Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Determination of small medicinal molecules by capillary electrophoresis mass spectrometry with on-line… Zhao, Tingting 2019

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2019_november_zhao_tingting.pdf [ 4.1MB ]
Metadata
JSON: 24-1.0380592.json
JSON-LD: 24-1.0380592-ld.json
RDF/XML (Pretty): 24-1.0380592-rdf.xml
RDF/JSON: 24-1.0380592-rdf.json
Turtle: 24-1.0380592-turtle.txt
N-Triples: 24-1.0380592-rdf-ntriples.txt
Original Record: 24-1.0380592-source.json
Full Text
24-1.0380592-fulltext.txt
Citation
24-1.0380592.ris

Full Text

DETERMINATION OF SMALL MEDICINAL MOLECULES BY CAPILLARY ELECTROPHORESIS MASS SPECTROMETRY WITH ON-LINE CONCENTRATION TECHNIQUES  by  Tingting Zhao  B.Sc., Fudan University, 2017  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Chemistry)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   August 2019  © Tingting Zhao, 2019   ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis/dissertation entitled:  Determination of Small Medicinal Molecules by Capillary Electrophoresis Mass Spectroscopy with On-line Concentration Techniques  submitted by Tingting Zhao in partial fulfillment of the requirements for the degree of Master of Science in Chemistry  Examining Committee: David Chen, Chemistry Supervisor  Russ Algar, Chemistry Supervisory Committee Member  Dan Bizzotto, Chemistry Supervisory Committee Member Takamasa Momose, Chemistry Additional Examiner    iii  Abstract Capillary electrophoresis (CE) is a powerful separation technology. CE has various advantages over other separation techniques (e.g. LC) including fast speed, separation efficiency and small solvent consumption. However, the small inner diameter of capillary and short optical path result in low sensitivity of CE for trace analytes. The poor detection limit restricts the wider application of CE. In this study, on-line preconcentration methods were developed to improve the detection sensitivity of normal capillary electrophoresis coupled with mass spectrometry (CE-MS). In Chapter 2, a field assisted sample stacking (FASS) technique was adopted for the quantification of imatinib (Gleevec), a drug approved for the treatment of chronic myeloid leukemia. In FASS, acetonitrile was used to dissolve the sample in order to enhance the electrical field strength in sample plug during initial CE separation. After optimization, baseline separations of imatinib with related compounds were obtained and detection sensitivity was improved by 8-fold. The sample pretreatment was simple and the LOD of 0.2 ng/mL was achieved for imatinib. Validation tests suggested this FASS-CE-MS method has a wide linearity range, high specificity, acceptable precision and accuracy.  In Chapter 3, acid barrage stacking (ABS) with CE-MS was used for the determination of alendronate sodium (Fosamax)- a drug used for the treatment of osteoporosis disease. This is the first ABS-CE-MS method for direct ALN analysis. Following sample injection, an acid barrage was introduced into separation capillary. The acid segment works as a barrier to stop the migration of negatively charged ALN ions and stacks them on the boundary. After optimization, the detection sensitivity was improved by 810-fold compared to normal capillary electrophoresis. The limit of detection achieved was 2 ng/mL for ALN. Relative peak area and concentration of ALN showed excellent linear relationship in the range of 8 - 2000 ng/mL (R2>0.9990). This iv  method was successfully used for quantification of ALN in drug tablet. Validation results showed good repeatability and accuracy of this method. Both online concentration methods -- FASS and ABS significantly improved the detection sensitivity of CE-MS. In addition, these two methods are easy-to-use and compatible with normal capillary electrophoresis system, which make them applicable for routine biological sample analysis. v  Lay Summary Capillary electrophoresis (CE) is an analytical separation technique, which separate analytes based on their charge to size ratio under strong electrical field in a capillary. As an inexpensive, fast and powerful technique, CE has been used in various application--proteomics, metabolism, pharmacokinetics, quality control and environmental science. However, low concentration sensitivity restricted CE in trace analysis. In this research, two on-line concentration techniques were used to enrich analytes within separation capillary. Compared to conventional CE, this method greatly simplified sample pretreatment and improved signal intensity. On the application side, quantifications of two drugs -- Fosamax for osteoporosis treatment and Gleevec for chronic myeloid leukemia treatment have been successfully achieved for drug tablets and serum analysis. As a result, CE-MS with online stacking techniques provide new method for complex mixtures in the pharmaceutical and clinical industry. vi  Preface The majority of the research included in this dissertation was conducted by the author, Tingting Zhao. The contribution of other researchers and collaborators are listed below. Contributions from other researchers: Chapter 2: The imatinib work was helped by Dr. Lingyu Wang in Dr. Chen’s group.  Chapter 3:  The alendronate sodium, pamidronate sodium and bovine serum were provided by Dr. Rizhi Wang and his student Christina Chen.  vii  Table of Contents  Abstract ......................................................................................................................................... iii Lay Summary ................................................................................................................................. v Preface ........................................................................................................................................... vi Table of Contents ......................................................................................................................... vii List of Tables ................................................................................................................................. xi List of Figures .............................................................................................................................. xii List of Equation .......................................................................................................................... xvi List of Symbols ........................................................................................................................... xvii List of Abbreviations .................................................................................................................. xix Acknowledgements ..................................................................................................................... xxi Dedication ................................................................................................................................... xxii Chapter 1: Introduction ................................................................................................................ 1 1.1 Capillary zone electrophoresis .................................................................................... 1 1.2 Online concentration methods ..................................................................................... 4 1.2.1 Field assisted sample stacking ................................................................................. 5 1.2.2 pH barrage stacking ................................................................................................. 7 1.3 Capillary electrophoresis-electrospray ionization-mass spectrometry ........................ 9 1.3.1 Triple quadrupole (QqQ) mass analyzer ................................................................. 9 1.3.2 Coupling of capillary electrophoresis with mass spectrometry ............................. 10 1.4 Tyrosine kinase inhibitor drug -- imatinib ................................................................. 12 1.5 Osteoporosis drug—alendronate sodium .................................................................. 12 viii  Chapter 2: Simultaneous determination of imatinib mesylate and related compounds by capillary electrophoresis mass spectrometry with field assisted sample stacking ..................... 15 2.1 Introduction ............................................................................................................... 15 2.2 Materials and methods ............................................................................................... 17 2.2.1 Chemicals and Reagents ........................................................................................ 17 2.2.2 Apparatus and instruments .................................................................................... 17 2.2.3 Standards ............................................................................................................... 18 2.2.4 Sample preparation ................................................................................................ 18 2.2.5 Capillary electrophoresis-MS procedure ............................................................... 19 2.3 Results and discussion ............................................................................................... 19 2.3.1 MRM parameters ................................................................................................... 19 2.3.2 Optimization of CZE conditions ........................................................................... 22 2.3.2.1 Influence of BGE pH on separation .............................................................. 22 2.3.2.2 Influence of BGE concentration (ammonium formate) on separation .......... 24 2.3.3 FASS Online concentration ................................................................................... 25 2.3.3.1 Influence of different sample solvent ............................................................ 26 2.3.3.2 Influence of the injection time ....................................................................... 27 2.3.4 Extraction efficiency of liquid-liquid-extraction ................................................... 28 2.3.5 Linearity and limit of detection ............................................................................. 29 2.3.6 Method validation .................................................................................................. 31 2.3.6.1 Specificity ...................................................................................................... 31 2.3.6.2 Intra-day assay ............................................................................................... 32 2.3.6.3 Inter-day assay ............................................................................................... 34 2.3.6.4 Stability study ................................................................................................ 35 ix  2.4 Conclusion ................................................................................................................. 37 Chapter 3: Quantitative determination of a bisphosphate drug - alendronate sodium by CE-MS assisted with acid barrage stacking technique ........................................................................... 38 3.1 Introduction ............................................................................................................... 38 3.2 Experimental .............................................................................................................. 39 3.2.1 Chemicals and reagents ......................................................................................... 39 3.2.2 Apparatus and instruments .................................................................................... 40 3.2.3 Standard solution preparation ................................................................................ 40 3.2.4 Drug tablets and serum sample pre-treatment ....................................................... 40 3.2.5 CE-MS procedure .................................................................................................. 41 3.3 Results and discussions ............................................................................................. 41 3.3.1 MRM parameters ................................................................................................... 41 3.3.2 Acid barrage stacking with large volume injection ............................................... 43 3.3.2.1 Influence of pH barrage length ...................................................................... 45 3.3.2.2 Sensitivity comparison between conventional CZE-MS and ABS-CE-MS. . 46 3.3.3 Influence of BGE pH and concentration on separation ......................................... 47 3.3.4 Influence of modifier flow rate .............................................................................. 48 3.3.5 Calibration curve for ALN in a linear range of 8-2000 ng/mL ............................. 49 3.3.6 Method validation .................................................................................................. 50 3.3.6.1 Specificity ...................................................................................................... 50 3.3.6.2 Intra-day assay ............................................................................................... 51 3.3.6.3 Inter-day assay ............................................................................................... 52 3.3.7 Analysis of drug tablets composition .................................................................... 52 3.3.8 Serum samples application .................................................................................... 54 x  3.4 Conclusion ................................................................................................................. 55 Chapter 4: Conclusion and future work ..................................................................................... 56 4.1 Concluding remarks ................................................................................................... 56 4.2 Future work ............................................................................................................... 57 Bibliography ................................................................................................................................. 58 Appendix ...................................................................................................................................... 70 xi  List of Tables  Table 2.1 Summary of MRM parameters for PYA, NDI, IM and d8 IM ..................................... 22 Table 2.2 Results of calibration curve for PYA, NID and IM ...................................................... 30 Table 2.3 The precision and accuracy of intra-day results. ........................................................... 33 Table 2.4 The precision and accuracy of inter-day test. ................................................................ 34 Table 2.5 Stabilities of three analytes after 3 freeze-and-thaw cycles. ......................................... 35 Table 2.6 Stability test of PYA. ..................................................................................................... 