UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Design and applications of improved hyphenation and separation strategies for multidimensional biomolecule… Maxwell, Elizabeth Jane 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2011_spring_maxwell_elizabeth.pdf [ 3.78MB ]
Metadata
JSON: 24-1.0060171.json
JSON-LD: 24-1.0060171-ld.json
RDF/XML (Pretty): 24-1.0060171-rdf.xml
RDF/JSON: 24-1.0060171-rdf.json
Turtle: 24-1.0060171-turtle.txt
N-Triples: 24-1.0060171-rdf-ntriples.txt
Original Record: 24-1.0060171-source.json
Full Text
24-1.0060171-fulltext.txt
Citation
24-1.0060171.ris

Full Text

DESIGN AND APPLICATIONS OF IMPROVED HYPHENATION AND SEPARATION STRATEGIES FOR MULTIDIMENSIONAL BIOMOLECULE CHARACTERIZATION by  Elizabeth Jane Maxwell B.Sc., McGill University, 2004  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) February 2011 © Elizabeth Jane Maxwell, 2011  ABSTRACT A novel interface for capillary electrophoresis – electrospray ionization – mass spectrometry (CE-ESI-MS) has been developed with the goal of providing a robust and easy-to-use hyphenation strategy that also provides high sensitivity. The interface uses a tapered stainless steel needle that surrounds the capillary terminus, so that the needle interior acts as the CE outlet vial and the outside tip is the electrospray emitter. A chemical modifier solution, introduced via a tee junction, serves to improve the compatibility of the background electrolyte with ESI and allows the interface to operate under conditions where the bulk flow in the capillary is suppressed or reversed. A major novel feature of the interface is the use of a beveled conductive needle as the ESI emitter. Because electric field is highest at the sharpest edge of an electrode, liquid exiting the needle is drawn to the sharpest point of the bevel, whereas for a conventional symmetrically tapered emitter the field is equally distributed around the circumference of the tip. Electrospray performance as a function of flow rate was investigated for a variety of emitter geometries. The beveled emitters provided a more stable and efficient electrospray over a wider range of flow rates than traditional emitters. The second part of the thesis describes the application of the CE-ESI-MS interface with the beveled emitter to a variety of biomolecular species, including peptides, proteins and carbohydrates. These applications include the use of a variety of capillary coatings and separation conditions where the electroosmotic flow is suppressed or even reversed, so that the bulk flow in the CE capillary is away from the interface. The interface was also used to interface capillary isoelectric focusing (cIEF) with MS  ii  detection, enabling free solution and fully automated protein analysis that is analogous to 2D gel electrophoresis. The final part of the thesis examines the viability of a two-marker correction strategy for improving retention time reproducibility in reversed-phase liquid chromatography. Although the strategy provided up to a ten-fold improvement in reproducibility for repeated separations, the correction was not successful in relating retention times obtained under different gradient conditions.  iii  PREFACE The majority of the results included in this dissertation and the entirety of the writing were performed by the author, Elizabeth Jane Maxwell. The contributions of other researchers and collaborations are detailed below. Contributions from other researchers: Chapter 2:  Xuefei Zhong and Nikita van Zeijl performed the experiments for the counter-EOF separation of SDS and BCA, and the comparison of the interface with the commercial sheath-flow interface.  Chapter 3:  Xuefei Zhong performed the modeling and electric field calculations.  Chapter 4:  The separations of proteins and of angiotensin I and II were performed collaboratively with Xuefei Zhong in the labs of Beckman Coulter (Fullerton, CA). The samples and coated capillaries were provided by Beckman Coulter. Trypsin-digested BSA was kindly provided by Dr. Leonard Foster’s Lab in the Department of Biochemistry and Molecular Biology at UBC.  Chapter 5:  The carbohydrate analysis kit was provided by Beckman Coulter.  Chapter 6:  The cIEF separations shown in this chapter were performed collaboratively with Xuefei Zhong. The cIEF materials, including proteins and carrier ampholytes and coated capillaries were provided by Beckman Coulter.  Chapter 7:  The separations shown in Figure 7.1 were performed by Hong Zhang. Rat plasma samples were provided by the UBC Brain Research Centre.  Publications arising from work presented in the dissertation: 1.  E. Jane Maxwell and David D.Y. Chen. Twenty years of interface development for capillary electrophoresis- electrospray ionization- mass spectrometry. Analytica Chimica Acta (2008) 627, 25-33. Material from this article is included in Chapter 1.  2.  E. Jane Maxwell, Xuefei Zhong, Hong Zhang, Nikita van Zeijl, and David D.Y. Chen. Decoupling capillary electrophoresis and electrospray ionization for a more robust interface with mass spectrometry. Electrophoresis (2010) 31, 1130-1137. Material from this article is included in Chapters 2 and 4.  3.  E. Jane Maxwell, Xuefei Zhong and David D.Y. Chen. Asymmetrical emitter geometries for increased range of stable electrospray flow rates. Analytical Chemistry (2010) 82: 8377-8381. Material from this article is included in Chapter 3.  iv  TABLE OF CONTENTS Abstract.............................................................................................................................. ii Preface............................................................................................................................... iv Table of contents ................................................................................................................v List of tables...................................................................................................................... ix List of figures......................................................................................................................x List of symbols and abbreviations ................................................................................ xiii Acknowledgements ..........................................................................................................xv Dedication ....................................................................................................................... xvi 1. Chapter 1: Introduction to multidimensional biomolecule characterization strategies .............................................................................................................................1 1.1. TOOLS FOR MULTIDIMENSIONAL BIOMOLECULE CHARACTERIZATION ........................ 2 1.2. THEORY OF MULTI-DIMENSIONAL SEPARATIONS ........................................................ 3 1.3. TWO-DIMENSIONAL GEL ELECTROPHORESIS............................................................... 4 1.4. MASS SPECTROMETRY ............................................................................................... 7 1.4.1. Electrospray ionization ................................................................................... 8 1.5. LIQUID CHROMATOGRAPHY – MASS SPECTROMETRY ............................................... 10 1.5.1. Liquid chromatography................................................................................. 10 1.5.2. Liquid chromatography – electrospray ionization – mass spectrometry ...... 11 1.5.2.1. Established multidimensional proteomic strategies using LC-MS....... 12 1.6. CAPILLARY ELECTROPHORESIS – MASS SPECTROMETRY .......................................... 14 1.6.1. Capillary electrophoresis .............................................................................. 14 1.6.2. Technology for interfacing capillary electrophoresis with ESI-MS ............. 18 1.6.2.1. Interfaces employing a sheath-flow or make-up liquid ........................ 20 1.6.2.2. Sheathless interfaces ............................................................................. 24 1.6.2.2.1 Tip shaping...................................................................................... 24 1.6.2.2.2 Electrical contact............................................................................. 26 1.6.2.2.3 Ease of use and longevity ............................................................... 29 1.6.3. Fundamental concerns for CE-MS interface design ..................................... 31 1.6.3.1. Electrospray ionization ......................................................................... 31 1.6.3.2. Flow rate considerations ....................................................................... 31 1.6.3.3. Emitter tip geometry ............................................................................. 33 1.6.3.4. CE current and spray currents............................................................... 35 1.6.3.5. Electrochemistry reactions in CE-ESI-MS ........................................... 35  v  1.6.3.6. Location of the electrical contact.......................................................... 37 1.7. RESEARCH OBJECTIVES ............................................................................................ 38 1.7.1. Design of a novel CE-MS hyphenation strategy........................................... 38 1.7.2. Application of the interface to problems in biomolecular characterization.. 39 1.7.3. Correction strategies for liquid chromatography .......................................... 39  Section A: Design of a novel interfacing strategy for capillary electrophoresis – electrospray ionization – mass spectrometry ...............................40 2. Chapter 2: Decoupling capillary electrophoresis and electrospray ionization for a more robust interface with mass spectrometry.............................................................41 INTRODUCTION ........................................................................................................ 42 MATERIALS AND METHODS ...................................................................................... 45 2.2.1. Materials ....................................................................................................... 45 2.2.2. Interface configuration.................................................................................. 46 2.2.3. Instrumentation. ............................................................................................ 47 2.3. RESULTS AND DISCUSSION ....................................................................................... 48 2.3.1. Interface design ............................................................................................. 48 2.3.2. Conductive electrode and tip geometry ........................................................ 49 2.3.3. Decoupling capillary electrophoresis and electrospray ionization solution requirements.................................................................................................. 51 2.3.4. Comparison with commercial sheath flow interface performance. .............. 54 2.4. CONCLUDING REMARKS ........................................................................................... 58 2.1. 2.2.  3. Chapter 3: Investigation of asymmetrical emitter geometries for increased range of electrospray flow rates.....................................................................................................59 3.1. 3.2.  INTRODUCTION ........................................................................................................ 60 MATERIALS AND METHODS ...................................................................................... 63 3.2.1. Chemicals and materials ............................................................................... 63 3.2.2. Instrumentation ............................................................................................. 65 3.2.3. Electric field calculations.............................................................................. 65 3.3. RESULTS AND DISCUSSION ....................................................................................... 66 3.3.1. Continuous infusion ESI-MS evaluation conditions..................................... 66 3.3.2. Peak-based ESI-MS evaluation..................................................................... 68 3.3.3. Electric field simulation results..................................................................... 70 3.4. CONCLUDING REMARKS ........................................................................................... 71  Section B: Application of the CE-MS interface to separations of biomolecules .............................................................................................................72 4. Chapter 4: Application of the CE-ESI-MS interface to the separation and characterization of peptides and proteins......................................................................73 4.1.  INTRODUCTION ........................................................................................................ 74  vi  4.2.  MATERIALS AND METHODS ...................................................................................... 76 4.2.1. Materials ....................................................................................................... 76 4.2.2. Instrumentation ............................................................................................. 77 4.2.3. CE-MS of intact proteins .............................................................................. 78 4.2.4. CE-MS of angiotensin I and II...................................................................... 78 4.2.5. CE-MS of bovine serum albumin digest peptides ........................................ 79 4.3. RESULTS AND DISCUSSIONS ..................................................................................... 79 4.3.1. Analysis of proteins ...................................................................................... 79 4.3.2. Analysis of angiotensins I and II .................................................................. 81 4.3.3. Analysis of bovine serum albumin tryptic peptides...................................... 83 4.4. CONCLUDING REMARKS ........................................................................................... 86 5. Chapter 5: Capillary electrophoresis – mass spectrometry for carbohydrate analysis ............................................................................................................................................87 INTRODUCTION ........................................................................................................ 88 MATERIALS AND METHODS ...................................................................................... 90 5.2.1. Materials ....................................................................................................... 90 5.2.2. APTS labeling ............................................................................................... 91 5.2.3. Capillary electrophoresis with laser-induced fluorescence detection ........... 91 5.2.4. Capillary electrophoresis – electrospray ionization – mass spectrometry .... 92 5.3. RESULTS AND DISCUSSION ....................................................................................... 92 5.3.1. Interfacing with reverse-EOF separations..................................................... 92 5.3.2. Optimization of BGE conditions for resolution of large glycans ................. 94 5.3.3. CE-ESI-MS of APTS-labeled glucose ladder standard ................................ 96 5.4. CONCLUDING REMARKS ......................................................................................... 101 5.1. 5.2.  6. Chapter 6: Two dimensional protein mapping as a function of isoelectric point and molecular mass using capillary isoelectric focusing – electrospray ionization – mass spectrometry...................................................................................................................102 6.1. 6.2.  INTRODUCTION ...................................................................................................... 103 MATERIALS AND METHODS .................................................................................... 107 6.2.1. Materials ..................................................................................................... 107 6.2.2. Capillary isoelectric focusing ..................................................................... 108 6.2.3. Mass spectrometry ...................................................................................... 110 6.3. RESULTS AND DISCUSSION ..................................................................................... 110 6.3.1. cIEF-MS using the capillary as the catholyte vial ...................................... 110 6.3.2. Chemical mobilization ................................................................................ 113 6.3.3. Neutral coated capillaries............................................................................ 116 6.3.4. Correlation of migration times with isoelectric point ................................. 117 6.4. CONCLUDING REMARKS ......................................................................................... 119  Section C: Strategies for improving the reproducibility of reversed-phase liquid chromatography separations...............................................121  vii  7. Chapter 7: Development of a standard hydrophobicity index for improved protein characterization by reversed-phase liquid chromatography.....................................122 7.1. 7.2.  INTRODUCTION ...................................................................................................... 123 THEORY ................................................................................................................. 125 7.2.1. Hydrophobicity ........................................................................................... 125 7.2.2. Theory of partition chromatography ........................................................... 125 7.2.3. Retention time correction strategies for a constant gradient....................... 126 7.2.4. Theory of gradient elution .......................................................................... 128 7.3. MATERIALS AND METHODS .................................................................................... 129 7.3.1. Materials ..................................................................................................... 129 7.3.2. Pre-fractionation of protein samples ........................................................... 129 7.3.3. Reversed phase separation of intact proteins .............................................. 130 7.4. RESULTS AND DISCUSSION ..................................................................................... 131 7.4.1. Evaluation of two-marker correction strategies .......................................... 131 7.4.2. Standardization of hydrophobicity for any linear gradient ......................... 132 7.5. CONCLUDING REMARKS ......................................................................................... 135 8. Chapter 8: Discussion and conclusions ........................................................................136 REALIZATION OF RESEARCH OBJECTIVES ............................................................... 137 8.1.1. Design of a novel CE-MS hyphenation strategy......................................... 137 8.1.2. Application of the interface to problems in biomolecular characterization 139 8.1.3. Correction strategies for liquid chromatography ........................................ 140 8.2. FUTURE RESEARCH DIRECTIONS ............................................................................. 140 8.2.1. Incorporation of pre-concentration strategies for increased sensitivity ...... 140 8.2.2. CE-MS with gel buffers for improved resolution of complex glycan structures ..................................................................................................... 141 8.2.3. Development of cIEF-MS as a reproducible, high resolution alternative to 2D gel electrophoresis ...................................................................................... 143 8.3. CONCLUDING REMARKS ......................................................................................... 143 8.1.  9. Bibliography ...................................................................................................................145  viii  LIST OF TABLES Table 2-1: Limits of detection for amino acid analysis with different configurations and MS instruments.............................................................................................. 57 Table 3-1: Dimensions of the stainless steel emitters involved in the investigation ........ 64 Table 3-2: Optimal electrospray potentials and calculated maximum electric field for the six emitters studied ........................................................................................ 70 Table 4-1: Limits of detection for protein analysis.......................................................... 81 Table 4-2: Limits of detection for angiotensin I and II..................................................... 82 Table 4-3: Selected BSA peptides .................................................................................... 85 Table 4-4: Amino acid sequence of bovine serum albumin ............................................. 85 Table 6-1: Correlation of migration time and isoelectric point based on protein markers, pI 5.1 - 9.45.................................................................................................. 118 Table 7-1: Gradient elution parameters determined by non-linear regression ............... 134  ix  LIST OF FIGURES Figure 1.1: Analyte migration during isoelectric focusing in an immobilized pH gradient (pH 3 to 6)........................................................................................................ 5 Figure 1.2: Schematic of a capillary electrophoresis instrument with optical detection .. 15 Figure 1.3: (A) Structure of the electrical double layer and resulting electroosmotic mobility at a fused silica capillary wall surface due to deprotonation of silanol groups. (B) Decrease in electrical potential with increasing distance from the capillary wall................................................................................... 17 Figure 1.4: Common sheath-flow interface arrangements. A: Coaxial sheath flow interface with sheath gas, B: Liquid junction interface, C: Pressurized liquid junction interface. .......................................................................................... 21 Figure 1.5: Methods for creating electrical contact in sheathless interfaces. A – Conductive coating applied to the emitter tip, B – Wire inserted at tip, C – Wire inserted through hole, D – Split-flow interface with a metal sheath, E – Porous etched capillary, F – Two-piece interface with metal sleeve, G- Twopiece interface with microdialysis junction, G – Two-piece interface with conductive emitter. ........................................................................................ 28 Figure 2.1: Schematic illustration of the interface apparatus, including a dissected view of the needle tip with inserted capillary (inset).................................................. 46 Figure 2.2: Continuous separation and infusion of sodium dodecylsulfate and 1,2,4benzene tricarboxylic acid with negative ion electrospray. (A) EOF flows toward the CE outlet. CE inlet voltage +30 kV, Sample concentration 0.1 mM in BGE. MS scan range m/z 250~280. (B) EOF flows towards the CE inlet. CE inlet voltage -30 kV, Sample concentration 0.5 mM in BGE. MS scan range m/z 200 ~ 220. In both cases, 75 µm ID, 75 cm bare fused silica capillary, ESI -3.3 kV, BGE and Modifier: methanol/20mM pH 7.3 NH4AC/ H2O (v/v/v) 75:15:10. .................................................................................... 52 Figure 2.3: Effect of modifier flow rate on normalized amino acid signal intensity. Conditions: BGE and modifier 0.2% formic acid, 50% methanol; Sample: 20 µM amino acids in BGE; CE inlet 30 kV; ESI 3.4 kV; Micromass Q-TOF MS.................................................................................................................. 55 Figure 2.4: Extracted ion electropherograms for a mixture of 18 amino acids. ............... 56 Figure 3.1: Typical effect of flow rate on normalized signal for a stainless steel µESI emitter for a mixture of 15 amino acids under continuous infusion.............. 61 Figure 3.2: Different shapes of the stainless steel electrospray emitters investigated: symmetrically tapered electrospray needle (1), blunt tapered tip (2), sharp tapered needle (3), 30° bevel tip made from 2 (4), 45° bevel tip made from 2(5), 35° bevel tip with a smaller surface area (6)......................................... 64  x  Figure 3.3: Average signal (A) and signal to noise ratio (B) for a continuous infusion of arginine (20 µM in BGE) as a function of ESI potential and flow rate for various emitter geometries............................................................................. 67 Figure 3.4: Relative peak height as a function of flow rate for the six different tip geometries investigated. Intra-day normalization was performed based on the height of the 0.15 µL/min peak for emitter 6. ............................................... 69 Figure 4.1: Total ion electropherogram (A), extracted ion electropherograms (B) and mass spectra (C) of proteins separated under positive polarity with a neutral coated capillary. CE capillary, 50 μm ID, 67 cm; BGE, 100 mM NH4Ac, pH 3.1 and MeOH (9:1); Modifier, 0.2% formic acid, 50% methanol, 0.3 µL/min. .......................................................................................................... 80 Figure 4.2: Base peak ion electropherogram of Angiotensin I and II separated under reverse polarity with a polyethyleneimine coated capillary. BGE: 1% formic acid, 25% methanol; Modifier, 0.1% formic acid, 75% isopropanol, 0.5 psi; Sample: 1 μM Angiotensin I and Angiotensin II........................................... 82 Figure 4.3: Summed ion electropherograms for 17 peptides from a BSA digest separated in an uncoated capillary under normal polarity (TOP) and in a PEI coated capillary under reverse polarity (BOTTOM). Peptide sequences are listed in Table 4-3........................................................................................................ 84 Figure 5.1: Carbohydrate labeling with the fluorophore 8-aminopyrene 1,3,6-trisulfonate (APTS). Reprinted with permission from Analytical Biochemistry157. ........ 89 Figure 5.2: Capillary electrophoresis separations of APTS-labeled glucose ladder standard with on-capillary laser induced fluorescence detection using different capillary arrangements. ................................................................... 93 Figure 5.3: Base peak electropherograms for APTS-labeled glucose ladder standard using different background electrolyte compositions, with CE potentials indicated in brackets. MS range 600-2000 m/z............................................................ 95 Figure 5.4: Base peak and extracted ion electropherograms for APTS-labeled glucose ladder standard (G1 to G24) using optimized conditions (BGE 2.0% formic acid, 30% methanol; separation potential -16 kV). ....................................... 97 Figure 5.5: Contour plot of migration time and mass-to-charge ratio (m/z) for separated APTS-labeled glucose ladder standard showing different charge states. Conditions as in Figure 5.4.......................................................................... 100 Figure 6.1: Injection techniques employed in this chapter. (A) Flanking method for bare fused silica capillaries; (B) Modified method for covalently coated capillaries..................................................................................................... 109 Figure 6.2: Effect of sample injection parameters on peak shapes and analysis time. Injection using the flanking method with an anolyte/sample/catholyte length ratio of 1:2:1 (A), 3:4:1 (B) and 2:1:1 (C). Conditions: Sample, protein mixture with 0.3% (w/v) Fluka® ampholytes; bare fused silica capillary, 50 µm ID, 68 cm; focusing, +30 kV for 20 min; pressure mobilization. ......... 112  xi  Figure 6.3: Total ion electropherogram (A) and extracted mass spectra (B) for a protein mixture separated by cIEF with chemical mobilization. Conditions: Protein mixture with 0.3% (w/v) Fluka® ampholytes filling ½ capillary length; bare fused silica capillary, 86 cm/50µm ID Focusing +30 kV for 8 min; Chemical/EOF mobilization by modifier solution, 2% acetic acid, 50% methanol at 0.5 µL/min. .............................................................................. 115 Figure 6.4: Total ion electropherogram for a protein mixture separated by cIEF with chemical mobilization in a neutral coated capillary. Conditions: Protein mixture with 0.6% (w/v) Fluka® ampholytes filling ½ capillary length; neutral coated NCHO capillary, 76 cm/50µm ID; Inlet +30 kV; ESI +4kV after 30 min; Chemical mobilization, 2% formic acid, 50% methanol at 1 psi. ..................................................................................................................... 117 Figure 6.5: Migration time as a function of isoelectric point for the separations shown in Figure 2B (green triangles), Figure 3 (blue squares) and Figure 4 (red circles). Protein and peptide markers: ribonuclease A, pI 9.45; myoglobin (two isoforms) pI 7.35 and 6.85; carbonic anhydrase II, pI 5.9; ßlactoglobulin, pI 5.1; CCK flanking peptide, pI 3.6 (blue trace only)......... 118 Figure 7.1: Gradient revered phase separation of pooled proteins with no correction applied (A), correction according to equations 7-6 (B), 7-7 (C) and 7-8 (D). ..................................................................................................................... 131 Figure 7.2: Reversed phase separation of mobile phase impurities: Original chromatograms (top) and chromatograms with the equation 7-8 correction applied (bottom)........................................................................................... 132 Figure 7.3: Separations of intact rat plasma proteins by reversed-phase HPLC with a 0100% acetonitrile gradient over 45 minutes (red) and 30 minutes (blue) ... 133 Figure 7.4: Mobile phase at elution as a function of the gradient speed for 13 peaks.... 