<%BANNER%>

Tandem High Performance Liquid Chromatography/Mass Spectromic Studies of Intact and Modified Ziconotide

Permanent Link: http://ufdc.ufl.edu/UFE0024845/00001

Material Information

Title: Tandem High Performance Liquid Chromatography/Mass Spectromic Studies of Intact and Modified Ziconotide
Physical Description: 1 online resource (170 p.)
Language: english
Creator: Mendoza, Jhoana
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

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

Notes

Abstract: STUDIES OF INTACT AND MODIFIED ZICONOTIDE BY LIQUID CHROMATOGRAPHY - TANDEM MASS SPECTROMETRY Ziconotide is the synthetic version of omega-conopeptide MVIIA, a 25-amino acid cone snail toxin from Conus magus. Its cysteine residues form 3 disulfide bridges, while its arginine, lysine residues and amidated C-terminal make it highly basic. It is a potent analgesic recently approved for the treatment of intense refractory pain in humans. The development of an analytical method for the qualitative and quantitative determination of this peptide in complex matrices is, therefore, of interest to many. High performance liquid chromatographic (HPLC)-tandem mass spectrometric (MS/MS)analyses of intact and modified ziconotide were carried out in this study. The chromatographic performance of various columns for reversed-phase HPLC of ziconotide was investigated. The peak shape and retention of ziconotide on the columns used were compared. The effects of solvent composition on the charge state distribution and response of ziconotide were investigated. Optimization of various mass spectrometric parameters such as spray voltage, sheath gas and auxiliary gas flows were also conducted. Analysis of intact ziconotide showed that the extensive cross-linking from multiple disulfide bridges reduces the structural information that can be derived from the MS/MS spectrum of ziconotide. Modification of the analyte by reducing the disulfide bonds and alkylating the resulting sulfhydryl groups eliminate cross-linking and increase the probability of fragmentation during collision-induced dissociation (CID). Fragment ions resulting from typical peptide bond cleavage (e.g., b and y ions) were observed after modification of ziconotide and the internal standard, omega-conotoxin GVIA. The limit of quantitation obtained from the modified peptide experiments was 0.50 ng on column. The HPLC-MS method developed for the analysis of intact ziconotide was successfully validated. The validated method was then used to evaluate the degradation of ziconotide in a test solution similar in composition to the commercially available drug formulation, Prialtregistered trademark (Elan Pharmaceutical, Inc.) The plot of the natural logarithm of ziconotide concentration versus storage time at 50 degree Celsius and half-life calculations indicated that the degradation of ziconotide could be a first- or pseudo first-order reaction. Results pertaining to the rate of degradation of a drug substance provide useful information for proper handling and storage of the ziconotide drug solution.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jhoana Mendoza.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Eyler, John R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024845:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024845/00001

Material Information

Title: Tandem High Performance Liquid Chromatography/Mass Spectromic Studies of Intact and Modified Ziconotide
Physical Description: 1 online resource (170 p.)
Language: english
Creator: Mendoza, Jhoana
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

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

Notes

Abstract: STUDIES OF INTACT AND MODIFIED ZICONOTIDE BY LIQUID CHROMATOGRAPHY - TANDEM MASS SPECTROMETRY Ziconotide is the synthetic version of omega-conopeptide MVIIA, a 25-amino acid cone snail toxin from Conus magus. Its cysteine residues form 3 disulfide bridges, while its arginine, lysine residues and amidated C-terminal make it highly basic. It is a potent analgesic recently approved for the treatment of intense refractory pain in humans. The development of an analytical method for the qualitative and quantitative determination of this peptide in complex matrices is, therefore, of interest to many. High performance liquid chromatographic (HPLC)-tandem mass spectrometric (MS/MS)analyses of intact and modified ziconotide were carried out in this study. The chromatographic performance of various columns for reversed-phase HPLC of ziconotide was investigated. The peak shape and retention of ziconotide on the columns used were compared. The effects of solvent composition on the charge state distribution and response of ziconotide were investigated. Optimization of various mass spectrometric parameters such as spray voltage, sheath gas and auxiliary gas flows were also conducted. Analysis of intact ziconotide showed that the extensive cross-linking from multiple disulfide bridges reduces the structural information that can be derived from the MS/MS spectrum of ziconotide. Modification of the analyte by reducing the disulfide bonds and alkylating the resulting sulfhydryl groups eliminate cross-linking and increase the probability of fragmentation during collision-induced dissociation (CID). Fragment ions resulting from typical peptide bond cleavage (e.g., b and y ions) were observed after modification of ziconotide and the internal standard, omega-conotoxin GVIA. The limit of quantitation obtained from the modified peptide experiments was 0.50 ng on column. The HPLC-MS method developed for the analysis of intact ziconotide was successfully validated. The validated method was then used to evaluate the degradation of ziconotide in a test solution similar in composition to the commercially available drug formulation, Prialtregistered trademark (Elan Pharmaceutical, Inc.) The plot of the natural logarithm of ziconotide concentration versus storage time at 50 degree Celsius and half-life calculations indicated that the degradation of ziconotide could be a first- or pseudo first-order reaction. Results pertaining to the rate of degradation of a drug substance provide useful information for proper handling and storage of the ziconotide drug solution.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Jhoana Mendoza.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Eyler, John R.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-02-28

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024845:00001


This item has the following downloads:


Full Text

PAGE 1

1 STUDIES OF INTACT AND MODIFIED ZICONOTIDE BY LIQUID CHROMATOGRAPHY TANDEM MASS SPECTROMETRY By JHOANA AVENDAO MENDOZA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMEN T OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Jhoana A. Mendoza

PAGE 3

3 To my parents and Dodge

PAGE 4

4 ACKNOWLEDGMENTS I am grateful to my research advisor, Dr. John R. Eyler, for his support and scientific guidance and belief in my capabilities. I would like to thank Dr. Richard Sams for allowing me to work on the ziconotide project and instill ing in me the importance of working systematically. I would also like to extend my gratitude to the rest of my committee members, Dr. Nicole Horenstein, Dr. David Powell and Dr. Richard Yost. I would like to thank Dr. Laszlo Prokai for paving the way to our collaboration with the College of Veterinary Medicine Racing Laboratory. Special thanks go to Dr. C ynthia Cole for supporting the collaboration and allowing me to work at the Racing Laboratory and for being a former member of my committee. I am also grateful to Dr. Nancy Szabo for her guidance a nd encouragement. I would like to thank Dr. Nancy Denslow for her generosity in equipment use. I also acknowledge the financial support of the Racing Medication and Testing Consortium. While at the Racing Laboratory, I was fortunate enough to work with Dr. Keith Zientek who taught me a lot of what I knew about sample preparation and the operation of the LTQ. I would like to thank him and the group of wonderful people who have been my labmates away from Chemistry Brooks, Vanessa, Christa, Julia and Craig. I would like to thank Pa trick Russell especially for his support and willingness to help whenever I need to troubleshoot LC/MS issues and Margaret Wilding for her encouragement and gentle support. Dr. Cristina Dancel and Dr. Sue Abbatiello have been very great resource persons and good friends too; I have learned a lot of science from them. I am also th ankful to the rest of the Eyler Polfer group for the friendship and helpful discussions during group meetings. My stay in Florida would not have been as fun and memorable as it i s if not for the great company of Filipino friends who have been my family away from home Jemy, Joy, Mario, Jen, Fair, Emilio, the Pabits, the Javelosas and the rest of Pinoy_UF group; my thanks to you all.

PAGE 5

5 I have been blessed to have a family whos sup portive of my dreams. I thank my dear parents, Rodolfo and Evelyn, for their hard work and countless sacrifices to give us a chance at a better life I thank my sister, Vanessa, for her strength and my brother BJ, for his joyful ways of encouraging me. I would also like to express my gratitude to the rest of the Avenda o and Mendoza families especially to Auntie Letty, Auntie Josie, Uncle Angel, Uncle Victor, Uncle Alex, Auntie Carmen and Auntie Tessie for their guidance and unwavering support Finally, I thank my husband, Dodge, for his support, patience, and sacrifice and for always encouraging and inspiring me to work hard to make it to this point.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .......................................................................................................................10 ABSTRACT ...................................................................................................................................14 CHAPTER 1 INTRODUCTION ..................................................................................................................16 Background of Study ..............................................................................................................16 Conotoxins and Ziconotide ..............................................................................................16 High Performance Liquid Chromatography ....................................................................17 Electrospray Ionization ....................................................................................................18 Quadrupole Ion Trap Mass Spectrometry .......................................................................19 Objectives of Research ...........................................................................................................19 Analysis of Intact Ziconotide ..........................................................................................20 Analysis of Modified Ziconotide .....................................................................................20 Development of a StabilityIndicating Assay for Ziconotide in CommercialType Formulations ................................................................................................................20 2 HPLC ESI MS/MS ANALYSIS OF INTACT ZICONOTIDE .............................................25 Introduction .............................................................................................................................25 Experimental Section ..............................................................................................................26 Chemicals and Reagents ..................................................................................................26 Instrumentation ................................................................................................................26 Optimization of Mass Spectrometric Para meters ............................................................27 Optimization of HPLC Parameters ..................................................................................28 Preparation of Calibration Standards ...............................................................................29 HPLC ESI MS/MS Analysis ...........................................................................................29 MS Data Analysis ............................................................................................................31 Results and Discussion ...........................................................................................................32 Optimized Mass Spectrometric Parameters for Ziconotide Analysis ..............................32 Optimized HPLC Parameters for Ziconotide Analysis ...................................................34 Effect of Various Solvent Systems and Flow Rates on Ziconotide Response .........34 Effect of Column Chemistry on Peak Shape and Retention of Ziconotide ..............38 Weighted Least Squares Linear Regression ....................................................................39 Generation of the Response Curve for Ziconotide ..........................................................40 Conclusion ..............................................................................................................................43

PAGE 7

7 3 HPLC ESI MSn ANALYSIS OF REDUCED AND ALKYLATED ZICONOTIDE ...........73 Introduction .............................................................................................................................73 Experimental Section ..............................................................................................................75 Chemicals and Reagents ..................................................................................................75 Evaluation of Ziconotide Extraction Recovery with ZipTip C18 Pipette Tips ................75 Optimization of Parameters Affecting Reduction ...........................................................76 Optimization of Parameters Affecting Alky lation ..........................................................78 Generation of the Response Curve for Reduced and Alkylated Ziconotide ...................79 Application of the Reduction and Alkylation M ethod to Ziconotide Plasma Extract .....79 HPLC ESI MS/MS Analysis ...........................................................................................81 MS Data Analysis ............................................................................................................82 Results and Discussion ...........................................................................................................83 Comparison of Extraction Recoveries with ZipTip C18 Pipette Tips ..............................83 Optimized Conditions for Ziconotide Reduction ............................................................84 Optimized Conditions for Ziconotide Alkylation ............................................................87 Identification of Potential Fragment Ions for Quantitation .............................................90 Generation of the Response Curve for Modified Ziconotide ..........................................92 Reduction and Alkylation of Ziconotide Plasma Extract ................................................95 Conclusion ..............................................................................................................................97 4 DEVELOPMENT OF A STABILITY INDICATING METHOD FOR ZICONOTIDE IN COMMERCIAL TYPE FORMULATIONS ..................................................................128 Introduction ...........................................................................................................................128 Experimental Section ............................................................................................................131 Chemicals and Reagents ................................................................................................131 Method Validation .........................................................................................................131 Effect of Time in Autosampler Rack on Ziconotide Stability in Simple Solution ........132 Effect of Harsh Oxidative Conditions on Ziconotide Stability in Commercial type Formulations ..............................................................................................................132 HPLC ESI MS/MS Analysis .........................................................................................134 MS Data Analysis ..........................................................................................................135 Results and Discussion .........................................................................................................135 Intra and Inter day Accuracy and Precision of Quality Control (QC) Samples ..........135 Post Preparative Stability of Ziconotide .......................................................................136 Effect of Harsh Oxidative Conditions on Ziconoti de Degradation ...............................137 Conclusions ...........................................................................................................................143 5 CONCLUSIONS AND FUTURE DIRECTIONS ...............................................................157 APPENDIX: LIST OF FRAGMENT IONS .............................................................................159 LIST OF REFERENCES .............................................................................................................163 BIOGRAPHICAL SKETCH .......................................................................................................170

PAGE 8

8 LIST OF TABLES Table page 21 Test for homoscedasticity using the F test. ........................................................................62 22 Regression parameters of the calibration curves generated for each weighting factor ( wi) and the respective sum of the relative errors ( %RE) for the inter day assay data. ....................................................................................................................................64 23 Regression parameters for the six calibrat ion curves of ziconotide. .................................66 24 Precision and accuracy of ziconotide calibration standards assayed during a six day period. ................................................................................................................................68 31 The m/z of the various charge states of ziconotide after reduction of its disulfide bonds. ...............................................................................................................................103 32 The m/z of the oxidation products of the various ions of intact and reduced ziconoti de. ........................................................................................................................103 33 The m/z of the various charge states of ziconotide after reduction and alkylation of its disulfide bonds. ................................................................................................................110 34 The m/z of the oxidation products of the various ions of reduced and alkylated ziconotide. ........................................................................................................................110 35 Precision and accuracy of calibration standards for the analysis of reduced and alkylated ziconotide. ........................................................................................................122 41 Intra and inter day accuracy and precision of quality control samples. .........................145 42 Regression parameters f or the time delayed analysis of the six calibration curves of ziconotide. ........................................................................................................................146 43 Precision and accuracy of ziconotide calibration standards which were reinjected after 12 and 24 hr of init ial analysis. ................................................................................146 44 Intra and inter day accuracy and precision of timedelayed analysis of quality control samples. ...............................................................................................................147 45 Results of analysis of test solution samples stored at 50 C and kept in contact with air at various times. ..........................................................................................................150 46 Room temperature stability derived from Q T calculations and stability d ata at 37 C. ..153 47 Room temperature stability derived from Q T calculations and stability data at 50 C. ..153

PAGE 9

9 48 Comparis on of experimental and calculated stability data at 50 C. ...............................153 A 1 Fragment ions of the +4 ion of modified ziconotide after CID at 35% collision energy. ..............................................................................................................................159 A 2 Product ions of the +4 ion of modified conotoxin GVIA after CID at 35% collision energy. ..............................................................................................................................161

PAGE 10

10 LIST OF FIGURES Figure page 11 3D structure of ziconotide ( left ) and its amino acid sequence ( right ). .............................22 12 Amino acid sequence of conotoxin GVIA. The disulfide bonds are shown in gold and X denotes hydroxyproline residues. ............................................................................23 13 Schematic of the electrospray ionization process. .............................................................24 21 Internal components of the LTQ mass spectr ometer. ........................................................45 22 Instrument setup for the HPLC low flow ESI MS analysis of ziconotide. ......................46 23 Response of ziconotide and TFA adduct with and without auxiliary gas flow. ................47 24 Effects of various tube lens offset voltages on the full scan mass spectra of ziconotide: ( A ) 100 V, ( B ) 160 V and ( C ) 200 V. ..............................................................48 25 Extracted ion chromatograms of m/z 661 from low flow isocratic elution of ziconotide using 0.5% HAc in ( A ) 70:30, ( B ) 50:50, ( C ) 20:80 and ( D ) 10:90 MeOH: H2O mixtures. ....................................................................................................................49 26 Mass spectra of ziconotide peaks that were eluted isocratically with 0.5% HAc in ( A ) 70:30, ( B ) 50:50, ( C ) 20:80 and ( D ) 10:90 MeOH: H2O mixtures. .............................50 27 E xtracted ion chromatograms of m/z 661 from low flow gradient elution of ziconotide with 0.5% HAc in ( A ) 70:30, ( B ) 20:80 and ( C ) 10:90 MeOH: H2O mixtures as starting mobile phase compositions. ...............................................................51 28 Extracted ion chromatograms of m/z 661 from low flow isocratic elution of ziconotide using 15:85 ACN:H2O with various acidic components. .................................52 29 Mass spectra of ziconotide peaks from isocratic elutions using 15:85 ACN:H2O with various acidic components. ................................................................................................53 210 Response of ziconotide ions and a TFA adduct with increasing TFA concentration in the mobile phases. ..............................................................................................................54 211 Response of conotoxin GVIA ions and a TFA adduct with increasing TFA concentration in the mobile phases. ...................................................................................55 212 Res ponse of ziconot ide ions and its TFA adducts with increasing HAc concentration in mobile phases with 0.01% TFA. ....................................................................................56 213 Response of conotoxin GVIA ions and a TFA adduct with increasing HAc concentration in mobile phases with 0.01% TFA.. ............................................................57

PAGE 11

11 214 Extracted ion chromatograms of m/z 661 for the lowest amount of ziconotide that can be detected on Atlantis dC18 columns with the same length but varying internal diameters. ...........................................................................................................................58 215 Mass spectra of peaks corresponding to the lowest amount of ziconotide that can be detected on 150 mm Atlantis dC18 columns with internal diameters of 1 mm (top panel ) and 2.1 mm ( bottom panel ). ....................................................................................59 216 Comparison of the peak shape of ziconotide obtained with various columns and acid composition of the HPLC solvents. ...................................................................................60 217 Plot of yresidual vs. ziconotide concentration. Data from the inter day assay of ziconotide standards were used. .........................................................................................61 218 Plots of percent relative error (%RE) vs. ziconotide concentration for unweighted (Model 1) and weighted (Models 2 5) linear regres sions of ziconotide standards. ...........63 219 Graphical representations of the area ratio of t he +4 ions of ziconotide to conotoxin GVIA vs. ziconotide concentration. ..................................................................65 220 Response of ziconotide and conotoxin GVIA (IS) across the calibration curve. ..........67 221 Extracted ion chromatograms ( A ) oxidized ziconotide, ( B ) ziconotide and ( C ) conotoxin GVIA for the 0.5 ng/L calibration standard. ..................................................69 222 F ull scan mass spectra of ( A ) oxidized ziconotide, ( B ) ziconotide and ( C ) conotoxin GVIA.. ...............................................................................................................70 223 Methionine oxidation under (a) mild and (b) strong oxidizing conditions. .......................71 224 Product ion spectra of the +4 ions of ziconotide at m/z 661.0 ( top) and conotoxin GVIA at m/z 760.4 ( bottom ) at 30% collision energy. ......................................................72 31 Mechanism involved in the reduction of disulfide bridges by dithiols. .............................99 32 Representation of disulfide bond reduction and alkylation of the resulting sulfhydryl groups. ..............................................................................................................................100 33 Extraction recoveries of ziconotide from ZipTip C18 pipette tips using various elution solvents. ...........................................................................................................................101 34 Representation of the net reactions for the reduction and alkylation of ziconotide .......102 35 Response of the various products of ziconotide reduction with increasing DTT:Z molar ratio. .......................................................................................................................104 36 Response of reduced and oxidized ziconotide with increasing reduction times. ...........105

PAGE 12

12 37 Response of reduced and oxidized ziconotide with increa sing reduction temperatures. ...................................................................................................................106 38 Response of reduced and oxidized ziconotide at 10and 30minute reduction times with increasing DTT concentrations. ..............................................................................107 39 Base peak chromatogram of the various products of ziconotide reduction. ....................108 310 Full scan mass spectra of reduced and oxidized ziconotide( A C ). .................................109 311 Average peak area (n = 6) for the reduction and alkylation products of ziconotide at increasing alkylation times. ............................................................................................111 312 Response of the side products of modified ziconotide at different alkylation times. ....112 313 Response of the various ions of reduced and alkylated ziconotide at different alkylatio n temperatures. .................................................................................................113 314 Response of the side products of modified ziconotide at various alkylation temperatures.. ...................................................................................................................114 315 Average peak areas (n = 6) of the triply reduced and alkylated ziconotide with increasing IAM: Z molar ratios at 15 and 60min. alkylation times. ............................115 316 Signal contributions of the side products of ziconotide modification with increasing IAM: Z molar ratios at 15 and 60min. alkylation times. ...............................................116 317 Response comparison of the major and side products of modified ziconot ide with and without the addition of NH4OH to enhance alkylation efficiency and the addition of TFA to quench the reaction. ........................................................................................117 318 Extracted ion chromatograms of the +4 charge states of th e reduced and alkylated ziconotide ( top ) and conotoxin GVIA ( bottom ).. ........................................................118 319 Full scan mass spectra of the modified conopeptides, ziconotide ( top) and conotoxin GVIA ( bottom ). ...............................................................................................119 320 Full scan MS/MS spectra of the +4 ions of modified ziconotide at m/z 748 ( top)and conotoxin GVIA at m/z 848 ( bottom ). ........................................................................120 321 Graphical representation of the area ratio of the +4 ions of modified ziconotide to conotoxin GVIA versus ziconotide concentration. .........................................................121 322 Representative base peak chromatograms of reduction and alkylation blank ( top) and control (bottom ) solutions. .............................................................................................123 323 Extracted ion chromatograms (EIC) of the 3 major ions of intact ziconotide for the ( A ) C18 solid phase ext raction (SPE) extract, ( B ) control and ( C ) matrix blank. ...........124

PAGE 13

13 324 Full scan mass spectra of the major peaks in the EIC of ( A ) ziconotide SPE extract, ( B ) control and ( C ) matrix blank. ....................................................................................125 325 Extracted ion chromatograms of the 3 major ions ( m/z 599, 748, and 997) of reduced and alkylated ziconotide from plasma extract ( top ) and matrix blank ( bottom ). ............126 326 Full scan mass spectra of modified ziconotide from plasma extract ( top) and matrix blank ( bottom ). .................................................................................................................127 41 Response of oxidized ziconotide from QC samples analyzed immediately and with time delay. ........................................................................................................................148 42 Changes in ziconotide concentration vs. time for test solutions stored at 50 C and kept in contact with air. ..................................................................................................149 43 Plot of ln [Ziconotide]undiluted (in molarity) versus time. .................................................151 44 Changes in ziconotide concentration vs. time for 0 to 216 hr to extrapolate time to reach 50% ziconotide. .....................................................................................................152 45 Extracted ion chromatograms of ( A ) oxidized ziconotide, ( B ) ziconotide and ( C ) conotoxin GVIA from a test solution sample stored at 50 C for 144 hr. ......................154 46 Full scan mass spectra of ( A ) oxidized ziconotide, ( B ) ziconotide and ( C ) conotoxin GVIA. ..............................................................................................................155 47 Summed response of the +3 and +4 charge states of ziconotide and oxidized ziconotide in test solutions stored at various times at 50 C. ........................................156

PAGE 14

14 Abstract of Dissertation Presented to the Graduate School of the Uni versity of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy STUDIES OF INTACT AND MODIFIED ZICONOTIDE BY LIQUID CHROMATOGRAPHY TANDEM MASS SPECTROMETRY By Jhoana Ave nda o Mendoza August 2009 Chair: John R. Eyler Major: Chemistry Zico notide is the synthetic version of conopeptide MVIIA, a 25amino acid cone snail toxin from Conus magus Its cysteine residues form 3 disulfide bridges, while its argini ne lysine residues and amidated C termina l make it highly basic. It is a potent analgesic recently approved for the treatment of intense refractory pain in humans. The development of an analytical method for the qualitative and quantitative determination of this peptide in complex matrices is, therefore, of interest to many. High performance l iquid c hromatography ( HPLC ) tandem mass spectrometric (MS/MS) analyses of intact and modified ziconotide were carried out in this study. The chromatographic performance of various columns for r eversed phase HPLC separation of ziconotide was investigated. T he peak shape and retention of ziconotide on the columns used were compar ed The effect s of solvent composition on the charge state distribution and response of ziconotide were investigated. Optimiz ation of various mass spectrometric parameters such as spray voltage, sheath gas and auxiliary gas flows was also conducted. Analysis of intact ziconotide showed that the ext ensive crosslinking from multiple disulfide bridges reduces the structural inform ation that can be derived from the MS/MS

PAGE 15

15 spectrum of ziconotide. Modification of the analyte by reducing the disulfide bonds and alkylating the resulting sulfhydryl groups eliminate cross linking and increase the probability of fragmentation during collisioninduced dissociation (CID). Fragment ions resulting from typical peptide bond cleavage (e.g., b and y ions) were observed after modification of ziconotide and the internal standard, conotoxin GVIA. The limit of quantitation obtained from the modifie d peptide experiments was 0.500 ng on column. The HPLC MS method developed for the analysis of intact ziconotide was successfully validated. The validated method was then used to evaluate the degradation of ziconotide in a test solution similar in compos ition to the commercially available drug formulation, Prialt (Elan Pharmaceutical, Inc.) The plot of the natural logarithm of ziconotide concentration versus storage time at 50 C and half life calculations indicated that the degradation of z iconotide cou ld be a first or pseudo first order reaction. Results pertaining to the rate of degradation of a drug substance provide useful information for proper handling and storage of the ziconotide drug solution.

PAGE 16

16 CHAPTER 1 INTRODUCTION Background of Study Cono toxins and Ziconotide Conopeptides or conotoxins are small peptide toxins produced by cone snails for prey capturing and defense.1, 2 There are about 100 different conotoxins in the venom of each of the 500 living species of Conus with very little sequence similarity between species.2 5 These unique conopeptides consist of 10 to 30 amino acids and most have multiple disulfide bonds.2 5 They are broadly categorized into two groups based on the number of disulfide bonds present and then further subdivided based on their pharmacological targets.4 Conopeptides which act selectively at a number of sites, including sodium and calcium channels and nicotinic acetylcholine receptors, have been reported.3, 6 The use of these conopeptides as tools in pharmacological research and as potential therapeutic agents in humans is currently being explored.6 8 Ziconotide (MW = 2.6 kDa) is the synthetic version of conopeptide MV IIA, a 25amino acid cone snail toxin from Conus magus (Figure 11).9 11 Its c ys teine residues form 3 disulfide bridges and are arranged in a 4 loop cysteine scaffold which is a characte ristic feature of most conotoxins, while its numerous lysine and arginine residues and amidated C terminal make it highly basic.2, 12 It is a potent peptide analgesic recently approved for the treatment of severe chronic pain in humans when administered intrathecally .10 Toxins belonging to the omega family of conopeptides block cell membrane calcium channels and ziconotide specifically blocks the N type voltage sensitive calcium channels.9, 11, 13 Analysis of these highly diverse cone snail toxins has been facilitated by reversed phase high performance liquid chromatography (RP HPLC ) with ultravioletvisible (UV vis)

PAGE 17

17 spectrophotometry2, 14, 15 and mass spectrometry (MS) for detection.1, 8, 16, 17 The presence of multiple disulfide bonds and post translational modifications in several of these conopeptides require s derivatization meth ods and tandem mass spectrometric (MS/MS) analysis for enhanced structural characterization.1, 1820 Nuclear magnetic resonance (NMR) spectroscopy has also been employed in the structural determ ination of conotoxins.8, 17, 21 However, the detection and quantitation of ziconotide, the conotoxin of interest in this study, in biological samples has only been carried out using a radioimmunoassay method.13 R adioimmunoassay technique s involve competition between unlabeled antigen in a sample matrix and a known amount of radiolabeled antigen for antibody binding sites.22 As the concentration of the unlabeled antigen increases, more radiolabeled antige n is displaced and the ratio of antibodybound radiolabeled antigen to unlabeled antigen decreases. The concentration in an unknown sample is quantified against the inhibition obtained with standard solutions containing known amounts of antigen. A 3 kDa peptide analog, conotoxin GVIA (Figure 12 ), was chosen as the internal standard (IS) in the ziconotide analysis discussed in the succeeding chapters. High Performance Liquid Chromatography Reversed phase high performance liquid chromatography has become an essential tool in the separation and analysis of peptides from complex sample matrices. RP HPLC separates peptides based on hydrophobicity with the non polar compounds retained longer in the stationary phase and eluted later than polar compounds.23 Eluate s from the HPLC column can then be directly introduced into a mass spectrometer and analyzed, a technique called online LC/MS.24 There are several advantages to separating a mixture by RP HPLC prior to mass spectrometric analysis. These include easier data i nterpretation from generation of mass spectra of single components and improved sensitivity through concentration of a large volume of dilute sample

PAGE 18

18 upon injection onto the column and separation of analytes from salts, buffers and other contaminants.25 The separation efficiency of RP HPLC is influenced by the type of resin, size of column and the length and slope of the gradient used.23 As the need to analyze limited volume of samples emerged, small diameter columns that employ reduced flow rates have become popular for RP HPLC with current methods using flow rates of 500 nL/min or less and column diameters between 50 100 m .26, 27 Electrospray Ionization The development of atmospheric pressure ionization techniques such as electrospray ionization (ESI) and chemical ionization (CI) has greatly facilitated LC/MS applications to bioanalytical studies.28 Electrospray ionization, in particular, involves the generation of charged droplets from a solution sprayed through a highvoltage capillary (Figure 1 3) .29 As the solvent evaporates upon application of drying gas, heat or both a nd the droplets shrink in size, analyte ions shrink and split through coulombic repulsion until they are transformed into gas phase ions.28, 29 The introduction of this technique has allowe d for the analysis of most large, nonvolatile and thermally labile compounds such as various classes of biomolecules directly from the liquid phase.30, 31 The generation of multiplycharged ions also enables mass spectrometers with limited mass to charge ( m/z ) ranges such as ion traps, to analyze high molecular weight molecule s.31 One of this techniques most practical attribute s is its ability to couple MS and liquid separation techniques. Signal enhancement from the use of small diameter capillary columns and low flow rates, coupled with improvements in sensitivity and selectivity provided by MS techniques, have facilitated the detection of peptides present at attomole level s in a complex sample mixture.32

PAGE 19

19 Quadrupole Ion Trap Mass Spectrometry The operation of the linear ion trap is based on the theory of the 3D quadrupole ion trap.33 The linear ion trap is composed of 4 parallel hyperbolic shaped rods, divided into 3 sections as opposed to the configuration of the 3D ion trap which consists of a ring and 2 endcap electrodes.34 For trapping of ions, a radio frequency (RF) electric field is appl ied radially and a static electric field using DC voltages is applied axially.34 The application of these voltages creates a homogenous field throughout the trapping region. These modifications have improved the ion storage capacity and trapping efficiency of the linear ion trap when compared to a 3D trap.35 Once ions are trapped, a particular m/z can be isolated and fragmented by applying the ap propriate supplementary frequency. Like its predecessor, the 3 D ion trap, mass analysis in the linear ion trap involves ejecting the trapped ions in an increasing m/z order by linearly ramping the RF amplitude. However, the trapped ions are ejected in the radial direction (instead of axially in the 3D trap) along the x axis, through the 2 parallel slots in the center of the xrod pairs.34 Ejected ions are then detected by a conversion dynode and an electron multiplier placed on each side of the trap. The use of a dual detector system ensures detection of the maximum number of ions, thus improving sensitivity. The enhanced detection efficiency together with the improved storage capacity and trapping efficiency of the linear ion trap offer advantages such as greater sensitivity, increased dynamic range and improved MSn performance. Objectives of Research The development of an HPLC ESI MSn method to detect ziconotide in complex biological matrices is a very promising alternative to radioimmunoassay techniques. The initial goals of this research project were to optimize the HPLC ESI MS method and to evaluate its figures of merit for the analysis of intact and modified ziconot ide in biological samples The evaluation of

PAGE 20

20 a stability indicating assay for ziconotide in commercial type formulations using the validated HPLC ESI MS method for the intact peptide was carried out when method limitations were encountered for the origina l application. Analysis of Intact Ziconotide The optimization of various HPLC and mass spectrometric parameters such as solvent composition, column chemistry, spray voltage, auxiliary and sheath gas flows for the a nalysis of intact ziconotide is discussed in Chapter 2. The use of small diameter columns and low flow rates to improve t he sensitivity of the method is explored. The figures of merit and limitations of the HPLC ESI MS method developed are also discussed in this chapter. Analysis of intact ziconotide showed that the extensive cross linking from its multiple disulfide bridges reduces the structural information that can be derived from its MS/MS spectrum. Analysis of Modified Ziconotide Modification of ziconotide by reduction and alkylation of its disulfide bonds eliminates crosslinking and yields fragment ions that are more structurally relevant. Optimization of the parameters affect ing reduction and alkylation is presented in Chapter 3 together with the figures of merit and limitations of th e HPLC ESI MS method for the modified conopeptide. Application of the reduction and alkylation method to ziconotide supplemented in horse plasma was also explored and is discussed in this chapter. Development of a Stability Indicating Assay for Ziconotide in Commercial Type Formulations The HPLC ESI MS method for the analysis of intact ziconotide was validated and evaluated as a stability indicating assay for ziconotide in test solutions similar in composition to the commercially available drug formulation, Prialt (Elan Pharmaceutical, Inc.). The results of

PAGE 21

21 the validation study and the determination of a possible rate of reaction for the degradation of ziconotide exposed to harsh oxidative conditions are discussed in Chapter 4.

