MASS SPECTROMETRY INTERFACED WITH ION MOBILITY OR
LIQUID CHROMATOGRAPHY SEPARATION FOR
THE ANALYSIS OF COMPLEX MIXTURES
A Dissertation
Presented to
The Graduate Faculty of The University of Akron
In Partial Fulfillment
of the Requirements for the Degree
Doctor of Philosophy
Danijela Smiljanic
December, 2011 MASS SPECTROMETRY INTERFACED WITH ION MOBILITY OR
LIQUID CHROMATOGRAPHY SEPARATION FOR
THE ANALYSIS OF COMPLEX MIXTURES
Dissertation
Danijela Smiljanic
Approved: Accepted:
______Advisor Department Chair Dr. Chrys Wesdemiotis Dr. Kim C. Calvo
______Committee Member Dean of the College Dr. Bi-min Zhang Newby Dr. Chand K. Midha
______Committee Member Dean of the Graduate School Dr. David Perry Dr. George R. Newkome
______Committee Member Date Dr. Yi Pang
______Committee Member Dr. Peter Rinaldi ii ABSTRACT
This dissertation focuses on coupling separation techniques such as ion mobility
(IM) and liquid chromatography (LC) to mass spectrometry and their application to characterization of complex mixtures. Non-covalent complexes between poly(ethylene imine) (PEI) and single stranded oligodeoxynucleotides (ODNs), as well as components from black raspberries, were characterized utilizing ion mobility mass spectrometry (IM-
MS) and liquid chromatography mass spectrometry (LC-MS), respectively. Interfacing these separation methods to mass spectrometry allows for detection and identification of isobaric species and species present in low concentration.
Non-covalent complexes between low molecular weight poly(ethylene imine)
(PEI 400 and 800) and single-stranded oligodeoxynucleotides (ODNs) were investigated for five ODNs, including d(TTTTT), d(CCCCC), d(AAAAA), d(GGGGG) and d(GCGAT). In chapter 4 the compositions, as well as solution and gas-phase stabilities of the complexes (termed polyplexes) were examined by electrospray ionization mass spectrometry (ESI-MS) and tandem mass spectrometry (MS2). Independent of the mixing ratio of the reactants, the polyplex with 1:1 polymer-to-nucleotide stoichiometry,
PN, is the dominant product. The gas-phase stabilities, assessed by MS2 and collisionally activated dissociation, follow the same order, providing evidence that the polyplex structures in aqueous solution and the more hydrophobic environment of the gas phase
iii are very similar. Non-covalent complexes with different composition but the same molecular mass were corroborated by ion mobility mass spectrometry (IM-MS).
In chapter 5 of this dissertation an investigation of the expanded non-covalent system, ternary complexes, of poly(ethylene imine), single-stranded oligodeoxynucleotides and glutamic acid entities, is discussed. The solution stabilities and gas-phase stabilities of the ternary complexes (termed terplexes) were examined by electrospray ionization mass spectrometry (ESI-MS) and tandem mass spectrometry
(MS2). In addition, higher order ternary structures with multiple units of polymer and/or nucleotide present within the terplex were identified utilizing IM-MS.
Finally, chapter 6 provides information on characterization of non-anthocyanin components from black raspberry fractions by interfacing liquid chromatography (LC) to mass spectrometry. The black raspberry extracts were provided from the lab of Dr.
Joseph C. Scheerens at Ohio Agricultural Research and Development Center, Wooster,
Ohio. Combination of LC-MS, ESI-MS and MS2 provided structural information for the corresponding components.
iv DEDICATION
To my parents Ana and Milutin Smiljanic and my brother Darijo Smiljanic, your love, support and sacrifice made this possible, and to my husband Josip Jovicic whose unconditional love and support helped me through the hard times.
v ACKNOWLEDGEMENTS
I would like to thank Dr. Wesdemiotis for his guidance, support and patience during my studies.
I would also like to thank my committee members Dr. Bi-min Zhang Newby, Dr.
David Perry, Dr. Yi Pang and Dr. Peter Rinaldi for their patience and helpful suggestions.
I especially want to thank my parents Ana and Milutin Smiljanic, my brother
Darijo and my husband Josip for always believing in me.
I would like to thank all the past and current group members Alyson Leigh, Dr.
Kittisak Chaicharoen, Dr. Sara Whitson, Dr. David Dabney, Dr. Bethany Subel, Madalis
Casiano, AleerYol, Vincenszo Scionti, Bryan Katzenmayer, Shi Chunxiao, Kai Guo, Nhu
Quynh Nguyen, Xiumin Liu, Nadrah Alawani and Dr. Xiaopeng Li for helpful discussions and friendships. Thank you for making the lab such a fun place.
I would like to thank my friend Madalis Casiano (Daly) for always making me laugh, and for making the days in the lab, that were not so great, more enjoyable.
I would also like to thank Liladher Pauldel for helpful discussions.
vi TABLE OF CONTENTS
Page
LIST OF TABLES ...... x
LIST OF FIGURES ...... xi
CHAPTER
I. INTRODUCTION ...... 1
II. INSTRUMENTAL METHODS BACKGROUND ...... 13
2.1 Mass Spectrometry ...... 13
2.2 Ionization Methods ...... 14
2.2.1 Electrospray Ionization (ESI)...... 15
2.3 Mass Analyzers ...... 19
2.3.1 Quadrupole Mass Analyzer ...... 20
2.3.2 Time-of-flight Mass Analyzer ...... 23
2.3.3 Quadrupole Ion Trap Mass Analyzer ...... 26
2.3.4 Quadrupole Time-of-Flight (Q/TOF) Mass Analyzer ...... 30
2.4 Detectors ...... 31
2.5 Ion Mobility Mass Spectrometry (IMMS) ...... 32
2.6 Liquid Chromatography Mass Spectrometry ...... 37
2.6.1 Column and Separation Efficiency ...... 39
III. MATERIALS AND INSTRUMENTATION ...... 44 vii 3.1 Materials ...... 44
3.2 Instrumentation ...... 45
IV. NON-COVALENT COMPLEXES BETWEEN POLY(ETHYLENE IMINE) ANDSINGLE STRANDED DEOXYNUCLEOTIDES ...... 49
4.1 Introduction ...... 49
4.2 Experimental Methods ...... 50
4.2.1 Mass Spectrometry Experiments ...... 50
4.3 Polyplexes with PEI 400 ...... 52
4.4 Comparison of PEI 400 vs. PEI 800 ...... 57
4.5 Intrinsic Stability of Polyplexes ...... 60
4.6 Ion Mobility Separation of the Polyplexes ...... 64
4.7 Conclusions ...... 67
V. TERNARY COMPLEXES OF POLY(ETHYLENE IMINE), SINGLE STRANDED OLIGONUCLEOTIDES AND GLUTAMIC ACID MOIETIES ...... 69
5.1 Introduction ...... 69
5.2 Materials and Methods...... 70
5.2.1 Mass Spectrometry Experiments ...... 71
5.3 Terplexes with PEI400...... 72
5.4 C/N Ratio Effect ...... 84
5.5 Tandem Mass Spectrometry Studies on the Terplexes ...... 88
5.6 Ion Mobility Separation of the Terplexes ...... 90
5.7 Conclusions ...... 95
VI. CHARACTERIZATION OF PHENOLIC COMPOUNDS IN BLACK RASPBERRIES BY LIQUID CHROMATOGRAPHY MASS SPECTROMETRY (LC-MS) ...... 96
