tide content of each fraction was determined by combining 5 L of each sample with 500 L of BCA working reagent (prepared according to manufacturer’s instructions, Pier ce Biotehcnology, Inc.). Each sample was transferred to a cuvette and heated with wa rm water (~90ºC). Fr actions 3 and 4 from each column eluate showed significant color change and were pooled into approximately 1.5 mL total volume, frozen and lyophilize d. The lyophilized cICAT labeled peptide sample was re-dissolved in the cleaving reagents provided by the manufacturer (190 L reagent A, 10 L reagent B) and allowed to react fo r 2 hours at 37ºC, then lyophilized to dryness. Strong Cation Exchange Fractionation of cICAT-Labeled Peptides Strong cation exchange (SCX) chromatography was used to fractionate the peptides prior to LC/MS/MS analysis. The cICA T-labeled MOLT-4 S and R peptides were dissolved in 100 L 0.1% formic acid. Ninety microliters of the sample was injected onto a polysulfoethyl A cation exchange co lumn (1 mm x 150 mm , PolyLC, Columbia, MD, USA), and separated at 50 L/min according to the following gradient: 3% mobile phase B for 5 minutes, 3-10% B over 35 minut es, 10-60% B in 46 mi nutes, 65-100% B in 1 minute, and held at 100% B for 9 minut es. Mobile phase A was 25% acetonitrile,

PAGE 158

139 mobile phase B was 25% acetonitrile, 0.5 a mmonium formate, pH 3.0. The separation was monitored by fluorescence, with ex = 266 nm and em = 340 nm. Strong cation exchange separation and fractionation were performed by Dr. King C. Chan. Fractions were collected every minute in a 96 well plate, lyophilized to dryness and reconstituted in 20 L of 0.1% TFA. Fractions were pooled into 38 consecutive pairs, from fraction 14 to 89, inclusive. . Strong Cation Exchange Fractionation of Label-Free MOLT-4 Nuclear Protein Digests MOLT-4 S and R nuclear protein digests we re separated by strong cation exchange chromatography, as described above. The re sulting 96 fractions for each sample were lyophilized and reconstituted in 20 L of 0.1% TFA (v/v). Fractions were pooled in groups of 3 between fractions 12 and 89, inclusive. Reversed-Phase Liquid Chromatography ESI MS/MS for cICAT Samples Pooled SCX fractions were se parated by on-line micro RPLC ( RPLC) ESI MS/MS. The end of a length of fused silica capillary, 75 m ID x 360 m OD (Polymicro Technologies, Phoenix, AZ ) was pulled to a fine tip (5-7 m) using a butane torch. The capillary was slurry packed with C18 Magic AQ reversed phase resin, with 5 m bead diameter and 200 Ã… pore size (Microm Bioresources, Inc, Auburn, CA), or Jupiter C18 reversed phase resin with 5 m bead diameter and 300 Ã… pore size (Phenomenex, Torrance, CA) to a bed length of 10 cm, as previously described.110 Solvent flow was supplied by an Agilent 1100 capillary LC system (Agilent Technologies, Palo Alto, CA). After loading 5 L of sample, the column was washed with 2% B for 30 minutes at a flow rate of 0.5 L/min, then the flow rate was decreased

PAGE 159

140 to 0.250 L/min before initiation of the gradient. The gradient was as follows: 2-40% B over 40 minutes, 40-98% B over 30 minutes, where mobile phase A was 0.1% formic acid and mobile phase B was 0.1% formic acid in acetonitrile (v/v). High voltage contact for electrospray ionization was provide d through a metal union connecting the microcapillary column to the LC pump. Th e linear ion trap-Fourier transform ion cyclotron resonance mass spectrometer ( LTQ-FTICR MS, Thermo Electron, San Jose, CA) was operated in data-dependent MS/MS mo de, where an initial parent-ion scan was acquired in the FTICR cell, and the five most intense ions were sequentially selected for collision-induced dissociation (C ID) in the linear ion trap. MS data collection began after the first 15 minutes of the HPLC method, a nd continued for a total of 150 minutes. Reversed-Phase Liquid Chromatography ESI MS/MS for Label-Free Nuclear Protein Digests Pooled SCX fractions were se parated by on-line micro RPLC ( RPLC) ESI MS/MS, as describe above with the following exceptions: The packing material used in the capillary column was Jupite r C18 resin, 300 Ã… pore size, 5 m bead diameter (Phenomenex, Torrance, CA, USA). The gradie nt was as follows: 0-20 minutes, 2% B; 20-160 minutes, 2-42% B; 160-170 minutes, 4298% B; 170-185 minutes, hold at 98% B; 185-190 minutes, 98-2% B; and re -equilibration for 20 minutes at 2% B. Solvent A was 0.1% formic acid in water, and solvent B was 0.1% formic acid in acetonitrile (v/v). The initial flow rate was 0.5 L/min at time zero, and was reduced at 20 minutes to 0.25 L/min. Flow was ramped back to 0.5 L/min at 160 minutes. The LTQ-FTICR was operated in data-dependent mode where the parent ion scan was acquired in the FTICR cell, and the seven most intense ions were se quentially selected for CID in the linear ion trap. In addition, the method was programmed for neutral loss scans to further fragment

PAGE 160

141 peptides that showed evidence of phosphoryla tion. MS data collec tion began after the first 0.1 minutes and continued for a total of 150 minutes. Peptide Identification and Quantitation Raw MS/MS data generated by the LTQ -FTICR MS were searched using SEQUEST against the Homo sapiens prot eome data base (June 2005 release) downloaded from the European Bioinformatics Institute (EBI, http://www.ebi.ac.uk/integr8/QuickSearc h.do?pageContext=201&action=doOrgSearch& geneName=&organismName=sapiens , last accessed April 4, 2006). For the cICAT samples, a static modification was set on cyst eine for mass addition of the light isotope cICAT label (227.13 Da) and a dynamic modifica tion (9.03 Da) was also set on cysteine for the heavy label (236.16 Da). SEQUEST criteria were set as follows: Xcorr 1.9 for [M+H]1+, 2.2 for [M+2H]2+ ions, and 2.9 for [M+3H]3+ ions, and Cn 0.08, for the identification of fully trypt ic, unique cICAT-labeled peptides. Only peptides that specifically identified a single protein were included in this study. Identified peptides were quantified using XPRESS (ThermoElect ron), which calculates the relative abundances of 13C9/13C0 peptide pairs based on the area of the extracted ion chromatograms. The label-free digest samples were also se arched against the Homo sapiens protein data base using SEQUEST. Dynamic m odifications included phosphorylation of tyrosine, serine and threonine. SEQUEST criteria used are described above. Quantitation of changes in the proteins identified in the label-free experiment was accomplished by taking the ratio of the spectra l count for peptides identifying unique proteins in the MOLT-4 R and S samples (Ratio R/S), and by subtrac tive analysis of the

PAGE 161

142 spectral count for the samples, providing an ab solute difference of the two samples, (|RS|). Processing of the data was accomplishe d by D. Aaron Lucas at the Laboratory of Proteomic and Analytical Technologies, SAIC-Frederick, NCI, Frederick, MD. Western Blot Analysis After cell lysis, protein desalt, and protein concentration determination, 50 g of each sample were separated on a 12% resolv ing SDS-PAGE gel with 5% stacking gel, then electrotransferred to a nitrocellulose membrane. The membrane was incubated in blocking solution (5% Carnation nonfat dry milk in 50 mM Tris, pH 7.5, 200 mM NaCl, 0.1% Tween 20) for 2 hours at ambient temperature with shaking. Anti-AS primary antibody (obtained from the Kilberg lab, Universi ty of Florida, Gainesville, Florida) was incubated with the membrane at a dilution of 1:200 in blocking solution for 2 hours at room temperature with shaking. Blots were washed three times for 5 minutes each in blocking solution, then incubated with horse radish peroxidase-conjugated goat antimouse secondary antibody (Pierce Biotechnol ogies, Inc) at a dilution of 1:10,000 for 45 minutes, and washed three times for 15 minut es each with 50 mM Tris, pH 7.5, 200 mM NaCl, 0.1% Tween. The peroxidase-conjugate d antibody was visualized with x-ray film (CL-XPosure film, Pierce Biotechnologies, Inc) after treatment with SuperSignal West Pico chemiluminescent substrate (Pie rce Biotechnologies, Inc), according to manufacturerÂ’s inst ructions. The x-ra y film was photographed and the image was analyzed by UN-SCAN-IT (Silk Scientific Corp oration) to determine the pixel intensity of each of the bands. Results and Discussion A number of studies have been conducte d with the MOLT-4 cell lines to determine what changes occur as a result of treatment with ASNase.4-6 An obvious change that has

PAGE 162

143 received much attention is the increase in both mRNA and protein expression of AS. This response from leukemia cells is believed to play a major role in the formation of the ASNase-resistant phenotype in vitro . This same response can be produced in multiple cancer cell lines by either treatment with AS Nase or amino acid deprivation, suggesting involvement of a more complicated pathway.2, 3, 159-161 While changes in gene expression are im portant to monitor, the relationship between changes in gene expression and ch anges in protein e xpression is not well established.27 Most methods of evaluation of th e MOLT-4 drug-resistan t cell line rely on mRNA arrays or targeted de tection of selected proteins by immunoassays. There have been few studies evaluating the proteome of the MOLT-4 cells;162 however, exploration of the amino acid response pathway, believed to be involved in the increased expression of AS after amino acid deprivation, has resulte d in identification of a number of proteins that may be involved in this response: tr anscription factors, metabolic enzymes, membrane transporters and growth factors.161 Changes in total pr otein expression in the MOLT-4 leukemia cell line were investigated using quantitative techniques in protein identification and detection through mass spectro metry. Particular attention was paid to the nuclear fraction of proteins, which contains a large percentage of transcription factors compared to the cytosolic or total protein fractions. The identification of additional proteins that change as a function of AS Nase treatment may provide new evidence of biological pathways that had been previously undetected. cICAT Analysis of MOLT-4 S a nd R Total Protein Fractions The procedure used in these experiments to evaluate changes in protein expression between the MOLT-4 S and R cell line s is illustrated in Figure 7.1.

PAGE 163

144 Figure 7.1. Schematic representation of th e cICAT workflow. Two cell samples were separately lysed and labeled with the light (12C9) or heavy (13C9) cICAT reagents. The samples were combined and digested, then cICAT labeled peptides were isolated using avidin chromatography. The biotin moiety of the tag wass removed by acid cleavage, and the peptides were fractionated by SCX prior to LC/MS/MS analysis. The first stage of MS was used to quantify the relative abundances of light a nd heavy peptide pairs based on peak intensity while the second stage of MS was used to fragment the ions and identify the sequence from th e fragmentation pattern. The first analysis of total protein ex tracted from the MOLT-4 S and R samples resulted in identification of 1953 unique protei ns (listed in Appendix C). The change in expression of these proteins was evaluate d by the comparison of the peak area of reconstructed ion chromatograms (RICs) for relative quantitation. Over 800 proteins were determined to have H:L ratios that were different from 1.0: 218 proteins had a ratio of 0.57 or less and were considered down-re gulated proteins, while 590 proteins had a

PAGE 164

145 ratio greater than or equal to 1.75 and were c onsidered up-regulated. Of all the proteins detected, 1145 did not exhib it a significant ch ange in expression, indicating the experiment was conducted successfully and ap proximately equal amounts of protein from each sample were analyzed. The modification in the workflow of the IC AT method was greatly beneficial to the number of proteins identified. The SCX fractionation step and subsequent mLC/MS/MS analysis produced a 6-fold incr ease in total proteins identifi ed using this method. Figure 7.2 illustrates the number of peptides and protei ns identified by each set of SCX fractions that was analyzed, which also corresponds well with the relative fluorescence intensity shown in the SCX chromatogram. This dras tic increase in protein identification, as compared to the number of proteins identified when using the manufact urerÂ’s inst ructions for the cICAT method illustrates the importa nce in optimization of sample separation prior to MS analysis. When using the cICAT method inst ructions, over 300 proteins were identified by 1 or more peptides, but with method optimization and SCX fractionation, nearly 2000 unique proteins were identified. This represents more than a 7fold increase in protein detection in the same type of sample. In addition to modification of the method workflow, the use of the LTQ -FTICR MS greatly enhanced the accuracy of the parent ion mass measurements and the speed of acquisition of the MS/MS spectra. The duty cycle of the LTQ-FTICR MS is cons iderably shorter than that of the QSTAR mass spectrometer, allowing for generation of more fragment ion spectra in the same period of time. These two features great ly improved the capacity of the method for peptide detection and protein identification.

PAGE 165

146 Figure 7.2. SCX chromatogram and corresponding histogram of peptides and proteins identified from the MOLT-4 total protein cICAT experiment. The SCX chromatogram (lower panel) reflects th e fluorescence intensity of the cICAT labeled peptides, with th e corresponding histogram (top panel) reflects the number of unique peptides (purple bars) and proteins (blue bars) identified in each of the SCX pooled fractions. The histogram is aligned to the approximate retention time of the fractions collected from the SCX chromatogram. The data generated by the ICAT analysis of total protein are plentiful and remarkably overwhelming. It is difficult to tease apart the importa nce of each of the 808

PAGE 166

147 proteins that exhibited a change in expre ssion as a function of the ASNase challenge. Before even evaluating the proteins with signi ficant changes in expr ession, it is important to recognize the source and function of each of the proteins identified. Figure 7.3 illustrates the distribution of proteins detect ed in the total protein cICAT experiment based on their cellular compartment, molecu lar function and biological process. Figure 7.3. Pie charts illustrati ng the types and location of pr oteins identified from the MOLT-4 S and R cICAT experiment. Th e proteins are categorized into Molecular Function, Cellular Componen t, and Biological Process, based on the definitions of the gene ontology consortium ( www.geneontology.org ). The proteins identified in the cICAT experiment of the MOLT-4 S and R total protein appear to be derived, to a large exte nt, from the cell membrane, with significant (> 45%) extent of the proteins identified as having “transport” listed as a biological process. This categorization provides a degree of focus wh en analyzing the changes in protein expression.

PAGE 167

148 When investigating the prot eins that were identified as involved in molecular transport, several were identified as being up-regulated. One such protein is 4F2 cellsurface antigen heavy chain (lymphocyte activ ation antigen 4F2 large subunit, accession number P08195). This protein was identifie d by 16 peptides and was found to be more abundant in the MOLT-4 R sample by 1.9 ± 0.4 – fold. The function of this protein is molecular transport of branched chain amino acids and it is a calcium:sodium antiporter. This protein exists as a heterodimer with the protein LAT1, a cationic amino acid transporter (accession number Q01650), which wa s also found to be up-regulated in the MOLT-4 R sample by 1.6 ± 0.4 fold, based on the identification of 4 peptides. These two proteins have been implicated in the am ino acid response pathway by providing increased amino acid-regulated transport in nutrient deficient conditions.161 Proteins that have been s uggested to be involved in th e increased expression of AS as a function of amino acid deprivation and/or treatment with ASNase include activating transcription factor 4 (ATF4) and DNAdamage-inducible transcript 3 (DDIT3).161, 163 Neither of these proteins were detected in the MOLT-4 total protein cICAT experiments, nor was AS even detected. There are a number of possible explanations for this outcome. First, the DDIT3 protein does not contain cyst eine residues, a necessary amino acid when using the cICAT methodology. Second, the pro cess of ionization and its relationship to the primary structure of a peptide are not we ll understood, so in the case of ATF4, its peptides may indeed be present in the sample , but due to competition of the analytes for ionization and subsequent MS detection, the pept ide ions may be suppressed as a result of the complexity of the sample. Finally, th e sequence of cysteine-containing peptides arising from the ATF4 protein sequence may not be unique to this enzyme. If the same

PAGE 168

149 sequence exists in a number of ot her proteins or a family of proteins, then it would not be identified as arising from ATF4, and the pr otein would not be identified. The optimum route for determining the presence of partic ular enzymes in samples of considerable complexity is orthogonal sample analysis or targ eted analysis of the proteins in question. Interestingly, AS was also not identifie d in the analysis of the MOLT-4 total protein fractions, even with the multi-dimensional LC separation, high-resolution MS detection, and MS/MS fragmentation. Based on previous work discussed in Chapter 4 of this dissertation, we know that AS is presen t at low femtomole (picogram) quantities in the MOLT-4 S sample, while the MOLT-4 R contai ns 50-100 fold more of this protein. This translates to AS contributing to between 0.00029% (MOLT-4 S) and 0.05% (MOLT-4 R) of the total protein in the samp le, as calculated by quantitative analyses. Even with targeted MS/MS analysis of these cell lines, the abundance of AS is very low, and perhaps its peptides can not compete with the complexity of the sample for ionization and/or detection by MS. Finally the number of proteins uniquely id entified was plotted ve rsus the number of unique peptides used to identify each protein (Figure 7.4). This histogram demonstrates that the majority of proteins were identifie d based on the detection of a single peptide. However, a considerable number of proteins were identified by more than one peptide. In this case, peptides are deemed as unique based on parent ion mass; that is, identical peptides arising from the MOLT-4 S and R samples would be considered two unique peptides because of the difference in the mass of the cICAT tag.