PAGE 22

22 CKGKGAKCSRLMYD CKGKGAKCSRLMYD CCTG CCTG SCRSGKC SCRSGKC NH2 NH2 15 20 25 1 5 10 CKGKGAKCSRLMYD CKGKGAKCSRLMYD CCTG CCTG SCRSGKC SCRSGKC NH2 NH2 15 20 25 1 5 10 CKGKGAKCSRLMYD CKGKGAKCSRLMYD CCTG CCTG SCRSGKC SCRSGKC NH2 NH2 15 20 25 1 5 10 Figure 11. 3D struc ture of ziconotide ( left ) and its amino acid sequence ( right ). The N terminus on the 3D structure is shown in red, C terminus in blue, and disulfide bonds in gold.36 The color scheme used for th e amino acids was matched to that of the 3 D structure.

PAGE 23

23 1 5 1 10 15 20 25CKSXGSSCSXTSYNCCRSCNXYTKRCY CKSXGSSCSXTSYNCCRSCNXYTKRCY NH NH2 21 5 1 10 15 20 25CKSXGSSCSXTSYNCCRSCNXYTKRCY CKSXGSSCSXTSYNCCRSCNXYTKRCY NH NH2 2 Figure 12. Amino acid sequence of conotoxin GVIA The disulfide bonds are shown in gold and X denotes hydroxyproline residues.

PAGE 24

24 Heated Metal Transfer Capillary Heated Metal Transfer Capillary Figure 13. Schematic of the electrospray ionization process. This ionization technique produce s gaseous analyte ions from a liquid flow introduced into a small capillary that is at a potential difference relative to a counter electrode from which ions are generated then sampled into the mass analyzer. A dapted from Cech and Enke29

PAGE 25

25 CHAPTER 2 HPLC ESI MS/MS ANALYSIS OF IN TACT ZICONOTIDE Introduction Ziconotide, the synthetic version of conotoxin MVIIA, is a potent peptide analgesic that has recently been approved for the treatment of severe chronic pain in humans.10 The development of an analytical method for determining this peptide in complex matrices, like biological fluids is of interest to many. At present, the only method available for the detection and quantitation of ziconotide in biological samples is radioimmunoassay which involves the use of a ziconotide specific antibody and a radioiodinated ziconotide derivative.13, 37 Radioimmunoassay is a sensitive and specific technique used mainly for the detection of hormones which are present in micromolar to picomolar range in the blood.38, 39 However, t he cost of generating specific antibodies and radioactive labeled antigens and safety issues involved in handling radioactive substances present some limitations to RIA. Reversed phase high performance liquid chromatography (RP HPLC) coupled to mass spectrometry (MS) has been widely employed for the analysis of peptides and proteins in biological samples.27, 32 The size of RP HPLC columns and flow rate s have been greatly reduced over the years to allow for the analysis of samples in limited amount. The lower flow rate required f or small diameter columns reduces the amount of solvent that r eaches the detector, resulting in a higher peptide concentration in the peak and hence, improved sensitivity.40 The development of an HPLC MS method for the qualitative and quantitative determination of ziconotide will be discussed in this chapter. HPLC tandem mass spectrometry (M S/MS) analysis of ziconotide will be presented. The optimization of various HPLC and mass spectrometric parameters will also be discussed together with the figures of merit of the method developed.

PAGE 26

26 Experimental Section Chemicals and Reagents The conopeptides conotoxin MVIIA (0.1 mg, 95% purity by HPLC) and conotoxin GVIA (0.1 mg, 98.3% purity by HPLC) were purchased from Sigma Aldrich (St. Louis, MO). Glacial acetic acid (HAc), trifluoroacetic acid (TFA), formic acid (FA) and HPLC grade acetonitri le (ACN) were obtained from Thermo Fisher (Fair L awn, NJ). Water used for solution preparation was purified by RiOs reverse osmosis and Milli Q polishing systems (Millipore, Billerica, MA). Water used for peptide solution preparation was also vacuum degassed to remove dissolved oxygen in order to minimize possible oxidation of ziconotide. Instrumentation A Surveyor Autosampler/HPLC system interfaced to a Thermo Finnigan (San Jose, CA) linear ion trap (LTQ) mass spectrometer equipped with an Ion Max elec trospray ionization (ESI) source was employed for HPLC MS/MS analyses. The HPLC system includes a quaternary, low pressure mixing pump with vacuum degassing, an autosampler with a temperature controlled tray and a column oven. The ESI needle is oriented a t 45o with respect to the axis of the ion transfer capillary. This configuration and the presence of the ion sweep cone prevent large droplets and particulates from entering the ion transfer capillary. A diagram showing the ion source interface, ion guide, mass analyzer and the dual detector system of the LTQ mass spectrometer is shown in Figure 21. The mass spectrometer and all peripheral components are controlled through the Xcalibur software (Thermo Scientific, San Jose, CA). Solvent flow from the HPLC system was split with a stainless steel tee (Upchurch Scientific, Oak Harbor, WA) in order to a chieve optimal flow rates ( 5 to 7 L/min) for the small diameter column used for low flow ESI MS analysis of ziconotide. A 0.5 m pre column filter (Upchur ch Scientific, Oak Harbor, WA) was placed before the split tee to prevent small particles

PAGE 27

27 from clogging the system. The f l ow coming out of the other end of the tee was connected to the divert valve on the LTQ mass spectrometer while the flow on the other end was sent to waste. The flow rate into the column was adjust ed by connecting 15 or 50 cm of a 50 m inner diameter (i.d.) Peeksil tubing (Upchurch Scientific, Oak Harbor, WA) at the waste end of the split tee. This instrumental set up is illustrated in Figure 2 2. B lack PEEK tubing (1/16 x 0.004) with a smaller i.d. than the red PEEK tubing (1/16 x 0.005) used before the column, was connect ed post column to minimize peak br oadening. This PEEK tubing was purchased from Thermo Fisher (Fair Lawn, NJ ). Optimization of Mass Spectrometric Parameters Mass calibration was performed with a calibration solution consisting of caffeine, MRFA (L Methionyl arginyl phenylalanyl alanine) and Ultramark 1621. The calibration solution was prepared according to the recommended procedure outlined in the Finnigan LTQ Getting Started manual and infused directly into the mass spectrometer at 3 L/min .41 Mass spectrometric parameters were optimized using the automatic, semi automa tic and manual tuning features of the LTQ. A 10 ng/L ziconotide solution in 40:60 ACN: H2O with 0.1% HAc: 0.01% TFA was used as the tune solution for experiments involving conventional electrospray. Conventional electr ospray in this case, refers to ana lyte solution flow rate s of 75 or 200 L/min The ziconotide solution was placed in a 250 L Hamilton syringe (Reno, NV) and flow from the syringe (10 to 20 L/min) was teed in with 75 or 200 L/min of 40:60 ACN: H2O with 0.1% HAc: 0.0 2% TFA before introd uction into the mass spectrometer. A 1 ng/L ziconotide solution in 50:50 ACN: H2O with 0.1% formic acid was used as the tune solution for low flow experiments. The tune solution, in this case, was infused directly into the mass spectrometer at 5 L/min. Signal intensity from the total ion current (TIC) or specific ions of ziconotide was monitored during tuning. Manual tuning was first performed to determine the optimum ESI

PAGE 28

28 probe position and gas flows. Automatic tuning was then employed to optimize the voltages for the capillary, tube lens and various multipoles and lenses making up the ion optics of the LTQ. Semi automatic tuning was used to determine if further improvement in signal could be achieved by moderate adjustments in the voltages obtained from automatic tuning. Optimization of HPLC Parameters Isocratic elution with solvents containing various concentrations of methanol with 0.5% HAc in water ( i.e., 10:90, 20:80, 50: 50 and 70:30 MeOH:H2O mixtures) was employed for the HPLC low flow ESIMS analysis of ziconotide. The peak shape and retention time obtained with the various solvents were evaluated. G radient elution with 0.5% HAc in 10:90, 20:80 and 30:70 MeOH: H2O mixtures a t the start of the gradient was also performed. Chromatographic separation was achieved with a Targa C18 column ( 0.3 x 100 mm, 5 m particles 120 pore size) from Higgins Analytical (Mountain View, CA). The effect s of using acetonitrile with 0.1% FA and other acids such as 0.01% TFA and a combination of HAc and TFA as the organic mobile phase on the peak shape and retention of ziconotide were also explor ed. The effects of varying the TFA concentration of the mobile phase on the charge state distribution and response of ziconotide and conotoxin GVIA were systematicall y investigated using conventional electrospray. A solution containing 1 ng/L each of ziconotide and the internal standard ( IS ) conotoxin GVIA, was used for the evaluation. Reversed phase HPLC was performed with an Atlantis dC18 column (2.1 x 150 mm, 5 m particles, 300 pore size) wit h a guard column (2.1 x 20 mm, 5 m particles) from Waters Incorporation (Milford, MA). The following TFA concentrations were used: 0.01, 0.02, 0.025, 0.03, 0.05, 0.075 and 0.1%. The effects of varying the acetic acid composition (0.01, 0.025, 0.05, 0.075 and 0.1%) of the mobile phase while keeping TFA at 0.01% were also investigated.

PAGE 29

29 The peak shape and retention time of ziconotide after elution from the Atlantis dC18 column using 0.1% HAc: 0.02% TFA as the acid com ponent of the mobile phase were compared to those from a Jupiter Proteo C12 column (Phenomenex, Torrance, CA). The Jupiter Proteo C12 column (2 x 150 mm, 4 m particles 90 pore size) was used with a Security Guard cartridge (4 x 2 mm, C12) to protect the analytical column. The chromatogram of ziconotide eluted with 0.1% formic acid i n the mobile phase from a new Jupiter Proteo C12 column was also compared to that obtained with 0.1% HAc: 0.02% TFA. A smaller diameter Atlantis dC18 column ( 1 x 150 mm, 5 m particles, 300 pore size) that required lower flow rates (75 L/min) of analyte solution was used to determine if improvement in the a mount of ziconotide detected c ould be achieved. Preparation of Calibration Standards Serial dilution of a 40 ng/ L aqueous ziconotide solution was carried out to prepare 11 calibration standards with concentrations ranging from 0.05 to 10 ng/L. A fixed amount of an aqueous internal standard, conotoxin GVIA, was added to each standard to achieve a final concentrat ion of 1 ng/L in solution. The calibration standards were prepared fresh in 1:99 ACN: H2O with 0.1 % HAc: 0.01% TFA before each analysis. Six calibration curves from separate sample dilutions were prepared and analyzed. Quality control (QC) samples with ziconotide concentrations at 0.06, 0.6 and 6 ng/L were prepared in duplicate and analyzed with each calibration curve. HPLCESI MS /MS Analysis Manual l oop injections of 2 L of 1 ng/L ziconotide in 1:99 MeOH:H2O with 0.5% HAc were carried out with is ocratic a nd gradient elution for analyses involving low flow electrospray Isocratic elution with acidified solvents was performe d for 30 min. Pre mixed solvents were

PAGE 30

30 used for isocratic elution to minimize variability in response caused by inefficient onl ine solvent mixing at low flow rates. Gradient elution with various acidified solvents was also employed for the analysis of ziconotide. The gradient included a 1 min aqueous hold at the start of the run, increase in percentage of the organic solvent fro m starting conditions to 90% for the next 10 min, and a 4 min organic hold. The column was then allowed to equilibrate to starting conditions for 15 min. The post column flow rate after splitting was 7 L/min. The column oven was set to 30 C while the autosampler tray was maintained at 12 C. Gradient elution with 0.1% HAc: 0.02% TFA in water as solvent A and 0.1% HAc: 0.02% TFA in ACN as solvent B, with solvent B raised from 2 to 70% in 9 min was carried out for conventional electrospray experiments T his was followed by a 6.5 min column equilibration to initial solvent composition. A flow rate of 200 L/min was used. Using the partial loop injection mode, 10 L of the sample was loaded into the column. External and internal syringe needle washes with 0.1% HAc: 0.01% TFA in 50:50 ACN: water were incorporated in the autosampler method to eliminate carry over. The column oven and autosampler tray temperatures were the same as those used in the low flow experiments. Similar gradient and instrument condi tions were employed in the evaluation of other solvent systems. Electrospray ionization in the positive mode was employed for both conventional and low flow experiments. Spray voltage for the fused silica capillary used during low flow electrospray analys es was set at 4.2 5kV while the capillary voltage was at 40 V. The capillary temperature was maintained at 16 0 C while a s heath gas flow of 25 arbitrary units was used A lower spray voltage, 3.75 kV, was used for the metal capillary installed for analys es at higher flow rate to prevent arcing. T he capillary voltage was set t o 45 V. The capillary temperature was maintained at 300 C. Sheath gas flow was set to 40 and that of the auxiliary gas to 1 0 arbitrary units.

PAGE 31

31 Mass spectra for low flow electrospray experiments were acquired for 30 min. The LTQ was programmed for full scan MS then full scan MS/MS ( m/z 400 1400) of m/z 661, the quadruplycharged ion of ziconotide. The precursor ion isolation width in full scan MS/MS mode was set to 3 m/z with col lision energy (CE) at 30% and activation energy (q) at 0.250. The LTQ mass spectrometer w as set to acquire for 10 min for conventional electrospray analyses Sample from the chromatographic column during the first 2 min and last 0.5 min was diverted to w aste. The LTQ was programmed for full scan MS ( m/z 500 1600) then full scan MS/MS ( m/z 500 1400) of m/z 661 and m/z 760, the +4 ions of ziconotide and con otoxin GVIA, respectively. The rest of the parameters used for full scan MS/MS of the precurso r ions were similar to those employed in low flow experiments MS Data Analysis Extracted ion chromatograms (EICs) for the various ions of ziconotide and the IS, conotoxin GVIA were generated using the Xcalibur software Chromatograms were processed using a 3point boxcar smoothing calculation which is available with the software The peaks for the various charge states of ziconotide and the IS, as well as that of oxidized ziconotide, were manually integrated to determine the area for each. The sum of the contributions of the +3 and +4 ions of oxidized ziconotide was used for its EIC. The ratio of the peak area of the +4 ion of ziconotide to that of the IS was plotted against ziconotide concentration to generate the calibration curve. Weighted lea st squares linear regression (WLSLR) was used for data fitting, with 1/x2 as the weighting scheme of choice.42, 43 GraphPad Prism Version 5.01 (San Diego, CA) was used for data processing. The need to apply weighting was determined by following the statistical evaluation scheme employed by Almeida et al. .43 The evaluation was conducted to establish i f a particular data set

PAGE 32

32 exhibited equal variance across the whole range of values being considered, a conditio n termed as homoscedasticity and to justify the use of weighting if homoscedasticity wa s not obeyed.43 An F test (see Equation 2 1) and a plot of yresidual (difference between experimental and fitted area ratios of the +4 ions of ziconotide to c onotoxin GVIA) versus concentration were evaluated to determine if homoscedasticity was obeyed. Fexp = s2 2/s1 2 (2 1) Ftab (f1, f2; 0.99) where s1 2: variance of the lowest calibration standard s2 2: variance of the highest calibration standard Ftab: value from F table at 99%confidence level f1, f2: degrees of freedom, equal to n 1 n: number of replicates The choice of the appropriate weighting scheme, on the other hand, is based on plots of percent relative error (%RE) versus concentration and the %RE sum.43 The percent relative error compares the concentration calculated (Ccalc) from the regression equation obtained for each weighting factor, wi, with the nominal standard concentration (Cnominal), see Equation 2 2. The %RE sum is defined as the sum of the absolute values of %RE and is found to be a useful indicator of goodness of fit of the resulting regression line after application of a weighting factor.43, 44 % RE = [(Ccalc Cnominal)/ Cnominal] x 100 (2 2) Results and Discussion Optimized Mass Spectrometric Parameters for Ziconotide Analysis When tuning the LTQ to optimize the signal for ziconotide, manual tuning to determine the best ESI probe position and optimum gas flows was first carrie d out. The total ion current for ziconotide was monitored while adjusting the ESI probe to a position that yielded the highest

PAGE 33

33 signal for ziconotide. The ESI probe was positioned closer to the ion transfer capillary for optimum response during low flow e lectrospray analyses. The same procedure of monitoring the TIC for ziconotide was carried out to determine the optimum gas flows A sheath gas flow of 40 (arbitrary units) and auxiliary gas flow of 20 (arbitrary units) showed the highest signal for zicon otide for experiments conducted at 200 L/min. The use of auxilliary gas was found to lessen the contribution of TFA adducts as seen in Figure 23. Nitrogen wa s applied as an inn er coaxial gas when sheath gas was used.41 It facilitate d nebulization of the sample solution. Auxiliary gas aid ed in focus ing the spray plume and improved desolvation.41 In contrast, auxiliary gas wa s not required for low flow electrospray. Com parisons of the normalized response of ziconotide (average of 200 scans) at various spray voltages, tube lens offset voltage s and capillary temperatures were carried out to determine the optimum setting for these parameters. The effect s of the tube lens offset voltage on the response and charge state distribution of ziconotide are illustrated in Figure 24. Comparison of these mass spectra and the normalized response of the +3 and +4 ions of ziconotide at m/z 661 and m/z 881, respectively, established 160 V as the optimum tube lens offset voltage for conventional electrospray. The response of m/z 661 was slightly higher at 100 V but contribution s from TFA adducts were more significant. Lower normalized response s were obtained for both the +3 and +4 ions of ziconotide at 200 V. The abundance of the higher m/z ions (+2 and +3) was observed to be greater than that of the lower m/z ions (+4 and +5) at higher tube lens offset voltage Tuning for the tube lens offset voltage ensures maximum transmission of io ns from the electrospray ion source to the mass analyzer.4 5 Comparison of the results obtained from various capillary temperature s ( i.e. 250, 275 and 300 C) showed an increase in the abundance of TFA adducts as the temperature was decreased. In order to achieve sufficient

PAGE 34

34 desolvation of the sample and minimi ze the contribution of TFA adducts, a capillary temperature of 300 C was used. A lower temperature was required for analyses involving low flow electrospray. Optimized HPLC Parameters for Ziconotide Analysis Effect of Various Solvent Systems and Flow Rat es on Ziconotide Response Isocratic elution of 1 ng/L ziconotide in 1:99 MeOH:H2O with 0.5% HAc using 70:30 MeOH: H2O with 0.5% HAc yielded the most satisfactory peak shape for ziconotide compared to acidified 10:90, 20:80 and 50:50 MeOH:H2O mixtures (Fig ure 25). Peaks eluted with 10:90, 20:80 and 50:50 MeOH:H2O mixtures were retained in the column extensively. The retention times of ziconotide peaks eluted with 0.5% HAc in 50:50 and 70:30 MeOH:H2O mixtures follow the expected order of elution based on the organic composition of the solvent. A higher organic composition of the mobile phase facilitates earlier elution of peptide peaks and this was observedwith the ziconotide peaks eluted with 0.5% HAc in 50:50 and 70:30 MeOH:H2O mixtures. This was not t he case, however, with the peaks obtained from 0.5% HAc in 10:90 and 20:80 MeOH:H2O mixtures. The retention time for ziconotide peak eluted with acidified 10:90 MeOH:H2O was shorter than that of acidified 20:80 MeOH:H2O. Inspection of the mass spectra of peaks eluted with 0.5% HAc in 10:90 and 20:80 MeOH:H2O mixtures (bottom panel in Figure 26) showed greater signal of background ions than ziconotide. The increased background signal could have caused the incorrect elution order of ziconotide observed wi th these solvents. Gradient elution of ziconotide at low flow rates with solvents containing 0.5% HAc and various methanol concentrations at the start of the gradient was also conducted to evaluate the effect of gradient elution on the peak shape and ret ention of ziconotide on the Targa column (0.3 x 100 mm, 5 m particles, 120 pore size). Peaks that sh owed signs of tailing were obtained even with gradient elution (F igure 2 7). The mechanism of peptide and protein retention is based

PAGE 35

35 on their adsorption to hydrophobic surfaces such as the stationary phase in C12 or C18 columns and desorption from the surface once a specific organic composition is reached.46 Retention of peptides and proteins on columns changes abruptly when the concentration of organic solvent required to desorb them from the hydrophobic surface is reached and this results to sharper peaks. The absence of narrow peaks from the gradient elution of ziconotide suggests potential silanol interactions with the stationary phase of the column which could be causing the tail ing observed. The mass spectra of ziconotide peaks from the gradient elutions showed the +4 ion ( m/z 661) as the most abundant ion. Interest in the methanol acetic acid solvent system was influenced by efforts to repeat the results of HPLC low flow ESI MS/MS experiments conducted at the University of North Texas (UNT) Health Science Center as a result of collaboration with the University of Florida Racing Laboratory. Good peak shape (minimal tailing) and retention of ziconotide and one of its MS/MS prod uct ions were obtained with loop injections of 3 L of 5 ng/L ziconotide, gradient elution and low flow electrospray analysis using the same Targa column and split flow system (data not shown). Loop injection of a 10 ng/L ziconotide solution yielded a peak with an area twice as much as that of the peak for 5 ng/L. These promising preliminary results led us to pursue and repeat the experiments that were conducted at UNT but differences in the HPLC system and instrumental set up details could have prevented a successful repetition of experimental results. Isocratic elution using 15:85 ACN:H2O with 0.1% FA, 0.01% TFA and 0.1% HAc: 0.05% TFA as acid components of the mobile phase was also explored. Improvement in peak shape was observed with the addition of TFA to the mobile phases but its ionization suppression effects were observed with a decrease in peak area of ziconotide (Figure 28). The corresponding mass

PAGE 36

36 spectra of ziconotide peaks from the isocratic elution with 15:85 ACN:H2O containing various acid components exhibited differences in the identity and abundance of ions present (Figure 2 9 ). The use of 0.1% HAc: 0.05% TFA as acid component of the mobile phase showed the +3 and +4 ions of ziconotide as the abundant ions in the mass spectrum and t he signal contribution from TFA adducts to be minimal. The inability to achieve significant retention of ziconotide on the Targa column and the challenges imposed by modifying a standard HPLC system to carry out gradient elution at low flow rates led us to study and carry out experiments at high flow rates to evaluate how ziconotide behaves at conditions compatible with the HP LC MS system available. There was a need to achieve significant column retention because polar components of biological samples like salts and other unwanted contaminants elute early during a gradient run. Longer retention time of the analyte of interest is desired to minimize contamination and/or potential signal suppression from these other sample components mentioned. The use of l ow flow r ates is often not compatible with many standard HPLC systems, especially in a gradient mode. The contribution of extracolumn volumes in the injector and connecting tubing to band spreading also becomes more significant at low flow rates.23 The succeeding sections, therefore, pertain to experiments conducted at high flow rates ( i.e. 75 and 200 L/min) With the tendency of ziconotide to be retained in the column and the ability of TFA to improve peak shape, solvent systems with varying concentrations of TFA and varying concentrations of acetic acid in a solvent mixture with TFA were evaluated for their effects on the charge state distribu tion and response of ziconotide and conotoxin GVIA. A solution containing 1 ng/L each of ziconotide and the IS was used for the evaluation. The optimized

PAGE 37

37 MS parameters were used in the analyses. The conopeptide conotoxin GVIA was chosen as the IS because of its similar ESI MS /MS characteristics to ziconotide Results from the various concentrations of TFA ( i.e., 0.01, 0.02, 0.025, 0.03, 0.05, 0.075 and 0.1%) us ed are i llustrated in Figure 2 10. The predominant charge state observed for ziconotide with TFA in the mobile phase was the +3 ion ( m/z 881). The +4 ( m/z 661) and +5 ( m/z 529) ions were also present. A TFA adduct at m/z 918 was also observed in the mass spectrum of ziconotide. The highest peak area for the +3 ion of ziconotide was observed at 0.025% TFA. There was a good trend in the peak area across the TFA concentrations used, increasing as 0.025% was approached and decreasing towards high concentrations. High concentrations of TFA are known to cause signal suppression when used with an electrospray interface.47 The +3 charge state ( m/z 1014) was also the most predominant ion for conotoxin GVIA (see Figure 2 11). The +4 ion ( m/z 760) and a TFA adduct ( m/z 1050) were also observed. The +5 ion ( m/z 609) was only detected with 0.01% TFA in the mobile phase. The highest signal for the +3 ion of the IS was obtained with 0.025% TFA as well. The use of HAc and TFA mixture as the acid component of mobile phases for HPLC MS studies of conopeptides has been reported.48, 49 The effects of varying the HAc composition (0.01, 0.025, 0.05, 0.075 and 0.1%) of the mobile phases while keeping TFA at the lowest concentration, 0.01%, were investigat e d in this study (see Figure 2 12). The highest response for ziconotide with minimal contribution from TFA adducts were found at 0.1% HAc. The average peak a rea s for the +3 and +4 ions was comparable at this HAc concentration and the only TFA adduct observed was at m/z 918. Significant contribution from other TFA adducts at m/z 689 and m/z 955 was observed at HAc concentrations lower than 0.1%. Similarly, the h ighest response for conotoxin GVIA with

PAGE 38

38 minimal contribution from TFA adducts was observed a t 0.1% HAc also (see Figure 2 13). The +3 and +4 charge states were the predominant ions in the mass spectra of the peptide at this HAc concentration. The signa l for the only TFA adduct observed at m/z 1013 decreased as the HAc concentration was increased. Based on the evaluation of the mobile phase acid composition conducted and considerations for minimizing signal suppression and system contamination with TFA, 0.1% HAc with 0.01% TFA proved to be a good choice for the acid component of the mobile phases. The use of a smaller diameter Atlantis dC18 column (1 x 150 mm, 5 m particles, 300 pore size) was explored to determine if improvements in sensitivity c ould be achieved. As shown in Figure 214, the lowest amount of ziconotide detected on the small diameter column (50 pg), was greater than the lowest amount detected on the larger diameter column (20 pg). There was no enhancement in detection sensitivity observed with the use of small diameter column, in this case. The corresponding mass spectra of the ziconotide peaks detected are presented in Figure 215. Effect of Column Chemistr y on Peak Shape and Retention of Ziconotide The need to use a column with m inimal silanol group interaction became evident as ziconotide exhibited a tendency to be retained in the column due to its basicity. Electrostatic interactions of residual silanols with cationic species cause peak tailing which may result in loss of chrom atographic resolution.50 Analysis of ziconotide with a new Jupiter Proteo C12 column using 0.1% HAc: 0.01% TFA as the acid component of the mobile phases showe d a peak with significant tailing (Figure 2 16). Its peak shape was improved when the TFA concentration was increased to 0.02%. With the reduction of peak tailing observed with 0.02% TFA, 0.1% HAc: 0.02% TFA was used as the acid component of the HPLC sol vents for the rest of the experiment conducted. The history of column use seemed to have an effect on the peak shape for ziconotide