viii 6.1 Introduction ...... 96
6.2 Sample Preparation ...... 98
6.3 Instrumental Conditions ...... 98
6.4 Characterization of Phenolic Compounds from Black Raspberries ...... 99
6.4.1 BRLC 1-H ...... 100
6.4.2 BRLC 1-E ...... 108
6.4.3 BRLC 1-K ...... 110
6.5 Conclusions ...... 119
VII. SUMMARY ...... 121
REFERENCES ...... 124
APPENDICES ...... 140
APPENDIX A. ADDITIONAL DATA ...... 141
APPENDIX B. COPYRIGHT PERMISSIONS ...... 145
ix LIST OF TABLES
Table Page
4.1 PEI-ODN (PN) Polyplexes Formed Between PEI 400 and Single-stranded Pentadeoxynucleotides ...... 55
2+ 4.2 E50 Values of PN Polyplexes Between PEI 400 (P) and Single-stranded Pentadeoxynucleotides (N) ...... 63
6.1 Gradient Elution Program for the Samples Obtained by HPLC Fraction Using UV-Vis Detection ...... 99
x LIST OF FIGURES
Figure Page
1.1 (a) Linear and (b) branched form of poly(ethylene imine) ...... 6
1.2 Structures of A, G, C and T bases in single-stranded deoxynucleotides ...... 8
2.1 General components of a mass spectrometer ...... 14
2.2 Droplet formation at the needle tip of the ESI capillary ...... 16
2.3 Decomposition of the droplet in the electrospray source according to the Rayleigh‘s equation ...... 17
2.4 Depiction of the Coulombic explosion of a charged droplet ...... 18
2.5 Quadrupole mass analyzer ...... 21
2.6 Schematic of a linear TOF mass analyzer ...... 23
2.7 Time of flight instrument equipped with a reflectron...... 24
2.8 Schematic of delayed extraction principle with TOF mass analyzer ...... 26
2.9 A quadrupole ion trap mass analyzer ...... 27
2.10 Stability diagram od the 3D quadrupole ion trap ...... 29
2.11 Microchannel plate detector and the electron multiplication within the channels ...... 31
2.12 Diagram of a Daly detector ...... 32
2.13 The Synapt HDMS, a Q/TOF mass spectrometer from Waters Corporation 2007 ...... 33
2.14 Triwave region of Synapt HDMS ...... 34
xi 2.15 Schematic of operation of traveling wave ion guide for the transfer of the ions through the buffer gas ...... 35
2.16 Basic components of HPLC-MS system ...... 38
2.17 (a) Pore structure of a stationary phase (b) Representation of eddy diffusion in a chromatographic column ...... 41
2.18 Flow distribution scheme for a longitudinal diffusion ...... 42
3.1 Background pressures in various segments of the Waters Synapt HDMS Q/TOF mass spectrometer Waters Corporation 2007 ...... 45
3.2 Diagram of the Bruker Daltonics Esquire-LC ESI-MS instrument ...... 46
3.3 Diagram of AgilentHPLC instrument ...... 47
4.1 ESI mass spectra of (a) PEI 400 and (b,c) the PEI-ODN polyplexes formed after mixing PEI 400 and 5'-d(TTTTT)-3' in the molar ratios 1:1 (b) and 5:1 (c) ...... 53
4.2 Relative intensity of the PEI-d(TTTTT) polyplexes with polymer-to-ODN stoichiometry of 1:1 (PN) vs. the total PEI 400 concentration in the PEI/ODN solution ...... 56
4.3 ESI mass spectra of (a) PEI 800 and (b,c) the PEI-ODN polyplexes formed after mixing PEI 800 and 5'-d(TTTTT)-3' in the molar ratios 1:1 (b) and 5:1 (c) ...... 58
4.4 MS2 (CAD) mass spectra of PN2+polyplex ions containing (a) d(TTTTTT) and (b) d(GCGAT) ...... 61
4.5 Fragmentation efficiency curves (relative abundance of PN2+ precursor ion vs. center-of-mass collision energy) of polyplexes containing a PEI with 5 repeat units and d(TTTTT), d(AAAAA) or d(GCGAT) ...... 62
4.6 MS2 (CAD) mass spectra of PN+ polyplex ions containing (a) d(TTTTTT), (b) d(AAAAA) and (c) d(GGGGG) ...... 64
4.7 Two-dimensional ESI-TWIM-MS plot (m/z vs. ion drift time) of the polyplexes formed after mixing PEI 400 and 5'-d(TTTTT)-3' in the molar ratio 5:1 ...... 65
4.8 Mass spectra of regions 9 (a), 10 (b) and 11 (c) in the ESI-TWIM-MS diagram of Figure 4.7 ...... 66
xii 5.1 Expanded view of the m/z 800-940 region of the ESI mass spectra of the PEI 400/5'-d(CCCCC)-3'/glutamic acid terplex aquired at an accelerating voltage of 35 V and backing pressure of (a) 2 mbars and (b) 6 mbars ...... 74
5.2 ESI mass spectra of PEI 400-d(TTTTT)-Glu terplexes formed by mixing the corresponding solutions in the ratio 10:1:1,using(a) 1mM and (b) 5mM solutions of Glu;(c) expanded view (m/z 900–1200) of the spectrum in part (b) ....76
5.3 Relative intensity of PEI-d(TTTTT)-Glu terplexes with PEI/d(TTTTT)/Glu stoichiometry of 1:1:1 vs. the total Glu concetration in the solution ...... 79
5.4 PEI-ODN-Glu ternary complexes formed from PEI 400, single-stranded pentadeoxynucleotides and glutamic acid solutions mixed in the ratio of 10:1:1 ....80
5.5 ESI mass spectra of a mixture of PEI400 and d(GGGGG) oligonucleotide a) before and b) after the addition of glutamic acid (E) ...... 82
5.6 MS2 (CAD) spectra of a) PE+ and b) P(EE)+ complexes ...... 83
5.7 ESI mass spectrum of PEI 400-d(TTTTT)-GluGlu terplexes formed by mixing the corresponding constituent solutions in the ratio 10:1:1,using a 5-mM GluGlu solution ...... 84
5.8 PEI-ODN-(Glu-Glu) ternary complexes formed from PEI 400, single-stranded pentadeoxynucleotides and glutamylglutamic acid solutions mixed in the ratio of 10:1:1 ...... 86
5.9 MS2 spectra of PNE2+terplexes containing a) d(TTTTT), b) d(CCCCC) and c) d(AAAAA) ...... 88
5.10 Fragmentation efficiency curves (relative precursor ion abundance versus center-of-mass collision energy) for PNE2+ ion from PEI-ODN-Glu terplexes containing d(TTTTT), d(AAAAA), d(CCCCC) and d(GCGAT) ...... 89
5.11 2-D ESI-IM-MS plot (m/zvs. drift time) of the species in a 10:1:1 (molar) solution of PEI 400, d(TTTTT) and GluGlu (EE, 5 mM) ...... 91
5.12 Mass spectra of regions a) 16 and b) 17in the ESI-IM-MS diagram of Figure 2+ 3+ 5.11, showing terplex compositions of P2N2(EE) and P2N(EE)2 , respectively ...... 93
5.13 Expanded plot of region 6 from the IM-MS diagram of Figure 5.11 ...... 94
6.1 LC-MS total ion chromatogram (TIC) of sample BRLC 1-H (bottom), and expanded trace of the species eluting between 26-33 minutes (top) ...... 100
xiii 6.2 LC-MS spectrum of the fraction eluting between 29.0-29.5 min in the TIC of BRLC 1-H ...... 101
6.3 MS2 spectrum of the ion at m/z 462.9 from fraction BRLC 1-H ...... 103
6.4 MS2 spectra of theions at m/z 300.9 (top) and m/z 475.0 (bottom) from fraction BRLC 1-H ...... 105
6.5 LC-MS spectrum of the fraction eluting between 29.5-30.0 min in the TIC of BRLC 1-H ...... 106
6.6 MS2spectrum of ion m/z 226.9 from fraction BRLC 1-H ...... 106
6.7 Chemical structures of the phenolic compounds identified in fraction BRLC 1-H ...... 107
