Silica Colloidal Crystals As Emerging Materials For The

Purdue UniversityPurdue e-Pubs

Open Access Dissertations Theses and Dissertations

Fall 2013

Silica Colloidal Crystals As Emerging Materials ForThe High-Throughput Sizing Of Proteins ByPacked Capillary ElectrophoresisNadine K. NjoyaPurdue University

Follow this and additional works at: https://docs.lib.purdue.edu/open_access_dissertations

Part of the Analytical Chemistry Commons

This document has been made available through Purdue e-Pubs, a service of the Purdue University Libraries. Please contact [email protected] foradditional information.

Recommended CitationNjoya, Nadine K., "Silica Colloidal Crystals As Emerging Materials For The High-Throughput Sizing Of Proteins By Packed CapillaryElectrophoresis" (2013). Open Access Dissertations. 84.https://docs.lib.purdue.edu/open_access_dissertations/84

https://docs.lib.purdue.edu?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F84&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/open_access_dissertations?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F84&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/etd?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F84&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/open_access_dissertations?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F84&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/132?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F84&utm_medium=PDF&utm_campaign=PDFCoverPages

https://docs.lib.purdue.edu/open_access_dissertations/84?utm_source=docs.lib.purdue.edu%2Fopen_access_dissertations%2F84&utm_medium=PDF&utm_campaign=PDFCoverPages

Graduate School ETD Form 9 (Revised 12/07)

PURDUE UNIVERSITY GRADUATE SCHOOL

Thesis/Dissertation Acceptance

This is to certify that the thesis/dissertation prepared

By

Entitled

For the degree of

Is approved by the final examining committee:

Chair

To the best of my knowledge and as understood by the student in the Research Integrity and Copyright Disclaimer (Graduate School Form 20), this thesis/dissertation adheres to the provisions of Purdue University’s “Policy on Integrity in Research” and the use of copyrighted material.

Approved by Major Professor(s): ____________________________________

____________________________________

Approved by: Head of the Graduate Program Date

Nadine K. Njoya

SILICA COLLOIDAL CRYSTALS AS EMERGING MATERIALS FOR THEHIGH-THROUGHPUT SIZING OF PROTEINS BY PACKED CAPILLARYELECTROPHORESIS

Doctor of Philosophy

Mary J. Wirth

David A. Colby

Kavita Shah

David H. Thompson

Mary J. Wirth

R. E. Wild 11-19-2013

i

SILICA COLLOIDAL CRYSTALS AS EMERGING MATERIALS FOR THE HIGH-

THROUGHPUT SIZING OF PROTEINS BY PACKED CAPILLARY

ELECTROPHORESIS

A Dissertation

Submitted to the Faculty

of

Purdue University

by

Nadine K. Njoya

In partial Fulfillment of the

Requirements of the Degree

of

Doctor of Philosophy

December 2013

Purdue University

West Lafayette, Indiana

ii

To My Family.

Despite the distance, I carry each and every one of you close to my heart.

iii

ACKNOWLEDGMENTS

I would like to thank my research advisor, Professor Mary J. Wirth, for her

guidance and mentoring. I wouldn’t have been able to achieve this milestone without her

faith in me and constant encouragements. My committee members are acknowledged for

their mentoring and accessibility whenever I needed advice regarding my graduate career.

I would like to thank Professor Kenttӓmaa and Professor Lipton for their guidance on

scientific projects I undertook while under their supervision. Thank you to all past and

currents members of the Wirth group: Dr. Yimin Hua, Dr. Bingchuan Wei, Dr. Zhaorui

Zhang, Dr. Brooke Koshel, Dr. Benjamin Rogers, Charu Yerneni, Oyeleye Alabi, Natalya

Khanina, Zhen Wu, Ao Zheng, Pei-Hsun Wei, Yao Wang, Xiang Cao, Ximo Zhang,

Gavin Jones, Alexis Huckabee and Yingxu Hao. I would like to especially thank Dr.

Robert Birdsall for all his help, guidance and patience. I definitely learned a lot from you

and your incredible work ethic. The following individuals have made my time here at

Purdue University even more enjoyable: Emmanuela Ohaeri, Christopher Benjamin,

James Riedeman, “Papa” Yawo Mondjinou (King of Togo), Dr. Nelson Vinueza, Dr.

Vanessa Gallardo, Dr. Daneli Lopez-Perez, Dr. Gurusankar Ramamoorthy, Dr. Joseph

Greene, Dr. Nawaporn Sanguatrakun, Dr. Sarah St John. Thank you for you friendship,

shoulders to cry on, honesty and most importantly, helping me keep my sanity during this

iv

endeavor. Thank you to all the Purdue Chemistry departmental staff for the wonderful job

you do at taking care of students’ needs. I would like to give a special thank you to

Robert Reason for always listening, all the encouragements and most importantly, the

coffee.

I cannot find the words to thank my family for their unlimited love and support.

You have been my rock and the only constant in my life for long as I can remember.

Thank you for being so awesome and even though we haven’t seen each other for years

I want you know that I am very grateful and humbled to have each and every one of you

in my life.

Finally I would like to thank Travis for his love and support. Many times, when I

felt like I couldn’t handle all of my workload, you have kept me motived and focused on

the bigger picture. Thank you for all your understanding, especially during the writing of

my thesis. You are amazing in many ways and thank you for putting up with me.

v

TABLE OF CONTENTS

Page

LIST OF TABLES ........................................................................................................... viii

LIST OF FIGURES ........................................................................................................... ix

LIST OF SCHEMES........................................................................................................ xiv

LIST OF ABBREVIATIONS .......................................................................................... xvi

ABSTRACT ................................................................................................................... xviii

PART ONE: SILICA COLLOIDAL CRYSTALS AS EMERGING MATERIALS FOR THE HIGH-THROUGHPUT SIZING OF PROTEINS BY PACKED CAPILLARY ELECTROPHORESIS .................................................................................1

CHAPTER 1: INTRODUCTION ........................................................................................2

1.1 Importance of Homogeneity in Protein Drug Formulations ....................................2 1.2 Current Analytical Techniques for Aggregate Characterization .............................3

1.2.1 Sodium Dodecyl Sulfate - Polyacrylamide Gel ..............................................4 1.2.2 Sodium Dodecyl Sulfate - Capillary Electrophoresis .....................................6 1.2.3 Size Exclusion Chromatography .....................................................................8

1.3 Gel Electrophoresis: Band Broadening and Resolution .........................................10 1.4 Research Goal ........................................................................................................13

CHAPTER 2: THERMAL TREATMENT, PACKING AND CHEMICAL MODIFICATION OF SILICA COLLOIDAL CRYSTALS FOR CAPILLARY ELECTROPHORESIS .......................................................................................................16

2.1 Background on silica colloidal crystals .................................................................16 2.2 Chemicals and Materials ........................................................................................18 2.3 Thermal Treatment of the SCCs and Rehydroxylation ..........................................18 2.4 Fused Silica Capillaries Preparation and Pressure Packing ...................................19 2.5 In situ Chemical Treatment of the SCCs Packed Bed ...........................................20

vi

Page

CHAPTER 3: HIGH-THROUGHPUT PROTEIN SIZING USING NON POROUS SILICA COLLOIDAL CRYSTALS BEARING A BRUSH LAYER OF POLYACRYLAMIDE ........................................................................26

3.1 Recent Developments in Electrophoresis of Proteins Using SCCs as the Sieving Medium .....................................................................................................26

3.2 Experimental ..........................................................................................................27 3.2.1 Chemicals and Materials ...............................................................................27 3.2.2 Preparation of the Silica Particles for Electrophoresis .................................28 3.2.3 PDMS Reservoirs Preparation ......................................................................29 3.2.4 Sample Preparation for Electrophoresis........................................................29 3.2.5 Capillary Sieving Electrophoresis Apparatus ...............................................30

3.3 Results and Discussion ..........................................................................................32 3.3.1 Rationale behind the Pore Size Determination for Antibody Aggregates

Detection .......................................................................................................32 3.3.2 Aggregation of the mAb and Confirmation by SEC .....................................40 3.3.3 Capillary Sieving Electrophoresis of Aggregated mAb ................................42 3.3.4 Particle Size, Migration Distance and Voltage Effects on Separation

Resolution .....................................................................................................43 3.3.5 Preliminary Results Using AGET ATRP .....................................................49

3.4 Concluding Remarks ..............................................................................................51

PART TWO: SYNTHESIS AND FRAGMENTATION STUDIES OF ASPHALTENE MODEL COMPOUNDS CAUSED BY COLLISION-ACTIVATED DISSOCIATION USING A LINEAR QUADRUPOLE ION TRAP ...............................52

CHAPTER 4: BACKGROUND ON ASPHALTENES, MODEL COMPOUNDS SYNTHESIS AND INSTRUMENTATION .....................................................................53

4.1 Asphaltenes Overview ...........................................................................................53 4.2 Synthesis of the Asphaltene Model Compounds ...................................................55

4.2.1 Synthesis of the Alkyl and Phenyl-Substituted Naphthalenes ......................55 4.2.2 Synthesis of the Alkyl and Phenyl-Substituted Isoquinolines ......................56

4.3 Linear Quadrupole Ion Trap (LQIT) Mass Spectrometry ......................................58 4.3.1 Atmospheric Pressure Chemical Ionization (APCI) .....................................58 4.3.2 LQIT Instrumentation ...................................................................................60

CHAPTER 5: STUDY OF COLLISION-ACTIVATED DISSOCIATION OF IONIZED ASPHALTENES’ MODEL COMPOUNDS USING A LINEAR QUADRUPOLE ION TRAP MASS SPECTROMETER .................................................64

5.1 Naphthalene-Based Model Compounds ................................................................64 5.2 Isoquinoline-Based Model Compounds .................................................................69 5.3 Concluding Remarks and Perspectives ..................................................................76

vii

Page

PART THREE: TOWARD THE SYNTHESIS THE PHOTOAFFINITY LABELED ANALOG AND MODE OF ACTION STUDIES OF CALLIPELTIN A ........................78 CHAPTER 6: SYNTHESIS OF THE β-MeO-tyr AZIDE ANALOG OF A NON-PROTEOGENIC AMINO ACID RESIDUE OF CALLIPELTIN A ................................79

6.1 Background on Callipeltins and Mirabamides .......................................................79 6.2 Results and Discussion ..........................................................................................82 6.3 Conclusion and Future Directions .........................................................................96

LIST OF REFERENCES ...................................................................................................98

APPENDICES

Appendix A: Supplemental Data for Chapters 4 and 5 ....................................................110 A.1 1H and 13C NMR of Asphaltene Model Compounds ..........................................111 A.2 NMR Spectra .......................................................................................................114 A.3 Asphaltene Model Compounds’ Mass Spectral Data .........................................133

Appendix B: Supplemental Data for Chapter 6 ...............................................................140 B.1 Experimental Procedures, 1H and 13C NMR Characterizations ..........................140 B.2 NMR Spectra .......................................................................................................145

VITA ................................................................................................................................155

PUBLICATION ...............................................................................................................156

viii

LIST OF TABLES

Table Page

1.1 Common Size Based Analytical Techniques for Detection of Protein Aggregates .....................................................................................................................4

4.1 Yields Obtained After Isoquinoline Syntheses ............................................................57

6.1 Conditions for the Reduction of Compound 19 ...........................................................85

6.2 Conditions for the Nitro Group Reduction of Compound 22 ......................................87

6.3 Conditions for the Nosyl Group Removal to Obtain Compound 26 ...........................88

6.4 Conditions for the Formation of Benzophenone Dimethyl Acetal ..............................94

Appendix Table

A.1 Asphaltene Model Compounds’ Mass Spectral Data ...............................................133

ix

LIST OF FIGURES

Figure Page

1.1 A. SDS mode of action and B. Migration illustration for the size dependent separation of three proteins through the porous network of a vertical slab SDS polyacrylamide gel. ........................................................................................................5

1.2 Schematic for capillary electrophoresis which shows charged analyte moving through the sieving material under an electric field produced when voltage is supplied. Analyte is detected when it passes through a detection window... ................7

1.3 A. SEC mechanism of action and B. Size exclusion chromatogram of an aged monoclonal antibody (74oC, 10 minutes) using a Sepax Technologies Zenix SEC-300, 3 µm, 300 Å, 4.6x250 mm. Mobile phase: 150 mM sodium phosphate buffer, pH: 7.0. Flow rate: 0.4 mL/min. Temperature: 25oC ........................9

1.4 Depiction of the differences in motion paths represented by arrows for two protein zones in A. Homogeneous medium and B. Heterogeneous medium. The differing red and green paths are caused by micro-heterogeneity in the gel pore network and cause broadening .............................................................................11

1.5 Simulated plots of Gaussian functions vs. separation length to illustrate effect of plate height, H, on resolution. A. For a plate height of 10 μm, the two Gaussians are unresolved in 1 cm of separation length, but they are baseline resolved in 10 cm (Image of heterogeneous PAAm medium is adapted from reference (1). B. For a plate height of 1 μm, the Gaussians are baseline resolved in only 1 cm ...................................................................................................14

2.1 High order packing of colloid crystals shown in A. Naturally occurring gem opal exhibiting bright colors due to Bragg diffraction (2) and B. Scanning electron microscope of SCCs in a planar channel. (Courtesy of Dr. Robert Birdsall)........................................................................................................................17

2.2 Illustration of the pressure packing of a fused silica capillary filled with a SCC slurry. The sonication is needed for a uniform and denser packed bed .......................20

x

Figure Page

2.3 Zoom microscope images of A. Densely packed plug with no obvious cracks. The black portion is dry while the brown one is wet. B and C show cracks and gaps where drying has occurred and which result from fast depressurization. ...........23

2.4 2.4 A. Zoom microscope image of a 75 µm i.d. capillary packed with 250 nm nonporous silica; the blue color is attributed to the Bragg diffraction of water which fills the free fractional volume in the packed bed. B. SEM image of the cross section of a 75 µm i.d. capillary packed with 470 nm SCCs which shows localized regions of high crystalline order (Image 2.4 B reproduced from reference (3)) C. SEM image of the cross section of a 75 µm i.d. capillary pressure packed with 700 nm SCCs. D. SEM image of the cross section of a 100 µm i.d. capillary pressure packed with 950 nm SCCs. Larger particles pressure packing is mostly random, however some long range order can be observed at the wall... ..................................................................................................24

3.1 A. Illustration of the sieving electrophoresis set-up. B. Actual laboratory set-up..................................................................................................................................31

3.2 Electrophoresis data for four proteins using a silica colloidal crystal as the medium. Proteins A–D are lysozyme, trypsin inhibitor, carbonic anhydrase, and bovine serum albumin, respectively. A. Electropherogram of the proteins, with data points (circles), and a fit to four Gaussians (line). B. Plot of electrophoresis data using Equation 3.1. Pore radius is indicated by ξ. Both panels are adapted from Birdsall and Wirth (4). ..........................................................33

3.3 Gel electropherogram of a monoclonal antibody initially (lane 4) and after storage (lanes 5–9). The size standards are shown in lane 2. The figure is adapted from Mahler et al. (5) .....................................................................................34

3.4 Comparison of selectivity for gels and colloidal crystals in protein electrophoresis. A. Data points (red circles) for the size standards of the gel in Figure 3.3 and fit to Equation 3.2 (red line). Plot of Equation 3.1 (blue line, blue circles). B. Theoretical plots for gel (red line) and colloidal crystal (blue line) for pore sizes optimized for resolving an antibody from its dimer. In both panels, the antibody molecular weight (MW) (150 kDa) and its dimer MW (300 kDa) are indicated by the dotted lines .................................................................36

3.5 A. function of pore radius to show the trade-off. B. Plot of plate height vs. the square of the broadening function of Equation 3.3 for a pore radius of 27 nm. The line is least-square fits of data points ....................................................................38

3.6 Illustration of the mobility dependence of an analyte on the pore size of the sieving medium A. For a protein of low hydrodynamic radius. B. For a protein of large hydrodynamic radius ......................................................................................40

xi

Figure Page

3.7 Size exclusion chromatograph of A. Alexafluor 546 labeled mAb. B. Alexafluor 546 labeled aggregated mAb (74.5 oC, 10 min) ........................................41

3.8 Calibration curve for the Zenix SEC-300 (3 µm, 300 Å, 4.6x250 mm) column .........42

3.9 Separation of monoclonal antibody monomer and dimer. A. False-color fluorescence image of antibody followed by its dimer as they electromigrate in SDS and buffer through a 100-μm-i.d. capillary packed with silica colloidal crystal and reach separation length of 6.5 mm in 72 s. B. corresponding electropherogram. ........................................................................................................44

3.10 Separation of monoclonal antibody monomer and dimer. A. False-color fluorescence image of antibody followed by its dimer as they electromigrate in SDS and buffer through a 75-μm-i.d. capillary packed with silica colloidal crystal and reach separation length of 2.3 cm in about 4 minutes. Corresponding electropherogram where data points are presented as dots and the lines represent the Gaussian fits of the protein zones... .........................................45

3.11 Voltage effects on the separation of monoclonal antibody monomer and dimer at a constant detection point of 1.7 cm. .............................................................46

3.12 A. Plot of the electric field (E) versus electrophoretic mobility. B. Plot of E versus separation time ..................................................................................................48

3.13 Electropherograms of Alexafluor 488 labeled A. Bovine albumin serum (BSA), B. Carbonic anhydrase (CA) and C. a mixture of both using 500 nm SCCs bearing a brush layer of PAAm packed in 100-id fused silica capillaries. Operating voltage and separation length (L) are indicated ..........................................50

4.1 APCI illustration ..........................................................................................................59

4.2 Illustration of the Thermo Finnigan LQIT mass spectrometer with some representative operating conditions. (Figure adapted from the Finnigan LTQ Hardware Manual 970055-97013 Revision B, Chapter 2, Page 14 and ThermoFisher Scientific LTQ Operation Courses from the Thermo Scientific Training Institute) ..............................................................................................................61

4.3 Illustration of the linear quadrupole ion trap (Figure reproduced from reference (6)) ......................................................................................................................62

5.1 CAD (collision energy of 20) mass spectra of isolated molecular ions of regiomeric model compounds (APCI, CS2) A. 1-butyl-3-propylnaphthalene and B. 3-butyl-1-propylnaphthalene ............................................................................65

xii

Figure Page

5.2 CAD (collision energy of 20) mass spectra of isolated molecular ion of 3-butyl-2-ethyl-1-phenylnaphthalene (APCI, CS2) .........................................................68

5.3 CAD (collision energy of 25) mass spectra of isolated molecular ion of 1-benzylisoquinoline (APCI, CS2) .................................................................................70

5.4 CAD (collision energy of 30) mass spectra of isolated molecular ion of 4-methyl-3-phenylisoquinoline (APCI, CS2). .................................................................72

5.5 CAD (collision energy of 25) mass spectra of isolated molecular ion of 4-ethyl-3-phenylisoquinoline (APCI, CS2). ....................................................................72

5.6 CAD (collision energy of 40) mass spectra of isolated molecular ion of 3,4-diphenylisoquinoline (APCI, CS2). ..............................................................................74

5.7 CAD (collision energy of 45) mass spectra of isolated molecular ions of regiomeric model compounds (APCI, CS2) A. 3-phenyl-4-(phenylethynyl)isoquinoline and B. 4-phenyl-3-(phenylethynyl)isoquinoline ............75

6.1 Structure of Callipeltin A .............................................................................................80

6.2 Structures of Mirabamides (11a-d) ..............................................................................81

6.3 Biotinylated analog of Callipeltin A ............................................................................82

6.4 Blown-up Area of the H1, H2 and H3 peaks from the 1H NMR spectrum of 25 (conditions “a”) ............................................................................................................90

6.5 Blown-up area of the H1, H2 and H3 peaks from the 1H NMR spectrum of 25 (conditions “b”) ...........................................................................................................91

6.6 Blown-up Area of the H1, H2 and H3 peaks from the 1H NMR spectrum of 16 ..........92

6.7 19F NMR for the R and S enantiomers of compound 33 ..............................................96

Appendix Figure

A.1 300 MHz 1H NMR of compound 1 in CDCl3 ...........................................................115

A.2 75 MHz 13C NMR of compound 1 in CDCl3. ...........................................................116

A.3 300 MHz 1H NMR of compound 2 in CDCl3. ..........................................................117


xiii

Appendix Figure Page

A.5 300 MHz 1H NMR of compound 3 in CDCl3. ..........................................................119


A.7 300 MHz 1H NMR of compound 4 in CDCl3 ...........................................................121


A.9 100 MHz 13C NMR of compound 5 in CDCl3. .........................................................123


A.11 300 MHz 13C NMR of compound 6 in CDCl3. .......................................................125








B.1 300 MHz 1H NMR of compound 24 in CDCl3. ........................................................146


B.3 75 MHz 13C NMR of compound 25 in CDCl3. .........................................................148


B.5 75 MHz 13C NMR of compound 25 in CDCl3 ..........................................................150

B.6 300 MHz 1H NMR of compound 33 in CDCl3 .........................................................151

B.7 282 MHz 19F NMR of compound 33 in CDCl3. ........................................................152

B.8 300 MHz 1H NMR of enantiomer of compound 33 in CDCl3. .................................153

B.9 282 MHz 19F NMR of enantiomer of compound 33 in CDCl3. ................................154

xiv

LIST OF SCHEMES

Scheme Page

2.1 Silane modification and atom transfer radical polymerization (ATRP) (7) ................22

4.1 Synthesis of the alkyl and phenyl-substituted naphthalenes... .....................................55

4.2 Synthesis of the alkyl and phenyl-substituted isoquinolines. For compounds 5-9, a palladium-catalyzed annulation of internal alkynes with an imine intermediate is used to obtain the isoquinolines ..........................................................56

4.3 Ionization process in APCI where CS2 is used as solvent ...........................................60

5.1 Proposed mechanism for the formation of fragment of m/z 184 yielded by CAD of 3-butyl-1-propylnaphthalene (1) ....................................................................65

5.2 Proposed mechanism for the formation of fragments of m/z 246 and 231 yielded by CAD of 3-butyl-2-ethyl-1-phenylnaphthalene (3). ....................................68

5.3 A. Proposed mechanism for the formation of the isoquinolium fragment (m/z 218) by hydrogen radical loss from the molecular ion of 1-benzylisoquinoline (4). B. Proposed mechanism for the formation of fragments of m/z 142 and 160................................................................................................................................71

5.4 A. Proposed mechanism for the formation of the ortho-pyridyne fragment (m/z 304) by hydrogen radical loss from the molecular ion of 3-phenyl-4-(phenylethynyl)isoquinoline (8). B. Proposed mechanism for the formation of fragments of m/z 228 ...................................................................................................76

6.1 Synthesis of the TBS protected 4-hydroxy methylcinnamate......................................83

6.2 Aziridination of compound 15 .....................................................................................83

6.3 Synthesis of the β-MeOTyr moiety and determination of the enantiomeric excess. ..........................................................................................................................84

6.4 Synthesis of 4-Azidobenzaldehyde ..............................................................................84

xv

Scheme Page

6.5. Resonance Mechanism of 20. .....................................................................................85

6.6 Alternate route for the synthesis of 4-azido methylcinnamate ....................................86

6.7 Synthesis of the β-MeOTyr azide analog 26 ...............................................................87

6.8. Conditions for the separation of the diastereomers of 25. ..........................................89

6.9 Formation of CAD 28 ..................................................................................................93

6.10 Conditions attempted for the direct formation of compound 28a. .............................93

6.11 Other conditions attempted for the synthesis of compound 28a. ...............................94

6.12 Enantiomeric ratio determination of compound 26 ...................................................95

xvi

LIST OF ABBREVIATIONS

AAm Acrylamide

AGET Activator generated by electron transfer

ATRP Atom transfer radical polymerization

BC (chloromethyl)phenylethyl‐trichlorosilane

C1 methyltrichlorosilane

CCD Charged-coupled device

CGE Capillary gel electrophoresis

CSA Camphorsulfonic acid

DBU 1,8-Diazabicyclo[5.4.0]undec-7-ene

DIEA N,N-Diisopropylethylamine

DMF N,N-Dimethylformamide

DNA Deoxyribonucleic acid

dsDNA Double stranded deoxyribonucleic acid

ED50 Median effective dose

EDTA Ethylenediaminetetraacetic acid

EOF Electroosmotic flow

xvii

EtOH Ethanol

EtOAc Ethyl acetate

FCC Face centered cubic

FDA Food and Drug Administration

HOAc Acetic acid

IgG Immunoglobulin (antibody)

Me6TREN Tris[2-(dimethylamino)ethyl]amine

NMR Nuclear magnetic resonance

PAAm Polyacrylamide

PAGE Polyacrylamide gel electrophoresis

PDMS Polydimethylsiloxane

p-TsOH Paratoluenesulfonic acid

SDS Sodium dodecyl sulfate

SEM Scanning electron microscope

ssDNA Single stranded deoxyribonucleic acid

TBAB Tetra-n-butylammonium bromide

THF Tetrahydrofuran

UHPLC Ultra high performance liquid chromatography

xviii

ABSTRACT

Njoya, Nadine K. Ph.D., Purdue University, December 2013. Silica Colloidal Crystals as Emerging Materials for the High-Throughput Sizing of Proteins by Packed Capillary Electrophoresis. Major Professor: Mary J. Wirth.