PAGE 169

150 Figure 7.4. Histogram of the number of unique proteins identified versus the number of peptides detected in the MOLT-4 to tal protein cICAT experiment. cICAT Analysis of MOLT-4 S a nd R Nuclear Protein Fractions A similar analysis of the MOLT-4 S and R nuclear protein fractions resulted in identification of a large numb er of proteins. Eight-hundred eighty-five pr oteins were uniquely identified from both the MOLT-4 S a nd R nuclear samples, with 277 proteins showing up-regulation and 80 showing down-re gulation as a function of the ASNase challenge (see Appendix D). A smaller number of overall proteins were detected from the nuclear samples when compared to the total protein experiment, likely due to the decreased amount of protein analyzed (500 g versus 1 mg, respectively, from each the MOLT-4 S and R samples). The categories of proteins identified from the nuclear cICAT experiment are shown in Figure 7.5.

PAGE 170

151 Figure 7.5. Pie charts illustrati ng the types and location of pr oteins identified from the MOLT-4 S and R nuclear protein cICA T experiment. The proteins are categorized into Cellul ar Component, Molecular Function and Biological Process. The majority of the proteins identified in this experiment appear to be located within the cell and in cell membrane regions , but the overall distri bution of molecular functions appears to be the same as th at found in the MOLT-4 total protein ICAT experiment. By isolating the nuclear pr otein fraction, we had hoped to detect and quantify a larger percentage of transcription factors a nd proteins responsible for transcription activation, but this does not appear to be the ca se. The shift in distribution of cellular component is likely due to the enri chment of the nuclear proteins and removal

PAGE 171

152 of cytosolic and cell-surface membrane proteins . A large percentage of proteins were identified that contribute to the biological process of transport, similar to the total protein cICAT results. This may indicate that inde ed, proteins involved in transport play an important role in the cellular response to asparagine depriv ation, and if the proteins are up-regulated, there is an increasing chance they will be positively identified. Asparagine synthetase was positively identif ied in this nuclear cICAT experiment, with a single peptide detected in the MOLT4 R (heavy cICAT labeled) sample. This is an interesting result because th ere is little evidence supporti ng the localization of AS in the nucleus of the cell. During the nuclear pr otein fractionation step, it is possible that isolation of the nuclear proteins from the cytosolic and other prot eins may not be 100% efficient, but it is interesting to note that AS was indeed positively identified in the nuclear protein fraction. A histogram displaying the number of unique proteins identified by peptide number is shown in Figure 7.6. The dist ribution of proteins is very similar to that seen for the total protein MOLT-4 cICAT experiment, demons trating that most pr oteins are identified by a single, unique peptide. When the cICAT results from the total pr otein and nuclear protein fractions from the MOLT-4 cells are compared for protein iden tification, there is a si gnificant overlap in data. The number of proteins identified in both analyses is equa l to 558 (Figure 7.7). This overlap indicates the cI CAT method results in reproduci ble identification of a large number of proteins, and is a su itable technique for broad proteo mic analysis of cell lines.

PAGE 172

153 Figure 7.6. Histogram of the number of unique proteins identified versus the number of peptides detected in the MOLT-4 nuc lear protein cICAT experiment. Figure 7.7. Venn diagram of the overlap in uni que proteins identifie d by ICAT analyses of the total (global) and nuclear prot ein fractions of the MOLT-4 S and R cell lines. Even though the global ICAT anal yses resulted in a larger number of proteins identified, a large portion of them were also identified as present in the nuclear protein fraction.

PAGE 173

154 The results of the cICAT analyses of th e MOLT-4 S and R protein samples have provided a long list of proteins that have b een positively identified in these leukemia cell lines. A number of these proteins have exhi bited a change in expr ession as a function of treatment with ASNase, resulting in an inve ntory of potential biomar kers or targets for further research to better understand the m echanism behind ASNase drug resistance. The MOLT-4 S (parental) cell line that was analyzed in these studies is considered ASNase resistant, because when exposed to hi gh doses of the drug, the cells usually die. The MOLT-4 R cell line that was investigat ed here was sub-cloned from the MOLT-4 parental cell line by in cubation with increasing sub-lethal doses of ASNase. These cell lines have each been passaged (split and dilu ted to grow an increasing number of cells at a low cell/volume density) a number of times, bu t due to the slow growth rate of the MOLT-4 R cell line, the treatment of the cells ha s not been identical. It is not possible to simply assume that all changes in protei n expression are solely due to the ASNase challenge, but instead to survey the list of changed proteins and use it as a reference for further studies. The cICAT method is not desi gned to selectively single out one protein from the thousands present in the system of study, but is instead used to gather a global picture of changes occurring in the cell as th e result of some stimulus, in this case the ASNase. Further exploration of the presence of proteins of interest should be pursued by Western blot analyses or ta rgeted protein identification using quantitative MS methods suitable for measuring changes in protein concentration. Label-Free Analysis of MOLT-4 S and R Nuclear Protein Fractions One major drawback to using the cICAT method is that it is only suitable for the detection of proteins contai ning one or more cysteine re sidues. Fortunately, most proteins, about 95% in the human proteome, based on a search of the Swiss database,

PAGE 174

155 contain at least one cysteine re sidue. However, we have evid ence that even proteins such as AS that contain 10 cysteine residues are often not detected by this method. While use of the cICAT reagents reduces the overall samp le complexity to contribute to the broad variety of peptides and protei ns identified, there are still limitations to the method that cause a number of proteins to go undetected. A label-free proteomic investigation was carried out on the nuclear protein fracti on from the MOLT-4 S and R cell lines to determine if additional proteins could be id entified and relatively quantitated based on the MS spectral count. In these experiments, af ter trypsin digestion of the MOLT-4 S and R nuclear protein fractions, the digests were separately fract ionated by SCX (Figure 7.8). What is interesting to note from this figur e is the reproducibility of the chromatograms from one sample to the other. There shoul d not be large differences in the way the peptides are separated using this method, b ecause the bulk of the peptides should be present in about equal amount s in each sample. Even if hundreds of proteins demonstrated a change in expression from one sample to the other, because of the overall quantity of peptides being analyzed (10s of thousands or more), such changes would not have a large impact on the retention of th e mixture. After SCX fractionation of each sample, the fractions were pooled separate ly in groups of three and analyzed by RPLC/MS/MS using the LTQ-FTICR mass spect rometer. Over 7500 unique proteins were identified as present in the MOLT-4 S or R samples based on the label-free experiment (see Appendix E), an 8.5-fold increa se in the number of proteins identified from the same sample pool using the cI CAT method. Of that list of proteins, 1662 proteins were positively identified by two or more peptides in both the MOLT-4 S and R samples. This indicates a significant overlap in proteins present in both samples.

PAGE 175

156 Figure 7.8. Strong cation exchange chromatogr ams of MOLT-4 S and R nuclear protein digests. Two hundred micrograms of each sample were digested with trypsin and separated by SCX monitored by fl uorescence detection. The overall shape of the chromatograms and peak retention is very similar.

PAGE 176

157 The varieties of proteins identified from the label-free experiment with the nuclear proteins are shown in Figure 7.9. Figure 7.9. Pie charts illustrati ng the types and location of pr oteins identified from the MOLT-4 S and R nuclear protein labelfree experiment. The proteins are categorized into Molecular Function, Cellular Component, and Biological Process. Relative quantitation of protei ns detected in the label-fr ee experiment is slightly more challenging than in the cICAT experime nts. The principle behind quantitation is based on the number of times a protein is iden tified by a unique peptide. Therefore, the number of unique peptides used to identif y a protein from each sample is directly compared (see Appendix E). These data may also imply the abundance of certain proteins in the sample, where proteins th at are identified by a large number of unique

PAGE 177

158 peptides may be present in higher concentration than prot eins that are identified by a smaller number of unique peptides. One way to analyze the data to determin e which proteins are exhibiting a large extent of change between the MOLT-4 S and R sa mples is to identify in each cell sample set the proteins that have been identified by 3 or more unique peptides, and are only present in one of the cell states. This analysis resulted in a list of 31 proteins that are only present in the MOLT-4 R sample, and only 10 proteins that are only present in the MOLT-4 S sample. These data are summar ized in Tables 7.1 and 7.2. What is interesting to note is that there are more pr oteins detected only in the MOLT-4 R sample (Table 7.2), indicating they might be up-regulated. Thes e proteins also seem to have more of a role in cell cycle regulation, DNA repair, and signaling pathways. The proteins identified as only present in the MOLT-4 S sa mple (Table 7.1) are fewer in number, and do not seem to be the same types of protei ns as found in theMOLT-4 R sample. It is possible to deduce from these data that tr eatment of the MOLT-4 leukemia cells with ASNase appears to have more of a stimu lating effect on the expression of certain proteins, which may ultimately result in the drug-resist ant phenotype. Asparagine synthetase was i ndeed detected in the labelfree proteomic investigation of the MOLT-4 S and R nuclear samples. Four unique peptides were identified in the MOLT-4 R sample, while only one peptide wa s identified in the MOLT-4 S sample. This is of particular importance when eval uating the cICAT versus label-free methods, because AS was not detected in the MO LT-4 S nuclear sample using cICAT.

PAGE 178

159 Table 7.1. Proteins Identified in MOLT4 S Nuclear Protein Faction Only. Protein Reference Accession Number Number of Unique Peptides Golgi-specific brefeldin A-resistance guanine nucleotide exchange factor 1 Q92538 5 Keratin, type I cytosk eletal 9 P35527 5 Tetratricopeptide repeat protein KIAA0103 Q15006 5 Tripartite motif protein 22 (RING finger protein 94) Q8IYM9 5 T-cell surface antigen CD 2 precursor P06729 4 Intersectin-1 (SH3 domain-containing protein 1A) Q15811 4 Myosin-5A (Myosin Va) Q9Y4I1 4 OTTHUMP00000016771 Q5VVH4 4 SET-binding protein (SEB) Q9Y6X0 4 ATPase WRNIP1 Q96S55 4 Table 7.2. Proteins Identified in MO LT-4 R Nuclear Protein Faction Only. Protein Reference Accession Number Number of Unique Peptides Myristoylated alanine-rich C-kinase substrate P29966 8 FLJ11200 protein Q6IA77 7 CDK-activating kinase assembly factor MAT1 P51948 6 Anaphase promoting complex subunit 2 (APC2) Q9UJX6 5 Complex I intermediate-associated protein 30, Q9Y375 5 Ubiquitin-specific proteinase 34 Q70CQ2 5 Adaptor protein Q99570 5 KIAA1542 protein (Fragment) Q9P1Y6 5 Selenocysteine-specific elongation factor P57772 5 DNA-repair protein complementing XP-C cells Q01831 5 Apoptotic protease-activating factor 1 (Apaf-1) O14727 4 Putative Polycomb group protein ASXL1 Q8IXJ9 4 COP9 signalosome complex subunit 6 Q7L5N1 4 mRNA decapping enzyme 1B (EC 3.-.-.-) Q8IZD4 4 Early endosome antigen 1 Q15075 4 Separin (EC 3.4.22.49) (Separase) Q14674 4 Glucocorticoid receptor DNA-binding factor 1 Q9NRY4 4 Glutamate--cysteine ligase regulatory subunit P48507 4 Importin-9 (Imp9) (Ran-binding protein 9) (RanbP9) Q96P70 4 Loss of heterozygosity 11 chromosomal region 2 gene A protein (Breast cancer suppressor candidate 1) O00534 4 Dedicator of cytokinesis 11 Q5JSL3 4 OTTHUMP00000017175 Q5TF21 4 Novel gene (KIAA1797) Q5VW36 4 Hypothetical protein FLJ37794 Q8N9C0 4 Megakaryocyte stimulating factor Q92954 4 Leucine rich repeat containing 45 Q96CN5 4 Hypothetical protein C14orf150 (human) Q96FK6 4 Hypothetical protein FLJ13144 (FLJ10415 protein) Q9H8Y5 4 KIAA1204 protein (Fragment) Q9ULL6 4 X-like 1 protein Q9Y485 4 UDP-glucose 6-dehydrogenase O60701 4

PAGE 179

160 As a final comparison of the difference in protein coverage of the cICAT and labelfree proteomic studies of the MOLT-4 S a nd R nuclear protein fractions, Figure 7.10 illustrates the great amount of overlap in pr otein identification between the two methods, and emphasizes the large increase in protei n identification acquire d in the label-free experiments. Figure 7.10. Venn diagram compar ing the proteins identified in the MOLT-4 S and R nuclear protein fractions by cICAT and La bel Free proteomic investigations. Conclusions The proteomic investigations of the MOLT-4 cell lineÂ’s response to ASNase challenge have been conducted to determine wh at proteins exhibit an increase or decrease in expression. These data provide a valuab le foundation for future studies investigating the pathway of ASNase response in the cell, and may provide new protein targets or biochemical pathways through which the AS Nase-resistant phenotype may be better understood.

PAGE 180

161 The large number of proteins identified in these studies illustrates the increasing abilities of protein derivatization, multi-dime nsional separation, and MS detection for pursuing changes in the proteome of cellular samples. The ch oice of cICAT or label-free methodologies provides the researcher with an ample supply of tools through which important biological questions may be answere d. Each method of proteomic analysis has its strengths and weaknesses, and it is up to the investigator to determine what data are important and should be obtaine d, and what data may be wort h sacrificing. While the label-free method provides a greater extent of protein identification, the degree of change in protein expression between samples is not as well quantitated as in the cICAT experiments. Further studies and interpre tation of the label-free methodology may result in greatly enhanced relative quantitation of protein expres sion, but currently the cICAT method is superior. Finally, the detection of AS in the nuclear protein fractions raises the question as to where AS is localized in the cell. To date, th ere are no studies that specifically target the cellular location of the enzyme. Further inves tigation into the locati on of AS before and during drug resistance may offer additional evidence as to why the protein expression changes as a result of amino acid deprivation.

PAGE 181

162 CHAPTER 8 CONCLUDING REMARKS The primary focus of this work was to better understand the enzyme asparagine synthetase (AS) and its role in acute ly mphoblastic leukemia (ALL). The available technologies for detection and quantitation of AS in complex sample mixtures have shown inconsistencies an d disagreements between in vitro and in vivo studies, and even within different in vivo studies, making determination of the function of AS in drug resistance difficult. The research disc ussed here was centered on the technologies currently available to detect AS and to characterize its primary structure using mass spectrometry as the main method of detection. Utility of MS in Proteomics Studies The field of mass spectrometry has evolved over the past several decades, resulting in the development of new hybrid mass anal yzers that are increasingly sensitive and selective for the measurement of target molecules. These technologies have aided in detection of low abundance molecules found in complex mixtures, as long as sufficient sample preparation and separation have been carried out. Mass spectrometry relies on ionization to introduce th e samples into the mass analyzer. Without suitable precautions taken to remove salt and contaminating small molecules, the ionization of peptides can be suppressed. Even if the samples are void of salt, the peptides themselves can compete with one another for ionization. Mass spectrometr y is a powerful tool for the detection of peptides, but gains a wealth of its selectiv ity from the separations techniques that are conducted prior to MS analysis.