PAGE 39

39 analyzed using the Atlantis dC18 column. Good peak shape was not obtained with a new Atlantis dC18 column as compared to one that had been previously used for analysis of plasma extracts of small drug molecules (data not shown). The peak shape obtained with this particular column even with 0.1% HAc: 0.02% TFA in the solvents was not as good as that from the Jupiter Proteo C1 2 co lumn (Figure 2 16 ). The use of formic acid was also evaluated on the Jupiter Proteo C12 column prior to any column treatment with TFA and results showed extensive retention of zico notide on the column (Figure 2 16). The importance of the purity of s ilica in the choice of column for ziconotide analysis was emphasized by comparing the peak shape obtained with the Jupiter Proteo C12 column and Atlantis dC18 column using the optimum solvent system The minimal tailing of the ziconotide peak obtained wit h the Jupiter Proteo C12 column using 0.1% HAc: 0.02% TFA in the HPLC solvents, compared to that from the Atlantis dC18 column, indicates that fewer residual silanol groups are present in this particular column. Improvement in peak shape was also observed with a combination of acetic acid and trifluoroacetic acid in the solvents compared to formic acid. Weighted Least Squares Linear Regression The use of weighted least squares linear regression counteracts the greater influence of the higher concentrations on the fitted regression line for analyses of drug molecules in biological samples.4244 The need to apply weighting, in this case, is supported by the residual versus concentration plot (Figure 217) and results from the F test (Table 2 1). The residual plot showed that the variance was not randomly distributed around the concentration axis. Greater variance was observed for ziconotide standards with higher concentrations compared to those at lower concentration s. Results from the interday assay of 6 calibration curves consisting of 11 standards with concentrations ranging from 0.05 to 10 ng/L were included in the plot. Data for 0.08, 0.3 and 0.8 ng/L standards were excluded to show the distribution of the l ow

PAGE 40

40 concentration calibrators more clearly. A data entry for the 3.0 ng/L standard was excluded due to pipetting error. The experimental F value (Fexp = 1.73 x 104) was also significantly higher than the tabulated value (Ftab = 10.97). These results suggested that homoscedasticity was not met and supported the need to use WLSLR. Figure 218 shows the percent relative error plots for unweighted (Model 1) and weighted (Models 2 5) regressions of ziconotide standards from the inter day assay. The %REs for concentrations less than 0.5 ng/L obtained with Model 1 were much greater than those obtained for the other models. The best accuracy and %RE distribution scatter especially at the low concentration range of the curve was obtained with models that employed 1/x2 (Model 2) and 1/y2 (Model 3) as the weighting factors. The regression parameters of the calibration curve generated for each weighting factor (with all the calibration standards included) and the respective sums of the relative errors are summari zed in Table 2 2. The most adequate approximation of variance was achieved with the weighting factor 1/x2 (Model 2) which provided the least %RE sum for the data. Considering this result and that of the %RE distribution scatter, 1/x2 was chosen as the weighting factor. Generation of the Response Curve for Ziconotide Eleven ziconotide standards with concentrations ranging from 0.05 to 10 ng/L, containing 1 ng/L of the IS, conotoxin GVIA, were analyzed using HPLC ESI MS/MS. The area ratio of the quadr uplycharged ions of ziconotide at m/z 661 to conotoxin GVIA at m/z 760 was plotted against ziconotide concentration to generate the calibration or response curve (Figure 2 19). Reagent blanks were not analyzed with the calibration curve standards due t o carry over issues especially after injection of the highest concentration calibrator. Solvent blanks of 50:50 ACN: H2O with 0.1% HAc: 0.01% TFA were injected instead. Calibration curves from six

PAGE 41

41 independent sample dilutions were prepared and analyzed, one set per day, for six different days. These calibration curves were employed for the determination of the figures of merit of the method and the validation study discussed in detail in Chapter 4. Quality control (QC) samples were analyzed with each calibration curve and the quantitation results for the QCs were used in the validation study. The regression parameters for the six response curves are shown in Table 2 3. A data entry for the 3.0 ng/L standard was excluded due to pipetting error. The s lope and y intercept values obtained were reproducible. The range of yintercepts obtained, however, did not contain zero. Possible suppression of ziconotides signal due to the IS signal could occur at very low concentrations and account for the negative yintercepts.51 Figure 220 shows the response of the +4 ion of the IS at m/z 760 with increasing ziconotide concentration. Analysis of back calculated concentrations of calibration standards yielded percent relative standard deviations (%RSDs) of less than 5.9% and %REs of l ess than 3.3% (Table 2 4). All the calibration curves satisfied the Food and Drug Administrations (FDA) requirement that at least 67% of the calibrators should have %REs and %RSDs of 15% except at the lower limit of quantitation (LLOQ) where a value of 20% is still considered passing. The limit of detection (LOD) and LLOQ were determined from peak area values for m/z 661 recorded from 10 blank analyses. The blank solution, in this case, is the solvent used in preparing the ziconotide standards, 0.1% HAc: 0.01% TFA in water. The average (Sbl) and standard deviation (sbl) of the blank measurements were calculated and incorporated into equation 2 3, where SLOD corresponds to the signal required for the LOD. SLOD = Sbl + 3(sbl) (2 3)

PAGE 42

42 The cons tant 3 in equation 23 represents a confidence level of detection of 89% or greater, and is the value recommended in literature.5254 The signal at LOD is then incorporated into equation 2 4 to determine the concentration at LOD (CLOD), wherein m is the slope of the calibration plot for the peak area of m/z 661 vs. ziconotide concentration (data not shown). CLOD = (SLOD Sbl)/m (2 4) The LOD for intact ziconotide using the HPLC ESI MS method developed was found to be 6 pg on column. The lower limit of quantitation was calculated as 10 times the standard deviation of the blank measurement, sbl (Equation 25). CL L OQ = (10 x sbl)/m (2 5) The LLOQ for intact ziconotide obtained by using the above equation was det ermined to be 20 pg on column. Higher values for LOD and L LOQ were obtained when the FDA guidelines were used to evaluate the ziconotide standards analyzed and results from injections of up to 0.001 ng/L ziconotide solution were considered. The LLOQ ac cording to the FDA is the lowest concentration that can be determined with acceptable accuracy (%RE 20%) and precision (%RSD 20%).55 The lowest calibrator, 0.05 ng/L, satisfied these criteria and was determined to be the LLOQ. This concentration corresponds to 0.5 ng on column, compared to the 20 pg obtained from blank measurements. A 0.01 ng/L ziconotide s olution was also analyzed with initial calibration curve runs but it did not satisfy the FDAs guidelines for precision and accuracy after performing calculations. The limit of detection was determined from the ziconotide concentration that exhibited a si gnal to noise ratio (S/N) of at least 3. The LOD using this definition corresponds to 20 pg on column. The presence of other components in the

PAGE 43

43 conopeptide solutions analyzed could have contributed to the difference s in LOD and LLOQ values obtained. The response of ziconotide was linear from 0.05 to 10 ng/L. Representative extracted ion chromatograms of ziconotide and the internal standard, conotoxin GVIA, are shown in Figure 221 with the corresponding mass spectra in Figure 222. A peak that eluted earlier than ziconotide was identified as its oxidized form. Its chromatogram and mass spectrum were also included in Figures 2 21 and 222, respectively. One of the identified degradation pathways of ziconotide involves oxidation of its methionine res idue.13, 56 Methionine is first oxidized to its sulfoxide and under strong oxidizing conditions, to a sulfone (Figure 223).57 Reagents like 95% performic acid oxidize methionine to methionine sulfone.58 The mass of the early eluting peak was found to be 16 Da more than the mass of ziconotide. The addition of 16 Da in mass is commonly associated to oxidation or the incorporation of oxygen to the original structure of a compound. The mass shift observed, in this case, corresponds most likely to the oxidation of the methionine residue of ziconotide to its sulfoxide form. The product ion spectra of the +4 ions of zi c onotide and the IS are shown in Figure 224. The e xtensive cross linking from the multiple disulfide bonds of ziconotide makes it less susceptible to collisional fragmentation. The only product ions identified from its MS/MS spectrum resul ted from loss of water molecules. Without the aid of analysis software, identifying the rest of the fragments obtained will be difficult. Conclusion A robust and sensitive HPLC ESI MS method was developed for the analysis of intact ziconotide. The use o f a high purity silica column with low concentration of trifluoroacetic in the HPLC solvents improved the peak shape of ziconotide. The use of acetic and TFA as acid components of the mobile phases enhanced the specificity of detection for ziconotide. Tw o of

PAGE 44

44 the characteristic ions of ziconotide we re present in significant abundance in its mass spectrum whe n a combination of HAc and TFA wa s used compared to a single predominant ion when TFA only was added to the mobile phases. Increasing the percentage of HAc improved the response of the ziconotide ions Optimization of mass spectrometric parameters such as gas flows, tube lens offset voltage and capillary temperature help reduce the signal contribution of TFA adducts which resulted to enhanced signal for ziconotide ions. The limit of quantitation for this method was found to be 0.5 ng of intact ziconotide on column while the limit of detection was 0.02 ng. The presence of multiple disulfide linkages however, limits the structural information that can be derived from its MS/MS spectrum.

PAGE 45

45 Figure 21. Internal components of the LTQ mass spectrometer. Adapted from the Finnigan LTQ Hardware Manual.45

PAGE 46

46 Microbore column Electrospray source Mass spectrometer Sample loop Microbore column Electrospray source Mass spectrometer Sample loop Figure 22. Instrument set up for the HPLC low flow ESI MS analysis of ziconotide. Solvent flow from the HPLC is split to achieve optimum flow rates for th e Targa C18 column (0.3 x 100 mm, 5 m particle size). Sample is loaded onto a 2 L sample loop and during injection, solvent flow transports the sample into the column and electrospray source then into the mass spectrometer. Figure is not drawn to scale.

PAGE 47

47 0 60 120 180 No auxilliary gas With auxilliary gas Peak Area x 104 m/z 661 m/z 881 m/z 918 (TFA adduct) Figure 23. Response of ziconotide and TFA adduct with and without auxiliary gas flow. Signal contribution from the predominant TFA adduct of ziconotide was reduced when auxiliary gas was used with the sheath gas. The data are an average of 3 rep licate analyses and error bars represent the standard deviation of the replicates.

PAGE 48

48 400 600 800 1000 1200 1400 m/z 0 50 100 0 50 100Relative Abundance 0 50 100 660.91 880.68 917.86 688.83 955.26 528.88 1319.81 660.93 880.83 1319.82 528.89 880.77 660.90 1319.87 NL: 6.44E5 NL: 5.82E5 NL: 4.31E5 A B C[Z+5H]5+[Z+4H]4+[Z+3H]3+[(Z+TFA)+4H]4+[(Z+TFA)+3H]3+[Z+2H]2+[(Z+2TFA)+3H]3+[Z+5H]5+[Z+4H]4+[Z+3H]3+[Z+2H]2+[Z+4H]4+[Z+3H]3+[Z+2H]2+ 400 600 800 1000 1200 1400 m/z 0 50 100 0 50 100Relative Abundance 0 50 100 660.91 880.68 917.86 688.83 955.26 528.88 1319.81 660.93 880.83 1319.82 528.89 880.77 660.90 1319.87 NL: 6.44E5 NL: 5.82E5 NL: 4.31E5 A B C[Z+5H]5+[Z+4H]4+[Z+3H]3+[(Z+TFA)+4H]4+[(Z+TFA)+3H]3+[Z+2H]2+[(Z+2TFA)+3H]3+[Z+5H]5+[Z+4H]4+[Z+3H]3+[Z+2H]2+[Z+4H]4+[Z+3H]3+[Z+2H]2+ Figure 24. Effects of various tube lens offset voltages on the full scan mass spectra of ziconotide: ( A ) 100 V, ( B ) 160 V and ( C ) 200 V. The optimum tube lens offset voltage was determined at 160 V where the signal for both the +3 and +4 ions of ziconotide are abundant and contribution from TFA adducts is minimal. spray voltage: 3.75 kV, capillary temperature: 300 C, sheath gas: 40, auxiliary gas: 20, NL: normalized level

PAGE 49

49 0 5 10 15 20 25 30 Time (min) 0 50 100 50 100 50 100Relative Abundance 50 100 RT: 2.43 MA: 13744574 RT: 8.74 MA: 16346852 RT: 20.20 MA: 3883410 RT: 13.15 MA: 2953089 A B C D 0 5 10 15 20 25 30 Time (min) 0 50 100 50 100 50 100Relative Abundance 50 100 RT: 2.43 MA: 13744574 RT: 8.74 MA: 16346852 RT: 20.20 MA: 3883410 RT: 13.15 MA: 2953089 A B C D Figure 25. Extracted ion chromatograms of m/z 661 from low flow isocratic elution of ziconotide using 0.5% HAc in ( A ) 70:30, ( B ) 50:50, ( C ) 20:80 and ( D ) 10:90 MeOH: H2O mixtures. The best peak shape was obtained with 0.5% HAc in 70:30 MeOH:H2O as the mobile phase. The peak, however, elutes early in the chromatographic run. RT: retention time, MA: manually integrated area

PAGE 50

50 400 600 800 1000 1200 1400 m/z 0 50 100 0 50 100 0 50 100Relative Abundance 0 50 100 660.82 880.24 660.79 518.89 660.75 444.94 528.90 881.17 518.85 444.95 660.77 NL: 2.05E5 NL: 3.43E4 NL: 3.27E4 NL: 3.48E4 [Z+4H]4+[Z+5H]5+[Z+4H]4+[Z+3H]3+[Z+4H]4+[Z+3H]3+[Z+4H]4+A B C D 400 600 800 1000 1200 1400 m/z 0 50 100 0 50 100 0 50 100Relative Abundance 0 50 100 660.82 880.24 660.79 518.89 660.75 444.94 528.90 881.17 518.85 444.95 660.77 NL: 2.05E5 NL: 3.43E4 NL: 3.27E4 NL: 3.48E4 [Z+4H]4+[Z+5H]5+[Z+4H]4+[Z+3H]3+[Z+4H]4+[Z+3H]3+[Z+4H]4+A B C D Figure 26. Mass spectra of ziconotide peaks that were eluted isocratically with 0.5% HAc in ( A ) 70:30, ( B ) 50:50, ( C ) 20:80 and ( D ) 10:90 MeOH: H2O mixtures.

PAGE 51

51 0 5 10 15 20 25 30 Time (min) 50 100 50 100Relative Abundance 50 100 RT: 2.93 MA: 11711619 RT: 4.01 MA: 9591657 RT: 11.01 MA: 7062902 A B C 0 5 10 15 20 25 30 Time (min) 50 100 50 100Relative Abundance 50 100 RT: 2.93 MA: 11711619 RT: 4.01 MA: 9591657 RT: 11.01 MA: 7062902 A B C Figure 27. Extracted ion chromatograms of m/z 661 from low flow gradient elution of ziconotide with 0.5% HAc in ( A ) 70:30, ( B ) 20:80 and ( C ) 10:90 MeOH: H2O mixtures as starting mobile phase compositions. B etter peak shape was obtained with 0.5% HAc in 70:30 and 20:80 MeOH:H2O compared to acidified 10:90 MeOH: H2O. All peaks, however, exhibited tailing. RT: retention time, MA: manually integrated area

PAGE 52

52 0 2 4 6 8 10 12 14 Time (min) 50 100 0 50 100Relative Abundance 50 100 RT: 2.17 MA: 8475658 RT: 2.67 MA: 195741 RT: 2.24 MA: 553590 0.1 % formic acid 0.01 % TFA 0.1 % HAc : 0.05% TFA 0 2 4 6 8 10 12 14 Time (min) 50 100 0 50 100Relative Abundance 50 100 RT: 2.17 MA: 8475658 RT: 2.67 MA: 195741 RT: 2.24 MA: 553590 0.1 % formic acid 0 2 4 6 8 10 12 14 Time (min) 50 100 0 50 100Relative Abundance 50 100 RT: 2.17 MA: 8475658 RT: 2.67 MA: 195741 RT: 2.24 MA: 553590 0.1 % formic acid 0.01 % TFA 0.1 % HAc : 0.05% TFA Figure 28. Extracted ion chromatograms of m/z 661 from low flow isocratic elution of ziconotide using 15:85 ACN:H2O with various acidic components. The best peak shape was obtained with 0.1% HAc: 0.05% TFA in the mobile phases. Lower peak areas were obtained with solvents containing TFA compared to th at with formic acid only. RT: retention time, MA: manually integrated area

PAGE 53

53 400 600 800 1000 1200 1400 m/z 0 50 100 50 100Relative Abundance 50 100 528.94 660.85 518.95 880.21 518.98 955.35 917.77 688.51 660.62 880.58 660.87 880.61 917.87 518.84 956.11 NL: 1.78E5 NL: 3.59E4 NL: 3.22E40.1 % formic acid 0.01 % TFA 0.1 % HAc : 0.05% TFA [Z+5H]5+[Z+4H]4+[Z+3H]3+[(Z+TFA)+4H]4+[(Z+2TFA)+3H]3+[Z+4H]4+[(Z+TFA)+3H]3+[Z+3H]3+ [Z+4H]4+[Z+3H]3+[(Z+2TFA)+3H]3+[(Z+TFA)+3H]3+ 400 600 800 1000 1200 1400 m/z 0 50 100 50 100Relative Abundance 50 100 528.94 660.85 518.95 880.21 518.98 955.35 917.77 688.51 660.62 880.58 660.87 880.61 917.87 518.84 956.11 NL: 1.78E5 NL: 3.59E4 NL: 3.22E40.1 % formic acid 0.01 % TFA 0.1 % HAc : 0.05% TFA [Z+5H]5+[Z+4H]4+[Z+3H]3+[(Z+TFA)+4H]4+[(Z+2TFA)+3H]3+[Z+4H]4+[(Z+TFA)+3H]3+[Z+3H]3+ [Z+4H]4+[Z+3H]3+[(Z+2TFA)+3H]3+[(Z+TFA)+3H]3+ Figure 29. Mass spectra of ziconotide peaks from isocratic elutions using 15:85 ACN:H2O with various acidic components. Differences in the identity and abundance of ions pre sent in each mass spectrum were observed. The use of 0.1% HAc: 0.05% TFA as acid component of the mobile phase yields a good abundance of ziconotide ions and minimal signal contribution from TFA adducts.

PAGE 54

54 0 50 100 150 0.01 0.02 0.025 0.03 0.05 0.075 0.1 % TFA Peak Area x 104 m/z 529 (+5) m/z 661 (+4) m/z 881 (+3) m/z 918 (+3, TFA adduct) Figure 210. Response of ziconotide ions and a TFA adduct with increasing TFA c oncentration in the mobile phases. The +3 ion of ziconotide at m/z 881 was the most predominant charge state. The highest response for this ion was obtained with 0.025% TFA. The data are an average of 3 replicate analyses and error bars represent standard deviation of the replicates.

PAGE 55

55 0 20 40 60 0.01 0.02 0.025 0.03 0.05 0.075 0.1 % TFA Peak Area x 104 m/z 609 (+5) m/z 760 (+4) m/z 1014 (+3) m/z 1050 (+3, TFA adduct) Figure 211. Response of conotoxin GVIA ions and a TFA adduct with increasing TFA c oncentration in the mobile phases. The +3 ion of the IS at m/z 1014 was the most predominant charge state. The highest response for this ion was obtained with 0.025% TFA. The data are an average of 3 replicate analyses and error bars represent standard deviation of the replicates.

PAGE 56

56 0 50 100 150 0.01 0.025 0.05 0.075 0.1 % Acetic acid Peak Area x 104 m/z 529 (+5) m/z 661 (+4) m/z 689 (+4, TFA Adduct) m/z 881 (+3) m/z 918 (+3, TFA Adduct) m/z 955 (+3, TFA Adduct) Figure 212. Response of ziconotide ions and its TFA adducts wit h increasing HAc c oncentration in mobile phases with 0.01% TFA. The highest response for ziconotide with minimum contribution from TFA adducts was obtained with 0.1% HAc. The signal for the +3 and +4 ions of ziconotide were comparable in this case. The data are an average of 3 replicate analyses and error bars represent standard deviation of the replicates.

PAGE 57

57 0 20 40 60 80 0.01 0.025 0.05 0.075 0.1 % TFA Peak Area x 104 m/z 609 (+5) m/z 760 (+4) m/z 1014 (+3) m/z 1050 (+3, TFA adduct) Figure 213. Response of conotoxin GVIA ions and a TFA adduct with increasing HAc c oncentration in mobile phases with 0.01% TFA. The highest response for the +3 and +4 ions of the IS with minimum contribution from the only TFA adduct observed was obtained with 0.1% HAc. The data are an average of 3 replicate analyses and error bars represent standard deviation of the replicates.

PAGE 58

58 4.0 5.0 6.0 7.0 8.0 9.0 Time (min) 0 50 100 0 50 100Relative Abundance RT: 7.15 MA: 3805 RT: 4.62 MA: 1728Atlantis dC18column 1 x 150 mm, 75 L/min flow 50 pg ziconotide on column Atlantis dC18column 2.1 x 150 mm, 200 L/min flow 20 pg ziconotide on column 4.0 5.0 6.0 7.0 8.0 9.0 Time (min) 0 50 100 0 50 100Relative Abundance RT: 7.15 MA: 3805 RT: 4.62 MA: 1728Atlantis dC18column 1 x 150 mm, 75 L/min flow 50 pg ziconotide on column Atlantis dC18column 2.1 x 150 mm, 200 L/min flow 20 pg ziconotide on column Figure 214. Extracted ion chromatograms of m/z 661 for the lowest amount of ziconotide that can be detected on Atlantis dC18 columns with the same length but varying internal diameters. Improvement in the detection sensitivity of ziconotide was not observed with use of a smaller diameter column (top panel ), in this case. The limit of detection obtained with the larger diameter column ( bottom panel ) was less than that of the small diameter column.

PAGE 59

59 600 800 1000 1200 1400 m/z 0 50 100 0 50 100Relative Abundance 660.75 880.67 660.87 NL: 5.23E2 NL: 4.20E2 [Z+4H]4+[Z+3H]3+[Z+4H]4+ 600 800 1000 1200 1400 m/z 0 50 100 0 50 100Relative Abundance 660.75 880.67 660.87 NL: 5.23E2 NL: 4.20E2 [Z+4H]4+[Z+3H]3+[Z+4H]4+ Figure 215. Mass spectra of peaks corresponding to the lowest amount of ziconotide that can be detected on 150 mm Atlantis dC18 columns with internal diameters of 1 mm (top panel ) and 2.1 mm ( bottom panel ).

PAGE 60

60 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 Time (min) 1000 2000 3000 2000 4000 6000 5000Intensity 10000 20000 30000 RT: 4.94 MA: 139922 RT: 4.58 MA: 115056 RT: 5.68 MA: 71465 RT: 4.33 MA: 211893 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 Time (min) 1000 2000 3000 2000 4000 6000 5000Intensity 10000 20000 30000 RT: 4.94 MA: 139922 RT: 4.58 MA: 115056 RT: 5.68 MA: 71465 RT: 4.33 MA: 211893 Jupiter Proteo C120.1% HAc : 0.02% TFA Jupiter Proteo C120.1% HAc : 0.01% TFA Jupiter Proteo C120.1% FA Atlantis dC180.1% HAc : 0.02% TFA 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 Time (min) 1000 2000 3000 2000 4000 6000 5000Intensity 10000 20000 30000 RT: 4.94 MA: 139922 RT: 4.58 MA: 115056 RT: 5.68 MA: 71465 RT: 4.33 MA: 211893 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 Time (min) 1000 2000 3000 2000 4000 6000 5000Intensity 10000 20000 30000 RT: 4.94 MA: 139922 RT: 4.58 MA: 115056 RT: 5.68 MA: 71465 RT: 4.33 MA: 211893 Jupiter Proteo C120.1% HAc : 0.02% TFA Jupiter Proteo C120.1% HAc : 0.01% TFA Jupiter Proteo C120.1% FA Atlantis dC180.1% HAc : 0.02% TFA Figure 216. Comparison of the peak shape of ziconotide obtained with various columns and acid composition of the HPLC solvents.

PAGE 61

61 0 3 6 9 12 -1.0 -0.5 0.0 0.5 1.0 Ziconotide Concentration, ng/uLResidual (yactualyfitted) Figure 217. Plot of yresidual vs. ziconotide concentration. Data from the inter day assay of ziconotide standards were used. The greater variance of the standards with higher concentrations compared to those with lower conce ntrations supported the need for weighting.

PAGE 62

62 Table 2 1. Test for homoscedasticity using the F test. [Ziconotide Standard] ng/L Azic AIS Response (A zic /A IS ) s2 0.05 58891 951096 0.0479 5.44 x 10 6 68344 917115 0.0530 66233 911003 0.0507 59962 787271 0.0533 67539 913997 0.0541 67095 937598 0.0504 10.0 15627335 977477 10.001 9.40 x 10 2 15686759 955908 9.972 15417215 962997 9.605 14488562 837191 10.395 14553122 938112 9.704 15233215 936254 9.607 F exp = s 10.0 2 /s 0.05 2 = 9.40 x 10 2 /5.44 x 10 6 = 1.73 x 104 F tab (5, 5, 0.99) = 10.97 Azic: peak area of the +4 ion of ziconotide ( m/z 661) AIS: peak area of the +4 ion of conotoxin GVIA ( m/z 760)

PAGE 63

63 Model 1 ( wi = 1) 0 3 6 9 12 -90 -60 -30 0 30 60 90 Ziconotide Concentration, ng/uL% RE Model 2 ( wi = 1/x2) 0 3 6 9 12 -30 -20 -10 0 10 20 30 Ziconotide Concentration, ng/uL% RE Model 3 ( wi = 1/y2) 0 3 6 9 12 -30 -20 -10 0 10 20 30 Ziconotide Concentration, ng/uL% RE Model 4 ( wi= 1/x) 0 3 6 9 12 -30 -20 -10 0 10 20 30 Ziconotide Concentration, ng/uL% RE Model 5 ( wi = 1/y) 0 3 6 9 12 -30 -20 -10 0 10 20 30 Ziconotide Concentration, ng/uL% RE Figure 218. Plots of percent relative error (%RE ) vs. ziconotide concentration for unweighted (Model 1) and weighted (Models 2 5) linear regressions of ziconotide standards. Data from the interday assay of ziconotide standards were used. The best accuracy and %RE distribution scatter was achieved with 1/x2 (Model 2) as the weighting model.

PAGE 64

64 Table 2 2. Regression parameters of the calibration curves generated for each weighting factor ( wi) and the respective sum of the relative errors ( %RE) for the inter day assay data. Model w i m std. error b std. error R 2 %RE 1 1 1.653 0.012 0.001 0.050 0.997 598.16 2 1/x 2 1.647 0.012 0.0132 0.0015 0.997 243.76 3 1/y 2 1.645 0.013 0.0128 0.0015 0.997 245.00 4 1/x 1.659 0.010 0.016 0.007 0.997 252.75 5 1/y 1.659 0.010 0.016 0.007 0.997 251.22 n = 65, m = slope, b = yintercept, R2 = correlation coefficient

PAGE 65

65 Ziconotide Calibration Curve 0 5 10 15 0 5 10 15 20 y = (1.647 0.012)x (0.0132 0.0015) R2= 0.997 Ziconotide Concentration, ng/uLAverage Area Ratio of m/z 661/760 Ziconotide Calibration Curves 0 5 10 15 0 5 10 15 20 Set 1 Set 2 Set 3 Set 4 Set 5 Set 6 Ziconotide Concentration, ng/uLArea Ratio of m/z 661/760 Figure 219. Graphical representations of the area ratio of the +4 ions of ziconotide to conotoxin GVIA vs. ziconotide concentration. The top panel shows the aver age of six separate dilutions and the error bars represent standard deviation of the measurements. The line equation and standard deviation of the slope and yintercept are shown. The bottom panel shows each of the six response curves, results of which we re averaged to generate the first plot. The inset shows a blow up of the calibration range from 0.05 to 0.8 ng/L. 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.5 1.0 1.5

PAGE 66

66 Table 2 3. Regression parameters for the six calibration curves of ziconotide. Calibration Curve m std. error b std. error R 2 1 1.600 0.019 0.0147 0.0023 0.9993 2 1.647 0.026 0.013 0.003 0.9994 3 1.668 0.019 0.0119 0.0024 0.9976 4 1.67 0.05 0.013 0.006 0.9901 5 1.60 0.03 0.013 0.004 0.9991 6 1.695 0.016 0.0138 0.0020 0.9981 n = 11 except for CC4 with n=10, m = slope, b = yintercept, R2 = correlation coefficient

PAGE 67

67 70 80 90 100 110 120 0 2 4 6 8 10 Ziconotide Concentration, ng/ L Peak Area of IS x 104 0 400 800 1200 1600 2000 Peak Area of Ziconotide x 104 m/z 760 m/z 661 Figure 220. Response of ziconotide and conotoxin GVIA (IS) across the calibration curve. The data are an average of 6 separate dilutions corresponding to each of th e calibrator analyzed in 6 separate days. The error bars represent the standard deviation of the measurements.