Protein heterogeneity is currently one of the leading problems in formulation of

therapeutic proteins. The effectiveness of therapeutic proteins is diminished when they

exhibit heterogeneity due to aggregates and degradation by-products. Furthermore,

guidelines set forth by the FDA require therapeutic proteins to have high stability as well

as high purity. With development time on the order of years and costs incurred in the

millions, pharmaceutical companies are investing more time in finding fast and efficient

ways to screen for protein heterogeneity. Current methods such as native polyacrylamide

gel electrophoresis (PAGE), size-exclusion chromatography (SEC) and capillary

electrophoresis (CE) are time consuming, require large amount of samples, and are not

always easily implementable with instrumentation. Our approach to these issues involves

the use of silica colloidal crystals bearing brush layers of polyacrylamide. Using

established correlations between molecular weight and reduced mobility, selectivity is

compared in silica colloidal crystals and gels. Due to low broadening effects in colloidal

crystals, fast sieving separations of proteins have been observed in capillaries over

distances as short as 6.5 mm in 72 seconds with exceptionally low plate heights. When

adapted to a 96 well plate format, throughput can be increased a 100 fold. Our work

xix

explores the feasibility of high throughput characterization of protein aggregates and

degradation products using a clinical antibody donated by the Eli Lilly Corporation.

Chapter 4 and 5 cover a project related to the investigation of the structural

backbone of asphaltenes, which are a cause of serious concern to the petroleum industry.

They have many undesirable effects such as precipitating in the pipelines and some of

their metal content will foul catalysts used in oil processing. There are several endeavors

dedicated to elucidating their structures. The one our group undertook involved

synthesizing smaller asphaltene compounds that retained some of the key structural

features associated with real asphaltenes. Fragmentation studies of these model

compounds using tandem mass spectrometry revealed an unexpected fragmentation

pattern which is meant to be used in elucidating the structure of real asphaltenes

compounds.

In Chapter 6, work toward the synthesis of the photoaffinity labeled analog of

Callipeltin A is presented. Capilleltin A is a natural product with desirable anti-HIV

activity. Unfortunately, not much is known about its mode of action with its therapeutic

targets. In order to learn more, we decided to replace the hydroxyl group of one of the

non-proteogenic amino acid residues by a highly reactive group such as azide. From past

studies, such a replacement is not expected to affect the bioactivity of Callipeltin A. A β-

MeOTyr azide analog was synthesized and fully characterized as part of the progress

made.

1

PART ONE

SILICA COLLOIDAL CRYSTALS AS EMERGING MATERIALS FOR THE HIGH-THROUGHPUT SIZING OF MONOCLONAL ANTIBODIES BY PACKED

CAPILLARY ELECTROPHORESIS

2

CHAPTER 1: INTRODUCTION

1.1 Importance of Homogeneity in Protein Drug Formulations

Over the years, the pharmaceutical industry has seen an increasing shift towards

biopharmaceuticals and monoclonal antibody (mAb) drugs. Such rapid growth has been

fueled by advances in biotechnology and an exceptional selectivity towards their intended

therapeutic targets as compared to small molecule drugs (8-10). It therefore comes as no

surprise that the global market for biologic drugs is projected to reach $239 billion by

2015 with therapeutic monoclonal antibodies accounting for 36% of sales (11).

Heterogeneity in monoclonal antibody drug formulations can stem from several

chemical and physical factors imparted during bioengineering, transport and storage

processes. Such factors combined with the intricate structural profile of those bio entities

can lead to degradation events such as photodegradation, disulfide scrambling, oxidation,

dissociation, deamination, precipitation and aggregation. Of those various highly

deleterious processes, aggregation stands out due to the fact that it can occur

spontaneously in the formulated drug and endanger patient safety through a wide range of

immunogenic responses (12). Therefore, in order for these therapeutics to perform at their

optimum efficacy, be devoid of potentially dangerous side effects and meet the FDA drug

specifications, it is imperative that their formulation remain aggregate free (5).

3

Protein aggregation is currently the research focus of many biopharmaceutical and

academic laboratories. They are faced with developing high throughput methods for

aggregate detection and quantification. The urgency for such methods is heightened by

the fact that there are several challenges associated with identifying a matrix that

minimizes aggregate formation at various stages of the production pipeline. Such

obstacles arise from the complex physiochemical backbone (primary, secondary, tertiary

or quaternary structure) of biological entities as opposed to small molecules and their

sensitivity to many bioengineering mechanisms and factors such as (5, 12-16):

Freeze and thaw cycles: used during production to ensure proteins retain

their native conformation and biological activity

Shaking/stirring: encountered during drug processing and transport

Purification: needed for the formulation to remain contaminant free

Solution conditions: concentration, pH, ionic strength, temperature

Exposure to air-air, air-liquid, liquid-solid interfaces

Contact with metal containers, chromatography columns, pipes, etc.

Assessing the homogeneity of protein formulations is therefore of the utmost

importance in order to avoid diminished bioactivity and off-target occurrences.

1.2 Current Analytical Techniques for Aggregate Characterization

There are several methods currently used to establish the aggregate profile of

proteins. To stay within the realm of this project, only the main size based separation

techniques will be discussed (Table 1.1).

4

Table 1.1 Common Size Based Analytical Techniques for Detection of Protein Aggregates

Technique Theory Advantages Drawbacks

SDS-PAGE (17-20)

Polymeric gel is used to separate analytes by size in the presence of an electric field

Cheap Easy to perform Compatible with most bio-analytes

Very long running times Low reproducibility Broadening effects Staining required

SDS-CGE (21, 22)

Size based separation of analytes in a capillary filled with a polymeric sieving material in the presence of an electric field.

High resolution High sensitivity Fast separation No staining required Coupling capability with other separation methods

Long running times Low reproducibility Broadening effects

SEC (22)

A porous medium is used to selectively exclude analytes from its pores

High reproducibility Good resolution Non trapping Coupling capability with other separation methods

Long running times Low capacity (not amendable for large scale processes) Particle size range limitation

1.2.1 Sodium Dodecyl Sulfate - Polyacrylamide Gel

Polyacrylamide gel is the medium of choice for size dependent protein

fractionations by electrophoresis and is made by crosslinking acrylamide with N,N’-

methylenebisacrylamide (bis-acrylamide). The pore gradient is controlled by the total gel

concentration as well as the amount of cross-linker used (23) . SDS, which is a surfactant,

binds most proteins at a constant ratio (1.4 g of SDS per 1 g of protein). The SDS binding

masks the intrinsic charge of the protein resulting in an even distribution of net negative

charge per unit mass (Figure 1.1 A). In the presence of an electric field, this separation

technique is based entirely on the differential motion of analytes through the gel matrix

i.e. the migration distance is solely relative to the approximate analyte size as shown in

Figure 1.1 B (16, 22)

se

h

w

ar

fl

Figure 1.1 Aseparati

As ca

eparation of

ave a higher

weights coun

re made visi

luorescent dy

A. SDS modion of three

an be observ

f proteins thr

r migration

nterparts (24

ible to the h

yes (28, 29).

de of action aproteins thro

po

ved in Figur

rough a siev

velocity and

). Once the

human eye u

.

and B. Migraough the porolyacrylamid

e 1.1, the po

ving mechan

d traverse m

electrophore

using silver

ation illustrarous networkde gel.

ores within

nism. The low

more volume

esis is comp

(25, 26), Co

ation for the k of a vertica

the gel mat

w molecular

e than their

pleted, the s

oomassie Bl

size dependal slab SDS

trix allow fo

r weight ana

higher mole

eparated pro

lue stains (2

5

dent

or the

alytes

ecular

oteins

27) or

6

SDS-PAGE is a very well established separation method which has been used for

over half a century to successfully identify thousands of proteins (30). However, there are

several drawbacks associated with this technique. The gel making process is tedious and

time consuming; depending on the complexity of the sample the running times can be

quite long (anywhere from 45 minutes to 3 hours) and even though smaller proteins are

well resolved, the uneven pore distribution of the gel adds to the broadening factors that

plague larger proteins separations (31). This technique is therefore inadequate to separate

antibody monomers from their aggregates.

1.2.2 Sodium Dodecyl Sulfate - Capillary Electrophoresis

Capillary and slab gel electrophoresis have the same working principle: analyte

separation based on mobility differences in a given electric field with the gel serving as a

sieving matrix (Figure 1.2) (21). However, an increased throughput is achieved with

capillary gel electrophoresis due to the use of narrow bore fused silica capillaries. These

capillaries allow for higher voltages resulting in faster, high resolution separations and

reduced band-broadening. Joule heating is minimized by the rapid dissipation of heat in

fused silica capillaries as their walls can be made very thin (32). Most importantly, this

method has the advantages of on-column detection, full automation, precise protein

quantification and molecular weight assessment, which make protein analysis much

simpler (33, 34). However, the preparation of the gel inside the capillaries would often

lead to poor material robustness shortening the lifetime of the column (34). This led to

the use of non-rigid replaceable linear polymers (in situ prepared) as sieving matrices for

the size-based separation of proteins and sequencing of ssDNA and dsDNA (35, 36).

H

p

ti

ag

F

However, de

olymerizatio

imes and d

ggregates (3

Figure 1.2 Sthrough the

spite strong

on control,

istances do

7).

chematic forsieving matAnalyte is d

g points of

possible po

not make

r capillary elterial under adetected whe

this techniq

olymer conta

it ideal fo

lectrophoresan electric fi

en it passes th

que such as

amination a

or separating

sis which shoield producehrough a det

more robu

as well as l

g larger pr

ows chargeded when volttection wind

ustness and

longer separ

roteins and

d analyte motage is suppldow.

7

more

ration

their

ving ied.

8

1.2.3 Size Exclusion Chromatography

Size exclusion chromatography (SEC) is a separation method that predominantly

uses silica-based polymeric beads to discriminate between analytes based on their size.

The larger species are excluded from the intrapore of the sieving material and only travel

through the interstitial volume while the smaller proteins can go through the intrapores

and traverse more volume. As a result, the larger proteins have a shorter elution time

compared to the smaller ones as illustrated in Figure 1.3 A. (38). Although SEC shows

high efficiency (22), issues such as protein adsorption are prevalent due the high

dependence of the separation mechanism on strong interactions between analyte and

stationary phase. This may impose a size cut-off for the aggregates that can actually be

detected by SEC (39). Figure 1.3 B. shows the size exclusion chromatogram for an aged

monoclonal antibody with poor resolution between the aggregates. Grimm et al. showed

that a better resolution is achieved if the sample elutes longer and were able to baseline

resolve an IgG monomer from its dimer and higher molecular weight aggregates in about

30 minutes (40). This shows that long separation times are required for the efficient

separation of antibodies and their aggregates.

Fm

µ

sh

re

w

g

co

co

re

Figure 1.3 Amonoclonal

µm, 300 Å, 4

Of the

hows the mo

esulting in s

would allow

el are plagu

ompensate f

ould use a m

eplacement f

A. SEC mechantibody (74.6x250 mm.

e techniques

ost room for

smaller sepa

for simultan

ed by slow v

for the broa

more compa

for gel.

hanism of act4oC, 10 minu Mobile pharate: 0.4 mL

s discussed

r throughput

aration distan

neous separa

velocities of

adening effec

act, homoge

tion and B. Sutes) using aase: 150 mML/min. Temp

above, it ap

t improveme

nces, capilla

ations. Howe

f the protein

cts. Therefo

eneous mate

Size exclusioa Sepax Tech

M sodium phoperature: 25o

ppears that

ent: if separa

aries can be

ever, as men

n zones whic

ore, to circum

erial that exh

on chromatohnologies Zeosphate buffoC.

capillary ge

ation times

multiplexed

ntioned abov

ch require lo

mvent these

hibits long

ogram of an enix SEC-30fer, pH: 7.0.

el electropho

can be short

d into arrays

ve, separatio

onger distanc

e drawbacks

range order

9

aged 00, 3 Flow

oresis

tened

s that

ons in

ces to

s, one

r as a

10

1.3 Gel Electrophoresis: Band Broadening and Resolution

In gel electrophoresis, there are two major causative forces of band broadening

processes: longitudinal diffusion and Joule heating.

Longitudinal diffusion occurs when a protein zone spreads away from its center

along the migration direction (41). This phenomenon is represented by the peak

dispersion, σ2, which proportionally relates the analyte diffusion constant, D, in the

separation medium over a period of time, t.

2 (Equation 1.1)

Diffusion happens during staining and visualization which contributes to

broadening and increased analysis time (42). Equation 1.1 shows that the extent of

protein zone spreading will be reduced if the time during which this phenomenon takes

place is shortened. This can happen one of two ways:

1) In an electrophoresis gel, the degree of cross-linking is not perfectly uniform

on the microscopic scale, and the resultant velocity distribution creates a

broader zone. If the sieving medium is homogenous and has an ordered pore

distribution, the path traversed by the analyte will be even all throughout the

material as well as the migration velocities reducing band broadening (Figure

1.4)

11

Figure 1.4 Depiction of the differences in motion paths represented by arrows for two protein zones in A. Homogeneous medium and B. Heterogeneous medium. The differing red and green paths are caused by micro-heterogeneity in the gel pore network and cause

broadening.

2) Higher voltages will decrease the separation time and consequently the peak

dispersion of the protein as shown in equations 1.2 and 1.3 where L is the

separation distance, V is the voltage, v is the zone velocity and µapp is the

apparent electrophoretic mobility.

(Equation 1.2)

Homogeneous

Medium, (small σ)

Even velocity distribution

Heterogeneous

Medium, (large σ)

Random velocity distribution

A.

B.

12

(Equation 1.3)

However, applying higher voltages will also lead to Joule heating (43). This stems

from Ohm’s law which states that the potential difference, V, in a medium is proportional

to its resistance, R, and current, I, flowing through the material (Equation 1.4) and its

relation to heat generated over time, t (Equation 1.5):

(Equation 1.4)

Heat = VIt (Equation 1.5)

Therefore, increasing the voltage applied is a trade-off between increasing the

separation speed and minimizing Joule heating. Joule heating lowers temperature

dependent physical factors such as dynamic viscosity, which consequently increases the

analyte’s diffusion constant. As mentioned in section 1.2.2, thin fused silica capillaries

are used to minimize Joule heating by improving the heat dissipation process. However,

how does using smaller i.d.’s capillaries affect the separation resolution?

Resolution, Rs, is proportional to the selectivity, Δv/v, where H is the plate height

and will be discussed in the next section (Equation 1.6). This equation applies to any type

of separation, showing that both zone sharpness, described by the first term, and

selectivity combine to give resolution, and the two effects are multiplicative. The

limitation of capillary gel electrophoresis is that the selectivity is much lower because the

pores are large. The lower selectivity has to be compensated by longer lengths to give the

same resolution. The net effect is an increase by only about a factor of two in overall

speed for capillary gel electrophoresis of proteins (34, 44, 45). Selectivity needs to be

combined with a thin, homogeneous medium for faster resolution.

13

/

∙ ∆ (Equation 1. 6)

Low resolution and band broadening effects such as longitudinal diffusion, micro

heterogeneity in the pore network and a need for improved resolution call for the use of a

uniformly sized, compact and highly order material in place of gel.

1.4 Research Goal

The goal of this project is to replace gel inside capillaries with a medium more

suitable for the high throughput separation of larger proteins. The idea is to reduce the

broadening processes in the separation medium of capillary electrophoresis by decreasing

the separation length and consequently increasing the speed of separation.

The principle behind this idea uses the plate height, H, which is the amount of

zone variance, σ2 over the separation distance, L.

(Equation1.7)

Conceptually, the physical separation arising from mobility differences would

occur much earlier if the medium contributes negligible broadening to the protein zones.

This concept is illustrated in Figure 1.5: two zones separate over a tenfold shorter length

when the plate height is tenfold smaller.

F

uo

ad

v

se

d

Figure 1.5 Siof plate heigunresolved i

of heterogene

The d

dvantages fo

elocity is no

eparation de

ispensing.

This p

1) Si

ca

imulated ploght, H, on rein 1 cm of seeous PAAm

μm, the

design of a

or throughpu

ot changed

evice, which

project was c

ilica colloid

apillaries we

ots of Gaussiesolution. A.eparation lenmedium is a

e Gaussians

a material w

ut, one being

by the mat

h would fac

carried out in

crystals whi

re thermally

ian functions For a plate

ngth, but theyadapted fromare baseline

with a smal

g a tenfold sh

erial, and th

cilitate the

n three parts

ich were use

y and chemic

s vs. separatheight of 10y are baselin

m reference ( resolved in

ller plate h

horter separ

he other bei

use of mic

s:

ed as a replac

cally stabiliz

tion length to0 μm, the twone resolved i(1). B. For aonly 1 cm.

height would

ration time, a

ing a tenfol

croplate rob

cement for g

zed. A dense

o illustrate eo Gaussiansin 10 cm (Im

a plate height

d thus give

assuming tha

ld more com

otics for sa

gel in fused

e polyacryla

14

ffect are

mage t of 1

e two

at the

mpact

ample

silica

amide

15

brush layer was grown unto the surface of the non-porous silica beads to

minimize protein absorptivity and enhance biocompatibility with the

bioentities.

2) Thermal aggregation of a therapeutic monoclonal antibody provided by the Eli

Lilly Corporation followed by electrophoretic separation from its dimer based

on previous work published by Birdsall and Wirth (4).

3) Voltage, separation length and diffusion coefficient studies were also

conducted to take further improve the separation efficiency of silica colloidal

crystals.

16

CHAPTER 2: THERMAL TREATMENT, PACKING AND CHEMICAL MODIFICATION OF SILICA COLLOIDAL CRYSTALS FOR CAPILLARY

ELECTROPHORESIS

2.1 Background on Silica Colloidal Crystals

The initial interest in silica colloidal crystals (SCCs) was spurred by their

photonic properties resulting from the periodic change in their refractive index (46-48).

Indeed, colloidal crystals exhibit minimal polydispersity and can self-assemble into

highly ordered 3D arrays of colloidal particles (49). In nature, such highly ordered

structures are known as gem opals and are characterized by fascinating colors due to their

opalescence (Figure 2.1 A.) (50). Colloid crystals photonic properties are currently used

in sensors (51), electro-optic devices (52), energy conversion and storage products (53).

Furthermore, they have been used as patterned supports for microarrays applied to

biomedical processes (54). However, the most relevant application of SCCs to this

project is their use as media for biomolecules separations (55).

Monodisperse SCCs self-assembled layers exhibit a face-centered cubic (FCC)

conformation with a maximum void volume fraction of .26 (Figure 2.1 B.) (56). Our

research group uses this characteristic property to efficiently pack these particles in

stainless steel, fused silica capillary columns and planar channels to effect high resolution

and high-throughput protein separations (3, 4, 57-59). However, before use, some thermal

and chemical treatment is required to make the SCCs compatible with proteins.

Figure 2.1 Hopal exhib

micro

High order pbiting bright

oscope of SC

acking of cot colors due t

CCs in a plan

olloid crystalto Bragg dif

nar channel.

ls shown in Affraction (2)(Courtesy o

A. Naturallyand B. Scanf Dr. Robert

y occurring gnning electrot Birdsall)

17

gem on

18

2.2 Chemicals and Materials

Uniform nonporous silica particles were obtained from Fiber Optic Center, Inc.

(New Bedford, MA). Fused silica capillaries were purchased from Polymicro

Technologies (Phoenix, Arizona). The trichorosilanes,

(chloromethyl)phenylethyltricholorosilane (BC) and methyltrichlorosilane (C1) were

obtained from Gelest, Inc. (Morrisville, PA). CuCl (99.9%) was purchased from Alfa

Aesar (Ward Hill, MA), CuCl2 was obtained from Acros Organics (Morris Plains, NJ)

and acrylamide, from Sigma Aldrich (St. Louis, MO). Toluene and N,N-

dimethylformamide (DMF) were purchased from Avantor Performance Materials

(Center Valley, PA) and Tris[2-(dimethylamino)ethyl] amine (Me6TREN) was

synthesized according to literature (60).

2.3 Thermal Treatment of the SCCs and Rehydroxylation

Prior to their use, SCCs must be heated at high temperatures. This process which

is known as calcination, shrinks their volume and avoids cracks in their crystalline form

(61). As a result, the particles have a higher density and an increased stability. For this

project, we proceeded by calcining the SCCs three times in a 600 oC in furnace oven.

Each of the calcining steps was followed by suspension of the particles in ethanol and

sonication until all visible aggregates were broken up. The sonication step was followed

by centrifugation, to separate the solvent from the particles which were then allowed to

dry overnight in a 60 oC vacuum oven. After calcination, the particles were annealed at

1050 oC in a furnace oven for three hours. The purpose of the annealing was to decrease

the isolated silanols concentration after rehydroxylation by smoothing the particles

19

surface beforehand (62). After cooling down, the particles were suspended in ethanol,

sonicated and passed through 2 µm metallic mesh filters to get rid of any

aggregates/impurities. After centrifugation, the particles were allowed to dry overnight in

a 60 oC vacuum oven. Rehydroxylation involved suspending the annealed particles in a

50/50 mixture of ultrapure water (18 MΩ) and nitric acid (HNO3), followed by sonication

and reflux at 250 oC overnight. After reflux, the particles were distributed into centrifuge

tubes and washed with ultrapure water until the pH of the supernatant was about 6. The

final wash was done with ethanol and the particles were dried in a 60 oC vacuum oven.

Slurries for packing into fused silica capillaries were prepared at varying w/w

concentrations in 1 mL of ultrapure water and sonicated as needed for all particles

conglomerates removal.

2.4 Fused Silica Capillaries Preparation and Pressure Packing

A 75 and 100 µm i.d. fused silica capillary of about 180 cm in length was

conditioned with 0.1 M NaOH, ultrapure water and ethanol successively for 20 minutes

each time using a syringe pump (Harvard apparatus, Holliston, MA). The capillary was

then cut into 12 cm segments which were allowed to dry in a 60 oC vacuum oven for 30

minutes. A 35 % w/w slurry of SCCs in ultrapure water was wicked into the capillaries

which were packed against a frit under 7000 psi delivered by a packing pump (Series

1500 Scientific Systems Inc, State College, PA) with sonication for about 20 minutes.

The capillaries were then allowed to dry under a vacuum desiccator until the plug formed

went from translucent to bright white. An illustration of the fused silica capillary packing

is

th

F

co

u

T

s shown in F

he packed be

Figure 2.2 Illslur

Plastic

onnected to

sing a syring

The capillarie

Figure 2.2. It

ed.

lustration of rry. The son

2.5 In sit

c syringes f

the packed

ge pump (H

es were then

t is importan

the pressureication is ne

tu Chemical

filled with

capillaries a

Harvard appa

n rinsed with

nt to note tha

e packing of eeded for a u

Treatment o

a 20:1 (v/v

and the solu

aratus, Hollis

h dry toluen

at slow depr

a fused silicuniform and

of the SCCs

v) of BC an

ution was pu

ston, MA) fi

ne for 20 mi

ressurization

ca capillary fdenser pack

Packed Bed

nd C1 in d

ushed throug

itted with a

nutes using

n avoids crac

filled with a ed bed.

d

dry toluene

gh the silica

cursor overn

a pressure p

20

cks in

SCC

were

a plug

night.

pump

21

(Series 1500 Scientific Systems Inc, State College, PA) to get rid of the unreacted silanes

and dried for two hours in a 120 oC vacuum oven for the condensation of any unreacted

silanols. The surface stabilization of the SCCs with the trichlorosilanes was followed by

the surface initiated atom transfer radical polymerization (ATRP) (7, 63, 64) which was

used to grow the acrylamide layer on the SCCs surface to minimize protein adsorption.