PAGE 182

163 A number of mass analyzers are available for analysis of peptide mixtures and proteomics-based samples of extremely comp lex matrices. The important question for proteomics research isnÂ’t what mass analyzer to use, but instead what biological question needs to be answered. Once this question is established, then questions regarding selection of the mass analyzer follow. Is resolution important? Is MS/MS a necessary feature for identification of the target molecules? Is sensitivity in measurement of the molecule a limiting factor? Is there a financial limitation on the research, restricting purchase of instrumentation to a certain leve l? Will on-line separation of the sample be necessary for detection of the analyte? Answ ers to these questions are necessary before suitable MS analysis can be conducted. Fo r example, even though FTICR MS analysis provides the highest resolution av ailable, it is unable to disc riminate between peptides of identical amino acid composition with diffe ring primary structure without MS/MS or fragmentation capabilities. In addition, purchasing and maintenance of FTICR MS instrumentation is expensive, with commercia lly available instruments selling for near $1 million. At the other end of the resolution spectrum, ion trap mass spectrometers have the lowest resolution available among the co mmonly used mass analyzers for protein and proteomics studies. However, the MS/MS fragmentation capabilities provide peptide sequence information, and the determination of site-specific location of post-translational modifications such as oxidation and phosphor ylation. The ion trap mass spectrometers (both 3-D ion trap and linear ion trap) are readily coupled to LC separations, allowing analysis of extremely complex peptide mixtures such as digests of cell lysates. The duty cycle of ion trap mass spectrometers, especially the newer linear ion trap, is sufficiently fast such that during the elution time of a pe ptide of interest, 15-25 data points (MS and

PAGE 183

164 MS/MS spectra) can be accumulated across the profile of the peak. Acquisition of such data is conducive to the development of relative quantitation methods based on stableisotope internal standards. During the next decade, I pred ict there will be continuous improvement in the quantitative measurements made by ion trap mass spectrometers. Characterization of rhAS by Mass Spectrometry The characterization of rhAS was carried out using a number of different mass analyzers. The bulk of the data was provi ded by a 4.7 T ESI-FTICR mass spectrometer. These data verified the primary structure of rhAS, based on that predicted by the cDNA sequence. While MS/MS was not employed in the initial characterization of the sequence of rhAS, the FTICR provided suitable resoluti on and mass accuracy to detect 71% of the sequence without sample separation. Additi onal sample separation, using a simple step gradient of acetonitrile on a bed of C18 resi n, resulted in identifica tion of an additional 20% of the sequence. Without sample fractionation, MALDI-TOF MS provided an additional 6% sequence coverage compared to ESI-FTICR MS. This difference is most likely due to the decreased re solution of MALDI-TO F analyses, where peptides with the same nominal mass could not be resolved using this method. Successful Method Development for the Quan titation of AS Protein in Cancer Cell Lines and Leukemia Patient Samples A method for the quantitation of AS prot ein in cancer cell lines has been developed. This method utilizes a heavyisotope peptide standard for the relative quantitation of endogenous prot ein. Analysis of the samples was conducted on a linear ion trap mass spectrometer, providing MS/MS data and allowing the added selectivity of using fragment ions for the generation of reconstructed ion chro matograms (RICs) for comparison of the heavy-isotope internal standard with the endogenous protein. The

PAGE 184

165 method was able to categorize seven different cancer cell lines into four different ASexpression populations of cells. While initially this may not seem impressive, it is the most direct method of quantita tion of AS protein in cancer ce ll lines currently available. The application of this method to four sa mples of blood from human leukemia patients resulted in quantitative detection of AS, demonstrating agreement with Western blotting analysis. This is the first example of anal ysis of human samples for the presence of AS using MS detection. Other methods of detec tion of AS either provide semi-quantitative data from the protein (Western blotting) or quantitative measure of the mRNA, which encodes the enzyme. The method developed a nd described here provi des direct detection of peptides arising from AS, and through the use of a peptide isotopomer, quantitation is accomplished. Efforts of quantitating proteins using MS have long been under scrutiny, and will remain so for a long time to come. Mass spectrometry is not a method conducive to quantitation. There is no inhe rent feature of molecules that is predicted for ionization and detection. Analogous methods would be absorbance and fluorescence described by extinction coefficients. Until further characteri zation of peptides is conducted to describe the features that contribute to their ionization and detecti on, we must rely on the relative quantitation achieved through the use of h eavy-isotope internal standards. We are fortunate in the studies of peptid es and proteins that the fragmentation patterns generated by CID or ECD are predic table. These fragmentation patterns can further be exploited by examining the ratio of fragment ions w ithin a fragmentation spectrum to determine if there are interf erences in the MS/MS spectrum that will contribute to the incorrect calculation of anal yte. In the example shown here, a suitable

PAGE 185

166 peptide that did not appear to degrade a nd was adequately resolved and detected by MS/MS in control samples exhibited a contaminant fragmentation ion in the MS/MS spectrum that inflated the calculated value of the peptide, based on the internal standard, which did not have such a contaminant. This method of data scrutiny will serve well for robust and reproducible assays, and will prov ide sufficient criteria to ascertain the integrity of the data. Additional efforts are underway to improve the sensitivity a nd reproducibility of the method, which include modifying the LC gr adient used to separate the digested protein samples as well as targeted MS/MS an alysis of the parent ions to increase the number of data points across the peaks as they elute from the LC. With the analysis of human samples to determine the biologically re levant level of AS, it may be possible to remove the SDS-PAGE enrichment step in th e procedure. Substitution of the linear ion trap with a triple quadrupole mass spectrometer, an MS that is well-suited for quantitation by MS, is being considered to increase the selectivity and sensitivity of the method. Evaluation of Proteomics Met hods for Detection of AS There are a number of proteomics based methods available to detect the relative changes in protein expression between two cellular samples. These methods were designed to increase the overall number of pr oteins detected by se lected separation of proteins based on certain unique features of the molecules prior to MS detection. The cell lines we had been analyzing had a known and established difference in the expression of AS; however, the abundance of AS and the extent of change from the MOLT-4 S to the MOLT-4 R samples was not known. Therefore, we used two commonly used proteomics methods to determin e if the changes in AS were detectable, and if the extent of change could be quantified. Unfo rtunately, neither method was

PAGE 186

167 successful. The ICAT method of sample de rivatization and relativ e MS/MS quantitation and detection was only able to identify th e presence of a single AS peptide from the MOLT-4 R sample, and quantitation was not ach ieved. While this method was meant to target Cys-containing peptides (of which AS has 7), AS was not quantified, implying the AS present in the MOLT-4 S sample was below the threshold of the MS detector, or that the method could not relatively quantify large (> 10-fold) changes in protein expression. Several hundred proteins were identified in these samples, but only about 15 of these were found to be significantly upor down-regul ated, with a H:L ratio of >1.5 or <0.5. The 2D-DIGE method was able to resolve a large number of protein spots, but none of the spots that were obviously differe nt in expression from the MOLT-4 S and R samples were found to be identified as AS . Additional 2D-GE experiments suggested that AS was not efficiently transferred thr ough the procedure to the second dimension of electrophoresis. However, a number of protei ns were identified as being upor downregulated as a result of Lasparaginase challenge. Exploration of Protein Expression Be tween MOLT-4 S and R Cell Lines as a Function of L-Asparaginase Challenge Two large scale ICAT experiments were conducted to determine what changes in the global proteome might be linked to the Lasparaginase resistance seen in the MOLT-4 R cell line. In addition, a clos er examination of changes in the nuclear proteins between the MOLT-4 S and R cells was made, both with ICAT derivatization and with label-free analysis of the proteins. The results were overwhelming, with 1953 proteins identified in the ICAT experiment of the total cell lysa te, over 2000 proteins identified from the nuclear fraction using ICAT, and over 1600 proteins identified by the label-free technique. These experiments, conducted with multi-dimensional LC separation and

PAGE 187

168 analyzed with a hybrid linear ion trap-FTI CR mass spectrometer, provided a wealth of data that might aid in unders tanding the changes the cancer ce ll must undergo to adjust to deprivation of the amino acid asparagine by exposure to L-asparaginase. In summary, these experiments were conduc ted to further the information available regarding the cellÂ’s re sponse to L-asparaginase, and to provide additional methods for detection and quantitation of AS in complex mixtures. It is my hope that this work will help provide new, alternative means for unde rstanding the response of leukemic cells to drug treatment.

PAGE 188

169 APPENDIX A TWO-DIMENSIONAL DIFFERENT IAL GEL ELECTROPHORESIS FLUORESCENCE AND SILV ER-STAINED GEL IMAGES Two-dimensional differential imaging gel electrophoresis (2D-DI GE) analysis of MOLT-4 drug-sensitive (S) and drug-resistant (R) cell line s was carried out using 5 different pH gradients in the first dimension (described in Chapter 5). The images shown in figures A.1-A.10 represent the overlaid fl uorescent signals from the fluorescentlylabeled proteins as well as silver stained images of each gel (stained after fluorescence detection). The fluorescence images were us ed to target proteins that had significant changes in expression levels between the MOLT-4 S and R cell lines. Protein spots in figures A.1, A.3, A.5, A.7, and A.10 appearing gr een indicate that prot ein is present in a larger abundance in the MOLT-4 S sample. Pr otein spots appearing red indicate that protein is present in a larger abundance in the MOLT-4 R sample. Proteins present in approximately equal abundances in both samples appear as yellow spots. The silver stained images of each gel (f igures A.2, A.4, A.6, A.8, and A.10) were compared to the corresponding fluorescence imag e, and the spots that were obviously red or green were located on the silver stained ge l image. If the spot was visible, it was excised and labeled (G or R) according to its fluorescence image color. The number associated with each gel spot is based on its location in the gel from left to right. The samples were digested and the proteins iden tified by LC/MS/MS analysis. The data are summarized in Chapter 5, Table 5.2.

PAGE 189

170 Figure A.1. Overlaid fluorescence image of 2D SDS-PAGE with pH gradient 3-11, non linear.

PAGE 190

171 Figure A.2. Silver stained 2D SDS-P AGE with pH gradient 3-11, non linear.

PAGE 191

172 Figure A.3. Overlaid fluorescence image of 2D SDS-PAGE with pH gradient 3-5.6. Improved resolution of the gel spots is demonstrated when comparing this (and the following gels) to the same pH region on the pH 3-11 2D-DIGE gel (Figure A.1).

PAGE 192

173 Figure A.4. Silver stained 2D SDSPAGE with pH gradient 3-5.6.

PAGE 193

174 Figure A.5. Overlaid fluorescence image of SDS-PAGE with pH gradient 5.3-6.7. This gel includes the pH region around wh ere we would expect asparagine synthetase to migrate. There are a number of red spots clearly visible, however many of them were not visibly st ained with silver stain, and were not excised (Figure A.6).

PAGE 194

175 Figure A.6. Silver stained 2D SDSPAGE with pH gradient 5.3-6.7.

PAGE 195

176 Figure A.7. Overlaid fluorescence image of SDS-PAGE with pH gradient 6.2-7.5.

PAGE 196

177 Figure A.8. Silver stained 2D SDSPAGE with pH gradient 6.2-7.5.

PAGE 197

178 Figure A.9. Overlaid fluorescence image of SDS-PAGE with pH gradient 7-11. The cause of the “trails” left by the prot ein spots in the second-dimension is unknown. There were no spots on this ge l that appeared obviously red or green, therefore none were excised for analysis.

PAGE 198

179 Figure A.10. Silver stained 2D SDS-PAGE with pH gradient 7-11. The visible spots in this gel appeared as yellow in the fl uorescence image (fi gure A.9), indicating equal protein abundances from the MOLT-4 S and R cell lines. No spots were analyzed from this gel.

PAGE 199

180 APPENDIX B PROTEINS EXHIBITING CHANGE IN EXPRESSION IN MOLT-4 CELL LINES AS A FUNCTION OF L-ASPARAGINASE CHALLENGE The following list of proteins was compile d from the ICAT experiments discussed in Chapter 6. These proteins were identified and quantitated based on MS/MS fragmentation and comparison of the light and heavy ICAT labeled peptides. The proteins are sorted by the numbe r of peptides found in the sample. The average H:L ratio reflects the average fold-change in pr otein expression between the MOLT-4 drugsensitive (S) and drug-resistant (R) cell lines, as a function of L-asparaginase challenge. Ratios greater than 2 are believed to reflect a significant increase in expression of those proteins, while ratios of 0.5 or less are beli eved to reflect a si gnificant decrease in expression, as a function of L-asparaginase. Th e standard deviation re flects the deviation in the relative quantitation measurements made , if multiple peptides were detected. Table B.1. List of Proteins Identified and Quantitated in MOLT-4 S and R Cytosolic Protein Samples using ICAT. Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|1351907 SERUM ALBUMIN PRECURSOR 14 0.4 0.1 gi|2493464 L-PLASTIN (LYMPHOCYTE CYTOSOLIC PROTEIN 1) 5 3.3 0.0 gi|2500576 HETEROGENEOUS NUCLEAR RIBONUCLEOPROTEIN H' 4 0.5 0.0 gi|135482 TUBULIN BETA CHAIN 4 6.4 0.0 gi|114312 SARCOPLASMIC 4 6.1 0.7 gi|17865718 HEAT SHOCK PROTEIN HSP 90-BETA (HSP 84) 4 5.3 0.0 gi|118102 PEPTIDYL-PROLYL CIS-TRANS ISOMERASE A (PPIASE 4 5.1 gi|1168996 COFILIN, NON-MUSCLE ISOFORM 4 2.5 0.2 gi|112910 ALPHA-2-HS-GLYCOPROTEIN PRECURSOR (FETUIN-A) 4 2.2 1.8

PAGE 200

181 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|6648067 MALATE DEHYDROGENASE, MITOCHONDRIAL PRECURSOR 4 1.6 0.0 gi|129379 60 KDA HEAT SHOCK PROTEIN, MITOCHONDRIAL PRECURSOR 4 1.3 0.0 gi|14548316 HYPOTHETICAL ZINC FINGER PROTEIN KIAA0296 4 0.0 0.0 gi|8134540 LIM 4 0.0 0.0 gi|135464 TUBULIN BETA-3 CHAIN 3 6.3 0.6 gi|3122836 40S RIBOSOMAL PROTEIN S3A 3 4.7 gi|130980 PROFILIN 3 4.4 gi|1174596 TUBULIN BETA CHAIN 3 4.1 1.8 gi|1708596 ADENYLATE KINASE ISOENZYME 2, MITOCHONDRIAL 3 1.6 0.0 gi|3024705 TUBULIN BETA-4 CHAIN 3 0.0 2.1 gi|3334157 PEPTIDYL-PROLYL CIS-TRANS ISOMERASE 3 3.0 0.0 gi|417155 HEAT SHOCK COGNATE PROTEIN HSP 90BETA 3 2.8 3.0 gi|13124489 HETEROGENEOUS NUCLEAR RIBONUCLEOPROTEIN D0) 3 4.4 0.0 gi|6226865 SERINE HYDROXYMETHYLTRANSFERASE, MITOCHONDRIAL PRECURSOR 3 3.7 gi|6094272 D-3-PHOSPHOGLYCERATE DEHYDROGENASE) 3 3.5 3.0 gi|112983 ASPARTATE AMINOTRANSFERASE, MITOCHONDRIAL PRECURSOR 3 2.0 0.0 gi|131806 RAS-RELATED C3 BOTULINUM TOXIN SUBSTRATE 2 3 1.6 0.9 gi|462384 ISOCITRATE DEHYDROGENASE [NADP], MITOCHONDRIAL PRECURSOR 3 1.4 0.0 gi|117759 CYTOCHROME C1, HEME PROTEIN, MITOCHONDRIAL PRECURSOR 3 1.3 0.0 gi|2507249 TRYPSINOGEN, CATIONIC PRECURSOR 3 0.9 0.0 gi|1730229 GUANINE NUCLEOTIDE-BINDING PROTEIN G(I), ALPHA-2 SUBUNIT 3 0.9 0.2 gi|231469 ALPHA-2-HS-GLYCOPROTEIN PRECURSOR (FETUIN-A) 3 0.8 0.0 gi|12585192 BACULOVIRAL IAP REPEAT-CONTAINING PROTEIN 6 3 0.0 0.0 gi|6225584 CAMP-DEPENDENT PROTEIN KINASE TYPE II REGULATORY CHAIN 3 0.0 0.0 gi|3915204 EXCINUCLEASE ABC SUBUNIT 3 0.0 0.0 gi|6094511 TUMOR NECROSIS FACTOR RECEPTOR TYPE 1 ASSOCIATED DEATH DOMAIN PROTEIN 3 0.0 0.0 gi|129694 PROLIFERATING CELL NUCLEAR ANTIGEN 2 6.7 0.0 gi|129902 PHOSPHOGLYCERATE KINASE 1 2 5.6 0.0 gi|120649 GLYCERALDEHYDE 3-PHOSPHATE DEHYDROGENASE, LIVER 2 5.4 1.1 gi|730318 POTENT HEAT-STABLE PROTEIN PHOSPHATASE 2A INHIBITOR I1PP2A 2 5.0 0.0