PAGE 68

68 Table 2 4. Precision and accuracy of ziconotide calibration standards assayed during a sixday period. Nominal Concentration, ng/L n Averag e Calculated Concentration, ng/L %RSD %RE 0.050 6 0.051 6 4.5 3.2 0.080 6 0.077 6 5.8 3.0 0.10 6 0.097 9 5.3 2.1 0.30 6 0.29 6 2.2 1.2 0.50 6 0.49 9 4.4 0.075 0.80 6 0.79 9 4.7 0.91 1.0 6 0.99 4 2.1 0.62 3.0 5 3.0 3 3.0 0.84 5.0 6 5.1 0 3.0 2.0 8.0 6 8.2 0 3.6 2.5 10 6 9.8 8 3.1 1.2 %RSD = (SD/ X ) x 100 where SD: standard deviation and X : mean

PAGE 69

69 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Time (min) 0 50 100 0 50 100Relative Abundance 0 50 100 RT: 4.68 MA: 8429 RT: 4.98 MA: 782673 RT: 5.14 MA: 944513 A B C 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Time (min) 0 50 100 0 50 100Relative Abundance 0 50 100 RT: 4.68 MA: 8429 RT: 4.98 MA: 782673 RT: 5.14 MA: 944513 A B C Figure 221. Extracted ion chromatograms ( A ) oxidized ziconotide, ( B ) ziconotide and ( C ) conotoxin GVIA for the 0.5 ng/L calibration standard. Chromatograms were processed using a 3 point boxcar smoothing calculation. RT: retention time in min, MA: manually integrated area

PAGE 70

70 600 800 1000 1200 1400 1600 m/z 0 50 100 0 50 100Relative Abundance 0 50 100 664.92 885.75 660.99 880.80 917.71 529.01 1319.95 689.14 760.51 1013.62 1050.79 1518.98 788.27 608.47 A B C [(Z+O)+4H]4+[(Z+O)+3H]3+[Z+5H]5+[Z+4H]4+[Z+3H]3+[(Z+TFA)+4H]4+[(Z+TFA)+3H]3+[Z+2H]2+[G+5H]5+[G+4H]4+[(G+TFA)+4H]4+[G+3H]3+[(G+TFA)+3H]3+[G+2H]2+660.72 660.56 660.94 660.38 661.18 661.42 760.24 760.08 760.46 760.70 759.80 600 800 1000 1200 1400 1600 m/z 0 50 100 0 50 100Relative Abundance 0 50 100 664.92 885.75 660.99 880.80 917.71 529.01 1319.95 689.14 760.51 1013.62 1050.79 1518.98 788.27 608.47 A B C [(Z+O)+4H]4+[(Z+O)+3H]3+[Z+5H]5+[Z+4H]4+[Z+3H]3+[(Z+TFA)+4H]4+[(Z+TFA)+3H]3+[Z+2H]2+[G+5H]5+[G+4H]4+[(G+TFA)+4H]4+[G+3H]3+[(G+TFA)+3H]3+[G+2H]2+660.72 660.56 660.94 660.38 661.18 661.42 760.24 760.08 760.46 760.70 759.80 Figure 222. Full scan mass spectra of ( A ) oxidized ziconotide, ( B ) ziconotide a nd ( C ) conotoxin GVIA. Insets show the isotopic distributions of the +4 charge state of ziconotide (Z) and conotoxin GVIA (G) obtained with zoom scan.

PAGE 71

71 Methionine sulfoxide Methionine sulfone Methionine residue (a) H2O2(b) Performic acid Methionine sulfoxide Methionine sulfone Methionine residue (a) H2O2(b) Performic acid Figure 223. Methionine oxidation under (a) mild and (b) strong oxidizing conditions. Hydrogen peroxide, iodine and dimethylsulfoxide are examples of mild oxidants while 95% performic acid is a strong oxidant.58 Adapted from Reubsaet et al .57

PAGE 72

72 500 600 700 800 900 1000 1100 1200 m/z 0 20 40 60 80 100 0 20 40 60 80 100Relative Abundance 652.04 819.35 634.84 776.19 851.65 611.45 751.13 738.93 918.74 947.00 907.84 984.08 896.23 733.84 1007.56 1119.87 647.76[(Z 2H2O)+4H]4+[(G 2H2O)+4H]4+[(Z -3H2O)+4H]4+ 500 600 700 800 900 1000 1100 1200 m/z 0 20 40 60 80 100 0 20 40 60 80 100Relative Abundance 652.04 819.35 634.84 776.19 851.65 611.45 751.13 738.93 918.74 947.00 907.84 984.08 896.23 733.84 1007.56 1119.87 647.76[(Z 2H2O)+4H]4+[(G 2H2O)+4H]4+[(Z -3H2O)+4H]4+ Figure 224. Product ion spectra of the +4 ions of ziconotide at m/z 661.0 ( top) and conotoxin GVIA at m/z 760.4 ( bottom ) at 30% collision energy. The p roduct ions identified were mostly due to water losses.

PAGE 73

73 CHAPTER 3 HP LC ESI MSn ANALYSIS OF REDUCED AND ALKYLATED ZICONOTIDE Introduction The extensive cross linking of ziconotide and the internal standard (IS), conotoxin GVIA, via the six cys teine residues in each molecule, reduces the structural information that can be derived from t heir collision induced dissociation (CID) spectra. Experiments which involved r eduction of disulfide bonds linking the cysteine residues of the conopeptides and alkylation of the resulting sulfhydryl groups in order to eliminate cross linking and increase the likelihood of fragmentation during CID will be reported in this chapter. The presence of six cysteine residues, which form three disulfide bridges with a distinct pattern, is a common structural feature of conopeptides.2, 12, 20, 59 Disulfide bond formation is a post translational modification that plays an important role in imparting stability to the peptide conformations and maintaining certain biological functions such as specific receptor interactions.18, 48, 60 This modification is especially abundant in some hormones, enzymes, plasma proteins and venoms such as those from spiders, snakes or cone snails.7, 61 Techniques e mployed to elucidate the structure of these disulfide rich peptides include conventional methods8 such as Edman degradation and amino acid analysis, cDNA sequencing1, 49 and mass spectrometry (MS ).1, 16, 18, 48, 62. High energy CID, in particular, has been used to characterize and sequence disulfide linked peptides.63, 64 Results for low energ y CID conditions, however, indicate that the presence of disulfide bonds makes the peptides less susceptible to collisional fragmentation, making reduction and alkylation prior to MS based structure confirmation essential.49, 65, 66 Reduction, which is typically performed with dithiothreitol (DTT) or tris (2 carboxyethyl) phosphine (TCEP), involves the cleavage of disulfide bonds and addition of hydrogen to the

PAGE 74

74 thiolate anion resulting in 2 sulfhydryl groups per disulfide bond. Dithiothreitol, which is the reducing agent used in the experiments described in this chapter, react as the thiolate ion and reduce disulfide bridges through a thiol interchange.67 The mechanism involved in the reduction process involving dithiols is illustrated in Figure 3 1. Thiols, like DTT, work best for these reactions at high pH compared to TCEP, which can be employed at low pH.61, 68 Thiol disulfide exchange which is responsible for disulfide scrambling in peptides with tightly clustered disulfides, however, is minimized in acidic conditions.61 It is for this reason that the use of TCEP was explored and evaluated in various applications.18, 60, 61 In order to minimize disulfide scrambling for reduction conducted with DTT at basic pH, alkylation is performed immediately after reduction. Alkylating agents like iodoacetamide (IAM) attack the thiolate ion and in the process, compete with disulfide exchange.61 Alkylation protects the sulfhydryl groups formed and prevents them from reforming disulfide bonds. The alkylation process is also favored in basic conditions.60, 68 A general representation of reduction and alkylation is shown in Figure 3 2. Reduction and alkylation combined with high performance liquid chromatography (HPLC) mass spectrometric analysis has emerged as an important technique in the analysis of conopeptides. Jakubowski et al compared the mass profiles and CID spectra before and after the global reduction and alkylation of unpurified cone snail venom to determine the amino acid sequence of several conopeptides and ascertain those which contain cyste ine.48 Using the same global reduction and alkylation procedure and HPLC/MS, screening for post translational modifications in conot oxins was achieved by the same group.19 A differential alkylation procedure using substituted maleimides followed by HPLC and tandem mass spectrometric

PAGE 75

75 (MS/MS) analy sis was used by Jones et al to establish disulfide bond connectivity in a group of conopeptides .18 The optimization of parameters (e.g., reagent concentration, reaction time, temperature) affecting reduction and alkylation will be discussed in the sections that follow Analysis of the major products of reduction and alkylation of the conopeptides by HPLC/MS/MS was carried out in order to identify potential fragment ions for quantitation. The optimized reduction and alkylation method was then evaluated for the analysis of ziconotide plasma extract. Experimental Section Chemicals and Reagents The conopeptides conotoxin MVIIA (0.1 mg, 95% purity by HPLC) and conotoxin GVIA (0.1 mg, 98.3% purity by HPLC), DL dithiothreitol, and iodoacetamide were purchased from Sigma Aldrich (St. Louis, MO). Glacial acetic acid (HAc), ammonium hydroxide (NH4OH), trifluoroacetic acid (TFA), HPLC grade aceto nitrile (ACN), methanol (MeOH) and isopropanol (IPA) were obtained from Thermo Fisher (Fair L awn, NJ). Water used for solution pr eparation was purified by RiOs reverse osmosis and Milli Q polishing systems (Millipore, Billerica, MA). Water used for peptide solution preparation was also vacuum degassed to remove dissolved oxygen in order to minimize possible oxidation of ziconotide Evaluation of Ziconotide Extraction Recovery with ZipTip C18 Pipette Tips The extraction recoveries of a ziconotide standard solution from ZipTip C18 pipette tips (Millipore, Billerica, MA) using various elution solvents were evaluated. Prior to sample loading, the resin in the tip was wet with 20 L of 100% ACN and equilibrated with 20 L of 0.02% (v/v) TFA in water. The sample, 10 L of 0.234 M ziconotide in 0.02% TFA(aq), was applied to the resin bed by aspirating and dispensing 5 times to ensure m aximum binding of the

PAGE 76

76 peptide. Contaminants were removed by washing the resin with 10 L of 0.02% TFA in water, 45 times. The purified conopeptide was eluted with 10 L of solvent, aspirated and dispensed 3 times. Ten microliters of 0.183 M of the int ernal standard (IS), conotoxin GVIA, in the corresponding elution solvent was added before final dilution to 40 L. Percent extraction recoveries from elution with 0 .02% TFA in 15:85, 20:80, 25:75, 50:50 and 80:20 ACN: H2O and 0.02% TFA in 50:50 IPA: H2O mixtures were then compared. To avoid peaksplitting on analysis, t he final ACN concentration from elution with 0.02 % TFA in 50:50 ACN:H2O was adjusted to 25% before final dilution. ZipTip extract of 10 L of 0.02% TFA in water mixed with 10 L each of 0.234 M ziconotide and 0.183 M conotoxin GVIA in the appropriate elution solvent, and 10 L of elution solvent constituted the control solution. Ten microliters of 0.02% TFA in water was added to the control solution for extraction with 0.02% TFA i n 50:50 ACN: H2O, instead of 10 L of elution solvent, to adjust the ACN concentration to 25%. Response s of the ziconotide extracts were compared to the average response of the control solutions to determine percent recoveries. The solvent that yielded t he best recovery was used for the elution of reduced and alkylated ziconotide from the ZipTip resin. Optimization of Parameters Affecting Reduction The effects of reducing agent (DTT) concentration, reduction time and temperature on the response and identi ty of ziconotide reduction products were investigated. Reduction of 30 L of 2.95 M aqueous ziconotide solution with 20 L of aqueous DTT solution at various concentrations ( i.e., 0.60, 1.2, 2.4, 4.8 and 9.6 mM) was carried out in a water bath held at 60 C for 30 min. The mixture was allowed to equilibriate to room temperature for 1 min. Six

PAGE 77

77 microliters of 0.1% TFA in water was then added to a 30L aliquot of the reduction mixture to quench the reaction and reduce disulfide scrambling Thirty microliters of the reduction mixture treated with 0.1% TFA in water were applied to the resin bed of pre conditioned C18 ZipTips by aspirating and dispensing 10 L at a time of the 30 L sample for 5 times. The resin bed was washed with 10 L of 0.02% TFA in wa ter, 4 5 times to remove remaining contaminants. The purified conopeptide was then eluted with 30 L of 0.02% TFA in 50:50 ACN: H2O by aspirating and dispensing 10 L thrice each time. Finally, the organ ic composition in the final solutio n was adjusted to 25% by adding 30 L of 0.02% TFA in water to the extract before LC MS analysis. Following the same reduction and ZipTip extraction procedures described above, the effect of carrying out reduction at 5, 10, 30 and 60 min to the extent of reaction was explored. The effect of various temperatures ( i.e. 30, 45, 60 and 75 C) on the extent of reduction and formation of unwanted side products was also tested. Reduction time for conotoxin GVIA was also optimized. Blank and control solutions analyzed wit h each run were processed identically to the ziconotide samples. Blank solutions contain ed all reagents present in the sample except ziconotide. Control solutions contained all reagents present in the sample except DTT. The blank, in this case, was comp osed of 30 L of degassed, deionized water mixed with 20 L of the appropriate DTT(aq) solution while the control was made up of 2.95 M ziconotide solution mixed with 20 L of degassed, deionized water. Both were extracted with ZipTips which were eluted with 0.02% TFA in 50:50 ACN: H2O with the final ACN concentration adjusted to 25%. Blank, control and sample solutions were processed and analyzed one after the other to minimize response differences due to time variability.

PAGE 78

78 Optimization of Parameters Affecting Alkylation Using the optimal conditions determined for the reduction of ziconotide, parameter s that may affect the results of alkylation were also optimized. These include the concentration of the alkylating agent, iodoacetamide, alkylation time and temperature. Reduction of 30 L of 2.95 M aqueous ziconotide solution with 20 L of 0.60 mM DTT(aq) was carried out at 60 C for 10 min. A 30L aliquot of the reduction mixture was made basic with 0.1 M NH4OH (2 L, pH 10.5) before alkylation. Ten mi croliters of 3.2 mM IAM(aq) was then added and alkylation proceeded for 15, 30, 45, and 60 min in the dark at room temperature The reaction was quenched with the addition of 8 L of 0.1% TFA in water and exposure to light. The resulting products were ext racted using pre conditioned C18 ZipTip s with 0.02% TFA in 50:50 ACN:H2O as eluent. The final ACN concentration in the extract was adjusted to 25% prior to HPLC MS analysis. The same procedures for sample loading, washing away of contaminants and sample elution used in the optimization of reduction parameters were followed. Alkylation reactions without NH4OH and without 0.1% TFA in water for quenching the reaction were also performed. The alkylation temperature ( i.e., 22, 30, and 60 C) and IAM concent rations ( i.e., 3.2, 6.4, 12.8 and 25.6 mM) were also varied to assess the effects of these variables on the response and identity of the products of reduction and alkylation of ziconotide. Blank and control solutions were processed identically to the zico notide samples. The blank consisted of 30 L of degassed, deionized water mixed with 20 L of 0.60 mM DTT(aq) solution while the control had 2.95 M ziconotide solution mixed with 20 L of degassed, deionized water. Reduction, alkylation, extraction and analysis of all solutions were conducted one after the other to minimize response differences due to variations in time.

PAGE 79

79 Generation of the Response Curve for Reduced and Alkylated Ziconotide Six standard solutions of aqueous ziconotide with concentrations ranging from 0.035 to 1.8 M were prepared and mixed with a fixed amount of internal standard. Thirty microliters of each solution and 30 L of 0.66 M conotoxin GVIA were reduced with 40 L of 0.60 mM DTT for 20 min at 60 C. The mixture was allowed to equil ibrate to room temperature for one minute. Sixty microliters of the reduction mixture were made basic with the addition of 4 L of 0.1 M NH4OH. Twen ty microliters of 3.2 mM IAM were then added and alkylation proceeded for 15 min in the dark at r oom temperature. The alkylation mixture was exposed to light and 30 L of it was applied to the resin bed of a pre conditioned C18 ZipTip by aspirating and dispensing 10 L at a time of the 30 L sample for 5 times. Contaminants were washed away with 10 L of 0.02% TFA in water, aspirated and dispensed 45 times. The purified conopeptide was then eluted with 30 L of 0.02% TFA in 25:75 ACN: H2O by aspirating and dispensing 10 L thrice each time. Ten microliters more of 0.02% TFA in 25:75 ACN: H2O was ad ded to provide sufficient sample volume for LC MS analysis. Blank and control solutions were analyzed identically to the ziconotide samples. Application of the Reduction and Alkylation Method to Ziconotide Plasma Extract The optimized reduction and alkylation method was used for a preliminary analysis of blank horse plasma supplemented with ziconotide. The extraction method developed for ziconotide in horse plasma at Dr. Laszlo Prokais laboratory in the University of North Texas Health Science Center w as slightly modified and employed in this experiment.69 One hundred microliters of 25 ng/L aqueous ziconotide solution was added to 1 mL of blank horse plasma and diluted to 2.5 mL with 0.5% acetic acid before loading onto a pre conditioned PD 10 desalting column (GE Healthcare, Piscataway, NJ ). The column was washed with 3.5 mL of

PAGE 80

80 0.5% HAc to remove high molecular weight compounds, followed by 5 mL of 0.5% HAc to elute the low molecular weight fraction containing ziconotide A Waters C18 Sep Pak cartridge ( Waters Corporation, Milford, MA ) was washed with 3 mL ACN and 5 mL of 0.5% HAc for conditioni ng. The low molecular weight (MW) eluate from the PD 10 column was then loaded onto the C18 cartridge and the cartridge was washed with 5 mL of 0.5% HAc. The cartridge was dried for 10 min at a pressure of 20 in Hg. Ziconotide was eluted afterwards with 1% HAc in 40:60 MeOH:H2O and dried under N2(g) at 6 0 C. The residue was dissolved in 100 L of degassed, deionized water. A portion (50 L) was placed in a 250 L, 11 mm polypropylene autosampler vial covered with crimp seals that include Teflon/red ru bber septa (National Scientific, Rockwood, TN) for LC MS analysis. Another portion (30 L) was reduced with 20 L of 0.60 mM DTT for 10 min at 60 C. The mixture was allowed to equilibrate to room temperature for a minute. The reduction mixture (30 L) was then made basic with 2 L of 0.1 M NH4OH and alkylated with 10 L of 3.2 mM IAM for 15 min in the dark at room temperature. The alkylation mixture was exposed to light and 30 L was extracted with a pre conditioned C18 ZipTip following the optimized sample loading, washing and sample elution procedure described in the analysis of calibration standards. Considering 100% recovery and taking into account the dilutions in the sample treatment, 100 L of 15 ng/L ziconotide solution was added to 3 mL of 1% HAc in 40:60 MeOH:H2O, dried under N2(g) at 6 0 C and reconstituted with 100 L of degassed, deionized water to estimate extraction recovery. This resulting solution served as the control solution and the amount of ziconotide in solution is equivalent to the ideal amount (1.5 g) present in the extract before drying the sample.

PAGE 81

81 HPLCESI MS/MS Analysis A Surveyor Autosampler/HPLC system interfaced to a Thermo Finnigan (San Jose, CA) linear ion trap (LTQ) mass spectrometer equipped with an electrospray ioni zation source (ESI) was employed for LC MS analyses. Chromatographic separation was achieved with a Jupiter Proteo C12 column (2 x 150 mm, 4 m particle size, 90 pore size) from Phenomenex (Torrance, CA) with a Security Guard cartridge (2 x 4 mm, C12) to protect the analytical column. The column oven was set to 30 C while the autosampler was maintained at 12 C. Mobile phase A was 0.1% HAc: 0.02% TFA in water while mobile phase B was 0.1% HAc: 0.02% TFA in acetonitrile. Gradient elution with solven t B raised from 2 to 70% in 9 minutes was employed for the ZipTip extraction recovery and reduction and alkylation analyses. A longer gradient, with solvent B raised from 2 to 60% in 20 min was used to resolve the reduction products of ziconotide. Both gr adients were followed by a 6.5 min column equilibration to initial solvent conditions. A flow rate of 200 L/minute was used. Using the partial loop injection mode, 10 L of the sample was loaded onto the column. External and internal syringe needle washes with 50:50 ACN: water were incorporated in the autosampler method to prevent carry over from previous injections. Electrospray ionization in the positive mode was employed. The spray needle voltage was set at 3.75kV while the capillary voltage was at 45 V. The capillary temperature was maintained at 300 C. Sheath gas flow was set to 40 and that of the auxiliary gas to 10 arbitrary units. The LTQ mass spectrometer was set to acquire data for 10 min for the 9min gradient and 20 min for the longer gradient used to resolve the products of ziconotide reduction. Sample from the chromatography column during the first 2 min and last 0.5 min was diverted to waste. The LTQ was programmed for full scan MS then full scan MS/MS ( m/z 500 1400) of m/z 661 and m/z

PAGE 82

82 760, the quadruplycharged ions of ziconotide and conotoxin GVIA, respectively. The precursor ion isolation width in full scan MS/MS mode was set to 3 m/z with collision energy (CE) at 30% and activation energy (q) at 0.250. Full scan MS and f ull scan MS/MS ( m/z 400 1500) of m/z 748 and m/z 848, the +4 ions of the reduced and alkylated ziconotide and conotoxin GVIA, respectively, were also performed. Collision energy of 35% and activation energy of 0.250 were employed for its analysis. A precursor ion isolation width of 2 m/z was used. The same scan modes and MS/MS parameters were used for analyzing the reduction products of ziconotide. An nth order triple play (full scan, zoom scan, CID at 35% CE) data dependent acquisition mode was employed for charge state confirmation and MS3 analysis was performed to identify product ions whose m/z correspond to more than 1 ion when compared to theoretical values. MS Data Analysis Base peak chromatograms (BPCs) of the products of ziconotide reduction and extracted ion chromatograms (EICs) of intact and modified ziconotide and conotoxin GVIA were generated using the Xcalibur software. Chromatograms were processed using a 3point boxcar smoothing calculation. Extracted ion chromatograms were genera ted by plotting the response of each charge state or product ion versus time using the full scan MS or MS/MS filters, resulting in a peak area for each ion. Peak areas for the various ions were then manually integrated. The area ratio s of the sum of the +3 and +4 ions of ziconotide to IS were used for ZipTip extraction recovery evaluation. The response of the reduction products of zi conotide was obtained from base peak chromatograms while those of reduction and alkylation were from the sum of the peak ar eas for the +3, +4 and +5 ions of modified ziconotide and conotoxin GVIA.

PAGE 83

83 The ratio of the peak area of the +4 ion of modified ziconotide to that of the IS was plotted against ziconotide concentration to generate the response curve. Weighted least squares linear regression was used for data fitting, with 1/x2 as the weighting scheme of choice. GraphPad Prism Version 5.01 (San Diego, CA) was used for data processing. One or two way analysis of variance (ANOVA) in Microsoft Excel (Redwood, WA) established statistical differences at 95% confidence level for the experimental parameters. Twoway ANOVA with replication was used in this study. Results and Discussion Comparison of Extraction Recoveries with ZipTip C18 Pipette Tips Extraction recoveries of i ntact ziconotide from ZipTip C18 resin using the following elution solvents were compared: 0.02% TFA in 15:85, 20:80, 25:75 and 50:50 ACN: H2O mixtures. Response of the ziconoti de extracts was compared to the average response of the control solutions and percent recoveries were calculated according to Equation 3 1. The ratio of the total areas of the peaks for the +3 and +4 cha rge states of ziconotide to those f or the +3 and +4 charge states of the IS was used to determine the response of the extract and control solutions. Percent recovery = Area ratio of m/z 661+881 760+1014 x 100 extract Average area ratio of m/z 661+881 760+1014 control Percent recovery = Area ratio of m/z 661+881 760+1014 x 100 extract Average area ratio of m/z 661+881 760+1014 control A graph of the average extraction recoveries obtained from the different sol vents is presented in Figure 33. The average percent recovery with 0.02% TFA in 50:50ACN:H2O was slightly higher (88.3 17.5) compared to that of 0.02% T FA in 25:75 ACN:H2O (81.1 5.2). The final organic composition in extracts eluted with 0.02% TFA in 50:50ACN:H2O was adjusted to 25% before analysis because splitting of the ziconotide peak was observed at ACN (3 1)

PAGE 84

84 concentrations of more than 40% in the reconstitution solvent (data not shown). Values obtained for the seven extractions with 0.02% TFA in 50:50 ACN:H2O, however, were less precise as indicated by the higher standard deviation of its percent recovery. Average percent recoveries with 0.02% TFA in 15:85 and 20:80 ACN:H2O were lower at 65.3 7.1 and 68.0 3.5, respectively, compared to that obtained with 0.02% TFA in 25:75 ACN:H2O. The use of 0.02% TFA in 80:20 ACN: H2O and 50:50 IPA: H2O, a more nonpolar solv ent, for ziconotide elution was also evaluated but no improvement in percent recoveries was observed for either Since the highest recovery was obtained with 0.02% TFA in 50:50ACN:H2O, it was chosen as the elution solvent for the succeeding experiments in optimizing reduction and alkylation parameters. Optimized Conditions for Ziconotide Reduction The starting amounts of reagents used in the reduction and alkylation experiments for this study were based from the work of Yen et al. .60 The same reference was used as the basis for the reduction and alkylation procedure employed by Moller et al. in the determination of a novel fram ework or pattern of cysteine residues in conotoxins.21 Yen and coworkers used a 25fold molar excess of DTT and a 75 fold molar excess of IAM versus cysteine residues for reduction and alkylation of 5 to 30 L of 220 pmol/L peptide or protein solution (0.01 to 0.6 nmol).60 The identity and response of t he products of ziconotide reduction were determined and compared at the various reaction conditions investigated. The net reactions undergone by ziconotide (Z) during reduction and alkylati on are represented in Figure 34. The m/z of the various charge states of ziconotide after reduction of 1, 2, or 3 disulfide bonds were calculated and summarized in Table 3 1. The m/z of the corresponding oxidation products of the intact and reduced ions are shown in Table 32. The resulting m/z after the addition of 1 or 2 oxygen atoms were determined to account for all the possible products of oxidation. Knowing the expected m/z

PAGE 85

85 for the various reduction and oxidation products of ziconotide helped in the identification of the ions observed from the mass spectra of t he reduced peptide. Reductions of 2.95 M ziconotide with increasing DTT: Z molar ratios at 136, 271, 542, 1085 and 2169, for 30 minutes at 60 C were compared. These molar ratios correspond to DTT concentrations of 0.60, 1.2, 2.4, 4.8 and 9.6 mM, respect ively. The plot of average peak area versus DTT: Z molar ratio for the products of ziconotide reduction is shown in Figure 35. The peak areas obtained for the product of interest, the triplyreduced ziconotide, were not significantly different across th e various concentrations used. Doublyreduced and oxidized ziconotide were observed as side products. The peak areas for the doublyreduced ziconotide were also not significantly different but those for oxidized ziconotide decreased with increasing DTT c oncentrations. Since no statistical difference was established for th e response with increasing DTT: Z molar ratios, the lowest DTT concentration (0.60 mM), was used for subsequent analysis. Statistical difference at the 95% confidence level of the various responses being compared was established with one way ANOVA. When the results of the analysis showed P value > 0.05 and F < Fcritical, the group averages for the particular parameter being investigated are not significantly different, while P value < 0. 05 and F > Fcritical signify t hat the group averages are sta tistically different. The results for reduction with 0.60 mM DTT performed at 5, 10, 30 and 60 min at 60 C are illustrated in Figure 3 6. Peak areas of the triply reduced ziconotide were not s tatistically different at various reduction times but contribution from the oxidized product was less at 5 and 10 min. Peak areas of the doublyreduced peptide were not significantly different at the reduction times investigated. Since a good yield for t he desired product and minimal contribution from side products were obtained with reduction at 10 min, subsequent experiments

PAGE 86

86 were performed at this reaction time. The choice between 5 and 10min reduction times was based on ensuring a more complete reac tion at 10 versus 5 min. Comparison of the response for reduction conducted at various temperatures showed that yields for the desired product were comparable at 60 and 75 C but the least amount of side products was obtained at 60 C (Figure 3 7). Reduct ion carried out at 30 C only showed the unreduced ziconotide while reduction at 45 C yielded some unreduced, doublyand triply reduced peptide. Based on the results of this evaluation, 60 C was chosen as the reduction temperature. Finally, results of re duction conducted for 10 min at 60 C with increasing DTT: Z molar ratios were compared to those obtained at 30 min (Figure 3 8 ). Two way ANOVA was used to determine if there is interaction between DTT concentration and reduction time that could affect the response of the various products. Average peak areas of the triplyreduced ziconotide at various DTT concentrations (0.60, 1.2, 2.4, 4.8 and 9.6 mM ) were significantly higher for reduction at 30 min compared to those at 10 min. However, the average peak areas for the doublyreduced and oxidized product s were also higher at 30 min. When responses for the various DTT concentrations used were compared, results showed no significant differences. Considering the effect of concentration and reduction tim e, results from the two way ANOVA of the response for the completely reduced peptide showed no interaction between the two variables. Results from this study supported further the choice of 10 min as the reduction time. Analysis with the internal standard conotoxin GVIA, required the alkylation time to be adjusted to 20 min due to incomplete reduction of the IS at 10 min. The base peak chromatogram and mass spectra of the products of ziconotide red uction are shown in Figures 39 and 3 10, respectively. Reduction was conducted with 20 L of 0.60 mM

PAGE 87

87 DTT, for 10 min at 60 C, which are the optimal conditions determined. The m/z of the +3 and +4 ions of the oxidation product observed corresponds to the m/z of ziconotide ions with a singlyreduced disulfide bond and an additional 16 daltons (Da), which is attributed to oxygen. The base peak chromatogram of blank solutions did not show any peaks at the retention times where the product s of ziconotide reduction elute, while that of the control solutions showe d a peak at 6.57 min that corresponds to the intact ziconotide. The blank solution consisted of the ZipTip C18 extract of 30 L of degassed, deionized water mixed with 20 L of 0.60 mM DTT(aq) while the control was the extract of 2.95 M ziconotide solution mixed with 20 L of degassed, deionized water. Both were processed and analyzed identically as the ziconotide samples. Optimized Conditions for Ziconotide Alkylation Optimization of alkylation time, temperature and concentration of the alkylating agent iodoacetamide were also carried out. The addition of ammonia solution before alkylation was carried out to enhance alkylation efficiency. The use of ammonia at this step was reported by Yen and coworkers.60 To aid in the identification of the alkylation products m/z of the various ions of ziconotide after reduction and addition of 2, 4, and 6 carbamidomethyl moieties were calculated and summarized in Table 3 3. The m/z of the corresponding oxidation products of modified ziconotide ions are presented in Table 3 4. Using the optimal conditions for reduction, alkylation at different time s was performed with 3.2 mM IAM in the dark and at room temperature. The response of the completely reduced and alkylated ziconotide was comparable at 15, 30, 45 and 60 min. alkylation times (Figures 311). Response of the completely reduced and alkylated ziconotide, in all the optimization studies, pertains to the sum of the peak areas of the +3, +4 and +5 ions of the completely modified peptide. These ions correspond to m/z 997, 748 and 599, respectively and are

PAGE 88

88 represented as [ZA6+3H]3+, [ZA6+4H]4+ an d [ZA5+5H]5+. Contribution from the TFA adduct of the oxidation product of modified ziconotide ([(ZA6+O+112)+4H]4+, m/z 780), was highest at 60 min. Its signal, however, was quite var iable as indicated by the error bar on the plot. The response of two other side products of the reduction and alkylation of ziconotide were also plotted against alkylation time; these are the doublyreduced and alkylated ziconotide ([ZA6+4H]4+, m/z 719) and the oxidation product of the modified peptide ([(ZA6+O)+4H]4+, m/z 752). The signal s from these two were found to be small. Responses of the side products were also plotted separately from the most predominant signal to provide a better illustration of their abundance and any trends in response (Figure 3 12). Since the response of the desired product at 15 min was not significantly different from that at 1 hr, 15 min was used as the alkylation time for the succeeding analyses. The effect of alkylation temperature on the extent of reaction and identity of products formed was investigated next. Alkylation was carried out with 3.2 mM IAM, for 15 min. in the dark at 21, 30 and 60 C. There was no statistical difference in the peak areas obtained for the completely reduced and alkylated ziconotide at the temperatures evalua ted (Figure 3 13 ). The average peak area for the major product at 30 C seemed h igher than at room temperature (21 C) but variability in its signal was also greater. Ions corresponding to the doublyreduced and alkylated peptide and oxidation product of the completely modified ziconotide were observed at all temperatures. The signal due to the doublyreduced and alkylated ziconotide at 30 C was also highly variable. The TFA adduct of [(ZA6+O)+4H]4+, in this case, was only detected at 60 C. The better precision of the response for the completely reduced and alkylated ziconotide and the absence of TFA adduct at room temperature supported the choice for alkylation temperature.