Care must be applied because this reaction is highly sensitive to oxygen therefore, the

entire process was carried under inert conditions. A 3 M solution of acrylamide in DMF

was degassed for 30 to 45 minutes and added to 74.25 mg of CuCl and 10.05 mg of

CuCl2 under argon (Ar). After dissolution of the copper salts, 100 µL of Me6TREN was

added and about 2 mL of the reaction mixture was transferred to the capillaries in a

Schlenk flask under Ar. After the solution had wicked into the capillaries and all the

packed beds were wet, the capillaries were transferred to the syringe pump (Harvard

apparatus, Holliston, MA) and more solution was pushed through the silica plugs

overnight. The capillaries were rinsed successively with a 0.1 M solution of EDTA and

ultrapure water for 20 minutes each time using the syringe pump (Series 1500 Scientific

Systems Inc, State College, PA) and dried in a vacuum desiccator overnight. The dried

capillaries were then stored in a 4 oC fridge until electrophoresis. Scheme 2.1 shows the

silane stabilization and polymerization.

m

d

b

cu

o

Scheme 2.1

Right

microscope (

efects as th

eds. Cracks

utting off th

f solvent and

1 Silane mod

after press

(Nikon SMZ

e solvent ev

in the pack

e flow rate a

d cut off the

dification and

sure packing

Z1500, Lou

vaporates. F

ked bed mo

after the plug

sonication s

d atom trans

g, the pack

uisville, KY

Figure 2.3 sh

ostly occur

g is formed,

so as to not d

sfer radical p

ked bed can

Y) to observ

hows examp

from rapid

it is advised

disturb the p

polymerizati

n be imaged

ve potential

ples of unifo

depressuriz

d to slowly d

packed bed.

on (ATRP) (

d under a z

cracks or

form and cra

zation. Instea

decrease the

22

(7).

zoom

other

acked

ad of

e flow

FT

ar

p

p

p

9

ob

Figure 2.3 ZThe black po

and C sho

Figure

re pressure

acked bed c

article size.

acked in a 7

50 nm in va

bserved at th

oom microscortion is dry ow cracks an

e 2.4 A. sho

packed in a

changes the B

470 nm par

5 i.d capillar

arying i.d. ca

he capillary w

cope imageswhile the br

nd gaps wherd

ows high ord

a 75 i.d. ca

Bragg angle

rticles also s

ry as is show

apillaries is a

walls (Figur

s of A. Densrown one is wre drying hadepressurizat

der packing

apillary. Wat

e of light res

show localiz

wn in Figure

a little bit mo

re 2.4 C. and

ely packed pwet. B (Cou

as occurred aation.

that is exhib

ter inside th

sulting in a b

zed regions o

e 2.4 B. How

ore random

d D.)

plug with nourtesy of Dr. and which re

bited when

he interstitia

blue hue tha

of high orde

wever, the pa

but some lo

o obvious cra

Brooke Kosesult from fas

250 nm par

al volume o

at varies wit

er when pre

acking of 700

ong range ord

23

acks. shel) st

rticles

of the

th the

essure

0 and

der is

nt

c

Figure 2.4 nonporous sithe free frac

µm i.d. cacrystalline orsection of a the cross se

particles p

A. Zoom milica; the blutional volum

apillary packrder (Image 275 µm i.d. cction of a 10

pressure pack

microscope imue color is atme in the packed with 4702.4 B reprodcapillary pre00 µm i.d. caking is mostl

ob

mage of a 75tributed to th

cked bed. B. 0 nm SCCs wduced from rssure packed

apillary presly random, hserved at the

µm i.d. caphe Bragg difSEM image

which showsreference (3)d with 700 nsure packed

however some wall.

pillary packeffraction of we of the crosss localized re)) C. SEM imnm SCCs. Dd with 950 nmme long rang

d with 250 nwater whichs section of egions of higmage of the c. SEM imagm SCCs. Large order can b

24

nm h fills a 75 gh cross e of rger be

25

The high crystalline order exhibited by SCCs pressure-packed in small i.d. fused

silica capillaries is used in this research to model size-based protein separations within

the packed bed. Figure 2.1 B. shows that pores for sieving exist everywhere where

particles meet. Flow versus pressure as well as protein diffusion in packed and open

capillary experiments were carried out in our lab to determine the pore radius formed by

the SCCs and were shown to be suitable for the size-exclusion of proteins. Chapter 3 will

discuss the sieving electrophoresis experiment which is facilitated by the use of small

diameter fused silica capillaries that minimize Joule heating and consequently zone

broadening (65). The sieving medium has a high thermal conductivity and the

polyacrylamide coating on the surface of the SCCs not only decreases the protein

adsorption but also suppresses the EOF by masking the charges at the surface (21).

26

CHAPTER 3: HIGH-THROUGHPUT PROTEIN SIZING USING NON POROUS SILICA COLLOIDAL CRYSTALS BEARING A BRUSH LAYER OF

POLYACRYLAMIDE

3.1 Recent Developments in Electrophoresis of Proteins Using SCCs as the Sieving Medium

The replacement of gel with a compact material that reduces broadening effects

due to random pore distribution and provides faster Joule dissipation was discussed in

Chapter 1. A material with such characteristics packed in small i.d. fused silica capillaries

could significantly increase the throughput of protein sizing by electrophoresis. Our

group has investigated the ultra-fast separations of peptides and dyes by electrophoresis

using SCCs vertically deposited on a fused silica slide (66). However, the use of silica

colloidal crystals as sieves for biomolecules separation using PDMS microchips was first

described in 2007 by Zeng and Harrison (55). Subsequently, they effected the

fractionation of 2-50 kbp DNA inside crystalline nanoarrays over a few millimeters (67).

Densely packed nanoarrays generated from monodisperse colloidal crystals could be

implemented into automated devices for high-throughput protein sizing. Such devices

would be more efficient if the sieving mechanism of the biomolecules inside the colloidal

self-assembly (CSA) could be modeled as a function of non-fitting parameters. This task

was undertaken by Birdsall and Wirth (4) who derived and experimentally validated a

model based on the Ogston model theory for sieving in colloidal crystal (68, 69). The

27

purpose of this work is to model an electrophoretic separation of a monoclonal antibody

and its dimer using data and equations from the literature, and to carry out such a

separation using a silica colloidal crystal for which the particle diameter is selected based

on the modeling. Sodium dodecyl sulfate (SDS) denaturing is used to impart a charge on

the monoclonal antibodies, which is necessary because these proteins typically have pI

values near neutral pH. Covalent aggregation is ubiquitous for monoclonal antibodies due

to scrambling of disulfide bonds. The technology would be applicable to characterizing

covalent aggregates since SDS breaks up noncovalent aggregates. The plate height will

be measured to assess speed and miniaturization.

3.2 Experimental

3.2.1 Chemicals and Materials

Uniform nonporous silica particles were obtained from Fiber Optic Center, Inc.

(New Bedford, MA). Fused silica capillaries were purchased from Polymicro

Technologies (Phoenix, Arizona). The trichorosilanes,

(chloromethyl)phenylethyltricholorosilane (BC) and methyltrichlorosilane (C1) were

obtained from Gelest, Inc. (Morrisville, PA). CuCl (99.9%) was purchased from Alfa

Aesar (Ward Hill, MA), CuCl2 was obtained from Acros Organics (Morris Plains, NJ)

and acrylamide, from Sigma Aldrich (St. Louis, MO). Toluene and N,N-

dimethylformamide (DMF) were purchased from Avantor Performance Materials

(Center Valley, PA) and tris[2-(dimethylamino)ethyl] amine (Me6TREN) was

synthesized according to literature (60). 10X Tris-glycine-SDS buffer was purchased

28

from Sigma Aldrich (St. Louis, MO); ethylenediaminetetraacetic acid, Avantor

Performance Materials (Center Valley, PA); sucrose and

tris(hydroxymethyl)aminomethane, Mallinckrodt Baker, Inc. (Phillipsburg, NJ); SDS,

Sigma Aldrich (St. Louis, MO); SYLGARD® silicone elastomer curing agent and base,

Dow Corning Corp. (Midland, MI); sodium phosphate monobasic and dibasic, Avantor

Performance Materials (Center Valley, PA). A sample of pharmaceutical-grade

monoclonal antibody, obtained from Eli Lilly (Indianapolis, IN), had gone through drug

development but is no longer an active candidate. The antibody was labeled with

AlexaFluor 546 (Invitrogen, Carlsbad, CA) according to the vendor’s instructions.

Lysozyme from chicken egg white, carbonic anhydrase from bovine erythrocytes,

albumin from chicken egg white, human insulin solution, trypsin inhibitor from glycine

max (soybean) and albumin from bovine serum were obtained from Sigma Aldrich (St.

Louis, MO), Mouse anti IgG Fc antibody (anti-IgG2) was purchased from Fitzgerald

Industries (Acton, MA).

3.2.2 Preparation of the Silica Particles for Electrophoresis

750 and 1000 nm silica particles were calcined, annealed, rehydroxylated,

chemically modified and packed into 75 and 100 i.d. Teflon® coated fused silica

capillaries respectively as described in Chapter 2. After annealing, SEM imaging

revealed the particle diameters to be 700 ± 10 nm and 850 ± 20 nm respectively.

29

3.2.3 PDMS Reservoirs Preparation

A 40 g mixture of a 9:1 ratio of SYLGARD® silicone elastomer curing agent and

base was gently stirred, poured onto a Petri dish with a silicon wafer bottom which was

then put in a vacuum dessicator for degassing. Once the mixture was bubble free, the

Petri dish was removed from vacuum and a lab made mold with the reservoir shapes was

strategically set on top of the gel. The polymerization was allowed to occur overnight

undisturbed. The reservoirs where then cut out into the desired shape and used for the

electrophoresis experiment.

3.2.4 Sample Preparation for Electrophoresis

A sample of pharmaceutical-grade monoclonal antibody, obtained from Eli Lilly

(Indianapolis, IN), was labeled with AlexaFluor 546 (Invitrogen, Carlsbad, CA)

according to the vendor’s instructions. The concentration after labeling was 1mg/mL. For

the electrophoresis experiment, a 50 µL, 0.05 mg/mL aliquot was prepared by diluting

the labeled IgG with a stock loading buffer made with 62 mM Tris (base), 1 mM EDTA,

3% sucrose and 2% SDS at pH 8.00. Aggregation was effected by heating the sample at

74 oC for 10 minutes and confirmed using a commercial SEC column from Sepax

(Newark, DE) with the following characteristics: Zenix SEC-300 (3 µm, 300 Å, 4.6x250

mm). The vendor’s recommended mobile phase, sample concentration and flow rate were

used.

30

3.2.5 Capillary Sieving Electrophoresis Apparatus

The set-up for the electrophoresis experiment is shown in Figure 3.1. A fused

silica capillary packed with PAAm coated silica particles was conditioned overnight with

a commercial electrophoresis running buffer made with 25 mM tris, 192 mM glycine and

0.1 % w/v SDS at pH 8.6 from Sigma Aldrich (St. Louis, MO). The capillary was cut to

about 2 cm and set between PDMS reservoirs filled with the running buffer. The current

required for the analyte migration was provided by platinum electrodes connected to a

voltage supply (10A25A Series, Ultravolt, Ronkonkoma, NY) controlled by the Labview

software. The in-column fluorescence detection was possible using a an inverted Nikon

Eclipse TE2000-U microscope fitted with a 2X objective (CFI Plan Apo 2X from Nikon

Instruments Inc., Melville, NY) and connected to a ProEM 512 CCD camera (Princeton

Instruments, Trenton, NJ). Excitation was provided through a Hg lamp (Nikon

Intensilight C-HGFI, Nikon Instruments Inc., Melville, NY) and emission at the

appropriate wavelength was recorded using filter cubes from Omega Optical

(Brattleboro, VT). During electrophoresis, the data was acquired with the Winview 32

software (Princeton Instruments, Trenton, NJ).

FFigure 3.1 AA. Illustrationn of the sievving electropup.

phoresis set-uup. B. Actuaal laboratory

31

set-

32

3.3 Results and Discussion

3.3.1 Rationale behind the Pore Size Determination for Antibody Aggregates Detection

Electrophoresis results reported recently by Birdsall and Wirth (4) are analyzed

further here to understand how these materials can be used to study aggregation of

monoclonal antibodies. Figure 3.2 A. shows an electropherogram, where a silica colloidal

crystal made of a particle 483 nm in diameter was used for electrophoresis of lysozyme

(14.7 kDa), trypsin inhibitor (20 kDa), carbonic anhydrase (29 kDa), and bovine serum

albumin (67 kDa). The electropherogram shows narrow zones as the proteins

electromigrate past a fixed detection point, which was only 4 mm from the injection

point. The zones broaden as protein size increases. This is typical behavior for

electrophoresis, where plate height increases with protein radius. The plate heights were

not calculated in the original work, but are essential in the context of the present work,

and thus are calculated here: the plate heights increase in the order 0.33, 1.0, 1.6, and 3.3

μm for the smallest to the largest protein. These are all smaller than the typical plate

height of 10 μm for medium-sized proteins in gels (70). The unusually small plate heights

demonstrate that the silica colloidal crystals are promising for miniaturization. Figure 3.2

B. shows a plot of molecular weight vs. normalized migration rate through the medium.

The shape of the curve is similar to such curves for gel electrophoresis, where the

mobility goes to zero when the molecular weight exceeds the pore size, and the velocity

is no longer dependent on molecular weight when the protein radius becomes negligible

in size compared to the pore radius.

(

w

ch

w

an

pr

is

im

ot

Figure 3.2 medium. Pro

serum albcircles), and3.1. Pore ra

The so

was derived

haracterized

The m

which was m

nd porosity

rotein radius

s equivalent

mportant out

ther protein

Electrophoroteins A–D aumin, respec

d a fit to fourdius is indic

olid curve is

d from firs

d (no fitting p

mobility, μ, i

measured by o

, ε, were in

s is presente

to the ratio

tcome is tha

s, i.e., mono

resis data forare lysozymctively. A. Er Gaussians (ated by ξ. B

s a plot of th

st principle

parameters r

in the mediu

open capillar

ndependentl

ed by R. Sinc

o Δv/<v> us

at we can use

oclonal antib

r four proteinme, trypsin inElectrophero(line). B. Plooth panels a

he predicted m

es (4), and

equired).

um is norma

ry (free solu

ly determine

ce velocity i

sed in Equa

e Equation 3

bodies and a

ns using a sinhibitor, carbgram of the ot of electropare adapted f

migration ra

d each par

alized to the

ution) electro

ed from flo

is proportion

ation 1.6 for

3.1 to predic

aggregates,

lica colloidabonic anhydrproteins, wiphoresis datfrom Birdsal

ate using Equ

rameter wa

mobility in

ophoresis. Th

ow rate me

nal to mobili

r expressing

ct the selectiv

and other p

al crystal as trase, and bovith data poina using Equall and Wirth

uation 3.1, w

as independ

(Equation

free solution

he pore radi

easurements.

ity, then Δμ/

g selectivity

vity, Δμ/<μ>

ore radii tha

33

the vine

nts ation (4).

which

dently

n 3.1)

n, μ0,

ius, ξ,

The

/<μ>,

. The

>, for

at are

m

th

el

an

E

m

by

more optima

heoretical tr

lectrophores

In this

nd the other

Equation 3.2

monoclonal a

y Mahler et

Figure 3.3 storage (lan

al for the l

reatment by

sis.

s equation, r

r terms have

is valid, dat

antibody and

al (5).

Gel electropnes 5–9). Th

arger protei

y Kopecka

r is the radiu

e the same m

a from the li

d its aggrega

pherogram ohe size stand

fr

ins. For co

et al. (71)

us of the ge

meanings as

iterature are

ates and frag

f a monoclondards are shorom referenc

omparison o

provides a

l fibers, e.g.

s for Equati

used. Figur

gments by g

nal antibodyown in lane 2ce (5).

of selectivity

an equation

., the polyac

ion 3.1. To

re 3.3 shows

gel electroph

y initially (la2. The figure

y with gels

to describe

(Equation

crylamide ch

demonstrate

a separation

horesis, publ

ane 4) and afe is reproduc

34

s, the

e gel

n 3.2)

hains,

e that

n of a

lished

fter ced

35

The image illustrates the benefit of gel electrophoresis: a wide range of molecular

weights is separated, and many lanes can be run simultaneously. The first lane, which has

the size standards, allows one to identify the antibody, which is the most intense band in

the subsequent lanes. In the subsequent lanes, the aggregates of the antibody are visible at

higher molecular weight, and fragments are visible at lower molecular weights. The data

for the size standards are analyzed here by Equation 3.2, and the results are plotted in

Figure 3.4 A. The data in red are for the gel, and the curve is from Equation 3.2, with r

and ξ adjusted for best fit.

Figure 3.4 A. also uses Equation 3.1 for colloidal crystals to make a comparison

of selectivity, which is the blue curve. The circles show the theoretical positions of the

same size standards for comparison. The mobility was divided by porosity to make the x-

axes coincide. The graph reveals that the selectivity is similar for the gel and the colloidal

crystal, with selectivity being somewhat higher for the gel.

The two equations allow a theoretical comparison between a gel and a colloidal

crystal when each have a pore size better suited for resolving a monoclonal antibody from

its dimer with higher velocity. The resulting plot is in Figure 3.4 B. where the pore sizes

were adjusted to give the same velocity for the dimer. The comparison shows that the gel

has a 10% higher selectivity. This is a small difference, but it seems surprising that the

disordered gel would give higher selectivity. The underlying reason is that the pore size

distribution enables the gel to separate a wider range of proteins, which translates to a

smaller slope in log(MW) vs. mobility. The analysis shows that the colloidal crystal is a

reasonable substitution for the gel from the standpoint of selectivity.

eFigure

electrophoresand fit toTheoretioptimize

molecular w

3.4. Comparsis. A. Data

o Equation 3ical plots fored for resolvweight (MW

rison of selepoints (red c.2 (red line).r gel (red linving an antibW) (150 kDa)

ctivity for gecircles) for th. Plot of Eque) and colloiody from its) and its dim

dotted line

els and collothe size standuation 3.1 (bidal crystal (s dimer. In b

mer MW (300es.

oidal crystalsdards of the

blue line, blu(blue line) fo

both panels, t0 kDa) are in

s in protein gel in Figur

ue circles). Bor pore sizesthe antibodyndicated by t

36

e 3.3 B. s y the

37

Since the goal is high throughput, the velocities must also be considered. In

Figure 3.4, the plots were made with respect to “relative mobility, “which is the

normalized mobility divided by the porosity. The porosity is not known for the gel, but it

is likely close to unity since the polyacrylamide fibers probably take up much less of a

negligible volume fraction of the gel. In fact, the best fit of the data for the size standards

gave a fiber radius of zero. For the colloidal crystal, the porosity is 0.2, and the porosity

is inherently low because the silica particles take up most of the volume. What this means

in practice is that the absolute velocity in the colloidal crystal will be about fivefold lower

than that of a gel. The colloidal crystal thus has two problems to overcome: somewhat

smaller selectivity and a much lower velocity. These limitations can potentially be

overcome by a much smaller plate height.

In choosing the particle diameter for the colloidal crystal, one must consider the

trade-off between speed and selectivity. To elaborate, Equation 3.1 shows that selectivity

is enhanced by having the pore size closer to the protein size, but the equation also shows

that the electrophoretic mobility is slowed as the pore size approaches the protein size.

Equation 3.1 can be used to examine the trade-off between speed and selectivity to select

an optimal pore size. A plot of both selectivity and speed with respect to pore size was

made by use of Equation 3 and is given in Figure 3.5 A. The plot reveals that selectivity

drops sublinearly whereas speed goes up nearly linearly until the pore radius exceeds 30

nm. These considerations would suggest a pore radius of 30 nm to be a good

compromise.

Fs

h

th

Figure 3.5 Asquare of the

Zone

eight increas

hat there is a

A. function ofe broadening

broadening

ses strongly

a pore size

f pore radiusg function of

least-sq

is also a c

with protein

distribution.

s to show thef Equation 3.quare fits of d

consideration

n size for a

. One can m

e trade-off. B.3 for a poredata points.

n in selectin

given pore s

model the de

B. Plot of plae radius of 27

ng pore size

size. Such b

ependence o

ate height vs7 nm. The lin

e since the

behavior indi

of plate heig

38

s. the ne is

plate

icates

ght on

39

protein radius by introducing a stochastic variable, ξ’, to account for random changes in

pore radius, and relating this to changes in μ. Adding ξ’ to ξ in Equation 3 and then

differentiating mobility with respect to ξ’ gives a simple relation when ξ>>R>>ξ’.

/ (Equation 3.3)

Since this variation in mobility with pore size would be proportional to zone

width, then its square would be proportional to plate height. If the pore size distribution is

the origin of the broadening, and the pore size variations are small, then plate height

should be proportional to R2. Such a plot is given in Figure 3.5 B., where the least

squares fit shows good agreement, and the intercept is zero. Extrapolating to higher

molecular weights gives calculated plate heights of 9, 20, and 32 μm for monomer,

dimer, and trimer of a monoclonal antibody. For detecting aggregation of antibodies,

larger pores are needed as is shown in Figure 3.6. For a protein of a given radius R, the

electrophoretic mobility through pores of varying radii ξ or ξ’ is high if R << ξ < ξ’.

However, as R approaches ξ or ξ’, the mobility is considerably slowed down. This led us

to choose 1,000 nm as the particle diameter, which is twice that of the 500 nm particles

that gave the pore radius of 27 nm. However, one cannot assume the new pore radius is

also twice as large because the polymer chains grow longer in larger pores.

co

ch

m

o

fo

h

Figure 3.6 Isieving med

Aggre

onfirmed usi

haracteristic

mobile phase

f aggregatio

or the therap

ad a confirm

Illustration odium A. For

3.3.2 Ag

egation was

ing a comme

s: Zenix SE

e, sample co

n was done

peutic mAb a

med dimer.

of the mobilia protein of

hyd

ggregation of

effected by

ercial SEC c

C-300 (3 µm

ncentration

using a calib

and its dimer

ity dependenf low hydrodydrodynamic

f the mAb an

y heating th

column from

m, 300 Å, 4.6

and flow ra

bration curv

r were match

nce of an anadynamic radiu

radius.

nd Confirma

he sample at

m Sepax (New

6x250 mm).

ate were use

ve as shown

hed to those

alyte on the pus. B. For a

ation by SEC

t 74 oC for

wark, DE) w

. The vendor

d (Figure 3.

in Figure 3.

e of mouse a

pore size of protein of la

C

10 minutes

with the follo

r’s recomme

.7). Confirm

8. Retention

anti IgG Fc w

40

f the arge

s and

owing

ended

mation

n time

which

Figure 33.7 Size excluAlexafluo

usion chromor 546 labele

matograph of ed aggregated

f A. Alexaflud mAb (74.5

uor 546 labe5 oC, 10 min

led mAb. B.n).

41

.

F

it

on

ca

T

ty

m

th

Figure 3.8 Ca

Figure

ts dimer. Th

nly 6.5 mm

alculated pla

These extrem

ype of separa

monoclonal a

he packed c

alibration cu

3.3.3 Capi

e 3.9 A. sho

e brighter m

m and in a tim

ate heights a

mely low pla

ation. Wider

antibodies, b

capillary ele

urve for the Z

illary Sieving

ows an imag

monomer pea

me of only

are 0.15 and

ate heights a

r pore gels h

but the broad

ectrophoresis

Zenix SEC-3

g Electropho

e from the s

ak is separat

72 s. Based

d 0.42 μm fo

are extraord

have been us

dening is sev

s data of F

300 (3 µm, 3

oresis of Agg

separation of

ted from the

d on Gaussia

or the mono

dinary for m

sed to separa

vere (72). F

Figure 3.9 A

300 Å, 4.6x2

gregated mA

f a monoclo

e dimer peak

an fits of th

mer and dim

monoclonal a

ate monomer

Figure 3.9 B

A., where a

250 mm) col

Ab

onal antibody

k over a leng

he two zones

mer, respecti

antibodies in

rs from dime

. shows a pl

a fixed dete

42

umn.

y and

gth of

s, the

ively.

n any

ers of

lot of

ection

43

position of 0.65 cm was used and the intensity was monitored vs. time to generate an

electropherogram. Figure 3.7 B. shows a chromatogram from size-exclusion

chromatography for the same sample. The resolution is similar, slightly higher for

electrophoresis, but the time scale for electrophoresis is fivefold shorter. Both separations

show primarily the monomer and dimer peaks. There is evidence for a large aggregate in

the electrophoresis data. Both show evidence of species with smaller molecular weight,

but more precisely controlled samples will need to be studied as well as a variety of

particle sizes, to learn what additional information is obtainable. The fivefold higher

speed in separating monomer from dimer, combined with the present commercial

capability of running 24 capillaries simultaneously in electrophoresis, promises higher

throughput for assessing aggregation of monoclonal antibodies in the optimization of

formulations.

3.3.4 Particle Size, Migration Distance and Voltage Effects on Separation Resolution

Extremely small plates were obtained when 1000 nm particles were used for the

sieving of the mAb monomer and dimer. These exciting results led us to investigate

improvement in resolution using 750 nm particles. The mAb monomer was separated

from its dimer in about 4 minutes, 2.3 cm away from the capillary front as shown in

Figure 3.10 A. From the Gaussian fits of the two protein zones, extraordinary plate

heights of 0.09 and 0.3 µm were achieved for the monomer and dimer respectively

(Figure 3.10 B.).

fFigure 3.9

fluorescencebuffer thro

separ

9 Separatione image of anough a 100-μration length

n of monoclontibody folloμm-i.d. capilh of 6.5 mm

onal antibodyowed by its dllary packedin 72 s. B. c

y monomer adimer as they

d with silica ccorrespondin

and dimer. Ay electromigcolloidal cryng electrophe

A. False-colograte in SDSystal and reacerogram.

44

or S and ch

f

s

Figure 3.1fluorescence

buffer thrseparation ledata points

0 Separatione image of anough a 75-μ

ength of 2.3 are presente

n of monoclontibody folloμm-i.d. capillcm in about d as dots and

onal antibodowed by its dlary packed w4 minutes.

d the lines rezones.

dy monomer dimer as theywith silica cCorrespondepresent the

and dimer. Ay electromig

colloidal crysding electrop

Gaussian fit

A. False-colgrate in SDSstal and reacherogram wts of the prot

45

lor S and ch

where tein

E

se

th

nm

A dec

Equation 1.7

eparation dis

he plate heig

m SCCs pac

Figure 3.1

crease in pla

. However,

stance is kep

ght. Figure

cked in a 75-

1 Voltage efdim

ate height w

higher volt

pt constant,

3.11 shows

-id fused sili

ffects on themer at a cons

with increasin

tages decrea

it is expecte

s mAb separ

ca capillary

separation ostant detectio

ng separatio

ase the peak

ed that incre

ration of mo

at the same

of monoclonon point of 1

on distance

k variance.

easing the v

onomer and

detection po

nal antibody 1.7 cm.

is expected

Therefore, i

oltage will l

dimer using

oint.

monomer an

46

from

if the

lower

g 750

nd

47

As shown in Figure 3.11, 50, 75 and 100 V are the voltages that provide the best

separation at the fixed distance of 1.7 cm for this particle size as resolution is lost at

higher voltages. Figure 3.12 A. shows that as voltage/ cm increases, the electrophoretic

mobilities of monomer and dimer become equal while Figure 3.12 B. shows loss of

resolution as the separation time decreases with increasing electric field. These

perplexing results are under active investigation in our research group and avenues such

as:

Instability of the mAb dimer which is exacerbated at higher voltage

The fact that the polymer layer does not act like a hard boundary under

higher voltage

A uniformly thick polymer layer is not formed during ATRP or a

combination of both along the plug

Dimer reputation at higher voltages

are considered as possible causative factors for this phenomenon.