36 Table 2.7 Stability test of NDI. ..................................................................................................... 37 Table 2.8 Stability test of IM. ........................................................................................................ 37 Table 3.1 A summary of  calirabtion curve for ALN using CE-MS and CE-LIF methods. ......... 50 Table 3.2 A summary of intra-day test results. ............................................................................. 51 Table 3.3 A summary of inter-day test results. ............................................................................. 52 Table 3.4 A summary of ABS-CE-MS and CE-LIF results for commercial ALN tablets. ........... 53 Table 4.1 Summary of regression equation for ALN with CE-LIF methods ................................ 73 Table 4.2 Intra-assay precision and accuracy ................................................................................ 74 Table 4.3 Inter-assay precision and accuracy ................................................................................ 74  xii  List of Figures  Figure 1.1 Schematic diagram of a CE setup with optical detector. ............................................... 1 Figure 1.2 Effect of EOF and electrophoretic mobility on the migration velocity of three types of particles. ........................................................................................................................................... 3 Figure 1.3 Schematics for field assisted sample stacking (FASS). (a) positive ions migrate faster towards cathode in sample segment, (b) positive ions slow down up entering BGE, (c) positive ions stack up at the boundary, (d) focused ions are subject to separation under CZE mode. ......... 6 Figure 1.4 Schematics of pH barrage mediated stacking for weak acids. (a) negative analytes migrate towards anode, (b) analytes are protonated upon entering the acid barrage and stop migration, (c) analytes stack up at the boundary, (d) focused analytes are subject to separation under CZE mode. ............................................................................................................................. 8 Figure 1.5 Schematic diagram of a triple quadrupole spectrometer ................................................ 9 Figure 1.6 The configuration of CE-ESI-MS setup with a Tee-union interface. .......................... 11 Figure 1.7 Schematic diagram of bone rebuilding. Resorption phase: osteoclasts destroy and remove fractured bones. Formation phase: osteoblasts help to form new bone matrix. ............... 13 Figure 2.1 Molecular structures of IM, NDI, PYA and d8 IM. ..................................................... 16 Figure 2.2 Representative Q1/Q2 spectra obtained for IM(a), NDI(b), PYA(c) and d8 IM(d); Structures of precursor and fragments for IM(a), NDI(b), PYA(c) and d8 IM(d). ....................... 21 Figure 2.3 Influence of buffer pH on separation. Buffer: 10mM HCOONH4 dissolved in 75%MeOH and 25%H2O. Sample was injected at 5 psi for 5 s. ................................................... 22 Figure 2.4 Influence of different concentration of ammonium formate (HCOONH4) on separation. Buffer: 75%MeOH 25%H2O, pH=2.4. Sample injected at 5psi for 5s ....................... 24 xiii  Figure 2.5 FASS process of IM, NDI and PYA. Positively charged IM, NDI and PYA (a) migrate faster towards cathode in sample segment, (b) slow down up entering BGE, (c) stack up at the boundary, (d) focused ions are subject to separation under CZE mode. ............................. 25 Figure 2.6 Influence of different sample solvent on stacking efficiency. BGE: 5 mM HCOONH4 in 75%MeOH and 25% H2O at pH 2.4. Sample injected at 5 psi for 5 s. ..................................... 26 Figure 2.7 Influence of different injection pressure on stacking efficiency, given injection time of 5s. BGE: 5 mM HCOONH4 in 75%MeOH and 25% H2O at pH 2.4. ........................................... 27 Figure 2.8 Representative electropherogram obtained with (a) normal injection at 1 psi for 5 s; (b) FASS injected at 7.5 psi for 5 s. Sample contains 50 µg/mL PYA, 10 µg/mL NDI and 10 µg/mL IM. ..................................................................................................................................... 28 Figure 2.9 Calibration curves for PYA(a), NDI (b) and IM (c). ................................................... 29 Figure 2.10  Electropherograms of (a) blank serum sample without d8 IM, (b) blank serum sample spiked with d8 IM and (c) LLOQ sample. ........................................................................ 32 Figure 3.1 Molecular structures of (a)ALN and (b)PAM .............................................................. 38 Figure 3.2 Representative Q1/Q2 mass spectrum of ALN and PAM. .......................................... 42 Figure 3.3 Schematics of acid barrage stacking (ABS) process for ALN. (a) negative ALN migrate towards anode, (b) ALN are protonated and become neutral upon entering the acid barrage, (c) ALN  stack up at the boundary, (d) focused analytes are subject to separation under CZE mode. ..................................................................................................................................... 43 Figure 3.4 Acid dissociation constants of ALN. ........................................................................... 44 Figure 3.5 Influence of injection time of acid barrage at 30 psi pressure on sample stacking efficiency. Standard solution with 200 ng/mL ALN was used. .................................................... 45 xiv  Figure 3.6 Sensitivity comparison between (a) conventional CZE-MS where 2 µg/mL ALN was injected at 2psi for 10s and (b) ABS-CE-MS, where 2 µg/mL ALN was injected at 30 psi for 40s followed by a plug of acid barrage. Capillary electrophoresis separations were conducted with 30 kV voltage with 1.0 psi pressure. .................................................................................................. 46 Figure 3.7 Influence of BGE (a) pH and (b) concentration on separation. ................................... 47 Figure 3.8 Modifier flow rate effect on the signal intensity of ALN with 200 ng/mL concentration. ................................................................................................................................ 48 Figure 3.9 Calibration curve for ALN using 200 ng/mL PAM as internal standard. .................... 49 Figure 3.10 Electropherograms of (a) blank solution, (b) blank solution spiked with ALN, (c) blank solution spiked with PAM and (d) LLOQ sample. .............................................................. 51 Figure 3.11 Representative electropherograms of commercial ALN tablets (a) without PAM (internal standard) and (b) with PAM. .......................................................................................... 53 Figure 3.12 Electropherogram for extractants of serum samples containing (a) 10 µg/mL ALN and 10 µg/mL PAM (b) foregoing extracted serum sample spiked with 10 µg/mL ALN and 10 µg/mL PAM just before CE-MS analysis ..................................................................................... 54 Figure 4.1 Reaction of fluorogenic derivatization reagent FQ with primary amine. .................... 70 Figure 4.2 Influence of derivatization buffer pH ........................................................................... 71 Figure 4.3 Influence of BGE concentration (Na2HPO4) on separation. ........................................ 72 Figure 4.4 Influence of BGE pH on separation. ............................................................................ 72 Figure 4.5 Calibration curve for ALN with CE-LIF method. ....................................................... 73 Figure 4.6 Electropherograms of drug tablet extract ..................................................................... 75 Figure 4.7 Electropherogram of serum sample containing 100 µg/mL ALN and 100 µg/mL PAM (I.S.). .............................................................................................................................................. 75 xv   xvi  List of Equation  Equation 1.1 Electrophoretic velocity as a function of electrical field (𝑬), ion charge (Qeff), hydrodynamic radius (R) and the viscosity of medium (η). ............................................................ 2 Equation 1.2 Electrophoretic mobility as a function of ion charge (Qeff), hydrodynamic radius (R) and the viscosity of medium (η). ..................................................................................................... 2 Equation 1.3 Electroosmotic flow velocity as a function of electrical field, dielectric constant of solution (ε), zeta potential (ζ) and the viscosity of medium (η) ...................................................... 3 Equation 1.4 Electroosmotic flow mobility as a function of dielectric constant of solution(ε), zeta potential (ζ) and the viscosity of medium (η) .................................................................................. 3 Equation 1.5 Relationship of electric field strength between sample zone and BGE region in FASS. ............................................................................................................................................... 5 Equation 1.6 Relationship of ion concentration inside BGE and sample zone ............................... 6 Equation 1.7 The length of stacked sample (Lstacked sample) as a function of initial length of injected sample (Linitial sample) and resistivity of BGE (ρBGE) and sample (ρs). ............................................... 7 Equation 2.1 The calculation formula of bias for accuracy test. ................................................... 32 Equation 2.2 Calculation formula of relative recovery for stability test. ...................................... 35  xvii  List of Symbols °C Degrees Celsius µL Microliter µm Micrometer cps Counts per second E Electric field strength (Volts/cm) g Gram ġ Unit of gravitational acceleration h Hour kW Kilo-ohm kV Kilovolt M Molarity; M = mol/L mL Millilitre mol Mole µep Electrophoretic mobility µeof Electroosmotic mobility  nL Nanolitre nm Nanometre η Viscosity of medium pH Power of Hydrogen; pH = -log[H+] psi Pounds per square inch Qeff Effective charge of a spherical particle in solution R Hydrodynamic radius of a particle xviii   R2 Coefficient of determination for a fitted calibration curve t Migration time of an analyte V Volt ν Velocity of analyte  xix  List of Abbreviations ABS Acid barrage stacking ACN Acetonitrile ALN Alendronate sodium BGE Background electrolyte CE Capillary electrophoresis  CE-ESI-MS Capillary electrophoresis-electrospray ionization-mass spectrometry CE-MS Capillary electrophoresis-mass spectrometry  CGE Capillary gel electrophoresis CL Chemiluminescence CZE Capillary zone electrophoresis cIEF Capillary isoelectric focusing DC Direct current EOF Electroosmotic flow ESI Electrospray ionization FASS Field assisted sample stacking GC Gas chromatography GC-MS Gas chromatography-mass spectrometry i.d. Inner diameter (of capillary) IM Imatinib LC Liquid chromatography LC-MS Liquid chromatography-mass spectrometry LIF Laser-induced-fluorescence MEKC Micellar electrokinetic capillary chromatography MeOH Methanol MRM Multiple reaction monitoring MS Mass spectrometry o.d. Outer diameter (of capillary) QqQ Triple quadrupole RF Radio frequency R.S.D. Relative standard deviation xx    SPE Solid-phase extraction TKIs Tyrosine kinase inhibitors UV Ultraviolet xxi  Acknowledgements I sincerely give my gratitude to my supervisor, Dr. David Chen. He is always patient to my endless questions and encourage me to explore the frontier of science. In addition to instructions in academic research, Dr. Chen always generously shares his opinion about career and life. His open mind and interesting soul deeply impress me. Thank you to my lab-mates for the happiness and colorful life you all bring to me. Special thanks to Zi-Ao Huang, Lingyu, Jianhui Cheng and visiting scholar Noorfatimah for the share of your knowledge and experimental techniques in analytical chemistry field.  Special thanks to Dr. Rizhi Wang and his student Christina Chen for providing alendronate sodium and pamidronate sodium standards for the studies in Chapter 3. Thanks to Mitacs for offering me an opportunity to explore academic world in Canada during my undergraduate study and supporting my graduate study with Globalink Graduate Fellowship. I really appreciate my family’s solid support and companionship throughout my education. I will never forget your encouragement when I am overwhelmed by research and nearly gave up. Your love and faith are the priceless treasure of my whole life.   xxii  Dedication1  Chapter 1: Introduction 1.1 Capillary zone electrophoresis Capillary electrophoresis(CE) is a powerful and efficient separation technology since first demonstrated by Jorgenson in 19811. CE drives the migration of ionic analytes with high electrical field across the capillary. In terms of charge-to-size ratio characteristics, different species can be distinguished with CE technique. Compared to other chromatography (e.g. LC ), CE possesses several advantages: (1) versatility for a wide range of compounds under different separation mode (2) low sample (~ 10 nL) and organic solvent (~2 mL) consumption (3) a high potential for miniaturization devices (e.g. microfluidic chips) 2, 3.  Figure 1.1 Schematic diagram of a CE setup with optical detector. In terms of the chemical/physical properties of target species, CE can be operated under various mode, including capillary zone electrophoresis (CZE), micellar electrokinetic capillary chromatography (MEKC), capillary gel electrophoresis (CGE), capillary isoelectric focusing (cIEF). CZE is the most popular and conventional separation mode of CE, which realizes the separation of compounds of a mixture merely based on their charge-size-ratio. In CZE, a fused-silica capillary with a narrow inner diameter (less than 200 µm) is usually used as separation 2  column. A schematic diagram of the setup is demonstrated in Figure 1.1. The capillary is filled with conductive medium called background electrolyte (BGE). After a few nanoliters of sample was injected into capillary, a high voltage (e.g. 20 to 30 kV) is applied on electrodes, imposing a high electrical field across the capillary. Thus, analytes in the sample zone are driven to migrate by electrostatic force -- anions move towards the cathode, cations towards the anode and neutral molecules stay stationary. This process is called electrophoresis. The electrophoretic velocity (𝜈%&) is determined by Equation 1.1 4, 5. Electrophoretic mobility (µ(⃗ %&) is more common to describe electrophoretic behavior of ions independent of electric field. µ(⃗ %& is defined as Equation 1.2 4, 5. This equation demonstrates that µ(⃗ %& depends on the analyte’s charge-to-size ratio (Qeff / R).  𝜈%& = 𝑄%++𝐸(⃗6𝜋𝜂𝑅  Equation 1.1 Electrophoretic velocity as a function of electrical field (𝑬(⃗), ion charge (Qeff), hydrodynamic radius (R) and the viscosity of medium (η). 𝜇%& = 𝑄%++6𝜋𝜂𝑅	 Equation 1.2 Electrophoretic mobility as a function of ion charge (Qeff), hydrodynamic radius (R) and the viscosity of medium (η).  