PAGE 201

182 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|1170739 L-LACTATE DEHYDROGENASE B CHAIN (LDH-B) 2 4.8 0.7 gi|132559 DOLICHYL-DIPHOSPHOOLIGOSACCHARIDE-PROTEIN GLYCOSYLTRANSFERASE 67 KDA SUBUNIT PRECURSOR 2 4.8 3.0 gi|124945 INTEGRIN ALPHA-4 PRECURSOR 2 4.5 2.0 gi|12643329 PROTEIN CGI-51 2 4.4 0.0 gi|462494 L-LACTATE DEHYDROGENASE (LDH) 2 4.3 0.0 gi|3915092 TUBULIN ALPHA-2 CHAIN (ALPHA-II TUBULIN) 2 4.3 0.0 gi|6016267 HEAT SHOCK PROTEIN HSP 90-ALPHA (HSP 86) 2 3.0 4.3 gi|1706611 ELONGATION FACTOR TU, MITOCHONDRIAL PRECURSOR (P43) 2 2.8 0.0 gi|2506254 T-PLASTIN 2 2.6 0.0 gi|136062 TRIOSEPHOSPHATE ISOMERASE (TIM) 2 2.6 1.7 gi|267126 THIOREDOXIN 2 2.4 0.0 gi|14285643 MITOCHONDRIAL PRECURSOR PROTEINS IMPORT RECEPTOR 2 2.4 0.0 gi|115702 CATALASE 2 2.3 1.2 gi|2500530 PROBABLE ATP-DEPENDENT RNA HELICASE P47 2 2.2 0.0 gi|728810 ADP,ATP CARRIER PROTEIN, FIBROBLAST ISOFORM 2 1.9 1.2 gi|112985 ASPARTATE AMINOTRANSFERASE, MITOCHONDRIAL PRECURSOR 2 1.8 0.0 gi|2851395 ENOYL-COA HYDRATASE, MITOCHONDRIAL PRECURSOR 2 1.7 0.0 gi|544074 CATECHOL-O-METHYLTRANSFERASE, SOLUBLE FORM (S-COMT) 2 1.7 0.0 gi|17368369 PROBABLE G-PROTEIN-COUPLED RECEPTOR MTH-LIKE 12 PRECURSOR (METHUSELAHLIKE 12 PROTEIN) 2 1.6 0.0 gi|3023739 EXOSTOSIN-2 (PUTATIVE TUMOR SUPPRESSOR PROTEIN EXT2) 2 1.6 0.0 gi|6225175 CELLULAR NUCLEIC ACID BINDING PROTEIN 2 1.6 0.0 gi|129729 PROTEIN DISULFIDE ISOMERASE PRECURSOR (PDI) (PROLYL 4-HYDROXYLASE BETA SUBUNIT) 2 1.6 0.0 gi|1172552 VOLTAGE-DEPENDENT ANION-SELECTIVE CHANNEL PROTEIN 1 (VDAC-1) 2 1.6 0.0 gi|2497825 DNA REPLICATION LICENSING FACTOR MCM6 (MIS5 HOMOLOG) 2 1.4 0.0 gi|2493964 XANTHINE DEHYDROGENASE (XD) (ROSY LOCUS PROTEIN) 2 1.3 0.0 gi|3121839 CALMEGIN PRECURSOR 2 1.2 1.2 gi|5902732 ACYL CARRIER PROTEIN, MITOCHONDRIAL PRECURSOR (ACP) 1 + 1mod 1.2 0.0 gi|729704 DNA-BINDING PROTEIN HEXBP 2 1.1 0.0 gi|123668 HEAT SHOCK PROTEIN HSP 90-ALPHA 2 1.1 0.0

PAGE 202

183 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|120996 GUANINE NUCLEOTIDE-BINDING PROTEIN G 2 1.0 0.0 gi|136692 UBIQUINOL-CY TOCHROME C REDUCTASE COMPLEX 11 KDA PROTEIN 2 1.0 0.0 gi|12643558 52 KDA REPRESSO R OF THE INHIBITOR OF THE PROTEIN KINASE (P58IPK-INTERACTING PROTEIN) 2 1.0 0.0 gi|13633675 RHO-RELATED GTP-BINDING PROTEIN RHOE (RHO8) (RND3) 2 1.0 0.0 gi|2499862 TRYPSIN I-P1 PRECURSOR 2 0.9 0.0 gi|114776 BETA-2-MICROGLOBULIN PRECURSOR 2 0.8 0.0 gi|3122811 40S RIBOSOMAL PROTEIN S2 2 0.5 0.0 gi|133875 40S RIBOSOMAL PROTEIN S20 2 0.0 0.0 gi|1345615 BONE MORPHOGENETIC PROTEIN 1 HOMOLOG PRECURSOR (SUBMP) 2 0.0 0.0 gi|1708543 HISTIDINE TRIAD NUCLEOTIDE-BINDING PROTEIN (PROTEIN KINASE C INHIBITOR 1) 2 0.0 0.0 gi|137884 HYPOTHETICAL GENE 2 PROTEIN 2 0.0 0.0 gi|136097 TROPOMYOSIN 5, CY TOSKELETAL TYPE 2 0.0 0.0 gi|136096 TROPOMYOSIN, CYTOSKELETAL TYPE (TM30NM) 2 0.0 0.0 gi|417677 60S RIBOSOMAL PROTEIN L9 2 gi|3023362 BRANCHED-CHAIN AMINO ACID AMINOTRANSFERASE 2 gi|729000 CARG-BINDING FACTOR-A (CBF-A) 2 gi|113607 FRUCTOSE-BISPHOSPHATE ALDOLASE A 2 gi|1346126 GLIAL CELL LINE -DERIVED NEUROTROPHIC FACTOR PRECURSOR 1 9.6 0.0 gi|14916886 HYPOTHETICAL PROTEIN AQ_1824 1 8.9 0.0 gi|2499655 ACTIVIN RECEPTOR TYPE I PRECURSOR (ACTR-I) 1 8.4 0.0 gi|6166214 HEAT SHOCK PROTEIN HSP 90-BETA 1 7.2 0.0 gi|1351144 ALANYL-TRNA SYNTHETASE (ALANINE-TRNA LIGASE) 1 7.1 0.0 gi|266394 INTEGRIN ALPHA-4 PRECURSOR (INTEGRIN ALPHA-IV) (VLA-4) (CD49D) 1 6.5 0.0 gi|8928248 TELOMERASE-BINDI NG PROTEIN P23 (HSP90 CO-CHAPERONE) 1 6.4 0.0 gi|120643 GLYCERALDEHYDE 3-PHOSPHATE DEHYDROGENASE, 1 6.3 0.0 gi|1170955 MACROPHAGE MI GRATION INHIBITORY FACTOR (MIF) 1 6.1 0.0 gi|11135060 TYROSYL-TRNA SYNTHETASE (TYROSINE-TRNA LIGASE) (TYRRS) 1 6.0 0.0 gi|121027 GUANINE NUCLEOTIDE-BINDING PROTEIN BETA SUBUNIT-LIKE PROTEIN 12.3 (P205) 1 5.6 0.0 gi|1346664 NUCLEAR AUTOANTIGENIC SPERM PROTEIN (NASP) 1 5.3 0.0 gi|1709928 MULTIFUNCTIONAL PROTEIN ADE2 1 5.3 0.0 gi|5915835 CYTOCHROME P450 71D6 1 5.1 0.0

PAGE 203

184 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|586078 TUBULIN BETA CHAIN 1 5.1 0.0 gi|2828192 D-DOPACHROME TAUTOMERASE 1 5.0 0.0 gi|1730088 [Segment 1 of 2] LEUCINE-RICH ACIDIC NUCLEAR PROTEIN 1 5.0 0.0 gi|133825 40S RIBOSOMAL PROTEIN S17 1 5.0 0.0 gi|12230408 POLY(RC)-BINDING PROTEIN 1 (ALPHA-CP1) (HNRNP-E1) 1 4.9 0.0 gi|12230169 MITOCHONDRIAL IMPORT INNER MEMBRANE TRANSLOCASE SUBUNIT TIM13 A 1 4.9 0.0 gi|729422 ALPHA ENOLASE, LUNG SPECIFIC (2PHOSPHO-D-GLYCERATE HYDRO-LYASE) 1 4.9 0.0 gi|1172624 PROFILIN II 1 4.8 0.0 gi|2506440 GLYCERALDEHYDE 3-PHOSPHATE DEHYDROGENASE, MUSCLE (GAPDH) 1 4.7 0.0 gi|17368519 HYPOTHETICAL PROTEIN SLL0765 1 4.6 0.0 gi|231468 ALPHA-2-HS-GLYCOPROTEIN PRECURSOR (FETUIN-A) (GLYCOPROTEIN PP63) 1 4.6 0.0 gi|417246 LACTOYLGLUTATHIONE LYASE 1 4.5 0.0 gi|544337 FOLATE RECEPTOR ALPHA PRECURSOR (FRALPHA) 1 4.5 0.0 gi|6225870 PALMITOYL-PROTEIN THIOESTERASE PRECURSOR 1 4.4 0.0 gi|136056 TRIOSEPHOSPHATE ISOMERASE (TIM) 1 4.3 0.0 gi|6166599 MYOSIN HEAVY CHAIN, NONMUSCLE TYPE A 1 4.3 0.0 gi|15214271 PROBABLE MITOCHONDRIAL 40S RIBOSOMAL PROTEIN MRP2 1 4.3 0.0 gi|586270 XYLDLEGF OPERON TRANSCRIPTIONAL ACTIVATOR 1 4.3 0.0 gi|400851 PATHOGENESIS-RELATED PROTEIN P2 PRECURSOR 1 4.1 0.0 gi|549335 MAJOR CAPSID PROTEIN L1 1 4.1 0.0 gi|1350995 40S RIBOSOMAL PROTEIN S4 1 4.0 0.0 gi|5915686 Serine 1 4.0 0.0 gi|731652 HYPOTHETICAL 42.3 KDA PROTEIN IN VMA10SRB2 INTERGENIC REGION 1 4.0 0.0 gi|13638229 RAB GDP DISSOCIATION INHIBITOR BETA 1 3.9 0.0 gi|6685360 PROBABLE COBALAMIN BIOSYNTHESIS PROTEIN COBD 1 3.7 0.0 gi|416844 CALRETICULI N PRECURSOR (CRP55) (CALREGULIN) 1 3.7 0.0 gi|12229875 POLYADENYLATE-BINDING PROTEIN 4 1 3.7 0.0 gi|2500916 NUCLEAR HORMONE RECEPTOR NOR-2 1 3.7 0.0 gi|3123175 HYPOTHETICAL 54.2 KD PROTEIN F17H10.3 IN CHROMOSOME 1 3.7 0.0 gi|125731 ATP-DEPENDENT DNA HELICASE II, 80 KDA SUBUNIT (LUPUS KU AUTOANTIGEN PROTEIN P86) 1 3.7 0.0 gi|461429 14-3-3 PROTEIN ZETA 1 3.6 0.0

PAGE 204

185 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|17368027 NUCLEAR HORMONE RECEPTOR FAMILY MEMBER NHR-61 1 3.6 0.0 gi|3122783 40S RIBOSOMAL PROTEIN S12 1 3.6 0.0 gi|1707893 RHO GDP-DISSOCIATION INHIBITOR 2 (RHO GDI 2) (RHO-GDI BETA) 1 3.5 0.0 gi|1171014 40S RIBOSOMAL PROTEIN S27 1 3.5 0.0 gi|2829825 HYPOTHETICAL 33.2 KD PROTEIN IN IBPB 5'REGION 1 3.4 0.0 gi|730667 40S RIBOSOMAL PROTEIN S6 1 3.4 0.0 gi|2500625 DNA-DIRECTED RNA POLYMERASE I 135 KD POLYPEPTIDE 1 3.3 0.0 gi|6015024 DNA POLYMERASE 1 3.3 0.0 gi|124427 INOSINE-5'-MONOPHOSPHATE DEHYDROGENASE 2 1 3.3 0.0 gi|2498106 ALKYLDIHYDROXYACETONEPHOSPHATE SYNTHASE PRECURSOR 1 3.1 0.0 gi|118853 DNA POLYMERASE 1 3.1 0.0 gi|549387 NONSTRUCTURAL RNA-BINDING PROTEIN 35 (NS35) (NCVP3) 1 3.1 0.0 gi|130349 PHOSPHOGLYCERATE MUTASE, BRAIN FORM (PGAM-B) (BPG-DEPENDENT PGAM) 1 3.1 0.0 gi|585465 DNA REPLICATION LICENSING FACTOR MCM2 1 3.0 0.0 gi|2495108 GLUTATHIONE S-TRANSFERASE P 1 3.0 0.0 gi|6225751 NUCLEOSIDE DIPHOSPHATE KINASE A1 1 3.0 0.0 gi|461427 14-3-3 PROTEIN GAMMA (PROTEIN KINASE C INHIBITOR PROTEIN-1) (KCIP-1) 1 2.9 0.0 gi|1730590 HYPOTHETICAL 64.0 KD PROTEIN IN ACS2MPT4 INTERGENIC REGION 1 2.8 0.0 gi|464710 40S RIBOSOMAL PROTEIN S21 1 2.8 0.0 gi|1172479 PHNA PROTEIN HOMOLOG 1 2.8 0.0 gi|1705695 CYTOCHROME C-TYPE HEME LYASE (CCHL) 1 2.7 0.0 gi|1709183 UDP-NACETYLENOLPYRUVOYLGLUCOSAMINE 1 2.7 0.0 gi|417581 [Segment 3 of 4] RAS-RELATED C3 BOTULINUM TOXIN SUBSTRATE 1 (P21-RAC1) 1 2.7 0.0 gi|2842762 5-METHYLTETRAHYDROFOLATE-HOMOCYSTEINE METHYLTRANSFERASE 1 2.7 0.0 gi|3914915 40S RIBOSOMAL PROTEIN 1 2.7 0.0 gi|4033427 PYRUVATE KINASE, M2 ISOZYME 1 2.7 0.0 gi|120165 FIMBRIN 1 2.6 0.0 gi|2493466 I-PLASTIN (INTESTINE-SPECIFIC PLASTIN) 1 2.6 0.0 gi|729167 COMPLEMENT COMPONENT C8 ALPHA CHAIN PRECURSOR 1 2.6 0.0 gi|730052 PHOSPHATE CARRIER PROTEIN, MITOCHONDRIAL PRECURSOR (PTP) 1 2.6 0.0 gi|127287 MITOCHONDRIAL PROCESSING PEPTIDASE ALPHA SUBUNIT, MITOCHONDRIAL PRECURSOR (ALPHA-MPP) (P-55) 1 2.6 0.0