Figure 3.12 A. Plot of thhe electric fivers

eld (E) versusus separatio

us electrophon time.

horetic mobil

lity. B. Plot o

48

of E

49

3.3.5 Preliminary Results Using AGET ATRP

The current method for the in situ surface-initiated polymerization of SCCs is

cumbersome. This is caused by the stringent O2 free requirement for a successful

reaction; which can only be met through a thorough degassing of DMF and several argon/

vacuum cycles.

Our group is currently investigating an alternative polymerization known as

AGET ATRP. This method is not affected by O2 because unlike ATRP, CuCl is omitted.

Instead a reducing agent converts CuCl2 to CuCl. Using this technique, 500 nm SCCs are

coated with a brush layer of acrylamide and packed in 100-id fused silica capillaries as

described in Section 2.4. Preliminary separations for Alexafluor 488- labeled albumin

from bovine serum and carbonic anhydrase are shown in Figure 3.13. Further

investigation is ongoing in our laboratory.

Figure 3.1(BSA), Bbearing a

3 ElectropheB. Carbonic a

brush layer volt

erograms of anhydrase (Cof PAAm pa

tage and sepa

f Alexafluor 4CA) and C. aacked in 100aration lengt

488 labeled a mixture of 0-id fused silth (L) are in

A. Bovine aboth using 5lica capillaridicated.

albumin seru500 nm SCCies. Operatin

50

um Cs ng

51

3.4 Concluding Remarks

Silica colloidal crystals are an emerging material for the fast and efficient

separations of monoclonal antibodies and aggregates. The uniformity of the material

allows a much smaller footprint, a higher speed and less zonal broadening contributions

as compared to polyacrylamide gel. With the tiny separation lengths of less than 1 cm,

the media could be patterned for direct compatibility with the robotics of 96-well plates

for high throughput. The results presented here use SDS-denatured protein, which

assesses covalent aggregates, which are arguably more deleterious than noncovalent

aggregates. Further studies are needed to address native electrophoresis for assessing

noncovalent aggregates. The silica colloidal crystal makes it easy to control pore size for

sieving large proteins with minimal broadening. Larger particles could be used, in

principle, to assess even larger aggregates. The remaining issues for reduction to practice

are the design of such devices and label-free detection. Protein ladder separation will

require an optimized pore size and PAAm length for the high resolution separation of a

wide range of proteins. AGET ATRP not only simplifies the polymerization process but

preliminary data using SCCs modified by this method are very encouraging.

52

PART TWO

SYNTHESIS AND FRAGMENTATION STUDIES OF ASPHALTENE MODEL COMPOUNDS CAUSED BY COLLISION-ACTIVATED DISSOCIATION USING A

LINEAR QUADRUPOLE ION TRAP

53

CHAPTER 4: BACKGROUND ON ASPHALTENES, MODEL COMPOUNDS SYNTHESIS AND INSTRUMENTATION

4.1 Asphaltenes Overview

Asphaltenes, complex mixtures of organic compounds, make up the for heaviest,

most polar and highest boiling point fraction of crude oil. The steady exhaustion of

lighter crude reserves as well as the ever climbing cost of oil is shifting the focus of the

refining industry towards renewable energy sources and utilization of the heavier

fractions of petroleum which contain a higher percentage of asphaltenes (up to 20 %)

(73). However, asphaltenes are known to aggregate and precipitate in oil pipelines

making crude oil processing problematic (74). A better understanding of asphaltenes is

necessary to efficiently process heavier crude residuals (75). Prior research suggests

asphaltenes are made up of large polycyclic and aliphatic hydrocarbon networks (40 %

aromatic carbons) with a carbon to hydrogen ratio of 1:1.2, and are rich in heteroatoms

such as nitrogen, oxygen, sulfur, porphyrins as well as metals such as vanadium and

nickel (73, 74, 76-86). Two tentative motifs are currently used to describe the structural

configuration of asphaltenes: the island and archipelago models; the island or continental

type has one aromatic core with alkyl chains branching out and is believed to be the most

predominant model. The archipelago model consists of several fused aromatic and

partially saturated ring systems connected by alkyl chains stemming from the main core

(87-94). Significant efforts have been made to synthesize compounds that mimic the

54

structural and physicochemical properties of actual asphaltenes but these models fail to

encompass all known characteristics of asphaltene (95-97). Recent work compared mass

spectrometric fragmentation patterns of asphaltenes with island and archipelago-type

asphaltene model compounds by laser-induced acoustic desorption (LIAD) and electron

or chemical ionization (EI/ CI) (75, 98) as well as atmospheric pressure chemical

ionization (APCI) (98). Results favor the island model although it is likely that the

samples studied are composed of several isomeric and isobaric molecules (91, 98).

Herein, we present smaller island and archipelago model compounds. These compounds

are representative of the simpler yet similar aromatic and heteroaromatic cores present in

asphaltenes. APCI was performed in a linear quadrupole ion trap (LQIT) mass

spectrometer using carbon disulfide (CS2) as the preferred solvent and ionization reagent

because of the asphaltenes’ higher solubility (98) and extensive fragmentation in CS2.

This process resulted in the generation of molecular ions (M+) which were isolated and

fragmented by collisionally activated dissociation (CAD). The goal of this study is to

uncover some of the structural features present in asphaltenes by analyzing the

fragmentation pattern of simpler aromatic/heteroaromatic compounds.

55

4.2 Synthesis of the Asphaltene Model Compounds

All reagents used in this work were of commercial grade and used without further

purification.

4.2.1 Synthesis of the Alkyl and Phenyl-Substituted Naphthalenes

Compounds 1-3 were synthesized using a protocol available in the literature (99).

Scheme 4.1 shows the route employed.

Scheme 4.1 Synthesis of the alkyl and phenyl-substituted naphthalenes.

56

4.2.2 Synthesis of the Alkyl and Phenyl-Substituted Isoquinolines

The protocols describing the synthesis of compounds 4-9 are published elsewhere

(100-102). Scheme 4.2 shows the routes employed to synthesize those compounds and

Table 4.1 presents the corresponding yields.

Scheme 4.2 Synthesis of the alkyl and phenyl-substituted isoquinolines. For compounds 5-9, a palladium-catalyzed annulation of internal alkynes with an imine intermediate is

used to obtain the isoquinolines.

sm

in

st

(L

The r

maller and

nformation r

table molecu

LQIT) mass

Table 4.

rationale be

less compl

relevant to re

ular ions wh

spectromete

1 Yields Ob

hind the sy

lex model a

eal asphalten

hich were su

er.

btained After

ynthesis of

asphaltene p

nes, these an

ubjected to C

r Isoquinolin

the compou

polycyclic a

nalytes were

CAD using

ne Syntheses

unds above

aromatic co

e ionized by

a linear qua

s

was to an

ores. To o

APCI, leadi

adrupole ion

57

nalyze

obtain

ing to

n trap

58

4.3 Linear Quadrupole Ion Trap (LQIT) Mass Spectrometry

The resolution in ion trap mass spectrometers is achieved by the stability

of the trajectory of ions in a fluctuating electric field. The main advantage of using a

linear quadrupole ion trap mass spectrometer is the increased ion storage capability of the

linear trap compared to the three-dimensional quadrupole ion trap (QIT). This leads to

reduced space charge contributions that would otherwise cause low mass resolution

(103). The highly improved confinement capacity of such instruments coupled with mass

selection events known as tandem mass spectrometry allow for a better understanding of

gas-phase ion chemistry (104). There are several methods that can be used to induce

ionization. However, APCI is the only method that would be discussed as it was the

method of ionization chosen to carry out this project.

4.3.1 Atmospheric Pressure Chemical Ionization (APCI)

APCI is a soft ionization technique which means that at least some of the analyte

is converted into ions with retain their initial structural profile (105). APCI is very similar

to chemical ionization (CI) however, it occurs at atmospheric pressure instead of vacuum.

The working principle of APCI is shown in Figure 4.1. The solvated analyte is introduced

into the ion probe, through a heated ion transfer tube (150-225 oC) at a flow rate that is

kept between 200 µL/min and 2.0 mL/min for a stable spray. The tube is surrounded by a

nebulizer co-axial gas known as auxiliary gas (usually nitrogen). When needed, a

secondary outer gas known as the sheath gas (usually nitrogen) is also applied. The heat

of the transfer tube and the gases will nebulize the sample into a fine mist of droplets.

Once introduced into the vaporizer, the droplets in the mist are flash vaporized at

te

m

w

4

emperatures

molecules are

with the analy

.3.

usually high

e ionized wh

yte molecule

her than 400

hich proceed

es for their i

Figure

0 oC. When

d to ionize th

ionization (1

4.1 APCI il

exposed to

he solvent m

106). This p

llustration.

the Corona

molecules. T

process is de

a needle, nitr

Those will co

etailed in Sch

59

rogen

ollide

heme

60

Scheme 4.3 Ionization process in APCI where CS2 is used as solvent.

As a general characteristic, APCI is a better ionization method for low polar and

low molecular weight analytes.

4.3.2 LQIT Instrumentation

A Finnigan LTQ linear quadrupole ion trap mass spectrometer (Figure 4.2),

equipped with an atmospheric pressure ionization (API) source, was used to complete

APCI mass spectrometry experiments. Working solutions, made by dissolving the

asphaltene model compounds in CS2 at concentrations of ~20 µg/mL, were infused into

the ion source via the incorporated syringe line of the instrument using a 500 µL

Hamilton syringe. Standard APCI parameters were set: vaporizer temperature; 300 oC,

Corona Discharge Ionization: Ionization of nitrogen to primarily yield

Solvent Molecule Ionization: Formation of solvent molecules through a series of collisions with nitrogen gas ions

Analyte Molecule Ionization: Analyte is ionized from collisions with solvent

sh

sw

2

an

is

w

pr

M

heath gas (N

weep gas (N

50 oC, and c

n enhanced

solation win

were set at 2

redetermine

Figure 4.2representati

Manual 9700O

N2) flow, 40

N2) flow, 5 (

capillary vol

analyte sign

ndow and su

20 to 45. T

d by the LTQ

2 Illustrationve operating

055-97013 ROperation Co

0 (arbitrary u

(arbitrary un

ltage, 12 V.

nal. The m

ubjected to

The values a

Q software).

n of the Therg conditions.

Revision B, Courses from t

units); auxil

nits); dischar

The ion opt

molecular ion

CAD at dif

are fractions

.

rmo Finnigan. (Figure ada

Chapter 2, Pathe Thermo

liary gas (N

rge current,

tics voltages

ns were iso

fferent energ

s of the nor

n LQIT masapted from thage 14 and TScientific Tr

N2) flow, 10

5 µA; capil

s were adjust

lated using

gies (collisio

rmalized col

s spectromehe Finnigan ThermoFisheraining Insti

(arbitrary u

lary tempera

ted as neede

a 2 Da (±1

on energies

llision energ

ter with somLTQ Hardw

er Scientific itute).

61

units);

ature,

ed for

1 Da)

used

gy as

me ware

LTQ

se

ex

d

fa

th

on

m

v

F

Ions g

eries of ion

xcited by co

issociation (

The q

ashion at eq

hree sections

ne is 37 mm

mm) through

oltages axial

Figure 4.3 Illu

generated in

optics to rea

ollisions wit

(CAD) impar

quadrupole s

ual distance

s with the fr

m. The cente

h which the

lly (along th

ustration of

the API sou

ach the mass

th helium (H

rts energy to

system cons

e of each oth

ront and bac

er section al

ions are eje

he Z-axis) an

the linear qu

urce are pull

s analyzer w

He). This p

o the ions wh

sists of four

her as shown

ck sections b

lso has 2 ex

ected for det

nd RF voltag

uadrupole io(6).

led into the

where the ion

process know

hich leads to

r rods which

n in Figure

being 12 mm

xit slots in th

tection. Ions

ges radially.

on trap. Figur

heated capi

ns of interest

wn as collis

o their fragm

h are arrang

4.3. The ro

m in length

he X-Z plan

s are trapped

re reproduce

llary and tra

t are isolated

sionally activ

mentation.

ged in a pa

ds are divid

while the c

ne (0.25mm

d by applyin

ed from refer

62

avel a

d and

vated

arallel

ded in

center

x 30

ng dc

rence

63

Ion isolation waveform voltage, resonance excitation RF voltage, and resonance

ejection RF voltages are used to disrupt the stable trajectory of ions inside the trap. They

are applied to the exit rods to eject trapped ions and orient them towards the ion detection

systems.

The ejected ions converge towards the conversion dynode. Upon striking the

dynode, secondary particles are generated. These particles are accelerated into the

electron multiplier where more electrons are generated thereby amplifying the

corresponding ion signal-to-noise ratio.

64

CHAPTER 5: STUDY OF COLLISION-ACTIVATED DISSOCIATION OF IONIZED ASPHALTENES’ MODEL COMPOUNDS USING A LINEAR QUADRUPOLE ION

TRAP MASS SPECTROMETER

5.1 Naphthalene-Based Model Compounds

The fragmentation of the two regiomeric asphaltene model compounds was

similar although in both cases CAD of the molecular ion favored loss of even-electron

neutral molecules (alkenes, Figure 5.1). This was initially considered a surprising

departure from the conventional fragmentation pathway of alkyl-substituted aromatics

which is known to go through a cleavage at the benzylic position resulting in even -

electron (odd mass) fragments. However, benzene rings substituted by larger alkyl groups

(number of carbon atoms ≥ 3) can undergo a γ-H shift similar to the “true” McLafferty

rearrangement. Such a rearrangement is most likely driven by a highly favored 6-

membered transition state that results in a molecular ion that can be stabilized by loss of

H (107). This occurrence in the case of phenylalkanes can be extrapolated to alkyl

substituted naphthalenes and provides a plausible explanation for the observed

fragmentation pattern. A proposed mechanism for this McLafferty rearrangement is

shown in scheme 1 for 3-butyl-1-propylnaphthalene (1) (107).

Figure 5.1regiomeric

m

Scheme 5.1

1 CAD (collimodel comp

HH

m/z 226

1

2

34

Proposed mC

ision energypounds (APC

butyl-

H-shift

mechanism foCAD of 3-bu

y of 20) massCI, CS2) A. -1-propylnap

H H

for the formautyl-1-propyl

s spectra of i1-butyl-3-pr

phthalene.

-

-cleavag

ation of fragmlnaphthalene

isolated molropylnaphtha

H

ge

m/z 1

ment of m/z e (1).

lecular ions oalene and B.

H

184

184 yielded

65

of . 3-

d by

66

This mechanism assumes that position 2 on the naphthalene ring is more sterically

hindered than position 4 by the free degrees of rotation of the propyl group resulting in

the γ-H being mainly transferred to position 4 after ionization.

In addition to the McLafferty rearrangement, 1-butyl-3-propylnaphthalene (2)

undergoes the expected benzylic cleavage to yield an even electron/odd mass fragment as

the primary product (Figure 5.1.B). Unlike its isomer, 1-butyl-3-propylnaphthalene (2)

has a single receiving site for the γ-H transfer which happens be sterically encumbered

(position 2) hence possibly favoring the benzylic cleavage over the McLafferty

rearrangement.

MS3 experiments for both isomers result in further alkyl chain fragmentation with

both even and odd electron fragments being formed (Table A.1).

The CAD spectrum for the molecular ion of a more complex model compound, 3-

butyl-2-ethyl-1-phenylnaphthalene (3) is shown in Figure 5.2. As previously observed,

the McLafferty rearrangement appears to be the major fragmentation pathway in this case

as the only receiving site for the γ-H transfer is sterically unencumbered. This results in

an odd electron fragment that undergoes α-cleavage (Scheme 2). The small fragments at

m/z 183 and 245 (Figures 5.1A and 5.2) result from the benzylic cleavage at the butyl

chain of the respective molecular ions. These observations may be an indication that the

McLafferty rearrangement is the fragmentation pathway of choice for larger alkyl (≥ 3

carbon long) - substituted naphthalenes.

The fragmentation study of isolated protonated molecules of a Maya asphaltene

sample (APCI, toluene) carried out by Borton et al., showed molecules which fragment

67

by loss of their alkyl chains (methylene units) in an exponential decay pattern. As a

possible explanation for a dominant methyl loss, they inferred the presence of several

ethyl groups or some branching at the α-carbon (97). The model compounds we

synthesized are intended to represent smaller aromatic cores with pending alkyl chains of

different lengths. A common observation to both studies is that the main fragmentation

pathway occurs by loss of alkyl chains, however the molecular ions generated by our

naphthalene model compounds are not as robust as the protonated molecules upon CAD

at comparable collision energies. Also, as mentioned, the molecular ions mainly fragment

by loss of the largest alkene molecule (Scheme 5.1) rather than a major methyl group

followed by successive methylene units. This could possibly indicate a change in

fragmentation mechanism of the alkyl chains as a function of the aromatic core size

(number of aromatic rings) and varying lengths of the pending chains which is currently

under investigation in our group.

Figure 5.2

H

m/

Scheme 5.2

CAD (collisbuty

H

/z 288

2 Proposed myielded by

sion energy ol-2-ethyl-1-p

mechanism fCAD of 3-b

of 20) mass phenylnapht

H H

m/z 24

for the formbutyl-2-ethyl

spectra of isthalene (APC

46

-c

ation of fragl-1-phenylna

solated moleCI, CS2).

-CH3

cleavage

gments of m/aphthalene (

ecular ion of

H H

m/z 231

/z 246 and 23).

68

f 3-

231

69

5.2 Isoquinoline-Based Model Compounds

Nitrogen containing aromatics are well established components of asphaltenes.

We synthesized alkyl- and phenyl- substituted isoquinolines to encompass structural

features of hetero-asphaltenes. Those were subjected to APCI/CAD events and a different

fragmentation behavior was observed when compared to their naphthalene counterparts.

The simplest azanaphthalene compound studied was 1-phenylisoquinoline (4)

which mainly fragmented by hydrogen radical loss (Figure 5.3.A, scheme 5.3).

Surprisingly, higher collision energies resulted in the loss of benzyne (m/z 142) followed

by water addition (m/z 160) (Figure 5.3B, scheme 5.3B). This addition may be favorable

to restore the aromaticity of the ring and could be used as a benzylic cleavage indicator in

structural elucidation of asphaltenes.

Analyzing the MS2 spectrum of 3,4-diphenylisoquinoline (5). The main

fragmentation occured by H loss yielding an even electron fragment (m/z 280) which in

turn lost hydrogen cyanide and benzene (Figure 5.4). This fragmentation pattern could be

used as a structural elucidation tool for asphaltene compounds which behave similarly

upon CAD.

Figure 5.3 CAD (collission energy obenzyliso

of 25) mass oquinoline (

spectra of is(APCI, CS2)

solated mole.

ecular ion of

70

f 1-

71

Scheme 5.3 A. Proposed mechanism for the formation of the isoquinolium fragment (m/z 218) by hydrogen radical loss from the molecular ion of 1-benzylisoquinoline (4). B.

Proposed mechanism for the formation of fragments of m/z 142 and 160.

The MS/MS spectra of alkyl-substituted isoquinolines such as 4-methyl-3-

phenylisoquinoline (6) and 4-ethyl-3-phenylisoquinoline (7) showed base peaks formed

by H loss from the molecular ion (Figures 5.4 and 5.5). Hydrogen cyanide was expulsed

from the [M–H] + ion (m/z 218, Figure 5.4) while benzylic cleavage resulted in methyl

radical loss from the base peak (m/z 233, Figure 5.5).

Figure 5.4

Figure 5.5

CAD (collism

CAD (collise

sion energy omethyl-3-phe

sion energy oethyl-3-phen

of 30) mass nylisoquinol

of 25) mass nylisoquinoli

spectra of isline (APCI,

spectra of isine (APCI, C

solated moleCS2).

solated moleCS2).

ecular ion of

ecular ion of

72

f 4-

f 4-

73

These results show the influence of alkyl substituents on heterocyclic aromatic

cores; applying this perspective to asphaltenes, it is possible that the longer the alkyl

subtituent, fragmentations patterns such as benzylic cleavage and McLafferty

rearrangements predominate hydrogen cyanide loss. Supporting this argument, Sample et

al. showed that as methyl substituents on quinoline and isoquinoline rings were

increased, less hydrogen cyanide was loss during electron impact (108).The successive

hydrogen losses observed for compound 6 are not uncommon occurrences, mechanisms

involved for aromatic systems have been vastly studied (109, 110).

3,4-diphenylisoquinoline (5) was synthesized to represent a compact archipelago-

type asphaltene and its CAD spectrum (Figure 5.6) showed hydrogen cyanide and

benzene losses from the [M–H]+ ion (m/z 280). Despite a high collision energy of 40 and

activation Q of 0.5, the intensity of the fragment resulting from benzene cleavage was

rather small. This would suggest that an asphaltene with such features would have a

rather stable core and a maximum of one carbon linkage with other aromatic substituents.

4

m

b

ot

fo

fo

m

sc

p

th

Figure 5.6 C

Other

(phenylethyn

main differen

enzene radic

ther regiome

or the benze

ormation of

methyl group

crambled mo

ossible expl

hrough com

CAD (collisi

models f

nyl)isoquino

nce in the

cal from the

er at the sam

ene radical lo

a stable orth

p from bot

olecular ion

lanations for

mparison bet

ion energy odiphenyli

for archipe

oline (8) and

fragmentatio

e molecular

me collision

oss is shown

ho-pyridyne

th compoun

n. This argum

r the loss o

tween the c

of 40) mass ssoquinoline

lago aspha

d 4-phenyl

on pattern o

ion of comp

n energy valu

n in scheme

fragment.

nds which

ment was m

f a methyl

calculated a

spectra of iso(APCI, CS2

altenes incl

l-3-(phenyle

of these com

pound 8 whi

ue (Figure 5

5.4 and is p

Even more

most likely

made by Bow

radical from

and observe

olated molec2).

luded regio

ethynyl)isoqu

mpounds w

ich was not

5.7). A plau

proposed to

perplexing w

y arose from

wie and Whi

m the stilben

ed ratios fo

cular ion of 3

omers 3-ph

uinoline (9).

was the loss

observed fo

usible mecha

be driven b

was the loss

m a compl

ite as one of

ne molecula

or hydrogen

74

3,4-

henyl-

. The

of a

or the

anism

by the

s of a

letely

f two

ar ion

n and

d

d

th

th

fr

fr

b

m

sp

euterium ra

iphenylacety

hree-member

he molecula

ragment furt

ragment resu

ehavior sugg

molecules c

pectrometry.

Figure 5.7regiomeric m

andomization

ylene which

red ring bip

ar ion of co

ther. Howev

ulting from

gests the stru

ould be co

.

7 CAD (collimodel comp

and

n (111). Su

mainly frag

phenylene ra

ompound 9

ver, the M+

methyl loss

uctural eluc

omplicated

ision energypounds (APCB. 4-phenyl

urprisingly,

gments by ac

adical ion (1

led to a sta

+ peak of com

underwent

idation of as

by varying

y of 45) massCI, CS2) A. 3-3-(phenylet

methyl loss

cetylene loss

11). We can

able even e

mpound 8 lo

further frag

sphaltenes c

g fragment

s spectra of i3-phenyl-4-(thynyl)isoqu

s is not a

s to yield a

n infer that m

electron spec

ost a benzen

gmentations

containing se

tation profi

isolated mol(phenylethynuinoline.

major even

very stable

methyl loss

cies that did

ne radical an

upon CAD.

everal regiom

iles using

lecular ions onyl)isoquino

75

nt for

fused

from

d not

nd the

This

meric

mass

of line

76

Scheme 5.4 A. Proposed mechanism for the formation of the ortho-pyridyne fragment (m/z 304) by hydrogen radical loss from the molecular ion of 3-phenyl-4-

(phenylethynyl)isoquinoline (8). B. Proposed mechanism for the formation of fragments of m/z 228.

5.3 Concluding Remarks and Perspectives

The challenge posed by the structural elucidation of asphaltenes is tackled from

various angles by several research groups. Our approach to this pressing issue involved

the synthesis of naphthalene and isoquinoline- derived compounds that model compact

versions of archipelago and island-type asphaltenes. This was done in an effort to

decipher fundamental fragmentation pathways and mechanisms of asphaltenes.

Fragmentation studies by tandem mass spectrometry yielded the following observations:

77

(i) The island-type regiomeric naphthalenes favor the McLafferty rearrangement

as the major fragmentation pathway whenever steric factors are negligible

causing loss of even electron molecules . This suggests that structural

elucidation of asphaltenes should take into account regiomeric/isomeric

fragmentation differences when assessing possible structures.