In addition to electrophoretic mobility, electroosmotic flow (EOF) also plays prominent impact on the mobility of analytes. EOF is a bulk flow of solution arising from the movement of the diffuse layer in the electric double layer. Silicon hydroxyl group on the inner surface of capillary experience deprotonation above pH 3, making capillary wall carry negative charge at higher pHs. Driven by electrostatic force, cations will assemble close to the capillary wall. Then, a double layer forms at the boundary of BGE and capillary surface. With the application of high voltage, the positively charged double layer will move towards anode, dragging the bulk solution 3  mobilize in the same direction. This movement of bulk solution is known as electroosmotic flow (EOF). The velocity of electroosmotic flow (𝜈%3+) is determined by equation 1.36. Same as electrophoretic mobility, electroosmotic coefficient (µ(⃗ %3+) is commonly used to describe magnitude of EOF independent of electrical field (𝐸(⃗ ). µ(⃗ %3+ is determined by Helmultz equation as Equation 1.46. Obviously, µ(⃗ %3+ is proportional to zeta potential which depends on the charge density on the capillary inner wall. Higher pH and lower ionic strength of BGE contribute to a higher zeta potential. ?⃗?%3+ = 	 𝜀𝜁𝐸(⃗4𝜋𝜂 Equation 1.3 Electroosmotic flow velocity as a function of electrical field, dielectric constant of solution (ε), zeta potential (ζ) and the viscosity of medium (η) µ(⃗ %3+ = 𝜀𝜁4𝜋𝜂 Equation 1.4 Electroosmotic flow mobility as a function of dielectric constant of solution(ε), zeta potential (ζ) and the viscosity of medium (η)  Figure 1.2 Effect of EOF and electrophoretic mobility on the migration velocity of three types of particles. The apparent mobility of analytes is the net effect of electrophoretic mobility and electroosmotic mobility. Figure 1.2 illustrates net migration mobilities of positive-charged, negative-charged and neutral species. Positively charge particles will migrate fastest towards anode, negatively charged particles slowest. Neutral species will migrate along with bulk EOF. 4  Therefore, a mixture of species can be separated and distinguished based on their charge-to-size ratio. 1.2 Online concentration methods Although CE has been used as an analytical technique in recent years, low concentration sensitivity limit more wide application of CE in trace analysis. This problem mainly arises from the small inner diameter of capillary which restricts the volume of sample injected and offers a short optical path for online detection. In order to address this issue, more sensitive detection technologies such as laser induced fluorescence (LIF), mass spectrometry (MS), chemiluminescence (CL) and electrochemical detection, have been adopted as an alternative for UV detector. However, those method requires expensive and complex setups as well as software, which are not affordable to some academic laboratories.7-10 Another more feasible strategy is to pre-concentrate the target analytes pre-column or on-column. Pre-column concentration methods, also called off-line concentration, enrich analytes before loading sample into CE system. Common offline pre-concentration methods include liquid-liquid extraction,11, 12 dispersive micro-liquid-liquid-extraction,13 salting-out-liquid-liquid extraction14, and solid phase extraction.15, 16 On-line pre-concentration does not require special experimental instrument or capillary surface modification. They electrokinetically focus analytes with large sample volume within the separation capillary and amplify analytes’ concentration, thus significantly enhancing detection sensitivity. Compared to off-line concentration methods, on-line concentration is more convenient which help to decrease the sample preparation workload. Commonly used on-line concentration modes include field assisted sample stacking (FASS) 17, pH-mediated stacking18-21, isotachophoresis (ITP) 22-26, sweeping 27-30 and transient moving chemical reaction boundary method (tMCRBM) 31, 32. Usually, achievements of pre-concentration are based on differences 5  between sample zone and BGE in physicochemical properties -- conductivity, ion strength, pH, surface affinity and existence of addictive.  In this study, field assisted sample stacking (FASS) method was adopted to assist the determination of imatinib in serum, and acid barrage stacking (ABS) technique assisted the determination of alendronate sodium in serum and drug tablets. 1.2.1 Field assisted sample stacking Since first proposed by Mikkers et al. in 1979,33 field assisted sample stacking (FASS) became the most popular and easy-to-handle technique to enrich the sample on-line in CE. Conventional CZE requires the sample solvent to be as similar to BGE as possible in order to avoid asymmetric and distorted peaks. By contrast, samples are prepared in a lower concentration BGE or another less conductive solvent (e.g. methanol, acetonitrile) instead of BGE.  In FASS capillary electrophoresis, the capillary of length L is filled by less conductive sample with resistivity of ρs and BGE with resistivity of ρBGE (γ=ρs/ρBGE >1). With voltage applied on electrodes, the relation of local electric field strengths of two regions are as below 𝐸789&:% = γ	E=>? Equation 1.5 Relationship of electric field strength between sample zone and BGE region in FASS.  Equation 1.5 clearly shows that the field strength in sample zone is increased and equal to γ fold of EBGE. Hence, ions inside lower conductive sample zone will experience stronger electrostatic force and migrate faster than ions entering BGE region.  6   Figure 1.3 Schematics for field assisted sample stacking (FASS). (a) positive ions migrate towards the cathode in sample segment, (b) positive ions slow down up entering BGE, (c) positive ions stack up at the boundary, (d) focused ions are subject to separation under CZE mode. As illustrated in Figure 1.3, charged analytes migrate faster in sample zone and slow down when entering the higher conductive BGE zone. Decrease of ions velocity leads to the stacking of analytes across the concentration boundary. Given the flux of ions across the concentration boundary must be conserved, concentration of ions is distributed as below  𝐶=>? = 𝛾	𝐶7 Equation 1.6 Relationship of ion concentration inside BGE and sample zone It’s clearly that the concentration of ions will be amplified by factor γ after the focusing process. Additionally, because of the conservation of the total number of ions, the length of focused sample zone must decrease by the same factor γ.  L stacked sample=L initial sample /γ= L initial sample * ρBGE / ρs   7  Equation 1.7 The length of stacked sample (Lstacked sample) as a function of initial length of injected sample (Linitial sample) and resistivity of BGE (ρBGE) and sample solution (ρs). In summary, the change of the ions’ velocity across the concentration boundary resulted in a reduction of the ions’ zone length and amplification of ions concentration compared to the initial injected sample plug. Focused ions are subject to separation under CZE mode afterwards. Note that positive ions stack up in the front of initial sample plug and negative ions at the back of sample plug. The stacking efficiency can be described by the ratio of resistivity of sample zone and BGE region (γ).4, 34 Obviously, the greater conductivity difference between sample solution and BGE, higher stacking efficiency can be obtained. 10-80 fold enhancements have already reported using FASS.35 The most common practices in FASS are to dilute the sample with water or organic solvent (methanol, acetonitrile etc.).35, 36 1.2.2 pH barrage stacking In pH-mediated stacking, analytes are focused within the separation capillary by manipulating the ionization of analytes in multi-section electrolytes 37, 38. This method is usually feasible for weakly ionic analytes, but not applicable for strong acid or base whose electrophoretic mobilities (ionization) are independent of the pH of the surrounding liquid 18. A number of pH-mediated stacking methods were reported, such as dynamic pH junction 18, 37, 39, pH barrage stacking, moving reaction boundary 40, 41, transient moving chemical reaction boundary 32, 41. 8   Figure 1.4 Schematics of pH barrage mediated stacking for weak acids. (a) negative analytes migrate towards anode, (b) analytes are protonated upon entering the acid barrage and stop migration, (c) analytes stack up at the boundary, (d) focused analytes are subject to separation under CZE mode.  In pH barrage stacking, an extra acid or base segment is also introduced into CE setup except for sample plug and separation BGE. Like normal CZE, the sample is prepared in BGE solution. The configuration of the component inside the capillary for weak acid stacking is depicted in Figure 1.4(A). As illustrated in Figure 1.4(B), the sample segment and acid barrage segment of different pH are sandwiched between two plugs of BGE. The choice of acidic or basic barrage and the order of segment locations are both dependent on the electrophoretic direction of target analytes.  Wang et. al. simulated the focusing mechanism of pH barrage stacking in 2018. 20 Studies demonstrated a sharp pH boundary formed at the interface between pH barrage and sample segment. As shown in (B), deprotonated acid species migrate towards higher electric potential 9  (inlet end) and suffer protonation upon entering acid barrage. Due to the change of dissociation status, target species slow down at the sharp pH boundary which allows latter ones catch up and get focused. The band width of stacked species can get further narrowed before the disappearance of pH barrage. After the stacking process, focused analyte will be separated under CZE mode. By adjusting the pH difference between sample prepared in BGE and pH barrage, analytes with specific pKa can be selectively focused while leaving other species continue to migrate as a broad plug. pH barrage stacking is compatible with large sample volume on microliter level as it is not affected by the ionic strength and pH of sample20, 42, 43. With optimized separation conditions and the ratio of sample length to pH barrage length, femtomole level analyses of phosphoproteins 44 , hormones 45 even ions 46 have been reported.  1.3 Capillary electrophoresis-electrospray ionization-mass spectrometry 1.3.1 Triple quadrupole (QqQ) mass analyzer  Figure 1.5 Schematic diagram of a triple quadrupole spectrometer  Mass spectrometry differentiates and detects gaseous charged species based on their mass-to-charge ratio (m/z). The rapid, sensitive and information-rich characteristics contribute to the wide application of mass spectrometry in environmental science, pharmacokinetics, metabolomics and proteomics. Linear quadrupole ion trap47 is one type of mass analyzer which consists of four parallel hyperbolic electrodes. A voltage combining direct current (DC) and 10  radio frequency (RF) are applied to those four rods, creating an oscillating electric field within quadrupole region47. Only ions with specific mass-to-charge ratio will pass through the quadrupole region and get detected. The rest of ions will collide on the surface of quadrupole. In addition, an RF-only quadrupole can also serve as a non-selective ion guide. Triple quadrupole (QqQ) mass analyzer is composed of three linear quadrupoles as illustrated in Figure 1.5. The first stage quadrupole (Q1) imposed with DC + RF voltage can filter ions with target m/z. In the collision cell (Q2) with only RF voltage, precursor ions from Q1 are subject to covalent bond dissociation. The fragmentation of precursor ions is further scanned and filtered by the second-stage quadrupole (Q3) with DC + RF voltage. This process is called multiple reaction mode (MRM). MRM enables identification of structural information and quantification of target analytes simultaneously. 1.3.2 Coupling of capillary electrophoresis with mass spectrometry The combination of capillary electrophoresis with mass spectrometry (CE-MS) has exhibited powerful ability to realize two-dimensional separation of charged species. There are various coupling techniques used to introduce separated analytes from the capillary into the mass spectrometer, such as electrospray ionization (ESI)48, chemical ionization49, inductively coupled plasma 50 and matrix assisted laser desorption ionization 51. Among them, electrospray ionization (ESI) converts solvated ions into isolated gaseous ions at atmospheric pressure, significantly reducing the fragmentation of ions during the ionization process. Furthermore, ESI can be easily adapted for both volatile and non-volatile species including ions, single molecules and polymolecules. The soft ionization characteristic and its versatility for different types of compounds makes it become the most popular soft ionization technique in CE-MS application52, 11  53. Figure 1.6 The configuration of CE-ESI-MS setup with a Tee-union interface. To optimize CZE separation and ESI process independently, our group designed a micro-flow through interface in 2010 52, 53. Figure 1.5 shows the approximate configuration of CE-ESI-MS setup with a Tee-union interface. The apparatus is consisted of three parts -- a Tee junction into a steal-stainless needle, a separation capillary and a modifier capillary. As illustrated above, the outlet of separation capillary is inserted into a stainless-steel needle with a beveled tip through the Tee union. Then, a small dead space between the end of separation capillary and inner wall of needle tip. Another capillary is assembled through the lower port of the Tee union which supplies a flow of modifier solution. The modifier flow helps to sustain the flow rate of effluent from the electrospray needle and control the composition of electrosprayed liquid.  Upon the introduction of modifier solutions, separated analytes meet with modifier solution in the dead space, undergo electrospray ionization and enter the inlet of mass spectrometer subsequently. When appropriate solution flow rates are adopted (e.g. 0.1 µl/min) 52, 53, dilution effect of modifier solution can be significantly decreased with our Tee union interface. Additionally, since there is no requirement for pre-treatment of separation capillary 12  compared to sheathless electrospray interface54, the operation of CE-MS with a Tee union interface is robust, reproducible and suitable for routine analysis.  1.4 Tyrosine kinase inhibitor drug -- imatinib Tyrosine kinases play vital roles on cellular level, manipulating signal transduction, proliferation, differentiation and apoptosis. Researches demonstrate appearances of some tumors such as renal cell cancer, gastro intestinal stromal tumors and chronic myeloid leukemia, are closely related to the misfunction of tyrosine kinases. Until now, there are up to ~30 tyrosine kinase inhibitors (TKIs) mainly used as oral medication for targeted cancer therapy55, 56. TKIs target at specific molecules and stop the function of tyrosine kinases, which cause less side effect compared to chemotherapy.  Imatinib (IM) is the first TKI found by Novartis in Switzerland in 2000.55, 56 IM inhibits the Bcr-Abl tyrosine kinase in human body which leads to programmed death of unregulated myeloid cells.56 Nevertheless, IM treatment failed on some patients who present IM drug resistance. Studies suggest that the plasmatic IM level can provide more information about IM response of cancer patients, which help to customize IM dosage for each patient and guide the treatment in clinical practice. 