PAGE 205

186 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|125552 PROTEIN KINASE C, ALPHA TYPE (PKCALPHA) 1 2.6 0.0 gi|3183181 TRANSCRIPTION INTERMEDIARY FACTOR 1BETA (TIF1-BETA) (K RAB-A INTERACTING PROTEIN) (KRIP-1) 1 2.6 0.0 gi|112832 ASPARAGINYL-TRNA SYNTHETASE, CYTOPLASMIC (ASPARAGINE--TRNA LIGASE) 1 2.5 0.0 gi|1170519 INTERFERON GAMMA UP-REGULATED I-5111 PROTEIN PRECURSOR (IGUP I-5111) 1 2.5 0.0 gi|1169974 SERINE HYDROXYMETHYLTRANSFERASE (SERINE METHYLASE) (SHMT) 1 2.5 0.0 gi|729418 PROBABLE ENDONUCLEASE III (DNA(APURINIC OR APYRIMIDINIC SITE) LYASE) 1 2.4 0.0 gi|13878808 TRANSMEMBRANE 9 SUPERFAMILY PROTEIN MEMBER 3 PRECURSOR (SM-11044 BINDING PROTEIN) (EP70-P-ISO) 1 2.4 0.0 gi|6647591 MEMBRANE ASSOCIATED PROGESTERONE RECEPTOR COMPONENT 1 2.4 0.0 gi|6647832 PROGESTERONE MEMBRANE BINDING PROTEIN (STEROID RECEPTOR PROTEIN DG6) 1 2.4 0.0 gi|14286137 MYB PROTEIN 1 2.4 0.0 gi|1352510 CHITOOLIGOSACCHARIDE DEACETYLASE (NODULATION PROTEIN B) 1 2.4 0.0 gi|543922 CALNEXIN PRECURSOR 1 2.4 0.0 gi|2833261 IMPORTIN BETA-1 SUBUNIT (KARYOPHERIN BETA-1 SUBUNIT) (NUCLEAR FACTOR P97) (IMPORTIN 90) 1 2.4 0.0 gi|130228 1-PHOSPHATIDYLINOSITOL-4,5BISPHOSPHATE PHOSPHODIESTERASE DELTA 1 (PLC-DELTA-1) 1 2.4 0.0 gi|6707683 PUTATIVE EUKARYOTIC TRANSLATION INITIATION FACTOR 3 SUBUNIT (EIF-3) 1 2.3 0.0 gi|3915067 ASPARTYL-TRNA SYNTHETASE (ASPARTATE-TRNA LIGASE) (ASPRS) 1 2.3 0.0 gi|1705681 CELL DIVISION CO NTROL PROTEI N 1 2.3 0.0 gi|123805 IG HEAVY CHAIN V-I REGION WOL 1 2.3 0.0 gi|1171733 PROBABLE LIPOPROTEIN NLPC HOMOLOG PRECURSOR 1 2.3 0.0 gi|2829412 CRYPTDIN-16 PRECURSOR 1 2.3 0.0 gi|2500575 HETEROGENEOUS NUCLEAR RIBONUCLEOPROTEI N A1 (HELIXDESTABILIZING PROTEIN) 1 2.2 0.0 gi|1730078 130 KDA LEUCINE-RICH PROTEIN (LRP 130) (GP130) 1 2.2 0.0 gi|1708594 6-PHOSPHOFRUCTOKINASE (PHOSPHOFRUCTOKINASE) 1 2.1 0.0 gi|1707884 C-C CHEMOKINE RE CEPTOR TYPE 8 (C-C CKR8) (CC-CKR-8) (CCR-8) (GPR-CY6) (GPRCY6) 1 2.1 0.0 gi|116006 LEUKOCYTE COMMON ANTIGEN PRECURSOR (L-CA) (CD45 ANTIGEN) (T200) 1 2.1 0.0

PAGE 206

187 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|549809 GTP-BINDING PROTEIN YPTV2 1 2.1 0.0 gi|730510 RAS-RELATED PROTEIN RIC1 1 2.1 0.0 gi|549778 HYPOTHETICAL PROTEIN IN LYC 5'REGION 1 2.1 0.0 gi|730868 GLYCYL-TRNA SYNTHETASE (GLYCINE-TRNA LIGASE) (GLYRS) 1 2.1 0.0 gi|1730213 GUANINE NUCLEOTIDE-BINDING PROTEIN G 1 2.1 0.0 gi|728861 ATRIAL NATRIURE TIC PEPTIDE RECEPTOR A PRECURSOR (ANP-A) (ANPRA) (GC-A) 1 2.1 0.0 gi|728972 PROTEIN BMRU 1 2.0 0.0 gi|1730891 HYPOTHETICAL 137.4 KDA PROTEIN IN BCSADEGR INTERGENIC REGION 1 2.0 0.0 gi|731833 HYPOTHETICAL 41.6 KD PROTEIN IN SDS3THS1 INTERGENIC REGION 1 2.0 0.0 gi|585959 PROTEIN TRANSPORT PROTEIN SEC61 ALPHA SUBUNIT ISOFORM 1 (SEC61 ALPHA-1) 1 2.0 0.0 gi|8488997 NAD(P)-DEPENDENT STEROID DEHYDROGENASE (H105E3 PROTEIN) 1 2.0 0.0 gi|2851390 TRIOSEPHOSPHATE ISOMERASE (TIM) 1 1.9 0.0 gi|2494624 G PROTEIN PATHWAY SUPPRESSOR 1 (GPS1 PROTEIN) (MFH PROTEIN) 1 1.9 0.0 gi|2495218 HEXOKINASE TYPE II (HK II) 1 1.9 0.0 gi|1709381 NADH-UBIQUINONE OXIDOREDUCTASE CHAIN 1 1.9 0.0 gi|5921789 CITRATE SYNTHASE, MITOCHONDRIAL PRECURSOR 1 1.9 0.0 gi|120731 GLUCOSE-6-PHOSPHATE 1-DEHYDROGENASE (G6PD) 1 1.9 0.0 gi|121695 GLUTATHIONE S-TRANSFERASE I (GST-I) (GST-29) (GST CLASS-PHI) 1 1.9 0.0 gi|6165988 VOLTAGE-DEPENDENT R-TYPE CALCIUM CHANNEL ALPHA-1E SUBUNIT 1 1.9 0.0 gi|119960 FERREDOXIN IV (FDIV) (FERREDOXIN, PLANTTYPE) 1 1.8 0.0 gi|2500861 INSECTICIDAL TOXIN DTX12 1 1.8 0.0 gi|118601 DIHYDROPTERIDINE REDUCTASE (HDHPR) (QUINOID DIHYDROPTERIDINE REDUCTASE) 1 1.8 0.0 gi|5915887 CATHEPSIN L-LIKE PRECURSOR 1 1.8 0.0 gi|5921831 CATECHOL O-METHYLTRANSFERASE, MEMBRANE-BOUND FORM (MB-COMT) 1 1.8 0.0 gi|12644235 LEPTOMYCIN B RESISTANCE PROTEIN PMD1 1 1.8 0.0 gi|2501325 TRIOSEPHOSPHATE ISOMERASE (TIM) 1 1.8 0.0 gi|2833272 OLIGOPEPTIDE TRANSPORTER, KIDNEY ISOFORM (PEPTIDE TRANSPORTER 2) 1 1.8 0.0 gi|6647694 2-OXOGLUTARATE DEHYDROGENASE E1 COMPONENT (ALPHA-KETOGLUTARATE DEHYDROGENASE) 1 1.8 0.0 gi|2500067 RAS-RELATED PROTEIN RAB-21 1 1.8 0.0 gi|17368526 MAKORIN 3 (ZINC-FINGER PROTEIN 127) 1 1.8 0.0

PAGE 207

188 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|1176038 HYPOTHETICAL ABC TRANSPORTER ATPBINDING PROTEIN HI1618 1 1.7 0.0 gi|2494897 PERIODIC TRYPTOPHAN PROTEIN 1 HOMOLOG (KERATINOCYTE PROTEIN IEF SSP 9502) 1 1.7 0.0 gi|3122875 D-3-PHOSPHOGLYCERATE DEHYDROGENASE (PGDH) (A10) 1 1.7 0.0 gi|13638202 ATP-DEPENDENT HELICASE HRPA 1 1.7 0.0 gi|15214127 PROBABLE NA(+)-TRANSLOCATING NADHQUINONE REDUCTASE SUBUNIT A (NA(+)TRANSLOCATING NQR SUBUNIT A) 1 1.7 0.0 gi|2500181 TRANSFORMING PROTEIN RHOC/RHOA 1 1.7 0.0 gi|2500183 RAS-LIKE GTP-BINDING PROTEIN RHOA 1 1.7 0.0 gi|124240 INSULIN-LIKE GROWTH FACTOR I RECEPTOR PRECURSOR 1 1.7 0.0 gi|1168789 CATHEPSIN B PRECURSOR 1 1.7 0.0 gi|6094372 SURFEIT LOCUS PROTEIN 1 1.7 0.0 gi|1346187 DOLICHYL-DIPHOSPHOOLIGOSACCHARIDEPROTEIN GLYCOTRANSFERASE PRECURSOR 1 1.6 0.0 gi|1729913 TRANSCRIPTION FACTOR P65 (NUCLEAR FACTOR NF-KAPPA-B P65 SUBUNIT) 1 1.6 0.0 gi|232206 GLUTATHIONE S-TRANSFERASE (GST CLASSMU) 1 1.6 0.0 gi|2497823 DNA REPLICATION LICENSING FACTOR MCM5 HOMOLOG 1 1.6 0.0 gi|2506903 ES1 PROTEIN HOMOLOG, MITOCHONDRIAL PRECURSOR (PROTEIN KNP-I) (GT335 PROTEIN) 1 1.6 0.0 gi|729433 PROTEIN DISULFIDE ISOMERASE A3 PRECURSOR (DISULFIDE ISOMERASE ER-60) 1 1.5 0.0 gi|133320 DNA-DIRECTED RNA POLYMERASE I 135 KD POLYPEPTIDE (A135) (RNA POLYMERASE I SUBUNIT 2) 1 1.5 0.0 gi|1345961 FIBRILLIN 2 PRECURSOR 1 1.5 0.0 gi|117915 CYTOCHROME C5 1 1.5 0.0 gi|13124390 ORIGIN RECOGNITION COMPLEX SUBUNIT 1 1.5 0.0 gi|125396 HOMOSERINE KINASE (HK) 1 1.5 0.0 gi|1709269 NITROGENASE IRON PROTEIN 1 (NITROGENASE COMPONENT II) (NITROGENASE REDUCTASE) 1 1.5 0.0 gi|1717984 PROTEIN U65 1 1.5 0.0 gi|8928571 PROTEIN KINASE C-BINDING PROTEIN NELL1 PRECURSOR (NEL-LIKE PROTEIN 1) (NELRELATED PROTEIN 1) 1 1.5 0.0 gi|3913339 CYTOCHROME P450 CYP12A2 (CYPXIIA2) 1 1.5 0.0 gi|232130 GROWTH-ARREST-SP ECIFIC PROTEIN 1 (GAS1) 1 1.4 0.0 gi|266904 RETINOIC ACID-BINDING PROTEIN I, CELLULAR (CRABP-I) 1 1.4 0.0

PAGE 208

189 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|548749 60S RIBOSOMAL PROTEIN L18 1 1.4 0.0 gi|15214174 NA(+)-TRANSLOCATING NADH-QUINONE REDUCTASE SUBUNIT F (NA(+)TRANSLOCATING NQR SUBUNIT F) 1 1.4 0.0 gi|17369290 RETICULON 4 (NEURITE OUTGROWTH INHIBITOR) (NOGO PROTEIN) (FOOCEN) (NEUROENDOCRINE-SPE CIFIC PROTEIN) 1 1.4 0.0 gi|6225523 ISOCITRATE DEHYDROGENASE [NADP] (OXALOSUCCINATE DECARBOXYLASE) 1 1.4 0.0 gi|548856 40S RIBOSOMAL PROTEIN S3 1 1.4 0.0 gi|2494238 DELT A3,5-DELTA2,4-DIENOYL-COA ISOMERASE, MITOCHONDRIAL PRECURSOR 1 1.3 0.0 gi|13124727 CADHERIN-RELA TED TUMOR SUPPRESSOR PRECURSOR (FAT PROTEIN) 1 1.3 0.0 gi|6225515 TRANSCRIPTION FACTOR 12 (TRANSCRIPTION FACTOR HTF-4 1 1.3 0.0 gi|120816 GAG POLYPROTEIN [CONTAINS: CORE PROTEIN P15; MAJOR CORE PROTEIN P24; NUCLEIC ACID BINDING PROTEIN P1 1 1.3 0.0 gi|3915729 HYPERPLASTIC DISCS PROTEIN (HYD PROTEIN) 1 1.3 0.0 gi|1709335 NEUROGENIC LOCUS NOTCH PROTEIN HOMOLOG PRECURSOR (XOTCH PROTEIN) 1 1.3 0.0 gi|2494205 DYNEIN HEAVY CHAIN, CYTOSOLIC (DYHC) 1 1.3 0.0 gi|2501668 DYSTROPHIN-RELATED PROTEIN 1 1.3 0.0 gi|3914138 NADH-UBIQUINONE OXIDOREDUCTASE 15 KDA SUBUNIT (COMPLEX I-15 KDA) (CI-15 KDA) 1 1.3 0.0 gi|12643945 VOLTAGE-DEPENDENT ANION-SELECTIVE CHANNEL PROTEIN 3 (VDAC-3) (HVDAC3) (OUTER MITOCHONDRIAL MEMBRANE 1 1.3 0.0 gi|9972788 CATALASE 1 1.3 0.0 gi|418146 UBIQUINOL-CY TOCHROME C REDUCTASE COMPLEX CORE PROTEIN 2, MITOCHONDRIAL PRECURSOR 1 1.3 0.0 gi|1352094 CALITOXIN 2 PRECURSOR (CLX-2) (NEUROTOXIC PEPTIDE) 1 1.2 0.0 gi|2494097 NADP-SPECIFIC GLUTAMATE DEHYDROGENASE (NADP-GDH) (NAD(P)HDEPENDENT GLUTAMATE DEHYDROGENASE) 1 1.2 0.0 gi|2500182 RHO-RELATED GTP-BINDING PROTEIN RHO6 (RND1) 1 1.2 0.0 gi|2499316 NADH-UBIQUINONE OXIDOREDUCTASE 13 KDA-B SUBUNIT (COMPLEX I-13KD-B) (CI13KD-B) 1 1.2 0.0 gi|14286105 PLASMA MEMBRANE CALCIUMTRANSPORTING ATPASE 4 (PMCA4) 1 1.2 0.0 gi|1709213 NUCLEAR ENVELOPE PORE MEMBRANE PROTEIN POM 121 1 1.2 0.0