(ii) Cleavage of alkyl chains branching off the aromatic core overshadows ring

opening fragmentations which seem to mostly occur during MS3 and MS4

events.

(iii) The azanaphthalenes mainly fragment by hydrogen radical loss yielding very

stable [M–H] + fragments. Hydrogen cyanide loss is dependent on the nature

of the substituent and is not always observed.

(iv) Methyl loss from the molecular ions of the regiomeric isoquinolines was a

surprising observation that led to infer that such a pathway probably gives rise

to a stable fragment at least for one of the regiomers.

Understanding and interpreting fragmentation routes and associated mechanisms for the

model compounds presented herein brings a new insight into assessing asphaltene

structures; differences in fragmentation for regiomeric compounds as well as unexpected

losses caused by the McLafferty mechanism or the complete scrambling of the molecular

ion are possible occurrences when analyzing asphaltene samples by tandem mass

spectrometry. Further delving into the effects of the substituent’s nature (size, length,

conformation, degree of unsaturation) on the fragmentation of asphaltenes should take us

a step closer to elucidating their structure and is actively investigated in our research lab.

78

PART THREE

TOWARD THE SYNTHESIS THE PHOTOAFFINITY LABELED ANALOG AND

MODE OF ACTION STUDIES OF CALLIPELTIN A

79

CHAPTER 6: SYNTHESIS OF THE β-MeO-tyr AZIDE ANALOG OF A NON-PROTEOGENIC AMINO ACID RESIDUE OF CALLIPELTIN A

6.1 Background on Callipeltins and Mirabamides

The fight to annihilate HIV (human immunodeficiency virus), the causative agent

of the acquired immunodeficiency syndrome (AIDS), has been relentless over the last

two decades. This statement is backed by the fact that more treatments are available for

HIV compared to all the other viral illnesses. However, this constant advance is clouded

by factors such as toxicity and drug resistance which makes the quest for new treatment

of paramount importance (112).

In this regard, our group has laid its interest on Callipeltin A (10), a cyclic

depsidecapeptide isolated from the Lithistid sponges Callipelta Sp and Latrunculia sp.

whose structural backbone contains three non-proteogenic amino acid residues:

(2R,3R,4S)-4-amino-7-guanidino-2,3-dihydroxyheptanoic acid (AGDHE), β-methoxy

tyrosine (β-OMeTyr) and (3S,4R)-3,4-dimethyl-L-glutamine (diMeGln) branching off a

lactone macrocycle. Its antiviral activity was measured on HIV-1 (Lai strain) infected-

CEM4 lymphocytic cell lines (ED50 =0.01µg/mL) and it was also shown to have

cytotoxic activity against various cancerous cell lines (NSCLC-N6 non-small-cell-lung

carcinoma, E39 renal carcinoma, P388 murine leukemia and M96 melanoma). Callipeltin

A

gu

m

ou

2

(1

m

m

(1

an

A (10) also d

uinea pig lef

Althou

mechanism o

ur group.

Mirab

007, it was s

11) was not

mirabamide A

most potent o

115). So tha

ntagonistic e

displays ant

ft atria (113,

ugh a lot is

of its mode o

bamides (11)

shown that a

t essential fo

A (11), whic

of its family

at if the 4’

effect on the

ifungal activ

, 114).

known abo

of action hav

Figure 6.1

) are cyclic

a free hydrox

or anti-viral

ch has a rha

y in an asses

position of

e antiviral a

vity and inh

out the biolo

ve not been

Structure of

depsipeptide

xyl group at

activity. Th

mnosylated

sment of the

f the β-MeO

activity in m

hibits the Na

ogical activi

elucidated y

f Callipeltin

es that also

the 4’ posit

hat assertion

β-MeOTyr

e inhibition

OTyr residu

mirabamides,

a+/Ca2+ exch

ity of Callip

yet and is ac

A.

exhibit anti-

tion of the β-

n derived fr

residue, wa

of different

ue can be m

, such modif

hanger (NCX

peltin A (10)

ctively pursu

-HIV potenc

-MeOTyr re

rom the fact

s found to b

strains of H

modified wi

fication cou

80

X) in

), the

ued in

cy. In

esidue

t that

be the

HIV-1

ithout

uld be

ef

ph

in

an

re

an

C

is

si

ffected on C

hotoaffinity

nteracting wi

With

nalog that o

eactivity upo

nd will sub

Callipeltin A

solation of th

itu (Figure 6

Callipeltin A

label that

ith the macro

Figu

that knowle

nce synthesi

on photochem

bstitute the

A (10). Also,

he “click” p

6.3).

A (10) and p

will serve t

ocycle .

ure 6.2 Struc

edge in hand

ized will she

mical excita

hydroxyl g

, a biotin un

product resul

provide the

the purpose

ctures of Mi

d, we came u

ed some ligh

ation, an azid

group at 4’

nit will be a

lting from it

ideal loca

e of identify

irabamides (

up with a de

ht on its mo

de will be us

position of

attached to

ts reaction w

ation for the

ying the rec

11a-d).

esign of a C

ode of action

sed as the ph

f the β-Me

the analog

with interact

e appendage

ceptor mole

Callipeltin A

n. Due to its

hotoaffinity

OTyr residu

to allow a f

ting molecul

81

e of a

ecules

A (10)

s high

label

ue in

facile

les in

82

Figure 6.3 Biotinylated analog of Callipeltin A.

Upon successful conversion of the hydroxyl group of the β-MeOTyr unit to an

azide, the latter fragment will be incorporated in the general structure of Callipeltin A by

solid phase peptide synthesis to obtain Callipeltin analog 12.

6.2 Results and Discussion

The diastereo- and enantioselective synthesis of the β-MeOTyr has been

accomplished by Dr. David Cranfill, starting from 4-hydroxy methylcinnamic acid (116).

We decided to carry out the same synthesis to familiarize ourselves with the chemistries

involved.

We treated 4-hydroxy methylcinnamic acid (13) with sulfuric acid in methanol to

obtain the ester 14 whose hydroxyl substituent was subsequently protected with the tert-

butyldimethylsilyl (TBS) group to afford 15 in 89% yield (89% yield over 2 steps,

Scheme 6.1).

83

Scheme 6.1 Synthesis of the TBS protected 4-hydroxy methylcinnamate.

Having the desired starting material in hand, we proceeded with the key step of

our synthesis: the aziridination of compound 15. This was done by reacting it with a

catalytic amount of Cu(O2OSCF3)2 and (-)-2,2’-isopropylidene[(4S)-4-phenyl-2-

oxazoline] as the ligand. This resulted in the formation of 16 in 50% yield (6:1

diastereomeric ratio (dr) and 7:1 enantiomeric ratio (er)) (Scheme 6.2) and as it was

observed by Dr. Cranfill, when the aziridination was carried out in acetonitrile, the

diastereomeric ratio was lowered to about 2:1.

Scheme 6.2 Aziridination of compound 15.

84

Compound 16 then underwent removal of the nosyl group using 4-

methoxythiophenol to yield the amine 17, which was then coupled with both enantiomers

of the Mosher acid chloride for the enantiomeric excess determination (Scheme 6.3)

(116).

Scheme 6.3 Synthesis of the β-MeOTyr moiety and determination of the enantiomeric excess.

For the synthesis of the β-MeOTyr azide analog, 4-azido benzaldehyde was used

as the starting material, which could be obtained from 4-nitrobenzaldehyde in three steps

(Scheme 6.4) (117).

Scheme 6.4 Synthesis of 4-Azidobenzaldehyde.

85

The protection of the aldehyde group of 4-nitrobenzaldehyde occurred in

quantitative yield. More difficulties were encountered in trying to reduce the nitro group

of 19. Several conditions were attempted, however, the desired product was not obtained

(Table 6.1).

Table 6.1 Conditions for the Reduction of Compound 19

Entry SM Reaction conditions Results 1 19 H2, Pd/C, EtOH/THF 1:1 traces of product 2 19 TBAB, SnCl2.2H2O no desired product 3 19 Zn dust, HCOONH4, MeOH/ EtOAc 1:1 no desired product 4 19 Raney Nickel, HCOOH, H2 no desired product 5 19 H2, Pd/C, EtOH no desired product

We postulated that the resonance product in Scheme 6.5 may undergo

polymerization under the reaction conditions thereby reducing the yield of the amine.

Scheme 6.5 Resonance Mechanism of 20.

86

An alternate route for the synthesis of 4-azido methylcinnamate was performed

(Scheme 6.6).

Scheme 6.6 Alternate route for the synthesis of 4-azido methylcinnamate.

The key step of Scheme 6.6 is the selective reduction of the nitro group of

compound 22 in the presence of the carbon-carbon double bond. The conditions we tried

(hydrazine/formic acid complex in the presence of zinc powder) did not give us the

desired product during later trials. We decided to make other attempts using the reaction

conditions listed in Table 6.2.

87

Table 6.2 Conditions for the Nitro Group Reduction of Compound 22

Entry SM Reaction conditions Results 1 22 Pd(OH)2/C, THF:EtOH (1:1), H2, 30 mins pilot conditions* 2 22 Pd/ BaSO4, EtOAc, Quinoline, H2 no desired product 3 22 Zn dust, HCOONH4, MeOH/ EtOAc 1:1 no desired product 4 22 Hydrazine glyoxylate, Zn dust no desired product 5 22 H2NNH2• HCOOH, Zn, EtOAc no desired product * These conditions reduced both the nitro group and the carbon-carbon double bond. The product was used

in TLC analysis to assess the reduction pattern of the other conditions.

Upon several trials, we settled for the direct reduction of 4-nitrobenzaldehyde to

4-aminobenzaldehyde using SnCl2•2H2O in ethanol at reflux (118). The amine was then

converted into an azide followed by the Horner-Wadsworth-Emmons reaction to yield 24.

(Scheme 6.7)

Scheme 6.7 Synthesis of the β-MeOTyr azide analog 26.

The yield of the aziridination/methanol opening steps was lower than with the 4-

hydroxy methylcinnamate counterpart. We postulated that the electron-donating nature of

the hydroxyl group compared to the azide may facilitate the aziridine formation and

consequently account for the higher yield. Also, the conditions used for the removal of

the nosyl group are presented below (Table 6.3).

88

Table 6.3 Conditions for the Nosyl Group Removal to Obtain Compound 26

Entry SM Reaction conditions Results

(yield/observation)

1 25 HSCH2CH2OH, DBU,DMF Azide reduction

observed

2 25 Thiophenol, K2CO3, CH3CN/DMSO (49:1), reflux at 50 oC (119)

26

3 25 Thiophenol, K2CO3, CH3CN/DMSO (49:1) 21

4 25 4-Methoxythiophenol, K2CO3, CH3CN/DMSO (49:1)

20

We have been intrigued by the results presented in Table 6.3 as we expected 4-

methoxythiophenol to improve the yield of the reaction compared to thiophenol. We

suspected that upon silica gel column chromatography purification some of the amine

product binds irreversibly to the silica gel and this could explain why the yield has been

so constantly low.

At that point in time, our next goal was the determination of the diastereomeric

and enantiomeric ratios of 26 as well as establishing the stereochemistry of its two

stereocenters. In order to do so, we decided to carry out the aziridination reaction in

acetonitrile as it was shown to reduce the diasteromeric ratio of the nosyl-protected β-

MeOTyr while only one diastereomer was obtained when using dichloromethane (116).

89

Scheme 6.8. Conditions for the separation of the diastereomers of 25.

Conditions “a” did not lead to the separation of the two diastereomers while

conditions “b” did. Upon analysis of the 1H NMR spectrum of the diastereomeric

mixture, we came up with a diastereomeric ratio of 1.4:1 by integration.

The stereochemistry of the two stereocenters of 25 was determined by 1H NMR

comparison of 16 and 25 (as well as the mixture obtained using conditions “b”) (Scheme

6.8).

F

Figure 6.4 BBlown-up Area of the H1

(, H2 and H3 p(conditions “

peaks from t“a”).

the 1H NMRR spectrum o

90

of 25

ca

d

tw

d

as

sa

w

F

Figure 6.5 B

From

an confident

ichlorometh

wo stereocen

iffer by the

ssumed that

ame or vary

would be quit

igures 6.5 a

Blown-up are

the data (m

tly say that t

hane is the m

nters in com

e nature of t

t the couplin

y slightly fo

te different f

and 6.6, we

ea of the H1,(

mostly chemic

the single dia

major diaster

mpound 16 w

their substit

ng constants

or the two m

for the mino

can see tha

, H2 and H3 p(conditions “

cal shifts) pr

astereomer w

reomer from

were determi

tuent at the

s between H

major diaste

or diastereom

at the couplin

peaks from t“b”).

resented in b

we obtained

m the mixture

ined to be 2

para-positi

H1, H2 and H

ereomers of

mer. So, look

ng constant

the 1H NMR

both Figures

d upon runnin

e. The stereo

2R,3R. Since

ion of the b

H3 would be

f compounds

king at the ta

s of H1, H2

R spectrum o

s 6.4 and 6.5

ng the reacti

ochemistry o

e 16 and 25

benzene ring

pretty muc

s 16 and 25

ables present

and H3 are

91

of 25

5, we

ion in

of the

only

g, we

ch the

5 and

ted in

very

m

d

2R

ri

F

co

w

ac

much close f

isparity is o

R,3R stereoc

ing.

Figure 6.6 B

Once

ompound 25

was. At that

cid) was no

for compoun

observed wit

chemistry is

lown-up Are

we had de

5, the next lo

time, α-me

ot immediate

nd 16 and th

th the minor

s conserved

ea of the H1,

etermined th

ogical step w

ethoxy-α-(tr

ely available

he major dia

r one. We th

with the az

, H2 and H3 p

he diastereom

was for us t

ifluoromethy

e to us so,

astereomer o

herefore cam

zide at the p

peaks from t

meric ratio

to find out w

yl)phenylace

we decided

of compoun

me to the co

para position

the 1H NMR

and the ste

what the ena

etic acid (M

d to synthes

d 25 while

onclusion tha

n of the ben

R spectrum o

ereochemistr

antiomeric ex

MTPA, Mos

size a new c

92

some

at the

nzene

of 16.

ry of

xcess

sher’s

chiral

93

derivatizing agent (CAD): a tartaric acid derivative, 28a, which would be prepared

according to Scheme (6.9) (120).

Scheme 6.9 Formation of CAD 28.

Diethyl tartrate (DET) was used as the starting material instead of the dimethyl

analog. We attempted the direct acetal formation of benzophenone with DET using Dean-

Stark conditions and varying the acid catalyst without success (Scheme 6.10).

Scheme 6.10 Conditions Attempted for the Direct Formation of Compound 28a.

We proceeded to convert benzophenone to benzophenone dimethyl acetal and

react it with L- or D-DET. Below are the different conditions used to achieve that goal.

94

Table 6.4 Conditions for the Formation of Benzophenone Dimethyl Acetal

Entry SM Reaction conditions Results (yield)

1 29

HC(OCH3)3, p-TsOH, MeOH

SM recovered

2 29 *HC(OCH3)3, Ce(OTf)3, MeOH, Et3N (121)

SM recovered

3 29 HC(OCH3)3, LiBF4, MeOH (122)

SM recovered

4 29 * *HC(OCH3)3, MeNO2, triflic acid, MeOH (123)

43%

* Ce(OTf)3 may have not been in usable conditions due to its yellowish physical appearance instead of the normal white.

** Yield obtained after recrystallization

Once we obtained the benzophenone dimethyl acetal, we attempted to make

compound 28a (Scheme 6.11).

Scheme 6.11 Other conditions attempted for the synthesis of compound 28a.

It became evident that the transacetalization reaction was extremely water

sensitive. The conditions “a” and “b” were repeated several times, each time trying to

su

an

u

M

g

β

uppress wate

nd unreacted

By th

s so, we dec

Mosher meth

S

Of the

ave us the b

-MeOTyr az

er by any w

d DET.

en the R and

cided to aban

od (Scheme

Scheme 6.12

e three cond

best result (1

zide counterp

way possible

d S enantiom

ndon the syn

6.12).

2 Enantiome

ditions used

24) and we

part to be 22

but we alw

mers of the M

nthesis of th

ric ratio dete

d to make th

were able to

2:1 (Figure 6

ways ended u

Mosher’s ac

he CAD 28a

ermination o

he Mosher a

o determine

6.7)

up recoverin

cids had bec

a and rely on

of compound

acid chloride

the enantiom

ng benzophe

come availab

n the well-kn

d 26.

e, condition

meric ratio o

95

enone

ble to

nown

ns “c”

of the

an

m

fo

si

w

th

F

We h

nalog of th

milestone in

ollowed by

imulate the

with the gp12

he biological

igure 6.7 19F

have efficien

he non-prote

uncovering

a model “C

covalent lab

20 and gp41

l mode of ac

A test for

remains th

F NMR for th

6.3 Conclu

ntly synthes

eogenic am

g the mode

Click reactio

bel of the bi

and light irr

ction studies

r viral fusio

he same rega

he R and S e

usion and Fut

sized and fu

mino acid re

of action o

on” on 26 w

iotinylated a

radiation. Fin

of Callipelti

on inhibition

ardless of the

enantiomers

ture Directio

ully charact

esidue β-Me

of Callipeltin

with an alky

analog of C

nally we wil

in A. Those

n to ensure

e modificatio

of compoun

ons

terized the

eOTyr. Th

n A. This e

yne of our c

allipeltin A

ll synthesize

studies will

that the bi

on.

nd 33.

β-MeOTyr

his constitu

endeavor wi

choice. This

upon incub

e 12 and carr

consist of:

iological ac

96

azide

utes a

ill be

s will

bation

ry out

ctivity

97

Following a successful testing, incubation of the biotinylated analog of

Callipeltin A with the glycoproteins gp120 and gp41 with irradiation will

be carried out.

Identification, extraction and characterization will then be effected

LIST OF REFERENCES

98

LIST OF REFERENCES

1. Wu Y, Zhou Z, Fan Q, Chen L, Zhu M. Facile in-situ fabrication of novel organic nanoparticle hydrogels with excellent mechanical properties. Journal of Materials Chemistry. 2009;19(39):7340-6.

2. http://www.fellins.com/blog/opal-octobers-birthstone-2.

3. Wei B, Rogers BJ, Wirth MJ. Slip Flow in Colloidal Crystals for Ultraefficient Chromatography. Journal of the American Chemical Society. 2012;134(26):10780-2.

4. Birdsall RE, Koshel BM, Hua Y, Ratnayaka SN, Wirth MJ. Modeling of protein electrophoresis in silica colloidal crystals having brush layers of polyacrylamide. Electrophoresis. 2013.

5. Mahler H-C, Friess W, Grauschopf U, Kiese S. Protein aggregation: Pathways, induction factors and analysis. Journal of Pharmaceutical Sciences. 2009;98(9):2909-34. doi: 10.1002/jps.21566.

6. Schwartz JC, Senko MW, Syka JEP. A two-dimensional quadrupole ion trap mass spectrometer. Journal of the American Society for Mass Spectrometry. 2002;13(6):659-69. doi: http://dx.doi.org/10.1016/S1044-0305(02)00384-7.

7. Huang X, Doneski LJ, Wirth MJ. Surface-confined living radical polymerization for coatings in capillary electrophoresis. Analytical chemistry. 1998;70(19):4023-9.

8. Hamrang Z, Rattray N, Pluen A. Proteins behaving badly: emerging technologies in profiling biopharmaceutical aggregation. Trends in biotechnology. 2013. doi: 10.1016/j.tibtech.2013.05.004.

9. Rosenberg A. Effects of protein aggregates: an immunologic perspective. The AAPS journal. 2006;8(3):7. doi: 10.1208/aapsj080359.

10. Strohl WR. Therapeutic monoclonal antibodies: past, present, and future. Therapeutic Monoclonal Antibodies: From Bench to Clinic. 2009:1-50.

99

11. Biosimilars and Follow-On Biologics Report: The Global Outlook 2010-2025. Visiongain Ltd. 2010.

12. Patel J, Kothari R, Tunga R, Ritter NM, Tunga BS. Stability considerations for biopharmaceuticals, part 1. BioProc Int. 2011;9:20-31.

13. Wang W, Nema S, Teagarden D. Protein aggregation—Pathways and influencing factors. International Journal of Pharmaceutics. 2010;390(2):89-99. doi: http://dx.doi.org/10.1016/j.ijpharm.2010.02.025.

14. Bhatnagar B, Bogner R, Pikal M. Protein stability during freezing: separation of stresses and mechanisms of protein stabilization. Pharmaceutical development and technology. 2007;12(5):505-23. doi: 10.1080/10837450701481157.

15. Hawe A, Kasper J, Friess W, Jiskoot W. Structural properties of monoclonal antibody aggregates induced by freeze-thawing and thermal stress. European journal of pharmaceutical sciences : official journal of the European Federation for Pharmaceutical Sciences. 2009;38(2):79-87. doi: 10.1016/j.ejps.2009.06.001.

16. Chi E, Krishnan S, Randolph T, Carpenter J. Physical stability of proteins in aqueous solution: mechanism and driving forces in nonnative protein aggregation. Pharm Res. 2003;20(9):1325-36. doi: 10.1023/A:1025771421906.

17. Halperin A. Polymer Brushes that Resist Adsorption of Model Proteins: Design Parameters. Langmuir. 1999;15. doi: 10.1021/la981356f.

18. Kizhakkedathu JN, Janzen J, Le Y, Kainthan RK, Brooks DE. Poly(oligo(ethylene glycol)acrylamide) Brushes by Surface Initiated Polymerization: Effect of Macromonomer Chain Length on Brush Growth and Protein Adsorption from Blood Plasma. Langmuir. 2009;25(6):3794-801. doi: 10.1021/la803690q.

19. Wang D, Shi J, Carlson SR, Cregan PB, Ward RW, Diers BW. A Low-Cost, High-Throughput Polyacrylamide Gel Electrophoresis System for Genotyping with Microsatellite DNA Markers. Crop Sci. 2003;43(5):1828-32. doi: 10.2135/cropsci2003.1828.

20. Görg A, Boguth G, Obermaier C, Posch A, Weiss W. Two-dimensional polyacrylamide gel electrophoresis with immobilized pH gradients in the first dimensin (IPG-Dalt): The state of the art and the controversy of vertical versus horizontal systems. ELECTROPHORESIS. 1995;16(1):1079-86. doi: 10.1002/elps.11501601183.

100

21. Manz A, Pamme N, Lossifidis D. Bioanalytical chemistry/Andreas Manz, Nicole Pamme, Dimitri Iossifidis: Imperial College Press; 2004.

22. Engelsman J, Garidel P, Smulders R, Koll H, Smith B, Bassarab S, et al. Strategies for the Assessment of Protein Aggregates in Pharmaceutical Biotech Product Development. Pharm Res. 2011;28(4):920-33. doi: 10.1007/s11095-010-0297-1.

23. Chrambach A, Rodbard D. Polyacrylamide Gel Electrophoresis. Science. 1971;172(3982):440-51. doi: 10.2307/1731589.

24. Hames BD. Gel Electrophoresis of Proteins: A Practical Approach: A Practical Approach: Oxford University Press; 1998.

25. Gromova I, Celis JE. Protein detection in gels by silver staining: A procedure compatible with mass-spectrometry. Cell biology: A laboratory Handbook. 2006;4:421-9.

26. Switzer III RC, Merril CR, Shifrin S. A highly sensitive silver stain for detecting proteins and peptides in polyacrylamide gels. Analytical biochemistry. 1979;98(1):231-7.

27. Neuhoff V, Arold N, Taube D, Ehrhardt W. Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brilliant Blue G-250 and R-250. ELECTROPHORESIS. 1988;9(6):255-62. doi: 10.1002/elps.1150090603.

28. Steinberg TH, Jones LJ, Haugland RP, Singer VL. SYPRO Orange and SYPRO Red Protein Gel Stains: One-Step Fluorescent Staining of Denaturing Gels for Detection of Nanogram Levels of Protein. Analytical Biochemistry. 1996;239(2):223-37. doi: http://dx.doi.org/10.1006/abio.1996.0319.

29. Patton WF. Detection technologies in proteome analysis. Journal of Chromatography B. 2002;771(1–2):3-31. doi: http://dx.doi.org/10.1016/S1570-0232(02)00043-0.

30. Guttman A. On the separation mechanism of capillary sodium dodecyl sulfate-gel electrophoresis of proteins. ELECTROPHORESIS. 1995;16(1):611-6. doi: 10.1002/elps.1150160198.

31. Stellwagen N. DNA Gel Electrophoresis. In: Tietz D, editor. Nucleic Acid Electrophoresis: Springer Berlin Heidelberg; 1998. p. 1-53.

101

32. Cohen AS, Karger BL. High-performance sodium dodecyl sulfate polyacrylamide gel capillary electrophoresis of peptides and proteins. Journal of Chromatography A. 1987;397(0):409-17. doi: http://dx.doi.org/10.1016/S0021-9673(01)85026-3.

33. Guttman A. Capillary sodium dodecyl sulfate-gel electrophoresis of proteins. ELECTROPHORESIS. 1996;17(8):1333-41. doi: 10.1002/elps.1150170810.

34. Zhu Z, Lu JJ, Liu S. Protein separation by capillary gel electrophoresis: A review. Analytica Chimica Acta. 2012;709(0):21-31. doi: http://dx.doi.org/10.1016/j.aca.2011.10.022.