57-59 Until now, IM drug quality control, pharmacokinetic studies and metabolism researches are mainly conducted on HPLC.60-62 There are limited studies that explore the feasibility of CE techniques for IM studies, and among which most are based on CE-UV.14, 55, 56, 63, 64 As an information-rich and powerful separation techniques, CE-MS is a good alternative for IM analysis. The study in Chapter 2 aims to develop and validate CE-MS method assisted with field assisted sample stacking for IM studies in serum. 1.5 Osteoporosis drug—alendronate sodium Bone is the special connective tissue of great importance in our body. Bone undergoes dynamic resorption and rebuilding continuously to substitute infantile bone and renew fractured 13  bone. As shown in Figure 1.7, The bone rebuilding involves the resorption of old bone performed by osteoclast cells and the formation of new bone performed by osteoblasts. The correct balance of activity between osteoclasts and osteoblasts is the pre-requisition for the maintenance of bone mass (or bone density). The deregulation and misfunction of either osteoblast or osteoclast can lead to including various bone pathologies. Osteoporosis is an osteolytic disease where bone is highly susceptible to fracture. This bone disease occurs when the bone formation by osteoblast cannot compensate for the bone destruction by osteoclast.65  Figure 1.7 Schematic diagram of bone rebuilding. Resorption phase: osteoclasts destroy and remove fractured bones. Formation phase: osteoblasts help to form new bone matrix. Alendronate sodium (ALN), brand name Fosamax, is the second generation of bisphosphate drugs for the treatment of osteoporosis. After incorporated into bone matrix, ALN inhibits the osteoclastic bone destruction while activate the osteoblastic bone building. Therefore, the process of bone remodeling gets corrected and the bone density gets enhance.66, 67  Due to the low bioavailability (~0.75%) of ALN in human, pharmacokinetic studies require highly sensitive analytical techniques for measurement of ALN. The high polarity and lack of chromophores of ALN present challenges for ALN quantitative analysis. Recent 14  strategies either exhibit poor sensitivity (e.g. indirect UV detection68 and conductivity detection69) or require tedious sample derivatization (e.g. fluorescence detection70-74 and mass spectrometry75-77).  Therefore, the study in Chapter 3 is aimed to develop and validate a fast and highly sensitive CE-MS method assisted with pH barrage stacking technique for ALN determination. 15  Chapter 2: Simultaneous determination of imatinib mesylate and related compounds by capillary electrophoresis mass spectrometry with field assisted sample stacking 2.1 Introduction Imatinib (IM), brand named Gleevec, is the first small-molecule tyrosine kinase inhibitor (TKIs) for the treatment of chronic myeloid leukemia found by Novartis in Switzerland.55, 56 Despite the anti-tumor efficacy of IM, some patients fail to respond to IM treatment and present drug resistance. In fact, different food intake, environmental condition and co-mediation among patients can result in the inter-individual variability of response to IM treatment. Studies found a large variability of plasmatic IM concentration among those drug-efficacy and drug-resistance patients. Researchers proposed that an adequate plasmatic level of IM is closely related to an effective IM therapeutic response.57-59 Therefore, the measurement of plasmatic IM level is a quite important indicator for monitoring therapy and individualizing the IM dosage for each cancer patient in clinical practice.78 It is of high demand to develop a sensitive and robust method for IM assay. The main metabolite of IM is N-desmethyl imatinib (NDI) as shown in Figure 2.1 (b). NDI is also one impurity/by-product of IM drug produced during the drug synthesis process.79 Additionally, IM is easy to hydrolyze in acidic or basic environment, converted to drug impurities PYA (N-(5-amino-2-methylphenyl)-4-(3-pyridyl)-2-pyrimidinamine). The molecular structure of PYA is represented Figure 2.1(c).80  16   Figure 2.1 Molecular structures of IM, NDI, PYA and d8 IM. Until now, researches about the separation and quantitative analysis of IM and its metabolite have been studied on GC-MS 81, LC-MS60, 61, HPLC-MS62. Applications of CE are mainly focused on CE-UV for IM drug analysis and plasmatic sample analysis.14, 55, 56, 63, 64 Capillary electrophoresis is an efficient separation technology possessing advantages of low sample volume requirement and versatility for different compounds, which make CE an alternative and complementary method to chromatography. In this study, CE-ESI-MS method was developed and validated as an alternative, robust and sensitive approach to simultaneously quantify IM and relevant species in serum. In order to improve sensitivity, field assisted sample stacking (FASS) method was also adopted during electrophoresis separation. 17  2.2 Materials and methods 2.2.1 Chemicals and Reagents NDI, PYA and d8 IM were all purchased from Toronto Research Chemicals (Toronto, ON, CA). Methanol (MeOH), acetonitrile (ACN), formic acid (FA), tertiary butyl methyl ether (TBME) and acetone were purchased from Thermo Fisher Scientific (Nepean, ON, CA). Analytical or better grade sodium hydroxide (NaOH), concentrated hydrogen chloride (HCl) and ammonium formate (HCOONH4) were purchased from Sigma Aldrich (Oakville, ON, CA). All solution was prepared with deionized water (18.2 kΩ•m-1).  Bovine serum mixed with cell culture medium were used to simulate the condition of human plasma which is not available for us. Dulbecco's Modified Eagle cell culture media and bovine serum albumin were purchased from Thermo Fisher Scientific (Nepean, ON, CA). 2.2.2 Apparatus and instruments Electrophoresis experiments were conducted on a PA800+ system (Beckman Coulter Inc., Fullerton, CA, USA) coupled to an API 4000 triple quadrupole mass spectrometer (AB SCIEX, Concord, ON, CA). 32 Karat software was used to control electrophoresis procedures. Analyst software was used to control and optimize MS parameters. Uncoated fused-silica capillaries with 50-µm i.d. and 75-µm i.d. (360 µm o.d.) both were purchased form Polymicro Technologies (Phoenix, AZ, USA). The syringe pump (Harvard apparatus, MA, USA) was used to control the flow rate of modifier solution. High-purity water (18.2 kΩ) for preparation of all aqueous solution was obtained from Milli-Q system (Millipore, Billerica, MA, USA). Φ350 pH meter was purchased from Beckman Coulter Inc (Fullerton, CA, USA). Centrifuge 5415D and Vacufuge instrument were supplied by Eppendorf Canada (Mississauga, ON, CA). 18  2.2.3 Standards Stock solutions of IM, NDI, and with a concentration of 1 mg/mL were prepared by dissolve certain amounts of analytes in methanol. A series of solutions were made by diluting the stock solution with methanol. In order to prepare the calibration solution and the quality control solution, the stock solution was added into drug-free bovine serum with volumes not exceeding 5% of serum volume. A 0.5 mg of deuterated-8 imatinib mesylate (d8 IM) solid was dissolved in 1 mL of methanol and further diluted into 25µg/mL concentration. A 5 µL of this prepared solution was added to 0.5 mL of serum sample serving as internal standard. This identical amount of deuterated-8 imatinib was used for all serum samples throughout the whole study. All solutions were stored at -20 ℃. 2.2.4 Sample preparation Serum samples were stored in the freezer at -20℃ and allowed to thaw at room temperature before sample pretreatment. A 5 µL aliquot of 25 µg/mL internal standard solution was added to 0.5 mL of sample serving as internal standard. A 20 µL of 1.0 M NaOH solution was added in order to adjust the pH of sample and make IM neutral. 0.8 mL of tert-butyl methyl ether (TBME) was added, and the polypropylene tube was vortexed for 2 minutes. Then the samples were centrifuged for 5 min at 6000 g and supernatant was collected. The aqueous part was extracted with 4.0 mL of TBME again. Those supernatants were dried under vacuum in Vacufuge machine. Then samples were constituted with 50 µL of specific solvent and ready for injection into CE instrument. 19  2.2.5 Capillary electrophoresis-MS procedure An uncoated fused-silica capillary (64 cm total length × 50 µm i.d.) served as the separation capillary. Initially, the capillary was flushed with 0.1 M HCl, deionized H2O, 0.1 M NaOH, deionized H2O and BGE solution in turn. Each flushing lasted for 10 min. To obtain high reproducibility, the capillary was rinsed and equilibrated with BGE solution for 5 min before each sample run. Sample solutions got injected over 5 s at a pressure of 7.5 psi. The separation voltage was set at 30 kV. To introduce separated analytes at the outlet of capillary into mass spectrometer, a micro-through ESI interface developed in our lab52 was used throughout the experiments. Except separation capillary, another bare fused-silica capillary (75 µm i.d.) was used to transfer modifier solution composed of 1% formic acid, 75% methanol and 24% deionized H2O. The flow rate of modifier solution was controlled by a syringe pump at 1.0 µL/min with a 250-µL HPLC syringe. Mass spectrometer was operated in positive ion mode with an + 4.5 kV ESI voltage. Initially, MS parameters for multiple reaction mode (MRM) were optimized by directly injecting analytes into mass spectrometer through the syringe pump.  2.3 Results and discussion 2.3.1 MRM parameters The MS instrument was operated in multiple reaction monitoring (MRM) scanning mode. The tuning was performed by direct injection of 100 µg/mL of IM, PYA, NDI and d8 IM dissolved in modifier solutions respectively.  The representative collision-induced-dissociation (CID) spectra and fragments structures for four analytes were illustrated in Figure 2.2. Peaks corresponding to the m/z transition 494.4 → 394.2 represent IM, 480.4 → 394.2 for NDI, 277.9 → 262.2 for PYA and 502.8 → 394.2 for d8 IM. It suggested that NDI, IM and d8 IM have the same product-ions signal at m/z 394.2. 20      21    Figure 2.2 Representative Q1/Q2 spectra obtained for IM(a), NDI(b), PYA(c) and d8 IM(d); Structures of precursor and fragments for IM(a), NDI(b), PYA(c) and d8 IM(d).  22  Analytes Q1 Q2 DP (volts) CE (volts) PYA 277.9 262.2 100 50 NDI 480.4 394.2 105 36 IM 494.4 394.2 100 32 d8 IM 502.8 394.2 113 38 Table 2.1 Summary of MRM parameters for PYA, NDI, IM and d8 IM Other parameters, including collision energy (CE), the declustering potential (DP) of studied compounds were tuned as shown in Table 2.1 in order to get sensitive signals. 2.3.2 Optimization of CZE conditions 2.3.2.1 Influence of BGE pH on separation  Figure 2.3 Influence of buffer pH on separation. Buffer: 10mM HCOONH4 dissolved in 75%MeOH and 25%H2O. Sample was injected at 5 psi for 5 s. Note that IM and NDI were hard to separate in an aqueous system due to the similarity of molecular structure. To address this issue, a non-aqueous system (BGE: a mixture of 75% MeOH and 25% H2O) was used in our study.  23  In CZE, apparent mobilities of analytes were co-regulated by electrophoretic mobility and EOF. pH of BGE solution greatly impacted the ionization of analytes as well as silanol groups on capillary inner wall, resulting in different migration time and separation efficiency. The pH effect was investigated from 1.8-5.1 using formic acid to adjust pH of BGE solution. As shown in Figure 2.3, the migration time of NDI, IM and PYA all increased from pH 1.4 to pH 3.2. However, the migration time of IM and NDI drop gradually while that of PYA still increased above pH of 3.2. The different migration behaviors can be attributed to the different predominant role of µep and EOF. Below pH 3.2, the EOF of the bulk solution got suppressed and µep dominate the movement properties of analytes. In the pH range from 1.4 to 3.2, the restrained protonation of three analytes at higher pH mainly contributed to lower apparent mobilities, resulting in longer migration time as pH increased. Above pH 3.2, the formation of double layer led to an increase of EOF at higher pH. For IM and NDI, increasing EOF played dominant role in apparent velocity at higher pH, leading to a shorter migration time from pH 3.2 to pH 5.1. On the contrary, the effect of deprotonation of PYA (lower µep) surpassed the effect of increasing EOF, resulting in a longer migration time.  In addition, the resolution between IM and NDI is greatest at pH 2.4. Considering the separation resolution and operation efficiency, the pH of 2.4 was used afterwards.   24  2.3.2.2 Influence of BGE concentration (ammonium formate) on separation  Figure 2.4 Influence of different concentration of ammonium formate (HCOONH4) on separation. Buffer: 75%MeOH 25%H2O, pH=2.4. Sample injected at 5psi for 5s The concentration of BGE solution determines the ionic strength in running buffer and influences the zeta potential of double layer, thus changing the EOF. Concentrations of ammonium formate varying from 5 to 30 mM were studied as illustrated in Figure 2.4. It showed that the migration time of three target analytes increased as the concentration increased gradually. Considering the short separation time, 5 mM of ammonium formate was applied to the rest of experiments. As a summary, a mixture of MeOH and H2O containing 5 mM ammonium formate with pH 2.4 was chosen because it gave the shortest separation time.  25  2.3.3 FASS Online concentration  Figure 2.5 FASS process of IM, NDI and PYA. Positively charged IM, NDI and PYA (a) migrate faster towards cathode in sample segment, (b) slow down upon entering BGE, (c) stack up at the boundary, (d) focused ions are subject to separation under CZE mode. Unlike normal CZE, sample was prepared in a series of non-aqueous solvent instead of background electrolyte (BGE) in this study. Since non-aqueous solvents have less conductivity (higher resistivity) than BGE, the electric field strength is greater in sample zone than BGE zone. The online stacking process was depicted in Figure 2.5. Initially, positive-charged PYA, NDI, IM migrated faster in sample zone and slowed down upon entering BGE solution. Thereby, the length of ions zone was shortened by a factor of g, and ions concentration were amplified by a factor of g at the concentration boundary (γ refers to the ratio of resistivity of sample solvent and BGE). Afterwards, stacked analytes were subject to separation under CZE mode before exiting the outlet of capillary. The type of sample solvent and injection time were optimized as follow respectively. 