PAGE 209

190 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|3915796 PHOSPHOLIPASE A2 ISOZYME I PRECURSOR (PHOSPHATIDYLCHOLINE 2ACYLHYDROLASE) (PLA2-I) 1 1.2 0.0 gi|9910799 PUTATIVE TRANSCRIPTION FACTOR OVOLIKE 1 (MOVO1) (MOVO1A) 1 1.2 0.0 gi|12643899 HEAT SHOCK FACTOR PROTEIN 4 (HSF 4) (HEAT SHOCK TRANSCRIPTION FACTOR 4) 1 1.1 0.0 gi|138345 M POLYPROTEIN PRECURSOR [CONTAINS: NONSTRUCTURAL PROTEIN NS-M; GLYCOPROTEIN G1; GLYCOPROTEIN G2] 1 1.1 0.0 gi|2499488 PYROPHOSPHATE--FRUCTOSE 6-PHOSPHATE 1-PHOSPHOTRANSFERASE ALPHA SUBUNIT (PFP) 1 1.1 0.0 gi|3915462 HYPOTHETICAL LIPOPROTEIN YFJS PRECURSOR 1 1.1 0.0 gi|730752 DNA-BINDING PROTEIN SMUBP-2 (IMMUNOGLOBULIN MU BINDING PROTEIN 2) (SMUBP-2) 1 1.1 0.0 gi|1708983 THIMET OLIGOPEP TIDASE (ENDOPEPTIDASE 24.15) (MP78) 1 1.1 0.0 gi|133962 40S RIBOSOMAL PROTEIN S17-A (RP51A) 1 1.1 0.0 gi|13431923 TRANSFORMING ACIDIC COILED-COILCONTAINING PROTEIN 2 (ANTI ZUAI-1) 1 1.1 0.0 gi|2497460 PROBABLE GERANYLTRANSTRANSFERASE (FARNESYL-DIPHOSPHATE SYNTHASE) 1 1.1 0.0 gi|461475 ADP,ATP CARRIER PROTEIN, HEART 1 1.1 0.0 gi|14424435 ARL-6 INTERACTING PROTEIN-1 (AIP-1) 1 1.1 0.0 gi|129049 PYRUVATE DEHYDROGENASE E1 COMPONENT 1 1.1 0.0 gi|2494219 DYNEIN 8 KD LIGHT CHAIN, FLAGELLAR OUTER ARM 1 1.1 0.0 gi|1709587 PROLIFERATION-ASSOCIATED PROTEIN 2G4 (PROLIFERATION-ASSOCIATED PROTEIN 1) (PROTEIN P38-2G4) 1 1.1 0.0 gi|120632 G25K GTP-BINDING PROTEIN, BRAIN ISOFORM (GP) (CDC42 HOMOLOG) 1 1.1 0.0 gi|2507302 CELL DIVISION CONT ROL PROTEIN 42 1 1.1 0.0 gi|464535 RAS-RELATED PROTEIN RAC1B 1 1.1 0.0 gi|464611 RHO-RELATED GTP-BINDING PROTEIN RHOG (SID10750) 1 1.1 0.0 gi|729077 CDC42 HOMOLOG 1 1.1 0.0 gi|9972758 CELL DIVISION CONTROL PROTEIN 42 HOMOLOG 1 1.1 0.0 gi|6093898 RAS-RELATED PROTEIN RAB-10 1 1.1 0.0 gi|5915811 SUCCINATE DEHYDROGENASE CYTOCHROME B560 SUBUNIT, MITOCHONDRIAL PRECURSOR (QPS1) (CII-3) 1 1.1 0.0 gi|135686 TRANSFORMING GROWTH FACTOR BETA 3 PRECURSOR (TGF-BETA 3) 1 1.1 0.0

PAGE 210

191 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|543978 CELL FUSION PROTEIN PRECURSOR (SYNCYTIAL PROTEIN) 1 1.1 0.0 gi|549227 REPLICATION PROTEIN E1 1 1.1 0.0 gi|12644089 ZINC FINGER PROTEIN SNAI1 (SNAIL PROTEIN HOMOLOG) (SNA PROTEIN) 1 1.0 0.0 gi|543875 ATP SYNTHA SE GAMMA CHAIN, MITOCHONDRIAL PRECURSOR 1 1.0 0.0 gi|11131487 F-ACTIN CAPPING PROTEIN ALPHA-3 SUBUNIT (CAPZ ALPHA-3) 1 1.0 0.0 gi|126843 METALLOCARBOXYPEPTIDASE INHIBITOR PRECURSOR (MCPI) (CARBOXYPEPTIDASE INHIBITOR) 1 1.0 0.0 gi|6015070 ELONGATION FACTOR G 1 (EF-G 1) 1 1.0 0.0 gi|128258 NITROGENASE IRON-MOLYBDENUM COFACTOR BIOSYNTHESIS PROTEIN NIFE 1 1.0 0.0 gi|130584 POL POLYPROTEIN [CONTAINS: REVERSE TRANSCRIPTASE ; ENDONUCLEASE] 1 1.0 0.0 gi|1703322 ANNEXIN A11 (ANNEXIN XI) (CALCYCLINASSOCIATED ANNEXIN 50) (CAP-50) (56 KDA AUTOANTIGEN) 1 1.0 0.0 gi|1170244 RNA POLYMERASE ASSOCIATED PROTEIN HOMOLOG (ATP-DEPENDENT HELICASE HEPA) 1 1.0 0.0 gi|13878602 DNA MISMATCH REPAIR PROTEIN MUTS 1 1.0 0.0 gi|585398 LIMULUS CLOTTING FACTOR C PRECURSOR (FC) 1 1.0 0.0 gi|1168212 HLA CLASS I HISTOCOMPATIBILITY ANTIGEN, AW-80(A-1) ALPHA CHAIN PRECURSOR 1 1.0 0.0 gi|1730230 GUANINE NUCLEOTIDE-BINDING PROTEIN G(I), ALPHA SUBUNIT 1 1.0 0.0 gi|548412 PYRUVATE DEHYDROGENASE E1 COMPONENT ALPHA SUBUNIT, TESTISSPECIFIC FORM, MITOCHONDRIAL PRECURSOR 1 1.0 0.0 gi|8039798 NK-TUMOR RECOGNITION PROTEIN (NATURAL-KILLER CELLS CYCLOPHILINRELATED PROTEIN) (NK-TR PROTEIN) 1 1.0 0.0 gi|134218 PROACTIVATOR POLYPEPTIDE PRECURSOR [CONTAINS: SAPOSIN A (PROTEIN A); SAPOSIN B (SPHINGOLIPID ACTIVAT 1 1.0 0.0 gi|3914939 SULFATED GLYCOPROTEIN 1 PRECURSOR (SGP-1) (PROSAPOSIN) 1 1.0 0.0 gi|6094466 TISSUE FACTOR PATHWAY INHIBITOR PRECURSOR (TFPI) (LIPOPROTEINASSOCIATED COAGULATION INHIBITOR) 1 1.0 0.0 gi|1170685 PHOSPHORYLASE B KINASE ALPHA REGULATORY CHAIN, LIVER ISOFORM (PHOSPHORYLASE KINASE ALPHA L SUBUNIT) 1 1.0 0.0 gi|547785 PROBABLE SERINE 1 1.0 0.0

PAGE 211

192 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|1705827 CHITIN SYNTHASE 1 (CHITIN-UDP ACETYLGLUCOSAMINYL TRANSFERASE 1) 1 1.0 0.0 gi|1706119 BETA CRYSTALLIN A2 (BETA-A2CRYSTALLIN) 1 1.0 0.0 gi|125302 CREATINE KINASE, FLAGELLAR 1 1.0 0.0 gi|3913675 FERRIC REDUCTASE TRANSMEMBRANE COMPONENT 5 PRECURSOR (FERRICCHELATE REDUCTASE 5) 1 1.0 0.0 gi|3023672 SULFITE REDUCTASE, DISSIMILATORY-TYPE ALPHA SUBUNIT (HYDROGENSULFITE REDUCTASE ALPHA SUBUNIT) 1 0.9 0.0 gi|3041717 GLYCOGEN PHOSPHORYLASE, MUSCLE FORM (MYOPHOSPHORYLASE) 1 0.9 0.0 gi|6226304 HYPOTHETICAL PROTEIN RP006 1 0.9 0.0 gi|6686356 TRANSPOSON TY1 PROTEIN 1 0.9 0.0 gi|1171777 NEUROGENIC LOCUS NOTCH PROTEIN HOMOLOG 1 PRECURSOR (TRANSLOCATIONASSOCIATED NOTCH PROTEIN TAN-1) 1 0.9 0.0 gi|1705455 FORKHEAD PROTEIN G1A (FORKHEADRELATED PROTEIN FKHL2) (TRANSCRIPTION FACTOR BF-2) 1 0.9 0.0 gi|6137308 ZINC FINGER PROTEIN 189 1 0.9 0.0 gi|1708084 EXOGLUCANASE B PRECURSOR (EXOCELLOBIOHYDROLASE B) (1,4-BETACELLOBIOHYDROLASE B) (CBP120) 1 0.9 0.0 gi|731497 NUCLEOPORIN NUP157 (NUCLEAR PORE PROTEIN NUP157) 1 0.9 0.0 gi|135303 CD81 ANTIGE N (26 KDA CELL SURFACE PROTEIN TAPA-1) (TARGET OF THE ANTIPROLIFERATIVE ANTIBODY 1) 1 0.9 0.0 gi|1346159 VIRULENCE PLAS MID INTEGRASE PGP7-D (PROTEIN P-11) 1 0.9 0.0 gi|12643234 GLYCERATE KINASE 1 0.9 0.0 gi|125105 KERATIN, TY PE II CYTOSKELETAL 5 (CYTOKERATIN 5) (K5) (CK 5) (58 KDA CYTOKERATIN) 1 0.9 0.0 gi|122858 HEMAGGLUTININ PRECURSOR [CONTAINS: HEMAGGLUTININ HA1 CHAIN; HEMAGGLUTININ HA2 CHAIN] 1 0.9 0.0 gi|731699 HYPOTHETICAL 49.4 KD PROTEIN IN CDC12ORC6 INTERGENIC REGION 1 0.9 0.0 gi|134180 LEVANSUCRASE AND SUCRASE SYNTHESIS OPERON ANTITERMINATOR 1 0.9 0.0 gi|13632816 PHOSPHATE TRAN SPORT SYSTEM PROTEIN PHOU 1 0.8 0.0 gi|1723501 HYPOTHETICAL 98.4 KDA PROTEIN C582.05C IN CHROMOSOME II 1 0.8 0.0 gi|3334269 NEURAL CELL ADHESION MOLECULE 2 PRECURSOR (N-CAM 2) (RB-8 NEURAL CELL ADHESION MOLECULE) (R4B12) 1 0.8 0.0

PAGE 212

193 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|2499467 COMPLEMENT COMPONENT C9 PRECURSOR 1 0.8 0.0 gi|7388530 HYPOTHETICAL PROTEIN AF2149 1 0.8 0.0 gi|1708158 HEPATOMA-DERIVED GROWTH FACTOR (HDGF) 1 0.8 0.0 gi|17380331 MAWD BINDING PROTEIN (UNKNOWN PROTEIN 32 FROM 2D-PAGE OF LIVER TISSUE) 1 0.8 0.0 gi|1173121 ATP-DEPENDENT RNA HELICASE ROK1 1 0.8 0.0 gi|116512 CLATHRIN HEAVY CHAIN 1 0.8 0.0 gi|1722837 WD40-REPEAT PROTEIN 1 0.8 0.0 gi|3023354 ATP SYNTHASE EPSILON CHAIN, MITOCHONDRIAL 1 0.8 0.0 gi|1723435 HYPOTHETICAL 170.7 KD PROTEIN C56F8.02 IN CHROMOSOME 1 0.8 0.0 gi|1708591 KERATIN, TY PE II CYTOSKELETAL 6 (CYTOKERATIN 6) (CK 6) (K6 KERATIN) 1 0.8 0.0 gi|729478 FERREDOXIN--NADP REDUCTASE, LEAF ISOZYME, CHLOROPLAST PRECURSOR (FNR) 1 0.8 0.0 gi|1352071 BETA-DEFENSIN 7 (BNDB-7) (BNBD-7) 1 0.8 0.0 gi|113576 SERUM ALBUMIN PRECURSOR 1 0.8 0.0 gi|114620 ATP SYNTHASE B CHAIN, MITOCHONDRIAL PRECURSOR 1 0.7 0.0 gi|1175376 PROBABLE EXOSOME COMPLEX EXONUCLEASE 1 0.7 0.0 gi|1351928 METHIONINE AMINOPEPTIDASE 1 PRECURSOR 1 0.7 0.0 gi|1706619 ELONGATION FACTOR TU (EF-TU) 1 0.7 0.0 gi|462375 INSULIN-LIKE GROWTH FACTOR BINDING PROTEIN 4 PRECURSOR (IGFBP-4) (IBP-4) (IGFBINDING PROTEIN 4) 1 0.7 0.0 gi|3914572 RIBULOSE BISPHOSPHATE CARBOXYLASE LARGE CHAIN PRECURSO R (RUBISCO LARGE SUBUNIT) 1 0.7 0.0 gi|136941 DNA HELICASE 1 0.7 0.0 gi|1176623 HYPOTHETICAL 93.9 KD PROTEIN T20B12.6 IN CHROMOSOME III 1 0.7 0.0 gi|119777 COAGULATION FACTOR IX (CHRISTMAS FACTOR) 1 0.6 0.0 gi|2500331 50S RIBOSOMAL PROTEIN L33 1 0.6 0.0 gi|13878603 UDP-N-ACETYLGLUCOSAMINE--NACETYLMURAMYL-(PENTAPEPTIDE) PYROPHOSPHORYL-UNDECAPRENOL NACETYLGLUCOSAM 1 0.6 0.0 gi|13124702 ADENYLATE CYCLASE, TYPE II (ATP PYROPHOSPHATE-LYASE) (ADENYLYL CYCLASE) 1 0.5 0.0 gi|6093676 PERIOD CIRCADIAN PROTEIN 1 0.5 0.0 gi|125117 KERATIN, TYPE II MICROFIBRILLAR, COMPONENT 1 0.4 0.0 gi|13431609 INTEGRIN BETA-5 PRECURSOR 1 0.4 0.0

PAGE 213

194 Table B.1 Continued Accession # Protein Name Peptides Found Avg. H:L Ratio Std. Dev gi|136421 BETA-TRYPTASE PRECURSOR (TRYPTASE 2) 1 0.4 0.0 gi|2506226 CERULOPLASMIN PRECURSOR (FERROXIDASE) 1 0.4 0.0 gi|117516 PHYTOENE DEHYDROGENASE, CHLOROPLAST PRECURSOR (PHYTOENE DESATURASE) 1 0.4 0.0 gi|123892 HEXOKINASE, TYPE I (HK I) (BRAIN FORM HEXOKINASE) 1 0.4 0.0 gi|1705733 CALCIUM-DEPENDENT PROTEIN KINASE, ISOFORM 1 (CDPK 1) 1 0.3 0.0 gi|11132282 ENOYL-[ACY L-CARRIER-PROTEIN] REDUCTASE [NADH 1 0.3 0.0 gi|2497609 LAMININ-LIKE PROTEIN C54D1.5 PRECURSOR 1 0.3 0.0 gi|462406 INHIBIN BETA A CHAIN PRECURSOR (ACTIVIN BETA-A CHAIN) 1 0.3 0.0 gi|14194461 A KINASE ANCHOR PROTEIN 9 (PROTEIN KINASE A ANCHORING PROTEIN 9) (PRKA9) 1 0.2 0.0 gi|3915436 HYPOTHETICAL 53.4 KD PROTEIN IN PRP9NAT1 INTERGENIC REGION 1 0.2 0.0 gi|13878448 ATRIAL NATRIUTERIC PEPTIDE-CONVERTING ENZYME (PRO-ANP-CONVERTING ENZYME) 1 0.2 0.0 gi|114291 ARGININOSUCCINATE SYNTHASE 1 0.2 0.0 gi|126730 MACROPHAGE MANNOSE RECEPTOR PRECURSOR 1 0.2 0.0 gi|416566 ASPARTATE AMINOTRANSFERASE, CYTOPLASMIC 1 0.2 0.0 gi|728994 B-CELL ANTIGEN RECEPTOR COMPLEX ASSOCIATED PROTEIN BETA-CHAIN PRECURSOR 1 0.1 0.0