6.8 LC-MS total ion chromatogram (TIC) of sample BRLC 1-E (top) and extracted mass spectrum for the peak eluting between 24.0-25.0 minutes (bottom) ...... 109
6.9 MS2 spectrum of the ion at m/z 288.9 from fraction BRLC 1-E ...... 109
6.10 LC-MS total ion chromatogram (TIC) of sample BRLC 1-K, (bottom) and expanded trace of the component eluting between 34-38 minutes (top) ...... 111
6.11 LC-MS spectra of the fractions eluting between 34.4-34.5min, 34.7-35.0 min and 35.1-35.2 min in the TIC of sample BRLC 1-K ...... 112
6.12 MS2spectra of the ions at m/z 725 and m/z 503 (top and bottom, respectively) from fraction BRLC 1-K ...... 114
6.13 LC-MS spectra of the fractions eluting between 35.5-35.6min, 35.6-35.8 min and 35.8-36.0 min in the TIC of sample BRLC 1-K ...... 115
6.14 MS2spectra of the ions at m/z 435.0(top) and m/z 272.9(bottom) from sample BRLC 1-K ...... 116
6.15 MS2 spectrum of m/z 479.0 form sample BRLC 1-K ...... 117
6.16 Chemical structures of the phenolic compounds tentatively identified in fraction BRLC 1-K ...... 118
6.17 Comparison of extracted chromatogram peaks for m/z 579.9 and m/z 462.9 from fraction BRLC 1-K and distinguishing between real and spike peaks ...... 119
xiv
CHAPTER I
INTRODUCTION
Non-covalent interactions play an important role in the areas of chemistry, biology and biochemistry. Their importance is reflected through some of the crucial processes they control which make our life possible. These weak forces are responsible for modulation of reactivity in chemistry, they control the secondary and tertiary structures of proteins, and they drive cellular processes such as cell division, cell signaling, gene transcription and translation. Commonly, the function of biomolecules is dependent on non-covalent interactions with other biomolecules. Proteins interact with other proteins, peptides, nucleic acids, oligonucleotides, metal ions and other species to fulfill their tasks.1,2 The study of non-covalent interactions in and between molecules is named supramolecular chemistry. The focus of study in this field is aimed at investigating not only single molecules (as single species or as a bulk material) but complex chemical assemblies of at least two molecules held together by non-covalent forces.3,4 The non-covalent interactions occur between single molecular building blocks which by molecular recognition and self-assembly lead to the formation of supramolecular entities.5 Self-assembly is responsible for the creation of complex structures such as enzymes and viruses. These well-defined aggregates can have special properties that are only found in the ensemble itself but not in the single participating
1 constituents. Gaining knowledge about supramolecular interactions is essential for understanding biological processes in diseases.3,6
Supramolecular chemistry is considered a young discipline which did not start its development as an independent field of research until the 1960s. One of the reasons is that early chemists considered strong covalent bonds to be the major determinant of molecular properties, while they neglected the influence of the surroundings. A shift in this perception has occurred after an increasing number of examples documented the importance of the environment on the properties of molecules.7 The significance of the non-covalent, intermolecular interactions then became the focus of research activities and
Jean-Marie Lehn, one of the founders of supramolecular chemistry, introduced the term
―supramolecule.‖8 The non-covalent forces within supramolecules are individually much weaker in nature than covalent bonds. They can roughly be classified in four main types: electrostatic or ionic interactions, hydrogen bonding interactions, Van-der-Waals and hydrophobic interactions.9 Another reason for the late development of supramolecular chemistry is the lack of a reliable method for studying these weak non-covalent forces.
Mass spectrometry has emerged as one of the major tools for studying non-covalent complexes after the development of the soft-ionization methods electrospray ionization
(ESI)10 and matrix assisted laser desorption ionization (MALDI).11
Mass spectrometry has played a major role in a variety of scientific fields by providing quantitative and qualitative information about the analyte. As an analytical technique which investigates the mass-to-charge ratio (m/z) of ions in high vacuum, of up to 10-10 torr, it is a method for studying isolated ions and can provide information on the
2 intrinsic properties of the analyte studied.12 Comparison of the intrinsic properties (in the gas phase) with the properties affected by the environment (solution phase) can significantly contribute to a better understanding of the nature of the non-covalent bonds.7 Various other analytical techniques have been applied for the study of non- covalent assemblies, more specifically for their structural characterization, stoichiometric determination and the measurement of their dissociation rate constants. These instrumental techniques include: Nuclear Magnetic Resonance (NMR), X-ray crystallography, and Infrared (IR) and Ultraviolet spectroscopy (UV), among others.2,13
Although soft ionization mass spectrometry is complementary to the other analytical methods for studying non-covalent complexes, it offers advantages in aspects which the other biophysical methods lack. McLafferty has termed these advantages of mass spectrometry as the three ―S‖, referring to specificity, speed and sensitivity.14 All of these can benefit the study of non-covalent complexes. Specificity is crucial in identifying specific interactions and relative binding constants.15,16 Speed and sensitivity are other noticeable advantages of mass spectrometry; they permit analysis at very low concentrations of the species of interest ( picomoles to femtomoles) and within a very short time.17
A great number of investigations on non-covalent complexes by mass spectrometry have been reported since the invention of electrospray and matrix assisted laser desorption ionization, ESI15,18,19 and MALDI,20,21,22,23 respectively. In MALDI, the solution of analyte is mixed with a solution of crystalline matrix which has the main role to absorb energy from the laser. After the mixture is deposited on the sample plate and dried, it is irradiated with a UV laser beam to induce the ionization of the analyte. 3 Consequently, fragmentation may occur to a significant level with this type of ionization.