PAGE 89

89 The major reduction and alkylation product was not included in the plot shown on Figure 314 to illustrate the response of the side products more clearly. The responses of triplyreduced and alkylated ziconotide obtained with increasing IAM:Z molar ratios at 15 min alkylation time was plotted and compared against the results at 60min (Figure 3 15 ). Results from the two way ANOVA showed that no significant difference was observed among the average peak areas for the desired product at various IAM: Z molar ratios and at the two alkylation times considered. Responses were also not st atistically differe nt when the effect of both IAM:Z molar ratios and alkylation times were considered together. These results suggest that no interaction between the two variables being investigated exists. A plot of the average peak areas for the side products observed with increasing IAM:Z molar ratios and at the two alkylation times being comp ared is presented in Figure 3 16. The signal s for both [ZA4+4H]4+ and [(ZA6+O)+4H]4+ were low and not significantly different across the various iodoacetamide c oncentrations used. The responses obtained at 15and 60min. alkylation times for these two side products were also similar. Greater variability was observed, however, for the response of the TFA adduct of the oxidation product of modified ziconotide ([ (ZA6+O+TFA)+4H]4+) at both 15and 60min. alkylation times. The TFA adduct signal was generally lower across the different IAM: Z ratios at 15 min. than 60min except at IAM concentration of 6.4 mM (1205 IAM: Z molar ratio). The highest signal for the T FA adduct was observed at the highest IAM concentration used ( i.e ., 25.6 mM). Standard deviation of these data was also the greatest. Since the response for the major product, triplyreduced and alkylated ziconotide, did not change significantly at incre asing IAM concentrations and alkylation times of 15and 60min., the optimized conditions for alkylation were determined to be 3.2 mM IAM at

PAGE 90

90 15 min. in the dark and at room temperature. Contributions from side products were also minimal at these conditio ns. Reduction and alkylation reactions were also conducted without adding NH4OH to make the solution basic and without adding 0.1% TFA in water to quench the reaction. The responses of the major and side products at these two conditions were compared to the products of reduction and alkylation carried out with NH4OH and with TFA added (Figure 317 ). The response of modified ziconotide obtained without NH4OH was significantly low er compared to that with NH4OH. This result showed that making the reaction mixture basic with NH4OH before alkylation helped enhance the efficiency of the reaction. On the other hand, the response of modified ziconotide without the addition of TFA was slightly higher than that with TFA present. The TFA adduct of the oxidation pr oduct of modified ziconotide was also not observed when alkylation was conducted without adding 0.1% TFA to quench the reaction. The importance of adding NH4OH to the reaction mixture before alkylation and the improvements observed without quenching the a lkylation process with TFA were established from the results of the analyses described. Identification of Potential Fragment Ions for Quantitation The extracted ion chromatograms and full scan mass spectra of modified ziconotide and conotoxi n GVIA are shown in Figures 318 and 3 19, respectively. The +3, +4 and +5 charge states of the completely reduced and alkylated conopeptides were observed in each mass spectrum with the +4 ion as the most predominant. The net reactions undergone by conotoxin GVI A during reduction and alkylation followed the representation adapted for ziconotide (Figure 34). The inclusion of other ions in the mass spectrum of the reduced and alkylated ziconotide such as the doublyreduced and alkylated ions of the peptide at m/z 719 (+4) and 958 (+3) and

PAGE 91

91 the oxidation product of the modified peptide at m/z 752 (+4) could have resulted from coelution of the peaks for these products and that of the product of interest. The ion at m/z 775 (+4) with a corresponding +3 ion at m/z 1033 was not successfully identified. The use of a shorter gradient for the chromatographic separation of reduction and alkylation products shortened the analysis time but limited the complete resolution of the various products. Obtaining the peak areas of the ions of interest from extracted ion chromatograms circumvented some of the issues associated with co elution and facilitated the determination of the response solely from the ion or ions of interest. The presence of other reduction and alkylation products was not observed in the mass spectrum for the internal standard. A small contribution from the +4 ion of modified ziconotide was observed in the mass spectrum of the IS due to some degree of overlap of the peaks of the 2 conopeptides. Isolation an d fragmentation of the +4 charge states of modified ziconotide and conotoxin GVIA through CID at 35% collision energy were performed to identify potential fragment ions for quantitation. The product ion spectra for the +4 ions of the modified conopeptid es are shown in Figure 3 20. The MS/MS pro duct ions were identified with the aid of ProteinProspectors MS Product software70 tha t provided theoretical average masses for peptide fragments which are needed for reliable mass assignment. The majority of the fragment ions were b and y ions from peptide bond cleavage. Table A 1 summarizes the observed and theoretical m/z of the various product ions of modified ziconotide together with their relative intensities and calculated mass differences (see Appendix A). The same information for the product ions of modified conotoxin GVIA is summarized in Table A 2 (see Appendix A). The diffe rence between the observed and theoretical m/z ( m) was limited to 0.5 Da for identification of the fragment ions. The 3 most intense product ions of m/z 748, the +4 ion of

PAGE 92

92 modified ziconotide, were m/z 900.42, 937.32, and 716.30, corresponding to y23 3+, y24H2O3+, and y12NH3 2+, respectively while those of m/z 848, the +4 ion of modified conotoxin GVIA, were m/z 1004.21, 1033.22 and 1033.92, corresponding to y24 3+, a18NH3 2+, and y25 3+, respectively. These ions could be used as diagnostic ions for the qualitative and quantitative determination of ziconotide and conotoxin GVIA. When the m/z of a product ion corresponded to that predicted f or several possible CID products the fragment ion actually responsible for the observed peak was confirmed by performing MS3 analysis. As an example, t he i on at m/z 900.42, the most abundant one seen following CID of m/z 748 might be attributed to 3 different product ions: b15H2O2+ ( m/z 900.09, m = 0.33), b15NH3 2+ ( m/z 900.59, m = 0.17), and y23 3+ ( m/z 900.73, m =0.31). Comparison of the product ions obtained after MS3 analysis of m/z 900.42 with theoretical m/z values generated using ProteinProspectors MS Product software confirmed its identity as y23 3+. Generation of the Response Curve for Modified Zico notide Six ziconotide standards with concentrations ranging from 0.035 to 1.8 M containing a fi xed concentration of 0.20 M conotoxin GVIA ( IS) in the reduction mixture were alkylated, extracted and analyzed using HPLC ESI MS and MS /MS. Peak areas of the most abundant ions ( i.e ., +4 charge states) in the full scan mass spectra of the modified conopeptides were obtained from corresponding extracted ion chromatograms. The ratios of the area of the +4 ion of ziconotide to that of the IS were plotted agai nst ziconotide concentrations, using a weighted linear fit, to generate the calibration curve for modified ziconotide (Figure 321). Since each solution had to be reduced, alkylated, extracted and analyzed one after the other, only 8 solutions including blank and control solutions were processed in duplicate and included in the run. Out of the 6 standard solutions, only 4 yielded quantifiable signal. No

PAGE 93

93 signal was detected for the +4 ion of modified ziconotide in the lowest calibrator (0.035 M) while i ts signal to noise ratio (S/N) was less than 10 in the next calibration standard (0.059 M). Analysis of back calculated concentrations of calibration standards yielded absolute values of 4.0% for the % relative error (%RE) and % relative standard devia tion ( % RSD) of 1. 0 to 16.3% (Table 3 5). With a % RSD of 16.3%, the standard at 1.2 M failed to satisfy the Food and Drug Administrations (FDA) 15% acceptance criteria for %RSD. The limited number of replicates analyzed for each standard contributed to the high %RSDs obtained for some calibrators. Using the results from back calculation of the calibration standard concentrations and the FDA guidelines, the lower limit of quantitation (LLOQ) for the reduced and alkylated ziconotide was found to be 0.50 ng on column. This corresponds to the lowest calibration standard analyzed that yielded a quantifiable signal (%RE (LOD) was determined from the ziconotide concentration that exhibited a signal to noise ratio (S /N) of at least 3 As mentioned in Chapter 2, the constant 3 represents a confidence level of detection of 89% or greater, and is the value recommended in literature.5254 The LOD, in this case, corresponds to 0.2 5 ng on column. The LOD and L LOQ determined from peak areas of the +4 ions of modified ziconotide ( m/z 748) and IS ( m/z 848), which were recorded from 10 blank analyses were also evaluated for comparison. The average (Sbl) of the area ratios of m/z 748 to m/z 848 and the resulting standard deviation (sbl) from the blank measurements were calculated and incorporated into equation 3 1, where SLOD corresponds to the signal required for the LOD. SLOD = Sbl + 3(sbl) (3 1) The signal at LOD is then incorpor ated into equation 32 to determine the concentration at LOD (CLOD), where m is the slope obtained from the plot shown in Figure 321.

PAGE 94

94 CLOD = (SLOD Sbl)/m (3 2) The LOD for the reduced and alkylated ziconotide was found to be 0.13 M or 0.57 ng on column. The lower limit of quantitation was calculated as 10 times the standard deviation of the blank measurement, sbl (Equation 33). CL L OQ = (10 x sbl)/m (3 3) The LL OQ for modified ziconotide obtained by using the above equation was determined to be 0.42 M or 1.8 ng on column. This calculated value, however, was higher than the L LOQ obtained from the experim ental results (0.50 ng) using the FDA guidelines. The LOD (0.57 ng) from the blank measurements was also higher than 0.25 ng, which corr esponds to the concentration with a S/N of at least 3 Using the ratio of the blank signals for m/z 748 to m/z 848 could have contributed to the difference in LOD and L LOQ values obtained from calculations and experimental results. The equations presented above which include the signal and standard deviation of blank measurements are intended for LOD and LLOQ determination involving linear response curves.71 The area ratio of m/z 748 to m/z 848 was used for calculations since plot of the area of m/z 748 versus ziconotide concentration did not yield a linear plot. A quadratic fit best describes the relationship between peak area and modified ziconotide concentration, in this case. The ratio of the total peak areas of the 3 major fragment ions of modified ziconotide ( m/z 716, 900.5, 937.5) to that of the IS ( m/z 1004, 1033, 1034) against ziconotide concentration was also plotted and considered for the determination of LLOQ and LOD. However, using the line equation for the plot described, the %RSD calculated for 2 of the calibration standar ds (0.59 and 1.8 M) failed the FDA acceptance criterion of 15%. There was also no improvement in the

PAGE 95

95 LOD and LLOQ when the response of product ions of modified ziconotide and IS were used. This is contrary to what was expected since isolation and fragmentation of a precursor ion should result to improved signal to noise ratio which helps increase the sensitivity of a method.72 This improvement, however, is associated with applications involving analyses of compounds in complex sample matrices where MS/MS facilitates the reduction of chemical noise contribution. Base peak chromatograms of blank and control solutions for reduction and alkylation are shown in Figure 322. The absence of any distinct peak in the bl ank chromatogram at the elution times of the modified conopeptides indicates the absence of possible interference from the blank solution. The peak that is starting to appear at 6 min in the blank chromatogram corresponds to a polyethyleneglycol (PEG) contaminant peak. The peak observed in the control chromatogram corresponds to the unmodified or intact ziconotide. Difference in peak areas observed for the modified and unmodified ziconotide could be attributed to differences in ionizat ion efficiency (see Figures 3 18 and 322 ). Reduction and Alkylatio n of Ziconotide Plasma Extract A preliminary analysis of ziconotide extract from blank horse plasma using the reduction and alkylation method developed was conducted. Extraction of the peptide from blank horse plasma using a PD 10 desalting column and C18 cartridge was shown to be fairly successful. Comparison of the total peak area of the +3 ( m/z 881), +4 ( m/z 661), and +5 ( m/z 529) charge states of intact ziconotide from the actual extract and control solution showed an average percent recovery of 71.41% 26.21% for a supplemented concentration of 2.5 g/mL ziconotide in 1 mL of horse plasma This average was only from duplicate analysis of both extract and control solutions. Representative extracted io n chromatograms of the 3 major ions of intact ziconotide for the extract, control and blank matrix and their respective mass spectra are shown in Figures

PAGE 96

96 323 and 324. A sample of blank horse plasma processed in the same way as the extract constituted the matrix blank. The full scan mass spectra of the extract and control solution showed the +3 and +4 ions of intact ziconotide as the most abundant ions. The presence of other ions, specifically m/z 627 and 835, was also observed. The masses of these io ns when compared to m/z 661 and 881, respectively, differ by 136 Da. This mass difference could correspond to the loss of 8 ammonia (17 Da) groups. The acidic conditions used during extraction would easily favor protonation and cleavage of the amine grou ps from the 4 lysine and 2 arginine residues, N terminal and amidated C terminal of ziconotide, thus accounting for the loss observed. The pH of peptide or protein solutions is said to greatly affect unfolding and denaturation.17 The possible unfolding of ziconotide, in this case, provides access to the basic amine groups for protonation. The presence of these ions ( i.e., m/z 627, 835) in the extracted matrix blank, however, may indicate contamination. Results from the HPLC/MS analysis of the reduction and alkylation product of the solid phase extract ion (SPE) extract showed the presence of reduced and alkylated ziconotide (Figures 325 and 326). However, the response for modified ziconotide was low. Failure to account for changes in peptide concentration during reduction and alkylation led to the use of the same DTT and IAM concentrations as those used during optimization wherein the peptide concentration was lower than that present in the extract analyzed. The low signal for modified ziconotide could also indicate the need to optimize reduction and alkylation in the actual sample matrix. Other ions of significant abundance ( i.e., m/z 705, 940) were also present in the mass spectra of the reduced and alkylated extract. These ions were found to correspond to the completely reduced and alkylated forms of m/z 627 and 835, respectively, which further lose 2

PAGE 97

97 water molecules. The m/z of the completely reduced and alkylated species of m/z 627 and 835, are 714 and 952. The detection of m/z 705 and 940 in the mass spectra of the blank could be due to contamination. Conclusion T he reduction and alkylation conditions for ziconotide were systematically optimized The reduction conditions were found to be optimal at 10 min and 60 C with the use of 0.60 mM dithiothreitol while complete alkylation was achiev ed with 3.2 mM iodoacetamide at room temperature for 15 min in the dark. Reduction time had to be adjusted to 20 min when the internal standard, conotoxin GVIA was added to the reaction mixture. The optimized conditions for alkylation were used for the analysis of a mixture of ziconotide and the IS with the volume of reagents doubled. Reduction of the disulfide bonds and alkylation of the resulting sulfhydryl groups eliminate cross linking and incr ease the probability of fragmen tation during CID of the conopeptides Tandem HPLC/ MS /MS analyses of modified ziconotide yielded structurally relevant fragment ions ( i.e.,b and y ions) as compared to that of the intact peptide. The availability of software such as ProteinProspectors MS Product software aide d in the reliable assignment of masses to the fragment ions generated. The 3 most intense product ions of the CID of ziconotide could be used as diagnostic ions for the qualitative and quantitative determination of the conopeptide. The limit of quantitat ion obtained for modified ziconotide however, restricted the application of the method to complex sample matrices. Because of the low effective dose of ziconotide, concentrations of the peptide in biological samples are expected to be in the ng/mL or pg/ mL range. Nevertheless, the feasibility of analyzing ziconotide from blank horse plasma

PAGE 98

98 extracts supplemented with ziconotide standard using the reduction and alkylation method developed was tested and positive results were obtained. Ziconotide was detec ted from the plasma extract although the intensity of the signal obtained was low compared to what was actually expected from the amount supplemented. Further optimization of the extraction, reduction and alkylation methods using plasma samples is recomme nded to improve product yield.

PAGE 99

99 SH SH -H+H+ S-SH RSSR RSH S SH SR -H+H+ RSH S S-SR S S SH SH -H+H+ S-SH RSSR RSH S SH SR -H+H+ RSH S S-SR S S Figure 31. Mechanism involved in the reduction of disulfide bridges by dithiols. The dithiol undergoes deprotonation forming a thiolate anion. The anion then attacks a disulfide bond, cleaving the disulfide bond in the process and forming sulfhydryl groups. Overall, the oxidation of the dithiol to the cyclic disulfide facilitates the reduction of other disulfides in solution. Adapted from Lamoureux, et al .67

PAGE 100

100 -S S+ SH HS -+ DTT Reducing Agent Peptide Reduced PeptideAlkylation O O SH + I CH2CNH2 S CH2CNH2+ HI IAM Peptide Alkylating Agent Peptide + (57g/Cys) Reduction -S S+ SH HS -+ DTT Reducing Agent Peptide Reduced PeptideAlkylation O O SH + I CH2CNH2 S CH2CNH2+ HI IAM Peptide Alkylating Agent Peptide + (57g/Cys) Alkylation O O SH + I CH2CNH2 S CH2CNH2+ HI IAM Peptide Alkylating Agent Peptide + (57g/Cys) O O SH + I CH2CNH2 S CH2CNH2+ HI IAM Peptide Alkylating Agent Peptide + (57g/Cys) Reduction Reduction Figure 32. Representation of disulfide bond reduction and alkylation of the resulting sulfhydryl groups.

PAGE 101

101 0 30 60 90120 15 20 25 50 Percentage of ACN in Elution Solvent Percent Recovery of Ziconotide Fig ure 33. Extraction recoveries of ziconotide from ZipTip C18 pipette tips using various elution solvents. Mixtures of acetonitrile and water with 0.02% TFA constituted the solvents. The 0.02% TFA in 50:50 ACN: H2O elution solvent exhibited a higher but comparable extraction recovery with 0.02% TFA in 25:75 ACN: H2O. The data are an average of 7 separate extractions and error bars represent standard deviation of the extractions.

PAGE 102

102 Z ZH2nDTT IAM ZH(2n x)Axn : number of disulfide bonds reduced x: number of carbamidomethylated cysteine residues A : carbamidomethyl group Z ZH2nDTT IAM ZH(2n x)Axn : number of disulfide bonds reduced x: number of carbamidomethylated cysteine residues A : carbamidomethyl group Figure 34. Representation of the net reactions for the reduction and a lkylation of ziconotide

PAGE 103

103 Table 3 1. The m/z of the various charge states of ziconotide after reduction of its disulfide bonds. Ave. MW of Z Charge State (CS) m/z of CS Added Mass per S S Reduced m/z of CS after Reduction of "X" S S 1 2 3 2639 .1 1+ 2640.1 2.02 2642.2 2644.2 2646.2 2+ 1320.6 1321.6 1322.6 1323.6 3+ 880.7 881.4 882.1 882.7 4+ 660.8 661.3 661.8 662.3 5+ 528.8 529.2 529.6 530.1 X: number of disulfide bonds S S : disulfide bond m/z of CS = (Ave. MW of Z + CS*1.01)/ CS m/z of CS after reduction = ( m/z of CS*CS + 2.02*X S S reduced)/CS Table 3 2. The m/z of the oxidation products of the various ions of intact and reduced ziconotide. CS m/z of CS m/z after Oxidation m/z of CS after Reduction of "X" S S and Oxida tion Add'n of 1 O Add'n of 2 O Addition of 1 O Addition of 2 O 1 2 3 1 2 3 1+ 2640.1 2656.1 2672 .1 2658.2 2660.2 2662.2 2674.2 2676.2 2678.2 2+ 1320.6 1328.6 1336 6 1329.6 1330.6 1331.6 1337.6 1338.6 1339.6 3+ 880.7 886.1 891 4 886.7 887.4 88 8.1 892.1 892.7 893.4 4+ 660.8 664.8 668 .8 665.3 665.8 666.3 669.3 669.8 670.3 5+ 528.8 532.0 535 2 532.4 532.9 533.3 53 5. 6 536.1 536.5 X: number of disulfide bonds S S : disulfide bond O: oxygen m/z of CS after addition of 1O = ( m/z of CS*CS + 16)/CS m/z of CS after addition of 2O = ( m/z of CS*CS + 32)/CS m/z of CS after reduction and addition of 1O = ( m/z of CS*CS + 2.02*X S S reduced + 16)/CS m/z of CS after reduction and addition of 2O = ( m/z of CS*CS + 2.02*X S S reduced + 32)/CS

PAGE 104

104 0 30 60 90 120 136 271 542 1085 2169 DTT:Z Molar Ratio Peak Area x 104 Doubly-reduced Triply-reduced Oxidized F igure 35. Response of the various products of ziconotide reduction with increasing DTT:Z molar ratio. Reduction of 2.95 M ziconotide was conducted for 30 minutes at 60C. There was no statistical difference in the peak areas for the triply and doublyreduced peptide but lower response was observed for the oxidized product at higher DTT: Z molar ratios. The data are an average of 6 replicate analyses and error bars represent standard deviation of the measurements.

PAGE 105

105 0 40 80 120 5 10 30 60 Reduction Time, min Peak Area x 104 Doubly-reduced Triply-reduced Oxidized Figure 36. Response of reduced and oxidized ziconotide with increasing reduction times. Average (n = 6) peak areas of the comp letely reduced ziconotide for 5, 10 and 30 minutes showed no significant difference but contribution from the oxidized product w as minimal at 5 and 10 minutes. Error bar s represent the standard deviation of the replicates.

PAGE 106

106 0 50 100 150 200 30 45 60 75 Temperature, oC Peak Area x 104 Unreduced Doubly-reduced Triply-reduced Oxidized Figure 37. Response of reduced and oxidized ziconotide with increasing reduction temperatures. Good yields of the desired product were observed at 60 and 75C but the least amount of side products was obtained at 60C. The data are an average of 6 replicate analyses and error bars represent standard deviation of the measurements.

PAGE 107

107 0 25 50 75 100 0 500 1000 1500 2000 DTT:Z Molar Ratio Peak Area x 104 10-min Triply-reduced 30-min Triply-reduced 10-min Doubly-reduced 30-min Doubly-reduced 10-min Oxidized 30-min Oxidized Figure 38. Response of reduced and oxidized ziconotide at 10and 30minute reduction times with in creasing DTT concentrations. Responses (n = 6) of the triply reduced ziconotide at 10and 30 minute reduction times were statistically different while there were no significant difference at various DTT concentrations. There was no interaction between the two variables based on statistical analysis.

PAGE 108

108 Time (min) RT: 7.69 MA: 841814 6.0 7.0 8.0 9.0 0 50 100 RT: 7.97 MA: 19038 RT: 7.30 MA: 21042 a b cRelative Abundance Time (min) RT: 7.69 MA: 841814 6.0 7.0 8.0 9.0 0 50 100 RT: 7.97 MA: 19038 RT: 7.30 MA: 21042 a b cRelative Abundance Figure 39. B ase peak chromatogram of the various products of ziconotide reduction Reduction was carried out with 0.60 mM DTT for 10 min at 60 C, the optimum conditions determined. Peaks a, b, and c c orrespond to mass spectra A, B and C, respectively. RT: retention time, MA: manually integrated area

PAGE 109

109 Relative Abundance100 0 50 661.97 881.84 [ZH4+4H]4+ 0 50 100 662.50 883.00 530.11 50 100 665.58 886.38 [(ZH2+O)+4H]4+m/z 500 600 700 800 900 1000 1100 1200 0 [ZH4+3H]3+[ZH6+5H]5+[ZH6+4H]4+[ZH6+3H]3+[(ZH2+O)+3H]3+Doubly reduced Triply reduced Oxidized A B CRelative Abundance100 0 50 661.97 881.84 [ZH4+4H]4+ 0 50 100 662.50 883.00 530.11 50 100 665.58 886.38 [(ZH2+O)+4H]4+m/z 500 600 700 800 900 1000 1100 1200 500 600 700 800 900 1000 1100 1200 0 [ZH4+3H]3+[ZH6+5H]5+[ZH6+4H]4+[ZH6+3H]3+[(ZH2+O)+3H]3+Doubly reduced Triply reduced Oxidized A B C Figure 310. Full scan mass spectra of reduced and oxidized ziconotide ( A C ). The average molecular mass of the oxidation product corresponds to the ad dition of 16 Da to the mass of a singly reduced peptide. A gradient of 260% B in 20 min was used to resolve the various products.

PAGE 110

110 Table 3 3. The m/z of the various charge states of ziconotide after reduction and alkylation of its disulfide bonds. CS m /z of CS m/z of CS after Reduction of "X" S S m/z of CS after Addition of "Y" IAM 1 2 3 2 4 6 1+ 2640.1 2642.2 2644.2 2646.2 2756.3 2872.4 2988.6 2+ 1320.6 1321. 6 1322.6 1323.6 1378.7 1436.7 1494.8 3+ 880.7 881.4 882.1 882.7 919.4 958.2 996.9 4+ 660.8 661.3 661.8 662.3 689.8 718.9 747.9 5+ 528.8 529.2 529.7 530.1 552.1 575.3 598.5 X: number of disulfide bonds S S : disulfide bond Y : number of carbamidomethyl moieties m/z of CS after reduction = ( m/z of CS*CS + 2.02*X S S reduced)/CS m/z of CS after alkylation = (m/z of CS after reduction*CS + 57.06*Y IAM added)/CS Table 3 4. The m/z of the oxidation products of the various ions of reduced and alkylated ziconotide. CS m/z of CS after Addition of "Y" IAM m/z of CS after Addition of "Y" IAM and Oxidation Addition of 1 O Addition of 2 O 2 4 6 2 4 6 2 4 6 1+ 2756.3 2872.4 2988. 6 2772.3 2888.4 3002.5 2788.3 2904.4 301 8.5 2+ 1378.7 1436.7 1494.8 1386.7 1444.7 1502.8 1394.7 1452.7 1510.8 3+ 919.4 958.2 996.9 924.8 963.5 1002.2 930.1 9 68.8 1007.5 4+ 689.8 718.9 747.9 693.8 722.9 751.9 697.8 726.9 755.9 5+ 552.1 575.3 598.5 555.3 578.5 601.7 558.5 581.7 604.9 Y: number of carbamidomethyl moieties m/z of CS after alkylation = (m/z of CS after reduction*CS + 57.06*Y IAM added)/CS m/z of CS after alkylation and addition of 1O = ( m/z of CS after reduction*CS + 57.06*Y IAM added + 16)/CS m/z of CS after alkylation and addition of 2O = ( m/z of CS after reduction*CS + 57.06*Y IAM added + 32)/CS

PAGE 111

111 0 50 100 150 200 15 30 45 60 Alkylation Time, min Peak Area x 104 [ZA4+4H]4+ [ZA6+3H]3+, [ZA6+4H]4+, [ZA6+5H]5+ [(ZA6+O)+4H]4+ [(ZA6+O+TFA)+4H]4+ Figure 311. Average peak area (n = 6) for the reduction and alkylation p roducts of ziconotide at increasing alkylation times. Responses for the completely reduced and alkylated product at various alkylation times were comparable. Error bars represent standard deviation of the replicates.

PAGE 112

112 0 10 20 30 15 30 45 60 Alkylation Time, min Peak Area x 104 [ZA4+4H]4+ [(ZA6+O)+4H]4+ [(ZA6+O+TFA)+4H]4+ Figure 312. Response of the side products of modified ziconotide at different alkylation times. Contribution from the TFA adduct of the oxidation product of modified ziconotide was highest at 60 min. but its signal was irreproducible.

PAGE 113

113 0 50 100 150 200 21 30 60 Alkylation Temperature, oC Peak Area x 104 [ZA4+4H]4+ [ZA6+3H]3+, [ZA6+4H]4+, [ZA6+5H]5+ [(ZA6+O)+4H]4+ [(ZA6+O+TFA)+4H]4+ Figure 313. Response of the various ions of reduced and alkylated ziconotide at different alkylation temperatures. There was no significant difference in the peak areas (n = 6) obtained for the major product at the temperatures evaluated. Response at 30 C seemed hi gher than at room temperature but the greater variability in signal resulted to the choice of room temperature for alkylation. Error bars represent SD.

PAGE 114

114 0 1 2 3 4 5 21 30 60 Alkylation Temperature, oC Peak Area x 104 [ZA4+4H]4+ [(ZA6+O)+4H]4+ [(ZA6+O+TFA)+4H]4+ Figure 314. Response of the side products of modified ziconotide at various alkylation temperatures. The average peak area for [ZA4+4H]4+ at 30 C was also variable. The TFA adduct of [(ZA6+O)+4H]4+ was only observed at 60 C.