26  2.3.3.1 Influence of different sample solvent  Figure 2.6 Influence of different sample solvent on stacking efficiency. BGE: 5 mM HCOONH4 in 75%MeOH and 25% H2O at pH 2.4. Sample injected at 5 psi for 5 s. Different solvents possessed different resistivities and viscosities, affecting the electrophoretic separation of analytes. To maximize the stacking effect, a series of non-aqueous solvent – methanol, acetonitrile and acetone, were studied and used to dissolve samples. During the experiments, 20% H2O was added to each non-aqueous solvent to maintain stable separation current. From Figure 2.6, MS signal of three analytes are all stronger in non-aqueous solvent than those dissolved in BGE solution. Additionally, the strongest signals were observed when acetonitrile (ACN)-water served as sample solvent, which was attributed to the maximum resistivity difference between acetonitrile solvent system and BGE solution. Therefore, a mixture of 80% acetonitrile and 20% H2O was adopted to dilute sample before loading into CE-MS system. 27  2.3.3.2 Influence of the injection time  Figure 2.7 Influence of different injection pressure on stacking efficiency, given injection time of 5s. BGE: 5 mM HCOONH4 in 75%MeOH and 25% H2O at pH 2.4.  With fixed separation conditions, the sample injection volume is proportional to the injection pressure and time when hydrodynamic injection method was employed. In order to improve the detection sensitivity as well as maintaining appropriate separation efficiency, injection volume was optimized. When injection time remained at 5 s, the injection pressure varied from 1 psi to 9.5 psi. As shown in Figure 2.7, peak intensities increased correspondingly with higher injection pressure. However, distorted and wider peaks were observed when the injection pressure increased beyond 7.5 psi, destroying the separation between three analytes. This phenomenon was attributed to two reasons: (1) Too large injection volume led to unstable separation current and poor separation since the conductivity of non-aqueous solvent was greatly smaller than BGE 82; (2) A laminar flow generated by the concentration difference in sample solvent and BGE broadened the length of focused ions.4, 83 As a result, 7.5 psi was chosen as the optimal injection pressure so as to not exceed the maximum online-stacking capacity of our proposed method. 28   Figure 2.8 Representative electropherogram obtained with (a) normal injection at 1 psi for 5 s; (b) FASS injected at 7.5 psi for 5 s. Sample contains 50 µg/mL PYA, 10 µg/mL NDI and 10 µg/mL IM. Experiments found that the sample volume was around 7 nL with normal injection at 1 psi for 5 s and 56 nL with FASS injected at 7.5 psi for 5 s. Figure 2.8 compared mass spectra with/without FASS. It’s obvious that signal intensities of each analytes were amplified by 8-fold assisted with the FASS technique.  2.3.4 Extraction efficiency of liquid-liquid-extraction TBME was adopted as extraction solvent to recover imatinib and related two analytes from serum. Note that it’s necessary to adjust the pH of serum samples to pH 13 prior to the addition of TBME. IM, NDI and PYA were all deprotonated and the decreased polarity under strong basic condition helped to improve extraction efficiency. As a result, recovery rates were 88%, 83% and 90% for PYA, NDI and IM respectively. 29  2.3.5 Linearity and limit of detection    Figure 2.9 Calibration curves for PYA(a), NDI (b) and IM (c).   30  Linear relationships between the relative peak areas (analyte peak areas/d8 IM peak area) and concentration of analytes were established in the concentration range of 40-2000 ng/mL for PYA, 6-400 ng/mL for NDI, 3-500 ng/mL for IM. The regression curves were shown in Figure 2.9.  Analytes (N=3) Regression equation R2 Linear range (ng/mL) LOD (ng/mL) PYA y=0.0032x-0.0057 0.9954 40 – 2000 4.0 NDI y=0.0267x+0.0492 0.9991 6 – 400 0.4 IM y=0.0758x+0.0259 0.9999 3 – 500 0.2 Table 2.2 Results of calibration curve for PYA, NID and IM Table 2.2 summarized results of calibration curves for 3 analytes. The regression equations were y=0.0032x-0.0057 (R2=0.9954) for PYA, y=0.0267x+0.0492 (R2=0.9991) for NDI, y=0.0758x+0.0259 (R2=0.9999) for IM. All three calibrations curve exhibited excellent linear relationships. For good linearity, the lower limit of quantification (LLOQ) were chosen as 40, 6, 3 ng/mL for PYA, NDI and IM respectively. The limit of detection was 4.0, 0.4, 0.2 ng/mL for PYA, NDI and IM respectively, which were satisfactory for analysis of both drug formula and human plasmatic sample according to previously reported results.14, 61, 79 LOD was determined as the lowest concentration of a target species giving a noticeable response (S/N=3). 31  2.3.6 Method validation 2.3.6.1 Specificity   32   Figure 2.10  Electropherograms of (a) blank serum sample without d8 IM, (b) blank serum sample spiked with d8 IM and (c) LLOQ sample. The specificity of this assay method was tested with (a) a blank serum sample, (b) a blank serum sample spiked with only internal standard d8 IM as well as (c) LLOQ sample. Electropherograms of three types of extractants were shown in Figure 2.10.  From the Figure 2.10 (a), no peaks were seen at the corresponding retention time of PYA, NDI, IM and d8 IM in the blank serum, indicating that no endogenous interference with three analytes and internal standard in drug-free serum. In addition, Figure 2.10 (b) demonstrated that the internal standard (d8 IM) did not influence the signal of either PYA, NDI or IM.  2.3.6.2 Intra-day assay The quality control samples were at four different concentration levels including LLOQ (40, 6, 3 ng/mL for PYA, NDI and IM) and low (120, 18, 9 ng/mL for PYA, NDI and IM), medium (700, 160, 175 ng/mL for PYA, NDI and IM) and high concentration (1500, 300 and 375 ng/mL for PYA, NDI and IM). Those samples were prepared by spiking specific amounts of 33  analytes into drug-free serum. The measured concentration was back-calculated in terms of calibration curve. The precision was defined by the relative standard deviation of the measured concentration for each quality control sample. The accuracy was represented by bias which was the percentage discrepancy between measured concentration and added concentration as shown in Equation 2.1. The bias should not exceed 10% for all QC samples except for LLOQ sample (<20%) according to the guideline on bioanalytical method validation issued by European Medicines Agency. Bias (%) = (Concentration measured – Concentration added)/ Concentration added * 100% Equation 2.1 The calculation formula of bias for accuracy test.   Analytes (N=5) Concentration (ng/mL)  R.S.D.  Bias  Added Measured  PYA 40 43 8.1% 7.1% 120 120 6.2% -0.8% 700 727 2.0% 3.9% 1500 1559 1.3% 3.9%  NDI 6 5 6.4% -11% 18 17 7.2% -5% 160 170 3.2% 6% 300 325 9.1% 8%  IM 3 3 4.4% -5% 9 9 6.3% -2% 175 178 1.3% 1.8% 375 357 1.1% -4.7% * N=number of determinations within a single day Table 2.3 The precision and accuracy of intra-day results. 34  The intra-day precision and accuracy of this method were validated with drug-spiked serum sample within the same day (N=5). Data shown in Table 2.3 illustrated the intra-day precision was better than 9% and the inaccuracy was within ±11% for all QC samples, which were both satisfactory. 2.3.6.3 Inter-day assay  Analytes (N=3) Concentration (ng/mL)  R.S.D.  Bias  Added Measured  PYA 40 43 0.7% 6.2% 120 116 5.0% -3.0% 700 716 2.1% 2.3% 1500 1525 2.2% 1.7%  NDI 6 5 1.3% -11% 18 19 8.5% 4.6% 160 173 1.3% 7.8% 300 314 4.4% 4.7%  IM 3 3 3.9% -2.4% 9 9 3.9% 2.7% 175 183 2.4% 4.6% 375 365 4.1% -2.6% * N=number of days evaluated Table 2.4 The precision and accuracy of inter-day test.  The inter-day precision and accuracy of assay were evaluated by testing LLOQ, low, medium and high QC samples over 3 consecutive days. Results were summarized in Table 2.4. The precision was better than 8.5% and the bias did not exceed ±11% at all levels, which were both acceptable.  35  2.3.6.4 Stability study The stabilities of IM and related substances in serum were evaluated with quality control samples at two concentration level (100 and 1500 ng/mL for PYA, 15 and 300 ng/mL for NDI, 9 and 375 ng/mL for IM) experiencing different storage conditions: (1) after 3 freeze-thaw cycles (2) after 8 hours at room temperature as well as (3) after 2 weeks at -20 ℃. The stabilities were validated by the relative recovery calculated from Equation 2.2. The relative recovery values obtained should within ±15% of nominal concentration for all QC samples according to the guideline on bioanalytical method validation issued by European Medicines Agency. Relative recovery (%) = Concentration measured / Concentration nominal * 100% Equation 2.2 Calculation formula of relative recovery for stability test. 2.3.6.4.1 Freeze-thaw stability  Analytes Nominal concentration (ng/mL) Measured concentration (ng/mL) Relative recovery PYA 100 104 104% 1500 1557 104% NDI 15 16 109% 300 309 103% IM 9 9 101% 375 346 92% Table 2.5 Stabilities of three analytes after 3 freeze-and-thaw cycles. QC samples were stored and frozen at -20℃ and subject to 3 freeze-and-thaw cycles. For each concentration level, 0.5 mL aliquots in duplicate were studied, and obtained results were averaged as shown in Table 2.5. The mean concentrations recovered at each level were 92%-109% of nominal values, indicating no significant degradation were observed after 3 freeze-and-thaw cycles.  36  2.3.6.4.2 Short- and long-term stability Low and high QC samples were processed and analyzed initially. Then the processed samples were left at room temperature (25 ℃) for 8 hours and analyzed again. The recovered concentrations were indicator of short-term stabilities of shown in Table 2.6-2.8. Measured values suggested three analytes were stable at ambient condition for at least 8 hours.  Long-term stability was evaluated where QC samples in serum were preserved in the freezer at -20℃ for 1 week and 2 weeks. Samples were then analyzed following the same pre-treatment and analysis procedures in 2.2.4. The percentage of recovered analytes ranged from 92-108% for PYA, from 97% to 109% for NDI and from 92% to 103% for IM. The measured concentrations of each level were well within ±15% of the nominal concentration. Hence, all analytes were stable at -20 ℃ for at least 2 weeks. Storage period & condition Nominal concentration (ng/mL) Measured concentration (ng/mL) Relative recovery  0 h, 25 ℃ 100 106 106%  1500 1552 104% 8 h, 25 ℃ 100 108 108%  1500 1608 107% 1 weeks, -20 ℃ 100 91 91%  1500 1600 107% 2 weeks, -20 ℃ 100 96 96%  1500 1587 106% Table 2.6 Stability test of PYA. Storage period & condition Nominal concentration (ng/mL) Measured concentration (ng/mL) Relative recovery  0 h, 25 ℃ 15 16 107% 37   300 317 106% 8 h, 25 ℃ 15 108 104%  300 16 98% 1 weeks, -20 ℃ 15 15 97%  300 328 110% 2 weeks, -20 ℃ 15 16 108%  300 317 106% Table 2.7 Stability test of NDI. Storage period & condition Nominal concentration (ng/mL) Measured concentration (ng/mL) Relative recovery  0 h, 25 ℃ 9 9 103%  375 355 95% 8 h, 25 ℃ 9 9 102%  375 358 95% 1 weeks, -20 ℃ 9 9 101%  375 346 92% 2 weeks, -20 ℃ 9 9 102%  375 351 94% Table 2.8 Stability test of IM. 2.4 Conclusion In this study, simultaneous separation and determination of IM, related metabolite (NDI) and drug impurity (PYA) were realized by CE-MS assisted with field assisted sample stacking (FASS) technique. Compared to normal CZE, three analytes signal intensity was amplified by 8-fold and LOD achieved 0.2 ng/mL for imatinib. Validation test suggested this FASS-CE-MS method has wide linearity ranges, satisfactory specificities, acceptable precisions and accuracies. Considering high detection sensitivity, simple sample pretreatment and fast analysis speed, FASS-CE-MS is suitable for drug administrations and clinical applications. 38  Chapter 3: Quantitative determination of a bisphosphate drug - alendronate sodium by CE-MS assisted with acid barrage stacking technique 3.1 Introduction Alendronate sodium (ALN), brand name Fosamax, is the second generation of an amino bisphosphate drug for various bone disease -- osteoporosis, Paget’s diseases, bone metastases induced by breast cancer and myeloma.66, 67  Figure 3.1 Molecular structures of (a)ALN and (b)PAMLack of chromophore and fluorophore groups and high polarity are main structural characteristics of ALN (shown in Fig.3.1 (a)), presenting great challenges for sample extraction and quantitative analysis. Recent researches reported various strategies to analyze ALN directly or indirectly (ALN-derivatives). However, both of them showed their disadvantages for routine application. Direct strategies (e.g. Ultra-Violet detection and conductivity detection) usually exhibited poor sensitivity, while more sensitive strategies such as laser-induced fluorescence (LIF)70-74 required tedious labelling process. Aiming at the introduction of fluorescent groups, fluorescent tags such as 3-(2-furoyl)-quinoline-2-carboxaldehyde (FQ), 9-fluorenylmethyl derivative (FMOC)71 and fluorescamine72, were commonly used to pre-column derivatization with amine group of ALN. But LIF has several limitations for routine analysis of ALN: (1) long derivatization reaction time (2) introduction of toxic chemical (e.g. KCN) for fluorescent reaction (3) poor stability of fluorescent tags or amine-derivative products.  39  Mass spectrometry (MS)75-77 is an another highly sensitive strategy. Although fluorescent tags were not necessary with MS-based assay methods (e.g. GC-MS, HPLC-MS), derivatization procedures were still indispensable in order to realized sample extraction avoiding suppressing effect of salt content on MS signal. Commonly used reagents such as diazomethane76, 84 and trimethyl orthoacetate (TMOA) 77 reacted with bisphosphate group of ALN forming non-polar compounds. Again, safety concerns and tedious sample pre-treatment were the main issues about those reported method. Hence, it’s of highly demand to improve detection limit as well as simplify sample pre-treatment for the determination of ALN. Until now, only one direct CE-MS method was reported for detection of amino bisphosphate drug without derivatization, which illustrated poor sensitivity (~80 µg/mL for clodronate).85  The aim of this study was to develop a fast and sensitive electrophoretic method for direct ALN analysis. Underivatized ALN was directly introduced to CE-MS system and an acid barrage stacking (ABS) techniques was adopted to focus target species and enhance sensitivity of ALN signal. The developed ABS-CE-MS strategy was successfully applied for determination of ALN in tablets and serum samples. As a comparison, quantitative analysis was performed with conventional CE-LIF method as well which introduce fluorogenic ALN derivatives into CE system. Comparison details were listed in Appendix.  3.2 Experimental  3.2.1 Chemicals and reagents High-purity water for preparation of all aqueous solution was obtained from Milli-Q system (Millipore, Billerica, MA, USA). Analytical or better grade sodium hydroxide (NaOH), concentrated hydrogen chloride water solution (HCl), sodium monophosphate (Na2HPO4), sodium acetate (CH3COONa), formic acid, methanol (MeOH) and acetonitrile (ACN) was purchased from Thermo Fisher Scientific (Nepean, ON, CA). Alendronate sodium trihydrate 40  (C4H12NaNO7P2·3H2O) and pamidronate disodium pentahydrate(C3H9NNa2O7P2·5H2O) was purchase from Sigma-Aldrich (Oakville, ON, CA). ALN tablets (CSPC Ouyi Pharmaceutical Co., Ltd., Shijiazhuang, CN) were purchased from a drug store in Leshan, China. Bond Elut DEA SPE cartridges (50 mg/1 mL) were purchased from Agilent Technologies Inc. (Mississauga, ON, CA) Bovine serum mixed with cell culture medium were used to simulate the condition of human plasma which was not available for us. Dulbecco's Modified Eagle cell culture media and bovine serum albumin were purchased from Thermo Fisher Scientific (Nepean, ON, CA). Fused-silica capillaries were purchased from Polymicro Technologies (Phoenix, AZ).  