In ESI, the solution of analyte flows continuously through a metal capillary to which a high electrical potential is applied. Upon evaporation of the solvent, the droplets undergo fission leading to single ions.12 The evaporating solvent molecules help to keep the internal energy of the ions low and suppress fragmentation which makes ESI a more suitable method for studying weak non-covalent interactions.7 Moreover, ESI possesses the ability to generate multiply charged ions which permits the analysis of high molecular weight molecules. The extensive literature on ESI-MS points out that ESI has benefitted the analysis of many types of biochemical systems. Systems studied by ESI-MS include protein-peptide interactions,15,24 polypeptide-metal ion complexes,25 and protein-nucleic acid/oligonucleotides complexes.26,27
Non-covalent interactions between bimolecular systems and synthetic polymers have gained great interest in the global research community. Synthetic polymers play one of the most important roles in the preparation of pharmaceutical products.28 The pharmaceutical applications of polymers extend in many areas, including material packaging, and the production and development of drug delivery systems in gene therapy.
Gene therapy can be defined as the treatment of human disease by delivering genetic material into specific cells of the patient.29 Development of polymers with special properties and refinement of their surface properties (such as hydrophilicity and smoothness) and bulk properties (such as molecular weight and solubility) can assist in the design of polymers for various drug delivery applications. Biodegradable polymers are extensively researched for this purpose as they can degrade to non-toxic monomers
4 inside of body.30,31 For a successful gene therapy, safe and efficient gene introduction is needed, which still is a key limitation for controllable human gene therapy.32,33
Two major categories for gene delivery vehicles have been established: viral vectors and synthetic vectors. In the case of viral vectors, a virus carries genes and delivers them into living cells. Although viral vectors are considered very efficient, their use is affected by the amount of genetic material they can carry and severe safety risks.
Because of these concerns, non-viral gene vehicles or synthetic vectors became promising alternative as they offer opportunities for increased safety, flexibility and easier production. In general, synthetic vectors are materials that by electrostatic binding spontaneously form complexes with DNA, called polyplexes. Furthermore, they condense genetic material into particles, protect the genes and mediate the cellular entry.33 Among synthetic vectors, synthetic polymers have emerged with several advantages such as ease of preparation and chemical modification as well as great stability. Major attention has been paid to cationic polymers as materials for the study of gene delivery. Although when compared to efficient viral vectors, they suffer from low gene transfer efficiency and toxicity, cationic vectors are able to carry large genes and mask the negative DNA charge, which is necessary for gene transfection.33,34,35
Polyethylene imine (PEI) has been widely used as a gene delivery vehicle and has emerged as one of the most popular cationic polymers for research in this area. Extensive investigation of this cationic vector has confirmed its efficiency in both in-vitro and in- vivo applications.36,37,38,39,40, 41,42 The property that makes PEI suitable as DNA vehicle is its structure. The basic monomer unit in the backbone of PEI has the composition
CH2CH2NH and the entire polymer is available in linear and branched form (Figure 1.1).
5 (a) (b)
NHCH2CH2 NHCH2CH2 NCH2CH2 n n m CH CH NH 2 2 2
Figure 1.1. (a) Linear and (b) branched form of poly(ethylene imine).
The branched form contains primary, secondary and tertiary amines, all of which groups can easily be protonated. High degree of protonation is an important asset of PEI that makes it a successful gene vector. The polymer has a very high density of amines where every third atom of the polymer is nitrogen, and at physiological pH 15-20% of them are protonated.43 The presence of multiple unprotonated amines in complexes are believed to allow the polymer to function as a buffering system during a sudden decrease in endolysosomal pH. After the cellular uptake via endocytosis, the polyplexes are entrapped in endosomes. In the endosome, the pH drops from 7 to 5 causing the overall protonation level of polymer to increase from 20% to 40%. This is important for the protection of DNA as it travels to the nucleus and allows the release of DNA into the plasmid before lysosomal degradation occurs. This unique property makes PEI a strong
―proton sponge‖ polymer. Once they are internalized into the cells via endocytosis, proton-sponge polymers are believed to attract more protons and become more highly charged. The high concentration of cations increases the osmotic pressure, ultimately causing rupture of the endosome membrane (due to influx of chlorine ions) and release of the polyplex into the cytoplasm. 33, 44 An advantage of PEI over other cationic polymers used in drug delivery is its low cost.45,46
6 Numerous studies have been performed on PEI/DNA or
PEI/oligodeoxynucleotide (ODN) complexes which describe transfection efficiencies and cytotoxicities.38,47,48,49,50 The effect of their size, shape, surface charge, N/P ratio (the ratio of nitrogen moieties in PEI to phosphate units in DNA/ODNs), and molecular weight has been examined. Since the efficacy of PEI-derived vectors depends on molecular weight and degree of branching of the polymer, some of these studies have evaluated the influence of these properties. Godbey et al. found increasing transfection efficiencies with increasing molecular weight in vitro,49 while Abdallah et al. found that molecular weight increase leads to reduction of transfection efficiencies in vivo.50
Somewhat contradictory results were also obtained for the influence of branching.
Although branched PEI has yielded significantly greater success in terms of cell transfection, and is the standard form used in gene delivery,44,51 Wightman et al. found that linear PEI was more effective than the branched PEI, both in vitro and in vivo.47
High MW PEI has been shown to provide higher stabilization to PEI/DNA polyplexes and a high transfection rate but at the same time it suffers from increased cyctotoxicity.
To reduce the cytotoxicity and still keep high transfection rates, studies with low molecular weight PEI vector have been performed, which require higher polymer concentrations to achieve comparable efficacy. Here, optimization of the N/P ratio parameter has been shown to have great influence on the effectiveness of the gene delivery system; more specifically, optimal conditions are achieved with a higher N/P ratio.46,52 In addition, the modification of low molecular PEIs has been studied to improve gene transfer efficiency while keeping cytotoxicity manageable.46,53
7 As mentioned above, many studies have investigated the cytotoxicity and transfection efficiency of polyplexes, however information about their compositions and binding interactions are limited. Chapter 4 of this dissertation reports the first such characterization for non-covalent complexes between single-stranded oligodeoxynucleotides (ODNs) and low molecular weight poly(ethylene imine) (PEI 400 and 800), which were studied by electrospray ionization mass spectrometry (ESI-MS).
Five ODNs were investigated (Figure 1.2) including: d(TTTTT), d(CCCCC), d(AAAAA), d(GGGGG) and d(GCGAT) which were mixed with PEI in different molar ratios. The stoichiometries and charge state distributions of the resulting polyplexes, together with their solution and gas-phase stabilities, were examined by ESI-MS and tandem MS.
NH2 O
N N N NH
N N H N H N NH2
A - Adenine G-Guanine
NH2 O
N NH
N O N O H H
C - Cytosine T - Thymine
Figure 1.2.Structures of A, G, C and T bases in single-stranded deoxynucleotides.