134  xii  LIST OF TERMS AND ABBREVIATIONS 2DE 2D-PAGE APCI APPI APTS BCA BGE BSA ß-lac CA-II CCK CE CGE cIEF CZE E EOF ESI  ε η f FD FAB G20  γ hp HPLC ID IEF k1 K L LC LIF LOD MALDI  two-dimensional gel electrophoresis two-dimensional polyacrylamide gel electrophoresis atmospheric pressure chemical ionization atmospheric pressure photoionization 8-aminopyrene 1,2,6-trisulfonate 1,2,4-benzenetricarboxylic acid background electrolyte bovine serum albumin ß-lactoglobulin carbonic anhydrase II CCK flanking peptide capillary electrophoresis capillary gel electrophoresis capillary isoelectric focusing capillary zone electrophoresis electric field electroosmotic flow electrospray ionization electric permittivity viscosity frictional coefficient drag force fast atom bombardment ionization glucose ladder standard surface tension hydrophobicity index high performance liquid chromatography inner diameter isoelectric focusing capacity factor equilibrium constant length of separation capillary or column liquid chromatography laser induced fluorescence limit of detection matrix-assisted laser desorption ionization xiii  MEKC MS MS/MS MudPIT Myo nc nanoESI OD PX PEEK PEI pI PNGase F π Q-TOF RS RNase A RP RSD  ρ SCX SDS SNR SS tM tm tR TFA TOF U UV μeo μ ep μESI  υ vep V ζ  micellar electrokinetic chromatography mass spectrometry tandem mass spectrometry multidimensional protein identification technology myoglobin peak capacity nanoelectrospray ionization outer diameter partition coefficient polyether ether ketone polyethyleneimine isoelectric point peptide N-glycosidase F hydrophobicity constant quadrupole time of flight mass spectrometer resolution ribonuclease A reversed-phase relative standard deviation surface tension strong cation exchange sodium dodecylsulfate signal-to-noise ratio stainless steel migration time chromatography void time retention time trifluoroacetic acid time of flight electrical potential ultraviolet electroosmotic mobility electrophoretic mobility micro-electrospray ionization Taylor cone angle electrophoretic velocity volume zeta potential xiv  ACKNOWLEDGEMENTS I would first like to thank my supervisor, Dr David Chen, for being a wonderful mentor for the past five years. I have learned so much from your thoughtful approach to subjects both research-related and beyond. Thank you for giving me every opportunity to develop and succeed and for having enough faith in me and my research to push me through the periods where I didn’t have it myself. This thesis certainly would not have been possible without your guidance and encouragement. Thank you to everyone who has provided mentorship through this process, particularly Dr. Eric Salin at McGill University, who first inspired me towards research and who has been a great supporter over the years, and Dr. Michael Blades, who has been a source of help and encouragement throughout my time at UBC. Thanks also to all of the members of my guidance committee. I have been fortunate to work alongside many exceptional people in the course of my research. Thank you to all of my group mates, past and present, who have shared their ideas and provided a friendly and supportive community in the lab: Koen, Sharon, Xuefei, Roxana, Joe, Chang, Sherry, Nikita, Hong, Ning and Alison. I would particularly like to acknowledge Xuefei Zhong and Hong Zhang, whose research contributions are integral to the work presented in this thesis. This research would not have been possible without the financial and logistical support of Beckman Coulter, or the expertise and dedication our collaborators there, particularly Chitra Ratnayake and Scott Mack. I am also grateful for financial support from the Natural Sciences and Engineering Research Council of Canada. My friends are the balancing force in my life, keeping a smile on my face regardless of how things are going in the lab. Thanks to all of you, both here in Vancouver and farther away, for your great sense of humour, for amazing adventures and for making any place where we’re together feel like home. B, thank you for making the last two years of my PhD the happiest. Finally, I am grateful to my wonderful family for their endless support and encouragement. Dave, Meg, Nika and Violet, thanks for your love and inspiration. Mom and dad, thank you for always believing in me and giving me the confidence to follow my own path. Your unconditional love and support have been the strongest things helping me through this process. The pages that follow are filled with my gratitude for having such wonderful parents. Thank you.  xv  to my parents…  xvi  1. Chapter 1: Introduction to multidimensional biomolecule characterization strategies  1  1.1. TOOLS FOR MULTIDIMENSIONAL BIOMOLECULE CHARACTERIZATION Due to the complexity of most biological samples, a multidimensional strategy is nearly always required to properly characterize them. The classic example of this is twodimensional gel electrophoresis (2DE), which can resolve thousands of components because of its two orthogonal dimensions of separation. More recently, mass spectrometry (MS) has become an essential tool for biomolecule characterization because it provides high resolution separation based on mass-to-charge ratio and the ability to positively identify analytes based on the analysis of fragment ions. However, most complex samples must be separated prior to MS analysis in order to prevent space-charge or suppression effects and to simplify the interpretation of MS spectra1. Free solution separation techniques, such as liquid chromatography (LC) and capillary electrophoresis (CE), have the advantage that their effluent can be directly coupled to MS, which saves time and maximizes the combined resolution of the hyphenated techniques. Capillary electrophoresis and liquid chromatography are complementary separation strategies, as their separation mechanisms are based on different analyte characteristics, and both techniques have found widespread application in the fields of proteomics, metabolomics and pharmaceutical research. While LC-MS is very well established as a hyphenated technique, CE-MS has lagged behind, due to the additional challenges involved in interfacing CE with electrospray ionization (ESI). Researchers wishing to use CE-MS analysis are currently faced with a choice between commercial instruments, which offer limited sensitivity, or a variety of “lab-made” hyphenation  2  strategies that require a high level of operator skill and care. As a result, the full potential of CE-MS as a tool for biomolecule characterization and analysis is not currently available to most of the research community. Despite the popularity of LC-MS in a variety of application areas, the full potential of LC-MS is also underexploited, as the LC portion of the analysis is often used only as a means of reducing the number of species simultaneously entering the MS. The retention time information, which is related to intrinsic analyte characteristics, is disregarded because poor reproducibility makes its incorporation and interpretation too complex. This thesis presents work on addressing these two deficiencies in the available tools for biomolecule characterization: the first part discusses the design of a novel interface for capillary electrophoresis – electrospray ionization – mass spectrometry. The second part demonstrates the utility of the interface for a variety of applications in biomolecular analysis. Finally, a two-marker strategy for improving the reproducibility of gradient reversed-phase LC separations, which relates retention times to the intrinsic hydrophobic character of the analytes, is proposed and evaluated. 1.2. THEORY OF MULTI-DIMENSIONAL SEPARATIONS Multidimensional analysis is required in order to characterize samples of high complexity. This is because the number of components that can be resolved based on a single separation is limited to the number of peaks that will fit into the chromatogram, electropherogram or physical space of the gel while maintaining a specified resolution2. This limiting number, known as the peak capacity (nc), can be calculated as  3  nc =  L 4σRS  (1-1)  where L is the length of the separation column, capillary or gel, σ is the standard deviation of the peak and RS is the required level of resolution. For baseline resolution of adjacent peaks, RS = 1.5. This limitation can be overcome by using multiple dimensions of separation, so that the separation takes place within a two-dimensional space, which has an area equal to the product of the individual dimensions’ lengths. Assuming that the dimensions are fully orthogonal, the peak capacity of the 2D separation is therefore the product of the peak capacities of the individual dimension:  nc = nx ny  (1-2)  Equation 1-2 also assumes that the method of coupling the two separation dimensions does not increase the widths of the peaks. While this is true for online coupling (LC-MS or CE-MS) and for separations performed on a 2D surface (2DE), it does not apply to methods that involve off-line coupling (fraction collection) or step-wise elution to transfer components from one dimension to the next. 1.3. TWO-DIMENSIONAL GEL ELECTROPHORESIS Historically, 2-dimensional polyacrylamide gel electrophoresis (2D-PAGE) has been the primary method of separation and comparison for complex protein mixtures. The advantage of 2D-PAGE is that it separates analytes based on two intrinsic properties: isoelectric point (pI) and molecular size. Separation based on the isoelectric point (pI), is achieved by gel isoelectric focusing (IEF). A linear pH gradient is formed by immobilizing buffering ampholytes in  4  a polyacrylamide gel matrix. Analytes loaded onto the gel will experience a localized pH that depends on their location along the gel, so that zwitterionic species, such as proteins, may simultaneously exist in both positively and negatively charged states in different parts of the gel, depending on whether the local pH is greater or less than the species pI. When an external electric field is applied, positively charged analytes will migrate towards the cathode, while negatively charged analytes migrate to the anode, as shown in the upper portion of Figure 1.1. However, as analytes migrate they experience a changing local pH, leading to corresponding changes in the analytes’ charge state. Eventually, each component will migrate into the region where the local pH is equal to its pI, resulting in a zero net charge state and ceasing the migration process. If a zwitterionic analyte diffuses away from this equilibrium position to a region with higher or lower pH, it will acquire a negative or positive charge, respectively, until it migrates back to the location where pH = pI, as depicted in the lower portion of Figure 1.1.  Figure 1.1: Analyte migration during isoelectric focusing in an immobilized pH gradient (pH 3 to 6) The result of this focusing process is that analytes are linearly arranged in the gel according to their pI. This is an extremely useful tool for the characterization of proteins  5  and peptides, since both the amino acid sequence and the degree and type of posttranslational modifications will have an affect on the analyte pI. For the second dimension separation, bands separated by IEF are subjected to sodium dodecylsulfate (SDS) polyacrylamide gel electrophoresis (SDS-PAGE), which provides separation based on molecular size. When SDS is added to denatured proteins or peptides, it binds to the polypeptide backbone with a constant weight ratio of 1.4 g per gram of polypeptide3. In this manner, the intrinsic charges of the polypeptide become negligible compared to the charges from the SDS, such that the protein-SDS complexes all have the same charge density. The repulsion of adjacent negative charges also causes the protein to stretch out into a long chain. Under the influence of an external electric field, the protein-SDS complex migrates towards the cathode. The presence of the branched gel structure retards the migration of the proteins through a sieving mechanism, so that larger proteins experience a greater resistance and reduced mobility . log( mobility ) ∝  1 molecular weight  (1-3)  The technique therefore allows the estimation of protein molecular weight, based on migration relative to markers of known molecular weight. 2D-PAGE has been critical in developing our understanding of the complexity and variety of proteins contained in cells and bodily fluids, including serum and plasma4. Although impressive technological improvements have been made, there remain some fundamental limitations. 2D-PAGE is time consuming, labour-intensive, lacks reproducibility and presents difficulties in detecting proteins with extremes in molecular mass and isoelectric point 5-7. As such, the focus of the proteomics community has  6  shifted to multidimensional platforms based on free-solution separations, which are more amenable to automation. 1.4. MASS SPECTROMETRY Mass spectrometry (MS) is an extremely powerful and versatile tool for qualitative and quantitative analysis of many types of biomolecules1. In addition to providing excellent resolution based on analyte mass-to-charge ratio, the information provided on the intact mass of the analyte, or by the fragmentation patterns achieved by tandem MS allows for positive identification of analytes. However, separation prior to MS analysis is still a necessary step in all but the simplest applications of this technology, since too many components entering the mass spectrometer at one time can lead to signal suppression, space-charge effects and overly complex mass spectra that make interpretation impossible8. From the perspective of those working in the field of separation science, mass spectrometry, as a detector coupled to chromatographic or electrophoretic separation techniques, offers an additional dimension of resolution and the potential for unambiguous identification of analytes. Since the instrumentation for mass spectrometry is now extremely sophisticated, the most challenging step is the hyphenation of the separation and detection techniques. For online coupling with liquid separation techniques, it is preferable that ionization occurs at atmospheric pressure in order to eliminate flow artifacts related to the pressure differential. Depending on the types of analytes involved and the properties of the separation effluent, there are a number of ionization techniques available, including electrospray ionization (ESI)9, 10, atmospheric pressure chemical ionization (APCI)11 and  7  atmospheric pressure photoionization (APPI)12. Matrix-assisted laser desorption ionization (MALDI)13, 14 is also commonly used, but this is an off-line technique, as eluent must be spotted onto a MALDI plate and analyzed at a later time. 1.4.1. Electrospray ionization Electrospray ionization (ESI) is a popular low-energy ionization technique for polar, non-volatile species. It has been the key enabling technology for the application of mass spectrometry to large biomolecules, such as peptides and proteins, because of its ability to ionize without causing fragmentation, and its tendency to form multiply charged ions. This allows macromolecules with masses of >100,000 Da to be observed as multiply charged ions, enabling the use of instruments with limited mass-to-charge ranges, such as quadrupole and ion trap mass analyzers. It also is capable of ionizing analytes directly from solution, making it ideal for online coupling with liquid chromatography or capillary electrophoresis. ESI was first proposed as an ion source for mass analysis by Dole et al.9 and Yamashita and Fenn helped to demonstrate the potential of ESI for mass spectrometry15, 16  . A few years later, the demonstration of its utility for ionizing large biomolecules  greatly increased its adoption as it allowed ESI-MS to be applied to a group of analytes previously considered unsuitable for MS17. Since then, ESI has become the most commonly used ionization technique for biomolecule characterization. ESI involves applying a high electrical potential to a liquid sample flowing through a capillary. The electric field at the capillary tip acts on the ions in solution, creating a charge separation between anions and cations. In positive ion mode ESI, positive ions migrate downstream towards the meniscus of the liquid at the tip of the  8  capillary. Electrostatic forces acting on the positive ions at the meniscus leads to the formation of a Taylor cone, balanced by the surface tension of the solution. When the electrostatic force exceeds the surface tension, a jet of charged droplets is ejected from the tip of the cone towards the counter electrode. The radius (re) of the droplets ejected from the Taylor cone depends on the solution density (ρ), surface tension (γ), the angle of the cone (υ), the ratio of the applied potential to the ESI onset potential (Ua/Ut) and the solution flow rate (dV/dt)18. ⎛ ρ re = ⎜ ⎜ 4 π 2γ tan(π 2 − υ ) (U U )2 −1 a T ⎝  [  1/ 3  ⎞ ⎟ ⎟ ⎠  ]  × (dV d t )  2/3  (1-4)  Following ejection from the Taylor cone, subsequent fissions19 and evaporation20 of the charged droplets results in the formation of single solvated gas phase ions21. These ions are then transmitted to the inlet aperture of the mass spectrometer for separation based on their mass to charge ratio, followed by detection. Although electrospray ionization can function over a wide range of flow rates, equation 1-4 indicates that the size of the ejected droplets will be larger for high flow rates, which will require a greater number of fissions and evaporations in order to fully desolvate ions before they enter the MS. Even at moderate flow rates, this is often not possible. It has been shown that more hydrophobic analytes, which have a higher surface activity (are more likely to reside at the surface of the droplet), are ionized with much greater efficiency than those residing in the bulk of the droplet, because ions closer to the droplet surface have a greater probability of ejection during fission events. As a result, when more than one analyte is present in ESI, the species with greater surface activity may significantly suppress the ionization of the less surface-active analyte.  9  Nanoelectrospray ionization (nanoESI), which refers to ESI performed at very low flow rates, has been shown to effectively combat this type of suppression. As demonstrated by Wilm and Mann in 1994, at very low flow rates and with a very small ESI emitter (theirs had an orifice of 1-3 μm and a flow rate of 25 nL/min), the droplets ejected from the Taylor cone are so small that each droplet will statistically contain less than one analyte molecule18. As a result, surface activity becomes irrelevant in determining ionization efficiency and ion transmission increases, due to the ease of desolvating droplets with such a large surface-to-volume ratio22-24. 1.5. LIQUID CHROMATOGRAPHY – MASS SPECTROMETRY 1.5.1. Liquid chromatography High performance liquid chromatography (HPLC, or LC) is the most widely used analytical separation technique for non-volatile biomolecules. A narrow sample plug is carried by a liquid mobile phase through a column containing a bonded liquid stationary phase. Separation is based on the varying strengths of analytes’ interactions with the stationary phase, as this will determine the proportion of time spent in the mobile phase and their average velocity in the column. Detection at the end of the column may be carried out by a variety of methods, including UV/visible absorbance, fluorescence, conductivity and mass spectrometry10, 25. Depending on the nature of the analytes, the interaction with the stationary phase may be based on solubility (normal and reversed phase partition chromatography), electrostatic interactions (ion exchange chromatography), size-based permeability (size exclusion chromatography) or physical adsorption of solutes onto a solid stationary phase  10  (adsorption chromatography)26. Because of this wide variety of available retention modes, LC separations have been successfully applied to all types of biomolecules. 1.5.2. Liquid chromatography – electrospray ionization – mass spectrometry The first electrospray ionization interface for liquid chromatography –mass spectrometry (LC -MS) was described in 198510. Early LC-ESI-MS interfaces were designed for flow rates in the range of 1 – 10 μL/min, which is significantly lower than the flow rates of most LC separations at the time. The ionspray interface, which combined ESI with pneumatic nebulization, increased the flow rates that could be processed up to 200 μL/min27. However, following the demonstration of nanoESI and its associated benefits, many LC-MS interfaces now use packed capillary columns, which can operate at flow rates below 1 μL/min and are easily coupled to micro- or nanoelectrospray ionization28. Reversed phase HPLC, where analytes are partitioned between a non-polar stationary phase and a more polar mobile phase, is the most popular option for coupling with ESI-MS, because the mobile phase composition - normally water mixed with a miscible organic solvent such as methanol or acetonitrile - is highly compatible with ESIMS7. Due to the complexity of biological samples, multiple dimensions of LC, employing different separation modes, are often combined. The most popular combination is ion exchange as the first dimension, followed by reversed phase as the second dimension. Although less popular, there are also examples of size-exclusion or affinity LC as the first dimension separation7. Compared to traditional 2D-PAGE methods, 2D-LC offers the advantages of higher loading capacity without band distortion and improved compatibility with  11  membrane proteins, highly hydrophobic proteins and proteins with low molecular weights5-7. It also offers enhanced reproducibility thanks to automation and the fact that the liquid format eliminates the need for additional procedures prior to protein identification by MS. 1.5.2.1. Established multidimensional proteomic strategies using LC-MS One of the most popular proteomic techniques currently in use for global proteomic analysis is multidimensional protein identification technology (MudPIT)29-31. Briefly, a capillary column which has been pulled to form a nano-electrospray tip at one end is packed with two different types of beads, such that the front end separates peptides in strong cation-exchange (SCX) mode and the back end separates in reversed phase (RP) mode. Trypsin is added to the sample in order to digest all proteins to peptides prior to any liquid separation. The digested sample is then injected onto the packed column in a low-ionic strength mobile phase, such that nearly all peptides are fully retained on the SCX stationary phase. The peptides are then eluted in step-wise manner using a series of increasing ionic strength elutions, each followed by reversed phase gradient elution on the RP section. As peptides elute from the RP section they are ionized by electrospray ionization into a tandem mass spectrometer for peptide identification. The power of MudPIT was demonstrated in its application to the yeast proteome, where it was used to detect and identify nearly 1500 proteins30, whereas 2D gel electrophoresis methods had previously only managed to identify a few hundred. One limitation of MudPIT and similar “top-down” proteomic approaches, in which mixtures of proteins are subjected to enzymatic digestion prior to separation, is that they rely on the assumption that the observation of a small number of peptides, representing a small  12  portion of the protein sequence, is sufficient for protein identification. Also, because significant portions of the protein sequence are lost, these techniques can not be used to monitor the dynamic properties of the proteome, such as the combinations of posttranslational modifications occurring simultaneously on a protein. In order to obtain a protein profile that is sensitive to post-translational modifications, proteins must be separated and measured intact. Recognizing this, a number of multi-dimensional liquid separation methods for intact proteins have been proposed32, 33. Among these, the most successful is a 2D-LC method developed in the group of David Lubman in 2000, which separates proteins based on their pI in the first dimension using chromatofocusing and based on hydrophobicity in the second dimension using reversed-phase HPLC. The separation is followed by mass analysis to give the intact protein molecular weight34, 35. The most novel feature of this 2D system is the use of chromatofocusing in the first dimension to separate proteins according to their pI. In order to accomplish this, proteins are equilibrated in a strong anion exchange column under basic conditions, so that nearly all proteins are negatively charged and retained by the stationary phase. The mobile phase is then switched to an acidic solution, which slowly titrates the pH inside the column to create a transient pH gradient. When the local pH matches the protein pI, the charge is neutralized and the protein loses its interaction with the stationary phase. Therefore, as the pH gradient sweeps through the column, it selectively “peels off” proteins with the appropriate pI, creating a focusing effect as proteins of similar pI over the entire column length are collected together into a band at the exit of the column. Fractions from the first dimension column are collected at regular pH intervals for  13  injection into the second dimension separation. There are also versions of this system where the chromatofocusing step has been replaced by a preparative free-solution isoelectric focusing device, such as the Rotofor®, which separates proteins into up to twenty fractions, according to their isoelectric point34. The second dimension fractionation is done by gradient reversed-phase chromatography, which uses a non-polar stationary phase and a water/acetonitrile mobile phase gradient to elute proteins in the order of increasing hydrophobicity. Trifluoroacetic acid is used as an ion suppressor agent to minimize non-specific interactions that may alter protein retention. Following the 2D separation, proteins are either eluted directly into an online ESI-MS or collected for off-line MS analysis to determine the intact protein molecular weight. Further details on this system, known as the PF2D, and the investigation of numerical methods for increasing the reproducibility and quality of information of the second-dimension separation, are discussed in chapter 7. 1.6. CAPILLARY ELECTROPHORESIS – MASS SPECTROMETRY 1.6.1. Capillary electrophoresis Capillary electrophoresis (CE), which was first demonstrated by Jorgenson and Lukacs in 1981, involves the application of a large electric potential across a narrow-bore fused silica capillary, such that a uniform electric field is generated along the length of the capillary36. Whereas separation in liquid chromatography is based on the partition equilibrium between a mobile and stationary phase, separation in capillary zone electrophoresis (CZE) is driven by the differential migration of analytes within a single phase, under the influence of an electric field. A typical capillary electrophoresis apparatus is shown in Figure 1.2. 14  Figure 1.2: Schematic of a capillary electrophoresis instrument with optical detection Ions in free solution in the presence of an electric field experience an electrostatic force, Fe, that is proportional to the charge on the ion, q, and the magnitude of the electric field, E. v v Fe = qE  (1-5)  As the ion accelerates in response to this force, it also experiences a drag force, FD, which ! is proportional to the radius of the solvated ion, R, the frictional coefficient of the  solution, f, and the velocity of the ion, v. v FD = f " v = 6#$R " v  (1-6)  where ! is the viscosity of the solution. The drag force quickly increases until its !  magnitude is equal to that of the electrophoretic force, so that the ion migrates with a steady-state electrophoretic velocity, vep. v qE v v ep = 6"#R  (1-7)  !  15  Therefore, in free solution capillary zone electrophoresis ions migrate with a velocity that is proportional to their charge-to-size ratio. In order to describe migration behaviour independent of the electric field, we can define the electrophoretic mobility (µep). μep ≡  v ep E  =  q 6πηR  (1-8)  In addition to the analyte’s electrophoretic mobility, there is also a bulk movement of solution in the fused silica capillary, called the electroosmotic flow (EOF), which is due to the acidic character of the silanol groups on the inner surface of the capillary wall. When the pH of the background electrolyte (BGE) is greater than 3, deprotonation of the silanol groups results in a negatively charged wall surface. This creates an electrical potential that attracts positive ions from the solution to the wall and generates an electrical double layer. The layer of ions closest to the wall is called the Stern, or fixed layer, because the ions in this region experience the maximum attractive force and are therefore held in place. The ions in the fixed layer counter-balance a portion of the wall potential, such that the layer beyond the fixed layer experiences a decreased potential, defined as the zeta potential (ζ). The second layer, known as the diffuse layer, contains both positive and negative ions, but has an excess of positive charge in order to counter-balance the residual zeta potential. The structure of this electrical double layer is shown in Figure 1.3. Because the magnitude of the electrostatic force in the diffuse layer is much less than in the fixed layer, the diffuse layer is free to move in response to external forces. Electroosmotic flow arises when an external field is applied across the capillary, because the excess of positive charge in the diffuse layer experiences an electrostatic force which moves the diffuse layer towards the cathode with an electroosmotic mobility of  16  μeo =  v eo εζ = E 4 πη  (1-9)  Beyond the diffuse layer is the bulk solution, which is completely shielded from the wall potential by the electrical double layer. The bulk solution is dragged along with the diffuse layer, so that the electroosmotic flow has a uniform velocity profile across the capillary.  Figure 1.3: (A) Structure of the electrical double layer and resulting electroosmotic mobility at a fused silica capillary wall surface due to deprotonation of silanol groups. (B) Decrease in electrical potential with increasing distance from the capillary wall The most obvious issue with capillary zone electrophoresis is that it is not capable of separating neutral compounds, because the separation is based purely on charge-to-size ratio. However, this can be circumvented through the addition of surfactant micelles to  17  the background electrolyte to act as a pseudo-stationary phase, a technique known as micellar electrokinetic chromatography (MEKC)37, 38. In this case, analytes are in equilibrium between the aqueous BGE and the hydrophobic interior of the micelles, which migrate at a different rate from the BGE. As a result, neutral compounds can be separated based on their different affinities for the micelle vs. BGE environments, while charged analytes’ migration is affected by both their electrophoretic mobility and equilibrium behaviour. It is also possible to use other types of BGE additives, such as cyclodextrans, to carry out chiral separations39. 1.6.2. Technology for interfacing capillary electrophoresis with ESI-MS When first introduced in 1987, the combination of capillary electrophoresis with mass spectrometry through an electrospray ionization interface provided an intriguing solution for combining two highly powerful analytical techniques. CE had already been shown to give excellent separation efficiencies, however, short optical path lengths provided by the small capillary inner diameters resulted in a concentration sensitivity that fell short of those achieved by other separation techniques. On the other hand, electrospray ionization mass spectrometry had previously been shown to offer low detection limits and fast response times as a detector for liquid chromatographic separations27, as well as providing valuable structural information on the analytes. Smith and coworkers performed the first demonstration of online electrospray ionization of CE effluent in 1987, using a metal sheath around the capillary terminus to replace the terminal electrode of a traditional CE setup, providing gaseous analyte ions for mass spectrometric analysis40. Unfortunately, the publications that followed on the subject seemed to highlight the limitations of the technique. The use of electrospray ionization  18  placed limitations on the composition of the background electrolyte that could be used in the CE separation and the flow rates required to maintain a stable electrospray could only be supplied under conditions giving a maximized electroosmotic flow41, 42. The use of a flowing sheath liquid was offered as a solution to these problems, however it came at the price of reduced sensitivity, as the sheath liquid diluted the CE effluent. The nature of CE lends some particular challenges when it comes to online MS detection. Both the CE and ESI processes require stable electrical contact of the solution with an electrode at the capillary outlet without interruption of the electroosmotic flow from the CE separation. In addition, the low volumetric flow rates used in CE place particular restrictions on the geometry of the tip if a stable electrospray is to be maintained43. Finally, compatibility of the background electrolyte with the electrospray process and on the resulting mass spectra must be considered. Over the past twenty years, a wide variety of interface designs has been proposed in order to maximize the potential of this hyphenated technique. These have been summarized in a number of excellent technology-based reviews on the subject44-52. However, commercially available CE-MS platforms continue to use interfaces that are far from optimal insofar as sensitivity is concerned, limiting them to applications where concentration is not a limiting factor53. This is in large part due to the fact that those interfaces offering the best sensitivity and lowest detection limits have also been among the most fragile, making them unattractive for commercialization. The compromise of robustness over sensitivity among commercial CE-MS platforms is one of the major reasons that the technique has still not been widely adopted as a routine method of analysis.  