35. Ruiz-Martinez MC, Berka J, Belenkii A, Foret F, Miller AW, Karger BL. DNA sequencing by capillary electrophoresis with replaceable linear polyacrylamide and laser-induced fluorescence detection. Analytical chemistry. 1993;65(20):2851-8.

36. Widhalm A, Schwer C, Blaas D, Kenndler E. Capillary zone electrophoresis with a linear, non-cross-linked polyacrylamide gel: separation of proteins according to molecular mass. Journal of Chromatography A. 1991;549:446-51.

37. Barron AE, Sunada WM, Blanch HW. The effects of polymer properties on DNA separations by capillary electrophoresis in uncross-linked polymer solutions. ELECTROPHORESIS. 1996;17(4):744-57. doi: 10.1002/elps.1150170421.

38. Hong P, Koza S, Bouvier ES. A review size-exclusion chromatography for the analysis of protein biotherapeutics and their aggregates. Journal of liquid chromatography & related technologies. 2012;35(20):2923-50.

39. Gabrielson JP, Brader ML, Pekar AH, Mathis KB, Winter G, Carpenter JF, et al. Quantitation of aggregate levels in a recombinant humanized monoclonal antibody formulation by size-exclusion chromatography, asymmetrical flow field flow fractionation, and sedimentation velocity. Journal of Pharmaceutical Sciences. 2007;96(2):268-79. doi: 10.1002/jps.20760.

40. Grimm R, Galatioto LE. Purification of Monoclonal and Polyclonal Antibodies by HPLC.

41. Hage DS, Carr JD. Analytical Chemistry and Quantitative Analysis: Prentice Hall PTR; 2011.

42. Xuan X, Li D. Joule heating effects on peak broadening in capillary zone electrophoresis. Journal of Micromechanics and Microengineering. 2004;14(8):1171.

102

43. Camilleri P. Capillary electrophoresis: theory and practice: CRC press; 1998.

44. Gomis DB, Junco S, Expósito Y, Gutiérrez D. Size‐based separations of proteins by capillary electrophoresis using linear polyacrylamide as a sieving medium: Model studies and analysis of cider proteins. Electrophoresis. 2003;24(9):1391-6.

45. Guttman A, Nolan J. Comparison of the separation of proteins by sodium dodecyl sulfate-slab gel electrophoresis and capillary sodium dodecyl sulfate-gel electrophoresis. Analytical biochemistry. 1994;221(2):285-9.

46. Koenderink AF, Johnson PM, Galisteo López JF, Vos WL. Three-dimensional photonic crystals as a cage for light. Comptes Rendus Physique. 2002;3(1):67-77. doi: http://dx.doi.org/10.1016/S1631-0705(02)01301-4.

47. Xia Y, Gates B, Yin Y, Lu Y. Monodispersed colloidal spheres: old materials with new applications. Advanced Materials. 2000;12(10):693-713.

48. Yablonovitch E. Inhibited spontaneous emission in solid-state physics and electronics. Physical review letters. 1987;58(20):2059.

49. Cong H, Yu B, Tang J, Li Z, Liu X. Current status and future developments in preparation and application of colloidal crystals. Chemical Society Reviews. 2013;42(19):7774-800. doi: 10.1039/C3CS60078E.

50. Sanders J. Colour of precious opal. 1964.

51. Pan Z, Ma J, Yan J, Zhou M, Gao J. Response of inverse-opal hydrogels to alcohols. Journal of Materials Chemistry. 2012;22(5):2018-25.

52. Yamada H, Nakamura T, Yamada Y, Yano K. Colloidal‐Crystal Laser Using Monodispersed Mesoporous Silica Spheres. Advanced Materials. 2009;21(41):4134-8.

53. Haugan E, Granlund H, Gjessing J, Marstein E. Colloidal crystals as templates for light harvesting structures in solar cells. Energy Procedia. 2011;10:292-6.

54. Kuo Y-C, Wang C-C. Guided differentiation of induced pluripotent stem cells into neuronal lineage in alginate–chitosan–gelatin hydrogels with surface neuron growth factor. Colloids and Surfaces B: Biointerfaces. 2013;104:194-9.

103

55. Zeng Y, Harrison DJ. Self-assembled colloidal arrays as three-dimensional nanofluidic sieves for separation of biomolecules on microchips. Analytical chemistry. 2007;79(6):2289-95.

56. Liu Y, Goebl J, Yin Y. Templated synthesis of nanostructured materials. Chemical Society Reviews. 2013;42(7):2610-53. doi: 10.1039/C2CS35369E.

57. Rogers BJ, Birdsall RE, Wu Z, Wirth MJ. RPLC of intact proteins using sub-0.5 um particles and commercial instrumentation. Analytical Chemistry.

58. Zhang Z, Wu Z, Wirth MJ. Polyacrylamide brush layer for Hydrophilic Interaction Liquid Chromatography of intact glycoproteins. Journal of Chromatography A. 2013.

59. Wei B, Malkin DS, Wirth MJ. Plate heights below 50 nm for protein electrochromatography using silica colloidal crystals. Analytical chemistry. 2010;82(24):10216-21.

60. Ciampolini M, Nardi N. Five-Coordinated High-Spin Complexes of Bivalent Cobalt, Nickel, andCopper with Tris (2-dimethylaminoethyl) amine. Inorganic chemistry. 1966;5(1):41-4.

61. Chabanov AA, Jun Y, Norris DJ. Avoiding cracks in self-assembled photonic band-gap crystals. Applied physics letters. 2004;84(18):3573-5.

62. Newby JJ, Legg MA, Rogers B, Wirth MJ. Annealing of silica to reduce the concentration of isolated silanols and peak tailing in reverse phase liquid chromatography. Journal of Chromatography A. 2011;1218(31):5131-5. doi: http://dx.doi.org/10.1016/j.chroma.2011.05.080.

63. Matyjaszewski K, Gaynor S, Greszta D, Mardare D, Shigemoto T. ‘Living’ and controlled radical polymerization. Journal of Physical Organic Chemistry. 1995;8(4):306-15. doi: 10.1002/poc.610080414.

64. Xiao D, Wirth MJ. Kinetics of surface-initiated atom transfer radical polymerization of acrylamide on silica. Macromolecules. 2002;35(8):2919-25.

65. Hua Y, Koshel BM, Wirth MJ. Field-free remobilization of proteins after isoelectric focusing in packed capillaries. Analytical chemistry. 2010;82(21):8910-5.

66. Zheng S, Ross E, Legg MA, Wirth MJ. High-speed electroseparations inside silica colloidal crystals. Journal of the American Chemical Society. 2006;128(28):9016-7.

104

67. Zeng Y, He M, Harrison DJ. Microfluidic Self-Patterning of Large-Scale Crystalline Nanoarrays for High-Throughput Continuous DNA Fractionation. Angewandte Chemie. 2008;120(34):6488-91. doi: 10.1002/ange.200800816.

68. Ogston A. The spaces in a uniform random suspension of fibres. Transactions of the Faraday Society. 1958;54:1754-7.

69. Ogston A, Phelps C. Exclusion of inulin from solutions of hyaluronic acid. Nature. 1960;187:1024.

70. Weiss R, Frazier AB. High lane density slab-gel electrophoresis using micromachined instrumentation. Electrophoresis. 2001;22:3908-15.

71. Kopecka K, Drouin G, Slater GW. Capillary electrophoresis sequencing of small ssDNA molecules versus the Ogston regime: Fitting data and interpreting parameters. ELECTROPHORESIS. 2004;25(14):2177-85. doi: 10.1002/elps.200305951.

72. Slavíčková A. Electrophoresis of nonreduced IgM and IgG monoclonal antibodies on mixed agarose‐polyacrylamide gels. Electrophoresis. 1990;11(10):892-.

73. Mullins OC, Sheu EY, Hammami A, Marshall AG. Asphaltenes, heavy oils, and petroleomics. New York: Springer,; 2007.

74. Akbarzadeh K HA, Kharrat A, Zhang D, Allenson S, Creek J, Kabir S, Jamaluddin A, Marshall AG, Rodgers RP, Mullins OC and Solbakken T. Asphaltenes, Problematic but Rich in Potential. Oilfield Review. 2007;19(2):22-43.

75. Pinkston DS, Duan P, Gallardo VA, Habicht SC, Tan XL, Qian KN, et al. Analysis of Asphaltenes and Asphaltene Model Compounds by Laser-Induced Acoustic Desorption/Fourier Transform Ion Cyclotron Resonance Mass Spectrometry. Energy & Fuels. 2009;23:5564-70.

76. Hughey CA, Rodgers RP, Marshall AG. Resolution of 11 000 compositionally distinct components in a single Electrospray ionization Fourier transform ion cyclotron resonance mass spectrum of crude oil. Analytical Chemistry. 2002;74(16):4145-9.

77. Klein GC, Kim S, Rodgers RP, Marshall AG. Mass spectral analysis of asphaltenes. I. Compositional differences between pressure-drop and solvent-drop asphaltenes determined by electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Energy & Fuels. 2006;20(5):1965-72.

105

78. Klein GC, Kim S, Rodgers RP, Marshall AG, Yen A. Mass spectral analysis of asphaltenes. II. Detailed compositional comparison of asphaltenes deposit to its crude oil counterpart for two geographically different crude oils by ESI FT-ICR MS. Energy & Fuels. 2006;20(5):1973-9.

79. Ancheyta J, Betancourt G, Centeno G, Marroquin G, Alonso F, Garciafigueroa E. Catalyst deactivation during hydroprocessing of Maya heavy crude oil. 1. Evaluation at constant operating conditions. Energy & Fuels. 2002;16(6):1438-43.

80. Ancheyta J, Betancourt G, Marroquin G, Centeno G, Castaneda LC, Alonso F, et al. Hydroprocessing of Maya heavy crude oil in two reaction stages. Applied Catalysis a-General. 2002;233(1-2):159-70.

81. Ancheyta J, Betancourt-Rivera G, Marroquin-Sanchez G, Perez-Arellano AM, Maity SK, Cortez T, et al. An exploratory study for obtaining synthetic crudes from heavy crude oils via hydrotreating. Energy & Fuels. 2001;15(1):120-7.

82. Trejo F, Centeno G, Ancheyta J. Precipitation, fractionation and characterization of asphaltenes from heavy and light crude oils. Fuel. 2004;83(16):2169-75.

83. Head IM, Jones DM, Larter SR. Biological activity in the deep subsurface and the origin of heavy oil. Nature. 2003;426(6964):344-52.

84. Mullins OC. The Asphaltenes. Annual Review of Analytical Chemistry, Vol 4. 2011;4. doi: 10.1146/annurev-anchem-061010-113849.

85. Souza RD, Nicodem DE, Garden SJ, Correa RJ. Study of the Asphaltene Aggregation Structure by Time-Resolved Fluorescence Spectroscopy. Energy & Fuels. 2010;24:1135-8.

86. Tanaka R, Sato E, Hunt JE, Winans RE, Sato S, Takanohashi T. Characterization of asphaltene aggregates using X-ray diffraction and small-angle X-ray scattering. Energy & Fuels. 2004;18(4). doi: 10.1021/ef034082z.

87. Groenzin H, Mullins OC. Molecular size and structure of asphaltenes. Petroleum Science and Technology. 2001;19(1-2):219-30.

88. Groenzin H, Mullins OC. Asphaltene Molecular Size and Structure. The Journal of Physical Chemistry A J Phys Chem A. 1999;103(50):11237-45.

89. Groenzin H, Mullins OC. Molecular size and structure of asphaltenes from various sources. Energy & Fuels. 2000;14(3):677-84.

106

90. Ruiz-Morales Y, Mullins OC. Polycyclic aromatic hydrocarbons of asphaltenes analyzed by molecular orbital calculations with optical spectroscopy. Energy & fuels. 2007;21(1):256-65.

91. Mullins OC. The Modified Yen Model. Energy & Fuels. 2010;24:2179-207.

92. Strausz OP, Lown EM. The chemistry of Alberta oil sands, bitumens and heavy oils: Alberta Energy Research Institute Calgary; 2003.

93. Pelet R, Behar F, Monin JC. Resins and Asphaltenes in the Generation and Migration of Petroleum. Org Geochem. 1986;10(1-3):481-98.

94. Alshareef AH, Scherer A, Tan XL, Azyat K, Stryker JM, Tykwinski RR, et al. Formation of Archipelago Structures during Thermal Cracking Implicates a Chemical Mechanism for the Formation of Petroleum Asphaltenes. Energy & Fuels. 2011;25(5):2130-6.

95. Rakotondradany F, Fenniri H, Rahimi P, Gawrys KL, Kilpatrick PK, Gray MR. Hexabenzocoronene model compounds for asphaltene fractions: Synthesis & characterization. Energy & fuels. 2006;20(6):2439-47.

96. Nordgård EL, Sørland G, Sjöblom J. Behavior of asphaltene model compounds at W/O interfaces. Langmuir. 2009;26(4):2352-60.

97. Francisco MA, Garcia R, Chawla B, Yung C, Qian K, Edwards KE, et al. Syntheses of Alkylated Aromatic Model Compounds To Facilitate Mass Spectral Characterization of Heavy Oils, Resids, and Bitumens. Energy & Fuels. 2011;25(10):4600-5.

98. Borton D, Pinkston DS, Hurt MR, Tan XL, Azyat K, Scherer A, et al. Molecular Structures of Asphaltenes Based on the Dissociation Reactions of Their Ions in Mass Spectrometry. Energy & Fuels. 2010;24:5548-59.

99. Kabalka GW, Ju YH, Wu ZZ. A new titanium tetrachloride mediated annulation of alpha-aryl-substituted carbonyl compounds with alkynes: A simple and highly efficient method for the regioselective synthesis of polysubstituted naphthalene derivatives. J Org Chem. 2003;68(20):7915-7.

100. Collado D, Perez-Inestrosa E, Suau R. Covalently linked acceptor-donor systems based on isoquinoline N-oxide acceptor: Photoinduced electron transfer produces dual-channel luminescent systems that evolve chemically to photohydroxylation of the aromatic donor. The Journal of organic chemistry. 2003;68(9):3574-84.

107

101. Roesch KR, Larock RC. Synthesis of Isoquinolines and Pyridines via Palladium-Catalyzed Iminoannulation of Internal Acetylenes. J Org Chem. 1998;63(16):5306-7.

102. Roesch KR, Zhang H, Larock RC. Synthesis of isoquinolines and pyridines by the palladium-catalyzed iminoannulation of internal alkynes. The Journal of Organic Chemistry. 2001;66(24):8042-51.

103. March RE, Todd JFJ. Linear Quadrupole Ion Trap Mass Spectrometer. Quadrupole Ion Trap Mass Spectrometry: John Wiley & Sons, Inc.; 2005. p. 161-87.

104. March RE. An introduction to quadrupole ion trap mass spectrometry. Journal of Mass Spectrometry. 1997;32(4):351-69.

105. Wong PS, Graham Cooks R. Ion trap mass spectrometry. Current separations. 1997;16:85-92.

106. Sunner J, Nicol G, Kebarle P. Factors determining relative sensitivity of analytes in positive mode atmospheric pressure ionization mass spectrometry. Analytical Chemistry. 1988;60(13):1300-7.

107. Lehmann WD. JÃ¼rgen H. Gross: Mass spectrometryâ€“A Textbook. Analytical and Bioanalytical Chemistry. 2011:1-2.

108. Sample SD, Lightner DA, Buchardt O, Djerassi C. Mass spectrometry in structural and stereochemical problems. CXXIV. Mass spectral fragmentation of alkylquinolines and isoquinolines. The Journal of Organic Chemistry. 1967;32(4):997-1005.

109. Pachuta SJ, Kenttamaa HI, Sack TM, Cerny RL, Tomer KB, Gross ML, et al. Excitation and dissociation of isolated ions derived from polycyclic aromatic hydrocarbons. Journal of the American Chemical Society. 1988;110(3):657-65. doi: 10.1021/ja00211a001.

110. Dibben MJ, Kage D, Szczepanski J, Eyler JR, Vala M. Photodissociation of the Fluorene Cation:â€‰ A Fourier Transform Ion Cyclotron Resonance Mass Spectrometric Study. The Journal of Physical Chemistry A. 2001;105(25):6024-9. doi: 10.1021/jp0100708.

111. Bowie JH, White PY. The loss of Mė from the stilbene molecular ion. A reassessment. Organic Mass Spectrometry. 1972;6(2):135-9. doi: 10.1002/oms.1210060203.

108

112. Mehellou Y, De Clercq E. Twenty-six years of anti-HIV drug discovery: where do we stand and where do we go? J Med Chem. 2010;53(2):521-38.

113. Zampella A, D'Auria MV, Paloma LG, Casapullo A, Minale L, Debitus C, et al. Callipeltin A, an anti-HIV cyclic depsipeptide from the New Caledonian Lithistida sponge Callipelta sp. Journal of the American Chemical Society. 1996;118(26):6202-9.

114. Trevisi L, Cargnelli G, Ceolotto G, Papparella I, Semplicini A, Zampella A, et al. Callipeltin A: sodium ionophore effect and tension development in vascular smooth muscle. Biochemical pharmacology. 2004;68(7):1331-8.

115. Plaza A, Gustchina E, Baker HL, Kelly M, Bewley CA. Mirabamides A–D, depsipeptides from the sponge Siliquariaspongia mirabilis that inhibit HIV-1 fusion. Journal of natural products. 2007;70(11):1753-60.

116. Cranfill DC, Lipton MA. Enantio-and Diastereoselective Synthesis of (R, R)-β-Methoxytyrosine. Org Lett. 2007;9(18):3511-3.

117. Walton R, Lahti PM. An efficient, simple synthesis of 4-azidobenzaldehyde. Synthetic communications. 1998;28(6):1087-92.

118. Bellamy F, Ou K. Selective reduction of aromatic nitro compounds with stannous chloride in non acidic and non aqueous medium. Tetrahedron Lett. 1984;25(8):839-42.

119. Mao H, Joly GJ, Peeters K, Hoornaert GJ, Compernolle F. Synthesis of 1-deoxymannojirimycin analogues using N-tosyl and N-nosyl activated aziridines derived from 1-amino-1-deoxyglucitol. Tetrahedron. 2001;57(32):6955-67.

120. Shim Y-J, Choi K. Cyclic ketals of tartaric acid: simple and tunable reagents for the determination of the absolute configuration of primary amines. Org Lett. 2010;12(4):880-2.

121. Ono F, Inatomi Y, Tada Y, Mori M, Sato T. A facile procedure for acetalization of aldehydes and ketones catalyzed by cerium (III) trifluoromethanesulfonate. Chemistry Letters. 2009;38(1):96-7.

122. Hamada N, Kazahaya K, Shimizu H, Sato T. An efficient and versatile procedure for the synthesis of acetals from aldehydes and ketones catalyzed by lithium tetrafluoroborate. Synlett. 2004;2004(06):1074-6.

123. Thurkauf A, Jacobson AE, Riee KC. An improved procedure for the preparation of acetals from diaryl ketones. Synthesis. 1988;1988(03):233-4.

109

124. Ward DE, Rhee CK. A simple method for the microscale preparation of Mosher's acid chloride. Tetrahedron Lett. 1991;32(49):7165-6.

125. Yang F, Zhang J, Wu Y. Facile synthesis of isoquinolines by imination and subsequent palladacycle-catalyzed iminoannulation of internal alkynes. Tetrahedron. 2011;67(16):2969-73. doi: http://dx.doi.org/10.1016/j.tet.2011.02.044.

126. Si C, Myers AG. A versatile synthesis of substituted isoquinolines. Angewandte Chemie International Edition. 2011;50(44):10409-13.

127. Korivi RP, Cheng CH. Highly Efficient Synthesis of Isoquinolines via Nickel-Catalyzed Annulation of 2-Iodobenzaldimines with Alkynes: Evidence for Dual Pathways of Alkyne Insertion. Org Lett. 2005;7(23):5179-82.

128. Mandadapu AK, Dathi MD, Arigela RK, Kundu B. Synthesis of 8-aryl substituted benzo [a] phenanthridine derivatives by consecutive three component tandem reaction and 6- endo carbocyclization. Tetrahedron. 2012.

APPENDICES

110

Appendix A: Supplemental Data for Chapters 4 and 5

A.1 1H and 13C NMR of Asphaltene Model Compounds


purification. 1H NMR spectra were recorded at 300 MHz while 75 MHz was used for 13 C

nuclei. 230-400 mesh silica gel was used for purification by flash chromatography.

Compounds 1-3 were synthesized using a protocol published elsewhere (99).

3-Butyl-1-propylnaphthalene (1). Off white oil. 1H NMR (300 MHz, CDCl3) δ

(ppm) 0.97 (t, J = 7.0 Hz, 3H), 0.98 (t, J = 7.0 Hz, 3H), 1.46 (m, 2H), 1.73 (m,

4H), 2.74 (t, J = 7.5 Hz, 2H), 3.04 (t, J = 7.7 Hz, 2H), 7.18 (d, J = 1.6 Hz, 1H),

7.43 (dt , J = 3.0 Hz, 2H), 7.47 (d, J = 1.7 Hz, 1H), 7.78 (dd, J = 3.5 and 5.7 Hz,

1H), 8.00 (dd, J = 3.3 and 6.2 Hz, 1H); 13C NMR (75 MHz, CDCl3) δ (ppm)

13.83, 13.92, 22.81, 24.33, 32.73, 32.95, 38.08, 124.63, 124.70, 125.22, 127.43,

127.48, 128.11, 130.27, 134.00, 138.63, 139.65.

1-Butyl-3-propylnaphthalene (2). Colorless oil 1H NMR (300 MHz, CDCl3) δ

(ppm) 0.95 (t, J = 9 Hz, 3H), 1.03 (t, J = 6 Hz, 3H), 1.40 (m, 2H), 1.69 (m, 2H),

1.77 m, 2H), 2.74 (t, J = 9 Hz, 2H), 3.02 (t, J = 6 Hz, 2H), 7.18 (d, J = 1.8 Hz,

1H), 7.42 (dt, J = 3.5 Hz, 2H), 7.48 (s, 1H), 7.78 (dd, J = 3.4 and 6.3 Hz, 1H),

8.00 (dd, J = 3.4 and 6.3 Hz, 1H); 13C NMR (75 MHz, CDCl3) δ (ppm) 13.91,

14.22, 22.35, 23.86, 33.42, 35.09, 35.68, 123.63, 124.63, 125.22, 127.53, 127.58,

128.08, 130.27, 134.02, 138.38, 139.84.

111

3-Butyl-2-ethyl-1-phenylnaphthalene (3). Yellow oil 1H NMR (300 MHz,

CDCl3) δ (ppm) 1.01 (t, J = 7.3 Hz, 3H), 1.02 (t, J = 7.5 Hz, 3H), 1.51 (m, 2H),

1.74 (m, 2H), 2.59 (q, J = 7.4 Hz, 2H), 2.84 (t, J = 7.8 Hz, 2H), 7.25-7.50 (m,

8H), 7.69 (s, 1H), 7.77 (d, J = 8.1 Hz, 1H); 13C NMR (75 MHz, CDCl3) δ (ppm)

14.02, 15.52, 22.90, 23.25, 32.78, 33.61, 124.7, 126.22, 126.41, 126.72, 126.91,

127.37, 127.77, 128.00, 128.40, 130.21, 131.40, 131.77, 138.30, 138.58, 139.12,

140.20.

1-Benzylisoquinoline (4) (100). This compound was prepared using a method

previously reported [29]. 1H NMR (300 MHz, CDCl3) δ (ppm) 4.72 (s, 2H), 7.26

(m, 5H), 7.62 (m, 3H), 7.84 (d, J = 6.0 Hz, 1H), 8.19 (d, J = 9.0 Hz, 1H), 8. 51 (d,

J = 3.0 Hz, 1H); 13C NMR (75 MHz, CDCl3) δ (ppm) 41.85, 119.84, 125.76,

126.19, 127.07, 127.27, 128.43, 128.50, 129.90, 136.51, 139.24, 141.64, 160.00.

The protocol describing the synthesis of compounds 6-10 is published elsewhere

(101, 102)

3,4-Diphenylisoquinoline (5) (125, 126). 1H NMR (300 MHz, CDCl3) δ (ppm)

7.20–7.23 (m, 3H), 7.24–7.27 (m, 2H), 7.36-7.40 (m, 5H), 7.62-7.65 (m, 2H),

7.68 (dd, J = 4.0 and J = 5.7 Hz, 1H), 8.06 (dd, J = 5.0 and 4.7 Hz, 1H) 9.40 (s,

1H); 13C NMR (75 MHz, CDCl3) δ (ppm) 125.64, 127.08, 127.09, 127.26, 127.43,

127.68, 127.80, 128.29, 131.17, 131.04, 131.22, 135.7, 137.13, 140.80, 150.89,

151.92.

112

4-Methyl-3-phenylisoquinoline (6) (126). 1H NMR (300 MHz, CDCl3) δ (ppm)

2.67 (s, 3H), 7.39–7.52 (m, 3H), 7.57–7.63 (m, 3H), 7.77 (ddd, J = 1.4, 6.8 and

8.4 Hz, 1H), 8.02 (d, J = 8.1 Hz, 1H), 8.08 (dd, J = 1 and 8.4 Hz, 1H), 9.22 (s,

1H); 13C NMR (75 MHz, CDCl3) δ (ppm) 15.46, 123.55, 124.35, 126.75,

127.04, 127.65, 128.07, 128.16, 129.78, 130.68, 136.29, 141.72, 150.45, 151.90.