PAGE 214

195 APPENDIX C RESULTS OF ICAT INVESTIGATION OF MOLT-4 S AND R PROTEINS FROM TOTAL CELL LYSATE Total protein lysate from MOLT-4 S and R (drug-sensitive and dr ug-resistant) cells was investigated using the cleavable isotope-coded affi nity tag (cICAT) method of quantitative proteomic analysis (described in chapter 7). Proteins are listed alphabetically. “Peptide Number” is the numbe r of unique peptides that were detected and used to identify the protein. “AvgOfRatio ” represents the fold change in expression of that protein between the MOLT-4 R (heavylabeled samples) and the MOLT-4 S (light labeled samples). “StDevOfRatio” represents the standard deviat ion of the calculated average H:L ratio. The “Accession” is the ac cession number assigned to the protein in the human database, and can be used to search its sequence ( http://www.expasy.ch/cgibin/sprot-search-ful , last accessed April 4, 2006). Due to the length of this list, it is included as an electronic pdf file (Object 1). Object 1. ICAT Investigation of MOLT-4 S and R Proteins from Total Cell Lysate

PAGE 215

196 APPENDIX D RESULTS OF ICAT INVESTIGATION OF MOLT-4 S AND R PROTEINS FROM NUCLEAR PROTEIN FRACTION The nuclear protein fraction from MOLT-4 S and R (drug-sensitive and drugresistant) cells was investigated using the cleavable isotope-coded affinity tag (cICAT) method of quantitative proteomic analysis (des cribed in chapter 7). Proteins are listed alphabetically. “Peptide Number” is the numbe r of unique peptides that were detected and used to identify the protein. “AvgOfRatio ” represents the fold change in expression of that protein between the MOLT-4 R (heavylabeled samples) and the MOLT-4 S (light labeled samples). “StDevOfRatio” represents the standard deviat ion of the calculated average H:L ratio. The “Accession” is the ac cession number assigned to the protein in the human database, and can be used to search its sequence ( http://www.expasy.ch/cgibin/sprot-search-ful , last accessed April 4, 2006). Due to the length of this list, it is included as an electronic pdf file (Object 2). Object 2. ICAT Investigation of MOLT-4 S and R Proteins from Nuclear Fraction

PAGE 216

197 APPENDIX E PROTEINS IDENTIFIED BY LABEL FREE PROTEOMIC INVESTIGATION OF MOLT-4 S AND R NUCLEAR PROTEINS The nuclear protein fractions from MOLT-4 S and R (drug-sensitive and drugresistant) cells was investig ated using a label free method of quantitative proteomic analysis (described in chapter 7). Prot eins are identified by name and “Accession” number, which can be used to search its se quence in protein data bases (for example: http://www.expasy.ch/cg i-bin/sprot-search-ful , last accessed April 4, 2006). The headings “MOLT4R” and “MOLT4S” refer to the number of unique peptides that were detected and used to identify the protein in each the MOLT-4 S and R samples. Proteins in which the MOLT4R value is greater th an the MOLT4S value may indicate that protein’s elevated presence in the MOLT-4 R sa mple, and vice versa. Due to the length of this list, it is included as an electronic pdf file (Object 3). Object 3. Label-Free Investig ation of MOLT-4 S and R Prot eins from Nuclear Fraction

PAGE 217

198 LIST OF REFERENCES (1) American Cancer Society, "Cancer F acts and Figures, 2005", Altanta, 2005, pp 164. (2) Kiriyama, Y.; Kubota, M.; Takimoto, T. ; Kitoh, T.; Tanizawa, A.; Akiyama, Y.; Mikawa, H. Leukemia 1989, 3, 294-297. (3) Martin, J. K.; Sun, W.; Moraga-Amador, D.; Schuster, S. M.; Wylie, D. E. Amino Acids 1993, 5, 51-69. (4) Hutson, R. G.; Kitoh, T.; Moraga-Amador , D. A.; Cosic, S.; Schuster, S. M.; Kilberg, M. S. American Jour nal of Physiology 1997, 272, C1691-1699. (5) Aslanian, A. M.; Fletcher, B. S.; K ilberg, M. S. Biochemical Journal 2001, 357, 321-328. (6) Aslanian, A. M.; Kilberg, M. S. The Biochemical Journal 2001, 358, 59-67. (7) Minowada, J.; Onuma, T.; Moore, G. E. Journal of the National Cancer Institute 1972, 49, 891-895. (8) Hahn, W. C. Mol ecular Cells 2002, 13, 351-361. (9) Barbosa-Tessmann, I. P.; Chen, C.; Zhong, C.; Schuster, S. M.; Nick, H. S.; Kilberg, M. S. Journal of Bi ological Chemistry 1999, 274, 31139-31144. (10) Barbosa-Tessmann, I. P.; Chen, C.; Zhong, C.; Siu, F.; Schuster, S. M.; Nick, H. S.; Kilberg, M. S. Journal of Bi ological Chemistry 2000, 275, 26976-26985. (11) Haskell, C. M.; Canellos, G. P. Biochemical Pharmacology 1969, 18, 2578-2580. (12) Pieters, R.; denBoer, M. L.; Durian, M. ; Janka, G.; Schmiegelow, K.; Kaspers, G. J. L.; vanWering, E. R.; Weerman, A. J. P. Leukemia 1998, 12, 1344-1348. (13) Dubbers, A.; Wurthwein, G.; Muller, H. J.; Schulze-Westhoff, P.; Winkelhorst, M.; Kurzknabe, E.; Lanvers, C.; Pieters, R.; Kasp ers, G. J. L.; Creutzig, U.; Ritter, J.; Boos, J. British Journal of Haematology 2000, 109, 427-429. (14) Stams, W. A. G.; denBoer, M. L.; Beverl oo, H. B.; Meijerink, J. P. P.; Stigter, R. L.; vanWering, E. R.; Janka-Schaub, G. E.; Slater, R.; Pieters, R. Blood 2003, 101, 2743-2747.

PAGE 218

199 (15) Panosyan, E. H.; Grigorvan, R. S.; Avramis, I. A.; Seibel, N. L.; Gaynon, P. S.; Siegel, E. S.; Fingert, H. J.; Avramis, V. I. Anticancer Research 2004, 24, 11211125. (16) Irino, T.; Kitoh, T.; Koami, K.; Kashima, T.; Mukai, K.; Takeuchi, E.; Hongo, T.; Nakahata, T.; Schuster, S. M.; Osaka, M. Journal of Molecular Diagnostics 2004, 6, 217-224. (17) Krejci, O.; Starkova, J.; Otova, B.; Ma dzo, J.; Kalinova, M.; Hrusak, O.; Trka, J. Leukemia 2004, 18, 434-441. (18) Fine, B. M.; Kaspers, G. J. L.; Ho, M.; Loonen, A. H.; Boxer, L. M. Cancer Research 2005, 65, 291-299. (19) Leslie, M.; Case, M. C.; Hall, A. G.; Coulthard, S. A. British Journal of Haematology 2006, 132, 740-742. (20) Holleman, A.; denBoer, M. L.; deMenezes, R. X.; Cheok, M. H.; Cheng, C.; Kazemier, K. M.; Janka-Schuab, G. E.; Gobe l, U.; Graubner, U. B.; Evans, W. E.; Pieters, R. Blood 2006, 107, 769-776. (21) Stams, W. A. G.; denBoer, M. L.; Ho lleman, A.; Appel, I. M.; BernaBeverloo, H.; vanWering, E. R.; Janka-Schuab, G. T.; Evans, W. E.; Pieters, R. Blood 2005, 105, 4223-4225. (22) McLean, T. W.; Ringold, S.; Neuberg, D.; Stegmaier, K.; Tantravahi, R.; Ritz, J.; Koeffler, H. P.; Takeuchi, S.; Janssen, J. W. G.; Seriu, T.; Bartram, C. R.; Sallan, S. E.; Gilliland, D. G.; Golub, T. R. Blood 2000, 88, 4252-4258. (23) Morrow, M.; Horton, S.; Kioussis, D. ; Brady, H. J. M.; Williams, O. Blood 2004, 103, 3890-3896. (24) Appel, I. M.; denBoer, M. L.; Meijerink, J. P. P.; Veerman, A. J. P.; Reniers, N. C. M.; Pieters, R. Blood 2006. (25) Kaspers, G. J.; Veerman, A. J.; Pieter s, R.; Broekema, G. J.; Huismans, D. R.; Kazemier, K. M.; Loonen, A. H.; Rottier, M. A. British Journal of Cancer 1994, 70, 1047-1052. (26) Liang, P.; Pardee, A. B. Science 1992, 257, 967-971. (27) Greenbaum, D.; Colangelo, C.; Willia ms, K.; Gerstein, M. Genome Biology 2003, 4, 117. (28) Hayakawa, T.; Nakamura, T.; Hattori, F.; Mae, T.; Ojima, K.; Yamaya, T. Planta 1994, 193.

PAGE 219

200 (29) Nakano, K.; Suzuki, T.; Hayakawa, T. ; Yamaya, T. Plant & Cell Physiology 2000, 41, 874-880. (30) Dickinson, J.; Fowler, S. J. Quantific ation of Proteins on Western Blots using ECL., 2nd Edition ed.; Humana Press Inc.: Totowa, NJ, 2002. (31) Luehr, C. A.; Schuster, S. M. Jour nal of Biochemical and Biophysical Methods 1980, 3, 151-161. (32) Unnithan, S.; Moraga, D. A.; Schuste r, S. M. Analytical Biochemistry 1984, 136, 195-201. (33) Romagni, J. G.; Dayan, F. E. Journa l of Agriculture and Food Chemistry 2000, 48, 1692-1696. (34) Ciustea, M.; Gutierrez, J. A.; Abbatiell o, S. E.; Eyler, J. R.; Richards, N. G. J. Archives of Biochemistry and Biophysics 2005, 440, 18-27. (35) Andrulis, I. L.; Chen, J.; Ray, P. N. Molecular and Ce llular Biology 1987, 7, 24352443. (36) Gygi, S. P.; Rist, B.; Gerber, S. A.; Tu recek, F.; Gelb, M. H.; Aebersold, R. Nature Biotechnology 1999, 10, 994-999. (37) Cohen, J. Technology Review 2001, 104, 54-61. (38) Sperling, K. Electrophoresis 2001, 22, 2835-2837. (39) O'Farrell, P. H. Journal of Biological Chemistry 1975, 250, 4007-4021. (40) Anderson, N. L.; Anderson, N. G. Molecular and Cellular Proteomics 2002, 1, 845867. (41) Gygi, S. P.; Corthals, G. L.; Zhang, Y.; Rochon, Y.; Aebersold, R. Proceedings of the National Academy of Sciences 2000, 97, 9390-9395. (42) Panisko, E. A.; Conrads, T. P.; Goshe, M. B.; Veenstra, T. D. Experimental Hematology 2002, 30, 97-107. (43) Klose, J.; Kobalz, U. Electrophoresis 1995, 16, 1034. (44) Delahunty, C.; Yates, J. R. Methods 2005, 35, 248-255. (45) Blomberg, A.; Blomberg, L.; Norbeck, J. ; Fey, S. J.; Mose-Larsen, P.; Roepstorff, P.; Degand, H.; Boutry, M.; Posch, A.; Gorg, A. Electrophoresis 1995, 16, 19351945. (46) Fey, S. J.; Larsen, P. M. Current Opinion in Chemical Biology 2001, 5, 26-33.

PAGE 220

201 (47) Unlu, M.; Morgan, M. E.; Minde n, J. S. Electrophoresis 1997, 18, 2071-2077. (48) Marouga, R.; David, S.; Hawkins, E. Analytical and Bioanalytical Chemistry 2005, 382, 669-678. (49) Davis, M. T.; Stahl, D. C.; Hefta, S. A.; Lee, T. D. Analytical Chemistry 1995, 67, 4549-4556. (50) vanderHoeven, R. A.; Hofte, A. J.; Frenay, M.; Irth, H.; Tjaden, U. R.; vanderGreef, J.; Rudolphi, A.; Boos, K. S.; Marko-Varga, G.; Edholm, L. E. Journal of Chromatography A 1997, 762, 193-200. (51) Emmett, M. R.; Caprioli, R. M. Journal of the American Society for Mass Spectrometry 1994, 5, 605-613. (52) Shen, Y.; Zhao, R.; Belov, M. E.; Conr ads, T. P.; Anderson, G. A.; Tang, K.; PasaTolic, L.; Veenstra, T. D.; Lipton, M. S.; Udseth, H. R.; Smith, R. D. Analytical Chemistry 2001, 73, 1766-1775. (53) Stacey, C. C.; Kruppa, G. H.; Watson, C. H.; Wronka, J.; Laukien, F. H.; Banks, J. F.; Whitehouse, C. M. Rapid Communi cations in Mass Spectrometry 1994, 8, 513516. (54) Emmett, M. R.; White, F. M.; Hendrickson, C. L.; Shi, S. D. H.; Marshall, A. G. Journal of the American Societ y for Mass Spectrometry 1998, 9, 333-340. (55) Giddings, J. C. Journal of High Re solution Chromatography and Chromatography Communications 1987, 10, 319-323. (56) Yates, J. R. Annual Review of Bi ophysics and Biomolecular Structure 2004, 33, 297-316. (57) Link, A. J.; Eng, J.; Schieltz, D. M.; Carmack, E.; Mize, G. J.; Morris, D. R.; Garvik, B. M.; Yates, J. R. Nature Biotechnology 1999, 17, 676-682. (58) Wang, S.; Regnier, F. E. Jour nal of Chromatography A 2001, 924, 345-357. (59) Riggs, L.; Sioma, C.; Regnier, F. E. Journal of Chromatography A 2001, 924, 359368. (60) Wang, S.; Zhang, X.; Regnier, F. E. Journal of Chromatography A 2002, 949, 153162. (61) Macht, M.; Marquard t, A.; Deininger, S.-O.; Damoc, E.; Kohlmann, M.; Przybylski, M. Analytical and Bioanalytical Chemistry 2004, 378, 1102-1111. (62) Zhang, R.; Sioma, C. S.; Thompson, R. A.; Xiong, L.; Regnier, F. E. Analytical Chemistry 2002, 73, 5142-5149.

PAGE 221

202 (63) Hsu, J.-L.; Huang, S.-Y.; Chow, N.-H .; Chen, S.-H. Analytical Chemistry 2003, 75, 6843-6852. (64) Goodlett, D. R.; Keller, A.; Watts, J. D. ; Newitt, R.; Yi, E. C.; Purvine, S.; Eng, J. K.; vonHaller, P.; Aebersold, R.; Kolk er, E. Rapid Communications in Mass Spectrometry 2001, 15, 1214-1221. (65) Ong, S.-E.; Blagoev, B.; Kratchmarova , I.; Kristensen, D. B.; Steen, H.; Pandey, A.; Mann, M. Molecular and Cellular Proteomics 2002, 1, 376-386. (66) Smolka, M. B.; Zhou, H.; Purkayastha, S.; Aebersold, R. Analytical Biochemistry 2001, 297, 25-31. (67) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64-71. (68) Watson, J. T. In Introduction to Mass Spectrometry, 3rd ed.; Placito, M., Ed.; Lippincott-Raven: Philadelphia, 1997, pp 303-319. (69) Gomez, A.; Tang, K. Ph ysics of Fluids 1994, 6, 404-414. (70) Li, Y.; Cole, R. B. Analytical Chemistry 2003, 75, 5739-5746. (71) Wilm, M. S.; Mann, M. Internationa l Journal of Mass Spectrometry and Ion Processes 1994. (72) Smith, R. D.; Anderson, G. A.; Lipton, M. S.; Pasa-Tolic, L.; Shen, Y.; Conrads, T. P.; Veenstra, T. D.; Udseth, H. R. Proteomics 2002, 2, 513-523. (73) Tholey, A.; Toll, H.; Huber, C. G. Analytical Chemistry 2005, 77, 4618-4625. (74) Flora, J. W.; Null, A. P.; Muddiman, D. C. Analytical and Bioanalytical Chemistry 2002, 373, 538-546. (75) Shen, Y.; Kim, J.; Strittmatter, E. F.; Jacobs, J. M.; Camp, D. G.; Fang, R.; Tolie, N.; Moore, R. J.; Smith, R. D. Proteomics 2005, 5, 4034-4045. (76) Watson, J. T. In Introduction to Mass Spectrometry, 3rd ed.; Placito, M., Ed.; Lippincott-Raven: Philadelphia, 1997, pp 278-302. (77) Qin, J.; Ruud, J.; Chait, B. T. Analytical Chemistry 1996, 68, 1784-1791. (78) Laiko, V. V.; Moyer, S. C.; Cotter, R. J. Analytical Chemistry 2000, 72, 52395243. (79) Cameron, D.; Hemberger, P. H.; Alarid, J. E.; Leibman, C. P.; Williams, J. D. Journal of the American Societ y for Mass Spectrometry 1993, 4, 774-781.