PAGE 115

115 0 50 100 150 200 0 1000 2000 3000 4000 5000 IAM:Z Molar Ratio Peak Area x 104 15 min_[ZA6+3H]3+, [ZA6+4H]4+, [ZA6+5H]5+ 60 min_[ZA6+3H]3+, [ZA6+4H]4+, [ZA6+5H]5+ Figure 315. Average peak areas (n = 6) of the triplyreduced and alkylated ziconotide with increasing IAM: Z molar ra tios at 15 and 60min. alkylation times. The responses were not significantly different at various IAM: Z molar ratios and at the alkylation times considered. Results from two way ANOVA did not indicate interaction between the two variables.

PAGE 116

116 0 6 12 18 24 0 1000 2000 3000 4000 5000 IAM:Z Molar Ratio Peak Area x 104 15 min_[ZA4+4H]4+ 60 min_[ZA4+4H]4+ 15 min_[(ZA6+O)+4H]4+ 60 min_[(ZA6+O)+4H]4+ 15 min_[(ZA6+O+TFA)+4H]4+ 60 min_[(ZA6+O+TFA)+4H]4+ Figure 316. Signal contributions of the side products of ziconotide modification with increasing IAM: Z molar ratios at 15 and 60min. alkylation times. The responses (n = 6) for [ZA4+4H]4+ and [(ZA6+O)+4H]4+ were low and not significantly different at vario us IAM: Z molar ratios and at the alkylation times considered. Response for [(ZA6+O+TFA)+4H]4+ was generally variable and its highest contribution was observed at the highest IAM concentration used. Variability of this data was also the highest. Error bars represent standard deviation.

PAGE 117

117 0 50 100 150 200 no NH4OH, with TFA with NH4OH, no TFA with NH4OH, with TFA Peak Area x 104 [ZA6+4H]4+ [ZA6+3H]3+, [ZA6+4H]4+, [ZA6+5H]5+ [(ZA6+O)+4H]4+ [(ZA6+O+TFA)+4H]4+ Figure 317. Response comparison of the major and side products of modified ziconotide with and without the addition of NH4OH to enhance alkylation efficiency and the addition of TFA to quench the reaction. The data a re an average of 3 replicate analyses and error bars represent standard deviation of the measurements.

PAGE 118

118 RT: 4.82 MA: 320516 0 50 100 Relative Abundance 4.0 4.5 5.0 5.5 6.0 RT: 4.89 MA: 127226 0 50 100 Time (min) Ziconotide m/z 748Conotoxin GVIA m/z 848 RT: 4.82 MA: 320516 0 50 100 RT: 4.82 MA: 320516 0 50 100 Relative Abundance 4.0 4.5 5.0 5.5 6.0 RT: 4.89 MA: 127226 0 50 100 Time (min) 4.0 4.5 5.0 5.5 6.0 RT: 4.89 MA: 127226 0 50 100 Time (min) Ziconotide m/z 748Conotoxin GVIA m/z 848 Figure 318. Extracted ion chromatograms of the +4 charge states of the reduced and alkylated ziconotide ( top ) and conotoxin GVIA ( bottom ). These data correspond to initial concentrations of 1.8 M for ziconotide and 0.66 M for conotoxin GVIA.

PAGE 119

119 Relative Abundance 0 50 100 748.02 996.56 774.84 598.56 718.92 957.88 [ZA6+4H]4+[ZA6+5H]5+751.78 [ZA4+4H]4+[ZA4+3H]3+[ZA6+3H]3+[(ZA6+O)+4H]4+1032.57 847.46 748.17 1129.65 678.77 0 50 100 m/z 500600 700 800 900 1000 1100 1200 [GA6+4H]4+[GA6+5H]5+[GA6+3H]3+ZiconotideConotoxin GVIA 747.78 748.02 747.52 748.26 747.30 748.48 748.74 847.46 847.26 847.02 846.74 847.74 847.98 848.24 Relative Abundance 0 50 100 748.02 996.56 774.84 598.56 718.92 957.88 [ZA6+4H]4+[ZA6+5H]5+751.78 [ZA4+4H]4+[ZA4+3H]3+[ZA6+3H]3+[(ZA6+O)+4H]4+1032.57 847.46 748.17 1129.65 678.77 0 50 100 m/z 500600 700 800 900 1000 1100 1200 [GA6+4H]4+[GA6+5H]5+[GA6+3H]3+ZiconotideConotoxin GVIARelative Abundance 0 50 100 748.02 996.56 774.84 598.56 718.92 957.88 [ZA6+4H]4+[ZA6+5H]5+751.78 [ZA4+4H]4+[ZA4+3H]3+[ZA6+3H]3+[(ZA6+O)+4H]4+1032.57 847.46 748.17 1129.65 678.77 0 50 100 m/z 500600 700 800 900 1000 1100 1200 [GA6+4H]4+[GA6+5H]5+[GA6+3H]3+Relative Abundance 0 50 100 748.02 996.56 774.84 598.56 718.92 957.88 [ZA6+4H]4+[ZA6+5H]5+751.78 [ZA4+4H]4+[ZA4+3H]3+[ZA6+3H]3+[(ZA6+O)+4H]4+1032.57 847.46 748.17 1129.65 678.77 0 50 100 m/z 500600 700 800 900 1000 1100 1200 [GA6+4H]4+[GA6+5H]5+[GA6+3H]3+ZiconotideConotoxin GVIA 747.78 748.02 747.52 748.26 747.30 748.48 748.74 747.78 748.02 747.52 748.26 747.30 748.48 748.74 847.46 847.26 847.02 846.74 847.74 847.98 848.24 847.46 847.26 847.02 846.74 847.74 847.98 848.24 Figure 319. Full scan mass spectra of the modified conopeptides, ziconotide ( top) and conotoxin GVIA ( bottom ). The +4 charge states of the complet ely reduced and alkyalted conopeptides were the most aboundant ions. The +3 and +5 ions of the modified conopeptides were also present. The doublyreduced and alkylated ziconotide ([ZA4+4H]4+, [ZA4+3H]3+) and the oxidation product of the completely modif ied ziconotide were also observed in its mass spectrum. Insets show the isotopic distributions of the +4 ions of the modified ziconotide and conotoxin GVIA.

PAGE 120

120 0Relative Abundance 50 100 900.42 937.32 716.30 666.45 657.81796.26553.06 1261.8 474.31 1154.4 937.96 b14 3+y11 2+y12-NH3 2+y19 3+y23 3+y24-H2O3+y10-H2O b21-NH3 2+712.17 b18 3+819.78 838.59 894.79828.57b14 2+y21 3+y24C K GK GA KCSRLMY D C K GK GA KCSRLMY D C CTG C CTG S CR SGK C S CR SGK C NH NH2 2b15 y12 y21b4 b14y11 b24 b19 y23y19 y10 b21 b18400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 m/z707.84 931.55 0Relative Abundance 50 100 900.42 937.32 716.30 666.45 657.81796.26553.06 1261.8 474.31 1154.4 937.96 b14 3+y11 2+y12-NH3 2+y19 3+y23 3+y24-H2O3+y10-H2O b21-NH3 2+712.17 b18 3+819.78 838.59 894.79828.57b14 2+y21 3+y24C K GK GA KCSRLMY D C K GK GA KCSRLMY D C CTG C CTG S CR SGK C S CR SGK C NH NH2 2b15 y12 y21b4 b14y11 b24 b19 y23y19 y10 b21 b18y24C K GK GA KCSRLMY D C K GK GA KCSRLMY D C CTG C CTG S CR SGK C S CR SGK C NH NH2 2b15 y12 y21b4 b14y11 b24 b19 y23y19 y10 b21 b18400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 m/z707.84 931.55 600 700 800 900 1000 1100 1200 1300 1400 1500 m/z 0 50 100 1004.21 1033.22 807.03 1210.39998.15 835.94 1254.27 1059.27 967.08 801.22 1192.99 625.731334.07 1420.93Relative Abundancey4a8NH3a17 2+y23 3+ y24 3+ y15 2+y18 2+y19 2+ y20 2+y22 2+ b20 2+C K S X G S S C S XT S Y NCCRSCN XY T KRC Y C K S X G S S C S XT S Y NCCRSCN XY T KRC Y NH2 NH2 y25 y16 b26 y26 b20 y24y23y22y21y20y18y14y4y5a18-NH3 2+1033.92 y25 3+ 600 700 800 900 1000 1100 1200 1300 1400 1500 m/z 0 50 100 1004.21 1033.22 807.03 1210.39998.15 835.94 1254.27 1059.27 967.08 801.22 1192.99 625.731334.07 1420.93Relative Abundancey4a8NH3a17 2+y23 3+ y24 3+ y15 2+y18 2+y19 2+ y20 2+y22 2+ b20 2+C K S X G S S C S XT S Y NCCRSCN XY T KRC Y C K S X G S S C S XT S Y NCCRSCN XY T KRC Y NH2 NH2 y25 y16 b26 y26 b20 y24y23y22y21y20y18y14y4y5 600 700 800 900 1000 1100 1200 1300 1400 1500 m/z 0 50 100 1004.21 1033.22 807.03 1210.39998.15 835.94 1254.27 1059.27 967.08 801.22 1192.99 625.731334.07 1420.93Relative Abundancey4a8NH3a17 2+y23 3+ y24 3+ y15 2+y18 2+y19 2+ y20 2+y22 2+ b20 2+C K S X G S S C S XT S Y NCCRSCN XY T KRC Y C K S X G S S C S XT S Y NCCRSCN XY T KRC Y NH2 NH2 y25 y16 b26 y26 b20 y24y23y22y21y20y18y14y4y5y25 y16 b26 y26 b20 y24y23y22y21y20y18y14y4y5a18-NH3 2+1033.92 y25 3+ Figure 320. Full scan MS/MS spectra of the +4 ions of modified ziconotide at m/z 748 ( top)and conotoxin GVIA at m/z 848 ( bottom ). Product ions matching the theoretical product ions proposed for the peptides are labeled in red and marked with red stars; those in black and marked with black stars correspond to more than one product ion. T he observed b and y ions are also shown in the amino acid sequence s

PAGE 121

121 0.0 0.5 1.0 1.5 2.0 0 1 2 3 y = (1.506 0.073)x (0.098 0.015) R2 = 0.9991Ziconotide Concentration, MArea Ratio of m/z 748/848 Figure 321. Graphical representation of the area ratio of the +4 ions of modified ziconotide to conotoxin GVIA versus ziconotide concentration. The data are an average of two s eparate extractions and the error bars represent standard deviation of the duplicate extractions. The line equation and standard deviation of the slope and yintercept are shown.

PAGE 122

122 Table 3 5. Precision and accuracy of calibration standards for the analys is of reduced and alkylated ziconotide. [Ziconotide] Amt. on Column Peak Area Calculated [Z] % Relative %RSD M ng of m/z 748/848 M Error 0.035 0.15 nd 0.059 0.25 0.03704* 0.12 0.50 0.07935 0.118 0.39 4.6 0.59 2.5 0.8152 0.593 1.9 14.7 1.2 5.0 1.757 1.22 5.3 16 .3 1.8 7.5 2.505 1.70 2.4 0.9 Amt. on Column = [Z] x Z sol n vol. x reduction mix vol. x vol. of alkylation mix extracted x injection vol. x MW of Z x 1/1000 total vol. of reduction mix total vol. of alkylation mix total vol. of extract nd : not detected *: S/N < 10Amt. on Column = 0.12 pmol x 30 L x 60 L x 30 L x 10 L x 2639.13 pg x 1 ng L 100 L 84 L 40 L pmol 1000 pg A 100 percent recovery at each step of the extraction was consid ered when the amount of peptide on column was calculated.Amt. on Column = [Z] x Z sol n vol. x reduction mix vol. x vol. of alkylation mix extracted x injection vol. x MW of Z x 1/1000 total vol. of reduction mix total vol. of alkylation mix total vol. of extract nd : not detected *: S/N < 10Amt. on Column = 0.12 pmol x 30 L x 60 L x 30 L x 10 L x 2639.13 pg x 1 ng L 100 L 84 L 40 L pmol 1000 pg A 100 percent recovery at each step of the extraction was consid ered when the amount of peptide on column was calculated.

PAGE 123

123 Relative Abundance 4.0 4.5 5.0 5.5 6.0 Time (min) 20 40 60 80 100 20 40 60 80 100 RT: 4.94 MA: 567037 Blank ControlRelative Abundance 4.0 4.5 5.0 5.5 6.0 Time (min) 20 40 60 80 100 20 40 60 80 100 RT: 4.94 MA: 567037 Blank Control Figure 322. Representative base peak chromatograms of reduction and alkylation blank ( top) and control ( bottom ) solutions. The blank contained all components of the reduction and alkylation mixture except the conopeptides while the control contained all reagents used in the modification except DTT and IAM. Only the time window where the modified conopeptides eluted is shown. The peak that is starting to appear at 6 min in the blank chromatogra m corresponds to PEG contaminant. The peak in the control chromatogram corresponds to the unmodified ziconotide.

PAGE 124

124 4.0 5.0 6.0 Time (min) 0 50 100 0 50 100Relative Abundance 0 50 100 RT: 4.95 MA: 4438053 RT: 4.65 MA: 7571 RT: 4.93 MA: 7189957 RT: 4.62 MA: 63720 NL: 1.37E6 NL: 2.09E6 NL: 3.80E3 4.5 5.5 SPE Extract Control Matrix blank A B C 4.0 5.0 6.0 Time (min) 0 50 100 0 50 100Relative Abundance 0 50 100 RT: 4.95 MA: 4438053 RT: 4.65 MA: 7571 RT: 4.93 MA: 7189957 RT: 4.62 MA: 63720 NL: 1.37E6 NL: 2.09E6 NL: 3.80E3 4.5 5.5 SPE Extract Control Matrix blank A B C Figure 323. Extracted ion chromatograms (EIC) of the 3 major ions of intact ziconotide for the ( A ) C18 solid phase extraction (SPE) extract, ( B ) control and ( C ) matrix blank. A 1mL blank horse plasma supplemented with 2.5 g ziconotide and diluted to 2.5 mL with 0.5% HAc(aq) was loaded onto a PD 10 desalting column. The low MW eluate was then loaded onto a pre conditioned C18 SPE cartr idge and eluted with 1% HAc in 40:60 MeOH: H2O. The extract was dried and the residue was dissolved in 100 L of degassed, deionized water. An aqueous solution containing 1.5 g of ziconotide added to 3 mL of 1% HAc in 40:60 MeOH: H2O and processed ident ically as the extract comprised the control solution. The peaks at 4.65 and 4.62 min correspond to oxidized ziconotide. RT: retention time, MA: manually integrated area

PAGE 125

125 400 600 800 1000 1200 1400 m/z 0 50 100 0 50 100Relative Abundance 0 50 100 660.96 880.66 626.88 917.74 528.95 1320.66 660.95 880.53 917.41 528.96 1320.70 626.89 835.30 1366.82 835.31 SPE Extract Control Matrix blank [Z+4H]4+[Z+5H]5+[Z+3H]3+[Z+3H+TFA]3+[Z+2H]2+[Z+4H]4+[Z+5H]5+[Z+3H]3+[Z+3H+TFA]3+[Z+2H]2+[Z+4H -136]4+[Z+4H -136]4+[Z+3H 136]3+[Z+3H -136]3+A B C 400 600 800 1000 1200 1400 m/z 0 50 100 0 50 100Relative Abundance 0 50 100 660.96 880.66 626.88 917.74 528.95 1320.66 660.95 880.53 917.41 528.96 1320.70 626.89 835.30 1366.82 835.31 SPE Extract Control Matrix blank [Z+4H]4+[Z+5H]5+[Z+3H]3+[Z+3H+TFA]3+[Z+2H]2+[Z+4H]4+[Z+5H]5+[Z+3H]3+[Z+3H+TFA]3+[Z+2H]2+[Z+4H -136]4+[Z+4H -136]4+[Z+3H 136]3+[Z+3H -136]3+A B C Figure 324. Full scan mass spectra of the major peaks in the EIC of ( A ) ziconotide SPE extract, ( B ) control and ( C ) matrix blank. The most abundant ions in the mass spectra of the extract and control solution correspond to the +3 and +4 charge states of ziconotide. Ions corresponding to loss of 136 Da from m/z 661 and 881 were als o observed. The presence of these ions in the mass spectrum of the blank could indicate contamination.

PAGE 126

126 4.0 4.5 5.0 5.5 6.0 Time (min) 0 50 100 0 50 100Relative Abundance RT: 4.80 MA: 36131 NL: 4.20E3 NL: 2.78E3 SPE Extract Matrix blank 4.0 4.5 5.0 5.5 6.0 Time (min) 0 50 100 0 50 100Relative Abundance RT: 4.80 MA: 36131 NL: 4.20E3 NL: 2.78E3 SPE Extract Matrix blank Figure 325. Extracted ion chromatograms of the 3 major ions ( m/z 599, 748, and 997) of reduced and alkylated ziconotide from plasma extract ( top) and matrix blank ( bottom ). A supplemented concentration of 2.5 g/mL ziconotide in 1 mL of horse plasma was extracted, reduced and alkylated, then analyzed. Peak for the modified ziconotide was detected but its response was low.

PAGE 127

127 500 600 700 800 900 1000 1100 1200 1300 1400 m/z 0 50 100 0 50 100Relative Abundance 747.79 705.37 939.92 774.44 533.95 996.43 705.37 939.76 550.85 1359.52 759.39 [ZA6+4H]4+[ZA6+3H]3+SPE Extract Matrix blank 500 600 700 800 900 1000 1100 1200 1300 1400 m/z 0 50 100 0 50 100Relative Abundance 747.79 705.37 939.92 774.44 533.95 996.43 705.37 939.76 550.85 1359.52 759.39 [ZA6+4H]4+[ZA6+3H]3+SPE Extract Matrix blank Figure 326. Full scan mass spectra of modified ziconotide from plasma extract ( top) and matrix blank ( bottom ). The +3 and +4 charge states of the completely reduced and alkylated ziconotide were present in the mass spectra of the extract. The presence of ions at m/z 705 and 939 in the mass spectrum of the blank could be due to contamination.

PAGE 128

128 CHAPTER 4 DEVELOPMENT OF A STA BILITY INDICATING METHOD FOR ZICONOTIDE IN COMMERCIAL TYPE FORMULATIONS Introduction The limited sensitivity of the reduction and alkylation procedur e employed with the high performance liquid chromatography (HPLC) electrospray ionization (ESI)mass spectrometry (MS) method for detecting modified ziconotide described in the last chapter, led us to pursue the development of a stabilityindicating assay for intact ziconotide in commercial type formulations. In order to establish that the performance characteristics of the HPLC ESI MS method developed we re reliable and suitable for the analysis of i ntact ziconotide, the method had to be validated. The Bi opharmaceutics Coordinating Committee in the Center for Drug Evaluation and Research, in cooperation with the Center for Veterinary Medicine at the Food and Drug Administration (FDA), has prepared a guidance for industry document that provides general reco mmendations for bioanalytical method validation.55 The important parameters in a validation study include accuracy, precision, sensitivity, selectivity, reproducibility and stability.55 Accuracy describes the closeness o f the measured value (Mv) to the true value (Tv) (i.e., concentration) and is often reported as percent relative error or % RE (see equation below).42, 73 % RE = [(Mv Tv)/Tv] x 100 (4 1) Precision, on the other hand, describes the closeness of the results obtained from replicate measurements. It is commonly expressed as the percent coeff icient of variation (%CV) or percent relative standard deviation (%RSD) (see Equation 42).4 2, 73 % CV = %RSD = (SD/ X ) x 100 (4 2)

PAGE 129

129 where SD: standard deviation and X : mean The guidance document recommends analyzing a minimum of three concentrations within the methods linear dynamic range, wi th at least five replicates per concentration, to determine accuracy and precision.55 The mean value for each s et of quality control (QC) samples should have a difference of 15% from the actual value except at the lower limit of quantitation (LLOQ), where a deviation of 20% is allowable.55 At least 67% of the QC samples should satisfy the acceptance criteria for the analysis to be considered successful.55 Accuracy and precision during a single run are assessed as intra day statistics while values from analyses performed on different days are considered as inter day statistics. The highest and lowest concentrations that can be determined with acceptable accuracy and precision define the limits of quantitation (LOQ) or working dynamic range of an analytical method.55, 73 The acceptance criteria are set at 20% for LLOQ and 15% for the upper limit of quantitation (ULOQ).55 The limit o f detection (LOD) is the lowest concentration or weight of analyte that can be detected at a known confidence level.52, 73 A simplified definition of this limit which is often encountered in literature is the analyte concentration with a signal to noise ratio (S/N) of at least three.74 A selective analytical method allows for the differentiation and quantitation of an analyte in the presence of other components in the sample.55, 73 A method is said to be sensitive if it can detect small differences in analyte concentration or if small changes in concentration produce great chang es in the response function.52, 73 C hemical stability refers to the extent to which a compound retains its chemical integrity.75 Environmental factors such as temperature, light, air and humidity influence stability as well as other factors like nature of the container, pH of the solution and presence of other chemicals.75, 76

PAGE 130

130 The effect s of harsh oxidative conditions like exposure to high temperature and presence of oxygen on the degradati on of ziconotide were investigated in the work reported in this c hapter. The conditions chosen were aimed at simulat in g actual situations to which ziconotide drug formulations might be exposed during transit by personnel who are unaware of the stability considerations for the drug. In the commercially available Prialt formulation of ziconotide (25 or 100 g/mL) which contains lmethionine as a protective oxygen scavenger at a concentration of 50 g/mL, one of this formulations identified degradation pathways involves oxidation of its methionine residue.56 As discussed in Chapter 2, m ethionine is first oxidized to its sulfoxi de and under strong oxidizing conditions, to a sulf one .57 Stability studies on ziconotide drug formulations or admixtures with other analgesics have employed HPL C with ultravioletvisible (UV Vis) sp ectrophotometry for ziconotide detection.56, 77 In these studies, stability was defined as the time period during which the samples retained at least 90% of the initial ziconot ide concentration. Generally, 90% of the labeled potency is recognized as the minimum level of acceptable potency for drugs.76 UVVis spectrophotometric detection relies on the absorption of light by certain functional groups known as chromophores.52 The lesser degree of specifi city associated with assigning structures based on their retention times and UVVis absorption maxima prompted us to evaluate HPLC ESI MS as a stability indicating method for ziconotide analysis. Mass spectrometry enhances detection specificity through it s ability to identify structures based on massto charge ( m/z ) ratios and mass shifts following modification. The validation of an ESI HPLC MS method and its application to the analysis of ziconotide solutions similar in composition to the commercial drug formulation are discussed in

PAGE 131

131 the sections that follow. Calculations involved in determining the rate constant of degradation of ziconotide at 50 C are also presented and discussed Experimental Section Chemicals and Reagents The conopeptides, conotoxi n MVIIA (0.1 mg, 95% purity by HPLC) and conotoxin GVIA (0.1 mg, 98.3% purity by HPLC), and the amino acid L methionine (BioChemika Ultra division) were purchased from Sigma Aldrich (St. Louis, MO). Glacial acetic acid (HAc), trifluoroacetic acid (TFA ), sodium hydroxide (NaOH) pellets, hydrochloric acid (HCl), isopropanol (IPA) and HPLC grade acetonitrile (ACN) were obtained from Thermo Fisher (Fair L awn, NJ). A sterile, preservative free, saline solution (0.9% w/v sodium chloride) was acquired from B axter Healthcare Corporation (Deerfield, Il). Water used for solution preparation was purified by RiOs reverse osmosis and Milli Q polishing systems (Millipore, Billerica, MA). Water used for peptide solution preparation was also vacuum degassed to remo ve dissolved oxygen in order to minimize possible oxidation of ziconotide. Method Validation Quality control samples at concentrations of 0.06, 0.6, and 6 ng/L (n = 5) were analyzed with each of the 6 calibration curves discussed in Chapter 2. Eleven c alibration standards with ziconotide concentrations ranging from 0.05 to 10 ng/L, and containing 1 ng/L of the internal standard (IS), conotoxin GVIA, were included in each calibration curve. The intra and inter day accu racy and precision of the opti mized HPLC ESI MS method for intact ziconotide were evaluated from the results of the QC samples and the 6 calibration curve runs. Results from analyses conducted within 24 hr constituted the intra day statistics while those from six

PAGE 132

132 independent experiments constituted the inter day statistics. Concentrations were calculated from the equation of the line derived from the corresponding calibration curve. Effect of Time in Autosampler Rack on Ziconotide Stability in Simple Solution The e ffect of time spent in the a utosampler (thermostatted to12 C) on the concentration of ziconotide was determined to assess for post preparative stability. The formation of degradation products was also monitored. The calibrators and quality control samples were recapped at t he end of the first set of injections and those from the first 3 sets of calibration curves were reanalyzed after sitting in the autosampler for 12 hr while those from the last 3 sets were reanalyzed after 24 hr. Effect of Harsh Oxidative Conditions on Ziconotide Stability in Commercial type Formulations A ziconotide solution similar in composition to the commercially available drug formulation, Prialt (Elan Pharmaceutical, Inc.), was prepared and used for evaluating the effect of harsh oxidative condi tions on the degradation of ziconotide. Prialt is formulated as a sterile, preservative free, isotonic solution for intrathecal administration.78 Prialt has to be delivered directly to the spinal fluid via an external or implanted pump due to the potential adverse effects of intravenous administration.10 A 20 mL vial of Prialt contains 25 g/mL of ziconotide acetate, 50 g/mL of L methionine (a protective oxygen scavenger) and 9 mg/mL of sodium chloride with the pH adjusted to between 4.0 and 5.0.56, 78 L methionine (5 mg ) was first dissolved in sterile, preservative free 0.9% (w/v) saline solution to prepare a 50 g/mL methionine solution. Preparation of the test solution involved dissolution of 0.1 mg ziconotide standard in 4.0 mL of st erile, preservative free 0.9% (w/v) saline solution containing 50 g/mL L methionine The pH of the solution was adjusted to 4.05.0 with the addition of a drop of 0.01 M HCl. An aseptic technique that involved wiping the

PAGE 133

133 opening of a container with an a lcohol (70% IPA) swab prior to solution transfer was conducted to prevent bacterial contamination. Sampling and appropriate dilution of the test solution to 0.5 ng/L ziconotide were performed within minutes of preparation to determine the initial ziconot ide concentration. A fixed amount of the IS was added to the test solution samples to achieve a final concentration of 1 ng/L in solution. The solution was diluted with 0.1 % HAc: 0.01% TFA in 1:99 ACN: H2O. Concentration of ziconotide in this sample w as quantified against a calibration curve determined using 9 standards ranging in concentration from 0.05 to 0.60 ng/L. Quality control samples at concentrations of 0.10, 0.30 and 0.50 ng/L were prepared in duplicate and analyzed with the calibration cu rve. Using aseptic technique, 250 L aliquots of the test solution were pipetted into limited volume glass inserts with polypropylene bottom spring (Restek, Bellefonte, PA). The inserts were then placed in 4 mL amber screw thread vials (Fisher, Fair Law n, NJ) and tightly capped with poly(tetrafluoroethylene) or PTFE/silicone closures 13 425 (Fisher, Fair L awn NJ). Vials were placed in a calibrated Robbins Scientific hybridization oven model 1000 (SciGene, Sunnyvale, CA) set at 50.0 C 0.2 C. The glass insert top is designed to be flush with the vial cap. The inserts used in this experiment were smaller than those specified for the vial size, leaving the ziconotide solution in contact with 4.21 to 4.34 cm3 of air. The volume of air inside the vial s was approximated using the volume displacement method. The volume of water displaced by the glass insert and the volume of ziconotide solution in the insert were subtracted from the volume of water in the vial to de termine the volume of air. Vials wer e removed from the oven at the following time points: 0.5, 1, 2, 4, 6, 8, 16, 24, 48, 72, 96, 120, 144, 168 and 192 hr. The vials were allowed to cool for 5 min before aseptically transferring five 10 L aliquots into five separate 2 mL plastic microcentr ifuge tubes (Thermo

PAGE 134

134 Fisher, Fair L awn, NJ) containing 440 L of 1:99 ACN: water with 0.1% HAc: 0.01% TFA. The centrifuge tubes were placed in a refrigerator kept at 5 C until sample collection and dilution for the vial k ept in the oven for 48 hr were com pleted. Aliquots from a zero hr sample vial were also diluted, refrigerated and submitted for HPLC MS analysis like the other samples. A fixed amount of the IS was added to each sample before analysis to achi eve a final concentration of 1 ng/L. HPLCESI MS /MS Analysis A Surveyor Autosampler/HPLC system interfaced to a Thermo Finnigan (San Jose, CA) linear ion trap (LTQ) mass spectrometer equipped with an electrospray ionization source (ESI) was employed for LC/MS analyses. The same instrument was used in the studies reported in Chapters 2 and 3. Chromatographic separation was achieved with a Jupiter Proteo C12 column (2 x 150 mm, 4 m particle size, 90 pore size) from Phenomenex (Torrance, CA) with a Security Guard cartridge (4 x 2 mm, C12) to prote ct the analytical column. The column was set to 30 C while the autosampler was maintained at 12 C. Gradient elution with 0.1% HAc: 0.02% TFA in water as solvent A and 0.1% HAc: 0.02% TFA in acetonitrile as solvent B, with solvent B raised from 2 to 70% in 9 min was carried out. This was follow ed by a 6.5 min column equilibration to initial solvent composition. A flow rate of 200 L/min was used. Using the partial loop injection mode, 10 L of the sample was loaded into the column. External and interna l syringe needle washes with 0.1% HAc: 0.01% TFA in 50:50 ACN: water were incorporated in the autosampler met hod to eliminate carry over. Electrospray ionization in the positive mode was employed. The spray needle voltage was set at 3.75kV while the capillary voltage was at 45 V. The capillary temperature was maintained