4-Ethyl-3-phenylisoquinoline (7) (127). 1H NMR (300 MHz, CDCl3) δ (ppm)

1.31 (t, J = 7.5 Hz, 3H), 3.08 (q, J = 12.5 and 7.5 Hz, 2H), 7.43 (dt, J = 6.5 and

.7 Hz, 1H), 7.48 (t, J = 7.2 Hz, 2H), 7.53 (d, J = 8.0 Hz, 2H), 7.63 (ddd, J = 1.1,

6.9 and 8.1 Hz, 1H), 7.78 (ddd, J =1.4, 6.8 and 8.4 Hz, 1H), 8.03 (d, J = 8.1 Hz,

1H), 8.1 (d, J = 8.5 Hz, 1H), 9.23 (s, 1H); 13C NMR (75 MHz, CDCl3) δ (ppm)

15.52, 21.72, 123.56, 126.66, 127.62, 127.78, 127.94, 128.11, 128.42, 128.59,

128.81, 129.11, 130.09, 130.62, 135.23, 141.40, 149.73, 151.54.

3-Phenyl-4-(phenylethynyl)isoquinoline (8) (128). 1H NMR (300 MHz,

CDCl3) δ (ppm) 7.36-7.40 (m, 3H), 7.43-7.57 (m, 5H), 7.68 (ddd, J= 1.2, 6.9 and

8.1 Hz, 1H), 7.85 (ddd, J = 1.4, 6.9 and 8.3 Hz, 1H), 8.05 (dd, J = 0.9 and 8.1

Hz, 1H) 8.13-8.17 (m, 2H), 8.50 (dd, J = 1.0 and 8.4 Hz, 1H), 9.30 (d, J = 0.8,

1H); 13C NMR (75 MHz, CDCl3) δ (ppm) 85.61, 99.13, 112.54, 123.23, 125.62,

126.48, 127.52, 127.81, 128.35, 128.48, 128.57, 129.80, 131.26, 131.36, 139.76,

151.27, 154.22.

113

4-Phenyl-3-(phenylethynyl)isoquinoline (9) (128). 1H NMR (300 MHz,

CDCl3) δ (ppm) 7.25 (br s, 5H), 7.50-7.69 (m, 8H), 8.03-8.06 (m, 1H), 9.28 (s,

1H); 13C NMR (75 MHz, CDCl3) 89.41, 92.63, 122.57, 125.38, 127.42, 127.60,

127.76, 128.03, 128.08, 128.11, 128.42, 130.57, 130.96, 131.63, 134.92, 135.27,

136.17, 136.36, 151.88.

114

A.2: NMR Spectra

Fig

ure

A.1

300

MH

z 1 H

NM

R o

f co

mpo

und

1 in

CD

Cl 3

115

Fig

ure

A.2

75

MH

z 13

C N

MR

of

com

poun

d 1

in C

DC

l 3

116

Fig

ure

A.3

300

MH

z 1 H

NM

R o

f co

mpo

und

2 in

CD

Cl 3

117

Fig

ure

A.4

75

MH

z 13

C N

MR

of

com

poun

d 2

in C

DC

l 3

118

Fig

ure

A.5

300

MH

z 1 H

NM

R o

f co

mpo

und

3 in

CD

Cl 3

119

Fig

ure

A.6

75

MH

z 13

C N

MR

of

com

poun

d 3

in C

DC

l 3

120

Fig

ure

A.7

300

MH

z 1 H

NM

R o

f co

mpo

und

4 in

CD

Cl 3

121

F

igur

e A

.8 7

5 M

Hz

13C

NM

R o

f co

mpo

und

4 in

CD

Cl 3

122

F

igur

e A

.9 3

00 M

Hz

1 H N

MR

of

com

poun

d 5

in C

DC

l 3

123

F

igur

e A

.10

75 M

Hz

13C

NM

R o

f co

mpo

und

5 in

CD

Cl 3

124

F

igur

e A

.11

300

MH

z 1 H

NM

R o

f co

mpo

und

6 in

CD

Cl 3

125

Fig

ure

A.1

2 75

MH

z 13

C N

MR

of

com

poun

d 6

in C

DC

l 3

126

Fig

ure

A.1

3 30

0 M

Hz

1 H N

MR

of

com

poun

d 7

in C

DC

l 3

127

F

igur

e A

.14

75 M

Hz

13C

NM

R o

f co

mpo

und

7 in

CD

Cl 3

128

Fig

ure

A.1

5 30

0 M

Hz

1 H N

MR

of

com

poun

d 8

in C

DC

l 3

129

F

igur

e A

.16

75 M

Hz

13C

NM

R o

f co

mpo

und

8 in

CD

Cl 3

130

Fig

ure

A.1

7 30

0 M

Hz

1 H N

MR

of

com

poun

d 8

in C

DC

l 3

131

F

igur

e A

.18

75M

Hz

13C

NM

R o

f co

mpo

und

8 in

CD

Cl 3

132

133

A.3 Asphaltene Model Compounds’ Mass Spectral Data

Table A.1 Msn Spectra Measured for Asphaltene Model Compounds Ionized by (+) APCI/CS2 and Subjected To CAD

Analyte (MW) MS (m/z)

MS2 CAD fragments (m/z) [branching ratio]



3-butyl-1-propylnaphthalene (226)

226 [M•+]

226 – CH2CH3 (197) [6%]a

197 – CH2=CH2 (169) [7%]a

197 – CH2=CHCH3(155) [88%]a 197 – 56g (141) [5%]a

169 – CH3 (154)[13%]a 169 – CH2=CH2 (141) [87%]a 155 – 2H (153) [54%]b 155 – 40g,h (115) [46%]b

226 – CH2=CHCH3 (184)

[89%]a

184 – CH3 (169) [6%]a

184 – CH2CH3 (155) [38%]a 184 – CH2=CHCH3 (142) [29%]a 184 – CH2CH2CH3 (141) [27%]a

169 – CH3 (154) [5%]c 169 – CH2=CH2 (141) [95%]c 155 – 40g ,h (115) [100%]b 142 – H (141) [100%]a no fragmentation (141) [100%]d

226 – CH2CH2-CH3 (183) [5%]a

134

1-butyl-3-propylnaphthalene (226)

226 [M•+]

226 – CH2=CH2 (198) [3%]a 226 – CH2CH3 (197) [5%]a 226 – CH2=CHCH3(184) [38%]a

226 – CH2CH2CH3 (183) [48%]a

226 – 84 (142)

[2%]a

226 – 85 (141) [4%]a

197 – CH2=CH2

(169) [7%]c 197 – CH2=CHCH3

(155) [68%]c 197 – 56g (141) [25%]c 184 – CH3 (169) [12%]a 184 – CH2=CH2 (156) [7%]a

184 – CH2CH3 (155) [43%]a 184 – CH2=CHCH3

(142) [27%]a 184 – CH2CH2CH3 (141) [11%]a 183 – CH2=CH2 (155) [17%]a 183 – CH2=CHCH3

(141) [83%]a 141 – 26g (115) [100%]d

169 – CH2=CH2 (141) [100%]b 155 – H (154) [9%]b 155 – 2H (153) [79%]b 155 – 40g (115) [12%]b 169 – CH3 (154) [12%]c 169 – CH2=CH2 (141) [88%]c no fragmentation (155) [100%]b 155 – 26g (129) [64%]b 155 – 40g (115) [36%]b

135

3-butyl-2-ethyl-1-phenylnaphthalene (288)

1-benzylisoquinoline (219)

288 [M•+] 219 [M•+]

288 – CH2CH3 (259) [5%]c

288 – CH2=CHCH3(246) [67%]c 288 – CH2CH2CH3 (245) [9%]c 288 – 57g (231) [13%]c 288 – 71g (217) [6%]c 219 – H (218) [100%]c

259 – CH2=CH2 (231) [11%]a 259 – CH2CH3 (230) [4%]a 259 – CH2=CHCH3

(217) [80%]a 259 – Benzene (181) [5%]a

246 – CH3 (231) [74%]c 246 – CH2CH3 (217) [26%]c

231 – 2 H (229) [10%]c 231 – CH3 (216) [85%]c 231 – CH2=CH2 (203) [5%]c 217 – 2 H (215) [6%]b 217 – CH3 (202) [94%]b

218 – 76 + H2O (160) [49%]e 218 – 64 (154) [8%] e 218 – 76 (142) [26%]e 218 – benzene (140) [9 %]e 218 – 84 (134) [85]e

217 – 2H (215) [10%]c 217 – CH3 (202) [90%]c 216 – H (215) [100%]b

136

3,4-diphenylisoquinoline (281)

281 [M•+]

281 – H (280) [59%]e

281 – 2 H (279) [7%]e 281 – 3 H (278) [17%]e 281 – HCN (254) [3%]e 281 – H+HCN (253) [13%]e 281 – 2H+HCN or CH2CH3 (252) [1%]e

280 – H (279) [2%]f

280 – 2H (278) [15%]f 280 – CH3 (265) [5%]f 280 – 26g (254) [6%]f 280 – HCN (253) [30%]f 280 – H+HCN (252) [30%]f 280 – 60g (220) [5%]f 280 – Benzene (202) [7%]f 278 – H (277) [4%]e

278 – 9g (269) [26%]e 278 – 23g

(255) [5%]e 278 – 26g (252) [5%]e 278 – HCN (251) [52%]e 278 – H+HCN (250) [8%]e

137

4-methyl-3-phenylisoquinoline (219)

4-ethyl-3-phenylisoquinoline (233)

219 [M•+] 233 [M•+]

219 – H (218) [36%]b 219 – 2 H (217) [28%]b 219 – 3H (216) [29%]b 219 – H+HCN (191) [4%]b 219 – 2H+HCN or CH2CH3(190) [8%]b 233 – H (232) [56%]c 233 – CH3 (218) [13%]c 233 – 16g (217) [31%]c

218 – H (217) [36%]b,h 218 – 2 H (216) [28%]b,h 218 – 3 H (215) [7%]b,h 218 – HCN (191) [7%]b,h 218 – H+HCN (190) [14%]b,h 218 – 2H+HCN or CH2CH3 (189) [8%]b,h 191 – H (190) [26%]b,h 191 – 2 H (189) [60%]b,h 191 – 3 H (188) [5%]b,h 191 – 26g (165) [4%]b,h 191 – HCN (164) [5%]b,h 232 – CH3 (217) [100%]c

217 – H (216) [35%]b,h 217 – 2 H (215) [9%]b,h 217 – 3 H (214) [7%]b,h 217 – 26g (191) [5%]b,h 217 – HCN (190) [19%]b,h 217 – H+HCN(189) [25%]b,h

190 – H (189) [45%]b,h

190 – 2 H (188) [555]b,h

138

3-phenyl-4-(phenylethynyl)isoquinoline (305.12)

305 [M•+]

305 – H (304) [52%]f 305 – 14g (291) [4%]f 305 –15 (CH3) (290) [24%]f 305 – 77 (228) [20%]f

304 – H (303) [23%]b,h 304 – 2 H (302) [31%]b,h 304 – 3 H (301) [18%]b,h 304 – HCN (277) [6%]b,h 304 – CH2=CH2 (276) [22%]b,h 290 – 2 H (288) [73%]b,h 290 – 3 H (287) [4%]d,h 290 – 26g (264) [4%]d,h 290 – HCN (263) [14%]d,h 290 – CH2=CH2 (262) [5%]d,h 228 – H (227) [49%]b,h 228 – HCN (201) [39%]b,h 228 – CH2=CH2 (200) [12%]b,h

276 – HCN (249) [100%]b,h 288 – 2 H (286) [67%]d,h 288 – HCN (261) [29%]d,h 288 – CH2=CH2 (260) [4%]d,h 263 – 2 H (261) [37%]d,h 263 – 26g (237) [63%]d,h

139

Collision energies: 20a, 30b, 25c, 35d, 40e, 45f, gpossible ring opening fragments or unusual loss, hActivation Q = 0.5, ibranching ratios only given for fragments with a relative intensity 5 % or higher relative to the base peak.

4-phenyl-3-(phenylethynyl)isoquinoline (305.12)

305 [M•+]

305 – H (304) [74%]f 305 – 15 (CH3) (290) [21%]f 305 – HCN (278) [5%]f

304 – H (303) [3%]d,h 304 – 2 H (302) [36%]d,h 304 – 3 H (301) [37%]d,h 304 – HCN (277) [4%]d,h 304 – CH2=CH2 (276) [15%]d,h 304 – CH2-CH3 (275) [5%]d,h 278 – H (277) [25%]b,h 278 – 2 H (276) [55%]b,h 278 – 3 H (275) [5%]b,h 278 – CH3 (263) [7%]b,h 278 – 26g (252) [3%]b,h 278 – 58g (220) [5%]b,h

276 – H (275) [32%]d,h 276 – 2 H (274) [68%]d,h

140

Appendix B: Supplemental Data for Chapter 6

B.1 Experimental Procedures, 1H and 13C NMR Characterizations


purification. 1H NMR spectra were recorded at 300 MHz while 75 MHz was used for 13 C

nuclei. 230-400 mesh silica gel was used for purification by flash chromatography.

Mosher acids (0.0 Hz) were used as external references for the 19F NMR experiments.

4-Aminobenzaldehyde. 4-Nitrobenzaldehyde (100 mg, 0.66 mmol) and

SnCl2•2H2O (746.5 mg, 3.31 mmol) in 1.32 mL of absolute ethanol were stirred at

70 oC for three hours. Upon consumption of the starting material (observed by

TLC), the reaction mixture was cooled down, poured into ice and made slightly

basic (pH 7-8) by addition of approximately 15 mL of 5% NaHCO3. The mixture

was then extracted 3X with 20 mL portions of ethyl acetate. The combined

organic layers were then washed with brine, treated with decolorizing carbon,

filtered and dried with sodium sulfate. The solvent was removed upon rotary

evaporation to yield a dark orange oil (82 mg, quantitative yield) which was used

right away in the next step without further purification. 1H NMR (300 MHz,

CDCl3) δ (ppm) 3.73 (2H, br s), 6.69 (2H, d, J = 9.0 Hz), 8.21 (2H, d, J = 9.0 Hz),

9.73 (1H, s).

141

4-Azidobenzaldehyde (21). To a solution of glacial acetic acid (0.7 mL, 12.2

mmol) and sulfuric acid (70 µL, .7 mmol) stirring at 0 oC were added 4-

aminobenzaldehyde (160.32 mg, 1.32 mmol) yielding a black tar. Sodium nitrite

(90.4 mg, 1.32 mmol) dissolved in a minimum amount of cold water was added to

the reaction mixture. 10 mins later, sodium azide (90.3 mg, 1.39 mmol) dissolved

in a minimum amount of cold water was added to the reaction producing a

considerable amount of froth. The reaction mixture was then allowed to warm up

to room temperature and extracted 3X with 10 mL portions of ether. The

combined organic layers were then washed with saturated aqueous sodium

bicarbonate, dried with sodium sulfate followed by solvent evaporation under

minimum light exposure at less than 40 oC. Upon purification by column

chromatography (3:1 hexane/ethyl acetate), the desired product was obtained as a

dark yellow solid (181 mg, 93%). 1H NMR (300 MHz, CDCl3) δ (ppm) 7.18 (2H,

d, J = 9.0 Hz), 7.90 (2H, d, J = 9.0 Hz), 9.94 (1H, s).

4-Azido Methylcinnamate (24). To a stirred solution of 4-azidobenzaldehyde

(142.3 mg, 0.9 mmol) in 4.2 mL of a 1:1 mixture of H2O/dioxane were added

trimethylphosphonoacetate (209 µL, 1.45 mmol) and potassium carbonate (401

mg, 2.90 mmol). The mixture was stirred to completion, diluted with water and

extracted with 3X 10 mL portions of ethyl acetate. The combined organic layers

were washed 3X with saturated sodium bicarbonate, once with water and brine

and dried with sodium sulfate. Evaporation of the solvent yielded the desired

product as a light yellow solid (147.3 mg, 58%) which was taken to the next step

142

without purification. 1H NMR (300MHz, CDCl3) δ (ppm) 3.80 (3H, s), 6.41 (1H,

d, J = 15 Hz), 7.05 (2H, d, J = 9.0 Hz), 7.53 (2H, d, J = 9.0 Hz), 7.67 (1H, d, J =

15 Hz); IR 2118.5 cm-1.

(2R,3R)-3-(4-azidophenyl)-3-methoxy-2-(4-nitrophenylsulonamido) methyl

Propanoate (25) To copper triflate (4.6 mg, 0.012 mmol) were added (-)-2,2’-

isopropylidene[(4S)-4-phenyl-2-oxazoline] (PhINNs) (8.5 mg, 0.12 mmol), 4 Å

molecular sieves and anhydrous dichloromethane (1.3 mL). The reaction mixture

was allowed to stir for 15 minutes under nitrogen during which it turned green.

Then, 4-azido methylcinnamate (150 mg, 0.74 mmol) and [N-

(phenylsulfonyl)imido]phenyliodinane (85.3mg, 0.21 mmol) were added followed

by nitrogen flushing to rid the reaction flask of any moisture. It took about 11-13

hours of stirring for PhINNs to dissolve. The reaction was stopped by quick

filtration through a small pad of silica gel (eluent, 1:1 hexane/ethyl acetate); the

solvent was evaporated and the residue was redissolved in methanol and stirred

overnight. Concentration under reduced pressure followed by purification by flash

column chromatography (from 3:1 to 2:1 hexane/ethyl acetate) gave the desired

product as a yellow oil (91.8 mg, 77%). 1H NMR (300 MHz, CDCl3) δ (ppm)

3.20 (3H, s), 3.51 (3H, s), 4.33 (1H, dd, J = 3.0, 6.0Hz), 4.49 (1H, d, J = 6.0 Hz),

5.57 (1H, d, J = 9.0 Hz), 6.96 (2H, d, J = 9.0 Hz), 7.13 (2H, d, J = 9.0 Hz), 7.95

(2H, d, J = 9.0 Hz), 8.29 (2H, d, J = 9.0 Hz). 13C NMR(75 MHz, CDCl3) δ (ppm)

52.4, 57.3, 60.5, 82.8, 119.1, 123.9, 128.2, 128.3, 132.2, 140.6, 145.6, 149.8,

169.0; ESI-MS (m/z) C17H17N5O7S 437.02 (M++1); IR 2123.4 cm-1.

143

In order to separate the two diastereomers, we followed the same procedure but

changed the solvent to acetonitrile and when the residue was redissolved in MeOH, we

added a couple of drops of sulfuric acid. Minor diastereomer, 1H NMR (300 MHz,

CDCl3) δ (ppm) 3.21 (3H, s), 3.72 (3H,s), 4.13 (1H, dd, J = 3.0, 3.0 Hz), 4.72 (1H, d, J =

3.0 Hz), 5.49 (1H, d, J = 12 Hz), 6.88 (2H, d, J = 9.0 Hz), 7.18 (2H, d, J = 9.0 Hz), 7.73

(2H, d, J = 9.0 Hz), 8.19 (2H, d, J = 9.0 Hz).

(2R,3R)-Methyl 2-Amino-3-(4-azidophenyl)-3-methoxypropanoate (26). To a

stirred solution of (2R,3R)-3-(4-azidophenyl)-3-methoxy-2-(4-

nitrophenylsulonamidomethyl propanoate (16) (20 mg, 0.05 mol) in 0.4 mL of

MeOH/DMSO (49:1) were added thiophenol (23.3 µL, .2 mmol) and potassium

carbonate (25.1 mg, .2 mmol). The reaction mixture was stirred at 50 oC for two

hours. After cooling down, the reaction was quenched with 3 mL of saturated

ammonium chloride and extracted 3X with 5 mL portions of dichloromethane.

The organic phase was dried in anhydrous magnesium sulfate. After evaporation

and silica gel column chromatography (2:1 hexane/ ethyl acetate to 100% ethyl

acetate), the desired product was obtained as a brown oil (3 mg, 26%). 1H NMR

(300 MHz, CDCl3) δ (ppm) 3.25 (3H, s), 3.72 (3H, s) 3.77 (1H, d, J = 6Hz), 4.39

(1H, d, J = 6.0 Hz), 7.05 (2H, d, J = 9.0 Hz), 7.30 (2H, d, J = 9.0 Hz). 13C NMR

(75 MHz, CDCl3) δ (ppm) 51.8, 56.6, 59.9, 85.1, 114.9, 126.6, 128.5, 146.5,

173.9; IR 2122.3 cm-1.

144

(2R,3R)-methyl 3-(4-azidophenyl)-3-methoxy-2-((S)-3,3,3-trifluoro-2-

methoxy-2-phenylpropanamido)propanoate (33). To a solution of (S)-MTPA

(3 mg, 0.013 mmol) and DMF (1.1 µL, 0.013 mmol) in hexane (0.5 mL) at room

temperature was added oxalyl chloride (5µL, 0.057 mmol). A white precipitate

formed spontaneously. The mixture was allowed to stir for 1 hour followed by

vacuum filtration and concentration of the filtrate under reduced pressure yielded

the acid chloride. Compound 26 (3 mg, 0.012 mmol), triethylamine (4 µL, 0.03

mmol) and a crystal of 4- dimethylaminopyridine (~ 1 mg) in 100 µL of CDCl3

were added to the acid chloride and allowed to react for 1 hour and the

completion of the reaction was confirmed by 1H NMR (124). 1H NMR (300 MHz,

CDCl3) δ (ppm) 3.34 (6H, br s), 3.64 (3H, s) 4.65 (1H, d, J = 3Hz), 4.94 (1H, dd,

J = 9.0 and 9.0 Hz), 7.03 (2H, d, J = 9.0 Hz), 7.31 (2H, d, J = 9.0 Hz), 7.39 (m,

3H), 7.49 (m, 2H).

145

B.2: NMR Spectra

Fig

ure

B.1

300

MH

z 1 H

NM

R o

f co

mpo

und

24 in

CD

Cl 3

146

F

igur

e B

.2 3

00 M

Hz

1 H N

MR

of

com

poun

d 25

in C

DC

l 3

147

F

igur

e B

.3 7

5 M

Hz

13C

NM

R o

f co

mpo

und

25 in

CD

Cl 3

N3

CO

2Me

NH

Ns

OM

e

148

F

igur

e B

.4 3

00 M

Hz

1 H N

MR

of

com

poun

d 26

in C

DC

l 3

Acet

one

impu

rity

149

F

igur

e B

.5 7

5 M

Hz

13C

NM

R o

f co

mpo

und

25 in

CD

Cl 3

150

F

igur

e B

.6 3

00 M

Hz

1 H N

MR

of

com

poun

d 33

in C

DC

l 3

151

F

igur

e B

.7 2

82 M

Hz

19F

NM

R o

f co

mpo

und

33 in

CD

Cl 3

152

F

igur

e B

.8 3

00 M

Hz

1 H N

MR

of

enan

tiom

er o

f co

mpo

und

33 in

CD

Cl 3

153

F

igur

e B

.9 2

82 M

Hz

19F

NM

R o

f en

anti

omer

of

com

poun

d 33

in C

DC

l 3

154

VITA

155

VITA

Nadine Njoya was born in Cameroon, West Africa. After High school, she

attended the University of Buea where she obtained her Bachelor in chemistry with a

minor in material sciences. In 2006, she joined East Tennessee State University where,

under the supervision of Dr. Joseph Young, obtained a Masters in Chemistry working on

expansion-contraction protocols. At Purdue, she was working in Professor Mary Wirth

research laboratory with a focus on protein separations by sieving electrophoresis. Upon

completing her Ph.D. program, Nadine hopes to work as a research scientist in a

biopharmaceutical company.

PUBLICATION

Research ArticleTheme: Aggregation and Interactions of Therapeutic ProteinsGuest Editor: Craig Svensson

Silica Colloidal Crystals as Emerging Materials for High-Throughput ProteinElectrophoresis

Nadine K. Njoya,1 Robert E. Birdsall,1 and Mary J. Wirth1,2

Received 16 February 2013; accepted 30 May 2013; published online 26 June 2013

Abstract. Silica colloidal crystals are a new type of media for protein electrophoresis, and they areassessed for their promise in rapidly measuring aggregation of monoclonal antibodies. The nature ofsilica colloidal crystals is described in the context of the need for a high-throughput separation tool foroptimizing the formulations of protein drugs for minimal aggregation. The fundamental relationsbetween molecular weight and mobility in electrophoresis are used to make a theoretical comparison ofselectivity between gels and colloidal crystals. The results show that the selectivity is similar for thesemedia, but slightly higher, 10%, for gels, and the velocity is inherently lower than that for gels due to thesmaller free volume fraction. These factors are more than compensated for by lower broadening incolloidal crystals. These new media give plate heights of only 0.15 μm for the antibody monomer and0.42 μm for the antibody dimer. The monoclonal antibody is separated from its dimer in 72 s over adistance of only 6.5 mm. This is five times faster than size-exclusion chromatography, with more thantenfold miniaturization, and amenable to parallel separations, all of which are promising for the design ofhigh-throughput devices for optimizing protein drug formulations.

KEY WORDS: colloidal crystal; electrophoresis; high throughput; miniaturization; monoclonal antibody;plate height.