PAGE 222

203 (80) Quarmby, S. T.; Yost, R. A. Intern ational Journal of Mass Spectrometry 1999, 190/191, 81-102. (81) McLuckey, S. A.; VanBerkel, G. J.; G lish, G. L.; Schwartz, J. C. Electrospray and the Quadrupole Ion Trap; CRC Press: New York, 1995. (82) Todd, J. F. J.; Penman, A. D. Interna tional Journal of Mass Spectrometry and Ion Processes 1991, 106, 1. (83) Schwartz, J. C.; Senko, M. W.; Syka, J. E. P. Journal of the American Society for Mass Spectrometry 2002, 13, 659-669. (84) Yost, R. A.; Enke, C. G. Analytical Chemistry 1979, 51, 1251A-1262A. (85) Watson, J. T. In Introduction to Mass Spectrometry, 3rd ed.; Lippincott-Raven Publishers: Philadelphia, 1997. (86) Hillenkamp, F.; Karas, M.; Beavis, R. C.; Chait, B. T. Analytical Chemistry 1991, 63, 1193A-1203A. (87) Bienvenut, W. V.; Deon, C.; Pasquarello, C.; Campbell, J. M.; Sanchez, J.-C.; Vestal, M. L.; Hochstrasser, D. F. Proteomics 2002, 2, 868-876. (88) Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Mass Spectrometry Reviews 1998, 17, 1-35. (89) Gorshkov, M. V.; Pasa-Tolic, L.; Udseth, H. R.; Anderson, G. A.; Huang, B. M.; Bruce, J. E.; Prior, D. C.; Hofstadler, S. A.; Tang, L.; Cheng, L. Z.; Willett, J. A.; Rockwood, A. L.; Sherman, M. S.; Smith, R. D. Journal of the American Society for Mass Spectrometry 1998, 9, 692-700. (90) Zubarev, R. A.; Horn, D. M.; Fridrikss on, E. K.; Kelleher, N. L.; Kruger, N. A.; Lewis, M. A.; Carpenter, B. A.; McLafferty, F. W. Analytical Chemistry 2000, 72, 563-573. (91) Ens, W.; Standing, K. G. In Methods in Enzymology; Burlingame, A. L., Ed.; Elsevier: Boston, 2005; Vol. 402, pp 49-77. (92) Baldwin, M. A. In Methods in Enzymo logy; Burlingame, A. L., Ed.; Elsevier: Boston, 2005; Vol. 402, pp 3-48. (93) Senko, M. W.; Hendrickson, C. L.; Emmett, M. R.; Shi, S. D.-H.; Marshall, A. G. Journal of the American Societ y for Mass Spectrometry 1997, 8, 970-976. (94) Zhang, Y. P.; Lambert, M. A.; Cairney, A. E.; Wills, D.; Ray, P. N.; Andrulis, I. L. Genomics 1989, 4, 259-265. (95) Ciustea, M. Doctoral dissertation, University of Florida, Gainesville, 2003.

PAGE 223

204 (96) Compton, S. J.; Jones, C. G. Analytical Biochemistry 1985, 151, 369-374. (97) Schevchenko, A.; Wilm, M.; Vorm, O.; Mann, M. Analytical Chemistry 1996, 68, 850-858. (98) Senko, M. W.; Canterbury, J. D.; Guan, S.; Marshall, A. G. Rapid Communications in Mass Spectrometry 1996, 10, 1839-1844. (99) Martin, B. M.; Tsuji, S.; LaMarca, M. E.; Maysak, K.; Eliason, W.; Ginns, E. I. DNA 1988, 7, 99-106. (100) Kroiher, M.; Raffioni, S.; Steele, R. E. Biochimi ca et Biophysica Acta 1995, 1250, 29-34. (101) Poulsen, S. A.; Gates, P. J.; Cousins, G. R.; Sanders, J. K. Rapid Communications in Mass Spectrometry 2000, 14, 44-48. (102) Herbert, B.; Galvani, M.; Hamdan, M. ; Olivieri, E.; MacCarthy, J.; Pedersen, S.; Righetti, P. G. Electrophoresis 2001, 22, 2046-2057. (103) Smith, R. D.; Loo, J. A.; Loo, R. R. O.; Busman, M.; Udseth, H. R. Mass Spectrometry Reviews 1991, 10, 359-453. (104) Wahl, J. H.; Goodlett, D. R.; Udset h, H. R.; Smith, R. D. Electrophoresis 1993, 14, 448-457. (105) Hunt, D. F.; Henderson, R. A.; Shabanow itz, J.; Sakaguchi, K.; Michel, H.; Sevilir, N.; Cox, A. L.; Appella, E.; Enge lhard, V. H. Science 1992, 255, 1261-1263. (106) Edwards, E.; Thomas-Oates, J. The Analyst 2005, 130, 13-17. (107) Wilcox, B. E.; Hendrickson, C. L.; Mars hall, A. G. Journal of the American Society for Mass Spectrometry 2002, 13, 1304-1312. (108) Emmett, M. R. Journal of Chromatography A 2003, 1013, 203-213. (109) Ronghe, M.; Burke, G. A. A.; Lowis, S. P.; Estlin, E. J. Cancer Treatment Reviews 2001, 27, 327-337. (110) Conrads, K. A.; Yu, L.; Lucas, D. A.; Zhou, M.; Chan, K. C.; Simpson, K. A.; Shaefer, C. F.; Issaq, H. J.; Veenstra, T. D.; Beck, G. R.; Conrads, T. P. Electrophoresis 2004, 25, 1342-1352. (111) Moore, L. J.; Machlan, L. A. Analytical Chemistry 1972, 44, 2291-2296. (112) Cali, J. P.; Bowers, G. N.; Young, D. S. Clinical Chemistry 1973, 19, 1208-1213. (113) Schulman, M. F.; Abramson, F. P. Biomedical Mass Spectrometry 1975, 2, 9-14.

PAGE 224

205 (114) Cohen, A.; Hertz, H. S.; Mandel, J.; Pa ule, R. C.; Schaffer, R.; Sniegoski, L. T.; Sun, T.; Welch, M. J.; White, E. Clinical Chemistry 1980, 26, 854-860. (115) Ballmer, P. E.; McNurlan, M. A.; Milne, E.; Heys, S. D.; Buchan, V.; Calder, A. G.; Garlick, P. J. The Americal Journal of Physiology 1990, 259, E797-E803. (116) Barr, J. R.; Maggio, V. L.; Patterson , D. G.; Cooper, G. R.; Henderson, L. O.; Turner, W. E.; Smith, S. J.; Hannon, W. H.; Needham, L. L.; Sampson, E. J. Clinical Chemistry 1996, 42, 1676-1682. (117) Barnidge, D. R.; Dratz, E. A.; Martin, T.; Bonilla, L. E.; Moran, L. B.; Lindall, A. Analytical Chemistry 2003, 75, 445-451. (118) Gerber, S. A.; Rush, J.; Stemman, O.; Kirs chner, M. W.; Gygi, S. P. Proceedings of the National Academy of Sciences 2003, 100, 6490-6495. (119) Kuhn, E.; Wu, J.; Karl, J.; Liao, H. ; Zolg, W.; Guild, B. Proteomics 2004, 4, 11751186. (120) Hawkridge, A. M.; Heublein, D. M.; Berge n, H. R.; Cataliotti, A.; Burnett, J. C.; Muddiman, D. C. Proceedings of the National Academy of Sciences 2005, 102, 17442-17447. (121) Yost, R. A.; Enke, C. G. Analytical Chemistry 1979, 51, 1251A-1262A. (122) Eng, J. K.; McCormack, A. L.; Yates, J. R. Journal of the American Society for Mass Spectrometry 1994, 5, 976-989. (123) Mathrubutham, M.; Vattem, K. In Pierce Application Notes; Pierce Biotechnologies, Inc: Rockford, 2005, pp 1-3. (124) Graham, D. R. M.; Elliot, S. T.; Va nEyk, J. E. Journal of Physiology 2005, 563, 19. (125) Leitner, A.; Lindner, W. Jour nal of Chromatography B 2004, 813, 1-26. (126) Sechi, S.; Oda, Y. Current Op inion in Chemical Biology 2003, 7, 70-77. (127) Rabilloud, T. Analytical Chemistry 2000, 72, 48A-55A. (128) Vietor, I.; Huber, L. A. Bi ochimica et Biophysica Acta 1997, 1359, 187-199. (129) Quadroni, M.; James, P. Electrophoresis 1999, 20, 664-677. (130) Mujumdar, R. B.; Ernst, L. A.; Mujumdar, S. R.; Lewis, C. J.; Waggoner, A. S. Bioconjugate Chemistry 1993, 4, 105-111.

PAGE 225

206 (131) Conrads, K. A.; Yu, L. R.; Lucas, D. A.; Zhou, M.; Chan, K. C.; Simpson, K. A.; Schaefer, C. F.; Issaq, H. J.; Veenstra, T. D.; Beck, G. R.; Conrads, T. P. Electrophoresis 2004, 25, 1342-1352. (132) Ducret, A.; VanOostveen, I.; Eng, J. K.; Yates, J. R.; Aebersold, R. Protein Science 1998, 7, 706-719. (133) Karty, J.; Ireland, M. M. E.; Brun, Y. V. ; Reilly, J. P. Journal of Chromatography, B 2002, 782, 363-383. (134) Zhang, Y.; Figeys, D.; Aebersold, R. Analytical Biochemistry 1998, 261, 124-127. (135) Parker, K. C.; Garrels, J. I.; Hines, W. ; Butler, E. M.; McKee, A. H. Z.; Patterson, D.; Martin, S. Electrophoresis 1998, 19, 1920-1932. (136) Vestling, M. M.; Murphy, C. M.; Fenselau, C. Analytical Chemistry 1990, 62, 2391-2394. (137) Herbert, B. Electrophoresis 1999, 20, 660-663. (138) Lognonne, J. L. Cellular a nd Molecular Biology 1994, 40, 41-55. (139) Switzer, R. C.; Merrill, C. R.; Shif rin, S. Analytical Biochemistry 1979, 98, 231237. (140) Lauber, W. M.; Carroll, J. A.; Dufiel d, D. R.; Kiesel, J. R.; Radabaugh, M. R.; Malone, J. P. Electrophoresis 2001, 22, 906-918. (141) AppliedBiosystems Cleavable ICAT Reagent Kit for Protein Labeling (monoplex version), 2003. (142) AppliedBiosystems In Inside iScience, 2004; Vol. 1, pp 1-2. (143) Vogt, W. Free Radical Biology & Medicine 1994, 18, 93-105. (144) Boehlein, S. K.; Richards, N. G. J.; Sc huster, S. M. Journal of Biological Chemistry 1994, 269, 7450-7457. (145) Zhou, M.; Abbatiello, S. E. Percentage of AS in MOLT-4 R samples, personal communication: Gainesville, FL, 2004. (146) Sun, W.; Wu, S.; Wang, X.; Zheng, D. ; Gao, Y. European Journal of Mass Spectrometry 2005, 11, 575-580. (147) Wienkoop, S.; Glinski, M.; Tanaka, N. ; Tolstikov, V.; Fiehn, O.; Weckwerth, W. Rapid Communications in Mass Spectrometry 2004, 18, 643-650. (148) Mei, H.; Hsieh, Y.; Nardo, C.; Xu, X. ; Wang, S.; Ng, K.; Korfmacher, W. A. Rapid Communications in Mass Spectrometry 2003, 17, 97-103.

PAGE 226

207 (149) Crimmins, D. L. Analytica Chimica Acta 1997, 352, 21-30. (150) Wolters, D. A.; Washburn, M. P.; Yates, J. R. Analytical Chemistry 2001, 73, 5683-5690. (151) Paoletti, A. C.; Zybailov, B.; Washbur n, M. P. Expert Review of Proteomics 2004, 1, 275-282. (152) Righetti, P. G.; Castagna, A.; Antoni oli, P.; Boschetti, E. Electrophoresis 2005, 26, 297-319. (153) Cech, N. B.; Enke, C. G. Analytical Chemistry 2000, 72, 2717-2723. (154) Yu, L.; Conrads, T. P.; Uo, T.; Issaq, H. J.; Morrison, R. S.; Veenstra, T. D. Journal of Proteome Research 2004, 3, 469-477. (155) Bondarenko, P. V.; Chelius, D.; Shaler, T. A. Analytical Chemistry 2002, 74, 47414749. (156) Chelius, D.; Bondarenko, P. V. J ournal of Proteome Research 2002, 1, 317-323. (157) Wang, W.; Zhou, H.; Lin, H.; Roy, S.; Shaler , T. A.; Hill, L. R.; Norton, S.; Kumar, P.; Anderle, M.; Becker, C. H. Analytical Chemistry 2003, 75, 4818-4826. (158) Old, W. M.; Meyer-Arendt, K.; Avelin e-Wolf, L.; Pierce, K. G.; Mendoza, A.; Sevinsky, J. R.; Resing, K. A.; Ahn, N. G. Molecular and Cellular Proteomics 2005, 4, 1487-1502. (159) Iiboshi, Y.; Papst, P. J.; Hunger, S. P.; Terada, N. Biochemical and Biophysical Research Communications 1999, 260, 534-539. (160) Chen, C.; Dudenhausen, E.; Chen, H.; Pan, Y.; Gjymishka, A.; Kilberg, M. S. Biochemical Journal 2005, 391, 649-658. (161) Kilberg, M. S.; Pan, Y.; Chen, H.; Le ung-Pineda, V. Annual Review of Nutrition 2005, 25, 59-85. (162) Szkanderova, S.; Vavrova, J.; Hernyc hova, L.; Neubauerova, V.; Lenco, J.; Stulik, J. Radiation Research 2005, 163, 307-315. (163) Sarraj, J. A.; Vinson, C.; Thie l, G. Biological Chemistry 2005, 386, 873-879.

PAGE 227

208 BIOGRAPHICAL SKETCH Susan Eugenia Lindyberg Abbatiello was born March 10, 1974, to Jan and Eugenia Lindyberg of Scotia, NY. At the age of 5, Susan was diagnosed with Type I diabetes, and took the first unknowing step into the world of science. While pursuing her childhood dream of becoming a pediatric endocri nologist, Susan developed a certain love for chemistry and the lab work associated w ith it, and found she was better suited to bench chemistry. She graduated from The College of the Holy Cross in Worcester, Massachusetts, in May of 1996 and began workin g as a Research Associate at Genetics Institute, a small biopharmaceutical company specializing in protein pharmaceuticals. On April 19, 1997, Susan married Russell A. Ab batiello, a dear friend from college who has always brought out the best in her. After five years at Genetics Institute, Susan began her graduate studies at the Univ ersity of Florida. Her goal wa s to become an expert in the field of mass spectrometry with emphasis on the analysis of pr oteins and peptides. While at UF, her work lead her to the area of cancer research. This res earch, supported by Dr. John R. Eyler and Dr. Nigel G. J. Richards, culminated in a new motivation to use her skills to better understand the ro le of proteins in cancer.