PAGE 135

135 at 300 C. Sheath gas flow was set to 40 and that of the auxiliary gas to 20 arbitrary units. The use of auxi liary gas was found to lessen the contribution of TFA adducts The LTQ mas s spectrometer was set to acquire for 10 min. Sample from the chromatographic column during the first 2 min and last 0.5 min was diverted to waste. The LTQ was programmed for full scan MS then full scan MS/MS ( m/z 500 1600) of m/z 661 and m/z 760, the quadruplycharged ions of ziconotide and conotoxin GVIA, respectively. The precursor ion isolation width in full scan MS/MS mode was set to 3 m/z with collision energy at 35% and activation energy (q) at 0.250. MS Data Analysis Extracted ion chromat ograms (EICs) for the various ions of ziconotide and the internal standard, conotoxin GVIA were generated using the Xcalibur software (Thermo Scientific, San Jose, CA). Chromatograms were processed using a 3point boxcar smoothing calculation. The peak s for the various charge states of ziconotide and the IS, as well as that of oxidized ziconotide, were manually integrated to determine the area for each. The sum of the contributions of the +3 and +4 ions of oxidized ziconotide was used for its EIC. The ratio of the peak area of the +4 ion of ziconotide to that of the IS was plotted against ziconotide concentration. Weighted least squares linear regression (WLSLR) analysis was used for data fitting, with 1/x2 as the weighting scheme of choice.42, 43 GraphPad Prism Version 5.01 was used for data processing. Results and Discussion Intra and Inter day Accuracy and Precision of Q uality C ontrol (QC) Samples The intra and inter day accuracy and precision of qualit y control samples at ziconotide concentrations of 0.06, 0.6 and 6 ng/L (n =5) analyzed with each of the 6 calibration curves

PAGE 136

136 shown and discussed in Chapter 2 are summarized in Table 4 1. All th e QC samples for the six analyses perf ormed yielded %RSDs % except for the 0.06 ng/L QC in calibration curves 4 and 5 which yielded 12% and 11% absolute relative error, respectively. The higher %REs obtained for the QCs mentioned were due to inclusion in the calculations of a replicate not satisfying the acceptance criterion for %RE of 15% for the purpose of data set completeness. The inter day statistics also exhibited very good agreement among all 30 replicate measurements of QC samples (%RSD 6.5); the concentrations calculated were very close to those of the nominal, as well (% absolute RE 5.0 ). Overall, the results of the analysis of QC samples and calibrators signify the accuracy, precision and robustness of the HPLC ESI MS method developed for the analysis of intact ziconotide. Post Preparative Stability of Ziconotide The post preparative stability of ziconotide standards and quality control samples after remaining in the autosampler (thermostatted to 12 C ) for 12 and 24 hr were evaluated. The first three calibration cu rves used for method validation were re analyzed 12 hr after initial analysis while the last three curves were re analyzed after 24 hr. The slopes of the curves from the time delayed analysis were lower than those analyzed immediately (Table 4 2). This decrease can be attributed to the lower peak areas of ziconotide in the re injected samples and slight changes in the IS response. The precision and accuracy of all the calibrators and QC samples still pass the FDA criteria of 15%, even at the LLOQ, signifying that ziconotide prepared in 1:99 ACN: water with 0.1% HAc: 0.01% TFA remains stable for at least 24 hr at 12 C. Tables 4 3 and 4 4 summarize the results for the timedelayed analysis of the calibration standards and quality control samples, resp ectively. The low %RSDs and % REs obtained for the calibrators were comparable to those of the calibration curves analyzed immediately. The

PAGE 137

137 slightly higher %RSDs obtained for the re injected QCs can be attributed to the oxidation of ziconotide over time. A comparison of the response of oxidized ziconotide from quality control samples which were analyzed immediately and those with time delays in the analysis is shown in Figure 41. Effect of Harsh Oxidative Conditions on Ziconotide Degradation The validat ed method was then used to evaluate the degradation of ziconotide in a test solution similar in composition to the commercially available drug formulation. Aliquots of the test solution in smaller glass inserts placed in 4 mL, capped vials were exposed to harsh oxidative conditions ( i.e., temperature of 50 C, presence of atmospheric oxygen) and removed from storage at various times for analysis. Samples were diluted fifty fold in 1:99 ACN: H2O with 0.1% HAc: 0.01% TFA and stored at 5 C before analysis at the end of the second day. Sampling w as continued until 192 hr for the initial run and 144 hr for the repeat experiment due to the limited sample available Ziconotide concentrations in the samples were quantified against a cal ibration curve from 0.05 to 0.6 ng/L. The undiluted ziconotide concentrations were obtained from the average calculated concentrations (n = 5, 9 or 10) with the use of appropriate dilution factor. These were divided by the initial ziconotide concentration which was obtained fr om a sample analyzed right after preparation of the test solution. The resulting percentages of initial concentration were plotted against storage time using an exponential function (Figure 42), specifically at 0.5, 1, 2, 8, 16, 24, 48, 96, 120, and 144 hr. On ly those time points with n = 9 or 10 were included in the plot. Q test at 95% confidence level was performed to determine if a certain replicate would be excluded. The rate constant for the degradation of ziconotide at 50 C was determined from k and it was found to be 4.8 0.8 x 103/hr. The time during which the samples retained at least

PAGE 138

138 90% of the initial ziconotide concentration was determined from the inte rsection of the lower and upper 95% confidence limit (CL) and 90% ziconotide and this was found to be approximately equal to 17 to 48 hr. The 95% confidence limit takes into account the uncertainty associated with the measurements made.76 This result indicates that it would take 17 to 48 hr of exposure at 50 C for the concentration of a 25 g/mL ziconotide solution to be reduced to 90% of the initial, which is generall y recognized as the minimum acceptable potency level for a drug. Table 4 5 summarizes the results of monitoring the changes in ziconotide concentration in test solution samples exposed to harsh oxidative conditions at various times. The test solution ini tially contained 25 g/mL of ziconotide and 50 g/mL of L methionine in sterile, preservative free 0.9% saline solution with pH adjusted to between 4.0 and 5.0. Calculations were based on the measured initial concentration. Assuming a first order rate of decomposition for ziconotide, a plot of the natural logarithm (ln) of the undiluted ziconotide concentration in molarity versus time was constructed (Figure 4 3). The test solutions had to be diluted for analysis but their original concentrations were use d for plotting. Figure 43 shows several points lying on or near the straight line that described the relationship between ln ziconotide concent ration and time. This supports the assumption that the degradation of ziconotide could be a first order or pse udo first order reaction. The slope of this plot is the negative of the rate constant ( k ) of the reaction of interest.79 Therefore, the rate constant for the degradation of ziconotide at 50 C was determined to be 4.9 0.7 x 103/hr. This value agreed well with the rate constant obtained from the exponential plot, 4.8 0.8 x 103/hr. The possibility that the degradation of ziconotide could be a pseudo first order reaction as well is supported by having an excess of L methionine in solution. Performing the experiment with a

PAGE 139

139 higher concentration of L methionine would help establish if the rate of degradation of ziconotide is affected by the amount of L methionine present in solution. Using the equation for the half life (t1/2) of a first order reaction (see Equation 43) and the rate constant obtained, the half life of the degradation of ziconotide was calculated to be 141 hr. The time it takes for the percentage of initial ziconotide concentration to reach 50% was estimat ed from the intersection of the lower and upper 95% CL and 50% ziconotide (Figure 44 ) and was found t o be approximately equal to 120 214 hr. Data for time points 168, 192 and 216 hr were calculated using the line equation for the % of initial ziconotide concentration vs. time (Figure 4 2) and included in the plot shown in Figure 44 to extend the 95% CL band for extrapolating the time when 50% of the initial ziconotide concentration had been reached. The calculated half life value of 141 hr is with in the range of half life values estimated from the intersection of the lower and upper 95% confidence limit and 50% of initial ziconotide concentration. This comparability in half life values also supported the assumpti on that the degradation of ziconotide could be a first order or pseudo first order reaction. The wide range of values obtained was a result of taking the average of the measurements from 2 separate experiments which were conducted several months apart. E ffort was undertaken to repeat the exact conditions in the initial experiment but comparison of the results from the 2 experiments suggested that an underlying instrumental or solution preparation condition could account for the differences observed. t1/ 2 = ln 2/ k (4 3) In order to estimate the effect of temperature on the shelf life of a material, Q T is calculated (Equation 4 4).76 The Q T value is a multiplier or divisor used to estimate the changes in the reaction rate constant with changes in temperature ( T ). The estimate that for every 10 C

PAGE 140

140 decrease in storage temperat ure, the shelf life doubles is based on Q T calculations, for instance. However, this statement is only true if the activation energy (Ea) of the reaction(s) that causes degradation is 15 kc al/mol .76 The activation energy for many chemical processes involved in the degradation of a drug substance or product is typically within the rang e of 10 to 25 kcal/mol .76, 80 Q T = exp [Ea/RT ( T /T + T )] (4 4) where R: universal gas constant, 1.987 cal/deg mole T: temperature (in Kelvin) whose effect is being investigated The stability of ziconotide at room tempera ture and at various activation energies was determined from calculations involving Q T and published stability data for ziconotide stored in nave pumps at 37 C (Table 4 6).77 Nave pumps are pumps that have not been exposed to ziconotide. These results were compared to calculations performed on stability data at 50 C which were obtained from the experiment described in this chapter (Table 4 7). Since the activation energy of the reaction involved in the degradation of ziconotide is not known, calculations were carried out at activation energie s of 10, 15, 20 and 25 kcal/mol The room temperature used in Q T calculations is the designated controlled room temperature which is 25 C.81 The stabilities of ziconotide at room temperature obtained from data for a diluted 25 g/ml ziconotide formulation at 37 C turned out to be greater than those calculated using the experimental stability data at 50 C. The stability at room temperat ure for a degradation reaction with Ea of 25 kcal/mole calculated from data at 37 C, for example, is 59 days compared to 12 35 days if calculations were based from stability data at 50 C. Comparison of these values yielded an absolute relative error o f 44 80%. This difference in results could be attributed to the variability of the results obtained and the possible effects of temperature on peptide

PAGE 141

141 conformation which may result in nonArrhenius behavior.82 The prediction of shelf life using results from stability studies conducted at high temperatures tends to be complicated more by the effect of temperature on protein conformation. Reliance on stabil ity studies carried out at the recommended storage conditions, therefore, is often required for establishing the expiry of biotechnologyderived products.82 The calculations performed in this chapter involving rate constants and Q Ts are based on the assumption that the system under study follows the Arrhenius equation. The Arrhenius equation (Equation 4 5) defines the relationship between the rate at whic h a reaction proceeds and its temperature.79 k = A exp ( Ea/RT) where k: rate constant (4 5) A: pre exponential factor or frequency factor Ea: activation energy R: universal gas constant, 8.314 x 103 kJ/mol K T: temperature in Kelvin When the corresponding 90% stability at 50 C was calculated at various activation energies using the published stability data at 37 C, a fair ag reement between one of the experimental and calculated result s at activation energy of 25 kcal/mol was obtained (Table 48). Comparison of the calculated and experimental 90% stabilities, 3 vs. 0.71 to 2.0 days, respectively yielded an absolute relative error of 31 to 75%. This result could indicate that the activation energy for the reaction involved in the degradation of ziconotide at 50 C is approximately 25 kcal/mol or greater A peak that elutes at the same time as oxidized ziconotide was detected on the chromatograms of the test solution samples. Figure 45 shows the extracted ion chromatograms for ziconotide, its oxidized form and the internal standard, conotoxin GVIA, for the test solution stored at 50C for 144 hr while their mass spectra are presented in Figure 46. The mass

PAGE 142

142 spectra of the early eluting peak confirmed the presence of oxidized peptide. A mass shift of 16 daltons (Da) was observed for both the +3 and +4 charge states of ziconotide. The addition of 16 Da in mass is commonly a ssociated to oxidation or the incorporation of oxygen to the original structure of a compound. The mass shift observed, in this case, corresponds most likely to the oxidation of the methionine residue of ziconotide to its sulfoxide form. Changes in resp onse of the oxidized form, reported as percentage of the total ion signal for ziconotide, were plotted with the changes in response of the +3 and +4 charge states of ziconotide (Figure 47). The total ion signal was taken as the sum of the peak areas of t he various charge states of ziconotide, its TFA adducts and oxidized ziconotide. The sum of the response of the +3 ( m/z 886) and +4 ( m/z 665) ions of oxidized ziconotide obtained from extracted ion chromatograms was used in calculating its contribution to the total ion signal. The decrease in response of ziconotide was observed to be accompanied by slight increases in the response of the oxidized form. The percentage decrease in ziconotide over time, however, was not compensated for by the increase in oxidized product. Several possibilities could have contributed to the insignificant increase of the amount of oxidation product. These include the presence of L methionine which reduces the oxidation of ziconotide, degradation of ziconotide through pathways other than oxidation, formation of degradation products that could not be separated or even detected by the method employed and loss of material due to adsorption. The appearance of additional peaks in the chromatograms of the various samples was monit ored for the formation of other possible degradation products of ziconotide. As shown in Figure 45, no additional peaks were observed even for the longest stored sample at 50 C. The mass spectra of oxidized ziconotide for the various samples were also monitored to determine if methionine oxidation had proceeded to the formation of the sulfone derivative. With the ability

PAGE 143

143 of mass spectrometry to differentiate compounds based on their masses, the addition of one or two oxygen molecules to the methionine residue of ziconotide should be easily distinguished. However, mass spectra of the peak corresponding to oxidized ziconotide for all test solution samples only confirmed the presence of methionine sulfoxide (1 oxygen added) and not methionine sulfone. Conc lusions The HPLC ESI MS method optimized for the analysis of intact ziconotide was successfully validated and applied to the evaluation of ziconotide degradation. The intra and interday accuracy and precision of quality control samples exhibited low per cent relative errors and percent relative standard deviations that satisfy the method validation guidelines required by the FDA. All six calibration curves satisfied the guidelines for the minimum number of standards that needs to meet the accuracy and pr ecision criteria for an analysis to be considered passing. Post stability studies on ziconotide determined that solutions prepared in 1:99 ACN: H2O with 0.1% HAc: 0.01% TFA remain stable at 12 C for at least 24 hr. Application of this validated method to the analysis of test solution samples similar in composition to the commercial drug formulation, which were exposed to harsh oxidative conditions at various times, facilitated the determination of a rate constant for the degradation of ziconotide at 50 C This rate c o nstant was determined to be 4.9 0.7 x 103/hr. This result indicates that exposure of a 25 g/mL ziconotide solution to 50 C for 17 to 48 hr will reduce the concentration to 90% of the initial. The resulting plot of ln ziconotide concentration versus storage time at 50 C and agreement between the half life calculated using the derived rate constant and that determined from changes in percent initial concentration supported the assumption that the degradation of ziconotide could be a first or pseudo first order reaction. Additional experiments need to be

PAGE 144

144 performed to establish the reproducibility of the results obtained for the degradation of ziconotide at 50 C. Results pertaining to the rate of degradation of a drug substance, such a s the ones presented here, provide useful information for proper handling and storage of ziconotide drug solution. Compared to UV vis spectrophotometry, mass spectrometry as a detection technique also offered greater specificity for the identification of compounds that elute from a column. Identification is based on massto charge ratios of compounds and mass shifts which often correspond to structure modification as opposed to structure assignments based on UV vis absorption maxima and retention times w hich may not be unique for a particular compound and requires the use of standards for identity confirmation.

PAGE 145

145 Table 4 1. Intra and inter day accuracy and precision of quality control samples. Calibration Intra day Nominal Concentration ( ng/L) Curve Statistics (n=5) 0.060 0.60 6.0 1 Mean 0.060 1 0.57 2 5.7 7 SD 0.001 7 0.04 2 0.2 1 %RSD 2.8 7.3 3.6 %RE 0.13 4.6 3.9 2 Mean 0.056 7 0.57 3 5.9 8 SD 0.002 0 0.02 2 0.0 6 %RSD 3.5 3.8 1.0 %RE 5.4 4.6 0.352 3 Mean 0. 059 5 0.59 1 6.0 8 SD 0.003 6 0.00 9 0.1 3 %RSD 6.1 1.5 2.2 %RE 0.89 1.5 1.3 4 Mean 0.052 8 0.58 3 6.0 5 SD 0.001 6 0.02 8 0.1 6 %RSD 3.0 4.8 2.6 %RE 12 2.8 0.90 5 Mean 0.053 5 0.60 7 6.0 5 SD 0.002 3 0.02 0 0.1 6 %RSD 4.4 3.2 2.7 %RE 11 1.2 0.86 6 Mean 0.059 6 0.59 1 5.9 5 SD 0.002 5 0.01 9 0.2 6 %RSD 4.2 3.1 4.4 %RE 0.68 1.5 0.91 Inter day Statistics (n=30) Mean 0.057 0 0.58 6 5.9 8 SD 0.003 7 0.02 6 0.1 9 %RSD 6.5 4.4 3.2 %RE 5.0 2.3 0.35 Intra day statistics were obtained from 5 replicate samples for each concentration analyzed on a single run (n=5). Inter day statistics were obtained by analyzing 6 batches of the QCs consisting of 5 replicate samples per batch (n=30). All r eplicates were prepared individually.

PAGE 146

146 Table 4 2. Regression parameters for the timedelayed analysis of the six calibration curves of ziconotide. Calibration Curve m std. error b std. error R 2 Time Delay, hours 1 1.453 0.028 0.010 0.004 0.9940 12 2 1.52 0.03 0.015 0.004 0.9985 12 3 1.561 0.018 0.0132 0.0023 0.9981 12 4 1.584 0.025 0.021 0.003 0.9989 24 5 1.570 0.023 0.0191 0.0030 0.9989 24 6 1.600 0.025 0.014 0.003 0.9996 24 n = 11 except for CC4 w ith n=10, m = slope, b = yintercept, R2 = correlation coefficient Table 4 3. Precision and accuracy of ziconotide calibration standards which were reinjected after 12 and 24 hr of initial analysis. Nominal Concentration, ng/L n Average Calculated C oncentration, ng/L %RSD %RE 0.050 6 0.051 8 4.6 3.6 0.080 6 0.076 4 5.7 4.6 0.10 6 0.098 4 2.9 1.6 0.30 6 0.29 6 2.7 1.3 0.50 6 0.51 1 2.8 2.2 0.80 6 0.79 6 3.0 0.51 1.0 6 0.99 7 2.8 0.29 3.0 5 3.0 8 2.2 2.6 5.0 6 5.1 2 3.5 2.3 8.0 6 7.9 5 3.7 0.63 10 6 9.8 9 3.3 1.5 Calibration curves 1 3 were reanalyzed after 12 hr of initial run while 4 6 were analyzed after 24 hr.

PAGE 147

147 Table 4 4. Intra and inter day accuracy and precision of time delayed analysis of quality control samples. Set Intra day Nominal Concentration (ng/L) Statistics (n=5) 0.060 0.60 6.0 1 Mean 0.059 2 0.58 9 6.0 3 SD 0.002 7 0.04 5 0.2 2 %RSD 4.5 7.6 3.5 %RE 1.4 1.9 0.48 2 Mean 0.058 5 0.58 5 5.9 4 SD 0.004 1 0.021 7 0.2 6 %RSD 7.0 3.7 4.3 %RE 2.5 2 .6 1.0 3 Mean 0.056 3 0.60 2 5.8 8 SD 0.002 8 0.02 2 0.2 6 %RSD 4.9 3.6 4.5 %RE 6.2 0.3 2.1 4 Mean 0.055 6 0.57 7 5.9 4 SD 0.002 4 0.01 7 0.0 9 %RSD 4.4 3.0 1.6 %RE 7.4 3.9 1.1 5 Mean 0.056 0 0.59 4 5.9 3 SD 0.005 1 0.02 6 0.3 0 %RSD 9.1 4.4 5.0 %RE 6.6 0.98 1.2 6 Mean 0.059 0 0.60 4 5.9 1 SD 0.001 0 0.01 2 0.1 6 %RSD 1.7 2.0 2.7 %RE 1.7 0.70 1.6 Inter day Statistics (n=30) Mean 0.057 4 0.59 2 5.9 4 SD 0.003 3 0.025 6 0.2 1 % RSD 5.8 4.3 3.5 %RE 4.3 1.4 1.1 The first 3 sets of QCs were analyzed after 12 hours of initial run while sets 4 6 were analyzed after 24 hr. Intra day statistics were obtained from 5 replicate samples for each concentration analyzed on a single run (n=5). Inter day statistics were obtained by analyzing 6 batches of the QCs consisting of 5 replicate samples per batch (n=30). All replicates were prepared individually.

PAGE 148

148 0 3 6 9 0.06 0.6 6 QC Concentration, ng/uL % Total Ion Signal of Oxidized Ziconotide Immediate analysis Time-delayed analysis Figure 41. Response of oxidized ziconotide from QC samples analyzed immedia tely and with time delay. Samples analyzed with calibration curves 1 3 were reanalyzed after 12 hr while those with curves 4 6 were reanalyzed after 24 hr. The increase in oxidized ziconotide response of those analyzed with time delay contributed to the lower precision of QC samples in these batches.

PAGE 149

149 0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 160 Y = Yo e(-k*t)Y = (102.5 3.7) e(-0.0048 0.0008)(t)Storage Time at 50 C, hr% of Initial [Ziconotide] Figure 42. Changes in ziconotide concentration vs. time for test solutions stored at 50 C and kept in contact with air. The percentage of initial ziconotide concentration was plotted against sto rage time using an exponential function. The data are an average of 910 dilutions from 2 sample vials obtained from 2 separate experiments. Error bars represent the standard deviation of the replicates. The value for 90% stability was estimated to be 1 7 to 48 hr from the intersection of 90% ziconotide and the lower and upper 95% CL. The equation for the exponential decay and standard deviations of Yo and k are also shown. Yo: % of initial ziconotide concentration at time = 0 hr; k: rate constant

PAGE 150

150 Ta ble 45. Results of analysis of test solution samples stored at 50C and kept in contact with air at various times. Time, hr n Ave. Calcd [Zic], ng/ L Undiluted [Zic], ng/ L % of [Zic]i % of [Zic] i SD Undiluted [Zic], x 106 M 0 10 0.4 2 0 21 0 100. 0 25. 2 7.9 5 0.5 10 0.43 0 21. 5 102. 6 28. 9 8.1 5 1 10 0.42 6 21. 3 101. 6 27. 4 8.0 8 2 10 0.41 6 20. 8 99. 1 29. 0 7.8 8 4 5 0.30 8 15. 4 73. 5 1.8 0 5.8 4 6 5 0.53 4 26. 7 127. 3 2.6 0 10. 1 8 10 0.42 3 21. 1 100. 7 29. 4 8.0 1 16 10 0.40 4 20. 2 96. 4 30. 0 9. 7 0 24 10 0.39 7 19. 8 94. 6 33. 7 7.5 2 48 9 0.33 5 16. 8 79. 9 37. 9 6.3 6 72 5 0.35 8 17. 9 85. 3 7. 0 6.7 8 96 10 0.27 4 13. 7 65. 3 22. 5 5. 2 0 120 10 0.25 1 12. 6 59. 8 27. 5 4.7 5 144 9 0.20 1 10. 1 47. 9 31. 1 3.8 1 168 5 0.18 7 9.3 7 44. 6 1. 7 3.5 5 192 5 0.05 5 2.7 3 13. 0 1. 0 1.0 4 [Zic]undiluted = calculated [Zic] x 500 L/10 L [Zic] in molarity = [Zic] in ng per L/(FW of ziconotide x 1000) FW of ziconotide = 2639.13 ng/nmol

PAGE 151

151 0 30 60 90 120 150 -13.5 -13.0 -12.5 -12.0 -11.5 -11.0 -10.5 y = (-0.0049 0.0007)x (11.71 0.05) R2 = 0.9867Storage Time at 50C, hrln [Zic]undiluted, M Figure 43. Plot of ln [Ziconotide]undiluted (in molarity) versus time. The rate constant for the degradation of ziconotide at 50 C corresponds to the negative of the slope of the plot and was determined to be 4.9 0.7 x 103/hr. The data are an average of 9 10 dilutions from 2 sample vials obtained from 2 separate experiments. Error bars represent the stand ard deviation of the replicates. The line equation and standard deviations of the slope and yintercept are also shown.

PAGE 152

152 0 20 40 60 80 100 120 140 160 180 200 220 0 20 40 60 80 100 120 140 160 Y = Yo e(-k*t)Y = (102.5 3.6) e(-0.0048 0.0008)(t)Storage Time at 50 C, hr% of Initial [Ziconotide] Figure 44. Changes in ziconotide concentration vs. time for 0 to 216 hr to extrapolate time to reach 50% ziconotide. Data for 168, 192 and 216 hr were calculated from the line equation of % initial [ziconotide] vs. time for 0 to 144 hr in order to extend the 95% CL band. The time it takes to reach 50% of the initial ziconotide concentration was extrapolated from the intersection of the lower and upper 95% CL and 50% ziconotide. This was determined to be 120 214 hr.

PAGE 153

153 Table 4 6. Room temperature stability derived from Q T calculations and stability data at 37 C. Ea, kcal/mol Q12 Ziconotide Stability at 37 C in Nave Pumps, day s Calculated Stability at 25 C, days 10 1.83 13 24 15 2.48 32 20 3.35 44 25 4.54 59 The 90% stability of ziconotide at 37 C was multiplied with the Q12 values for the various activation energies considered to obtain an estimate of the stability a t room temperature. Stability was determined from the intersection of the lower 95% CL and 90% ziconotide.77 Table 4 7. Room temperature stability derived from Q T calculations and stability data at 50 C. E a Ziconotide Stability Calculated Stability kcal/mol Q 25 at 50C, days at 25C, days Lower 95% CL Upper 95% CL Lower 95% CL Upper 95% CL 10 3.06 0.71 2.0 2 6 15 5.36 4 11 20 9.38 7 19 25 16.42 12 33 The 90% stability of ziconotide at 50 C was multiplied with the Q25 values for the various activation ene rgies considered to obtain an estimate of the stability at room temperature. Stability was determined from the intersection of the lower and upper 95% CL and 90% ziconotide. Table 4 8. Comparison of experimental and calculated stability data at 50 C. E a Zic Stability Calcd Stability Exp'tl Stability kcal/mol Q 13 at 37C, days at 50C, days at 50C, days Lower 95% CL Upper 95% CL 10 1.83 13 7 0.71 2.0 15 2.47 5 20 3.34 4 25 4.51 3 The experimental value for 90% stability was ob tained from the intersection of the lower 95% confidence limit and 90% ziconotide in Figure 4 2. Calculated stability data were obtained by dividing the published stability of ziconotide at 37 C with the Q13 values for various activation energies.

PAGE 154

154 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Time (min) 0 50 100 0 50 100Relative Abundance RT: 4.91 MA: 104187 RT: 5.06 MA: 1122924 NL: 2.25E4 NL: 3.02E5 B C RT: 4.60 MA: 6282 NL: 1.32E3 0 50 100 A 3.5 4.0 4.5 5.0 5.5 6.0 6.5 Time (min) 0 50 100 0 50 100Relative Abundance RT: 4.91 MA: 104187 RT: 5.06 MA: 1122924 NL: 2.25E4 NL: 3.02E5 B C RT: 4.60 MA: 6282 NL: 1.32E3 0 50 100 A F igure 45. Extracted ion chromatograms of ( A ) oxidized ziconotide, ( B ) ziconotide and ( C ) conotoxin GVIA from a test solution sample stored at 50 C for 144 hr. RT: retention time in minutes, MA: manually integrated area

PAGE 155

155 600 800 1000 1200 1400 1600 m/z 0 50 100 0 50 100Relative Abundance 0 50 100 664.84 885.53 660.97 880.52918.441320.33 528.88 688.65 760.48 1013.41 1050.85 1518.84 788.22 608.45 [(Z+O)+4H]4+[(Z+O)+3H]3+[Z+5H]5+[Z+4H]4+[Z+3H]3+[(Z+TFA)+4H]4+[(Z+TFA)+3H]3+[Z+2H]2+[G+5H]5+[G+4H]4+[(G+TFA)+4H]4+[G+3H]3+[(G+TFA)+3H]3+[G+2H]2+A B C 600 800 1000 1200 1400 1600 m/z 0 50 100 0 50 100Relative Abundance 0 50 100 664.84 885.53 660.97 880.52918.441320.33 528.88 688.65 760.48 1013.41 1050.85 1518.84 788.22 608.45 [(Z+O)+4H]4+[(Z+O)+3H]3+[Z+5H]5+[Z+4H]4+[Z+3H]3+[(Z+TFA)+4H]4+[(Z+TFA)+3H]3+[Z+2H]2+[G+5H]5+[G+4H]4+[(G+TFA)+4H]4+[G+3H]3+[(G+TFA)+3H]3+[G+2H]2+A B C Figure 46. Full scan m ass spectra of ( A ) oxidized ziconotide, ( B ) ziconotide and ( C ) conotoxin GVIA.

PAGE 156

156 Percentage of Total Ion Signal 82 86 90 94 98 0 30 60 90 120 150 Storage Time at 50 C, hr %Ziconotide 0.7 1.4 2.1 2.8 3.5 %Oxidized Ziconotide Figure 47. Summed response of the +3 and +4 charge states of ziconotide and oxidized ziconotide in test solutions stored at various times at 50 C. The decrease in response of ziconotide was observed to be accompanied by slight increases in the response of the oxidized form. The total ion signal corresponds to the sum of the peak areas of the various charge states of ziconotide [ i.e ., m/z 1320 (+2), m/z 881 (+3), m/z 661 (+4) and m/z 528 (+5)], its TFA adducts [ i.e ., m/z 689 (+4), m/z 918 (+3) and m/z 955.5 (+3)] and its oxidized form [ i.e ., m/z 665 (+4) and m/z 886 (+3)].