3.2.2 Apparatus and instruments The same CE-MS system described in Chapter 2 was used. 3.2.3 Standard solution preparation Stock solution of ALN (1 mg/mL) and PAM (1 mg/mL) were prepared by dissolving C4H12NaNO7P2·3H2O and C3H9NNa2O7P2·5H2O in deionized water respectively. The standard solution and quality control solutions were prepared by further diluting stock solution by different dilution factor. 3.2.4 Drug tablets and serum sample pre-treatment One ALN drug tablet was weighted, powered and extracted with 100 mL distilled water in an ultrasonic bath for 5h. Prior to derivatization, the extractant was filtered through a 0.45 µm membrane filter and further diluted 200 times with background electrolyte (BGE).   Serum samples were pre-treated following the widely-used Ca2+ precipitation method.70-74 1 mL of serum sample was pre-treated with 0.5 mL of acetonitrile to denature proteins and centrifuged for 10 min at 6000 g. The supernatant was collected. 110 µL of 0.1 M KH2PO4, 110 µL of 0.1 M CaCl2 and 200 µl 1 M NaOH were added into the supernatant to form ALN-Ca 41  precipitate. Then the solution was centrifuged at 2000 g over 3 min. The ALN-Ca precipitate was re-dissolved in 1 mL of 0.01 M sodium acetate (pH=4) and purified on DEA SPE cartridge. The cartridge was first activated with methanol and H2O sequentially, and then pre-conditioned with 1mL of 0.01 M sodium acetate (pH=4) and dried for 2 min. The re-dissolved sample was loaded onto DEA cartridge and eluted by 1 mL of 0.2 M Na2HPO4. The elute was subject to following CE-MS analysis. 3.2.5 CE-MS procedure An uncoated fused-silica 120-cm capillary with a 50-μm inner diameter was used for separation of bisphosphate drugs. First, the capillary was flushed with HCl, H2O, NaOH, H2O and BGE solution sequentially. Sample solutions got injected over 40 s at a pressure of 30 psi. Then, acid barrage solution was loaded over 40 s at a pressure of 30 psi. The separation voltage was set at 30 kV with a pressure of 1 psi.   The separated analytes were introduced into MS using the same interface as that in Chapter 2. The modifier reservoir contains a mixture of 75% MeOH and 25% H2O with 0.2 M FA. 3.3 Results and discussions 3.3.1 MRM parameters To compensate the possible loss during sample pretreatments and variation of ionization efficiency in MS, an internal standard of known concentration was added into initial samples. Since isotopic ALN was not available to our lab, PAM was chosen as the replacement of the deuterated. As displayed in Figure 3.1, PAM molecular structure contained one less -CH2 group than ALN making PAM have similar physiochemical properties to ALN. Thus, PAM is suitable to normalize ALN signals. 42  The existence of an amino group helped both ALN and PAM to carry positive charge when in an acidic environment. This property made it possible for them to enter the mass spectrometer under positive ESI mode. In practice, formic acid of 0.2 M concentration was added into the modifier reservoir to guarantee ESI efficiency of ALN and PAM. The tuning was performed by direct injection of ALN and PAM (20 µg/mL) dissolved in the modifier reservoir.   Figure 3.2 Representative Q1/Q2 mass spectrum of ALN and PAM. Collected Q1 MS spectra are shown in Figure 3.2. Dominant ions have m/z 250.3 for ALN while m/z 236.3 for PAM. These masses correspond to protonated ALN and PAM. In addition, other stronger peaks appearing at m/z 272.7 and 258.6 were sodium adduct for ALN and PAM respectively. Hence, parent ions with m/z 250.3 and 236.3 were selected for ALN and 43  PAM respectively. According to Q2 MS spectra above, the most intense product ions were detected at m/z 150.1 for ALN and m/z 154.2 for PAM. Declustering potential (DP) and collision energy (CE) were tuned as well. Finally, quantitative determinations were performed in multiple reaction monitoring (MRM) scanning mode using the following m/z transitions: 250.3 → 150.1 for ALN, and 236.3 → 154.0 for PAM. The collision energy (CE) was set at 21 V (ALN) and 20 V (PAM), and the declustering potential (DP) was kept at 60 V for both analytes. 3.3.2 Acid barrage stacking with large volume injection  Figure 3.3 Schematics of acid barrage stacking (ABS) process for ALN. (a) negative ALN migrate towards anode, (b) ALN are protonated and become neutral upon entering the acid barrage, (c) ALN  stack up at the boundary, (d) focused analytes are subject to separation under CZE mode. Figure 3.3 illustrated the schematics of acid barrage stacking process for ALN. According to the acid dissociation constant of alendronic acid (shown in Figure 3.4),86 major ALN analytes 44  contained net 2 negative charge in basic BGE (pH=8) due to its pKa2. In addition, ALN became apparently neutral in acidic pH barrage zone (pH=1.8) due to its pKa1. Thus, with the onset of voltage, negatively charged ALN first migrated towards cathode and then stopped moving upon entering acid barrage. Online stacking of ALN analyte was realized at the boundary between sample zone and acidic pH barrage. After the disappearance of acid barrage, focused analytes were separated under CZE mode before reaching detector.   Figure 3.4 Acid dissociation constants of ALN. The ratio of the length of sample segment to pH barrage depends on the mobility of analytes. Shoulder peaks might show up in uncomplete focusing. Usually, larger sample volume requires longer pH barrage length to avoid any loss of analytes under the condition of baseline separation.20 In this study, the length of pH barrage was optimized with a given sample volume as follows. The total length of separation capillary was 120 cm. 45  3.3.2.1 Influence of pH barrage length   Figure 3.5 Influence of injection time of acid barrage at 30 psi pressure on sample stacking efficiency. Standard solution with 200 ng/mL ALN was used.  The sample segment length was kept at 36.3% of effective capillary length (injected at 30 psi over 40s). The acidic barrage length was optimized and controlled by the injection time under 30 psi pressure. The injection period ranged from 10 s to 50 s. As shown in Figure 3.5, a longer pH barrage length resulted in shorter migration time, higher peak heights and narrower peak width. The decreased migration time was attributed to the fact that the load of acid barrage pushes sample segment closer to outlet of capillary and MS detector. And a longer acid barrage meant a longer stopping barrier and resulted in better focusing effect with amplified signal intensity. Considering that signal intensities were stable when injection period longer than 40 s, 40 s was chosen afterwards. The sample segment and the pH barrage took up ca. 36.3% of whole capillary length separately. 0 2 4 6 8 10040080004000080000040000800000400008000004000080000 Time (min) 10 s  20 s Intensity (cps) 30 s  40 s   50 s46  3.3.2.2 Sensitivity comparison between conventional CZE-MS and ABS-CE-MS.  Figure 3.6 Sensitivity comparison between (a) conventional CZE-MS where 2 µg/mL ALN was injected at 2psi for 10s and (b) ABS-CE-MS, where 2 µg/mL ALN was injected at 30 psi for 40s followed by a plug of acid barrage. Capillary electrophoresis separations were conducted with 30 kV voltage with 1.0 psi pressure.  Under conventional CZE, sample was injected at 2 psi over 10 s (~10 nL). The LOD obtained for ALN was around 2 µg/mL. With the introduction of an extra acid barrage, the volume of sample increased up to 0.85 µl (injected at 30 psi over 40 s) without peak distortion. As shown in Figure 3.6, signal intensity was enhanced by approximate 810-fold in terms of peak heights. 47  3.3.3 Influence of BGE pH and concentration on separation          Figure 3.7 Influence of BGE (a) pH and (b) concentration on separation.  The influence of BGE pH on the migration of ALN was studied in the range from pH 5.0 to pH 8.0. With increasing pH, the apparent mobility of ALN increased gradually due to higher EOF (shown in Figure 3.7(a)). In order to realize shortest analysis time, BGE with pH 8.0 was chosen to perform the separation.  In addition, the concentration of HCOONH4 was investigated ranging from 5 to 20 mM.  Figure 3.7(b) demonstrated that small changes in the ALN migration time were observed with different BGE concentration. Considering concentration above 10 mM can deteriorate signal intensities, HCOONH4 concentration of 5 mM was adopted.  48  3.3.4 Influence of modifier flow rate  Figure 3.8 Modifier flow rate effect on the signal intensity of ALN with 200 ng/mL concentration. The proper performance of ESI process relied on the modifier’s composition and flow rate. A high methanol content (75% MeOH) of modifier reservoir was chosen since the low surface tension of methanol helped to promote analyte’s electrospray ionization efficiency.  Modifier flow rate would impact the stability and sensitivity of MS signal. Flow rates between 0.5 and 2.0 μl/min were investigated. From Figure 3.8, the strongest signal was obtained at 1.0 μl/min. ALN signal intensity was quite low when flow rate was 0.5 μl/min due to the poor ESI efficiency. On the other hand, the dilution factor of the modifier at too high flow rate decrease the signal intensity. We therefore selected a rate of 1.0 μl/min for subsequent experiments so as to obtain best signal sensitivity and stability. Under the optimized conditions, the whole analysis was completed in less than 6 min at the end. A representative electropherogram for 8 ng/mL ALN and 200 ng/mL PAM was shown in Figure 3.10(d).   49  3.3.5 Calibration curve for ALN in a linear range of 8-2000 ng/mL  Figure 3.9 Calibration curve for ALN using 200 ng/mL PAM as internal standard. Seven standard solutions of ALN in the concentration range of 8-2000 ng/mL were analyzed under the optimized analysis conditions. Each working solution contained 200 ng/mL PAM as internal standard. Working solutions were determined 5 times for each concentration level. Relative peak areas (RPA) of ALN to PAM were used to build a calibration curve (as shown in Figure 3.9). As summarized in Table 3.1, the linear regression equation was y= 2.33623x + 0.00304 (R2=0.99948) for ALN (N=5).  The detection limit (S/N=3) for ALN was found to be as low as 2 ng/mL. The lower limit of quantification (LLOQ) was 8 ng/mL for the sake of good linearity.  Meanwhile, six standard solutions of ALN in the concentration range of 0.6-20 μg/mL were analyzed under the optimized CE-LIF separation conditions. Each working solution contained 10 μg/mL PAM as internal standard. Using the relative peak area of ALN to PAM, the linear equations were 𝑦 =0.2687 𝑥 +0.00864 (R2=0.99983) for ALN. The detection limit (S/N=3) was 0.2 µg/mL for ALN. Experimental details can be found in Appendix. 50  By comparison, the sensitivity of ABS-CE-MS developed by us was 100 times as much as that of CE-LIF method while maintaining excellent correlation coefficient.  Methods Regression equation R2 Linear range  LOD  CE-MS y=2.33623x+0.00304 0.99948 8 - 1500 ng/mL 2 ng/mL CE-LIF y=0.26880x+0.00864 0.99983 0.6 - 20 µg/mL 0.2 µg/mL Table 3.1 A summary of  calirabtion curve for ALN using CE-MS and CE-LIF methods. 3.3.6 Method validation 3.3.6.1 Specificity The specificity of assay was tested with blank solutions spiked with only ALN and PAM as well as LLOQ sample. The obtained electropherograms were shown in Figure 3.10. No peaks were seen at the corresponding migration time of either ALN or PAM for the blank solution in Figure 3.10 (a). In addition, Figure 3.10 (b) and (c) demonstrated that PAM (internal standard) and ALN did not influence each other’s MS signal.   51    Figure 3.10 Electropherograms of (a) blank solution, (b) blank solution spiked with ALN, (c) blank solution spiked with PAM and (d) LLOQ sample. 3.3.6.2 Intra-day assay The precision and accuracy of this method were evaluated by the analyses of at four concentration level – LLOQ (8 ng/mL), low QC (30 ng/mL), medium QC (600 ng/mL) and high concentration QC (1500 ng/mL). The precision was described by relative standard deviation (R.S.D), and the bias was used to represent accuracy which was the percentage difference between measured concentration and the added concentration. Concentration (ng/mL) R.S.D. Bias Added Measured 8 6.9 8.9% -14% 30 27 0.1% -9.8% 600 593 5.5% -1.1% 1500 1523 1.8% 1.5% *Number of determinations N=5 Table 3.2 A summary of intra-day test results. 52  The intra-day tests for each concentration level were conducted in quintuplicate within the same day (N=5). As listed in Table 3.2, R.S.D for relative peak areas were less than 5.5% and bias value were within 10% for three QC samples. Bias for LLOQ was within 15% which was acceptable in terms of the guideline on bioanalytical method validation issued by European Medicines Agency. 3.3.6.3 Inter-day assay Concentration (ng/mL) R.S.D. Bias Added Measured 8 7.9 3.2% -1.9% 30 28 0.1% -5.2% 600 614 2.0% 2.3% 1500 1528 0.5% 1.9% *Number of days N=3 Table 3.3 A summary of inter-day test results. Inter-assay precision and accuracy was evaluated over consecutive 3 days. The data were given in Table 3.3. The R.S.D values were at most 3.2% and the inaccuracy was better than 6% at all levels. Both precision and accuracy were satisfactory.  3.3.7 Analysis of drug tablets composition  The developed CE-MS method was used for the determination of ALN content in commercial drug tablet. One ALN drug tablet was powered and extracted with 100 mL distilled water in an ultrasonic bath for 3 hours. Prior to loaded into the CE system, the extractant was filtered through a 0.45 µm membrane filter and diluted 200 times further with BGE solution.  53   Figure 3.11 Representative electropherograms of commercial ALN tablets (a) without PAM (internal standard) and (b) with PAM. The typical electropherograms for the analysis of samples with/without internal standard were presented in Figure 3.11. Based on the regression equation, the amount of ALN found is 12.99 mg/tablet. Counted as C4H12NNaO7P2∙3H2O, the determination result was close to both the labelled amount of 13.05 mg/tablet and the result of 13.20 mg/tablet through CE-LIF method, indicating the reliability and feasibility of developed ABS-CE-MS method for the pharmaceutical application. Note that ALN drug tablet preparation was the same except that filtered extract was further diluted 20 times in CE-LIF instead of 200 times considering the different linear range. Analysis method Relative peak area Measured concentration (µg/mL) Dilution factor Measured content * (mg/tablet) ABS-CE-MS 1.52 0.649 2000 12.99 CE-LIF 1.78 6.60 200 13.20 *Labeled content is 13.05 mg/tablet (counted as C4H12NNaO7P2∙3H2O)   Table 3.4 A summary of ABS-CE-MS and CE-LIF results for commercial ALN tablets. 