Cationic PEI carriers show high degree of transfection in many cases, but in the presence of serum their efficiency decreases greatly which hinders their in vivo
8 applications.54,55 The resulting net positive charge of the polyplexes is beneficial for binding and penetrating negatively charged cell surface in cell culture, but this positive surface charge can cause disadvantages when injected into the blood stream. The polyplex can bind non-specifically to the negatively charged components, such as plasma proteins (immunoglobulin M, etc.) or blood cells, and the resulting aggregates can clog the capillaries. Additionally, they can also activate innate defense mechanism against foreign particles in the blood stream. This can further lead to rapid elimination from the blood and adverse effects such as inflammatory reactions. The interaction of polycations with blood proteins, such as albumin, can also cause the complex to disassemble, thus exposing the DNA to degrading enzymes.56,57 In order to reduce cytotoxicity and overcome aggulation with blood components, several strategies have been reported for masking the polyplex‘s positive charge. To improve the biocompatibility of these systems a frequent approach used was chemical modification of PEI by covalently binding it to hydrophilic polymers such as poly(ethylene glycol) (PEG). Grafting of the
PEG onto the PEI has been shown to efficiently shield polyplexes from non-specific interactions with blood components. Polyplexes protected in this way have reduced cytotoxicity and improved transfection activity.56,58 However, as PEGylated PEI stabilized DNA polyplexes, a high degree of PEGylation was found to significantly reduce cellular uptake of these cationic complexes.59
An alternative way to modify DNA/PEI assemblies is to bind an anionic polymer to their surface, such as poly (glutamic acid) (PGA). PGA is a hydrophilic, biodegradable polymer with properties such as nontoxicity and biocompatibility which make it a good candidate for gene delivery studies.60 DNA/PEI/PGA complexes, called
9 terplexes, have successfully been applied to overcome the serum inhibitory effect and decrease toxicity.61,62,63 The carboxyl groups of PGA may self-assemble with PEI, and combined PEI/PGA carriers have shown higher cellular import of DNA than that by PEI alone under serum-containing conditions. Various analytical methods such as agarose gel electrophoresis and circular dichroism(CD) were utilized to examine the properties of terplexes, and their transfection efficacies were the subject of numerous investigations.
Chapter 5 of this dissertation examines the compostion and structures of
PEI/oligodeoxynucleotide/glutamic acid or GluGlu dipeptide terplexes (abbreviated as
PNE or PN(EE) respectively; P = polymer, N = ODN, E = Glu) by electrospray ionization mass spectrometry (ESI-MS).The stoichiometries of these ternary complexes are studied, together with the interactions developing among the constituents within the terplex.
The significance of diet in regards to human helath has increased the need for more information about food of nutritional value, especially fruits and vegetables.64
Various studies have reported that a diet rich in frutis and vegetables reduces the risk of numerous dieseases, among which are cardivascular diseases and cancer.65,66,67
Compounds that are greatly responsible for these effects are phytochemicals(with protective or disease preventative properties) some of which have strongantioxidant properties. The antioxidant benefit of phytochemicals in promoting health is related to their ability to quench reactive oxygen species such as hydroxyl, peroxide radicals and others, thus reducing and repairing damage from oxidative stress and inflamation.
Berries, among other fruits, are one of the most important sources of antioxidants in our diets.68,69 A major group of phytochemicals in berries are greatly diverese phenolic
10 compounds. They differ with respect to structure and molecular weight but possess a common characteristic of having at least one or more aromatic rings with hydroxyl groups. Berry phenolics are classified into four major categories which include flavonoids, stilenes, tannins and phenolic acids.70,71 Furthermore, flavonoids can be divided in flavanols (or catechins), flavonols, flavones, flavanones, isoflavonoids and anthocyanins, while two main sub-categories of the phenolic acids class are derivatives of hydroxy-cinnamic and hydroxy-benzoic acids72.
Anthocyanins are the group of phenolics in berries which are responsible for the pigmentations found in the skins of fruits, while the group that contributes to the taste of the berries are tannins.64,68 The content of phenolics, and anthocyanins, in berries can be affected by many factors such as species type, ripeness, geographic region, climate, storage conditions and others. Various studies examined the impact of these factors, reporting on the importance of climate fluctuations on the quality and quantity of berries, or on an increased content of anthocyanins with maturity of the fruitsbut decreased antioxidant activity at the same time.73,74,75,76
Among berries, the black raspberry specifically has been identified to contain high anthocyanin levels. Production of black raspberries in the United Stated has been declining since the 1900s, which has been correlated with increase in diseases.77 Some studies have shown that black raspberries could inhibit the growth of cancer in rodents and be a promising alternative as a natural chemopreventative option with various cancers, including esophageal, oral, and coloncancers.78,79,80 Studies on black raspberries increased markedly because they are a rich source of anthocyanins and their consumption can result in health benefits.81
11 Among the various methods used for the characterization of phenolic compunds, liquid chromatography coupled to mass spectrometry (LC-MS) has been recognized as a powerful tool.82,83 Crude samples of berries can be complex by nature and have a wide range of mixtures of phenolic compounds which can complicate their analysis and identifications by mass spectrometry alone. Liquid chromatography is the most appropriate tool to assist in the identification of such samples, as it combines a high separation power with the capability to characterize individual componentes. Separation via LC is achieved based on the interactions of the compounds with the stationary and mobile phases present in the column. Coupling mass spectrometry to a separation technique like liquid chromatography provides an analythical method with both specificity and selectivity.12,84
Many studies aiming at identifying the components in berries have used the power of LC-MS to determine the compound identity,phenolic content, or antioxidant capacity.82,85,86 The identification of single components in berries is necessary in order to assess their contribution to the overall health benefits of their consumption. Chapter 6 of this dissertation focuses on the identification of phenolic compounds in black raspberries.
The samples were first fractionated by HPLC and single fractions were then analyzed by
LC/ESI-MS and MS/MS methods. Collected fractions of the samples were divided in two series: Series 1 and Series 2. In this dissertation, the compounds in three representative samples of Series 1: samples BRLC-1H, BRLC-1K, and BRLC 1-E, were identified.
12
CHAPTER II
INSTRUMENTAL METHODS BACKGROUND
2.1 Mass Spectrometry
Mass spectrometry is an anlaytical method that is utilized in drug discovery, forensic science, archeology, identification of natural products, environmental monitoring and a wide range of other fields. Mass spectrometry is used for determining the structures of both individual components as well as complex mixtures. This is accomplished by converting the analyte molecules into charged ions, so that the mass spectrometer can separate them based on their mass-to-charge ratios (m/z). Basic components of every mass spectrometer are an inlet system, an ion source, a mass analyzer, a detector and a data analysis system (Figure 2.1).84 The inlet system introduces the sample into the ion source. Sample introduction can be achieved via direct injection or chromatograhy (such as liquid chromatography (LC) or gas chromatography
(GC)). The ionization source converts the analyte into gaseous ions. The mass analyzer seperates these ionic species according to their mass-to-charge (m/z) ratio. The detector then detects these ions, and it measures and amplifies the ion current from each ion generated from the analyte. Finally, the data system records this signal and displays the data in the from of mass spectra.12
13 Sample inlet
Ion source Mass analyzer Detector
Vacuum system Data system
Figure 2.1.General components of a mass spectrometer.
Most of the components, such as the analyzer and the detector, in some cases the ionization source, are kept under vacuum. Vacuum, achieved utilizing the pump system, maintains a low pressure in the mass spectrometer, 10-5- 10-10 torr. Because ions are reactive species, the vacuum is required to avoid collisions of the ions with other gaseous molecules.84 The most important step in mass spectrometric analysis is creation of ions.