19  1.6.2.1. Interfaces employing a sheath-flow or make-up liquid As previously mentioned, sheath-flow interfaces are the status quo for commercially available CE-MS interfaces and have been popular since the early years of CE-MS applications42. These interfaces use an additional flow of liquid, known as a sheath or make-up liquid, that mixes with the CE effluent as it exits the separation capillary. The added flow serves a number of purposes. The first is to establish electrical contact between an electrode and the background electrolyte (BGE) inside the capillary in order to drive the CE separation and the ESI process. The second purpose is to modify the composition of the CE electrolyte to make it more compatible with ESI and MS detection. In addition, in the early stages of CE-MS development most interfaces were adapted to fit into existing LC-MS setups, which required much higher flow rates than those delivered by CE. Therefore, the sheath liquid also served to increase the liquid flow to levels comparable to those found in liquid chromatography. The most common interface of this type is the coaxial sheath-flow arrangement, where a continuous flow of sheath liquid is delivered through a tube that surrounds the separation capillary terminus, as illustrated in Figure 1.4A. The advantage in this case is that there is no dead volume, so the capillary separation is undisturbed by additional flows until the analytes have exited the separation capillary. However, significant flow rates for the sheath liquid are required in order to operate stably, leading to significant dilution of the analytes. In addition, the flow of sheath liquid around the capillary terminus can create suction that may lead to a parabolic flow profile and reduced resolution. The currently available commercial CE-MS interfaces employ a three-tube coaxial design. The innermost tube is the separation capillary, which protrudes slightly  20  from the second tube through which the sheath liquid flows. The third, outer tube delivers a flow of sheath gas to help improve spray stability and solvent evaporation from the generated micro droplets.  Figure 1.4: Common sheath-flow interface arrangements. A: Coaxial sheath flow interface with sheath gas, B: Liquid junction interface, C: Pressurized liquid junction interface. Most sheath-flow interfaces use a straight tube to deliver sheath liquid around the separation capillary, however it is also possible to use a tapered tube with the CE capillary sitting slightly back from the tip. The advantage in this case is that mixing of CE effluent and sheath liquid is more complete, leading to improved stability of the  21  electrospray. Interfaces of this type have been constructed using wide-bore fused silica capillaries that have been pulled to a taper54, or commercial ESI emitters consisting of a tapered stainless steel needle55. The processes involved in shaping tapered tips are described in detail in section 1.6.2.2.1. In order to reduce the dilution factor while retaining the solution-modifying benefits of a sheath-flow interface, alternative arrangements have been demonstrated. One common interface type is the liquid junction, illustrated in Figure 1.4B, where sheath liquid is added to CE effluent at a tee junction through a narrow (25-50 μm) gap between the separation capillary and spray needle56-58. Unfortunately, the orthogonal mixing process can broaden peaks and lead to reduced efficiency for the separation44, 59. As a result, coaxial arrangements have remained the preferred design for the addition of sheath liquid. More recently, sheath-flow interfaces have been developed that use even lower flow rates (some less than 200 nL/min). One of these, the pressurized liquid junction, is similar to the original liquid junction design, however the junction is slightly wider (up to 300 μm) and is located in a pressurized reservoir of make-up liquid, as shown in Figure 1.4C. The addition of pressure helps to prevent defocusing of the CE effluent in the gap region that would lead to reduced resolution. To prevent back-flow due to the pressure differential across the separation capillary the inlet vial must also be pressurized. The conductive make-up liquid establishes electrical contact between the background electrolyte (BGE) and the shared electrode and also supplies a consistent flow to the electrospray tip in cases when the flow rate from CE is insufficient60. The additional flow  22  introduced in these ‘pressurized junction’ interfaces does add a dilution factor, however it is much less than in the case of more traditional sheath-flow interfaces. A sheath-flow nanospray interface has also been developed using a coaxial arrangement of silica capillaries54. The terminal end of the narrow separation capillary is coated with gold to create an electrical contact outside of the separation path. It is then inserted into a larger-diameter silica capillary with the end pulled to a taper. The coaxial capillary assembly is mounted in a standard ionspray interface. Sheath liquid is passed through the larger capillary and flows over the end of the separation capillary, carrying CE effluent to the tapered tip. The dilution factor with this arrangement is less than ½ and the total flow is approximately 500 nL/min. Another strategy for low volume sheath-flow electrospray interface uses a beveled tip to reduce the required flow rates for stable spray operation without significantly reducing the inner diameter of the emitter tip61. One application of the beveled tip uses a novel mixing arrangement that is neither coaxial nor a traditional liquid junction. The CE effluent and sheath liquid are delivered to the emitter tip in parallel capillaries and mixing occurs directly at the emitter orifice62. The rationale behind the beveled tip shape will be discussed later, as it does not relate exclusively to sheath-flow interfaces. Despite the dilution that is inherent to sheath-flow interfaces, they offer a number of important advantages. Because the solution exiting the interface is primarily made up of sheath liquid, it is possible to use background electrolytes or additives in the CE separation that might otherwise be incompatible with ESI-MS. It is also advantageous to use the sheath-liquid to create electrical contact at the CE capillary terminus, as this  23  keeps the electrolysis process away from the analyte path. Finally, sheath-flow interfaces are generally robust and well suited to commercialization. 1.6.2.2. Sheathless interfaces The first CE-ESI-MS interface, introduced by Olivares and coworkers in 1987, was in fact a sheathless interface40. However, that design was quickly abandoned in favour of a coaxial sheath-flow interface in order to increase the stability and compatibility of the hyphenated techniques. More recently, sheathless interfaces have regained popularity as researchers seek to improve detection limits. Although many different types of interfaces have been constructed, the improvements nearly always relate to one of three fundamental characteristics: the shaping of the emitter tip, the strategy for establishing electrical contact, and the assembly of the required components, when multiple pieces are used. 1.6.2.2.1 Tip shaping The shape of the emitter tip used in CE-MS is critical as it determines the conditions under which a stable electrospray can be achieved. In order to maintain a stable spray while accommodating the flow rates produced by the electroosmotic flow in CE, which are much lower than those typically used in electrospray applications, tip size must be reduced. The most common method for making emitters for CE-ESI-MS is to pull a fused silica capillary to a thin tip. This is achieved by suspending a small weight (10-50 g) at one end of the capillary and heating a section above the weight so that the capillary is stretched to form a fine filament. The filament can then be trimmed to give a clean tip edge with both the inner and outer diameters smaller than those of the original capillary dimensions. Often, this process is followed by etching with hydrofluoric acid in  24  order to refine the tip. Etching can also be used alone to shape capillary tips. Moini and coworkers have used etching with 49% hydrofluoric acid to reduce the thickness of relatively long sections of capillary walls, including the tip, to approximately 10 μm, rather than sharpening by mechanical means63. Capillary tips can also be shaped by mechanically grinding the capillary terminus to the desired shape on a rotating wet sanding machine60 or using a handheld drill and emery cloth64. Unlike the pulling procedure, etching and mechanical shaping retain the original inner diameter of the capillary, so that there is no distortion of flow at the tip, thus avoiding the undesired addition of backpressure into the separation capillary. These procedures can be carried out at the end of the separation capillary to give a one-piece CE-MS column that, with the addition of an electrical contact, may function as the electrospray emitter. They can be also used to make shorter, disposable emitter tips, which are connected to the separation capillary by butting together the two segments inside a sleeve to form a junction. The advantage of the two-piece arrangement is that the tips can be replaced without changing the separation capillary. Although this does introduce the issue of alignment of the different pieces, it offers the advantages of increased flexibility and the option of disposable emitters, which can be quickly replaced when performance begins to decline while retaining the original separation capillary65. Although glass and silica tips are by far the most commonly used materials, twopiece interfaces have also been introduced using emitters made of stainless steel. Generally, narrow bore stainless steel tubing is shaped into an emitter by either grinding with sand paper66, electro-polishing in an acid solution at low voltages67, 68 or by nipping with pincers69. Of these methods, only the nipping procedure significantly alters the  25  inner diameter of the tubing. Metal tips offer improved mechanical stability over capillary tips and simplify the issue of creating an electrical contact with the solution. However, redox reactions on the inner surface of the metal tip may lead to the formation of bubbles that can destabilize the electrospray. 1.6.2.2.2 Electrical contact One of the primary challenges in interfacing capillary electrophoresis with MS is creating a stable electrical contact at or near the capillary terminus that will serve as the terminal electrode for CE in the absence of an outlet vial, and also as the emitter electrode in the electrospray circuit. As previously mentioned, electrical contact can be simply created in the case where metal tips are used as the electrospray emitter. However, in the more common case of glass or silica emitters there are several different strategies that may be employed. Petersson and coworkers explored the possibility of using a thin film of static liquid between the capillary tip and a metal sheath pulled back slightly from the capillary tip70 to establish electrical contact. It has also been demonstrated that CE-ESI-MS can be performed with no electrode whatsoever at the capillary terminus. In this case electrical contact is established through gaseous ions in the space between the capillary tip and the orifice of the mass spectrometer71. Although this appears to offer an extremely simple solution to interfacing, the position of the capillary tip with respect to the mass spectrometer is critical and it is not possible to control the separation and spray voltages independently because voltage is distributed along the circuit based on resistance of the CE electrolyte and the ESI plume.  26  Perhaps the most widely used method for establishing electrical contact is to coat the outer surface of the (non-conductive) emitter tip with a conductive material such as gold72-74, silver75, copper76, nickel77 or graphite78-82, as shown in Figure 1.5A. Unfortunately, the lifetimes of coated tips are limited due to the high electrical fields applied to the metal coating at the tip. In most cases they can be used for a few days before the deterioration of the coating renders operation unstable. Stability may be improved by pre-treating the capillary surface or by mixing different materials into the coating72, 78, 83. An alternative to coating the capillary tip is to insert a wire electrode into the capillary channel in order to make electrical contact. Several different means to this end have been tested. When larger inner diameter capillaries are used, a thin wire electrode may be inserted into the end of the capillary channel84, as shown in Figure 1.5B, or into a small hole drilled near the capillary terminus85, as shown in Figure 1.5C. However, this creates turbulence and reduces the resolution of the CE separation. Turbulence can be reduced by using a hole filled with conductive gold epoxy rather than wire, however, as with any situation where electrolysis occurs within the separation channel this may lead to bubble formation inside the separation channel and destabilized spray86. Another strategy for creating electrical contact is to split the liquid flow from the capillary so that a portion of the flow contacts an outside electrode, creating an arrangement known as a split-flow interface. This arrangement is shown in Figure 1.5D. Splitting is achieved through a drilled hole or a small crack in a single capillary which serves both as the separation chamber and electrospray tip87. While this preserves the separation resolution, the difficulty in this strategy lies in creating reproducible holes or  27  cracks that give the desired split ratio between the two flow paths. Hydrofluoric acid may also be used to etch away sections of the outside surface of the fused silica capillary to the point where the capillary walls become porous. Electrical contact can then be made through the capillary wall, either by immersing the etched portion of the capillary in a buffer reservoir88, 89, or by inserting it into a metal sheath filled with a thin film of liquid63, 70, 90 as shown in Figure 1.5E. Although interfaces of this type have been shown to be quite successful, the production requires the use of concentrated hydrofluoric acid and the porous capillary tips are fragile.  Figure 1.5: Methods for creating electrical contact in sheathless interfaces. A – Conductive coating applied to the emitter tip, B – Wire inserted at tip, C – Wire inserted through hole, D – Split-flow interface with a metal sheath, E – Porous etched capillary, F – Two-piece interface with metal sleeve, G- Two-piece interface with microdialysis junction, G – Two-piece interface with conductive emitter. When the emitter and separation capillary are two distinct pieces, electrical contact can be made at the junction of the two sections. The separation capillary and emitter tip are closely butted together, such that no additional flow is introduced at the junction, and electrical contact is established through a surrounding electrolyte into which the terminal electrode is placed. Junctions of this type have been constructed using a 28  metal sleeve connected to the power source91 (Figure 1.5F), microdialysis tubing92 (Figure 1.5G) or a micro-tee93 to aid in the alignment of the two capillaries and to establish contact with an electrode. Junctions of this type can also be used to join the separation capillary with a stainless steel emitter tip68, as shown in Figure 1.5H. It has also been demonstrated that, when very narrow emitters are used, junctions can be formed by inserting the emitter capillary into the CE capillary terminus. A twopiece nanoelectrospray interface was constructed using an extremely fine glass emitter with an outer diameter smaller than the inner diameter of the CE capillary58. Inserting the emitter into the end of the CE capillary created a junction that would allow a fraction of the CE effluent to be transferred into the emitter for ionization by nanoelectrospray. The junction was located within an electrolyte reservoir, which served to establish electrical contact with the shared electrode, and a continuous flow of electrolyte through the reservoir was maintained in order to remove the products of electrolysis and, presumably, the analytes not transferred to the emitter. The electrolyte flow also generates a pressure within the junction reservoir that increases flow into the emitter in a manner similar to the pressurized junction described earlier. 1.6.2.2.3 Ease of use and longevity Although two-piece interfaces are convenient in the fact that the separation capillary can be replaced independently of the spray tip, or vice versa, alignment of the two pieces is not trivial and any misalignment will impact the quality of the CE separation. Often the junction is located within a short section of tubing, either metal or plastic, that is used as a sleeve to align the two sections, however because the inner  29  diameter of the sleeve is not a perfect match for the outer diameter of the capillaries this does not guarantee a straight path for the analyte. The use of nanoelectrospray tips offers advantages in terms of sensitivity, however fine emitter tips with very narrow orifices are far more susceptible to clogging and to deterioration of the conductive coating than larger-scale metal ESI needles. One proposed solution is to manufacture disposable tips that can be replaced when they begin to show signs of deterioration65. Disposable tips can also simplify the process of CE-MS using treated capillaries, since the separation capillary is prepared independently from the emitter tip. However, replacing a disposable tip still requires careful attention to alignment and positioning with respect to the mass spectrometer interface. As such, they require a skilled operator and are not practical for integration into systems that are intended for use in widespread applications. An automated nanoelectrospray system using disposable tips is commercially available, however no similar technology has been reported that would be compatible with CE-MS. In order for this approach to be successful the issues of reproducible positioning and alignment will need to be resolved. Metal spray tips offer excellent robustness as emitters, as they are more mechanically stable and require no additional coating, however they have not gained widespread popularity due to issues including electrolysis reactions inside the tip and an increased likelihood of discharge between the emitter and counter electrode. They also must be carefully aligned with the separation capillary. These issues are discussed in the later section of this review.  30  1.6.3. Fundamental concerns for CE-MS interface design 1.6.3.1. Electrospray ionization The preferred spraying mode of ESI for MS analysis is the cone-jet mode, which involves the formation of a stable Taylor cone that continuously ejects fine droplets from its apex. Although it is often overlooked in CE-MS literature, the formation of a stable cone-jet is not a trivial matter94, 95. In order to access this state the forces acting on the Taylor cone, including surface tension, charge density, electric field and flow rate, must be optimized. Improperly set conditions will quickly destabilize the Taylor cone and lead to a mode of spraying which gives inferior performance in terms of droplet size, ion flux stability and/or directionality of the spray96. 1.6.3.2. Flow rate considerations The bulk flow rate of a CE separation can vary widely depending on the capillary diameter, length, separation potential, background electrolyte composition and capillary pre-treatment that are used. Changing any of these factors, and therefore the flow rate, will affect the stability of the electrospray ionization process. The optimal flow rate of the ESI process is affected by a number of variables, including a large number of solution properties, the spray voltage and the tip geometry. Although the relationships can be difficult to elucidate, owing to the complex interplay between these variables, in general smaller tips lead to lower optimal flow rates23. When nano ESI is performed alone, the most stable mode of operation is to allow the electrospray process to determine the rate of sample flowing to the emitter tip. This is achieved by filling the sample into a capillary which is left open at the end and held level with the emitter tip to prevent flows driven by gravity95. However, when used in  31  conjunction with capillary electrophoresis the nanoelectrospray flow rate is forced to match that of the separation. Although increased flow rates do deliver more analyte ions to the end of the separation capillary, the gains in MS signal intensity are far from quantitative. In fact, the flow rate used CE-ESI-MS can affect the type of detection observed. At lower flow rates, generally below 150 nL/min, the ESI-MS acts as a mass-sensitive detector, with increasing signal for increased volumetric flow rates and analyte concentrations65, 77, 79, 97. However, as the flow rate continues to increase the signal intensity becomes independent from flow rate, acting as a concentration-sensitive detector. This change in detection behaviour has been observed using a number of different interfaces and is due to the fact that spray current increases with approximately the square root of the flow rate, so that the charging efficiency is decreased as the flow rate increases98. At the point where detection becomes concentration sensitive, the limiting factor for ionization is the ability of the electrospray needle to charge the analytes in solution. The optimal flow rate of any ESI system is limited by the total ESI current that can be produced. Often, a flow rate is chosen that lies just above the threshold of concentration-sensitive detection, so that small variations in EOF do not affect signal intensity. Although CE-MS interfaces have been designed for a range of possible flow rates, there are important advantages to using lower flow rates. Smaller tip radii and lower flow rates both contribute to decreasing the size of the initial droplet ejected from the Taylor cone18. This has several important effects on the ions observed in the MS analysis. Firstly, because smaller droplets have a much larger surface to volume ratio they require less solvent evaporation to convert the analytes into gaseous ions, leading to  32  a more effective ionization of the analytes in solution. The result of this more effective desolvation of ions is that surface activity of analytes becomes less important in determining ion intensity inside the MS22, 23. It has also been shown that nanoelectrospray has an improved tolerance to higher salt concentrations than traditional ESI and can operate in a stable manner with lower organic content24. These are both important advantages when coupling to CE. 1.6.3.3. Emitter tip geometry In the ideal cone-jet mode of electrospray, the dimensions of the Taylor cone are determined by the geometry of the needle tip. In nearly all cases, the outer diameter of the tip acts as the support for the cone base 95. As previously mentioned, it has also been shown that the size of droplets formed in cone-jet mode depends on the size of the Taylor cone, and therefore the emitter tip. The inner diameter of the tip will play a role in determining the flow rate exiting the electrospray emitter and therefore also on droplet size. The effect of the taper geometry on electrospray performance is more difficult to elucidate. Because electric field around a conductive surface depends on the sharpness of the surface features, sharper tips can be operated with lower spray potentials. Simulation software can be used to model the electric field for a given tip geometry 99, however it remains difficult to predict electrospray behaviour. Experimental studies have shown that longer tapers on nanoelectrospray tips lead, not surprisingly, to lower onset voltages for electrospray, but also to a greater ability to achieve control over low flow rate as a function spray potential 23. The work of Ishihama and coworkers using metal spray tips appears to demonstrate that optimal performance at lower flow rates depends strongly on  33  the presence of a tapered tip and improves with reduced inner and outer diameters of the emitter 68. An often-overlooked aspect of emitter tip design is that the shape need not be totally symmetrical. Although it is a topic that has not yet been widely explored, there are examples where alternative tip shapes led to improved performance. Her and coworkers created beveled tips for CE-MS by mechanically grinding 75 μm ID capillaries at an angle 79. The resulting beveled tips showed an optimal flow rate of 200 nL/min for concentration-sensitive detection, whereas the blunt tips showed no optimal value up to 500 nL/min. The flow-rate behaviour of the beveled tips is similar to that of a tapered tip with an inner diameter of 25 μm. Microscope images of the tip showed that a small Taylor cone forms at the sharpest point of the beveled tip. The angled shape is also more robust than drawn capillaries and is less prone to blockage of the orifice. In addition to geometry, when emitters are constructed from alternative materials the wettability of the material must be taken into consideration. It has been observed that, at lower flow rates and high electric field strengths, electrospray from a blunt stainless steel emitter occurs with a Taylor cone occupying only the area of the needle orifice rather than the entire outer diameter, due to the fact that stainless steel presents a significantly more hydrophobic surface than glass or silica 69. Although this observation has not been translated to CE-MS applications, it does suggest that the geometry of tips constructed from or coated with stainless steel, graphite or other metals may not require the ultra-fine wall dimensions that have led to the greatest success with emitters constructed from glass or silica.  34  1.6.3.4. CE current and spray currents Most CE-MS systems use two independent high voltage sources sharing a common ground to provide power to CE inlet vial and to the shared CE-ESI electrode at the capillary terminus. However, difficulty arises in accurately controlling the CE and spray potentials because CE currents are generally at least an order of magnitude larger than the currents generated and carried by the electrospray process. As a result, the power source at the shared electrode is required to simultaneously provide several kilovolts of positive potential and sink a significant current from the CE separation. The excess current passing through the internal resistance of the power source can cause the actual potential at the shared electrode to deviate significantly from the set value. One suggested solution to this problem is to use an adjustable resistor between the shared electrode and ground to sink current and set the desired potential at that point 57. The resistor must be adjustable and dynamically monitored in order to avoid drift over the course of separation due to heating and other factors. 1.6.3.5. Electrochemistry reactions in CE-ESI-MS Another complication of the current mismatch between the separation and spray is that the electrochemical reactions related to the CE circuit will be dominant over those resulting from the ESI circuit. When CE is operated in the positive polarity mode the major electrochemical reaction at the terminal electrode will be the reduction of water to produce hydrogen gas and hydroxide ions. This presents two potential problems: bubbles of hydrogen gas may reduce the resolution of the separation or stability of the spray as the solution exits the emitter. In extreme cases a build up of gas in the capillary could interrupt the current for either the CE or ESI process. The second problem is that the  35  addition of hydroxide ions can cause a significant change of the solution pH and affect the proportion of protonated analyte that can go on to form positive gaseous ions. Experiments by Smith and Moini using an inserted wire CE-ESI interface yielded several interesting and important observations on the nature of the electrochemical reactions occurring at the shared electrode 100. The first of these was that electrochemical reactions appeared to take place almost exclusively at the portion of the electrode that was nearest to the relevant counter electrode. That is to say that electrolysis related to the CE circuit occurs as far up-stream as possible, while the electrolysis due to the ESI circuit occurs independently at the downstream end of the electrode closest to the emitter tip. It was also observed that opposite redox reactions at the shared electrode do not cancel each other out, but rather occur side-by-side at different locations on the same metal surface. Therefore, care must be taken in using the net current at the shared electrode to estimate the extent of electrolysis when both oxidation and reduction are taking place. In this case the measured current will be representative only of the difference in the currents between the two processes. It has been demonstrated that the incorporation of redox agents such as pbenzoquinone or hydroquinone, depending on the separation polarity, into the background electrolyte can suppress undesirable electrochemical reactions of the shared electrode with analytes and other buffer components 101. However, the preferred strategy for avoiding complications due to electrolysis products is to keep the electrochemical reactions away from the path of analyte.  36  1.6.3.6. Location of the electrical contact Although strategies for establishing electrical contact with the CE BGE have already been discussed, it must be noted that the electrical circuits created by the various modes of contact are not all equivalent 102. When the shared CE-ESI electrode is placed outside of the direct flow path, the electrolytes in solution carry current to the separation capillary terminus and to the electrospray tip. However, there is a significant resistance to this motion of ions through solution, which increases with the distance from the electrode. Therefore, the potential applied to the terminal electrode will not be equal to the potential present on the capillary terminus, or at the emitter tip as it would in the case of a metalized emitter. The presence of solution resistance between the power source and emitter tip make the use of non-conducting needles for CE-ESI-MS electrically analogous to placing a resistor between the power source and conducting emitter. There have been reports of improved spray stability using non-conductive emitters in the ESI literature 103. However, precise control over the terminal electrode potential is crucial to achieving reproducible CE separations. Therefore, electrolyte composition, electrode positioning and temperature will need to be tightly controlled in order to maintain a constant solution resistance. When contact with the shared electrode is made through a crack, hole, an etched portion of capillary, or through a gap between capillaries, the size of the crack or hole, degree of porosity of the capillary wall, or the size of the gap will also play a role in determining the resistance between the power source and emitter tip. Care must therefore be taken to ensure that these modifications are performed reproducibly.  37  1.7. RESEARCH OBJECTIVES 1.7.1. Design of a novel CE-MS hyphenation strategy Despite the wide variety of tools available for biomolecule characterization, there are still areas where the available technology is not meeting the demands of research. While liquid chromatography – mass spectrometry has become a tool for routine analysis, capillary electrophoresis, which provides superior separations to LC in a number of applications, has not been hyphenated with mass spectrometry in a manner that has been suitable for widespread adoption. The currently available commercial interfaces have helped to bring the CE-MS to a wider audience. However, because they use a significant flow of sheath liquid, the sensitivity required for many biological applications cannot easily be met. There are also many types of background electrolytes and additives that are useful in achieving superior separation efficiencies that remain incompatible with CEESI-MS. A sensitive, commercial CE-ESI-MS platform would have the potential to become the technology of choice for many routine analyses as well as more challenging research and development activities because it offers excellent inter-lab reproducibility, since it is easy to recreate experimental conditions and there are no stationary phases involved. However, for this to be possible an interfacing approach that offers ease of use, robustness, versatility and sensitivity is required. The first part of this thesis describes our work on developing a new type of CE-MS interface with the goal of satisfying all of these criteria simultaneously.  