PAGE 157

157 CHAPTER 5 CONCLUSIONS AND FUTU RE DIRECTIONS A robust and sensitive HPLC ESI MS method for the analysis of intact ziconotide was developed and validated. The limit of detection for th is method was 0.02 ng of ziconotide on column. Tandem mass spectrometric analysis of intact ziconotide was also performed but the extensive crosslinking from its mu ltiple disulfide bridges limits the structural information that can be derived from its MS/MS spectra. Reduction and alkylation of the disulfide bonds of ziconotide was carried out to eliminate the crosslinking and render the conopeptide more susceptible to collision induced dissociation. Fragmentation of the reduced and alkylated ziconotide yielded more structurally relevant ions (i.e., b and y ions from typical peptide bond cleavage) compared to that of the intact peptide. The limit of quantitation (0 .50 ng on column) obtained for modified ziconotide, however, restricted the application of the method to complex sample matrices such as biological fluids A stability indicating assay for the quantitative determination of ziconotide in commercial type for mulations was successfully developed using the validated HPLC ESI MS method for intact ziconotide. The degradation of ziconotide in test solutions exposed to harsh oxidative conditions was evaluated. Results pertaining to the rate of degradation of ziconotide may provide useful information for proper handling and storage of the drug solution. The detection and quantitation of ziconotide in complex sample matrices will definitely benefit from the enhancement in detection sensitivity offered by nanoflow e lectrospray ionization. Nanoelectrospray ionization (NSI) decreases detection limits up to two orders of magnitude compared to ESI.83 The reduction of interference effects from salts and o ther species with the use of NSI has also provided improved sensitivity for peptides in samples contaminated by high levels of salts.84, 85 The use of a nanobore or nanoscale liquid chromatography will

PAGE 158

158 ensure accur ate delivery of specified solvent compositions especially when using gradient elution, which in turn will provide accuracy and precision f or experimental results. Care must be observed when modifying regular bore or standard HPLC equipment for low flow ra te experiments. This is because the contribution of extracolumn volumes in injectors and connecting tubing to bandspreading become s more significant.23 The reduction and alkylation method developed should be optimized for ziconotide in plasma extracts if ziconotide will be detected from plasma samples, for instance. The presence of more abundant peptides other than ziconotide in plasma samples w ould require the use of other reagents like denaturing agents, buffers, chelating agents, etc. to eliminate possible interference from other matrix components and prevent aggregation and other unwanted interactions.86, 87 Enzyme digestion of the reduced and alkylated ziconotide can also be explored for the generation of diagnostic fragment ions for the qualitative and quantitative determination of ziconotide. Lastly, monitoring the degradation of ziconotide at 50 C through several stages of half life will improve the accuracy of the half life determined. Performing the degradation study at another temperature will also provide relevant information about the rate of degradation of ziconotide which will help establish the accuracy of results obtained at 50 C. Repeating the experiment conducted with a higher concentration of the oxygen scavenger, L methionine, will help establish if the rate of degradation of ziconotide is affected by the amount of L methionine present in solution.

PAGE 159

159 APPENDIX LIST OF FRAGMENT ION S The foll owing list of fragment ions was obtained from collision induced dissociation (CID) of the +4 ions of the reduced and alkylated conopeptides: ziconotide and conotoxin GVIA. ProteinProspe ctors MS Product software was used to generate the theoretical m/z of the fragment ions observed. The difference between the observed and theoretical m/z ( m) was limited to 0.5 daltons (Da) for identification of the fragment ions (see Equation A 1) Fragment ions matching the theoretical fragment ions proposed for the peptide are labeled in red while those in black correspond to more than one fragment ion. The relative intensity of the 3 most abundant fragment ions are also in bold font. m = /(Obser ved m/z Theoretical m/z ) / ( A 1) Table A 1. Fragment ions of the +4 ion of modified ziconotide after CID at 35% collision energy. Observed Relative Theoretical Fragment ions m/z Intensity m/z m) 474.31 3.67 473.96, 474.57, 474.61, 474.76 b20N H3 5+ (0.35), y19H2O5 + (0.26), b4 (0.3), y19NH3 5+ (0.45) 553.06 4.74 553.01 b14 3+ (0.05) 606.19 8.87 606.4 b15 3+ (0.21) 657.81 10.36 657.77, 658.27 y11H2O2 + (0.04), y11NH3 2+ (0.46) 666.45 33.85 666.78 y11 2+ (0.33) 675.62 3.99 675.80, 675.87 y23 4+ (0.18), a12 2+ (0.25) 676.33 4.02 675.87 a12 2+ (0.46) 684.58 5.77 684.17 a17 3+ (0.41) 687.40 4.49 687.50, 687.83 b17H2O3+ (0.10), b17NH3 3+ (0.43) 693.32 7.66 693.50 b17 3+ (0.18) 699.46 5.41 699.08, 699.33 b24H2O4+ (0.38), b24NH3 4+ (0.13)

PAGE 160

160 T able A 1 Continued Observed Relative Theoretical Fragment ions m/z Intensity m/z m) 703.28 9.01 703.18, 703.34, 703.59 a 18 3+ (0.10), y 24 H 2 O 4 + (0.06), b 24 4+ or y 24 NH 3 4+ (0.31) 707.72 20.62 707.84 y24 4+ (0.12) 712.17 33.92 712.52 b18 3+ (0.35) 716.30 86.15 715.81 y12NH3 2+ (0.49) 796.26 7.05 796.28 y19 3+ (0.02) 739.00 52.24 819.78 17.15 819.97, 820.00 y20 3+ (0.19), b14H2O2+ (0.22) 828.57 38.61 829.00 b14 2+ (0.43) 829.25 14.83 829.00 b14 2+ (0.25) 832.98 12.57 832.98 y21H2O3+ (0.00) 838.59 40.68 838.99 y21 3+ (0.40) 839.35 20.53 838.99 y21 3+ (0.36) 846.91 8.73 847.01 b21 3+ (0.10) 869.99 5.91 870.03, 870.36 b22H2O3+ (0.04), b22NH3 3+ (0.37) 881.43 12.18 881.71 y22 3+ (0.28) 888.99 8.76 889.05, 889.38 b23H2O3+ (0.06), b23NH3 3+ (0.39) 889.88 4.70 889.38 b23NH3 3+ (0.50) 894.79 25.56 894.72, 895.05, 895.10 y23H2O3+ (0.07), b23 3+ or y23NH3 3+ (0.26), a15 2+ (0.31) *900.42 100.00 900.09, 900.59, 900.73 b15H2O2+ (0.33), b15NH3 2+ (0.17), y23 3+ (0.31) 901.13 34.72 900.73 y23 3+ (0.40) 931.55 41.83 931.77 b24H2O3+ (0.22) 932.23 26.19 932.10 b24NH3 3+ (0.13) 937.32 87.82 937.45 y24H2O3+ (0.13) 937.96 83.03 937.78 b24 3+ or y24NH3 3+ (0.18) 1154.43 5.13 1154.34 y10H2O (0.09) 1264.84 4.13 1261.50 b21NH3 2+ (0.3 4) MS3 of m/z 900.4 confirmed the identity of the fragment ion as y23 3+.

PAGE 161

161 Table A 2. Product ions of the +4 ion of modified conotoxin GVIA after CID at 35% collision energy. Observed Relative Theoretical Fragment ions m/z Intensity m/z m) 625.73 1.98 625.78 y4 (0.05) 726.49 1.17 726.88 y5 (0.39) 753.55 0.84 753.59 y24 4+ (0.04) 775.41 2.53 775.35 y25 4+ (0.06) 801.22 0.89 801.24, 801.57 y18H2O 3+ (0.02), y18NH3 3+ (0.35) 807.03 15.64 806.90, 807.24, 807.40 b 26 +H 2 O 4+ (0.13), y 18 3+ (0.21), y 26 4+ (0.37) 807.76 10.00 807.40 y26 4+ (0.36) 824.46 7.26 834.05 5.43 835.94 1.70 835.94 a8NH3 (0.00) 889.56 1.06 889.67 y20 3+ (0.11) 918.54 2.26 918.70 y21 3+ (0.16) 947.58 2.33 947.72 y22 3+ (0.14) 967.08 4.57 966.74 y23 3+ (0.34) 977.64 2.61 978.12 y14 2+ (0.48) 998.15 7.31 998.11 a17 2+ (0.04) 998.78 6.34 998.44, 998.77 y24H2O3+ (0.34), y24NH3 3+ (0.01) 1004.21 100.00 1044.44 y24 3+ (0.23) 1004.87 32.76 1027.25 17.83 1027.47 y25H2O3+ (0.22) 1028.16 5.85 1033.22 86.65 1033.13 a18NH3 2+ (0.09) 1033.92 27.05 1033.47 y25 3+ (0.45) 1034.66 6.56 1059.27 5.50 1059.71 y15 2+ (0.44) 1094.88 0.84 1094.74 y16NH3 2+ (0.14) 1154.22 1.61 1154.27 a11 (0.05) 1192.99 1.93 1192.80 b20 2+ (0.19) 1201.90 1.83 1201.85 y18NH3 2+ (0.05) 1209.68 10.01 1210.39 12.60 1210.36 y18 2+ (0.03)

PAGE 162

162 Table A 2 Continued Observed Relative Theoretical Fragment ions m/z Intensity m/z m) 1210.97 6.29 1253.23 6.4 1254.27 5.83 1253.90 y19 2+ (0.37) 1334.07 1.21 1334.00 y20 2+ (0.07) 1377.47 0.63 1377.54 y21 2+ (0.07) 1420.93 0.54 1421.08 y22 2+ (0.15) 1449.23 0.31 1449.60 y23 2+ (0.37)

PAGE 163

163 LIST OF REFERENCES 1. Craig, A. G., The Characterization of Conotoxins. Journal of Toxicology Toxin Reviews 2000, 19 (1), 5393. 2. Olivera, B. M., Rivier, J., Scott, J.K., Hillyard, D.R., Cruz, L.J. Conotoxins. Journal of Biological Chemistry 1991, 266, 2206722070. 3. Myers, R. A.; Cruz, L. J.; Rivier, J. E.; Olivera, B. M., Conus Peptides as Chemical Probes for Receptors and Ion Channels. Chemical Reviews 1993, 93 (5), 19231936. 4. Terlau, H.; Olivera, B. M., Conus Venoms: A Rich Source of Novel Ion Channel Targeted Peptides. Physiological Reviews 2004, 84 (1), 4168. 5. Olivera, B. M.; Rivier, J.; Clark, C.; Ramilo, C. A.; Corpuz, G. P.; Abogadie, F. C.; Mena, E. E.; Woodward, S. R.; Hillyard, D. R.; Cruz, L. J., Diversity of Conus Neuropeptides. Science 1990, 249 (4966), 257263. 6. Livett, B. G.; Gayler, K. R.; Khalil, Z., Drugs from the Sea: Conopeptides as Potential Therapeutics. Current Medicinal Chemistry 2004, 11 (13), 17151723. 7. Lewis, R. J.; Garcia, M. L., Therapeutic potential of venom peptides. Nature Reviews Drug Discovery 2003, 2 (10), 790802. 8. Adams, D. J.; Alewood, P. F.; Craik, D. J.; Drinkwater, R. D.; Lewis, R. J., C onotoxins and their Potential Pharmaceutical Applications. Drug Development Research 1999, 46 (3 4), 219234. 9. Newcomb, R.; Abbruscato, T. J.; Singh, T.; Nadasdi, L.; Davis, T. P.; Miljanich, G., Bioavailability of Ziconotide in Brain: Influx from Blood, Stability, and Diffusion. Peptides 2000, 21 (4), 491501. 10. Stix, G., A Toxin Against Pain. Scientific American 2005, 292 (4), 8893. 11. Wang, Y. X.; Bowersox, S. S., Analgesic Properties of Ziconotide, a Selective Blocker of N type Neuronal Calcium Channels. Cns Drug Reviews 2000, 6 (1), 120. 12. Olivera, B. M., Miljanich, G.P., Ramachandran, J., Adams, M.E. Calcium Channel Diver Agatoxins. Annual Review of Biochemistry 1994, 63, 823867. 13. Bowersox, S.; Mandema, J.; TarczyHornoch, K.; Miljanich, G.; Luther, R. R., Pharmacokinetics of SNX 111, a Selective N type Calcium Ch annel Blocker, in Rats and Cynomolgus Monkeys. Drug Metabolism and Disposition 1997, 25 (3), 379383. 14. Olivera, B. M.; Rivier, J.; Clark, C.; Ramilo, C. A.; Corpuz, G. P.; Abogadie, F. C.; Mena, E. E.; Woodward, S. R.; Hillyard, D. R.; Cruz, L. J., Dive rsity of Conus neuropeptides. Science 1990, 249 (4966), 25763.

PAGE 164

164 15. Olivera, B. M.; Cruz, L. J.; Desantos, V.; Lecheminant, G. W.; Griffin, D.; Zeikus, R.; Mcintosh, J. M.; Galyean, R.; Varga, J.; Gray, W. R.; Rivier, J., Neuronal Calcium Channel Antagonis ts Discrimination between Calcium Channel Subtypes Using Omega Conotoxin from Conus Magus Venom. Biochemistry 1987, 26 (8), 20862090. 16. Krishnamurthy, T.; Prabhakaran, M.; Long, S. R., Mass Spectrometric Investigations on Conus Peptides. Toxicon 1996, 34 (1112), 13451359. 17. Belva, H.; Lange, C., Conformational Studies of Omega Conotoxins Using Electrospray Mass Spectrometry. Rapid Communications in Mass Spectrometry 2000, 14 (15), 14331439. 18. Jones, A.; Bingham, J. P.; Gehrmann, J.; Bond, T.; Loughnan, M.; Atkins, A.; Lewis, R. J.; Alewood, P. F., Isolation and Characterization of Conopeptides by High Performance Liquid Chromatography Combined with Mass Spectrometry and Tandem Mass Spectrometry. Rapid Communications in Mass Spectrometry 1996, 10 (1), 138143. 19. Jakubowski, J. A.; Kelley, W. P.; Sweedler, J. V., Screening for Post Translational Modifications in Conotoxins using Liquid Chromatography/Mass Spectrometry: An Important Component of Conotoxin Discovery. Toxicon 2006, 47 (6), 688699. 20. Chung, D.; Gaur, S.; Bell, J. R.; Ramachandran, J.; Nadasdi, L., Determination of Disulfide Bridge Pattern in Omega Conopeptides. International Journal of Peptide and Protein Research 1995, 46 (3 4), 320325. 21. Moller, C.; Rahmankhah, S.; Lauer Fields J.; Bubis, J.; Fields, G. B.; Mari, F., A Novel Conotoxin Framework with a HelixLoopHelix (Cs Alpha/Alpha) Fold. Biochemistry 2005, 44 (49), 1598615996. 22. Berson, S. A.; Yalow, R. S., General Principles of Radioimmunoassay. Clinica Chimica Acta 2006, 369 125143. 23. Cunico, R. L.; Gooding, K. M.; Wehr, T., Basic HPLC and CE of Biomolecules Bay Bioanalytical Laboratory, Inc. : Richmond, CA, 1998. 24. Davis, M. T.; Stahl, D. C.; Hefta, S. A.; Lee, T. D., A Microscale Electrospray Interface for Onlin e, Capillary LiquidChromatography Tandem Mass Spectrometry of Complex Peptide Mixtures. Analytical Chemistry 1995, 67 (24), 45494556. 25. Stacey, C. C.; Kruppa, G. H.; Watson, C. H.; Wronka, J.; Laukien, F. H.; Banks, J. F.; Whitehouse, C. M., Reverse Ph ase LiquidChromatography ElectrosprayIonization Fourier Transform MassSpectrometry in the Analysis of Peptides. Rapid Communications in Mass Spectrometry 1994, 8 (7), 513516. 26. Emmett, M. R.; Caprioli, R. M., MicroElectrospray MassSpectrometry Ul tra High Sensitivity Analysis of Peptides and Proteins. Journal of the American Society for Mass Spectrometry 1994, 5 (7), 605613.

PAGE 165

165 27. Delahunty, C.; Yates, J. R., Protein Identification Using 2D LC MS/MS. Methods 2005, 35 (3), 248255. 28. Siuzdak, G., T he Expanding Role of Mass Spectrometry in Biotechnology MCC Press: San Diego, CA, 2003. 29. Cech, N. B.; Enke, C. G., Practical Implications of Some Recent Studies in Electrospray Ionization Fundamentals. Mass Spectrom Rev 2001, 20 (6), 36287. 30. Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M., Electrospray Ionization for Mass Spectrometry of Large Biomolecules. Science 1989, 246 (4926), 6471. 31. Glish, G. L.; Vachet, R. W., The Basics of Mass Spectrometry in the Twentyfirst Century. Nature Reviews Drug Discovery 2003, 2 (2), 140150. 32. Shen, Y. F.; Kim, J.; Strittmatter, E. F.; Jacobs, J. M.; Camp, D. G.; Fang, R. H.; Tolie, N.; Moore, R. J.; Smith, R. D., Characterization of the Human Blood Plasma Proteome. Proteomics 2005, 5 (15), 40344045. 33. March, R. E., An Introduction to Quadrupole Ion Trap Mass Spectrometry. Journal of Mass Spectrometry 1997, 32 (4), 351 369. 34. Schwartz, J. C.; Senko, M. W.; Syka, J. E. P., A TwoDimensional Quadrupole Ion Trap Mass Spectrometer. Journal of the American Society for Mass Spectrometry 2002, 13 (6), 659669. 35. Radial Ejection and Dual Detection on the Finnigan LTQ Linear Ion Trap Mass Spectrometer. Product Support Bulletin 117, Thermo Electron Corporation. 36. Omega conotoxin MVIIA (PDB I D: 1dw5); Protein Data Bank at the Research Collaboratory for Structural Informatics http://www.rcsb.org/pdb (accessed May 2007). 37. Barksdale, C. M.; Nordblom, G. D.; Miljanich, G. M.; TarczyHornoch, K.; Kristipa ti, R.; Kugler, A. R., Development of a Radioimmunoassay for the Novel N Type Calcium Channel Blocker. J. Clin. Ligand Assay 1996, 19 (4), 229233. 38. Lehninger, A. L.; Nelson, D. L.; Cox, M. M., Principles of Biochemistry. 2nd ed.; Worth Publishers: New York, 1993. 39. Parker, C. W., Radioimmunoassay. Annual Review of Pharmacology and Toxicology 1981, 21, 113132. 40. Poppe, H., Column Liquid Chromatography. In Chromatography 5th ed.; Heftmann, E., Ed. Elsevier: Amsterdam, 1992. 41. Finnigan LTQ Getting Started Manual (Revision B). Technical Publications, Thermo Electron Corporations: San Jose, CA, 2003.

PAGE 166

166 42. Miller, J. N.; Miller, J. C., Statistics and Chemometrics for Analytical Chemistry. 5th ed.; Pearson Prentice Hall: Harlow, England ; New York, 2005. 43. Almeida, A. M.; Castel Branco, M. M.; Falcao, A. C., Linear Regression for Calibration Lines Revisited: Weighting Schemes for Bioanalytical Methods. Journal of Chromatography B 2002, 774 (2), 215222. 44. Johnson, E. L., Reynolds, D.L., Wright, D.S., Pachla, L.A., Biological Sample Preparation and Data Reduction Concepts in Pharmaceutical Analysis. Journal of Chromatographic Science 1988, 26, 372379. 45. Finnigan LTQ Hardware Manual ( Revision B ). Technical Publications, Thermo Electron Corporations: S an Jose, CA, 2003. 46. Geng, X. D.; Regnier, F. E., Retention Model for Proteins in ReversedPhase Liquid Chromatography. Journal of Chromatography 1984, 296 (Jul), 15 30. 47. Apffel, A.; Fischer, S.; Goldberg, G.; Goodley, P. C.; Kuhlmann, F. E., Enhanced Sensitivity for Peptide Mapping with Electrospray Liquid Chromatography Mass Spectrometry in the Presence of Signal Suppression Due to Trifluoroacetic AcidContaining Mobile Phases. Journal of Chromatography A 1995, 712 (1), 177 190. 48. Jakubowski, J. A.; Sweedler, J. V., Sequencing and Mass Profiling Highly Modified Conotoxins Using Global Reduction/Alkylation Followed by Mass Spectrometry. Analytical Chemistry 2004, 76 (22), 65416547. 49. Jakubowski, J. A.; Keays, D. A.; Kelley, W. P.; Sandall, D. W.; Bingham, J. P.; Livett, B. G.; Gayler, K. R.; Sweedler, J. V., Determining Sequences and Post Translational Modifications of Novel Conotoxins in Conus victoriae using cDNA Sequencing and Mass Spectrometry. Journal of Mass Spectrometry 2004, 39 (5), 548 557. 50. Synder, L. R.; Kirkland, J. J.; Glaich, J. L., Practical HPLC Method Development John Wiley & Sons: NY, 1988. 51. Lagerwerf, F. M.; van Dongen, W. D.; Steenvoorden, R. J. J. M.; Honing, M.; Jonkman, J. H. G., Exploring the Boundaries of Bioanalytica l Quantitative LC MS MS. Trac Trends in Analytical Chemistry 2000, 19 (7), 418427. 52. Skoog, D. A.; Leary, J. J., Principles of Instrumental Analysis 4th ed.; Saunders College Publishing: 1992. 53. Kaiser, H., Quantitation in Elemental Analysis 2. Analy tical Chemistry 1970, 42 (4), A26. 54. Long, G. L.; Winefordner, J. D., Limit of Detection. Analytical Chemistry 1983, 55 (7), A712.

PAGE 167

167 55. Food and Drug Administration Guidance for Industry Bioanalytical Method Validation, 2001, pp. 120. http://www.fda.gov/CDER/GUIDANCE/4252fnl.htm (accessed January 2007). 56. Shields, D.; Montenegro, R.; Ragusa, M., Chemical Stability of Admixtures Combining Ziconotide with Morphine or Hydromorphone during Simula ted Intrathecal Administration. Neuromodulation 2005, 8 (4), 257263. 57. Reubsaet, J. L.; Beijnen, J. H.; Bult, A.; van Maanen, R. J.; Marchal, J. A.; Underberg, W. J., Analytical Techniques Used to Study the Degradation of Proteins and Peptides: Physical Instability. J Pharm Biomed Anal 1998, 17 (6 7), 979984. 58. Manning, M. C.; Patel, K.; Borchardt, R. T., Stability of Protein Pharmaceuticals. Pharmaceutical Research 1989, 6 (11), 903918. 59. Price Carter, M.; Hull, M. S.; Goldenberg, D. P., Roles of Individual Disulfide Bonds in the Stability and Folding of an Omega Conotoxin. Biochemistry 1998, 37 (27), 98519861. 60. Yen, T. Y.; Yan, H.; Macher, B. A., Characterizing Closely Spaced, Complex Disulfide Bond Patterns in Peptides and Proteins by Liquid Chromatography/Electrospray Ionization Tandem Mass Spectrometry. J Mass Spectrom 2002, 37 (1), 1530. 61. Gray, W. R., Disulfide Structures of Highly Bridged Peptides: A New Strategy for Analysis. Protein Sci 1993, 2 (10), 173248. 62. Neitz, S.; Jurgens, M.; Kellmann, M.; Schulz Knappe, P.; Schrader, M., Screening for Disulfide Rich Peptides in Biological Sources by Carboxyamidomethylation in Combination with Differential Matrix Assisted Laser Desorption/Ionization Timeof Flight Mass Spectrometry. Rapid C ommunications in Mass Spectrometry 2001, 15 (17), 15861592. 63. Bean, M. F.; Carr, S. A., Characterization of Disulfide Bond Position in Proteins and Sequence Analysis of Cystine Bridged Peptides by Tandem Mass Spectrometry. Analytical Biochemistry 1992, 201 (2), 216226. 64. Stults, J. T.; Bourell, J. H.; Canovadavis, E.; Ling, V. T.; Laramee, G. R.; Winslow, J. W.; Griffin, P. R.; Rinderknecht, E.; Vandlen, R. L., Structural Characterization by Mass Spectrometry of Native and Recombinant Human Relaxin. B iomedical and Environmental Mass Spectrometry 1990, 19 (11), 655664. 65. Stephenson, J. L.; Cargile, B. J.; McLuckey, S. A., Ion Trap Collisional Activation of Disulfide Linkage Intact and Reduced Multiply Protonated Polypeptides. Rapid Communications in Mass Spectrometry 1999, 13 (20), 20402048. 66. Fainzilber, M.; Nakamura, T.; Gaathon, A.; Lodder, J. C.; Kits, K. S.; Burlingame, A. L.; Zlotkin, E., A New Cysteine Framework in Sodium Channel Blocking Conotoxins. Biochemistry 1995, 34 (27), 86498656.

PAGE 168

168 67. Lamoureux, G. V.; Whitesides, G. M., Synthesis of Dithiols as Reducing Agents for Disulfides in Neutral Aqueous Solution and Comparison of Reduction Potentials. Journal of Organic Chemistry 1993, 58 (3), 633641. 68. Lundell, N.; Schreitmuller, T., Sampl e Preparation for Peptide Mapping A Pharmaceutical Quality Control Perspective. Analytical Biochemistry 1999, 266 (1), 3147. 69. Mendoza, J. A.; Zientek, K. D.; Eyler, J. R.; Prokai, L., Development of an LC/MSn Method for the Detection of Ziconotide in Horse Plasma. In Proceedings of the 55th ASMS Conference and Allied Topics Indianapolis, IN, 2007. 70. MS Product Software from ProteinProspector (v. 4.27.2 Basic) at the Mass Spectrometry Facility of the University of California, San Francisco. http://prospector.ucsf.edu/cgi bin/msform.cgi?form=msproduct (accessed April 2008). 71. Ingle, J. D.; Crouch, S. R., Spectrochemical Analysis. Prentice Hall: Upper Saddle River, N.J., 1988; p 171176. 72. Abbatiello, S. E. L. Mass Spectrometric Studies of Asparagine Synthetase and its Role in the DrugResistant Form of Acute Lymphoblastic Leukemia. University of Florida, Gainesville, FL, 2006. 73. Causon, R., Validation of Chromatographic Methods in Biomedical Analysis Viewpoint and Discussion. Journal of Chromatography B 1997, 689 (1), 175180. 74. Environmental Protection Agency, 1990, 40 Code of Federal Regulations Part 136, Appendix B. Definition and Procedure for the Determination o f the Method Detection Limit Revision 1.11. http://www.setonresourcecenter.com/CFR/40CFR/P136_008.HTM (accessed April 2009). 75. Stability Considerations in Dispensing Practice. I n United States Pharmacopeia, 30th ed.; United States Pharmacopeial Convention, Inc.: Rockville, MD, 2007. 76. O'Donnell, P. B.; Bokser, A. D., Stability of Pharmaceutical Products. In Remington, the Science and Practice of Pharmacy 21st ed.; Lippincott Wi lliams and Wilkins: Baltimore, MD, 2006; pp 10251036. 77. Shields, D. E.; Liu, W.; Gunning, K.; Montenegro, R., Statistical Evaluation of the Chemical Stability of Ziconotide Solutions during Simulated Intrathecal Administration. J Pain Symptom Manage 2008, 36 (1), e46. 78. PRIALT package insert. San Diego, CA: Elan Pharmaceuticals, Inc.: 2007. 79. Atkins, P. W., Physical Chemistry. 6th ed.; W.H. Freeman and Company: NY, 1998. 80. Kennon, L., Use of Models in Determining Chemical Pharmaceutical Stability. Journal of Pharmaceutical Sciences 1964, 53 (7), 815818.

PAGE 169

169 81. Pharmaceutical Stability. In United States Pharmacopeia, 30th ed.; United States Pharmacopeial Convention, Inc.: Rockville, MD, 2007. 82. Biotechnologyderived Articles. In United States Pharmacopeia, 23rd ed.; United States Pharmacopeial Convention, Inc.: Rockville, MD, 2005. 83. Wan, H.; Umstot, E. S.; Szeto, H. H.; Schiller, P. W.; Desiderio, D. M., Quantitative Analysis of [Dmt(1)]DALDA in Ovine Plasma by Capillary Liq uid ChromatographyNanospray IonTrap Mass Spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 2004, 803 (1), 8390. 84. Karas, M.; Bahr, U.; Dulcks, T., Nanoelectrospray Ionization Mass Spectrometry: Addressing Analytical Problems Beyond Routine. Fresenius Journal of Analytical Chemistry 2000, 366 (6 7), 669676. 85. Schmidt, A.; Karas, M.; Dulcks, T., Effect of Different Solution Flow Rates on Analyte Ion Signals in NanoESI MS, or: When Does ESI Turn Into NanoESI? Journal of the American Societ y for Mass Spectrometry 2003, 14 (5), 492 500. 86. Westermeier, R.; Naven, T., Proteomics in Practice. Wiley VCH: Weinheim, 2002. 87. Walker, J. M., The Protein Protocols Handbook Humana Press: Totowa, NJ, 2002.

PAGE 170

170 BIOGRAPHICAL SKETCH Jhoana Avendao Mend oza was born in Bataan, Philippines in 1977. Her acceptance as a scholar at the Philippine Science High School greatly nurtured her interest in the sciences. In April 1998, she completed her Bachelor of Science degree in Chemistry at the University of t he Philippines in Diliman, Quezon City, where she worked on the development and validation of a potentiometric method for the determination of organophosphates in natural waters under the direction of Dr. Cynthia Grace Gregorio She worked as a research a ssistant at the Seaweed Chemistry section of the Marine Science Institute, University of the Philippines where she tested the feasibility of a carrageenophyte as a test organism for toxicity testing before joining the Institute of Chemistry as a junior faculty. From 1999 to 2003, she taught and supervised laboratory classes in general, analytical and organic chemistry in the same institute where she earned her undergraduate degree. In August 2003, she started graduate studies at the Chemistry Department of the University of Florida where she joined the research group of Dr. John R. Eyler. Collaborative efforts between the Chemistry Department and the College of Veterinary Medicine Racing Laboratory paved the way for her to work on developing liquid chromatography and mass spectrometric methods to analyze small molecules and peptides. She plans to explore the vast field of research opportunities where she can use and further improve her knowledge and skills in liquid chromatography mass spectrometric m ethod development to study and analyze complex samples.