Thus, the different ionization methods along with the different mass analyzers and detectors used in the work of this dissertation will be disscussed.
2.2 Ionization Methods
A successful mass spectrometric analysis depends largely on successfully ionizing the analyte studied. Ionization of a moleule can be achieved by ejection or addition of an electron, or addition or subtraction of an ion. Depending on the analyte of interest, an appropriate ionization method should be selected. Ionization methods are generally classified as ―hard‖ and ―soft‖. Hard ionization methods cause extensive fragmentaion of
14 the analyte due to the high energy deposited during ionization. Fragmentation caused by this type of ionization can be extensive, to the degree that the molecular ion of the compound of interest would not be detected. Although extensive fragmentation can provide a good picture of the structures present in the sample, it can also result in a very complex spectrum and difficult interpretation. Examples of hard ionization methods include electron ionization (EI), chemical ionization (CI), fast atom bombardment (FAB), and field desorption (FD). Unlike hard ionization techniques, soft ionization methods allow detection of intact molecular ion species with minimum degradation. For a long time, mass spectrometry has been considered as a harsh and distructive method for compound analysis. The development of soft ionization methods,especially electrospray ionization (ESI) and matrix-assisted laser desorption ionization (MALDI), has changed this outlook by enabling direct ionization of intact molecules.
2.2.1 Electrospray Ionization (ESI)
The development of ESI is attributed to John Fenn when he demonstrated the formation of multiply charged ions from large protein molecules, in the late 1980s.106
The concept of electrospray ionization was first introduced by Malcolm Dole in 1968 when he sugested the formation of gaseous ions from a solution by the means of electrospray.87 The Nobel Prize in Chemistry in 2002was awareded to Fenn (Dole was deceased at that time) for contributing the first reported application of ESI to biological macromolecules.
Main steps in the process of electrospray ionization are ion formation, nebulization, desolvation and ion desorption.88 Electrospray ionization occurs when a 15 solution of the analyte is sprayed through a metalic capillary tube to which a high electric field is applied. This high electric field is obtained by a potential difference between the capillary and the counter electrode and is usually 3-6 kV in strength (Figure 2.2). The electric field leads to the formation of highly charged droplets at the capillray exit. The formation of these droplets is assisted with a nebulizing gas, usually nitrogen.
Nebulizing gas flow
Electrostatic ESI field capillary
Taylor c one
Figure 2.2. Droplet formation at the needle tip of the ESI capillary (Reproduced with permission from reference 88).
16
Figure 2.3. Decomposition of the droplet in the electrospray source according to the Rayleigh‘s equation (Reproduced with permission from reference 84).
At this high electric field the pressure of the charges accumulated in the droplet at the tip of the capilary overcomes the surface tension. Consequently this causes the droplet to become elongated and attain the shape of a Taylor cone that ultimately breaks into smaller droplets (Figure 2.3). Decomposition of the droplets occurs at the Rayleigh limit, which is described by the Rayleigh equation (Equation 2.1) where q is the charge, ɛ0 is permittivity of the environment, γ is the surface tension and D is the diameter of the spherical droplet. Gomez and Tang were able to photograph the droplet formation in the
ESI source and concluded that fragmentation of the droplets can occur before the
Rayleigh limit is reached because the droplets are mechanicaly distorted by the electrical field.89,90
2 2 3 q = 8π ɛ0γD Equation 2.1
On their way to the mass analyzer, charged droplets pass through a heated capillary where they undergo shrinkage as the solvent evaporates. This capillary is filled with 17 heated dry gas that is usually nitrogen. As the droplets become smaller, the charge that they carry remains the same, resulting in increased electrostatic tension at the surface of the droplets. Charge repulsion becomes equal to the surface tension force that is holding the droplet at the Rayleigh limit, as described in the equation above. At this point, formation of a Taylor cone occurs, which ultimately explodes resulting in the production of smaller droplets (Figure 2.4). This process is also called Coulombic exlposion and it is repeated until only a single molecule with chargesremains.12,84,88 When the electric field on the droplet surface becomes very high (due to solvent evaporation), desorption of multiply charged ions from the surface is also possible.
Figure 2.4. Depiction of the Coulombic explosion of a charged droplet (Adapted with permission from reference 88).
One of the advantages of electropray ionization (ESI) is the feasibility of the method for direct coupling to separation techniques such as high performance liquid chromatography (HPLC), as will be disscussed in the later sections. Another great
18 feature responsible for the success of this method is its abiltiy to produce multiple charged ions allowing the analysis of high molecular weight compounds with mass analyzers that have a limited mass range. Electrospray ionization is also a very sensitive method. Since charged molecules are released from the surface of a chraged droplet, the sensitivity for compounds whose concentration is high at the droplet‘s surface is higher.
With mixtures, molecules present in the core of the droplets, can not desorb, but may be released after sequential Coulombic fissions and solvent evaporation. The latter phenomenon usually applies to high molecular weight compounds (over 5000 Da).
2.3 Mass Analyzers
A crucial component of every mass spectrometer is the mass analyzer. The mass analyzer separates gasous ions according to their mass-to-charge ratio (m/z). Mass analyzers are held under vacuum in order to prevent any collisions of ions with other gaseous molecules before reaching the detector. While various types of mass analyzers have been developed, they can all be classified in two main categories. The first category corresponds to the scanning analyzers where the ions of different masses are scanned and transmitted at a given time, allowing only those ions of specific m/z to pass through. An example of such an analyzer is the quadrupole mass analyzer. The second category includes the analyzers that allow simultaneous transmissionor detection of all the ions.
The ion trap (IT) and time of flight (TOF) devices are examples of such analyzers.
Each mass analyzer has its own advantages and disadvantages, and their performance depends on the basic characteristics they possess. The upper mass limit is one of the important features of the mass analyzer. It is the maximum m/z ratio that the 19 mass anlyzer is able to measure. The next one is resolution which is the ability of a mass analyzer to distinctly separate two neighboring ions with a small m/z difference. In instruments such as ion traps (IT) and time of flight (TOF) mass spectrometers, two peaks are considered to be resolved if the valley between them is 50% of the height of the less intense peak. Transmission is the ratio of the number of ions reaching the detector and the number of ions entering the mass analyzer. Analysis speed is the number of spectra that can be measured per unit of time. Finally, there is the characterisitic of mass accuracy. Mass accuracy is the error between measured and theoretical accurate m/z ratio.
The next sections will discuss the mass analyzers used in this dissertation. These include the quadrupole, quadrupole ion trap (QIT),time-of-flight mass analyzer (TOF) and the Q/TOF hybrid.
2.3.1 Quadrupole Mass Analyzer
A quadrupole mass analyzer consists of four cylindrical (or hyperbolic) rods, that serve as electrodes (Figure 2.5).They are organized in a precisely parallel orientation.
Different direct-curent (dc) and radio-frequency (rf) potentials are applied to these rods to obtain an oscillating electric field. Positive dc and rf potentials are applied to one set of opposite rods, while the other set of rods is subjected to negative dc and rf potentials.