38  1.7.2. Application of the interface to problems in biomolecular characterization In order to evaluate the interface as a tool for biomolecular characterization, the second part of this thesis investigates its application to the analysis of peptides, proteins and glycans. Each of these samples presents unique challenges to interfacing, including the use of different flow rate conditions, capillary wall coatings, CE instrument polarities, and the use of negative-ion mode ESI-MS. Finally, while CE-MS usually refers to capillary zone electrophoresis separation, there are other modes of CE that are even more challenging to couple with MS detection. This section also demonstrates the interface as a means of hyphenating capillary isoelectric focusing with ESI-MS detection. 1.7.3. Correction strategies for liquid chromatography Another area which could be further exploited to give more confident biomolecule characterization is in gradient reverse-phase chromatography. While these separations are among the most commonly used prior to MS analysis, in many cases they are used only as a means of reducing the number of species entering the MS simultaneously and the retention times are not considered whatsoever. The reason for this is that the retention times are not directly related to any intrinsic characteristic of the biomolecule, and can vary run-to-run due to small variations in the column condition. Additionally, it is difficult to correlate results when different gradient programs are used. If it were possible to convert the reverse-phase retention times to a standard scale of hydrophobicity, based on retention relative to standard compounds, this could simultaneously correct for retention time variations and aid in biomolecule identification. The final part of this thesis presents an evaluation of two-marker correction strategies and explores the potential of using them to develop a standard hydrophobicity index.  39  Section A: Design of a novel interfacing strategy for capillary electrophoresis – electrospray ionization – mass spectrometry  40  2. Chapter 2: Decoupling capillary electrophoresis and electrospray ionization for a more robust interface with mass spectrometry  41  2.1. INTRODUCTION Capillary electrophoresis – electrospray ionization – mass spectrometry (CE-ESIMS) has been used as an analytical tool for proteomics50, 104, 105, metabolomic106-108, pharmaceutical109, and many other areas of research49, 51, 110. Despite its demonstrated potential, CE-ESI-MS has remained a specialist technique and has not been adopted by the wider research and development community. This is due in large part to the fact that most commercially available CE-ESI-MS interfaces cannot match the sensitivity and robustness achievable by liquid chromatography – mass spectrometry (LC-MS). Although CE and ESI both work well for polar and charged analytes ranging from small molecules to large proteins, reconciling the physical and chemical processes that drive the separation and ionization processes presents a challenge in interface design52. Both CE and ESI require a stable voltage applied on a pair of electrodes and a steady flow of electrolyte, but the magnitude of the required current and the amount of solution, as well as the solvent environment, differ significantly. Many types of interfaces have been proposed49, 51, 111-113, however, in most cases there is a compromise between sensitivity and robustness111. While sheathless interfaces offer superior sensitivity, the challenges involved in creating an interface with no dilution of the CE effluent often lead to the use of fragile tip materials and involve complicated fabrication52. More physically robust sheathless interfaces employing metal tips have also been constructed67, 68, however, because the metal tip acts as the terminal electrode for the CE separation, bubble formation inside the  42  tip can be problematic. Often, separations must be performed with an assisting pressure to ensure that bubbles are pushed out of the tip. Despite their lower sensitivity, interfaces with a liquid sheath-flow offer several attractive features. The flowing sheath liquid surrounding the capillary terminus establishes electrical contact, eliminating the need for special treatment of the capillary and keeping electrolysis products outside of the separation capillary. The use of a makeup solution and a nebulizer gas allows greater versatility in the choice of background electrolyte (BGE) and range of flow rates, because the make-up solution dilutes undesirable components in the CE effluent. However, because in this case electric field is not the only driving force in droplet formation, the analyte ionization efficiency may be compromised. Sheath-flow interfaces have found more applications since the early years of CEMS development42 because they can be adapted to existing LC-MS setups by increasing the liquid flow to levels comparable to those found in LC. Sheath-flow interfaces can be categorized based on the location of the mixing of effluent and sheath liquid59. In the liquid junction design a make-up solution, which serves similar purposes to the sheath liquid, is added to the CE effluent at a junction between the separation capillary terminus and the entrance to the channel of the electrospray emitter56, 58, 60, 114. In the more common coaxial design, found in many commercial CE-ESI-MS systems53, the sheath liquid surrounds the capillary terminus and mixing of the two solutions occurs within the Taylor cone, immediately before ionization and desolvation42, 115. A nebulizer gas is often introduced by a third concentric metal tube to aid in solvent evaporation.  43  Although there are benefits to using a sheath-flow setup, the addition of the sheath liquid dilutes the samples and leads to a loss in sensitivity116. Because of the small injection volumes used in CE, this additional loss often limits the use of CE-MS to situations where the concentration of analyte is relatively high. More recently, sheathflow interfaces have been developed to use much lower flow rates than traditional sheathflow interfaces (some less than 200 nL/min, compared to 1 to 10 μL/min). Such interfaces include the pressurized liquid junction60 and beveled tips constructed from silica capillaries that reduce the required flow rates for stable spray operation without significantly reducing the inner diameter of the emitter tip61, 62. Another design for low-dilution sheath-flow interfaces uses the concept of “junction-at-the-tip”, which uses a tapered outer tube surrounding the separation capillary so that mixing of the CE effluent and sheath liquid occurs within the tip of the electrospray emitter54, 55, 117-120. A sheath-flow nanospray interface of this type has been developed using a coaxial arrangement of silica capillaries54. The terminal end of the narrow separation capillary is coated with gold to create an electrical contact outside of the separation path. It is then inserted into a larger diameter silica capillary with the end pulled to a taper. The dilution factor with this arrangement is reported to be less than one half, and the total flow rate of the combined solution is approximately 500 nL/min. In this chapter we introduce a more physically robust junction-at-the-tip interface that builds on the successful strategies of several earlier designs. A standard 365 μm outer diameter separation capillary is inserted all the way into a tapered stainless steel hollow electrospray emitter. The small volume between the capillary end and the inner walls of the needle electrode tip forms a flow-through micro-vial that completes the CE  44  setup requirement by acting as both the outlet vial and terminal electrode and allows stable CE operation regardless of how the effluent is processed downstream. The flowthrough micro-vial also allows the addition of a chemical modifier solution at low flow rates in order to provide a stable flow of solution to the needle tip and to increase the compatibility of the CE effluent with electrospray ionization while minimizing sample dilution. The metal emitter tip is beveled to increase spray stability and also to move the site of ionization away from the CE outlet and the emitter tip orifice. 2.2. MATERIALS AND METHODS 2.2.1. Materials Individual amino acids, hydroxyproline, sodium dodecylsulfate and benzene tricarboxylic acid were purchased from Sigma Chemical Co. (St. Louis, MO, USA). All analytes were obtained in solid form and used without further purification. Formic acid, isopropanol and methanol (HPLC grade) were purchased from Fisher Scientific (Nepean, ON, Canada). Stock solutions of all analytes were prepared in deionized water and stored at -20 °C. Standard solutions were prepared by diluting stock solution with background electrolyte prior to analysis. Fused silica capillary (75 and 50 μm ID, 365 μm OD) was purchased from Polymicro Technologies (Phoenix AZ). Standard stainless steel tee-unions and PEEK fittings were purchased from Upchurch Scientific (Oak Harbor WA). Custom-designed stainless steel hollow electrodes were fabricated (Cadence Inc., Staunton VA) from a piece of stainless steel tubing (OD 0.028”, ID 0.016”) that is tapered both internally and externally at one end. The tip of the needle was beveled, such that the face of the tip  45  forms a 60° angle with the central axis of the electrode. The tapered end has an orifice (75 μm cross section ID) that allows solution to exit the needle. The other end is laser welded to 1/16” OD stainless steel tubing in order to facilitate connection with standard fittings. Note that the beveled tip has a large surface area compared to the effluent exit orifice, as illustrated in Figure 2.1.  Figure 2.1: Schematic illustration of the interface apparatus, including a dissected view of the needle tip with inserted capillary (inset). 2.2.2. Interface configuration Figure 2.1 shows the schematic of the interface apparatus. The terminal end of the separation capillary is inserted into the stainless steel needle as far as possible, so that the end of the capillary reaches the point where the diameter of the tapered inner surface is equal to the outer diameter of the capillary. The capillary and needle are secured through the opposing ports of a stainless steel tee union using standard fittings. A second  46  capillary (75 μm ID, 90 cm) is attached to the orthogonal port of the tee union in order to deliver a flow of modifier solution into the needle from a grounded modifier vial. 2.2.3. Instrumentation All separations were performed using a P/ACE MDQ or PA800 capillary electrophoresis system with a modified capillary cartridge (Beckman Coulter Inc., Fullerton CA). The modified cartridge allows the instrument to apply the separation potential and electrical ground at the inlet and modifier vials, respectively, and to provide pressure at the modifier vial to generate a low volumetric flow rate of solution to the needle tip. The potential at the stainless steel interface was controlled either by an external power supply (Spellman HV, Hauppauge NY) or through the electrospray power supply of the MS instrument. The needle tip was positioned 0.5 to 1 cm from the front of the MS inlet cone or heated capillary inlet, and fine control of the interface position was achieved through the MS instrument’s integrated positioning apparatus. Detection for most experiments was carried out using a Finnigan LCQ*DUO ion trap mass spectrometer (Thermo Scientific, Waltham MA), both with the present interface and with Beckman Coulter’s commercial CE-ESI-MS sheath-flow interface (provided with the LCQ*DUO), for comparison. The optimal sheath liquid flow rate for the commercial interface was found to be 1.0 μL/min, using a 22-guage (0.0285” O.D., 0.016” I.D.) ESI probe with the separation capillary extruding 1 mm from the ESI probe. At this low sheath-flow rate, the presence of a nebulizer gas did not improve performance, so none was used. For the analysis of amino acids, mass analysis was also carried out using a Micromass Q-TOF1 mass spectrometer (Waters, Milford MA) with the standard ESI ion  47  source removed. A potential of 20 V was applied to the inlet cone, which was maintained at a 60°C, and the mass analysis range was 70-250 m/z. 2.3. RESULTS AND DISCUSSION 2.3.1. Interface design The primary objective in designing the present interface was to effectively decouple the capillary electrophoresis and electrospray processes so that each could work under optimal conditions without adversely affecting the other. Other design priorities were that the interface be physically robust and provide maximum sensitivity for analyte detection. In order to satisfy these objectives it is necessary to arrange the interface in a way that allows the CE and ESI processes to operate independently in a side-by-side manner. A junction-at-the-tip arrangement55, where the separation capillary is inserted into a hollow metal electrode was chosen because the presence of the modifier liquid decouples the solution requirements of the two processes. As shown in Figure 2.1, the inner geometry of the needle tip is such that the inner diameter near the tip is less than the outer diameter of the CE separation capillary. When the capillary is fully inserted, it stops at the point where the internal diameter of the needle tapers to match the outer diameter of the polyimide coated capillary. This ensures both coaxial alignment of the capillary with the needle orifice and reproducible positioning of the capillary in the longitudinal direction. Standard fittings are used to hold the capillary in position, allowing the capillary to be easily replaced when necessary. The outer diameter of the capillary used in this interface is 365 μm, which allows the capillaries to retain their strength even if the outer coating is removed to create a  48  detection window for other detection techniques. The 150 μm outer diameter capillaries used in many other CE-MS interfaces may also be used, but are much more fragile. The volume contained between the capillary terminus and the inner walls of the needle tip constitutes a flow-through micro-vial that replaces both the outlet vial and terminal electrode of a typical capillary electrophoresis apparatus, providing electrical contact while allowing the analytes to pass through to the needle tip. Although the capillary is inserted as far as possible in the hollow needle, the junction of the two still allows passage of the modifier solution via the application of a low pressure to the modifier vial. The chemical modifier delivered through the perpendicular arm of the tee union passes around the end of the separation capillary and into the micro-vial. Filling or replenishment of the micro-vial is accomplished simply by flushing the modifier solution through the tee union, or by flushing the background electrolyte through the CE capillary prior to starting a separation. For a 365 μm OD capillary, the volume of the micro-vial is approximately 12 nanolitres. However, simulations of the fluid dynamics within the micro-vial show that the flowing modifier solution keeps analytes exiting the separation capillary focused along the central axis of the needle and moving towards the needle orifice. 2.3.2. Conductive electrode and tip geometry Although the metal needle acts as the shared electrode, providing contact both at the CE terminus and at the ESI site, these contacts are at different locations on the electrode. Based on previous reports observing the gas formation due to electrolysis100, it is expected that electrochemical reactions related to ESI should take place primarily on the outside surface of the needle, while those related to the CE process will take place  49  within the needle. The mismatch of currents between CE and ESI normally means that a significant current must be sunk at the shared electrode. Resistors have been previously used to control the ESI potential and sink excess current in CE-MS interfaces, replacing the ESI power source121. The grounded modifier capillary in the present interface acts as a liquid resistor that provides a sink for some of the excess CE current. This allows improved control and stability of the spray potential, as the power supply at the shared electrode is not well suited to act as a current sink when operating in positive polarity and provides unreliable potentials if forced to do so. Because the modifier capillary is only required to sink a portion of the CE current, the resistance of the liquid resistor does not need to be kept at an exact value, allowing flexibility in the composition of the modifier solution. Although it does not appear that the presence of the modifier capillary eliminates hydrolysis in the micro-vial, the steady flow of a modifier solution works to effectively push products of electrolysis out of the needle before their accumulation interrupts either the CE or ESI processes. Note that this pressure is not applied through the CE separation column. Although metal electrospray tips have been routinely used for LC-MS or infusion ESI-MS applications, there have been relatively few reports of metals tips used as electrospray emitters for CE-ESI-MS55, 67, 68, 122. Metal needles offer several desirable qualities. They can be micro-fabricated according to strict standards, so that batch-tobatch variability should not affect the interface performance. Although the fabrication of metal tips requires specialized machining tools, the cost of having custom needles produced is relatively low and, once made, the needles are extremely durable and simple to connect through standard fittings. The use of stainless steel allows experimentation  50  with novel geometries, such as beveled tips, that are difficult to achieve or maintain when fabricating spray tips from fused silica capillaries. It has also been demonstrated that the wetting properties of metal ESI emitters lead to different electrospray behaviours than those observed with emitters constructed from silica, such that the Taylor cone does not occupy the entire needle face at high spray potentials69. When a beveled tip shape is used, the Taylor cone is located at the sharpest point of the bevel rather than at the needle orifice. This results in a physical decoupling of the CE and ESI processes, such that small bubbles exiting the needle do not disrupt the stability of the Taylor cone. 2.3.3. Decoupling capillary electrophoresis and electrospray ionization solution requirements The addition of a modifier solution to the flow-through micro-vial gives additional versatility to the types of electrolytes and modes of separation that may be used. Capillary electrophoresis can be run in the absence of electrospray, such that the CE effluent exits the needle to form liquid drops. It is also possible to perform electrospray in the infusion mode, or while capillary electrophoresis is stopped, because the flow of modifier can supply solution to the electrospray tip even when the bulk solution within the separation capillary is static. The presence of the modifier solution also serves to prevent ESI-driven suction that has been previously reported for nanospray 95  . The stable mixing environment within the micro-vial also prevents the generation of  suction at the capillary terminus that is sometimes observed in traditional sheath-flow interfaces due to the high linear velocity of the sheath liquid or gas123. The addition of a modifier solution also allows this interface to be used for separations where the bulk flow is towards the capillary inlet, so long as the analytes of  51  interest migrate towards the capillary terminus. Figure 2.2 shows an example of this arrangement, where a solution of 1,2,4-benzenetricarboxylic acid (BCA) and sodium dodecylsulfate (SDS) is continuously infused under forward and reverse CE polarity with negative ion ESI-MS detection. Under normal CE polarity the EOF is strong enough to push the negatively-charged SDS out of the capillary, as seen in Figure 2.2A, despite the fact that its electrophoretic mobility is towards the inlet. When the polarity and EOF are reversed, SDS is pushed by the bulk flow towards the inlet. However, BCA, which carries a greater negative charge, migrates towards the outlet and is observed in the negative ion mass spectrum, as shown in Figure 2.2B.  Figure 2.2: Continuous separation and infusion of sodium dodecylsulfate and 1,2,4benzene tricarboxylic acid with negative ion electrospray. (A) EOF flows toward the CE outlet. CE inlet voltage +30 kV, Sample concentration 0.1 mM in BGE. MS scan range m/z 250~280. (B) EOF flows towards the CE inlet. CE inlet voltage -30 kV, Sample concentration 0.5 mM in BGE. MS scan range m/z 200 ~ 220. In both cases, 75 m ID, 75 cm bare fused silica capillary, ESI -3.3 kV, BGE and Modifier: methanol/20mM pH 7.3 NH4AC/ H2O (v/v/v) 75:15:10. 52  This type of separation is not possible in traditional sheath-flow arrangements, because the sheath liquid and CE effluent mix in the Taylor cone and the sheath liquid has a high linear velocity towards the MS inlet. In the present interface design the CE effluent and modifier solution come into contact within the stable environment of the flow-through micro-vial in a way that allows the modifier solution to enter the capillary outlet if necessary. Continuous infusion was used in this case because the combination of a long capillary and the small forward velocity of BCA resulted in prohibitively long analysis times when a normal injection was used. This experiment is a very simple example of how the chemical modifier can be used to alter the analyte environment between the CE and ESI processes. Capillary electrophoresis is such a powerful separation technique because the range of background electrolytes and additives available allow precise tuning of separation conditions to maximize selectivity and efficiency. However, many of the buffer and additive ingredients typically used in CE background electrolytes can have a negative impact on the quality of mass analysis95, 98, 124, 125. The utility of sheath and make-up liquids in modifying BGE composition to improve ESI-MS compatibility has been well established for traditional sheath-flow interfaces. Similar concepts exist in liquid chromatography. For example, a modifying solution has been added to LC effluent to counteract the ionization suppression due to trifluoroacetic acid in the mobile phase 126. Adjustment of this type to the chemical environment of the analytes can significantly increase the detection sensitivity by optimizing ionization conditions. In this new interface design, the modifier solution can be used in a variety of ways and over a wide range of flow rates to alter the composition of the CE effluent as it exits the separation capillary. As with  53  any separation where the composition of the electrolytes at the capillary inlet and outlet differ, it is important to remember that this may lead to the formation of ionic boundaries in the CE capillary127. 2.3.4. Comparison with commercial sheath flow interface performance A group of eighteen amino acids was separated and detected with CE-MS using both the present interface and a commercial sheath flow interface. Two different MS instruments, the Finnigan ion trap MS and a Micromass Q-TOF MS, were also compared using the present interface. The amino acids were chosen because they represent a group of molecules with very different properties, ranging from hydrophilic to hydrophobic and with side chains having different functionalities. The molecular masses of the amino acids range from 75 (glycine) to 240 Daltons for the Cys-Cys dimer, which is the oxidation product of Cys. A weak signal for the Cys monomer is observable at m/z 122, however it appears that the majority of the analyte is present in the oxidized form. Optimization of the modifier flow rate was carried out manually by infusing a solution of amino acids through the CE capillary at a constant rate while varying the modifier flow. As shown in Figure 2.3, the ESI-MS signal intensity for most amino acids increased as the flow rate of modifier solution decreased, due to the reduced dilution factor. The only amino acid to deviate significantly from this trend was glycine, as the result of an isobaric interference with the background electrolyte at m/z 76. At modifier flow rates below 0.1 μL/min the signal to noise ratio for all amino acids decreased dramatically, making this region poorly suited for quantitative analysis. The decrease in signal stability is attributed to both the limitations on the precision of the applied pressure required to generate such low flow rates, and the fact that the stainless steel electrospray  54  needles operate best at flow rates greater than 0.2 μL/min. A modifier flow rate of 0.2  1.1  1.1  1.0  1.0  0.9  0.9  0.8  0.8  0.7  0.7  0.6  0.6  0.5  0.5  0.4  0.4  0.3  0.3  0.2 0.0  0.1  0.2  0.3  0.4  0.5  Normalized SNR  Normalized ESI-MS intensity  μL/min was chosen for subsequent amino acid separations.  0.2  modifier flow rate [µL/min]  average SNR Ala Arg Asn Asp Cys-Cys Gln&Lys Glu Gly His Ile Met Phe Pro Ser Thr Trp Val  Figure 2.3: Effect of modifier flow rate on normalized amino acid signal intensity. Conditions: BGE and modifier 0.2% formic acid, 50% methanol; Sample: 20 µM amino acids in BGE; CE inlet 30 kV; ESI 3.4 kV; Micromass Q-TOF MS. Extracted ion electropherograms for all 18 amino acids are shown in Figure 2.4. The limit of detection (LOD) for each amino acid was determined according to the equation LOD =  3σ blank m  (2-1)  where σblank is the standard deviation of the background signal and m is the slope of the linear least squares fit of peak height as a function of concentration. The Q-TOF instrument provided the lowest LOD, as shown in Table 2-1, with an average of 0.3 µM. However, in a few cases (Asp, Pro) the linearity of the calibration curves was poor, as is commonly observed in ESI-MS analysis. In order to improve the linearity, an internal standard, hydroxyproline, was added to the amino acid solutions for analysis using the 55  ion trap MS. The signal to noise ratio and the peak heights over the concentration range 10 to 200 μM were used to estimate the limit of detection of individual amino acids for both our junction-at-the-tip interface and the commercially available sheath flow interface. As shown in Table 2-1, the junction-at-the-tip interface provided significantly improved detection limits (average 5-fold improvement) compared to the sheath-flow interface. This is a result of lower flow rates enabled by the junction-at-the tip arrangement compared to the sheath-flow interface, leading to reduced dilution of the analytes and increased ionization efficiency.  Figure 2.4: Extracted ion electropherograms for a mixture of 18 amino acids. Conditions: Sample concentration, Gly 20 μM, all other amino acids, 10 μM; injection 1.0 psi for 10 s; BGE and modifier, 0.2% formic acid, 50% methanol; CE inlet 30 kV; ESI 3.5 kV; modifier flow rate 0.2 μL/min; Micromass Q-TOF MS.  56  Amino acid Ala Arg Asn Asp Cys-Cys Gly Glu Gln His Ile Lys Met Phe Pro Ser Thr Trp Val  Limits of detection (µM) Ion trap with Q-TOF Ion trap sheath-flow interface 0.3 0.7 3.8 0.2 0.1 0.4 0.3 1.7 7.4 0.3 0.6 2.3 0.5 1.0 11 0.7 1.1 6.3 0.1 0.3 1.3 0.2 0.2 1.1 0.4 0.4 2.4 0.2 0.1 0.4 0.1 0.2 0.4 0.2 1.2 2.3 0.1 0.1 2.4 0.2 0.9 2.4 0.5 3.4 5.3 0.1 1.7 2.6 0.1 0.1 0.2 0.2 0.3 1.3  Fold improvement over sheath-flow interface 5 3 4 4 11 6 4 4 6 4 2 2 17 3 2 2 3 5  Table 2-1: Limits of detection for amino acid analysis with different configurations and MS instruments Q-TOF MS: conditions as in Figure 2.4. Calibration based on triplicate analysis of peak areas over the range 2 - 40 µmol/L Ion trap MS: Calibration based on triplicate analysis of peak heights over the range 10 – 200 µmol/L with Pro-OH internal standard Ion trap MS with present interface: CE inlet 24 kV: ESI 4.3 kV; CE capillary 75 cm; modifier capillary 90 cm; 0.4 psi applied to both the CE inlet vial and modifier vials; modifier flow rate 0.1µL/min; MS scan range m/z 70 – 250; heated capillary temperature 230 °C; ion trap injection time 120 mS. Ion trap MS with sheath flow interface: as for present interface, except CE inlet 25 kV, ES voltage, 5 kV sheath flow rate 1 μL/min.  57  2.4. CONCLUDING REMARKS The interface described in this work provides a physically robust and easy to use alternative to the interfaces used in current CE-MS hyphenation techniques. The interface can be easily integrated with commercial CE and MS instruments, making use of the simpler and more reproducible automated controls of sample injection, voltage ramping and modifier flow rate regulation. Because no special treatment of the separation capillary is required, it can be used with capillaries of any dimension and with any type of capillary coating or interior surface modification. The addition of the modifier solution provides a simple means for increasing the stability of both separation and detection with only minimal dilution of the analyte and enables the use of separation conditions that would not be possible with most low flow rate CE-MS interfaces. Additional research is being carried out to further investigate how the geometry of the conductive emitter affects electrospray performance and on the effect of varying the modifier flow rate on peak shapes, based on the fluid dynamics in the micro-vial.  58  3.  Chapter 3: Investigation of asymmetrical emitter geometries for increased range of electrospray flow rates  59  3.1. INTRODUCTION One of the challenges in designing effective interfaces for coupling liquid chromatography (LC) or capillary electrophoresis (CE) with electrospray ionization – mass spectrometry (ESI-MS) is that, in order to process effluent from a variety of separations, the interface must work over a wide range of flow rates. For capillary electrophoresis these flow rates range from near zero nL/min, when there is no electroosmotic flow (EOF) present, to a few hundred nL/min when there is a strong EOF. However, most electrospray emitters are designed to give optimal performance over a relatively narrow range of flow rates, determined primarily by the inner and outer diameters of the sprayer tip111. Although extensive research has helped to improve the theoretical understanding of the fundamental processes of ESI, it remains difficult or impossible to accurately predict electrospray behavior based on experimental parameters95. Depending on the applied potential, flow rate and solution properties, electrospray may proceed by a number of different modes, including stable and pulsating regimes96, 128. Nanoelectrospray (nanoESI) and microelectrospray (μESI) have not been well defined in terms of the range of flow rates that they encompass. Estimates of the flow rate at which nanoESI becomes µESI, as indicated by a change from concentrationsensitive to mass-sensitive detection98 and the presence of the associated benefits of nanoESI (decreased dependence of ionization efficiency on analyte surface activity and salt or surfactant concentration22), range from 20 to150 nL/min18, 22, 129, 130. For a given emitter the flow rate at which this transition occurs depends primarily on geometry. For  60  µESI and larger emitters, the stability of the electrospray signal at low flow rates may deteriorate before mass-sensitive detection can be observed. In this case, the optimal value is often the lowest possible flow rate that will maintain a stable ESI, as shown in Figure 3.1. Unlike nanoESI, there are no associated characteristics that can be used to define μESI. However, it is a useful distinction that indicates a range of flow rates, generally less than 1µL/min, that are more compatible with hyphenation to separation techniques such as CE and µHLPC than traditional ESI.  