54  3.3.8 Serum samples application  Figure 3.12 Electropherogram for extractants of serum samples containing (a) 10 µg/mL ALN and 10 µg/mL PAM (b) foregoing extracted serum sample spiked with 10 µg/mL ALN and 10 µg/mL PAM just before CE-MS analysis The developed CE-MS method was tested for the determination of ALN content in serum. Bisphosphate drugs in serum were recovered by Ca2+ precipitation method and purified by ion exchange SPE cartridge. Prior to loaded into CE setup, the extract was filtered through a 0.45 µm membrane filter. Nevertheless, weak signals were detected for serum samples containing 10 µg/mL ALN and PAM. Even though 10 µg/mL ALN and PAM were spiked into the foregoing extracted serum sample just before CE-MS process, signal intensities was significantly lower compared to results of standard solution (as shown in Figure 3.12). Therefore, the poor detection limit could be caused by the suppression effect of salt (i.e. Na+) introduced during SPE process. High ions concentration not only changed the conductivity of sample zone but also deteriorated focusing process.  55  Other methods, such as salting-out-liquid-liquid-extraction methods, were also tried but with unsatisfactory results. In the future, works will focus on two aspects: (1) other ALN extraction methods to avoid introduction of salt; (2) modification of ABS-CE-MS method to remove concentrated ions in sample zone.  3.4 Conclusion The first direct CE-MS method was developed for underivatized ALN analysis. Assisted with acid barrage stacking techniques, an 810-fold enhancement of sensitivity was achieved and LOD was as low as 2 ng/mL ALN. The operation and analysis of each sample could be finished within 8 min. Possessing simple and sensitive characteristics, this ABS-CE-MS was successfully applied to drug quality investigation in pharmaceutical industry. 56  Chapter 4: Conclusion and future work 4.1 Concluding remarks In this study, two online concentration techniques -- field assisted sample stacking (FASS) and acid barrage stacking (ABS) were incorporated into normal CE-MS for determination of two drugs respectively, including Gleevec (imatinib) for chronic myeloid leukemia treatment and Fosamax (alendronate) for osteoporosis treatment.  FASS was used with CE-MS for the quantification of imatinib along with its metabolite and related impurity in serum. A mixture of acetonitrile and water was used to dissolve sample to enhance the field strength in sample zone. After optimization, three analytes got baseline separation and detection sensitivity was amplified by 8-fold. Using liquid-liquid-extraction for sample pretreatment, LOD achieved 0.2 ng/mL for IM, 0.4 ng/mL for NDI and 4.0 ng/mL for PYA. They showed acceptable linearity between 3-500 ng/mL for IM, 6-400 ng/mL for NDI and 40-2000 ng/mL for PYA. This method was validated with spiked blank serum. The interday R.S.D. was 1% to 9% (N=5), intraday R.S.D. was 0.7%-8.5% (N=3). Recovery was 92%-109%. In addition, deuterated internal standards and three analytes were stable at room temperature for at least 6 h and stable at -20℃ for at least 2 weeks. This method possesses many advantages -- high detection sensitivity, fast analysis speed and simple sample pretreatment, which make it suitable for future daily biological sample analysis.  ABS was applied for determination of ALN in drug tablets by CE-MS. This was the first ABS-CE-MS method for direct ALN analysis. Following sample injection, an extra acid barrage was introduced into separation capillary. The acid barrage worked as a barrier to stop the migration of negatively charged ALN ions and stacked them. After optimization, the detection sensitivity was improved by 810-fold compared to normal sample injection. The limit of 57  detection achieved 2 ng/mL by ABS-CE-MS compared to 200 ng/mL by CE-LIF. Relative peak areas and concentration of ALN showed excellent linearity in the concentration range of 8-2000 ng/mL (R2>0.9995). Good repeatability and accuracy were verified by validation assay. This method was successfully used for quantification of ALN in drug tablet.  4.2 Future work Preliminary works in Chapter 2 and 3 have demonstrated the feasibility of FASS and ABS for biological sample with low concentration. If combine those two online stacking techniques, the sensitivity of new CE-MS strategy can be further improved in the application of trace analysis of diluted biological sample. In addition, although LOD for ALN tablets reached 2 ng/mL, the application of ABS-CE-MS for direct ALN analysis in serum sample need to be addressed. One promising strategy is to combine acid barrage stacking techniques with polarity switching to further enhance sensitivity.87, 88 In this modified technique, a reverse voltage was applied to CE system after injection of sample plug and prior to injection of acid barrage. The reverse electric field will create an EOF towards to the cathode while negative ALN migrate toward the anode. By carefully controlling the velocity of EOF and electrophoretic velocity of ALN, removal of majority sample matrix from target ALN may be realized. In this case, signal intensity can be sufficiently enhanced without the suppressing effect of salt introduced from sample pretreatment. 58  Bibliography 1. Jorgenson, J. W.; Lukacs, K. D., Zone electrophoresis in open-tubular glass capillaries. Analytical Chemistry 1981, 53, 1298-1302. 2. Lu, S.;  Dugan, C. E.; Kennedy, R. T., Microfluidic Chip with Integrated Electrophoretic Immunoassay for Investigating Cell-Cell Interactions. Analytical Chemistry 2018, 90 (8), 5171-5178. 3. Kameoka, J.;  Craighead, H. G.;  Zhang, H. W.; Henion, J., A polymeric microfluidic chip for CE/MS determination of small molecules. Analytical Chemistry 2001, 73, 1935-1941. 4. Chien, R. L.; Burgi, D. S., On-column sample concentration using field amplified in CZE. Analytical Chemistry 1992, 64, 489-496. 5. Oss Van, C. J., Electrokinetic separation methods. Journal of Separation and Purification Methods 1979, 8 (2), 119-198. 6. Hjerten, S., Free zone electrophoresis. Chromatographic Review 1967, 9. 7. Arce, L.;  Nozal, L.;  Simonet, B. M.;  Valcárcel, M.; Ríos, A., Liquid-phase microextraction techniques for simplifying sample treatment in capillary electrophoresis. TrAC Trends in Analytical Chemistry 2009, 28 (7), 842-853. 8. Sarafraz-Yazdi, A.; Amiri, A., Liquid-phase microextraction. TrAC Trends in Analytical Chemistry 2010, 29 (1), 1-14. 9. AlOthman, Z. A.;  Dawod, M.;  Kim, J.; Chung, D. S., Single-drop microextraction as a powerful pretreatment tool for capillary electrophoresis: A review. Analytica Chimica Acta 2012, 739, 14-24. 10. Hernandez-Borges, J.;  Rodriguez-Delgado, M. A.;  Garcia-Montelongo, F. J.; Cifuentes, A., Chiral analysis of pollutants and their metabolites by capillary electromigration methods. Electrophoresis 2005, 26 (20), 3799-813. 59  11. Yang, Q.;  Chen, B.;  He, M.; Hu, B., Sensitive determination of seven triazine herbicide in honey, tomato and environmental water samples by hollow fiber based liquid-liquid-liquid microextraction combined with sweeping micellar electrokinetic capillary chromatography. Talanta 2018, 186, 88-96. 12. Pedersen-Bjergaard, S.; Rasmussen, K. E., Liquid‐phase microextraction and capillary electrophoresis of acidic drugs. Electrophoresis 2000, 21, 579-585. 13. Chang, Y. T.;  Wang, Z. R.; Hsieh, M. M., Sensitive determination of warfarin and its metabolic enantiomers in body fluids via capillary electrophoresis combined with ultrasound-assisted dispersive liquid-liquid microextraction and online sample stacking. Microchemical Journal 2019, 146, 1276-1284. 14. Ahmed, O. S.;  Ladner, Y.;  Montels, J.;  Philibert, L.; Perrin, C., Coupling of salting-out assisted liquid-liquid extraction with on-line stacking for the analysis of tyrosine kinase inhibitors in human plasma by capillary zone electrophoresis. Journal of Chromatography A 2018, 1579, 121-128. 15. Qu, Y.;  Sun, L.;  Zhang, Z.; Dovichi, N. J., Site-specific glycan heterogeneity characterization by hydrophilic interaction liquid chromatography solid-phase extraction, reversed-phase liquid chromatography fractionation, and capillary zone electrophoresis-electrospray ionization-tandem mass spectrometry. Analytical Chemistry 2018, 90 (2), 1223-1233. 16. Pero-Gascon, R.;  Sanz-Nebot, V.;  Berezovski, M. V.; Benavente, F., Analysis of circulating micrornas and their post-transcriptional modifications in cancer serum by on-line solid-phase extraction-capillary electrophoresis-mass spectrometry. Analytical Chemistry 2018, 90 (11), 6618-6625. 60  17. Simpson, S. L., Jr.;  Quirino, J. P.; Terabe, S., On-line sample preconcentration in capillary electrophoresis: Fundamentals and applications. Journal of Chromatography A 2008, 1184, 504-41. 18. Kim, J. B.;  Britz-McKibbin, P.;  Hirokawa, T.; Terabe, S., Mechanistic study on analyte focusing by dynamic pH junction in capillary electrophoresis using computer simulation. Analytical Chemistry 2003, 75, 3986-3993. 19. Wang, L.;  MacDonald, D.;  Huang, X.; Chen, D. D., Capture efficiency of dynamic pH junction focusing in capillary electrophoresis. Electrophoresis 2016, 37 (9), 1143-50. 20. Wang, L.;  Tong, W.; Chen, D. D. Y., Optimization of dynamic pH barrage junction focusing for weakly alkaline or zwitterionic analytes in capillary electrophoresis. Journal of Chromatography B 2018, 1095, 94-102. 21. Meng, P.;  Wang, Y.; Meng, L., pH-mediated stacking in capillary electrophoresis for analysis of opiates in saliva. Analytical Methods 2012, 4 (11), 3695-3700. 22. Silvertand, L. H.;  Sastre Torano, J.;  van Bennekom, W. P.; de Jong, G. J., Recent developments in capillary isoelectric focusing. Journal of Chromatography A 2008, 1204 (2), 157-70. 23. Wang, T.;  Zhao, S.;  Shen, C.;  Tang, J.; Wang, D., Determination of iodate in table salt by transient isotachophoresis–capillary zone electrophoresis. Food Chemistry 2009, 112 (1), 215-220. 24. Thompson, T. J.;  Foret, F.;  Vouros, P.; Karger, B. L., Capillary electrophoresis/electrospray ionization mass spectrometry: Improvement of protein detection limits using on-column transient isotochophoretic sample preconcentration. Analytical Chemistry 1993, 65, 900-906. 61  25. Reinhoud, N. J.;  Tjaden, U. R.; van der Greef, J., Automated isotachophoretic analyte focusing for capillary zone electrophoresis in a single capillary using hydrodynamic back-pressure programming. Journal of Chromatography 1993, 641, 155-162. 26. Hirokawa, T.;  Okamoto, H.; Ikuta, N., Operational modes for transient isotachophoretic preconcentration-capilalry zone electrophoresis and detectable concentration of rare earth ions. Electrophoresis 2001, 22, 3483-3489. 27. Quirino, J. P.; Terabe, S., Exceeding 5000-fold concentration of dilute analytes in micellar electrokinetic chromatography. Science 1998, 282, 465-468. 28. Palmer, J.;  Muro, N. J.; Landers, J. P., A universal concept for stacking neural analytes in micellar capillary electrophoresis. Analytical Chemistry 1999, 71, 1679-1687. 29. Takagai, Y.; Igarashi, S., UV-detection capillary electrophoresis for benzo[a]pyrene and pyrene following a two-step concentration system using homogeneous liquid–liquid extraction and a sweeping method. Analyst 2001, 126 (5), 551-552. 30. Kim, J.; Terabe, S., Online sample preconcentration techniques in micellar electrokinetic chromatography. Journal of Pharmaceutical and Biomedical Analysis 2003, 30, 1625-1643. 31. Zhang, Z. X.;  Zhang, X. W.; Zhang, S. S., Heart-cut capillary electrophoresis for drug analysis in mouse blood with electrochemical detection. Analytical Biochemistry 2009, 387 (2), 171-177. 32. Cao, C. X.;  He, Y. Z.;  Li, M.;  Qian, Y. T.;  Gao, M. F.;  Ge, L. H.;  Zhou, S. L.;  Yang, L.; Qu, Q. S., Stacking ionizable analytes in a sample matrix with high salt by a transient moving chemical reaction boundary method in capillary zone electrophoresis. Analytical Chemistry 2002, 74, 4167-4174. 62  33. Mikkers, F. E. P.;  Everaerts, F. M.; Verheggen, T. P. E. M., Application of sample stacking to gravity injection in capillary electrophoresis. Journal of Chromatography 1979,  (169), 11. 34. Quirino, J. P.; Terabe, S., Sample stacking of cationic and anionic analytes in capillary electrophoresis. Journal of Chromatography A 2000, 902, 119-135. 35. Breadmore, M. C.;  Wuethrich, A.;  Li, F.;  Phung, S. C.;  Kalsoom, U.;  Cabot, J. M.;  Tehranirokh, M.;  Shallan, A. I.;  Abdul Keyon, A. S.;  See, H. H.;  Dawod, M.; Quirino, J. P., Recent advances in enhancing the sensitivity of electrophoresis and electrochromatography in capillaries and microchips (2014-2016). Electrophoresis 2017, 38 (1), 33-59. 36. Chien, R. L.; Burgi, D. S., Field amplified sample injection in high-performance capillary electrophoresis. Journal of chromatography 1991, 559, 141-152. 37. Britz-McKibbin, P.; Chen, D. D. Y., Selective focusing of catecholamines and weakly acidic compounds by capillary electrophoresis using a dynamic pH junction. Analytical Chemistry 2000, 72, 1242-1252. 38. Britz-McKibbin, P.;  Bebault, G. M.; Chen, D. D. Y., Velocity-difference induced focusing of necleotides in capillary electrophoresis with a dynamic pH junction. Analytical Chemistry 2000, 72, 1729-1735. 39. Imami, K.;  Monton, M. R.;  Ishihama, Y.; Terabe, S., Simple on-line sample preconcentration technique for peptides based on dynamic pH junction in capillary electrophoresis-mass spectrometry. Journal of Chromatography A 2007, 1148 (2), 250-255. 40. Cao, C. X., Moving chemical reaction boundary and isoelectric focusing I. Conditional equations for svensson-Teselius' differential equation of solute concentrtaion distrubution in idealized isoelectric focusing at steady state. Journal of Chromatography A 1998, 813, 153-171. 63  41. Cao, C. X.;  Fan, L. Y.; Zhang, W., Review on the theory of moving reaction boundary, electromigration reaction methods and applications in isoelectric focusing and sample pre-concentration. Analyst 2008, 133 (9), 1139-1157. 42. Feng, H. Y.;  Li, X. J.;  Hou, S. R.;  Zheng, N.;  Hu, Z. B.; Yuan, Z. B., On-line concentration of trace genistein by acid barrage stacking in capillary electrophoresis with UV detection. Chinese Chemical Letters 2008, 19 (8), 973-976. 43. Lu, Y.;  Bai, H.;  Kong, C.;  Zhong, H.; Breadmore, M. C., Analysis of brazilin and protosappanin B in sappan lignum by capillary zone electrophoresis with acid barrage stacking. Electrophoresis 2013, 34 (24), 3326-3332. 44. Bachmann, S.;  Bakry, R.;  Huck, C. W.;  Polato, F.;  Corradini, D.; Bonn, G. K., Peptide mapping using capillary electrophoresis offline coupled to matrix-assisted laser desorption ionization time of flight mass spectrometry. Electrophoresis 2011, 32 (20), 2830-2839. 45. Hasan, M. N.;  Park, S. H.;  Oh, E.;  Song, E. J.;  Ban, E.; Yoo, Y. S., Sensitivity enhancement of CE and CE-MS for the analysis of peptides by a dynamic pH junction. Journal of Separation Science 2010, 33 (23-24), 3701-3709. 46. Kamentsev, M. Y.;  Kamentsev, Y. S.;  Moskvin, L. N.; Yakimova, N. M., Determination of zinc(II) and cadmium(II) as their complexes with Xylenol Orange by capillary elecrophoresis using online preconcentration. Journal of Analytical Chemistry 2011, 66 (2), 212-214. 47. Douglas, D. J., Linear quadrupoles in mass spectrometry. Mass Spectrometry Reviews 2009, 28 (6), 937-60. 48. Soga, T., Amino acid analysis by capillary electrophoresis electrospray ionization mass spectrometry. Analytical Chemistry 2000, 72, 1236-1241. 49. Takada, Y.;  Sakairi, M.; Koizumi, H., Atmospheric pressure chemical ionzation interface for capillary electrophoresis/mass spectrometry. Analytical Chemistry 1995, 67, 1474-1476. 