Prior to entering the quadrupole mass analyzer, ions pass through a set of lenses and become accelerated in the direction of the z-axis. Inside of the quadrupole, ions travel along the z-axis, and due to the potentials applied to rods, oscillate in the x-y plane.
When a positive ion enters the space between the four rods, it is initially attracted to a 20 negatively charged pole. If the potential of the rod changes sign, the positive ion will change its direction and be pulled back; it is vice versa for a negative ion. This produces a zig-zag motion of the ions inside the quadrupole. The combination of these actions creates a stability window such that only ions of a specific m/z are able to travel through the rods. Unstable ions will hit the rods and become neutralized.12,91
Figure 2.5. Quadrupole mass analyzer. (Reprinted with permission from reference 12).
The electric field created in the quadrupole is made up of superimposed alternating field and constant fields produced from the potentials applied to the rods. The potential applied to one set of the rods has positive dc (U) and rf components (V cos ωt), where ω is the angular frequency of the rf voltage. The potential applied to the other set has negative dc (-U) and rf components but in opposite phase (thus the net potential is -
(U–Vcosωt)), as shown in equations 2.2-2.4.
Φ0 = (U – V cos ωt) Equation 2.2
21 - Φ0 = -(U – V cos ωt) Equation 2.3
ω = 2πƒ Equation 2.4
In these equations Φ0 is the potential applied to the rods, V is the amplitude of the rf potential and t is the time. A mass spectrum is obtained by changing the dc and rf voltages but keeping their ratio the same. The rf, f, isrelated to the angular frequency, ω, via equation 2.4. Only specific ions will oscillate through the rods and reach the detector at a specific setting of U and V.
The motion of the ions in the quadrupole is mathematically described by the
Mathieu equation (equation 2.5).
2 d u + (au – 2qu cosωt) = 0 Equation 2.5 d2(ωt/2)2
au = ax = -ay = 8zeU Equation 2.6 2 2 mω r0
qu = qx = -qy = 4zeV Equation 2.7 2 2 mω r0
In this equation u represents the x and y coordinates of the ion; au and qu are dimensionless parameters , which are obtained by solving the Mathieu equation. They are proportional to the dc potential U and rf amplitude V, respectively; and r0 is one half of the distance between opposing rods. A stable trajectory of an ion traveling through the quadrupole rods will be achieved only if its x and y coordinates remain smaller than r0.
22 In a given quadrupole ratio of U/V is kept constant, but the absolute values of U and V are varied so that only a specific m/z has a stable trajectory.84
2.3.2 Time-of-flight Mass Analyzer
The time-of-flight (TOF) mass analyzer consists of a field-free flight tube in which ions are separated according to their velocities. A pulsed laser ionization technique is the most suitable means of ionization for TOF measurments; thus, a matrix- assisted laser-desorption ionization (MALDI) source is usually coupled with TOF mass analyzers (Figure 2.6).
Figure 2.6.Schematic of a linear TOF mass analyzer. (Reprinted with permission from reference 84).
After ions are produced in the source (MALDI) they are expelled in pockets and accelerated towards the flight tube. Acceleration is achieved by the potential difference that is applied between the electrode and the extraction grid. Ions that have been 23 accelerated by this potential (V) gain potential energy (Eel) which is ultimately converted to kinetic energy Ek (Equation 2.9). After the ions reach the field-free region, they travel a distance L for a time t to reach a detector (Equation 2.10). The m/z ratios are established by measuring the time that the ion spends in field free region before it arrives at the detector:
2 Ek = zeV = ½ mv Equation 2.9 t = L / v Equation 2.10
½ ½ ½ ʋ = (2zeVs/m) —› t = (m/z) L/(2eVs) Equation 2.11 where ʋ is the velocity of the ion leaving the source, m is the mass of the ion, and ze is the charge of the ion. This equation also shows that the lighter ions will travel faster and reach the detector sooner than the heavier ions. Since not all ions are formed at the same location, at the same time and with the same initial kinetic energy, there is a spread of flight times for ions of the same m/z ratio which causes poor resolution. This problem was resolved by the utilization of the reflectron and delayed extraction techniques (Figure
2.7).
Figure 2.7. Time of flight instrument equipped with a reflectron (Reprinted from reference 84). 24
The reflectron acts as an electrostatic mirror, by which ions are deflected and sent back through the flight tube. It consists of a series of grid electrodes connected through a resistive network. Ions of a given m/z ratio but with higher kinetic energy (hence higher velocity) will travel deeper in the reflectron than ions of identical m/z but with lower kinetic energy. Therefore, the faster ions spend more time in the reflectron than the ions with lower kinetic energy, and reach the detector at the same time as the slower ions.
Outside of the reflectron, an ion‘s kinetic energy it reestablished to the exact same value as before but in the oposite direction. Although the reflectron increases the resolution, it does that at the expense of sensitivity and introduces an upper mass limit.
Delayed extraction is another way to reduce the kinetic energy spread between ions of the same m/z ratio. Delayed extraction introduces a time-delay at the source, allowing the ions to drift for a certain time (nanoseconds to microseconds), between the formation and extraction events (Figure 2.8). Before the extraction grid is pulsed, the ions formed during ionization are allowed to drift in a field-free region and to separate according to their initial kinetic energies. During this time the faster moving ions drift closer to the grid while slower ions are furhter away and spend more time in the source.
After the extraction grid pulse is applied, the slower moving ions that are drifting further away from the extraction gird will be subjected to the greater potential energy than will the faster moving ions. Consequently, the slower ions will receive more kinetic energy and will eventually catch up with the faster ions, and both will reach the detector at the same time.12,84
25
Figure 2.8. Schematic of delayed extraction principle with TOF mass analyzer. (Reprinted with permission from reference 84).
2.3.3 Quadrupole Ion Trap Mass Analyzer
The quadrupole ion trap mass analyzer (QIT) was developed by Wolfang Paul in
1958.92 In 1989 Paul received the Nobel Prize in Physics for his contribution.
A QIT is a storing device that uses an oscillating electric field to store the ions and a quadrupolar field that traps theions in space by controlling their motion. The ion trap is composed of three elctrodes, where one ring electrode is located between two end
26 cap electrodes. The shape and arrangement of these electrodes inside of the ion trap create a three-dimensional field (Figure 2.9).
Figure 2.9. A quadrupole ion trap mass analyzer (Reprinted with permission form reference 88).
The two end cap electrodes are held at groud potential while an rf potential is applied to the ring electrode. Due to the potential difference that is created between the ring electrode and the end caps, a 3D quadrupolar field is created which is used to trap the ions. Both of the end cap electrodes contain small holes. The hole on the entrance end cap allows the injection of ions into the ion trap, while the multiple holes present on the exit end cap allows for the efficient ejection of the ions so they can reach the detector.