Figure 3.1: Typical effect of flow rate on normalized signal for a stainless steel µESI emitter for a mixture of 15 amino acids under continuous infusion. When coupling ESI-MS online to LC or CE, the flow rate of electrospray is forced to match that used for the separation. As such, it is important to match the optimal flow rate and electrospray potential for a particular ESI emitter with the flow rate of the separation technique to maintain acceptable signal stability and sensitivity. This type of investigation is often carried out using a fixed value for the electrospray potential68, which may not completely characterize the performance of the emitter. Another common 61  practice is to adjust the spray potential at each flow rate in order to maintain a stable cone-jet electrospray, which has been generally regarded as the optimal electrospray regime for MS detection131. However, more recent studies have brought this assumption into doubt, suggesting that pulsating modes of electrospray may provide better sensitivity than the cone-jet mode128. When the flow rate is determined solely by the electrospray process, as in some nanoESI application, the applied ESI potential and the inner and outer diameters of the emitter appear to be the most significant determining factors132, 133. For fused-silica emitters, the outer diameter of the tip acts as the base for the Taylor cone, such that the wall thickness can significantly affect the optimal flow rate range95. For emitters constructed from stainless steel, or other materials that have a lower wettability than silica, the base of the Taylor cone may be restricted to the inner diameter of the emitter orifice under certain conditions69, 129. As a result, emitters constructed from stainless steel or other hydrophobic materials can operate at much lower flow rates than fused silica emitters of similar dimensions. For both stainless steel and fused silica emitters, it has been demonstrated that reducing the wall thickness leads to better stability at lower flow rates22, 68. Sprayers with a beveled tip may provide an alternative to small diameter symmetrical emitter because of the sharp tip and the gradual increase in cross section area. Her and coworkers demonstrated that beveled emitters provided improved spray stability at low flow rates61, 62, 79, 119. Although beveling the tip does not affect the inner diameter of the emitter, it generates an electric field distribution that causes the Taylor cone to form at the sharpest point of the beveled surface. We recently presented a CE-MS  62  interface using a stainless steel hollow needle with a beveled sprayer tip for low flow rates CE operations134. Observations of the Taylor cone varying in size in response to changes in flow rate have lead us to hypothesize that the beveled geometry may have the potential to provide stable ESI over a wider range of flow rates than symmetrical emitters. In the interest of quantifying the effect of emitter geometry, this chapter presents a comparison of electrospray performance as a function of flow rate for six different conductive emitters, including sharp, blunt and beveled tips. 3.2. MATERIALS AND METHODS 3.2.1. Chemicals and materials Individual amino acid standards were purchased from Sigma-Aldrich (St. Louis, MO, USA). Formic acid and methanol were purchased from Fisher Scientific (Nepean, ON, Canada). All chemicals were of analytical grade or better and were used without further purification. The background electrolyte of 0.2% (v/v) formic acid and 50% methanol in water was filtered thought a 0.45 µm pore size membrane prior to use. A sample solution of 20 μM each of proline, threonine, isoleucine and arginine was prepared by diluting the amino acid standards in the background electrolyte. This sample solution was used in all experiments. Fused silica capillary (75 and 50 µm ID, 365 μm OD) was purchased from Polymicro Technologies (Phoenix AZ). The six custom fabricated stainless steel electrospray needles studied are shown in Figure 3.2. Emitters 4 and 5 were modified inhouse from #2 tips by grinding at an angle to achieve the desired bevels (30° and 45°, respectively). Emitter 6 was beveled during the micro-fabrication process in order to  63  achieve a 35° angle with a significantly smaller tip surface area than emitters 4 and 5. The dimensions of all needles are listed in Table 3-1.  Table 3-1: Dimensions of the stainless steel emitters involved in the investigation shaft orifice diameter diameter (mm) (mm) 1 Taper 0.64 0.08 2 Blunt taper 0.72 0.10 3 Sharp 0.72 0.05 4 30° bevel 0.72 0.10 5 45° bevel 0.72 0.10 6 35º bevel 0.72 0.08 * with respect to needle cross-section ** with respect to needle axis Geometry  taper angle* (degrees) 16 14 14 14 14 16  bevel angle** (degrees)  face area (mm2)  0 0 0 30 45 35  0.01 0.05 0.007 0.04 0.07 0.01  Figure 3.2: Different shapes of the stainless steel electrospray emitters investigated: symmetrically tapered electrospray needle (1), blunt tapered tip (2), sharp tapered needle (3), 30° bevel tip made from 2 (4), 45° bevel tip made from 2(5), 35° bevel tip with a smaller surface area (6).  64  3.2.2. Instrumentation All experiments were performed using a P/ACE MDQ capillary electrophoresis system with a modified capillary cartridge (Beckman Coulter Inc., Fullerton CA) that enables external detection by ESI-MS. Time programs were created using Beckman Coulter’s 32 Karat software in order to carry out reproducible injections and infusions of the analyte solution. Detection was carried out using a Micromass Q-TOF-1 mass spectrometer (Waters, Milford MA) operating in TOF-MS mode. The potential and temperature of the MS inlet cone were set to 20 V and 100 °C, respectively. Data processing was performed using Masslynx software (version 4.0). The standard Micromass ESI interface was removed and replaced by the CE-ESI-MS interface developed in our laboratory134, described in the previous chapter. For pressure infusion ESI-MS experiments, the orthogonal port of the tee union, normally used to deliver the modifier solution, was closed with a PEEK plug. The relationship of flow rate as a function of applied pressure was calibrated by measuring the time required for an analyte plug to reach the MS detector over a range of applied pressures. A linear least squares fit of the resulting plot was then used to determine the pressure required for a desired flow rate. 3.2.3. Electric field calculations Calculations of the electric field strength with various tip geometries and potentials were carried out using COMSOL Multiphysics 3.4 software (ElectromagneticsElectrostatics module, COMSOL Inc., Los Angeles CA)135. Models were created using the measured dimensions of the emitter tips and of the mass spectrometer inlet.  65  3.3. RESULTS AND DISCUSSION 3.3.1. Continuous infusion ESI-MS evaluation conditions The sample solution containing four amino acids (20 µM each) was continuously infused while the flow rate was varied from 4 μL/min to 0.1 μL/min over 9 logarithmically spaced steps in 2-minute intervals, while holding the electrospray potential constant. The potential at the capillary inlet was matched to the ESI potential in order to ensure that no electroosmotic flow was present during the experiments. The ion intensity during the second minute of each flow-rate step was used to calculate the mean and the standard deviation of the MS signal. This experiment was repeated at 0.2 kV intervals over the working range of electrospray potentials for each needle. The resulting contour plots of the average signal and the signal-to-noise ratio (SNR) as a function of flow rate and electrospray potential are shown in Figure 3.3. Because all of the amino acids tested (proline, threonine, isoleucine and arginine) showed similar trends, only the signals from arginine are shown. The limited working flow rate ranges of geometries 1 (tapered) and 2 (blunt taper) can be clearly seen, as neither provides a stable electrospray (SNR > 10) at flow rates greater than 400 nL/min. The sharp tip (emitter 3) provided stable electrospray signals over a wider range of flow rates, and was the only emitter for which SNR did not decrease at flow rates < 0.15 µL/min. However, the quality of the spray was very sensitive to changes in flow rate or electrospray potential, which raises concerns of reproducibility in day-to-day operation, and the maximum signal intensity was less than for other tip geometries. The beveled needles appear to offer the widest regions of high signal and stable spray, with emitter 6 (35º bevel, small surface area) offering the best  66  Figure 3.3: Average signal (A) and signal to noise ratio (B) for a continuous infusion of arginine (20 µM in BGE) as a function of ESI potential and flow rate for various emitter geometries  67  performance, with SNR > 20 for flow rates from 100 to 1500 nL/min. The trend for these needles is also more predicable, with wide regions where signal remains high regardless of changes in flow rate or ESI potential and fewer regions where the spray behaviour changes abruptly. The results do somewhat justify the previously-mentioned common practice of evaluating ESI performance using only a fixed electrospray potential, since for several of the investigated tip geometries the best SNR across the range of flow rates can be achieved within a narrow range of ESI potentials. However, there are cases where the optimal potentials for high and low flow rates differ significantly. Due to the number of combinations of voltages and flow rates tested, the analysis of each tip geometry required an entire day. Therefore, the different tips were evaluated on different days, making it difficult to correct for possible inter-day drift of the mass spectrometer detector sensitivity. This is the rationale for including signal-to-noise ratio as the basis of comparison in Figure 3.3, as well as the raw signal. The results appear to support the use of the beveled needles when analytes are continuously introduced to the ESI source. However, the use of continuous infusion of analyte makes it impossible to differentiate between analyte and background signal at a particular m/z. 3.3.2. Peak-based ESI-MS evaluation In order to verify if the results obtained by continuous infusion are also applicable for transient analyte plugs, a peak-based evaluation was also carried out. A time program was created using the 32 Karat software in order to automatically inject and elute sample plugs at three different flow rates in an alternating fashion while carrying out ESI-MS detection. Sample plugs (10 s injection at 1.0 psi) were pushed through the capillary and  68  into the sprayer tip using hydrodynamically generated flow rates of 1.0, 0.40 and 0.15 μL/min (3.9 psi, 1.5 psi and 0.6 psi, respectively, in a 75 µm ID, 85 cm long capillary), with the optimal spray potential at each flow rate determined in advance. This cycle of three injections followed by hydrodynamic infusion at decreasing flow rates was repeated in triplicate, such that nine peaks were detected for each needle. The average and standard deviation of the peak heights and areas were calculated for each flow rate. Results obtained within a single day were normalized, in order to allow inter-day comparisons regardless of instrumental drift, and the experiment was repeated over six days.  Figure 3.4: Relative peak height as a function of flow rate for the six different tip geometries investigated. Intra-day normalization was performed based on the height of the 0.15 µL/min peak for emitter 6. Figure 3.4 shows a comparison of the normalized peak heights for each tip at 0.15, 0.40 and 1.0 μL/min. Consistent with the findings of the continuous infusion  69  experiments, all six tips showed the greatest sensitivity at the lowest flow rate examined. The three beveled tips also showed the greatest sensitivity over the entire flow rate range. There does not appear to be a statistically significant difference between the beveled tips at any of the flow rates investigated by this method. Additionally, the optimal electrospray potentials for the peak-based evaluation were in most cases lower than observed during the continuous infusion analysis. 3.3.3. Electric field simulation results In order to better understand the results of the investigations, the electric field around the various emitter tips were calculated using Comsol Multiphysics. Table 3-2 shows the maximum electric field (Emax) at the emitter tip, using the optimal ESI potentials from the peak-based evaluation. Table 3-2: Optimal electrospray potentials and calculated maximum electric field for the six emitters studied Geometry 1 2 3 4 5 6  Taper Blunt Sharp 30° bevel 45° bevel 35º bevel  Electrospray potential (kV) 3.2 3.8 3.0 3.6 3.4 3.2  Maximum electric field (x 107 V/m) 1.5 1.4 1.9 1.7 2.0 1.8  Because the electric field depends on both the applied potential and the sharpness of the conductive surface, the blunt tip (emitter 2) had the lowest electric field (Emax = 1.4 x 107 V/m), despite having the highest ESI potential (3.8 kV). This is ~ 30% less than emitter 5 (45° bevel, 3.4 kV), which had the highest Emax = 2.0 x 107 V/m. However, there is no correlation between the results in Table 3-2 and Emax, since the sharp tip (emitter 3), which has the second highest value of Emax, showed poor performance similar  70  to the blunt tip. It should be noted that the physical model used in the simulations is based on the sprayer tip geometry and does not take into account the thickness of the fluid layer, or the electrostatic and fluid dynamic properties of the Taylor cone or droplets. 3.4. CONCLUDING REMARKS The different sprayer tips used for this investigation demonstrate that optimum solution flow rates for ESI are related to both the size and specific geometry of the sprayer. While the size of a tapered sprayer tip can directly influence the optimum flow rate it can process, beveled sprayer tips can generally sustain stable ESI over a wider range of volumetric flow rates. The limited geometries available for this investigation prevent us from characterizing more precisely the dimensions that are most important in choosing a successful geometry for electrospray. There are also variables that contribute to ESI performance that are not within the scope of this investigation, particularly the properties of the solution, including chemical composition, viscosity, volatility and surface tension, and the geometry of the MS inlet. However, our observations support the hypothesis that beveled tips can effectively increase the working flow rate range of an electrospray emitter, while offering performance similar to or better than that of symmetrically tapered emitters even at low flow rates.  71  Section B: Application of the CE-MS interface to separations of biomolecules  72  4. Chapter 4: Application of the CE-ESI-MS interface to the separation and characterization of peptides and proteins  73  4.1. INTRODUCTION Although it is not as well established as LC-MS, CE-MS is a popular tool for protein and peptide characterization136-140. Compared to LC-MS, it offers several advantages, including a complementary separation mechanism, reduced sample consumption and the ability to operate under conditions that will retain physiological protein interactions139. In a side-by-side comparison of nanoLC-MS/MS (nanoESI ionization, 200 nL/min) with coaxial CE-ESI-MS/MS (“µESI” ionization, sheath liquid flow rate 1-3 µL/min) using the same mass spectrometer, Pelzing and Neusüß concluded that, while the sensitivity of nanoLC-MS was five times greater than that of CE-MS, the difference was much less than expected, based on the large dilution by the sheath-flow interface141. Considering that our own interface, described in Chapter 2, has shown a five-fold sensitivity improvement compared to a coaxial sheath flow interface (sheath liquid flow rate 1 µL/min) for the analysis of amino acids134, the potential exists that CEMS using our low-dilution hyphenation strategy could rival the performance of nanoLCMS. A key challenge in the application of CE-MS to proteins and peptides is these analytes’ particular tendency to interact both hydrophobically and electrostatically with the fused-silica capillary wall in ways that may broaden peaks and reduce sensitivity. In extreme cases, components may adsorb to the wall so strongly that a denaturant, such as urea, is required to rinse them out. The electrostatic interactions may be reduced either by employing a highly acidic BGE, in order to neutralize the charges on the bare fused silica surface, or by using a highly basic pH such that both analytes and wall are  74  negatively charged. However, there are a number of problems with these strategies. Many proteins are poorly soluble at extreme pH, so that they will not dissolve in the BGE, and even at high pH some proteins may have regions of positive charge that will be attracted to the wall104. These constraints on the pH of the background electrolyte can also compromise the separation efficiency by forcing the separation to occur under nonoptimal conditions137. Finally, this approach is not effective when the origin of the interactions is hydrophobic, rather than electrostatic. In order to prevent these interactions from interfering with the quality of the separation while maintaining the full versatility of separation pH, a large number of different capillary coatings have been developed, which modify charge and hydrophobicity characteristics of the capillary wall surface137, 142-144. For proteins and peptides, which can be either positively or negatively charged depending on the solution pH and the analyte pI, a neutral capillary surface has the best potential to eliminate all electrostatic interactions. Hydrophobic interactions can be avoided by using a surface coating that is also hydrophilic. As such, neutral, hydrophilic coatings, including those made from polyacrylamide145, poly(vinyl alcohol)146 and polyethylene glycol147 are the most popular choice for the analysis of proteins by capillary electrophoresis148. Precoated capillaries of this type are commercially available from a number of different manufacturers. When using neutral capillaries, the absence of charged groups on the wall surface means that no electroosmotic flow (EOF) is generated, regardless of the solution pH. Although this is not an issue when using optical detection, the absence of bulk flow makes these capillaries incompatible with many interfaces for ESI-MS, where a stable,  75  forward EOF is required to support the electrospray process149. It also poses a problem for traditional sheath-flow interfaces, where the >1 µL/min flow rate of sheath liquid will cause a major dilution of the near zero effluent flow rate exiting the separation capillary. A hydrodynamic bulk flow can be added to the CE separation by applying pressure to the inlet vial, but the parabolic flow profile generated reduces the separation efficiency and resolution. Positively charged capillary coatings, such as polyethyleneimine (PEI), are also a common option for peptide and protein analysis150. Used in conjunction with an acidic pH, such that proteins and peptides will have a net positive charge, coatings of this type prevent the majority of wall interactions and provide a strong EOF towards the anode. When used with reverse polarity, they enable fast separations where the positively charged analytes migrate towards the capillary inlet, but are carried towards the detector by the strong anodic EOF. This makes them especially popular for use with sheathless CE-MS interfaces, which often require a strong, forward EOF to maintain stable electrospray. The disadvantage of this approach is that the migration order is reversed, due to the use of reverse polarity. This chapter presents an investigation the characteristics of our junction-at-the-tip interface when used in conjunction with coated and uncoated capillaries for the characterization of proteins and peptides. 4.2. MATERIALS AND METHODS 4.2.1. Materials Formic acid, acetic acid, methanol, isopropanol and acetonitrile were purchased from Fisher Scientific (Nepean, ON). Ammonium acetate buffer (100 mM, pH 3.1), 76  standard protein (cytochrome C, lysozyme, ribonuclease A) and peptide (angiotensins I and II) mixtures were obtained from Beckman Coulter Inc. (Brea, CA). Peptide samples resulting from the tryptic digestion of bovine serum albumin (BSA) and stored on C18 solid phase extraction packing material in 10 µg aliquots were provided by Dr. Leonard Foster’s lab (Department of Biochemistry and Molecular Biology, University of British Columbia). Fused silica capillary (50 µm ID, 365 µm OD) was purchased from Polymicro Technologies (Phoenix, AZ). Fused silica capillaries treated with a polyacrylamide-based neutral, hydrophilic wall coating and with a positively charged polyethyleneimine (PEI) coating were obtained from Beckman Coulter Inc. Standard stainless steel tee-unions and PEEK fittings were purchased from Upchurch Scientific (Oak Harbor WA). 4.2.2. Instrumentation Separations were carried out on a P/ACE MDQ or PA800 capillary electrophoresis instrument from Beckman Coulter Inc. (Brea, CA) with a modified capillary cartridge. Interfacing with mass spectrometry was performed using the interface described in Chapter 2 with the 35° beveled needle described in chapter 3 (emitter 6). The potential at the interface was controlled either by an external power supply (Spellman HV, Hauppauge NY) or through the electrospray power supply of the MS instrument. The needle tip was positioned 0.5 to 1 cm from the front of the MS inlet cone or heated capillary inlet, and fine control of the interface position was achieved through the MS instrument’s integrated positioning apparatus. A Xevo TQ mass spectrometer (Waters Corporation, Milford MA) was used for the detection of angiotensins and intact proteins. Analysis was performed in MS1 scan  77  mode. Detection of the BSA digest peptides was performed using a Micromass Q-TOF-1 mass spectrometer (Water Corporation, Milford MA) operating in TOF-MS mode. Peptide identities were not confirmed by MS/MS. For both instruments, detection conditions were optimized prior to analysis by continuously infusing the analytes under CE conditions. 4.2.3. CE-MS of intact proteins Separations were carried out in a neutral coated capillary (50 μm ID, 365 μm OD, 67 cm long) with a BGE of (100 mM ammonium acetate, pH 3.1)/methanol, 9:1 (v/v) using a potential of +30 kV at the capillary inlet and no assisting pressure. The standard protein mixture was dissolved in a solution of 9:1 BGE/water prior to analysis and injected for 20 s at a pressure of 1 psi. A modifier solution of 0.2% formic acid, 50% methanol was delivered through a 90 cm, 75 µm ID capillary using a pressure of 1 psi. Due to the absence of bulk flow in the separation capillary, the solution composition and flow rate for the ESI process is determined by the modifier solution. 4.2.4. CE-MS of angiotensin I and II Separations were performed in a PEI-coated capillary (50 μm ID, 365 μm OD, 85 cm long) with a BGE of 1% formic acid, 25% methanol. Under reverse polarity (-30 kV at the inlet), the positively charged capillary surface generated a strong EOF towards the capillary outlet. The standard peptide mixture was dissolved in 9:1 water/BGE solution and injected for 5 s with 5 psi pressure at the inlet. A modifier solution of 0.5% formic acid, 75% isopropanol was delivered at a pressure of 0.5 psi and served to increase the organic content of the CE effluent prior to electrospray ionization.  78  4.2.5. CE-MS of bovine serum albumin digest peptides Prior to analysis, 10 µg aliquots of the BSA tryptic peptides were eluted from the C18 packing using 15 µL of a (0.5% acetic acid, 80% acetonitrile) solution, and further diluted with 15 µL of deionized water to yield a final concentration of approximately 5 µM peptides, assuming complete digestion and recovery. Separations were carried out using both unmodified bare fused silica (75 µm ID, 85 cm) and PEI-coated capillaries (50 µm ID, 70 cm). For the unmodified capillary, a potential of +30 kV was applied to the inlet with a BGE of (0.5% formic acid, 30% methanol) and modifier consisting of (0.1% formic acid, 75% methanol), provided to the needle tip using a pressure of 0.5 psi on the modifier vial. Separations using the PEI capillary were carried out in reverse polarity (30 kV) with a BGE of (0.5% formic acid, 50% methanol) and (0.1% formic acid 75% methanol) modifier delivered with a pressure of 1 psi. In both cases, the sample was injected at 1 psi for 20 s. 4.3. RESULTS AND DISCUSSIONS 4.3.1. Analysis of proteins Figure 4.1A shows the total ion electropherogram for the separation of cytochrome C, lysozyme and ribonuclease A in a neutral capillary under acidic conditions. A low flow of modifier solution (0.25 μL/min) provides a stable flow of solution to sustain the ESI-MS operation and maintains the electrical circuit, while the solution inside the CE capillary is static. The three proteins are well resolved and the mass spectrum for each peak shows two or three ions from the charge envelope of the individual protein within the scan range used for analysis.  79  To our knowledge, this is the first demonstration of CE-ESI-MS in the micro- or nano-electrospray regime using a neutral capillary with no assisting pressure. The extracted ion electropherograms are shown in Figure 4.1B. The absence of electroosmotic flow in the separation capillary may result in some of the observed peak tailing, as there is no bulk flow pushing the analytes from the capillary terminus to the needle orifice, except for the modifier solution. However, the varying degree of tailing between the different peaks indicates that the micro-vial is not the primary source of broadening and that other factors, such as residual interactions with the capillary wall or BGE components, are contributing to the observed asymmetry.  Figure 4.1: Total ion electropherogram (A), extracted ion electropherograms (B) and mass spectra (C) of proteins separated under positive polarity with a neutral coated capillary. CE capillary, 50 μm ID, 67 cm; BGE, 100 mM NH4Ac, pH 3.1 and MeOH (9:1); Modifier, 0.2% formic acid, 50% methanol, 0.3 µL/min. 80  The limits of detection (LOD) for both the total ion and extracted ion electropherograms are listed in Table 4-1. The LOD was estimated by extrapolating the linear least squares fit of signal-to-noise ratio (SNR, where the signal is the peak height and noise is the standard deviation of the electropherogram background) as a function of protein concentration to determine the concentration which gives SNR = 3. The total ion electropherogram was used rather than the base peak ion electropherogram due to a limitation of the data processing software that resulted in poor quality base peak data in the high-mass range. Table 4-1: Limits of detection for protein analysis LOD (µM) Total ion electropherogram Extracted ion electropherogram cytochrome C 0.08 0.03 lysozyme 0.1 0.02 ribonuclease A 0.4 0.09 LOD based on duplicate analysis over the concentration range 10 – 50 µg/mL Protein  4.3.2. Analysis of angiotensins I and II Whereas in the CE-MS of proteins the modifier solution serves to provide a minimum flow rate to support the ESI process, for CE-MS of angiotensins the role of the modifier is to improve the compatibility of the CE effluent with electrospray ionization by increasing the proportion of organic solvent. Figure 4.2 shows the base peak electropherogram of angiotensin I and II separated in a PEI coated capillary under reverse polarity. In this case, the peptides have electrophoretic mobilities that are towards the inlet, however they are carried to the outlet by the strong forward EOF generated by the positively coated capillary wall. The BGE contains only 25% methanol in order to maintain the speed and efficiency of the separation. The surface tension of the CE  81  effluent is reduced by the addition of modifier containing 75% isopropanol, in order to improve ESI-MS detection.  Figure 4.2: Base peak ion electropherogram of Angiotensin I and II separated under reverse polarity with a polyethyleneimine coated capillary. BGE: 1% formic acid, 25% methanol; Modifier, 0.1% formic acid, 75% isopropanol, 0.5 psi; Sample: 1 μM Angiotensin I and Angiotensin II Both peptides are detected as sharp, symmetrical peaks, indicating that the presence of the micro-vial between the capillary terminus and the needle orifice does not lead to increased band broadening. The limits of detection estimated from the signal-tonoise ratios in both the base peak and extracted ion electropherograms are listed in Table 4-2. Table 4-2: Limits of detection for angiotensin I and II LOD (µM) base peak extracted ion Angiotensin I 433 0.06 0.04 Angiotensin II 524 0.03 0.004 LOD based on duplicate analysis over the concentration range 0.05 to 1 µg/mL Peptide  m/z  82  4.3.3. Analysis of bovine serum albumin tryptic peptides The peptides resulting from a tryptic digest of BSA were separated using both bare fused silica and PEI-coated capillaries. In the first case, using normal polarity, both the electroosmotic flow and the electrophoretic migration of the peptides is towards the capillary outlet. In the case of the PEI-coated capillary, the combination of the positive capillary wall coating, acidic BGE and reverse polarity generate a strong EOF towards the outlet, while the positively-charged peptides migrate towards the inlet. However the magnitude of their electrophoretic velocities is less than that of the EOF, so that they are detected at the outlet in a much shorter time than in the case of the uncoated capillary. Due to the complexity of the sample, not all peptides were resolved. In order to give a clear visual comparison of the two capillaries, seventeen peptides with signals that could be observed in the base peak electropherogram were selected to carry out the comparison, shown in Figure 4.3. The selected peptides are listed in Table 4-3. Despite the use of acidic conditions, two of the seventeen selected peptides (#3,4) that were observed as sharp peaks using the PEI-coated capillary were observed as very slow, broad peaks long after the main group of peptides at ~34 min (not shown) when the bare fused silica capillary was used. This indicates that interactions between the positively charged peptides and the capillary wall have not been completely eliminated through the use of the acidic buffer. In addition to eliminating wall interactions, the PEIcoated capillary provided slightly better resolution, and the time required to complete the separations was approximately half that of the uncoated capillary. Although part of this difference can be attributed to the difference in capillary lengths (85 cm for the uncoated capillary, 70 cm for the PEI capillary), it is primarily due to the much faster EOF  83  generated at the positively-charged PEI coated wall surface. Because the separation was carried out in reverse polarity, the migration order of the peptides in the PEI capillary is the reverse of that observed for the uncoated capillary.  Figure 4.3: Summed ion electropherograms for 17 peptides from a BSA digest separated in an uncoated capillary under normal polarity (TOP) and in a PEI coated capillary under reverse polarity (BOTTOM). Peptide sequences are listed in Table 4.3. Of the 84 expected peptides from the tryptic digestion, 30 were observed as well as a number of peptides corresponding to incomplete digestion of BSA, as indicated in Table 4-4. This corresponds to an amino acid sequence coverage of 46% from a sample injection of only 0.3 picomoles, assuming complete digestion and recovery of the sample.  84  Most peptides were observed in the doubly charged state, however +1 and +3 ions were also present. Several peptides were observed with more than one charge state. Table 4-3: Selected BSA peptides # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17  m/z 269.1 395.2 404.2 432.2 461.7 464.2 480.8 487.7 501.8 507.8 518.9 582.3 653.4 740.4 756.4 785.1 978.8  Peptide sequence Position in amino acid sequence + [FWGK]+2H 157 - 170 [LVTDLTK]+2H+ 257 – 263 [SLGK]+H+ 452 – 455 [VGTR]+H+ 456 – 459 + [AEFVEVTK]+2H 249 – 257 [YLYEIAR]+2H+ 161 – 167 [EKVLASSAR]+2H+ 210 – 218 + [DLGEEHFK]+2H 37 – 44 [LVVSTQTALA]+2H+ 598 – 607 (C-terminal) + [QTALVELLK]+2H 549 – 557 [DDPHACYSTVFDK]+3H+ 387 – 399 + [LVNELTEFAK]+2H 66 – 75 [HLVDEPQNLIK]+2H+ 402 – 412 [LGEYGFQNALIVR]+2H+ 421 – 433 [VPQVSTPTLVEVSR]+2H+ 438 – 451 [DAFLGSFLYEYSR]+2H+ 347 – 359 + [DAIPENLPPLTADFAEDK]+2H 319 - 336  Table 4-4: Amino acid sequence of bovine serum albumin MKWVTFISLLLLFSSAYSRGVFRRDTHKK.SEIAHR.FFKDLGEEHFK GLVLIAFSQYLQQCPFDEHVKLVNELTEFAKTCVADESHAGCEK SLHTLFGDELCKVASLRETYGDMADCCEKQEPERNECFLSHKDDSPDLPK LKPDPNTLCDEFKADEKKFWGKYLYEIARRHPYFYAPELLYYANK YNGVFQECCQAEDKGACLLPKIETMREKVLASSARQRLRCASIQKFGERALK AWSVARLSQKFPKAEFVEVTKLVTDLTKVHKECCHGDLLECADDRADLAK YICDNQDTISSKLKECCDKPLLEKSHCIAEVEKDAIPENLPPLTADFAEDK DVCKNYQEAKDAFLGSFLYEYSRRHPEYAVSVLLRLAKEYEATLEECCAK DDPHACYSTVFDKLKHLVDEPQNLIKQNCDQFEKLGEYGFQNALIVRYTR KVPQVSTPTLVEVSRSLGKVGTRCCTKPESERMPCTEDYLSLILNRLCVLHEK TPVSEKVTKCCTESLVNRRPCFSALTPDETYVPKAFDEK LFTFHADICTLPDTEKQIKKQTALVELLKHKPKATEEQLKTVMENFVAFVDK CCAADDKEACFAVEGPKLVVSTQTALA  Red: expected tryptic peptide observed; Blue: peptide observed with one missed tryptic cleavage; Bold: peptide electropherogram shown in Figure 4.3  85  Although the separation resolution could be improved by further optimizing the composition of the BGE, the current separation using the PEI coated capillary provides peak shape and peak widths that are well suited for tandem mass spectrometry, and the very reproducible mobility values obtained from CE can assist in identifying changes in peptide characteristics, such as posttranslational modifications. 