INTRODUCTION

The growth of biotechnology in the pharmaceuticalindustry has spurred new interest among analytical chemistsfor faster characterization of protein heterogeneity. Of thevarious undesired forms that an engineered protein canadopt, aggregation is of particular concern because aggrega-tion occurs spontaneously in the formulated drug, causingimmunogenic response in patients (1). Minimizing formationof aggregates is an essential step in formulation of proteindrugs, which presently requires extensive trial and error toidentify a matrix that minimizes aggregation over the timeperiod of storage. Current techniques for characterizingprotein aggregation in trial formulations have low through-put, thus creating a demand for new technology that offershigher throughput.

Separations are presently the gold standard for charac-terizing aggregation because the phenomenon of aggregationnecessarily results in a mixture of proteins, and a physicalseparation facilitates quantitation of each component. Size-exclusion chromatography is widely used for such separa-tions, but it is lower in throughput than electrophoresisbecause the latter allows multiple samples to be run in

parallel. The material that has dominated electrophoresisover the past half century is polyacrylamide gel. The reasonwhy this technique is still low in throughput is that theseparations themselves are slow due to the slow velocities ofthe protein zones and the fact that the length needs to belarge to make up for zone broadening.

We are taking a new approach toward speeding upprotein separations, which is to decrease the separationlength drastically by reducing the broadening processes inthe medium. The separations would then be faster becausethe length is shorter. The idea is simple: if the mediumimparts little broadening to the zones, then the physicalseparation arising from the velocity differences would beapparent much earlier. The idea is illustrated in Fig. 1, whichuses plate height as the figure of merit for broadening by theseparation medium. The plate height is the amount of zonevariance, σ2, per separation distance, L.

H ¼ σ2

Lð1Þ

The graphs in Fig. 1 illustrate the essential concept behindthis work: two zones separate over a tenfold shorter length whenthe plate height is tenfold smaller. The design of a material witha smaller plate height would thus give two advantages forthroughput, one being a tenfold shorter separation time,

1 Department of Chemistry, Purdue University, 560 Oval Drive, WestLafayette, Indiana 49707, USA.

2 To whom correspondence should be addressed. (e-mail:[email protected])

The AAPS Journal, Vol. 15, No. 4, October 2013 (# 2013)DOI: 10.1208/s12248-013-9506-2

9621550-7416/13/0400-0962/0 # 2013 American Association of Pharmaceutical Scientists

156

assuming that the velocity is not changed by the material, andthe other being a tenfold more compact separation device,which would facilitate the use of microplate robotics for sampledispensing.

There are multiple factors that give rise to zone broadeningduring electrophoresis. Themain cause is the heterogeneity of themedium. In an electrophoresis gel, the extent of cross-linking isnot perfectly uniform on the microscopic scale, and the resultantvelocity distribution creates a broader zone. Another factor is thedistribution of velocities resulting from variations in viscosity dueto Joule heating from the ionic current through themedium. Jouleheating limits the voltage that can be applied, which slows theseparation. Diffusion of proteins during staining and imaging alsocontribute to broadening and add time to the analysis.

In genomic DNA sequencing, capillary gel electrophoresissignificantly increased throughput compared to slab gels (2).Soluble gels are used with this technique, where pores are createdby the use of a concentrated uncharged polymer solution. Effectivepore size is controlled by polymer concentration. Gel-filledcapillaries inherently avoid the problem of gel heterogeneitybecause the polymer is in a homogeneous solution. An additionaladvantage is the ability to electromigrate analytes past a UVdetector for quantitative detection and automation. Furthermore,higher electricfields increase speedbecause capillaries canbemadetenfold thinner than gels, e.g., 50 μm i.d. vs. 0.5 mm in thickness,which dramatically reduces zone broadening compared to slab gels(3). These factors, combinedwith the ability to runmany capillariesin parallel, enabled capillaries to replace slab gels in the firstsequencing of the human genome (2). This type of advance thatincreased throughput in DNA sequencing needs to be made inprotein sizing.

One might think the problem for rapidly sizing proteinswould be solved by adopting capillary gel electrophoresis, yetcapillaries are not widely used for protein separations. Thereason is that another factor besides plate height affectsresolution, which is indicated in the second term of Eq. 2.

Rs ¼ffiffiffiffiffiffiffiffiffiffiffiL=H

q4

⋅Δvv

ð2Þ

Resolution, Rs, is proportional to the selectivity, Δv/v,where v is the zone velocity. This equation applies to anytype of separation, showing that both zone sharpness,described by the first term, and selectivity combine to giveresolution, and the two effects are multiplicative. Thelimitation of capillary gel electrophoresis is that theselectivity is much lower because the pores are large.The lower selectivity has to be compensated by longerlengths to give the same resolution. The net effect is anincrease by only about a factor of two in overall speed forcapillary gel electrophoresis of proteins (4–6). Selectivityneeds to be combined with a thin, homogeneous mediumfor faster resolution.

Nanotechnology has advanced over the same time periodas biotechnology; hence, it is an auspicious time to revisit thedesign and development of porous media for proteinseparations. To circumvent heterogeneity of the porousmedium for electrophoresis, we use silica colloidal crystals(7). These materials are comprised of uniformly sized silicaparticles (8), which can be anywhere between a fewnanometers to 1,000 nm in diameter. The silica spheres arearranged to form crystals with a face-center cubic arrange-ment (9–11). Silica colloidal crystals have been demonstratedas media for DNA (12,13) and protein electrophoresis(12,14).

Figure 2a shows an SEM image of a silica colloidalcrystal formed inside of a microchannel by 470-nm silicaparticles. The image shows the expected crystal faces of theface-centered cubic lattice. Such high-quality colloidal crystalshave been made over wide areas (13,15). The orderrepresented in Fig. 2a extends the 2-cm length of the channel

Fig. 1. Simulated plots of Gaussian functions vs. separation length to illustrate effect ofplate height, H, on resolution. a For a plate height of 10 μm, the two Gaussians areunresolved in 1 cm of separation length, but they are baseline resolved in 10 cm. b For aplate height of 1 μm, the Gaussians are baseline resolved in only 1 cm

963Silica Colloidal Crystals as Emerging Materials

157

in the device shown in Fig. 2c. Figure 2b shows an SEM imageof the top surface of the colloidal crystal, which has thecharacteristic hexagonal geometry of the [111] crystal face.Pores for sieving exist everywhere where particles meet, andone can see from the expanded view in the inset of Fig. 2bthat these pores are tens of nanometers, which are suitablefor sieving or the size exclusion of proteins. Colloidal crystalcan also be formed in capillaries (17–20), and Fig. 2d shows aphotograph of a capillary packed with a silica colloidal crystal,along with an SEM image of the same capillary. Theopalescence owes to the lattice spacing being on the orderof the wavelength of light. Gem opals are also comprised ofcolloidal silica in face-centered cubic lattices, giving similarSEM images (21). The color is shifted when the index ofrefraction of the medium changes, and this so-called photoniceffect is a means of detecting biomolecules without labeling(22), which might be useful in optimizing formulations.

The purpose of this work is to model an electrophoreticseparation of a monoclonal antibody and its dimer using dataand equations from the literature, and to carry out such aseparation using a silica colloidal crystal for which theparticle diameter is selected based on the modeling. Sodiumdodecyl sulfate (SDS) denaturing is used to impart a chargeon the monoclonal antibodies, which is necessary becausethese proteins typically have pI values near neutral pH.Covalent aggregation is ubiquitous for monoclonal antibod-ies due to scrambling of disulfide bonds. The technology

would be applicable to characterizing covalent aggregatessince SDS breaks up noncovalent aggregates. The plateheight will be measured to assess speed and miniaturization.

METHODS

A sample of pharmaceutical-grade monoclonal antibody,obtained from Eli Lilly, had gone through drug developmentbut is no longer an active candidate. The antibody waslabeled with AlexaFluor 546 (Invitrogen, Carlsbad, CA)according to the vendor’s instructions. The antibody hadbeen given a 10-min heat treatment at 75°C to promoteaggregate formation, and size-exclusion chromatographyconfirmed that a small amount of dimer was formed. Aconcentration of 0.5 mg/mL was prepared in SDS andelectrophoresis buffer, as described earlier (12,14).

Uniform nonporous silica particles were obtained fromFiber Optic Center (New Bedford, MA); these were calcinedand annealed as described previously (14). The silica surfaceswere chemically modified to bear linear polyacrylamidechains, which were grown from the surface by using surface-initiated atom-transfer radical transfer polymerization (23).The material thus mimics a polyacrylamide gel, but is thinnerand more homogeneous. The details of how to prepare thesilica colloidal crystal in capillaries have been published (16),as well as the details of how to perform electrophoresis (14).

Fig. 2. Formats of colloidal crystals used for electrophoresis. a SEM image of channel cross sectionshowing high crystallinity. Image from Birdsall and Wirth, permission requested, © 2013 Wiley-VCHVerlag GmbH & Co. KGaA, Weinheim. b SEM image of the top of the colloidal crystal in the channel,showing ordered arrangement of silica spheres. The expanded view shows the pores formed among thespheres. c Photograph of a set of parallel channels, which are 2 cm in length. d Photograph of the end of acapillary packed with colloidal crystals, with SEM image superimposed to show polycrystalline packing,taken from Wei et al. (16) with permission, Copyright © 2012, American Chemical Society

964 Njoya et al.

158

RESULTS AND DISCUSSION

Electrophoresis results reported recently by our group (14)are analyzed further here to understand how these materials canbe used to study aggregation of monoclonal antibodies. Figure 3ashows an electropherogram, where a silica colloidal crystal madeof a particle 483 nm in diameter was used for electrophoresis oflysozyme (14.7 kDa), trypsin inhibitor (20 kDa), carbonicanhydrase (29 kDa), and bovine serum albumin (67 kDa). Theelectropherogram shows narrow zones as the proteinselectromigrate past a fixed detection point, which was only4 mm from the injection point. The zones broaden as proteinsize increases. This is typical behavior for electrophoresis, whereplate height increases with protein radius. The plate heights werenot calculated in the original work, but are essential in the contextof the present work, and thus are calculated here: the plateheights increase in the order 0.33, 1.0, 1.6, and 3.3 μm for thesmallest to the largest protein. These are all smaller than thetypical plate height of 10 μm for medium-sized proteins in gels(24). The unusually small plate heights demonstrate that the silicacolloidal crystals are promising for miniaturization. Figure 3bshows a plot of molecular weight vs. normalized migration ratethrough the medium. The shape of the curve is similar to suchcurves for gel electrophoresis, where the mobility goes to zerowhen themolecular weight exceeds the pore size, and the velocityis no longer dependent on molecular weight when the proteinradius becomes negligible in size compared to the pore radius.The solid curve is a plot of the predicted migration rate usingEq. 3, which was derived from first principles (14), and eachparameter was independently characterized.

μμ0

¼ εξ−Rξ

2

ð3Þ

Themobility, μ, in the medium is normalized to themobilityin free solution, μ0, which was measured by open capillaryelectrophoresis. The pore radius, ξ, and porosity, ε, wereindependently determined from flow rate measurements. Theprotein radius is presented byR. Since velocity is proportional tomobility, then Δμ/<μ>, is equivalent to the ratio Δv/<v> used in

Eq. 2 for expressing selectivity. The important outcome is that wecan use Eq. 3 to predict the selectivity, Δμ/<μ>, for otherproteins, i.e., monoclonal antibodies and aggregates, and otherpore radii that are more optimal for the larger proteins.

For comparison of selectivity with gels, the theoreticaltreatment by Kopecka et al. (25) provides an equation todescribe gel electrophoresis.

μμ0

¼ exp −π4

Rþ rξ

2 !

ð4Þ

Fig. 3. Electrophoresis data for four proteins using a silica colloidal crystal as the medium. Proteins A–Dare lysozyme, trypsin inhibitor, carbonic anhydrase, and bovine serum albumin, respectively. aElectropherogram of the proteins, with data points (circles), and a fit to four Gaussians (line). b Plot ofelectrophoresis data using Eq. 3. Pore radius is indicated by ξ. Both panels are adapted from Birdsall et al.(14) with permission requested, © 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Fig. 4. Gel electropherogram of a monoclonal antibody initially (lane 4)and after storage (lanes 5–9). The size standards are shown in lane 2. Thefigure is adapted from Mahler et al. (26) with permission requested, ©2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim


159

In this equation, r is the radius of the gel fibers, e.g., thepolyacrylamide chains, and the other terms have the samemeanings as for Eq. 3. To demonstrate that Eq. 4 is valid, datafrom the literature are used. Figure 4 shows a separation of amonoclonal antibody and its aggregates and fragments by gelelectrophoresis, published by Mahler et al. (26). The imageillustrates the benefit of gel electrophoresis: a wide range ofmolecular weights is separated, and many lanes can be runsimultaneously. The first lane, which has the size standards,allows one to identify the antibody, which is the most intense

band in the subsequent lanes. In the subsequent lanes, theaggregates of the antibody are visible at higher molecularweight, and fragments are visible at lower molecular weights.The data for the size standards are analyzed here by Eq. 4,and the results are plotted in Fig. 5a. The data in red are forthe gel, and the curve is from Eq. 4, with r and ξ adjusted forbest fit.

Figure 5a also uses Eq. 3 for colloidal crystals to make acomparison of selectivity, which is the blue curve. The circlesshow the theoretical positions of the same size standards for

Fig. 5. Comparison of selectivity for gels and colloidal crystals in protein electrophoresis. aData points (red circles) for the size standards of the gel in Fig. 4 and fit to Eq. 4 (red line).Plot of Eq. 3 (blue line, blue circles). b Theoretical plots for gel (red line) and colloidalcrystal (blue line) for pore sizes optimized for resolving an antibody from its dimer. In bothpanels, the antibody molecular weight (MW) (150 kDa) and its dimer MW (300 kDa) areindicated by the dotted lines

966 Njoya et al.

160

comparison. The mobility was divided by porosity to makethe x-axes coincide. The graph reveals that the selectivity issimilar for the gel and the colloidal crystal, with selectivitybeing somewhat higher for the gel.

The two equations allow a theoretical comparisonbetween a gel and a colloidal crystal when each have a poresize better suited for resolving a monoclonal antibody from itsdimer with higher velocity. The resulting plot is in Fig. 5b,where the pore sizes were adjusted to give the same velocityfor the dimer. The comparison shows that the gel has a 10%higher selectivity. This is a small difference, but it seemssurprising that the disordered gel would give higher selectiv-ity. The underlying reason is that the pore size distributionenables the gel to separate a wider range of proteins, whichtranslates to a smaller slope in log(MW) vs. mobility. Theanalysis shows that the colloidal crystal is a reasonablesubstitution for the gel from the standpoint of selectivity.

Since the goal is high throughput, the velocities must alsobe considered. In Fig. 5, the plots were made with respect to

“relative mobility,” which is the normalized mobility dividedby the porosity. The porosity is not known for the gel, but it islikely close to unity since the polyacrylamide fibers probablytake up much less of a negligible volume fraction of the gel.In fact, the best fit of the data for the size standards gave afiber radius of zero. For the colloidal crystal, the porosity is0.2, and the porosity is inherently low because the silicaparticles take up most of the volume. What this means inpractice is that the absolute velocity in the colloidal crystalwill be about fivefold lower than that of a gel. The colloidalcrystal thus has two problems to overcome: somewhat smallerselectivity and a much lower velocity. These limitations canpotentially be overcome by a much smaller plate height.

In choosing the particle diameter for the colloidal crystal,one must consider the trade-off between speed and selectivity.To elaborate, Eq. 3 shows that selectivity is enhanced by havingthe pore size closer to the protein size, but the equation alsoshows that the electrophoretic mobility is slowed as the pore sizeapproaches the protein size. Equation 3 can be used to examine

Fig. 6. a Plot of both selectivity (solid line) and relative mobility (dotted line) as a functionof pore radius to show the trade-off. b Plot of plate height vs. the square of the broadeningfunction of Eq. 5 for a pore radius of 27 nm. The line is least-square fits of data points


161

the trade-off between speed and selectivity to select an optimalpore size. A plot of both selectivity and speed with respect topore size was made by use of Eq. 3 and is given in Fig. 6a. Theplot reveals that selectivity drops sublinearly whereas speedgoes up nearly linearly until the pore radius exceeds 30 nm.These considerations would suggest a pore radius of 30 nm to bea good compromise.

Zone broadening is also a consideration in selecting poresize since the plate height increases strongly with protein sizefor a given pore size. Such behavior indicates that there is apore size distribution. One can model the dependence ofplate height on protein radius by introducing a stochasticvariable, ξ′, to account for random changes in pore radius,and relating this to changes in μ. Adding ξ′ to ξ in Eq. 3 andthen differentiating mobility with respect to ξ′ gives a simplerelation when ξ>>R>>ξ′.

∂ μ=μ0

∂ξ0

¼ ε2R

ξ2

ð5Þ

Since this variation in mobility with pore size would beproportional to zone width, then its square would beproportional to plate height. If the pore size distribution isthe origin of the broadening, and the pore size variations aresmall, then plate height should be proportional to R2. Such aplot is given in Fig. 6b, where the least squares fit shows goodagreement, and the intercept is zero. Extrapolating to highermolecular weights gives calculated plate heights of 9, 20, and32 μm for monomer, dimer, and trimer of a monoclonalantibody. For detecting aggregation of antibodies, largerpores are needed.

We chose 1,000 nm as the particle diameter, which istwice that of the 500-nm particles that gave the pore radius of27 nm. One cannot assume the new pore radius is also twiceas large because the polymer chains grow longer in largerpores. We perform electrophoresis to determine whethersmall plate heights are now observable for monoclonalantibodies and aggregates. Figure 7a shows an image fromthe separation of a monoclonal antibody and its dimer. Thebrighter monomer peak is separated from the dimer peakover a length of only 6.5 mm and in a time of only 72 s. Basedon Gaussian fits of the two zones, the calculated plate heightsare 0.15 and 0.42 μm for the monomer and dimer, respec-tively. These extremely low plate heights are extraordinaryfor monoclonal antibodies in any type of separation. Widerpore gels have been used to separate monomers from dimersof monoclonal antibodies, but the broadening is severe (27).Figure 7b shows a plot of the packed capillary electrophoresisdata of Fig. 7a, where a fixed detection position of 0.65 cmwas used and the intensity was monitored vs. time to generatean electropherogram. For comparison, Fig. 7c shows achromatogram from size-exclusion chromatography for thesame sample. The resolution is similar, slightly higher forelectrophoresis, but the time scale for electrophoresis isfivefold shorter. Both separations show primarily the mono-mer and dimer peaks. There is evidence for a large aggregatein the electrophoresis data. Both show evidence of specieswith smaller molecular weight, but more precisely controlledsamples will need to be studied as well as a variety of particlesizes, to learn what additional information is obtainable. The

fivefold higher speed in separating monomer from dimer,combined with the present commercial capability of running24 capillaries simultaneously in electrophoresis, promiseshigher throughput for assessing aggregation of monoclonalantibodies in the optimization of formulations.

Fig. 7. Separation of monoclonal antibody monomer and dimer. aFalse-color fluorescence image of antibody followed by its dimer as theyelectromigrate in SDS and buffer through a 100-μm-i.d. capillary packedwith silica colloidal crystal and reach a separation length of 6.5 mm in72 s. b Time trace of the same separation as in panel a. c Size-exclusionchromatogram of the same protein sample as in panels a and b

968 Njoya et al.

162

CONCLUSIONS

Silica colloidal crystals are an emerging material for fastseparations of monoclonal antibodies and aggregates. Theuniformity of the material allows a much smaller footprint anda higher speed. With the tiny separation lengths of less than1 cm, the media could be patterned for direct compatibility withthe robotics of 96-well plates for high throughput. The resultspresented here use SDS-denatured protein, which assessescovalent aggregates, which are arguably more deleterious thannoncovalent aggregates. Further studies are needed to addressnative electrophoresis for assessing noncovalent aggregates.The silica colloidal crystal makes it easy to control pore size forsieving large proteins with minimal broadening. Larger particlescould be used, in principle, to assess even larger aggregates. Theremaining issues for reduction to practice are the design of suchdevices and label-free detection.

ACKNOWLEDGMENTS

This work was supported by Eli Lilly, Inc. (NNJ) and byNIH under grant R21CA161772 (REB).

REFERENCES

1. An Z. Therapeutic monoclonal antibodies from bench to clinic.Hoboken: Wiley; 2009.

2. Dovichi NJ, Zhang JZ. How capillary electrophoresis sequencedthe human genome. Angew Chem-Int Edit. 2000;39(24):4463–8.

3. Jorgenson JW, Lukacs KD. Zone electrophoresis in open-tubularglass-capillaries. Anal Chem. 1981;53(8):1298–302.

4. Gomis DBL, Junco S, Exposito Y, Gutierrez D. Size-basedseparations of proteins by capillary electrophoresis using linearpolyacrylamide as a sieving medium: model studies and analysisof cider proteins. Electrophoresis. 2003;24(9):1391–6.

5. Guttman A, Nolan J. Comparison of the separation of proteins bysodium dodecyl-sulfate slab gel-electrophoresis and capillary sodiumdodecyl-sulfate gel-electrophoresis. Anal Biochem. 1994;221(2):285–9.

6. Zhu ZF, Lu JJ, Liu SR. Protein separation by capillary gelelectrophoresis: a review. Anal Chim Acta. 2012;709:21–31.

7. Colvin VL. From opals to optics: colloidal photonic crystals.MRS Bull. 2001;26(8):637–41.

8. Stober W, Fink A, Bohn E. Controlled growth of monodispersesilica spheres in micron size range. J Colloid Interface Sci.1968;26(1):62–9.

9. Jiang P, Bertone JF, Hwang KS, Colvin VL. Single-crystalcolloidal multilayers of controlled thickness. Chem Mat.1999;11(8):2132–40.

10. Miguez H, Meseguer F, Lopez C, Mifsud A, Moya JS, VazquezL. Evidence of FCC crystallization of SiO2 nanospheres.Langmuir. 1997;13(23):6009–11.

11. Pusey PN, Vanmegen W, Bartlett P, Ackerson BJ, Rarity JG,Underwood SM. Structure of crystals of hard colloidal spheres.Phys Rev Lett. 1989;63(25):2753–6.

12. Zeng Y, Harrison DJ. Self-assembled colloidal arrays as three-dimensional nanofluidic sieves for separation of biomolecules onmicrochips. Anal Chem. 2007;79(6):2289–95.

13. Zeng Y, He M, Harrison DJ. Microfluidic self-patterning oflarge-scale crystalline nanoarrays for high-throughput continuousDNA fractionation. Angew Chem-Int Edit. 2008;47(34):6388–91.

14. Birdsall RE, Koshel BM, Hua Y, Ratnayaka SN, Wirth MJ.Modeling of protein electrophoresis in silica colloidal crystalshaving brush layers of polyacrylamide. Electrophoresis. 2013.doi:10.1002/elps.201200413.

15. Jiang P, Prasad T, McFarland MJ, Colvin VL. Two-dimensionalnonclose-packed colloidal crystals formed by spincoating. ApplPhys Lett. 2006;89(1).

16. Wei BC, Rogers BJ, Wirth MJ. Slip flow in colloidal crystals forultraefficient chromatography. J Am Chem Soc. 2012;134(26):10780–2.

17. Kamp U, Kitaev V, von Freymann G, Ozin GA, Mabury SA.Colloidal crystal capillary columns—towards optical chromatog-raphy. Adv Mater. 2005;17(4):438–43.

18. Malkin DS, Wei BC, Fogiel AJ, Staats SL, Wirth MJ.Submicrometer plate heights for capillaries packed with silicacolloidal crystals. Anal Chem. 2010;82(6):2175–7.

19. Moon JH, Kim S, Yi GR, Lee YH, Yang SM. Fabrication ofordered macroporous cylinders by colloidal templating inmicrocapillaries. Langmuir. 2004;20(5):2033–5.

20. Wei B, Malkin DS, Wirth MJ. Plate heights below 50 nm forprotein electrochromatography using silica colloidal crystals.Anal Chem. 2010;82(24):10216–21.

21. Egas DA, Wirth MJ. Fundamentals of protein separations: 50years of nanotechnology, and growing. Annu Rev Anal Chem.2008;1:833–55.

22. Pal S, Fauchet PM, Miller BL. 1-D and 2-D photonic crystals asoptical methods for amplifying biomolecular recognition. AnalChem. 2012;84(21):8900–8.

23. Huang X, Wirth M. Surface-initiated radical polymerization onporous silica. Anal Chem. 1997;69:4577–80.

24. Papautsky I, Mohanty S, Weiss R, Frazier AB. High lane densityslab-gel electrophoresis using micromachined instrumentation.Electrophoresis. 2001;22(18):3908–15.

25. Kopecka K, Drounin G, Slater G. Capillary electrophoresissequencing of small ssDNA molecules versus Ogston regime:fitting data and interpreting parameters. Electrophoresis.2004;25:2177–85.

26. Mahler HC, Friess W, Grauschopf U, Kiese S. Protein aggrega-tion: pathways, induction factors and analysis. J Pharm Sci.2009;98(9):2909–34.

27. Slavickova A. Electrophoresis of nonreduced IgM and IgGmonoclonal-antibodies on mixed agarose-polyacrylamide gels.Electrophoresis. 1990;11(10):892.


163

Documents

Silica Colloidal Crystals As Emerging Materials For The