64  50. Liu, Y.; Lopez Avila, V., Capillary electrophoresis coupled on-line with inductively coupled plasma mass spectrometry for elemental speciation. Analytical Chemistry 1995, 67, 2020-2025. 51. Preisier, J.;  Hu, P.;  Rejtar, T.;  Moskovets, E.; Karger, B. L., Capillary array electrophoresis-MALDI mass spectrometry using a vacuum deposition interface. Analytical Chemistry 2002, 74, 17-25. 52. Zhong, X.;  Maxwell, E. J.; Chen, D. D., Mass transport in a micro flow-through vial of a junction-at-the-tip capillary electrophoresis-mass spectrometry interface. Analytical Chemistry 2011, 83 (12), 4916-4923. 53. Maxwell, E. J.;  Zhong, X.;  Zhang, H.;  van Zeijl, N.; Chen, D. D., Decoupling CE and ESI for a more robust interface with MS. Electrophoresis 2010, 31 (7), 1130-1137. 54. Janini, G. M.;  Conrads, T. P.;  Wilkens, K. L.;  Issaq, H. J.; Veenstra, T. D., A sheathless nanoflow electrospray interface for on-line capillary electrophoresis mass spectrometry. Analytical Chemistry 2003, 75, 1615-1619. 55. Wu, P.;  Nielsen, T. E.; Clausen, M. H., Small-molecule kinase inhibitors: an analysis of FDA-approved drugs. Drug Discovery Today 2016, 21 (1), 5-10. 56. Zhao, Z.; Bourne, P. E., Progress with covalent small-molecule kinase inhibitors. Drug Discovery Today 2018, 23 (3), 727-735. 57. Picard, S.;  Titier, K.;  Etienne, G.;  Teilhet, E.;  Ducint, D.;  Bernard, M. A.;  Lassalle, R.;  Marit, G.;  Reiffers, J.;  Begaud, B.;  Moore, N.;  Molimard, M.; Mahon, F. X., Trough imatinib plasma levels are associated with both cytogenetic and molecular responses to standard-dose imatinib in chronic myeloid leukemia. Blood 2007, 109 (8), 3496-3499. 58. Peng, B.;  Hayes, M.;  Resta, D.;  Racine-Poon, A.;  Druker, B. J.;  Talpaz, M.;  Sawyers, C. L.;  Rosamilia, M.;  Ford, J.;  Lloyd, P.; Capdeville, R., Pharmacokinetics and 65  pharmacodynamics of imatinib in a phase I trial with chronic myeloid leukemia patients. Journal of Clinical Oncology 2004, 22 (5), 935-942. 59. Larson, R. A.;  Druker, B. J.;  Guilhot, F.;  O'Brien, S. G.;  Riviere, G. J.;  Krahnke, T.;  Gathmann, I.; Wang, Y., Imatinib pharmacokinetics and its correlation with response and safety in chronic-phase chronic myeloid leukemia: a subanalysis of the IRIS study. Blood 2008, 111 (8), 4022-4028. 60. Bakhitiar, R.;  Lohne, J.;  Ramos, L.;  Khemani, L.;  Hayes, M.; Tse, F., High-throughput quantification of anti-leukemia drug STI571 and its main metabolite in human plasma using liquid-chromatography-tandem mass spectrometry. Journal of chromatography B 2002, 768, 325-340. 61. Zhuang, W.;  Qiu, H. B.;  Chen, X. M.;  Yuan, X. H.;  Yang, L. F.;  Sun, X. W.;  Zhou, X. J.;  Huang, M.;  Wang, X. D.; Zhou, Z. W., Simultaneous quantification of imatinib and its main metabolite N-demethyl-imatinib in human plasma by liquid chromatography-tandem mass spectrometry and its application to therapeutic drug monitoring in patients with gastrointestinal stromal tumor. Biomedical Chromatography 2017, 31, e4022. 62. Titier, K.;  Palmer, J.;  Ducint, D.;  Teilhet, E.;  Moore, N.;  Berthaud, P.;  Mahon, F. X.; Molimard, M., Quantification of Imatinib in human plasma by high-performance liquid chromatography-tandem mass spectrometry. Therapeutic Drug Monitoring 2005, 27, 634-640. 63. Dagher, D.;  Cohen, M.;  Williams, G.;  Rothmann, M.;  Gobburu, J.;  Robbie, G.;  Rahman, A.;  Chen, G.;  Staten, A.;  Griebel, D.; Pazdur, R., Approval summary imatinib mesylate in the treatment of metastatic and or unresectable malignant gastrointestinal stromal tumors. Clinical Cancer Research 2002, 8, 3034-3038. 64. Cohen, M.;  Williams, G.;  Johnson, J. R.;  Gobburu, J.;  Rahman, A.;  Benson, K.;  Leighton, J.;  Kim, S. K.;  Wood, R.;  Rothmann, M.;  Chen, G.;  Maung U, K.;  Staten, A. M.; 66  Pazdur, R., Approval summary for imatinib mesylate capsules in the treatment of chronic myelogenous leukemia. Clinical Cancer Research 2002, 8, 935-942. 65. Rosen, C. J.; Bouxsein, M. L., Mechanisms of disease: is osteoporosis the obesity of bone? Nature Clinical Practice Rheumatology 2006, 2 (1), 35-43. 66. Fleisch, H.;  Reszka, A.;  Rodan, G.; Rogers, M., Bisphosphate: mechanism of actions. In Principles of Bone Biology, Second ed.; Academic Press: 2002; pp 1361-1385. 67. Kim, H. S.;  Jung, H. Y.;  Kim, M. O.;  Joa, K. L.;  Kim, Y. J.;  Kwon, S. Y.; Kim, C. H., Successful conservative treatment: multiple atypical fractures in osteoporotic patients after bisphosphate medication: a unique case report. Medicine 2015, 94 (5), 1-4. 68. Huang, Y. S.;  Chen, S. N.; Whang, C. W., Capillary electrophoresis-electrochemiluminescence detection method for the analysis of ibandronate in drug formulations and human urine. Electrophoresis 2011, 32 (16), 2155-2160. 69. Zacharis, C. K.; Tzanavaras, P. D., Determination of bisphosphonate active pharmaceutical ingredients in pharmaceuticals and biological material: a review of analytical methods. Journal of Pharmaceutical and Biomedical Analysis 2008, 48 (3), 483-496. 70. Al Deeb, S. K.;  Hamdan, II; Al Najjar, S. M., Spectroscopic and HPLC methods for the determination of alendronate in tablets and urine. Talanta 2004, 64 (3), 695-702. 71. Ptacek, P.;  Klima, J.; Macek, J., Determination of alendronate in human urine as 9-fluorenylmethyl derivative by high-performance liquid chromatography. Journal of chromatography B 2002, 767, 111-116. 72. Flesch, G.; Hauffe, S. A., Determiantion of the bisphosphate pamidronate disodium in urine by pre-column derivatization with fluorescamine, high-performance liquid chromatography and fluorescence detection. Journal of Chromatography 1989, 489, 446-451. 67  73. Wu, D.;  Yang, J.;  Cai, Y.;  Lin, X.;  Yan, J.; Li, H., Simultaneous determination of two amino bisphosphonates drugs by micellar electrokinetic chromatography. Journal of Chromatography Science 2015, 53 (10), 1794-1800. 74. Su, S. W.;  Liao, Y. C.; Whang, C. W., Analysis of alendronate in human urine and plasma by magnetic solid-phase extraction and capillary electrophoresis with fluorescence detection. Journal of Separation Science 2012, 35 (5-6), 681-687. 75. Hasan, M.;  Schumacher, G.;  Seekamp, A.;  Taedken, T.;  Siegmund, W.; Oswald, S., LC–MS/MS method for the determination of clodronate in human plasma. Journal of Pharmaceutical and Biomedical Analysis 2014, 100, 341-347. 76. Zhu, L. S.;  Lapko, V. N.;  Lee, J. W.;  Basir, Y. J.;  Kafonek, C.;  Olsen, R.; Briscoe, C., A general approach for the quantitative analysis of bisphosphonates in human serum and urine by high-performance liquid chromatography/tandem mass spectrometry. Rapid Communications in Mass Spectrometry 2006, 20 (22), 3421-3426. 77. Tarcomnicu, I.;  Silvestro, L.;  Savu, S. R.;  Gherase, A.; Dulea, C., Development and application of a high-performance liquid chromatography-mass spectrometry method to determine alendronate in human urine. Journal of Chromatography A 2007, 1160 (1-2), 21-33. 78. Mahon, F. X., Pharmacologic monitoring and determinants of intracytoplasmic drug levels. Best Practice Research Clinical Haematology 2009, 22 (3), 381-386. 79. Solassol, I.;  Bressolle, F.;  Philibert, L.;  Charasson, V.;  Astre, C.; Pinguet, F., Liquid chromatography‐electrospray mass spectrometry determination of imatinib and its main metabolite,n‐desmethyl‐imatinib in human plasma. Journal of Liquid Chromatography & Related Technologies 2006, 29 (20), 2957-2974. 68  80. Szczepek, W. J.;  Kosmacinska, B.;  Bielejewska, A.;  Luniewski, W.;  Skarzynski, M.; Rozmarynowska, D., Identification of imatinib mesylate degradation products obtained under stress conditions. Journal of Pharmaceutical and Biomedical Analysis 2007, 43 (5), 1682-1691. 81. Zhang, C.;  Huang, L.;  Wu, Z.;  Chang, C.; Yang, Z., Determination of sulfonate ester genotoxic impurities in imatinib mesylate by gas chromatography with mass spectrometry. Journal of Separation Science 2016, 39 (18), 3558-3563. 82. Li, J.;  Huang, Y.;  Huang, L.;  Ye, L.;  Zhou, Z.;  Xiang, G.; Xu, L., Determination of imatinib mesylate and related compounds by field amplified sample stacking with large volume sample injection capillary electrophoresis. Journal of Pharmaceutical and Biomedical Analysis 2012, 70, 26-31. 83. Burgi, D. S.; Chien, R. L., Journal of Microcolumn Separations 1991, 3 (3), 199-202. 84. Sakiyama, N.;  Kataoka, H.; Makita, M., Gas chromatographic analysis of 3-amino-1-hydroxypropylidene-1,1-bisphosphate and related bisphosphate as their N-isobutoxycarbonyl methyl ester derivatives. Journal of chromatography A 1996, 724, 279-284. 85. Huikko, K.; Kostiainen, R., Analysis of bisphosphonates by capillary electrophoresis–electrospray ionization mass spectrometry. Journal of chromatography A 2000, 872, 289-298. 86. Cohen, H.;  Solomon, V.;  Alferiev, I. S.;  Breuer, E.;  Ornoy, A.;  Patlas, N.;  Eidelman, N.;  Hagele, G.; Golomb, G., Bisphosphonates and tetracycline experimental: models for their evaluation in calcium-related disorders. Pharmaceutical Research 1998, 15, 606-613. 87. Shao, Y. X.;  Chen, G. H.;  Fang, R.;  Zhang, L.;  Yi, L. X.; Meng, H. L., Analysis of six beta-lactam residues in milk and egg by micellar electrokinetic chromatography with large-volume sample stacking and polarity switching. Journal of Agriculture and Food Chemistry 2016, 64 (17), 3456-3461. 69  88. Zhao, Y.;  Wang, D.;  Li, X. J.; Liu, Y., Determination of rutin by large volume sample stacking with polarity switching and acid barrage stacking. Chemistry International 2016, 2 (3), 121-127. 70  Appendix  Appendix A  Experimental of CE-LIF method for the quantitative determination of ALN Standard solutions, ALN tablet and serum samples was prepared following the same procedure in Chapter 3. The only difference was that ALN tablet extract was further diluted 20 times after membrane filtration instead of 200 times in CE-MS.  Before introduced into CE capillary, as-resulted sample solutions were derivatized by fluorogenic labelling reagents FQ (3-(2-furoyl) quinoline-2-carboxaldehyde 2). Derivatization reaction mechanism was illustrated in Figure 4.1. ATTO-TAG™ FQ amine-derivatization kit was purchased from Thermo Fisher Scientific (Nepean, ON, CA). Derivatization was carried out by mixing a 10 µl aliquot of sample solution with 20 µl of 10 mM KCN solution and 10 µl of 10 mM FQ solution. The reaction vial was vortexed for 5 seconds and wrapped by aluminum foil to prevent light-induced degradation. The reaction was left at room temperature at for 3 h. The maximum excitation wavelength of fluorogenic products was at 488 nm with maximum emission wavelength at 600 nm.  Figure 4.1 Reaction of fluorogenic derivatization reagent FQ with primary amine. CE-LIF was performed on P/ACE MDQ Glycoprotein System equipped with 488 nm Argon-ion laser module (Beckman Coulter Cor., Fullerton, CA). The emission light was filtered by 600 nm bandpass filter (FWHM 20 nm, 10 mm diameter, Andover Corporation, NH, USA). The bare fused silica capillary (50-μm i.d. × 68 cm) was used for separation of bisphosphate 71  drugs. The effective length of the capillary was 57 cm to the detection window. The capillary was first rinsed and activated in the same way in Chapter 2. The separation voltage was 20 kV.  Appendix B  Results of CE-LIF method for the quantitative determination of ALN B.1 Optimization of separation conditions  Figure 4.2 Influence of derivatization buffer pH  Reaction pH impacts the derivatization efficiency of ALN and PAM. A range of pH from 7 to 12 was studied. Since higher pH led to the deprotonation of protonated amino group (pKa=12.2 for ALN and pKa=10.84 for PAM), the increasing of pH promotes the derivatization efficiency and result in higher fluorescent signal (as shown in Figure 4.2). However, too high pH led to more interfering side products and ruined electropherogram. At last, pH 10 was selected as the optimal pH value. 72   Figure 4.3 Influence of BGE concentration (Na2HPO4) on separation.  The BGE concentration impact the thickness of double layer and buffer viscosity, thus affecting zeta potential and EOF. The concentration of Na2HPO4 was investigated ranging from 5 to 40 mM. With the increasing of Na2HPO4 concentration, the apparent mobility of the analytes decreased owing to the suppressing of EOF. However, the resolution of ALN and PAM peaks increased to 2.0 above concentration of 20 mM. Therefore, 20 mM was adopted to realized fast analysis as well as baseline separation.  Figure 4.4 Influence of BGE pH on separation. 73  BGE pH was optimized in the range from 8.0 to 11.0. As shown in Figure 4.4, the mobility of ALN and PAM increased with increasing pH thanks to the contribution of EOF and maintained stable above pH 10. In order to shorten analysis time, BGE with pH 9.5 was chosen to perform the separation.  B.2 Calibration curve   Figure 4.5 Calibration curve for ALN with CE-LIF method. Six standard solutions of ALN in the concentration range of 0.6-20 μg/mL were analyzed under the optimized separation conditions. Each working solution contained 10 μg/mL of PAM (I.S.). Working solutions were analyzed 5 times each concentration to create a calibration curve (illustrated in Figure 4.5). Using the relative peak area of ALN to PAM, the linear equations were y=0.2687x+0.00864 (R2=0.99983) for ALN (N=5). The detection limit (S/N=3) was 0.2 µg/mL for ALN. Regression equation R2 Linear range (µg/mL) LOD (µg/mL) y=0.2688x+0.00864 0.99983 0.6 - 20  0.2 Table 4.1 Summary of regression equation for ALN with CE-LIF methods 74  B.3 Assay validation Concentration (µg/mL) R.S.D. Bias Added Measured 0.6 0.5 4.3% -6.3% 1.5 1.5 6.4% 2.6% 6.0 5.9 0.2% -0.8% 15.0 15.3 2.3% 0.6% Table 4.2 Intra-assay precision and accuracy LLOQ (0.6 µg/mL), low concentration (1.5 µg/mL), medium concentration (6.0 µg/mL) and high concentration (15.0 µg/mL) were chosen as quality control samples.  Intra-assay tests were conducted by analyzing each QC samples five times within the same day (N=5). Intra-assay precision and accuracy of the method are illustrated in Table 4.2. Results show that the precision was better than 6.4% and the bias did not exceed ±6.3% of added value at all levels.  Concentration (µg/mL) R.S.D. Bias Added Measured 0.6 0.6 1.7% -7.7% 1.5 1.5 1.6% 2.6% 6.0 5.9 2.6% -1.8% 15.0 15.3 1.6% 2.0% Table 4.3 Inter-assay precision and accuracy Inter-assay precision and accuracy was evaluated over 3 consecutive days (N=3).  The respective data was given in Table 4.3. The precision was at most 2.6% and the inaccuracy was better than ±7.7% at all levels.  75  B.4 Sample test  Figure 4.6 Electropherograms of drug tablet extract  Figure 4.7 Electropherogram of serum sample containing 100 µg/mL ALN and 100 µg/mL PAM (I.S.). The ALN content in drug tablet and serum samples were also tested with CE-LIF method. The typical electropherograms for the analysis of drug sample were shown in Figure 76  4.6. The content found in one drug tablet is 13.20 mg/tablet, which was close to the labeled amount of 13.05 mg/tablet counted as C4H12NaNO7P2·3H2O.  As shown in Figure 4.7, there was also a problem in analyzing serum samples with CE-LIF method.  

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}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            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:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0380592/manifest

Comment

Related Items