4.4. CONCLUDING REMARKS This chapter demonstrates the compatibility of our interface with a variety of commonly used capillary coatings, including neutral, positively-charged and hydrophilic. Of particular interest is the compatibility with neutral capillaries, where no bulk flow is generated in the capillary, which has been successfully used for the analysis of intact proteins by CE-MS. In this case the modifier solution serves to support the ESI while the analytes exit the separation capillary in the normal migration order. We are continuing to work on developing a better understanding of the flow dynamics in the emitter tip in order to optimize the transport of analytes from the capillary exit to the ionization site under low and zero bulk flow separation conditions. The interface is also useful for CE-MS analysis of peptides. Separations can be performed using either bare fused silica or PEI-modified positive capillaries, although the PEI capillaries allowed faster separations and were more effective at preventing wall interactions than the use of low pH with the unmodified capillary.  86  5. Chapter 5: Capillary electrophoresis – mass spectrometry for carbohydrate analysis  87  5.1. INTRODUCTION The growing interest in carbohydrate characterization and the development of glycomics in recent years has been driven by an increased understanding of the many different roles that carbohydrates play in biology. Carbohydrates attached to proteins, known as glycans, are important determinants in cell signaling, gene expression, immune response, protein function and protein folding151. Monitoring changes in glycosylation between healthy and diseased states, as well as establishing an understanding of how glycan structure is related to function, will be an essential complement to systems biology aimed at understanding disease and treatments152. Protein therapeutics make up an important and growing class of drugs. Unlike small-molecule drugs, recombinant proteins offer the ability to perform highly complex functions with greater specificity, and are often tolerated by the body better than synthetic drugs153. Many protein therapeutics, including erythropoietin (EPO), follicle stimulating hormone (FSH) and many antibodies (immunoglobilins), are glycoproteins, and glycosylation often plays a key role in determining the drug’s viability, potency, and potential side effects154. As the number of protein-based therapeutics continues to grow, so does the need for effective methods for assessing the quality and purity of recombinant proteins, including their glycosylation patterns. The first step in glycoprotein analysis is normally to remove N-linked glycans from the protein backbone. This can be achieved using a number of different enzymes, the most common of which is peptide N-glycosidase F (PNGase F), which cleaves the link between N-acetylglucosamine and asparagine. Cleaved glycans are often derivatized  88  with a fluorescent labeling dye in order to facilitate detection, as most underivatized glycans are not detectable by UV/visible absorption or laser-induced fluorescence (LIF). The fluorophore 8-aminopyrene 1,3,6-trisulfonate (APTS) is one of the most commonly used dyes for carbohydrate analysis and is available in a number of commercial carbohydrate analysis kits155, 156. The labeling reaction is shown in Figure 5.1.  Figure 5.1: Carbohydrate labeling with the fluorophore 8-aminopyrene 1,3,6trisulfonate (APTS). Reprinted with permission from Analytical Biochemistry157. Although it is desirable to use MS detection for quality control in pharmaceutical analysis, the amount of time and money that must be invested in order to get approval for new separation method development can make this inconvenient. Most established protocols for CE-LIF analysis of glycans involve the use of a gel-buffer to increase the resolution of large carbohydrates158. However, because these gel buffers are generally not compatible with mass spectrometry, CE-MS of glycans is normally performed in free solution159-162. Additionally, while bare fused silica capillaries are often used for CE-LIF, many CE-MS arrangements have the added requirement of using coated capillaries in order to provide a strong, forward EOF to maintain a stable electrospray signal and CE current162.  89  If the major characteristics of an established separation protocol (capillary surface, electrolyte, polarity) can be retained when hyphenating to MS, the approval process would be much simpler, saving time and other resources. A popular method for CE analysis of APTS-labeled glycans involves separation under reversed polarity in a bare fused silica capillary with an acidic (formic, acetic or ε-aminocaproic acid) electrolyte161. Although the low pH of the BGE means that the EOF is highly suppressed, the residual EOF is towards the capillary inlet. Therefore, in order to couple with ESI-MS, the interface must be capable of providing solution to back-fill the capillary from the terminal end while also providing a stable spray. This has been previously done with coaxial sheath-flow interfaces, however the high volumetric flow of sheath liquid leads to significant dilution of the analytes161. A sheathless, reverse polarity method for the analysis of unlabeled carbohydrates under similar conditions has also been proposed163, however the separations were extremely slow (>100 min) and showed poor reproducibility. This chapter presents a strategy for adapting standard reagents and protocols for free solution separations of APTS labeled carbohydrates with LIF detection to hyphenation with mass spectrometry detection, in order to facilitate identification of the carbohydrate species while minimizing sample dilution. 5.2. MATERIALS AND METHODS 5.2.1. Materials A carbohydrate analysis kit, including APTS labeling dye and glucose ladder (G20) standard, was obtained from Beckman Coulter Inc. (Brea, CA). Sodium cyanoborohydride (1M in tetrahydrofuran) was purchased from Sigma Aldrich (St. Louis  90  MO, USA). Formic acid, acetic acid, methanol, ethanol, acetone and isopropanol were purchased from Fischer Scientific (Nepean ON). Bare fused silica capillary (50 µm or 75 µm ID, 365 µm OD) was purchased from Polymicro Technologies (Phoenix AZ). 5.2.2. APTS labeling The G20 glycan standard was labeled with APTS according to the manufacturer’s instructions. A 20 µL aliquot of the APTS solution (100 mg/mL in 15% acetic acid) was added to 1.25 mg of dry glycans, followed by 20 µL of sodium cyanoborohydride (0.1M in THF). The resulting solution was mixed on a vortex mixer and incubated overnight at room temperature in the dark. Following incubation, the excess APTS reagent was removed by precipitating the labeled glycans with ice-cold acetone and centrifuging at 12000 rpm for 15 min. After removing the supernatant, the pellet was washed with additional cold acetone and reconstituted in 200 µL of distilled water. This sample solution was used directly for all CE-LIF and CE-MS experiments. 5.2.3. Capillary electrophoresis with laser-induced fluorescence detection All CE-LIF experiments were performed on a PA800 capillary electrophoresis instrument (Beckman Coulter, Brea CA) with a laser-induced fluorescence detection module. Excitation was carried out at 488 nm, and emission monitored at 520 nm, which provides optimal sensitivity for APTS labeled glycans. The LIF signal was monitored using both a standard capillary arrangement, where the capillary outlet is placed in a buffer vial (75 µm ID capillary, 50 cm total length, 40 cm length to detector), and with the arrangement enabling ESI-MS detection, where the capillary terminus is inside the interface (50 µm ID capillary, 85 cm total length, 20 cm to detector). In both cases,  91  separations were carried out in reverse-polarity mode using bare fused silica capillaries and a background electrolyte composed of 1.0% formic acid and 50% methanol. 5.2.4. Capillary electrophoresis – electrospray ionization – mass spectrometry MS detection was carried out using a Finnigan LCQ Duo ion trap mass spectrometer (Thermo Scientific, Waltham MA) operating in negative ion mode, with the standard electrospray ionization interface replaced by the interface described in Chapter 2. The modifier solution, which provides solution for electrospray and maintains the electrical connections with the CE circuit, was composed of 1:2:1 2% formic acid/isopropanol/methanol and infused at a flow rate of 0.3 to 0.4 µL/min. Optimization of the negative-ion mode ESI-MS parameters was done automatically using the LCQ Tune Plus software (Finnigan) while infusing a solution of APTS-labeled glucose ladder standard. 5.3. RESULTS AND DISCUSSION 5.3.1. Interfacing with reverse-EOF separations Initial experiments were carried out with on-capillary LIF detection in order to investigate the impact of replacing the CE outlet vial with the ESI-MS interface when the bulk flow in the capillary is towards the inlet. Separations performed with a traditional CE arrangement, where the capillary terminus is placed in the outlet vial, were compared with separations where the outlet vial was replaced by the CE-MS interface, positioned in front of the MS inlet. The peaks observed with the traditional CE arrangement (Figure 5.2A) are typical of the APTS-labeled glucose ladder standard sample, which contains a mixture of glucose oligomers with lengths ranging from one to more than twenty glucose units. The large 92  initial peaks are due to the free APTS-label, which is not completely removed by the acetone precipitation process. The subsequent peaks are for the labeled oligomers. In order to install the capillary in the ESI-MS interface, it was necessary to use a longer capillary and to reduce the length to the detector to only 20 cm. As a result, there are significant variations in migration times and peak resolution observed with the interface. Despite these differences, the peak pattern observed with the interface (Figure 5.2B) is similar to that observed with the traditional CE arrangement, although the resolution is much lower due to the reduced length between injection and the detector. In Figure 5.2C, with the electrospray potential turned on, the LIF signal is nearly identical to that in Figure 5.2B. Together these results indicate that the interface is successfully replacing the outlet vial by providing stable electrical contact via the modifier solution, and that the presence of the interface allows the separation to proceed unaffected by the hyphenation with electrospray ionization.  Figure 5.2: Capillary electrophoresis separations of APTS-labeled glucose ladder standard with on-capillary laser induced fluorescence detection using different capillary arrangements. 93  5.3.2. Optimization of BGE conditions for resolution of large glycans Having established that it is possible to separate glycans under reverse-EOF conditions despite the presence of the ESI-MS interface at the capillary outlet, the method was implemented with negative ion mode ESI-MS. The conditions used with the LIF detection (1.0% formic acid, 50% methanol) yielded the expected peak pattern, however, baseline resolution was only achieved for APTS-labeled glucose oligomers of five glucose units or fewer. The background electrolyte (BGE) conditions were therefore varied in order to find parameters that would give acceptable resolution for longer oligomer chain lengths, as shown in Figure 5.3. As there is no forward flow in the CE capillary, the BGE does not enter the electrospray ionization source and thus does not require the addition of any organic solvent. However, methanol was added in most cases for two reasons. First, when the BGE differs significantly from the modifier solution, an ionic boundary may form at the interface between the BGE and modifier solutions. When there is no forward flow in the capillary, this boundary will gradually enter the capillary from the terminal side and may lead to a deterioration of peak shapes and resolution. Therefore, methanol was added to the BGE to increase its similarity to the modifier solution. The second reason for adding methanol to the BGE is to reduce the CE current, which should be maintained at less than 15 μA with our interface arrangement. Acetic acid has a much lower conductivity than formic acid and did not require the addition of methanol. However, separations using acetic acid as the primary electrolyte provided low resolution for the entire range of oligomer lengths (Figure 5.3A and B). Using 0.5% formic acid as the electrolyte, it was observed that increasing the  94  methanol content from 10% to 30% increased the analysis time and increased the resolution (Figure 5.3C and D).  Figure 5.3: Base peak electropherograms for APTS-labeled glucose ladder standard using different background electrolyte compositions, with CE potentials indicated in brackets. MS range 600-2000 m/z. Decreasing the strength of the electric field also increased the resolution (Figure 5.3E). Although this is not expected based on the theory of CE, it is effective because it provides a better match between the peak widths and the slow acquisition speed of the  95  ion trap MS when using a wide scan range and a relatively long trap injection time (700 µs). Higher concentrations of formic acid also improved the resolution, but required higher proportions of methanol or reduced CE potential in order to limit the current, as seen by the high background and poor stability in Figure 5.3F. Of the conditions tested, the background electrolyte composed of 2% formic acid and 30% methanol provided the best separation, when used with a CE potential of -16 kV, as shown in Figure 5.3H. 5.3.3. CE-ESI-MS of APTS-labeled glucose ladder standard Figure 5.4 shows the base peak and extracted ion electropherograms for APTSlabeled glucose oligomers containing up to 24 glucose units from the glucose ladder standard. The separation is based on the size-to-charge ratio, where the size refers to the cross-sectional radius of the hydrated analyte. Because the oligomers are chain-like structures that increase in length, rather than cross-sectional area, as the number of glucose monomers increases, the increase in glucose chain length leads to smaller differences in electrophoretic mobility. As a result, the resolution for shorter chain lengths is much better than for longer chains. Using the optimized conditions, baseline resolution is achieved for oligomers containing up to 15 glucose units (G15). Applying a power function to fit the trend of migration time as a function of number of glucose units (nG) results in the relationship tm ∝ (nG)0.2. By extension, based on equation 1-8, we can interpret that the effective radius of the APTS-labeled glucose oligomers, which determines the magnitude of the drag force acting on the ion as it moves through the BGE, is also proportional to (nG)0.2.  96  Figure 5.4: Base peak and extracted ion electropherograms for APTS-labeled glucose ladder standard (G1 to G24) using optimized conditions (BGE 2.0% formic acid, 30% methanol; separation potential -16 kV).  97  The contour plot view in Figure 5.5 shows the two-dimensional separation of the glucose ladder standard based on migration time and mass-to-charge ratio. The range 600-2000 m/z was chosen in order to allow observation of all glucose oligomers while excluding the major part of the impurity peak due to the excess APTS labeling reagent (m/z 498). Despite this, a number of APTS-related impurities can be observed as a wide band from 14 and17 minutes. The APTS-labeled glycans appear in various charge states between 15.6 and 30.2 minutes. Because the glucose oligomers are neutral, the charge resides on the APTS label, which has three sulfonate groups and an amine linker that may be protonated. As seen in Figure 5.5, only singly-charged ions are observed for the G1G4 oligomers (15.6, 17.0, 17.9 and 18.9 minutes) because the multiply-charged ions for these analytes have a mass-to-charge ratio that falls below the observed range. The -2 charge states can be clearly seen as a trend moving to increasing migration times and increasing mass-to-charge ratios from G5 (19.8 min, m/z 634) to G18 (26.4 min, m/z 1688). A similar trend is observed for the -3 ions for G11 (23.5 min, m/z 747) to G33 (30.2 min, m/z 1935). Despite the presence of different charge states in the mass spectrum, the charge states in solution should be uniform for all analytes, as the sulfonate groups and the amine linker on the APTS label are the only sites with the potential to be ionized in an acidic solution. This is expectation is reinforced by the observed trend in migration times. In addition to the expected molecular ion trends, there are additional ions that appear at a higher m/z than the 2- charge states of the G6 to G9 oligomers. Although the masses of these ions do not correspond to any expected products, the pattern exhibits a  98  regular m/z change of 162 per glucose unit, which corresponds to a singly charged species. The most likely explanation for this additional peak pattern is that the observed ions are products of in-source fragmentation of the APTS-labeled G6 to G9 oligomers. There is also evidence of adduct formation, as seen in the pattern of one to three additional peaks at higher m/z for many of the molecular ion peaks. The difference in m/z is 27 for the -2 charge state, or 18 for the -3 charge state, indicating that the adductforming species has a mass of 54 Da. The identity of this adduct is currently unknown, but, based on the fact that up to three adducts form for each molecular ion, it is likely that the adduct molecule is associated with the sulfonic groups on the APTS label.  99  Figure 5.5: Contour plot of migration time and mass-to-charge ratio (m/z) for separated APTS-labeled glucose ladder standard showing different charge states. Conditions as in Figure 5.4.  100  5.4. CONCLUDING REMARKS A standard method for analyzing APTS-labeled glycans has been successfully adapted, using the interface described in Chapter 2, to allow online ESI-MS detection of the separated glycans. The separation is able to proceed in reverse polarity, despite the fact that the bulk flow in the separation capillary is away from the electrospray ionization interface. Stable negative-ion mode ESI is facilitated through the addition at low flow rates of a modifier solution containing a large proportion of organic solvent. The current method provides baseline resolution of oligomers containing up to 15 glucose units. Improved resolution may be possible through further optimization of the background electrolyte. However, CE separations carried out in free solution are fundamentally limited for analytes, such as glycans, where changes in length have only a small effect on the effective cross-sectional radius. Improved resolution may be attainable by using an aqueous gel buffer as the separation medium, as is common when using LIF detection. Although there are no existing examples of capillary gel electrophoresis hyphenated to ESI-MS, the present interface should make this possible, so long as the gel buffer ions are ESI-MS compatible and the gel is prevented from entering the electrospray, either by the use of reverse polarity or an EOF suppressing capillary coating. Future experiments will focus on establishing the feasibility of this combination, which could significantly impact the routine procedures used for glycan analysis.  101  6. Chapter 6: Two dimensional protein mapping as a function of isoelectric point and molecular mass using capillary isoelectric focusing – electrospray ionization – mass spectrometry  102  6.1. INTRODUCTION Two-dimensional gel electrophoresis (2DE) has been an essential tool for developing our understanding of the complexity of proteins contained in cells and biological fluids, including serum and plasma4. Unfortunately, it is a labour-intensive and time-consuming technique that cannot be fully automated. Despite these disadvantages, gel electrophoresis is still popular because of its high peak capacity and because both dimensions provide information on intrinsic protein characteristics (pI and mass) that can facilitate identification. As such, there has been great interest in the development of an automated, free-solution characterization system capable of providing analogous information to 2DE. The most obvious way to achieve this is to replace the first-dimension slab gel isoelectric focusing with capillary isoelectric focusing (cIEF), and the second dimension SDS-PAGE with mass spectrometry. Both of these systems are well established and capable of providing better resolution than their slab gel-based counterparts. However, the difficulty lies in successfully interfacing the two techniques. The two most practical options for doing so are electrospray ionization, for continuous online coupling, or MALDI-MS, which requires the collection of fractions as spots for later, off-line analysis. Capillary isoelectric focusing (cIEF) was first demonstrated by Hjertèn and coworkers in 1985164. The technique is analogous to gel isoelectric focusing, however the separation and detection are performed in free solution in a narrow-bore fused silica capillary, with on-capillary UV detection. Although this facilitates automation and  103  quantitation165, UV detection is not as information-rich as the traditional coupling of gel IEF with SDS-PAGE, which provides an estimate of protein mass. However, cIEF provides excellent resolution and is easily automated, since the equipment involved is standard to commercial CE instruments. Because cIEF is performed in a free solution aqueous gel buffer, the pH gradient must be formed dynamically, rather than being immobilized in a gel matrix. This is done by filling the capillary with a solution containing the sample, a mixture of carrier ampholytes and a soluble linear polymer. An acidic solution (the anolyte) is placed at the anodic end of the capillary, while a basic solution (the catholyte) is placed at the cathodic end. Upon the application of the external electric field, hydronium and hydroxide ions migrate into the capillary from the anodic and cathodic ends, respectively, titrating the carrier ampholytes from both sides and generating a pH gradient. Each species of carrier ampholyte is focused to the point in the pH gradient where the solution pH matches the ampholyte pI. The ampholytes also act as buffering agents, stabilizing the pH gradient. Ampholytic or zwitterionic analytes, such as proteins and peptides, will also be focused to the point where the local solution pH matches the analyte pI. Traditional cIEF uses strong inorganic acids and bases as the anolyte and catholyte, respectively, which are not compatibly with MS detection. However, these can easily be replaced by volatile organic acids and bases. More troubling is the presence in the capillary of a MS-incompatible anti-convective gel buffer, protein solubilizing agents and a high concentration (1-5%) of carrier ampholytes. Finally, there must be some means of replacing the outlet vial in order to provide a catholyte solution to the capillary  104  terminus for the focusing step and later, in the case of chemical mobilization, an acidic solution to mobilize the pH gradient past the detector. Despite these challenges, there have been several demonstrations of cIEF-MS using both MALDI and electrospray ionization137, 166. Early ESI interfacing attempts involved semi-online approaches, where the capillary terminus was placed in a vial of catholyte during the focusing stage, then installed into an ESI source prior to mobilization167-169. The next generation of strategies using coaxial sheath-flow interfaces required retracting the capillary into the sheath liquid tube during the focusing step, so that the sheath liquid could act as the catholyte solution170-172. Finally, more recent cIEFMS systems use a sheath liquid interface to deliver the catholyte for the focusing step in a manner that does not require repositioning the capillary173. Recently, Mokkadem and co-workers introduced a coaxial sheath-flow cIEF-MS system that uses a solution of 30% glycerol to replace the aqueous polymer gels employed in high-resolution cIEF174, 175. In addition to acting as an anti-convective, solubilizing agent, glycerol also has the benefit of suppressing the electroosmotic flow and protein-wall interactions, which enables the use of bare fused silica capillaries rather than the neutral coated capillaries that are standard in cIEF with UV detection. Because glycerol has a boiling point of 290°C, it will evaporate in the low-pressure environment inside the mass spectrometer and not build-up over time in a way that compromises MS performance. On the other hand, the presence of glycerol in the electrospray solution increases the viscosity and decreases the overall volatility of the solution, leading to decreased ionization efficiency. Another innovation introduced by this group was the use  105  of a partial filling technique, where the catholyte solution fills a portion of the cIEF capillary, rather than being provided by the sheath liquid interface. Following focusing, proteins must be mobilized towards the detector using pressure, or by changing the chemical composition of the catholyte in order to induce an electroosmotic flow or electrophoretic migration of the focused protein zones. In all cases, the electric field is applied throughout the mobilization process in order to maintain the focusing effect. Pressure-induced flow is the most popular option for cIEFMS, because it can quickly mobilize the focused zones to ESI-MS. However, the resulting parabolic flow profile deteriorates the resolution achieved during the focusing stage176. The sample zone can also be mobilized to the detector by replacing the basic catholyte solution with an acidic chemical mobilization solution. By cutting off the supply of hydroxide ions entering the sample zone and replacing them with a weak base, the entire pH gradient is mobilized towards the cathode while maintaining its focusing effect. Because the mobilization is based on electrophoretic, rather than hydrodynamic, movement of the ampholytes and sample ions, there is no increase in band broadening. Chemical mobilization is routinely used in traditional cIEF because it provides the highest resolution of all mobilization techniques. Despite this fact, there is only one previous example of its use for cIEF-MS. However, in that case a gravity-induced flow was used to stabilize the ESI operation173. In addition, the data showed that protein peaks were not well focused and that there was a significant bleed of protein both ahead of and behind the focused zone.  106  Electroosmotic flow has been used to carry out a step-wise (rather than continuous) mobilization of proteins for cIEF-MS177. Although the uniform (rather than parabolic) flow profile of the EOF does not contribute to band broadening, it is more difficult to control and carry out reproducibly than pressure or chemical mobilization because the average pH in the capillary varies as mobilization progresses178. A second complicating factor in managing the EOF in bare fused silica capillaries is the presence of the amphoteric ampholytes. Interactions of the ampholytes with the capillary wall can lead to suppression or even reversal of the EOF177. Therefore, the electroosmotic flow may also be affected by the changing concentration of ampholytes during the mobilization process. This chapter describes the early progress of our interface, described in Chapter 2, for coupling cIEF with ESI-MS, using glycerol as an anti-convective medium. Several features of the interface make it well suited for the challenge of hyphenating cIEF with ESI-MS. The junction-at-the-tip arrangement ensures stable electrical contact during the focusing and mobilization steps, while the modifier solution can provide a flow of catholyte during the focusing step, or an acidic solution to accomplish chemical mobilization of the focused analyte zones. Our interface can also accommodate the coated capillaries that are commonly used in cIEF to eliminate electroosmotic flow and minimize wall interactions. 6.2. MATERIALS AND METHODS 6.2.1. Materials Formic acid, acetic acid, methanol and ammonium hydroxide were purchased from Fisher Scientific (Nepean ON). Glycerol, glutamic acid, urea and lysozyme (from 107  chicken egg white) were purchased from Sigma Aldrich. Protein standards (ribonuclease A, pI 9.45, Mr 13676 Da; myoglobin, two isoforms pI 7.35 and 6.85, Mr 16946; carbonic anhydrase II, pI 5.9, Mr 29808 Da; ß-lactoglobulin, pI 5.1, Mr 18357), one peptide standard (CCK flanking peptide, pI 3.6) and carrier ampholyte mixtures for the pH range 3 to 10 (Fluka® and Pharmalyte® brands) were obtained from Beckman Coulter Inc. (Brea CA). Bare fused silica capillaries (50 µm ID) were purchased from Polymicro Technologies (Phoenix AZ). Neutral, hydrophilic “N-CHO” type capillaries were obtained from Beckman Coulter, Inc. (Brea CA). The carrier ampholyte solutions were diluted in 30% glycerol to a concentration double that desired for analysis and stored at 4°C. Mixtures of the protein standards were also prepared in 30% glycerol and stored in aliquots at -20°C (average protein concentration 0.2 mg/mL). Sample solutions were prepared daily by mixing equal volumes of the protein and ampholyte solutions. 6.2.2. Capillary isoelectric focusing Capillary isoelectric focusing was performed on a PA800 or PA800 plus capillary electrophoresis system (Beckman Coulter, Brea CA) using a modified capillary cartridge to allow external MS detection. Because the length of the capillary required to reach the mass spectrometer inlet is longer than that required for resolution of the focused protein bands, the capillary was only partially filled with sample solution. The injection parameters were determined empirically by observing the time required to push a plug of analyte through a capillary filled with the 30% glycerol solution using a pressure of 2 psi. Injection times and pressures were then calculated based on the proportion of the capillary to be filled with the sample solution. Bare fused silica capillaries were  108  conditioned prior to sample injection with 6 M urea, 0.1 M sodium hydroxide, water, 2% formic acid and 30% glycerol for 3-5 min each using a pressure of 40 psi. The sodium hydroxide wash step was omitted for the coated capillaries. Figure 6.1 illustrates the different injection techniques used for bare fused silica and coated capillaries. Catholyte (when required), sample and anolyte solutions were injected with a pressure of 16 psi for a duration determined by the length of the desired plug.  Figure 6.1: Injection techniques employed in this chapter. (A) Flanking method for bare fused silica capillaries; (B) Modified method for covalently coated capillaries. Isoelectric focusing was carried out with the capillary inlet placed in a vial containing the anolyte solution, composed of 50 mM formic acid and 1 mM glutamic acid, and the capillary terminus installed in the interface. The catholyte, which was either contained in the capillary (Figure 6.1A), or delivered through the modifier capillary  109  (Figure 6.1B), was a solution of 0.1 M ammonium hydroxide. A potential of +30 kV was applied at the capillary inlet, while the interface was grounded. Following focusing, the protein zones were mobilized towards the interface by applying a pressure of 1.2 psi to the inlet vial, or by using the composition of the modifier solution to chemically mobilize the contents of the capillary. Additional details on the chemical mobilization procedure are provided in Sections 6.3.2 and 6.3.3. 6.2.3. Mass spectrometry Hyphenation with mass spectrometry was carried out using the interface developed in our laboratory, described in Chapter 2. The modifier solution was composed of 2% formic acid, 50% methanol and delivered at a flow rate of 0.5 µL/min, unless otherwise stated. Mass analysis was carried out on either a Finnigan LCQ*Duo ion trap (Thermo Scientific, Waltham MA) or a Micromass Q-TOF 1 mass spectrometer (Waters, Milford MA) with the standard ESI interfaces removed. Detection parameters were optimized automatically for the Finnigan MS, or manually for the Q-TOF, while infusing a solution of lysozyme. Data acquisition was started at the beginning of the mobilization step, unless otherwise stated. 6.3. RESULTS AND DISCUSSION 6.3.1. cIEF-MS using the capillary as the catholyte vial The length of the capillary used for cIEF-MS is determined primarily by practical considerations, such as the length required to reach the MS inlet, and is much longer than what is required for effective resolution of the focused analyte bands. Therefore, the “sandwich” method, where the sample zone is flanked by zones of anolyte and catholyte  110  at the anodic and cathodic ends of the capillary, respectively, was used for sample injection (see Figure 6.1). The catholyte plug, which is located at the terminal end of the capillary, must contain enough base to establish the pH gradient by providing a flux of hydroxyl to the sample zone, while maintaining a high pH region that forms the boundary of the pH gradient. However, its length is also the distance that separates the sample zone from the detector, so that increasing the length of the catholyte zone may increase the time required for mobilization. Figure 6.2 shows the effect of varying the length of the catholyte plug from one quarter (A) to one eighth (B) of the capillary length, while the length of the sample zone was maintained at one half of the capillary length. Neither the timing, nor the spacing of the protein peaks is significantly affected, indicating that changes to the length of the catholyte zone within this range do not have a significant impact on the separation step. It is not clear if this is due to expansion of the sample zone, or due to the increased EOF that is present when a larger proportion of the capillary is filled with base, which would gradually move the sample zone towards the capillary terminus. The effect of the injected sample zone length was also investigated by filling one half (Figure 6.2A, B) to one quarter (C) of the capillary length with the mixture of proteins and carrier ampholytes. In this case, many important changes can be observed. The ribonuclease A (RNase A) peak is fully focused with the shorter sample zone, whereas with the longer sample zones (A and B) it appears as a pair of split anodic and cathodic peaks176. This is because analytes have a shorter distance to travel in the short sample zone and can therefore reach their equilibrium (focused) position more quickly. Reducing the length of the sample zone also allows all of the focused protein zones to  111  arrive at the detector more quickly, which reduces the time required for analysis. However, this comes at the cost of decreased resolution, since the protein zones are focused within a shorter region of the capillary, reducing the peak capacity. The sensitivity is also decreased, due to the smaller volume of sample loaded onto the capillary.  Figure 6.2: Effect of sample injection parameters on peak shapes and analysis time. Injection using the flanking method with an anolyte/sample/catholyte length ratio of 1:2:1 (A), 3:4:1 (B) and 2:1:1 (C). Conditions: Sample, protein mixture with 0.3% (w/v) Fluka® ampholytes; bare fused silica capillary, 50 µm ID, 68 cm; focusing, +30 kV for 20 min; pressure mobilization. In traditional cIEF arrangements, high-resolution separations can be carried out with a sample zone as short as a few centimeters. However, the resolution must be maintained during the mobilization step in order for that resolution to be realized during detection. Therefore, it is possible that the pressure mobilization is responsible for some  112  of the observed overlap. Alternatively, the low concentration of carrier ampholytes used (0.3% w/v) may not be sufficient to generate and maintain a stable pH gradient during the focusing step. Similar experiments using Pharmalyte® carrier ampholytes were unsuccessful, requiring much longer pressure mobilization times and yielding extremely broad protein peaks. Upon consulting the literature, it was discovered that this effect has been previously observed by other researchers, and is due to the Pharmalyte® ampholytes forming a dynamic coating layer on the capillary wall surface, which generates a reversed EOF in the capillary and pushes the sample zone back towards the capillary inlet177. Fluka® ampholytes were therefore used for all further comparisons. 6.3.2. Chemical mobilization Although pressure mobilization is a convenient means of removing the focused protein zones from the capillary, the parabolic flow profile that is generated in the capillary leads to reduced resolution. Chemical mobilization is the best choice for maintaining the highest possible resolution, but it requires replacing the catholyte solution with an acidic mobilization solution. A unique advantage of our interface is that the composition of the chemical modifier can be adjusted in order to serve this purpose. In this case, the modifier already contains acetic acid, which will function as the chemical mobilizer. In order to do so, it will have to neutralize the catholyte plug that lies between the sample plug and the capillary terminus before it can effect mobilization of the pH gradient. Therefore, the catholyte plug volume should be minimized, in order to reduce the time required for mobilization.  113  Injection was again performed using the sandwich method (Figure 6.1A) with a 3:4:1 anolyte/sample/catholyte zone length ratio. However, following the focusing step, no pressure was applied to the inlet vial, while the modifier solution (2% acetic acid, 50% methanol) was provided at a flow rate of 0.5 µL/min. Figure 6.3A shows the resulting total ion electropherogram. The acidic modifier solution was able to penetrate the catholyte and sample zones and initiate chemical mobilization, beginning with the basic end of the pH gradient at ~ 35 minutes. Although this process is slow compared to pressure mobilization, the resolution is drastically improved. Three distinct peaks, corresponding to different protein isoforms, are visible for myoglobin. However, carbonic anhydrase II and ß-lactoglobulin still appear as a single peak in the total ion electropherogram, as a result of a partial overlap between the protein zones. The small peaks (0-10% relative abundance) that appear intermittently are due to the presence of focused zones of carrier ampholytes with a mass-to-charge ratio greater than 1000. The extracted mass spectra for the focused protein zones in Figure 6.3B show the charge distributions for the different protein peaks. Although carbonic anhydrase II and ß-lactoglobulin are unresolved in time, the partial separation allows their mass spectra to be examined separately. The background signal due to the presence of carrier ampholytes is quite low, due to the fact that most have a mass of less than 1000 Da179. In addition to providing improved resolution, chemical mobilization provides a much more stable electrospray signal compared to pressure mobilization. Whereas in pressure mobilization the composition of the solution exiting the cIEF capillary is equivalent to the solution inside the capillary, the electrophoretic mechanism of chemical mobilization will affect ampholytic species such as proteins and carrier ampholytes, but  114  not neutral species, including glycerol. This significantly decreases the amount of glycerol that is introduced into the Taylor cone, which lowers the viscosity and increases the volatility of the electrosprayed solution.  Figure 6.3: Total ion electropherogram (A) and extracted mass spectra (B) for a protein mixture separated by cIEF with chemical mobilization. Conditions: Protein mixture with 0.3% (w/v) Fluka® ampholytes filling ½ capillary length; bare fused silica capillary, 86 cm/50µm ID Focusing +30 kV for 8 min; Chemical/EOF mobilization by modifier solution, 2% acetic acid, 50% methanol at 0.5 µL/min.  115  6.3.3. Neutral coated capillaries When using bare fused silica capillaries for cIEF, protein interactions with the negatively charged wall may interfere with the separation. Although the presence of glycerol in the sample solution should reduce these interactions, some residual interaction may remain. As mentioned in Chapter 4, capillaries with a neutral, hydrophilic coating eliminate electrostatic and hydrophobic interactions of proteins with the wall and eliminate EOF, regardless of the average capillary pH or the presence of different ampholyte species. Unfortunately, the covalent wall coating cannot be exposed to solutions with pH > 10, such as the catholyte. Rather than providing the catholyte as a plug in the cIEF capillary, it is possible to deliver it via the modifier capillary, as illustrated in Figure 6.1B(i). Using this alternative injection method, the sample zone was extended to the capillary terminus, while the catholyte solution was flushed as the chemical modifier in order to fill the needle interior and modifier capillary with catholyte. At the onset of focusing, the modifier capillary was placed in a vial containing the acidic modifier solution. By applying a low pressure (1 to 1.2 psi) to the modifier vial, the modifier solution eventually replaces the catholyte solution in the needle interior. The duration of the focusing step is therefore determined by the flow rate in the modifier capillary and the combined volume of the capillary and needle interior. Chemical mobilization is initiated when the acidic modifier solution completely displaces the catholyte (Figure 6.1B(ii)). Figure 6.4 shows a separation performed in this manner, with detection using the Q-TOF mass spectrometer. Although the Q-TOF instrument has limited desolvation capabilities at its inlet compared to the ion trap, it is possible to use it in conjunction with  116  chemical mobilization, since the concentration of glycerol in the electrospray solution is very low. Due to changes in capillary type, ampholyte concentration (0.6% w/v) and mass spectrometer type, the relative peak intensities in Figure 6.4 differ from those observed on the ion trap mass spectrometer. However, the pattern is similar to that achieved in bare fused silica capillaries, with the exception that carbonic anhydrase II (pI 5.9) and ßlactoglobulin (pI 5.1) are resolved in the total ion electropherogram. A third isoform of myoglobin is also visible at 51.5 minutes, which was not previously observed.  Figure 6.4: Total ion electropherogram for a protein mixture separated by cIEF with chemical mobilization in a neutral coated capillary. Conditions: Protein mixture with 0.6% (w/v) Fluka® ampholytes filling ½ capillary length; neutral coated NCHO capillary, 76 cm/50µm ID; Inlet +30 kV; ESI +4kV after 30 min; Chemical mobilization, 2% formic acid, 50% methanol at 1 psi. 6.3.4. Correlation of migration times with isoelectric point In theory, the detection times of zones focused by cIEF should be linearly related to the isoelectric point of the protein zone. However, there are a number of factors that can lead to non-linear relationships. Figure 6.5 shows the migration times for the  117  separations in Figures 2B, 3 and 4, plotted as a function of isoelectric point. The calculated slope and correlation coefficient, based on the protein markers only, are listed in Table 6-1.  Figure 6.5: Migration time as a function of isoelectric point for the separations shown in Figure 2B (green triangles), Figure 3 (blue squares) and Figure 4 (red circles). Protein and peptide markers: ribonuclease A, pI 9.45; myoglobin (two isoforms) pI 7.35 and 6.85; carbonic anhydrase II, pI 5.9; ß-lactoglobulin, pI 5.1; CCK flanking peptide, pI 3.6 (blue trace only) Table 6-1: Correlation of migration time and isoelectric point based on protein markers, pI 5.1 - 9.45 Slope (min/pH)  R2  Pressure (1psi)  -1.6  0.99  Bare fused silica  Chemical (2% acetic acid, 50% methanol)  -4.7  0.94  Neutral, hydrophilic coating  Chemical (2% formic acid, 50% methanol)  -5.5  0.98  Capillary  Mobilization  Bare fused silica  118  As expected, cIEF with pressure mobilization leads to shorter detection times and a lower absolute value of the slope. Because mobilization is driven by a constant pressure at the capillary inlet, it also provides a higher correlation coefficient. The use of chemical mobilization in a bare fused silica capillary yielded a larger absolute value for the slope, and thus better separation in time of the focused zones, but provided the worst correlation coefficient (0.94). This may be due to variations in the electroosmotic flow as the pH gradient leaves the capillary, or related to interactions of the proteins with the bare fused silica capillary wall. When chemical mobilization is applied in a neutral coated capillary, the correlation coefficient improves (0.98). The slope is also slightly larger, which reflects the fact that a higher concentration of carrier ampholytes was used, resulting in a slower mobilization process despite the use of a stronger acid (formic vs. acetic) in the modifier solution. The final marker, CCK flanking peptide (pI 3.6) arrives at the detector much later than would be expected, based on extrapolation of the linear least-squares line for the corresponding series. This is due to a gradual drift of the acidic portion of the pH gradient towards the capillary inlet, which can be mediated by incorporating an ampholytic spacer species that has an isoelectric point that is just below the range of the pH gradient176. However, that is beyond the current scope of our method development. 6.4. CONCLUDING REMARKS Capillary isoelectric focusing can be coupled with electrospray ionization mass spectrometry using the interface described in chapter 2. The addition of 30% glycerol to the sample solution allows the use of bare fused silica capillaries by suppressing both wall interactions and electroosmotic flow. However, it appears that neutral coated  119  capillaries provide slightly improved performance, due to the complete elimination of these effects. Compared to commercial sheath flow interfaces, ours offers the advantages of reduced dilution of the protein zones during the electrospray process and stable contact between the capillary and the modifier solution, which allows mobilization to be achieved through the introduction of an acidic solution at the capillary terminus. This chemical mobilization, which has not previously been demonstrated for cIEF-MS without supplemental pressure mobilization, provides superior resolution and improves electrospray stability by reducing the concentration of glycerol exiting the cIEF capillary. Additional work will be required in order to improve the reproducibility, resolution and linearity of the pH gradient. However, the progress that has already been made represents and exciting step towards demonstrating the feasibility of cIEF-MS as a viable alternative to 2D gel electrophoresis and the unique features of our interface that make it well suited for this task.  120  Section C: Strategies for improving the reproducibility of reversed-phase liquid chromatography separations  121  7. Chapter 7: Development of a standard hydrophobicity index for improved protein characterization by reversed-phase liquid chromatography  122  7.1. INTRODUCTION Liquid chromatography remains the most commonly used separation technique for the separation of peptides and proteins prior to analysis by mass spectrometry. Reversed phase HPLC, where analytes are partitioned between a non-polar stationary phase and a more polar mobile phase, is the most popular separation mode when coupling with ESI-MS, because the mobile phase composition - normally water mixed with a miscible organic solvent such as methanol or acetonitrile - is highly compatible with ESIMS7. As such, it is used as the second dimension separation in many popular 2D LC-MS proteomic strategies1, 30, 31, 34, 35, 180. The retention times observed in isocratic reversed phase LC separations can be related to the hydrophobicity of the analytes. However, due to the diverse characteristics of proteins and peptides, isocratic elution does not generally provide satisfactory separation: mobile phase conditions that give good resolution for early-eluting peaks tend to give very poor efficiencies and long retention times for analytes that are strongly retained by the stationary phase. This is commonly referred to as the “general elution problem” in chromatography. The solution is to alter the mobile phase conditions over the course of the separation, so that early conditions favour retention of analytes in the stationary phase, then change over time to shift the equilibrium towards the mobile phase and speed up the elution of highly retained compounds. This can be done in a step-wise manner, where the conditions change in increments at regular intervals, or it can be done gradually over time in a linear gradient. While this succeeds in improving the resolution  123  and efficiency, and reducing the time required for separation, it also makes it more difficult to relate the observed elution times to molecular characteristics. Although including retention times as criteria for positive identification of a suspected peptide or protein has been demonstrated to increase the number of highconfidence identifications181, 182, many LC-MS strategies do not use the retention time data as a part of their characterization, due in part to difficulties in obtaining reproducible retention times. More commonly, reversed phase separations of proteins or peptides are used only as a means of simplifying mass spectra by reducing the number of components entering the MS at one time. While retention times for a given instrument may be reproducible, changes in the sample matrix and column conditions can lead to variations over time. These variations are even more pronounced when attempting to compare separations performed on different instruments. Many groups working with reversed phase HPLC instruments correct for retention time variations using dynamic programming methods in which chromatograms are stretched or compressed point-by-point to give alignment with a reference chromatogram183-185. Although this method is effective for two chromatograms from different experiments, it ignores the physical meaning behind the protein elution times and distorts the original chromatograms. The degree of hydrophobicity of a peptide or protein is an intrinsic property determined by its amino acid sequence186. If reversed phase retention times could be successfully related to the intrinsic hydrophobic character of the analyte, this could simultaneously increase the reproducibility of separations and increase the quality of characterization. This chapter describes an investigation of the potential of converting  124  retention times to a scale of relative hydrophobicity, based on the retention times of marker compounds, as a means for simultaneously increasing separation reproducibility and linking separation results to intrinsic hydrophobicity characteristics. 7.2. THEORY 7.2.1. Hydrophobicity The hydrophobicity constant (π) for a functional group, X, can be quantitatively determined by measuring ratio of the octanol-water partition coefficients for a molecule containing the functional group (PR-X) versus a molecule where that group is replaced by hydrogen (PR-H)187:  π X = log  PR − X PR − H  (7-1)  This approach can be expanded to proteins by calculating the average protein hydrophobicity, based on the polarity of the amino acids that make up the protein sequence186. However, it has been shown that these theoretical values do not correlate well with empirically determined measurements of intact proteins based on interactions with a hydrophobic stationary phase or on solvent partition coefficients188. 7.2.2. Theory of partition chromatography Separation in partition chromatography is based on the partitioning of analyte between a mobile and stationary phase. Different analytes will have different affinities for the mobile and stationary phases, leading to a distribution between phases that is dictated by the analyte equilibrium constant, K. AM ⇌ AS  (7-2)  125  K=  [AS ] [AM ]  (7-3)  where AM and AS are the analyte in the mobile and stationary phases, respectively. For isocratic elution, when neither the composition nor the physical properties of the mobile phase are altered during the course of the separation, this leads to an average analyte velocity (in the chromatographic column) equal to:  vA = U  1 V 1+ K A S VM  (7-4)  where U is the velocity of the mobile phase, and VS and VM are the volumes of the stationary and mobile phases, respectively. Because the volumes are consistent for all analytes in a particular column, it is useful to group together the second term in the denominator and define it as the capacity factor, k’, which describes analyte retention in a chromatographic column based on both the partition coefficient and the volumes of the two phases. k' = K  VS VM  (7-5)  7.2.3. Retention time correction strategies for a constant gradient  In the interest of utilizing all of the information contained in each separation process we are working on the development of an alternative strategy which uses two small marker molecules added to a reverse-phase separation mixture to act as hydrophobicity standards. One marker is chosen to be hydrophilic (eluting early in the gradient) while the other is hydrophobic (late eluting), such that most proteins can be expected to elute between the two markers. The marker retention times will be used to  126  transform the time axis of a chromatogram into a normalized hydrophobicity index (hp), according to the equation: hp =  t R − t m1 t m 2 − t m1  (7-6)  where tR is the retention time at any point, and tm1 and tm2 are the retention times of the hydrophilic and hydrophobic markers, respectively. Therefore, the hp values of the marker compounds will be 0 and 1, for the early and late eluting markers, respectively. While equation 7-6 represents the simplest method of creating a hydrophobicity index, it may be beneficial to consider the elution behaviour in terms of the elution velocity (υx), rather than retention times, in order to eliminate the effect of the column length. Since υx=(L/tR), hp in terms of analyte velocities is calculated by: L L − t t m1 t R (t − t ) hp = = m 2 × R m1 L L t R (t m 2 − t m1 ) − t m 2 t m1  (7-7)  It must also be recognized that the concept of mobility standards in liquid separations is not new. Methods using up to four markers as internal standards have been described to improve the reproducibility in both LC and CE separations189, 190. Among these, a two-marker method originally developed by Dovichi et al. for migration time correction in capillary electrophoresis is of particular interest190. One fast-migrating compound (m1) and one slow-migrating compound (m2) are used to correct the migration times of the other components with respect to a reference chromatogram, according to the equation: t converted  ⎡ 1 1⎛ 1 1 ⎞⎤ = ⎢ − ⎜⎜ − ⎟⎟⎥ ⎣ t m1 γ ⎝ tˆm1 tˆx ⎠⎦  −1  (7-8)  127  1 1 − tˆ tˆ γ = m1 m 2 1 1 − t m1 t m 2  (7-9)  where tm1 and tm2 are the marker migration times and tx is the analyte migration time. Variables with hats correspond to migration times from the electropherogram being corrected, while those without refer to migration times in a reference electropherogram. The γ portion of the equation corrects for errors affecting electrophoretic mobilities, which will be more severe for slower migrating species, while the rest of Equation 7-8 corrects for errors in the electroosmotic flow, which affect all species equally. This is relevant to gradient HPLC, which also has errors that increase with retention time and others that are common to all peaks. Common errors include variations in the injection process and in the establishment of the gradient, while errors due to variations in the delivery of the gradient or in the retentive properties of the column would be more pronounced for later eluting species. All of the above equations can be applied to the entire time axis of a chromatogram in order to give detector intensity as a function of either hp or tconverted. It is also possible to apply Equation 7-6 to tconverted values obtained using Equation 7-8, in order express the result in terms of hp values. 7.2.4. Theory of gradient elution  The relationship between the mobile phase composition at elution (φelution) and intrinsic protein properties such as the capacity factor (k’) has been described mathematically as:  128  ϕ elution = S=  [  ]  1 log 2.3Smϕ (t 0 k 0 − t D ) − 1 S  Δ(log k ') Δϕ  (7-10)  (7-11)  where k0 is the capacity factor of the solute in the initial mobile phase composition, t0 is the dead time of the column, tD is the gradient delay time, and mφ is the rate of change the mobile phase gradient in % change per minute191. The capacity factor, k’, depends on the mobile phase composition and therefore changes over the course of the gradient. Equations 7-10 and 7-11 indicate that there are actually two intrinsic properties that determine elution in a reversed-phase gradient separation: the capacity factor of the protein in water and the effect of changing mobile phase composition on the value of k’. 7.3. MATERIALS AND METHODS 7.3.1. Materials  Rat plasma samples were obtained through a collaboration with the UBC Brain Research Centre and stored at -80°C prior to use. A ProteomeLab™ IgY-R7 LC2 Proteome Partitioning Kit, including avian IgY affinity column and buffers was obtained from Beckman Coulter. Columns and buffers for the ProteomeLab™ PF 2D protein fractionation system were also obtained from Beckman Coulter. Acetonitrile and trifluoroacetic acid were purchased from Fisher scientific. 7.3.2. Pre-fractionation of protein samples  A ProteomeLab™ IgY-R7 LC2 Proteome Partitioning system from Beckman Coulter (Brea, CA) was used to selectively deplete the seven most abundant proteins from the rat plasma samples, using the standard manufacturer’s protocol. The unbound  129  (low-abundance) fraction was collected and concentrated by centrifugation in a filter vial with a molecular weight cut-off of 5 kDa (Amicon Ultra 15, Millipore, Billerica MA). Following the depletion of high-abundance proteins, plasma samples were separated based on their isoelectric point by high performance chromatofocusing using standard procedures35. Briefly, protein samples diluted with the pH 8.5 start buffer were injected (1.0 - 2.0 mL injection) onto the chromatofocusing column equilibrated with the start buffer. The mobile phase was then switched to the elution buffer (pH 4), which generates an on-column pH gradient that gradually elutes proteins at a pH equal to the protein pI. Fractions were automatically collected at regular pH intervals (0.2 pH units, or 1.0 mL volume, whichever occurs first) according to the measured pH of the column effluent. 7.3.3. Reversed phase separation of intact proteins  Reversed phase separations were performed using a Beckman Coulter HPLC system, controlled using Beckman’s 32 Karat software. The second dimension reversedphase column uses a non-porous silica C-18 stationary phase column. Mobile phase A was a solution of 0.1% TFA in deionized water. Mobile phase B was a solution of 0.08% TFA in HPLC-grade acetonitrile. Separations were performed using a linear gradient from 0 to 100% B over 30 minutes. The temperature of the column was maintained at 50°C by a column heater. Detection was performed by monitoring the absorption at 214 nm. Pooled fractions from several 1st dimension chromatofocusing separations that shared a common pH were injected (200 µL injections) by the automated fraction collection/injection module.  130  7.4. RESULTS AND DISCUSSION 7.4.1. Evaluation of two-marker correction strategies  A series of four replicate separations of a pooled mixture of rat plasma proteins that were poorly aligned in the original chromatograms were standardized using all three of the methods described in the theory section. The results of the alignment are shown in Figure 7.1. Compared to the original chromatograms (RSD = 0.4%), equations 7-6 and 7-7 gave a factor of two improvement in the precision of the peak maxima (RSD = 0.2%), while in the case of equation 7-8 there was a ten-fold improvement in precision (RSD = 0.04%).  Figure 7.1: Gradient revered phase separation of pooled proteins with no correction applied (A), correction according to equations 7-6 (B), 7-7 (C) and 7-8 (D).  A series of fifteen replicate blank separations, showing characteristic peaks due to impurities in the mobile phase components and/or in the wash solution for the injection needle were also standardized using all three correction methods. As in the previous  131  experiments, equation 7-8 gave the largest improvement (RSD = 0.08%, compared to 0.6% for the original chromatograms), while equations 7-6 and 7-7 had an average RSD of 0.2%. The original and best corrected chromatograms are shown in Figure 7.2.  Figure 7.2: Reversed phase separation of mobile phase impurities: Original chromatograms (top) and chromatograms with the equation 7-8 correction applied (bottom) 7.4.2. Standardization of hydrophobicity for any linear gradient  In gradient reverse-phase chromatography, relative retention times are not directly related to the hydrophobicity, but depend also on a number of factors including the choice of stationary phase, mobile phase and the timing gradient elution program. Often, chromatograms will be reported with the x-axis in units of percent acetonitrile in the mobile phase (%B), however this does not reduce variations in retention times, nor does it help to standardize gradients run at different speeds, as shown in Figure 7.3.  132  45 min 30 min  20  30  40  50  60  70  80  % acetonitrile in mobile  Figure 7.3: Separations of intact rat plasma proteins by reversed-phase HPLC with a 0-100% acetonitrile gradient over 45 minutes (red) and 30 minutes (blue)  Although the hydrophobicity index standardization is successful for aligning chromatograms run under the same gradient conditions, it fails when applied to separations with different rates of change for the mobile phase gradient. This is because retention times in gradient HPLC do not vary linearly with rate of gradient change, as described earlier in Equation 7-10. In order to examine the relationship between elution times and the two factors a series of experiments were carried out where proteins which had been previously separated by chromatofocusing in the first dimension were separated in the second dimension using a series of different gradient speeds from 1.67 %B/min to 5 %B/min. The φelution for thirteen peaks were measured in each separation and plotted as a function of the gradient speed, as shown in Figure 7.4. A non-linear regression was performed using Equation 7-10 in order to calculate k0 and S for each peak. The results of the regression are shown in Table 7-1.  133  Figure 7.4: Mobile phase at elution as a function of the gradient speed for 13 peaks Table 7-1: Gradient elution parameters determined by non-linear regression  peak retention time in 30 minute gradient  capacity factor in 100% water  ⎛ Δ (log k1)⎞ ⎟ S = ⎜⎜ ⎟ ⎝ Δϕ ⎠  [min]  [calculated]  [calculated, min-1]  8.97 10.1 12.1 15.2 15.8 16.2 17.0 18.2 18.9 20.4 23.0 29.1 34.1  1  7.8 x 10 2.0 x 102 1.7 x 102 1.1 x 103 5.6 x 104 8.5 x 104 1.1 x 105 1.5 x 105 4.6 x 105 3.1 x 106 4.2 x 103 1.2 x 104 3.2 x 106  0.072 0.087 0.52 0.065 0.11 0.11 0.11 0.10 0.11 0.11 0.044 0.038 0.058  134  As expected, the k0 values vary widely, with the earliest eluting peaks having much lower k0 values than those eluting later. However, the order of elution under standard separation conditions does not match the order of the k0 values. This is due to the fact that there is a significant variation in the S values for different proteins. Unfortunately, this means that it is not currently possible to relate tR to k’0 without performing multiple separations. 7.5. CONCLUDING REMARKS  Retention time variations in gradient reversed phase chromatography can be significantly reduced by applying a number of different marker-peak based correction calculations. The most successful of those tested was a correction adopted from the capillary electrophoresis literature, which includes terms correcting for variations that affect all compounds equally and a separate correction for variations that become more pronounced over time, such as the delivery of the mobile phase gradient. Marker compounds can also be used to convert retention times to a relative hydrophobicity index, thus relating the chromatographic results to more intrinsic protein properties. Unfortunately, these strategies are not successful when applied to separations carried out with different gradient conditions. This is because gradient elution times depend on both the capacity factor of the analyte in the initial mobile phase composition and on the rate of change of the capacity factor as a function of the changing gradient composition. Without knowledge of both parameters (through experimental determination), it is not possible to relate the retention times of peaks in different gradient HPLC separations to a common scale of hydrophobicity.  135  8. Chapter 8: Discussion and conclusions  136  8.1. REALIZATION OF RESEARCH OBJECTIVES  The current state of CE-MS research involves two distinct groups of users: those who are working with interfaces of their own design, which often work well in the context of their own research, and those who are working with the commercially available coaxial sheath-liquid CE-MS interfaces. Our goal was to develop an alternative to the coaxial sheath-flow systems, which would provide novice users with an interface that shares the ease of use and physically robust character of the sheath-flow interface, while reducing the dilution to allow for greater detection sensitivity, similar to a sheathless interface. We also wanted to provide a flexible hyphenation strategy that would allow the use of a variety of capillary coatings and modes of operation that may not be possible with sheathless interfacing strategies. 8.1.1. Design of a novel CE-MS hyphenation strategy  Concerning ease of use, our interface is actually simpler to assemble than a coaxial sheath-liquid interface, as the internal taper of the emitter needle provides automatic alignment of the capillary, both coaxially and longitudinally, while the sheathflow interface requires more careful attention to the position of the capillary where it protrudes from the metal emitter needle. Once assembled, the steps required to optimize the spray conditions are similar to the coaxial sheath-liquid interface. In both cases, the composition and flow rate of the modifier or sheath liquid, as well as the electrospray potential must be optimized. The only difference is that, because the dilution by the modifier solution is much less for our interface than by a conventional sheath liquid, the composition of the BGE must be taken into account when optimizing the modifier  137  composition. For example, when a purely aqueous BGE is used with a significant forward EOF, the modifier solution must contain enough organic solvent and be delivered at a high enough flow rate that the organic content of the solution exiting the emitter (which is the combined BGE and modifier solutions) reaches ~ 40%, to ensure a stable electrospray. On the other hand, when the EOF in the separation capillary is suppressed, the solution exiting the emitter will be almost entirely composed of the modifier, so that a lower organic content will suffice. The second goal, of increasing the sensitivity to a level comparable to a sheathless interface, has been harder to test because no sheathless interfaces are commercially available for comparison. Comparisons based on values stated in the literature are not possible, because the MS instruments and conditions used are as important as the interfacing approach in determining the overall sensitivity and limits of detection. Therefore, our assessment of the sensitivity has been based on a side-by-side comparison with a commercial sheath liquid interface. As shown in chapter 2, our interface provided an average five-fold improvement in the limits of detection for the analysis of amino acids, compared to the sheath flow interface. Reports in the literature of similar side-byside comparisons of sheathless and coaxial sheath flow interfaces have shown improvements in limits of detection from one-fold (no improvement) to over fifty-fold, for peptides192 and proteins193, respectively. Although in the latter case the sheathless interface has a clear advantage, we have nonetheless succeeded in attaining a level of sensitivity that is closer to the range achieved by sheathless interfaces.  138  8.1.2. Application of the interface to problems in biomolecular characterization  The third goal, of providing maximum flexibility in the choice of conditions and separation modes for capillary electrophoresis, has been tested in the applications of the interface to different problems in biomolecular analysis. There are two fundamental restrictions on the CE conditions that can be used with this interface. First, the CE current should be limited to no more than 15 μA, or else the stability of both the CE and ESI processes can suffer as a result of bubble formation in the needle tip. While this can be managed through adjustments to the BGE composition, electric field strength and capillary dimensions, it is an inconvenience when translating existing CE methods to CEMS. Secondly, as in any CE-MS arrangement, the BGE may not contain ingredients that could damage the mass spectrometer, unless the overall mobility of those components is negative, i.e. towards the capillary inlet. Despite these limitations, which are not unique to our interface, the overall success of the interface with more challenging modes of separation has been remarkable. We have demonstrated for the first time the separation of proteins in a neutral coated capillary using a low-dilution or sheathless CE-MS interface with no assisting pressure. It has also been possible to perform reverse-polarity separations of labeled glycans in bare fused silica capillaries. The modifier solution makes both of these arrangements possible, despite the fact that the EOF under these conditions is suppressed or towards the capillary inlet, by maintaining the solution requirements of the CE separation while also providing stable electrospray of the analytes as they migrate out of the separation capillary. The interface has also enabled us to perform cIEF with chemical mobilization  139  of the analytes, in order to maintain the maximum resolution of the focused protein zones. All of these separations would be impossible with a sheathless interface, or complicated by sheath-liquid induced suction with a coaxial sheath liquid interface. This ability to integrate challenging separation conditions with ESI-MS is a unique advantage of our interface. I believe that this versatility will be crucial in allowing our interface to stand out among the alternative hyphenation strategies. 8.1.3. Correction strategies for liquid chromatography  The work in the chapter 7 demonstrates that the reproducibility of reversed-phase separations can be improved through the application of simple mathematical corrections. This type of correction could be applied by including marker compounds in the needle wash solution of the HPLC instrument, so that they would be automatically included in every separation. However, chapter 7 also identified a limitation of this strategy; retention times in gradient reversed-phase chromatography cannot be related to intrinsic protein characteristics based on the results of a single gradient program. Therefore, a hydrophobicity index based on a two-marker normalization strategy will not be successful in relating results obtained under different chromatographic conditions. 8.2. FUTURE RESEARCH DIRECTIONS 8.2.1. Incorporation of pre-concentration strategies for increased sensitivity  In order to focus on more fundamental aspects of the CE and ESI processes, the injection parameters in all of the applications presented in this thesis were kept as simple as possible. However, there are a number of online preconcentration strategies that can  140  be easily incorporated with capillary electrophoresis in order to increase the sensitivity, including field-enhanced sample injection, transient isotachophoresis (t-ITP), dynamic pH junctions, micellar sweeping, electrokinetic supercharging, etc. For example, adding a t-ITP step prior to CE separation has been shown to increase the sensitivity of sheathless CE-ESI-MS up to 86-fold194. Future incorporation of similar preconcentration strategies should make our interface appropriate for situations where the current limits of detection do not meet the demands of the application. 8.2.2. CE-MS with gel buffers for improved resolution of complex glycan structures  Although we were able to adapt one widely used method for glycan analysis to ESI-MS using our interface, there is a great deal more that could be done in this area. The majority of glycan analysis is performed using a gel buffer BGE. In this case, the separated peaks for a glucose ladder standard will be spaced evenly in time, whereas with free-solution separations provide diminishing resolution for larger sugars, as the difference in size-to-charge ratio becomes less significant. The gel buffer is therefore crucial in resolving large, branched glycans where several structural isomers may be present with very small differences in the hydrodynamic volume, and thus in electrophoretic mobility. Having demonstrated the feasibility of using glycerol as an anti-convective medium for cIEF, we would like to examine the use of aqueous glycerol solutions as a medium for glycan analysis under reverse polarity. As in cIEF, its presence will aid in the suppression of both EOF and wall interactions. More importantly, it will also provide  141  a higher viscosity medium that may provide increased resolution of large and branched glycans. If glycerol does not provide the required resolution, there appears to be no reason that the polyethyleneoxide/lithium acetate gel buffer, used in most carbohydrate analysis kits and many standard protocols, could not also be adapted to CE-MS. It should be possible to fill the CE capillary with the gel buffer, while using a gel-free modifier solution to complete the electrical circuit and facilitate ionization. While the lithium ions in the gel buffer are not MS-compatible, upon application of the reverse polarity they will migrate away from the MS, towards the capillary inlet, and will be replaced by cations from the modifier solution (hydronium or ammonium). The acetate ions, which will move towards the capillary terminus and enter the electrospray solution, are fully MS compatible. Glycans injected at the capillary inlet will migrate through the gel buffer until they reach the capillary terminus, where they will enter the modifier solution and exit the needle for analysis by negative ion ESI-MS. In theory, any gel buffer could be implemented with reverse-polarity CE-ESI-MS in this manner. However, those containing MS incompatible anions (phosphate, borate), would require an exchange/equilibrium step prior to sample injection, where the incompatible ions are allowed to migrate out of the gel buffer while ESI is disabled, and a MS compatible buffer solution at the capillary inlet provides suitable ions for CE-MS analysis. These additional steps can be automated by the current CE instrument and will not add significant amount of time or labor in the CE-MS operation. This ability to perform high-resolution capillary gel electrophoresis separations of glycans with online  142  ESI-MS detection would be an important new tool for the field of glycomics and for the testing of biopharmaceuticals. 8.2.3. Development of cIEF-MS as a reproducible, high resolution alternative to 2D gel electrophoresis  While the feasibility of coupling cIEF with MS has been demonstrated in chapter 6, that work represents only the early development stage of this application. The longterm goal is to develop our cIEF-MS method into a reproducible and high-resolution alternative to two-dimensional gel electrophoresis that can be easily automated using standard CE and MS instrumentation. A great deal of effort has already been dedicated to improving the reproducibility and quality of the first-dimension capillary isoelectric focusing separation176, which is in itself an extremely complex system for optimization. Our future challenge will be to adapt the conditions that have already been established for traditional cIEF so that they are compatible with ESI-MS detection, and to develop a thorough understanding of how the presence of the interface affects the focusing and mobilization stages. 8.3. CONCLUDING REMARKS  The most important test of the significance of the hyphenation strategies elaborated in this thesis will be in the coming years, based on whether this technology is adopted by other researchers. The level of interest thus far has been promising. Our collaborators at Beckman Coulter have begun testing our interface for specific applications at their own labs. We have also received requests from other research  143  institutions for access to our interface in order to assess its utility to their work. It is our hope that we will be able to provide these interfaces to others beginning next year. In the meantime, the merits of the research presented in this thesis can be judged based on its achievement of the goals that were set out in the introduction. We have designed and tested a novel CE-MS interface that is used on a daily basis in our own laboratory and whose simplicity, robustness and sensitivity make it suitable for commercial development. The fundamental processes of the interface have been explored, including an elucidation of the benefits of beveled conductive ESI emitters, which may be applicable to the broader field of electrospray ionization research. The interface is successful in common applications of CE-MS, including the analysis of amino acids and peptides, but also has enabled a number of more challenging and previously unexplored applications, including the separation of intact proteins in a neutral coated capillary, the reverse-polarity, counter-EOF separation of glycans, and capillary isoelectric focusing – mass spectrometry with chemical mobilization. I believe that these innovations, in design and in the unique applications of the interface, will have a significance that is equal to or greater than that of the interface itself.  144  9. Bibliography (1)  Yates, J. R.; Ruse, C. I.; Nakorchevsky, A. Annual Review of Biomedical Engineering 2009, 11, 49-79.  (2)  Giddings, J. C. Unified Separation Science, 1 ed.; John Wiley & Sons, Inc.: New York, 1991.  (3)  Reynolds, J. A.; Tanford, C. Journal of Biological Chemistry 1970, 245, 51615165.  (4)  Görg, A.; Weiss, W.; Dunn, M. J. Proteomics 2004, 4, 3665-3685.  (5)  Rabilloud, T. Proteomics 2002, 2, 3-10.  (6)  Blackstock, W. P.; Weir, M. P. Trends in Biotechnology 1999, 17, 121-127.  (7)  Shi, Y.; Xiang, R.; Horv·th, C.; Wilkins, J. A. Journal of Chromatography A 2004, 1053, 27-36.  (8)  Righetti, P. G.; Castagna, A.; Antonioli, P.; Boschetti, E. Electrophoresis 2005, 26, 297-319.  (9)  Dole, M.; Mack, L. L.; Hines, R. L.; Mobley, R. C.; Ferguson, L. D.; Alice, M. B. Journal of Chemical Physics 1968, 49, 2240-2249.  (10)  Whitehouse, C. M.; Dreyer, R. N.; Yamashita, M.; Fenn, J. B. Analytical Chemistry 1985, 57, 675-679.  (11)  Carroll, D. I.; Dzidic, I.; Stillwell, R. N.; Haegele, K. D.; Horning, E. C. Analytical Chemistry 1975, 47, 2369-2373.  (12)  Robb, D. B.; Covey, T. R.; Bruins, A. P. Analytical Chemistry 2000, 72, 36533659.  (13)  Tanaka, K.; Waki, H.; Ido, Y.; Akita, S.; Yoshida, Y.; Yoshida, T.; Matsuo, T. Rapid Communications in Mass Spectrometry 1988, 2, 151-153.  (14)  Karas, M.; Bachmann, D.; Bahr, U.; Hillenkamp, F. International Journal of Mass Spectrometry and Ion Processes 1987, 78, 53-68.  (15)  Yamashita, M.; Fenn, J. B. Journal of Physical Chemistry 1984, 88, 4451-4459.  (16)  Yamashita, M.; Fenn, J. B. Journal of Physical Chemistry 1984, 88, 4671-4675.  (17)  Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64-71.  145  (18)  Wilm, M. S.; Mann, M. International Journal of Mass Spectrometry and Ion Processes 1994, 136, 167-180.  (19)  Schmelzeisen, R. International Journal of Mass Spectrometry and Ion Processes 1989, 90, 139.  (20)  Iribarne, J. V.; Thomson, B. A. Journal of Chemical Physics 1976, 64, 22872294.  (21)  Kebarle, P.; Tang, L. Analytical Chemistry 1993, 65, 972A-986A.  (22)  Schmidt, A.; Karas, M.; Dülcks, T. Journal of the American Society for Mass Spectrometry 2003, 14, 492-500.  (23)  Valaskovic, G. A.; Utley, L.; Lee, M. S.; Wu, J.-T. Rapid Communications in Mass Spectrometry 2006, 20, 1087-1096.  (24)  Wilm, M.; Mann, M. Analytical Chemistry 1996, 68, 1-8.  (25)  Issaq, H. J.; Chan, K. C.; Blonder, J.; Ye, X.; Veenstra, T. D. Journal of Chromatography A 2009, 1216, 1825-1837.  (26)  Hanai, T. In Advanced Chromatographic and Electromigration Methods in BioSciences; Deyl, Z., Ed.; Elsevier Science B.V.: Amsterdam, The Netherlands, 1998.  (27)  Bruins, A. P.; Covey, T. R.; Henion, J. D. Analytical Chemistry 1987, 59, 26422646.  (28)  Tomer, K. B.; Arthur Moseley, M.; Deterding, L. J.; Parker, C. E. Mass Spectrometry Reviews 1994, 13, 431-457.  (29)  Link, A. J.; Eng, J.; Schieltz, D. M.; Carmack, E.; Mize, G. J.; Morris, D. R.; Garvik, B. M.; Yates, J. R., III Nature Biotechnology 1999, 17, 676-682.  (30)  Washburn, M. P.; Wolters, D.; Yates, J. R. Nature Biotechnology 2001, 19, 242247.  (31)  Wolters, D. A.; Washburn, M. P.; Yates, J. R., 3rd Analytical Chemistry 2001, 73, 5683-5690.  (32)  Liu, H.; Berger, S. J.; Chakraborty, A. B.; Plumb, R. S.; Cohen, S. A. Journal of Chromatography B 2002, 782, 267-289.  (33)  Wang, H.; Hanash, S. Mass Spectrometry Reviews 2005, 24, 413-426.  (34)  Wall, D. B.; Kachman, M. T.; Gong, S.; Hinderer, R.; Parus, S.; Misek, D. E.; Hanash, S. M.; Lubman, D. M. Analytical Chemistry 2000, 72, 1099-1111.  146  (35)  Chong, B. E.; Yan, F.; Lubman, D. M.; Miller, F. R. Rapid communications in mass spectrometry : RCM 2001, 15, 291-296.  (36)  Jorgenson, J. W.; Lukacs, K. D. Analytical Chemistry 1981, 53, 1298-1302.  (37)  Terabe, S.; Otsuka, K.; Ando, T. Analytical Chemistry 1985, 57, 834-841.  (38)  Terabe, S. Procedia Chemistry 2010, 2, 2-8.  (39)  Otsuka, K.; Terabe, S. Journal of Chromatography A 2000, 875, 163-178.  (40)  Olivares, J. A.; Nguyen, N. T.; Yonker, C. R.; Smith*, R. D. Analytical Chemistry 1987, 59, 1231.  (41)  Smith, R. D.; Olivares, J. A.; Nguyen, N. T.; Udseth, H. R. Analytical Chemistry 1988, 436-441.  (42)  Smith, R. D.; Barinaga, C. J.; Udseth, H. R. Analytical Chemistry 1988, 60, 19481952.  (43)  Manisali, I.; Chen, D. D. Y.; Schneider, B. B. Trends in Analytical Chemistry 2006, 25.  (44)  Niessen, W. M. A.; Tjaden, U. R.; Vandergreef, J. Journal of Chromatography 1993, 636, 3-19.  (45)  Cai, J. Y.; Henion, J. Journal of Chromatography A 1995, 703, 667-692.  (46)  Banks, J. F. Electrophoresis 1997, 18, 2255-2266.  (47)  von Brocke, A.; Nicholson, G.; Bayer, E. Electrophoresis 2001, 22, 1251-1266.  (48)  Moini, M. Analytical and Bioanalytical Chemistry 2002, 373, 466-480.  (49)  Schmitt-Kopplin, P.; Frommberger, M. Electrophoresis 2003, 24, 3837-3867.  (50)  Moini, M. In Capillary Electrophoresis of Proteins and Peptides; Humana Press, 2004; Vol. 276, pp 253-290.  (51)  Schmitt-Kopplin, P.; Englmann, M. Electrophoresis 2005, 26, 1209-1220.  (52)  Zamfir, A. D. Journal of Chromatography A 2007, 1159, 2-13.  (53)  Ross, G. A. LCGC Europe 2001, 14, 45-49.  (54)  Liu, C. C.; Zhang, J. Z.; Dovichi, N. J. Rapid Communications in Mass Spectrometry 2005, 19, 187-192.  147  (55)  Liu, C. C.; Alary, J. F.; Vollmerhaus, P.; Kadkhodayan, M. Electrophoresis 2005, 26, 1366-1375.  (56)  Wachs, T.; Sheppard, R. L.; Henion, J. Journal of Chromatography B 1996, 685, 335-342.  (57)  Zhang, B.; Liu, H.; Karger, B. L.; Foret, F. Analytical Chemistry 1999, 71, 32583264.  (58)  Jussila, M.; Sinervo, K.; Porras, S. P.; Riekkola, M. L. Electrophoresis 2000, 21, 3311-3317.  (59)  Pleasance, S.; Thibault, P.; Kelly, J. Journal of Chromatography 1992, 591, 325339.  (60)  Fanali, S.; D'Orazio, G.; Foret, F.; Kleparnik, K.; Aturki, Z. Electrophoresis 2006, 27, 4666-4673.  (61)  Tseng, M. C.; Chen, Y. R.; Her, G. R. Electrophoresis 2004, 25, 2084-2089.  (62)  Chen, Y. R.; Tseng, M. C.; Her, G. R. Electrophoresis 2005, 26, 1376-1382.  (63)  Moini, M. Analytical Chemistry 2007, 79, 4241-4246.  (64)  Barnidge, D. R.; Nilsson, S.; Markides, K. E.; Rapp, H.; Hjort, K. Rapid Communications in Mass Spectrometry 1999, 13, 994-1002.  (65)  Bateman, K. P.; White, R. L.; Thibault, P. Rapid Communications in Mass Spectrometry 1997, 11, 307-315.  (66)  Guzzetta, A. W.; Thakur, R. A.; Mylchreest, I. C. Rapid Communications in Mass Spectrometry 2002, 16, 2067-2072.  (67)  Alexander, J. N.; Schultz, G. A.; Poli, J. B. Rapid Communications in Mass Spectrometry 1998, 12, 1187-1191.  (68)  Ishihama, Y.; Katayama, H.; Asakawa, N.; Oda, Y. Rapid Communications in Mass Spectrometry 2002, 16, 913-918.  (69)  Shui, W.; Yu, Y.; Xu, X.; Huang, Z.; Xu, G.; Yang, P. Rapid Communications in Mass Spectrometry 2003, 17, 1541-1547.  (70)  Petersson, M. A.; Hulthe, G.; Fogelqvist, E. Journal of Chromatography A 1999, 854, 141-154.  (71)  Mazereeuw, M.; Hofte, A. J. P.; Tjaden, U. R.; vanderGreef, J. Rapid Communications in Mass Spectrometry 1997, 11, 981-986.  148  (72)  Ramsey, R. S.; McLuckey, S. A. Journal of Microcolumn Separations 1995, 7, 461-469.  (73)  Nilsson, S.; Klett, O.; Svedberg, M.; Amirkhani, A.; Nyholm, L. Rapid Communications in Mass Spectrometry 2003, 17, 1535-1540.  (74)  Kele, Z.; Ferenc, G.; Klement, T.; Toth, G. K.; Janaky, T. Rapid Communications in Mass Spectrometry 2005, 19, 881-885.  (75)  Chen, Y. R.; Her, G. R. Rapid Communications in Mass Spectrometry 2003, 17, 437-441.  (76)  Zamfir, A. D.; Dinca, N.; Sisu, E.; Peter-Katalinic, J. Journal of Separation Science 2006, 29, 414-422.  (77)  Bendahl, L.; Hansen, S. H.; Olsen, J. Rapid Communications in Mass Spectrometry 2002, 16, 2333-2340.  (78)  Chang, Y. Z.; Her, G. R. Analytical Chemistry 2000, 72, 626-630.  (79)  Chang, Y. Z.; Chen, Y. R.; Her, G. R. Analytical Chemistry 2001, 73, 5083-5087.  (80)  Nilsson, S.; Wetterhall, M.; Bergquist, J.; Nyholm, L.; Markides, K. E. Rapid Communications in Mass Spectrometry 2001, 15, 1997-2000.  (81)  Zhu, X. F.; Thiam, S.; Valle, B. C.; Warner, I. M. Analytical Chemistry 2002, 74, 5405-5409.  (82)  Dahlin, A. P.; Wetterhall, M.; Liljegren, G.; Bergstrom, S. K.; Andren, P.; Nyholm, L.; Markides, K. E.; Bergquist, J. Analyst 2005, 130, 193-199.  (83)  Trapp, O.; Pearce, E. W.; Kimmel, J. R.; Yoon, O. K.; Zuleta, I. A.; Zare, R. N. Electrophoresis 2005, 26, 1358-1365.  (84)  Fang, L.; Zhang, R.; Williams, E. R.; Zare, R. N. Analytical Chemistry 1994, 66, 3696.  (85)  Cao, P.; Moini, M. Journal of the American Society for Mass Spectrometry 1997, 8, 561-564.  (86)  Wahl, J. H.; Smith, R. D. Journal of Capillary Electrophoresis 1994, 1, 62-71.  (87)  Moini, M. Analytical Chemistry 2001, 73, 3497-3501.  (88)  Settlage, R. E.; Russo, P. S.; Shabanowitz, J.; Hunt, D. F. Journal of Microcolumn Separations 1998, 10, 281-285.  (89)  Janini, G. M.; Conrads, T. P.; Wilkens, K. L.; Issaq, H. J.; Veenstra, T. D. Analytical Chemistry 2003, 75, 1615-1619. 149  (90)  Whitt, J. T.; Moini, M. Analytical Chemistry 2003, 75, 2188-2191.  (91)  Figeys, D.; van Oostveen, I.; Ducret, A.; Aebersold, R. Analytical Chemistry 1996, 68, 1822-1828.  (92)  Severs, J. C.; Smith, R. D. Analytical Chemistry 1997, 69, 2154-2158.  (93)  Tong, W.; Link, A.; Eng, J. K.; Yates, J. R. Analytical Chemistry 1999, 71, 22702278.  (94)  Jaworek, A.; Krupa, A. Journal of Aerosol Science 1999, 30, 873-893.  (95)  Cech, N. B.; Enke, C. G. Mass Spectrometry Reviews 2001, 20, 362-387.  (96)  Juraschek, R.; Rollgen, F. W. International Journal of Mass Spectrometry 1998, 177, 1-15.  (97)  Gucek, M.; Vreeken, R. J.; Verheij, E. R. Rapid Communications in Mass Spectrometry 1999, 13, 612-619.  (98)  Bruins, A. P. Journal of Chromatography A 1998, 794, 345-357.  (99)  Sen, A. K.; Darabi, J.; Knapp, D. R.; Liu, J. Journal of Micromechanics and Microengineering 2006, 16, 620-630.  (100) Smith, A. D.; Moini, M. Analytical Chemistry 2001, 73, 240-246. (101) Moini, M.; Cao, P.; Bard, A. J. Analytical Chemistry 1999, 71, 1658-1661. (102) Jackson, G. S.; Enke, C. G. Analytical Chemistry 1999, 3777-3784. (103) Konig, S.; Fales, H. M.; Haegele, K. D. Analytical Chemistry 1998, 70, 44534455. (104) Simpson, D. C.; Smith, R. D. Electrophoresis 2005, 26, 1291-1305. (105) Schiffer, E.; Mischak, H.; Novak, J. Proteomics 2006, 6, 5615-5627. (106) Harada, K.; Fukusaki, E.; Kobayashi, A. Journal of Bioscience and Bioengineering 2006, 101, 403-409. (107) Ramautar, R.; Somsen, G. W.; Jong, G. J. d. Electrophoresis 2009, 30, 276-291. (108) Monton, M. R. N.; Soga, T. Journal of Chromatography A 2007, 1168, 237-246. (109) Smyth, W. F. Electrophoresis 2005, 26, 1334-1357. (110) Hernandez-Borges, J.; Neususs, C.; Cifuentes, A.; Pelzing, M. Electrophoresis 2004, 25, 2257-2281.  150  (111) Maxwell, E. J.; Chen, D. D. Y. Analytica Chimica Acta 2008, 627, 25-33. (112) Klampfl, C. W. Electrophoresis 2006, 27, 3-34. (113) Stutz, H. Electrophoresis 2005, 26, 1254-1290. (114) Johansson, I.; Huang, E.; Henion, J.; Zweigenbaum, J. Journal of Chromatography 1991, 554, 311-327. (115) Tetler, L. W.; Cooper, P. A.; Powell, B. Journal of Chromatography A 1995, 700, 21-26. (116) Mokaddem, M.; Gareil, P.; Belgaied, J.-E.; Varenne, A. Electrophoresis 2009, 30, 1692-1697. (117) Lewis, K. C.; Opiteck, G. J.; Jorgenson, J. W.; Sheeley, D. M. Journal of the American Society for Mass Spectrometry 1997, 8, 495-500. (118) Chen, Y. R.; Tseng, M. C.; Chang, Y. Z.; Her, G. R. Analytical Chemistry 2003, 75, 503-508. (119) Tseng, W. C.; Chen, Y. R.; Her, G. R. Analytical Chemistry 2004, 76, 6306-6312. (120) Hsieh, F.; Baronas, E.; Muir, C.; Martin, S. A. Rapid Communications in Mass Spectrometry 1999, 13, 67-72. (121) Foret, F.; Zhou, H.; Gangl, E.; Karger, B. L. Electrophoresis 2000, 21, 13631371. (122) Liu, C. C.; Jong, R.; Covey, T. Journal of Chromatography A 2003, 1013, 9-18. (123) Mokaddem, M.; Gareil, P.; Belgaied, J.-E.; Varenne, A. Electrophoresis 2008, 29, 1957-1964. (124) Constantopoulos, T. L.; Jackson, G. S.; Enke, C. G. Journal of the American Society for Mass Spectrometry 1999, 10, 625-634. (125) Dams, R.; Benijts, T.; Günther, W.; Lambert, W.; Leenheer, A. D. Rapid Communications in Mass Spectrometry 2002, 16, 1072-1077. (126) Apffel, A.; Fischer, S.; Goldberg, G.; Goodley, P. C.; Kuhlmann, F. E. Journal of Chromatography A 1995, 712, 177-190. (127) Foret, F.; Thompson, T. J.; Vouros, P.; Karger, B. L.; Gebauer, P.; Bocek, P. Analytical Chemistry 1994, 66, 4450-4458. (128) Marginean, I.; Kelly, R. T.; Prior, D. C.; LaMarche, B. L.; Tang, K.; Smith, R. D. Analytical Chemistry 2008, 80, 6573-6579.  151  (129) Gibson, G. T. T.; Mugo, S. M.; Oleschuk, R. D. Mass Spectrometry Reviews 2009, 28, 918-936. (130) Gaskell, S. J. Journal of Mass Spectrometry 1997, 32, 677-688. (131) Page, J. S.; Kelly, R. T.; Tang, K.; Smith, R. D. Journal of the American Society for Mass Spectrometry 2007, 18, 1582-1590. (132) Smith, K. L.; Alexander, M. S.; Stark, J. P. W. Physics of Fluids 2006, 18, 092104-092101 - 092104-092107. (133) Ryan, C. N.; Smith, K. L.; Alexander, M. S.; Stark, J. P. W. Journal of Physics D: Applied Physics 2009, 42, 155504-155511. (134) Maxwell, E. J.; Xuefei, Z.; Hong, Z.; Nikita van, Z.; David, D. Y. C. Electrophoresis 2010, 31, 1130-1137. (135) Zhong, X.; Yi, R.; Holliday, A. E.; Chen, D. D. Y. Rapid Communications in Mass Spectrometry 2009, 23, 689-697. (136) Monton, M. R. N.; Terabe, S. Analytical Sciences 2005, 21, 5-13. (137) Haselberg, R.; de Jong, G. J.; Somsen, G. W. Journal of Chromatography A 2007, 1159, 81-109. (138) Herrero, M.; Ibanez, E.; Cifuentes, A. Electrophoresis 2008, 29, 2148-2160. (139) Fonslow, B. R.; Yates, J. R. Journal of Separation Science 2009, 32, 1175-1188. (140) Metzger, J.; Schanstra, J.; Mischak, H. Analytical and Bioanalytical Chemistry 2009, 393, 1431-1442. (141) Pelzing, M.; Neusüß, C. Electrophoresis 2005, 26, 2717-2728. (142) Horvath, J.; DolnÌk, V. Electrophoresis 2001, 22, 644-655. (143) Lucy, C. A.; MacDonald, A. M.; Gulcev, M. D. Journal of Chromatography A 2008, 1184, 81-105. (144) Huhn, C.; Ramautar, R.; Wuhrer, M.; Somsen, G. Analytical and Bioanalytical Chemistry 2010, 396, 297-314. (145) Schmalzing, D.; Piggee, C. A.; Foret, F.; Carrilho, E.; Karger, B. L. Journal of Chromatography A 1993, 652, 149-159. (146) Detlev, B.; Alfred, D.; Heribert, H.; Frank, K.; Martin, L. Electrophoresis 2001, 22, 3813-3818.  152  (147) Bruin, G. J. M.; Chang, J. P.; Kuhlman, R. H.; Zegers, K.; Kraak, J. C.; Poppe, H. Journal of Chromatography A 1989, 471, 429-436. (148) DolnÌk, V. Electrophoresis 2006, 27, 126-141. (149) Katayama, H.; Ishihama, Y.; Asakawa, N. Analytical Chemistry 1998, 70, 52725277. (150) Towns, J. K.; Regnier, F. E. Journal of Chromatography A 1990, 516, 69-78. (151) Koch, K. E. Annual Review of Plant Physiology and Plant Molecular Biology 1996, 47, 509-540. (152) Marino, K.; Bones, J.; Kattla, J. J.; Rudd, P. M. Nature Chemical Biology 2010, 6, 713-723. (153) Leader, B.; Baca, Q. J.; Golan, D. E. Nature Reviews Drug Discovery 2008, 7, 2139. (154) Sinclair, A. M.; Elliott, S. Journal of Pharmaceutical Sciences 2005, 94, 16261635. (155) Chen, F. T. A.; Evangelista, R. A. Electrophoresis 1998, 19, 2639-2644. (156) Guttman, A. Journal of Chromatography A 1997, 763, 271-277. (157) Guttman, A. s.; Chen, F.-T. A.; Evangelista, R. A.; Cooke, N. Analytical Biochemistry 1996, 233, 234-242. (158) Guttman, A.; Pritchett, T. Electrophoresis 1995, 16, 1906-1911. (159) Zamfir, A.; Peter-Katalinic, J. Electrophoresis 2001, 22, 2448-2457. (160) Zamfir, A.; Peter-Katalinic, J. Electrophoresis 2004, 25, 1949-1963. (161) Campa, C.; Coslovi, A.; Flamigni, A.; Rossi, M. Electrophoresis 2006, 27, 20272050. (162) Amon, S.; Zamfir, A. D.; Rizzi, A. Electrophoresis 2008, 29, 2485-2507. (163) Bindila, L.; Peter-Katalinic, J.; Zamfir, A. Electrophoresis 2005, 26, 1488-1499. (164) Hjertèn, S.; Zhu, M.-D. Journal of Chromatography A 1985, 346, 265-270. (165) Rodriguez-Diaz, R.; Wehr, T.; Zhu, M. Electrophoresis 1997, 18, 2134-2144. (166) Wehr, T. LC-GC North America 2004, 22, 998-1004. (167) Tang, Q.; Harrata, A. K.; Lee, C. S. Analytical Chemistry 1995, 67, 3515-3519.  153  (168) Tang, Q.; Harrata, A. K.; Lee, C. S. Analytical Chemistry 1996, 68, 2482-2487. (169) Jensen, P. K.; Pasa-Tolic, L.; Anderson, G. A.; Horner, J. A.; Lipton, M. S.; Bruce, J. E.; Smith, R. D. Analytical Chemistry 1999, 71, 2076-2084. (170) Kirby, D. P.; Thorne, J. M.; Gotzinger, W. K.; Karger, B. L. Analytical Chemistry 1996, 68, 4451-4457. (171) Clarke, N. J.; Naylor, S. Biomedical Chromatography 2002, 16, 287-297. (172) Clarke, N. J.; Tomlinson, A. J.; Naylor, S. Journal of the American Society for Mass Spectrometry 1997, 8, 743-748. (173) Kuroda, Y.; Yukinaga, H.; Kitano, M.; Noguchi, T.; Nemati, M.; Shibukawa, A.; Nakagawa, T.; Matsuzaki, K. Journal of Pharmaceutical and Biomedical Analysis 2005, 37, 423-428. (174) Busnel, J.-M.; Varenne, A.; Descroix, S.; Peltre, G.; Gohon, Y.; Gareil, P. Electrophoresis 2005, 26, 3369-3379. (175) Mokaddem, M.; Gareil, P.; Varenne, A. Electrophoresis 2009, 30, 4040-4048. (176) Mack, S.; Cruzado-Park, I.; Chapman, J.; Ratnayake, C.; Vigh, G. Electrophoresis 2009, 30, 4049-4058. (177) Zhang, C.-X.; Xiang, F.; Pa≈°a-Toliƒá, L.; Anderson, G. A.; Veenstra, T. D.; Smith, R. D. Analytical Chemistry 2000, 72, 1462-1468. (178) Thormann, W.; Caslavska, J.; Mosher, R. A. Journal of Chromatography A 2007, 1155, 154-163. (179) Righetti, P. G.; SimÛ, C.; Sebastiano, R.; Citterio, A. Electrophoresis 2007, 28, 3799-3810. (180) Hu, L.; Ye, M.; Jiang, X.; Feng, S.; Zou, H. Analytica Chimica Acta 2007, 598, 193-204. (181) Strittmatter, E. F.; Kangas, L. J.; Petritis, K.; Mottaz, H. M.; Anderson, G. A.; Shen, Y.; Jacobs, J. M.; Camp, D. G.; Smith, R. D. Journal of Proteome Research 2004, 3, 760-769. (182) Silva, J. C.; Denny, R.; Dorschel, C. A.; Gorenstein, M.; Kass, I. J.; Li, G.-Z.; McKenna, T.; Nold, M. J.; Richardson, K.; Young, P.; Geromanos, S. Analytical Chemistry 2005, 77, 2187-2200. (183) Nordström, A.; O'Maille, G.; Qin, C.; Siuzdak, G. Analytical Chemistry 2006, 78, 3289-3295.  154  (184) Smith, C. A.; Want, E. J.; O'Maille, G.; Abagyan, R.; Siuzdak, G. Analytical Chemistry 2006, 78, 779-787. (185) Christin, C.; Smilde, A. K.; Hoefsloot, H. C. J.; Suits, F.; Bischoff, R.; Horvatovich, P. L. Analytical Chemistry 2008, 80, 7012-7021. (186) Bigelow, C. C. Journal of Theoretical Biology 1967, 16, 187-211. (187) Hansch, C.; Leo, A.; Nikaitani, D. The Journal of Organic Chemistry 1972, 37, 3090-3092. (188) Keshavarz, E.; Nakai, S. Biochimica et Biophysica Acta (BBA) - Protein Structure 1979, 576, 269-279. (189) Katajamaa, M.; Oresic, M. Journal of Chromatography A 2007, 1158, 318-328. (190) Li, X.-F., Hongji Ren, Xiaochun Le, Ming Qi, Ian D. Ireland and Norman J. Dovichi Journal of Chromatography A. 2000, 869, 375-384. (191) Verbruggen, E. M. J.; Klamer, H. C.; Villerius, L.; Brinkman, U. A. T.; Hermens, J. L. M. Journal of Chromatography A 1999, 835, 19-27. (192) Sanz-Nebot, V.; Balaguer, E.; Benavente, F.; Barbosa, J. Electrophoresis 2005, 26, 1457-1465. (193) Haselberg, R.; Ratnayake, C. K.; de Jong, G. J.; Somsen, G. W. Journal of Chromatography A 2010, 1217, 7605-7611. (194) Busnel, J.-M.; Schoenmaker, B.; Ramautar, R.; Carrasco-Pancorbo, A.; Ratnayake, C.; Feitelson, J. S.; Chapman, J. D.; Deelder, A. M.; Mayboroda, O. A. Analytical Chemistry 2010, ASAP article.  155  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0060171/manifest

Comment

Related Items