27 Once inside the ion trap, the ions are repelled from each other which causes theirtrajectories to expand and become unstable which can result in the ejection of the ions. To avoidsuch loss, a buffer gas (such as He) is introduced in the trap. The buffer gas collides with the ions and removes their excess energy while continually pushing them toward the center of the trap.84,88
The motion of anion in the quadrupole ion trap can also be described by theMathieu equation. As mentioned above, in the quadrupoleion motion was controlled by the potentials applied to the rods in the x and y direction only. However, in the quadrupole ion traps ion motion is determined by the potentials applied in three directions, x, y and z.
qz = qr = 8zeV Equation 2.12 2 2 m(r0+2z0 )Ω
Due to the cylindrical symmetry of the QIT, ion motion is described with coordinates z and r. The radial coordinate r is related to the x and y coordinates by r2 = x2 + y2. In equation 2.12, r0 is the inner radius of the ring electrode, ze is the charge of the ion, V is the amplitude of the rf potenial (Ω), and z0 is the distance from the center of the trap to an
2 2 end cap (commonly, r0 = 2z0 ). Application of the rf potential to the ring electrode creates an oscillatory field in the ion trap causing the trapped ions to oscillate along the r0 and z0directions. For the ions to have a stable trajectory, they can never reach or exceed the r0 and z0 coordinates. The ions remain in the traponly if their qz< 0.908 (Figure 2.10).
The maximum value of qz also defines the low-mass cut off. Modern quadrupole ion traps do not employ dc fields.84,92
28
Figure 2.10. Stability diagram od the 3D quadrupole ion trap (Reprinted with permission from reference 12).
In order to measure the m/z ratio of the trapped ions, they are ejected from the ion trap so that they can reach a detector. There are two modes of ejection, ions can be ejected at their stability limit or by resonance ejection. In the first mode, the rf potential is gradually increased so that ion trajectories are successively destabilized. This is followed by ejection of the ions from the trap in the z-direction where they strike the detector. Alternatively, there is resonant ejection which is achieved with an auxilary rf potential that is applied to the end cap electrodes. Trapped ions oscillate in the ion trap at their secular frequency (ω). When the rf potentialapplied to the end capsis tuned to the ions‘ secular frequency, theyabsorb energy and become excited. This leads to an increase
29 in their amplitudes and unstable trajectoriesuntil they are ejected in the z-direction toward the detector.
2.3.4 Quadrupole Time-of-Flight (Q/TOF) Mass Analyzer
Various mass analyzers can be combined to enhance the power of a mass spectrometer. These mass spectrometers are identified as hybrid. The goal of these mass spectrometers is to utilize the advatage of each mass analyzer and provide a better preformance than that obtained with an individual mass analyer. One example of such hybrid instrument used for the work in this dissertation is the quadrupole time-of-flight mass analyzer.
The Q/TOF analyzer is composed of a quadrupole (Q) placed orthogonally before a time-of-flight (TOF) analyzer. Generally this instrument contains a collision cell located between a quadrupole and time-of-flight tube. For MS experiments, only an rf potential is applied to the quadrupole so that it acts as an ion guide and transmits all ions from the source to the TOF tube. For MS/MS experiments, both rf and dc potentials are applied to the quadrupole, thus allowing only a specific precursor ion to pass through.
Once the selected ion travels through the quadrupole, it arrives to the collision cell where it is collided with a gas (such as Ar) to produce fragments. The fragment ions then travel to the TOF part for mass analysis.
Coupling quadrupole with time-of-flight mass analyzers allows for a higher resolution and mass accuracy. Resolutions higher than 10000 can be achieved with mass accuracies of 5-10 ppm.
30 2.4 Detectors
After the ions are mass-analyzed they travel to the detector. A detector converts ions into an electic current that is proportional to their abundance. The detectors used for the work in this dissertation were the microchannel plate (MCP) and Daly (photon multiplier) detectors.
The MCP detector consists of a plate with parallel cylindrical channels pierced inside. Each channel is 5-25 um in diameter and a few mm long. Each channel is covered by a semiconductor substance that emits electrons when it is struck by an ion beam. The beam will come in at an angle so that it can be magnified (Figure 2.11). The input of the plate is kept at the negative potential of about 1kV as compared to the output side. At the output, a metal anode is placed which collects the cascade of electons and the current is measured.84
Figure 2.11. Microchannel plate detector and the electron multiplication within the channels (Reprinted with a permission from reference 84).
31 The Daly detector is a type of electro-optical ion detector which operates by converting ions into electrons and photons. Daly detectors are composed of a conversion dynode, a scintillator and a photomultiplier tube (Figure 2.12).
Figure 2.12. Diagram of a Daly detector (Reprinted with permission from reference 84).
Ions from the mass analyzer will hit the conversion dynode; positve ions will strike the negatively charged dynode and negative ions will hit the dynode with a positive potential.
This will result in the production of secondary electrons. These are then accelerated onto a phosporescent screen which emits photons. The photons are detected and amplified by a photomultiplier.84
2.5 Ion Mobility Mass Spectrometry (IMMS)
Traditional mass spectrometers possess a limitation in regards to distinguishing between ions with the same m/z ratio but different shape and size. To surmount this
32 problem, the mass spectrometer can be interfaced with ion moblity (IM) separationto provide a new dimension in MS analysis. The separation by IM depends on the size/charge ratios of the ions as well as on their interactions with a buffer gas.93 The ion mobility mass spectrometer utlizied for the work presented in this dissertatoin is the
Synapt HDMS from Waters Co. It is a quadrupole time-of-flight (Q/TOF) instrument with the ion mobility chamber located between the quadrupole and time-of-flight mass analyzers (Figure 2.13).
Figure 2.13. The Synapt HDMS, a Q/TOF mass spectrometer from Waters Corporation 2007(Reprinted with permision from reference 94).
The basic components of the Synapt HDMS mass spectrometer include an ion guide
(TWIG), a quadrupole, the triwave collision cell, and a time-of-flight tube.94
Commonly,pulsed pockets of the ions travel from the ion guide, to the quadrupole in MS mode, the quadrupole is used as a guide and guides the ions into the triwave reagion. The triwave region is composed of a trap cell, ion mobility cell and transfer cell. The trap cell 33 is used to accumulate ions and then release them to the ion mobility cell for mobility separation. The transfer cell is usedfor transfer of the mobility-separated ions to the TOF tube for mass analysis.95,107
The triwave components trap, IM and trasfer cells(all also called ion guides) are composed of stacked ring electrodes (Figure 2.14). An rf voltage with different phases is applied to consecutive electrodes, enabling the ions to move through (ions are funneled through). A low pressurebuffer gas (nitrogen) flows through the IM cell (in the opposite direction to the moving ions), which significantly slows down the movement of the ions.107,108
Figure 2.14. Triwave region of Synapt HDMS (Reprinted with permission from reference 107).
To drive the ions through the buffer gas a short dc voltage is applied to adjacent electrodes in repeating sequence and at regular time intervals (Figure 2.15). The continious sequence of traveling waves that are created in this manner transfer the ions 34 through the cell. As ions travel on these waves, the larger ones will move slower and stay behind the waves longer; conversely an ion with a small size will travel faster.
Usually, traveling waves with heights of up to 25 V and velocities between 300-600 m/s are used in Synapt HDMS, but settings that lie slightly outside these ranges may also be employed to improve resolution and sensitivity.107,108
Figure 2.15. Schematic of operation of traveling wave ion guide for the transfer of the ions through the buffer gas (Reprinted with permission from reference 107).
In addition, the transfer cell region is exposed to a weak traveling wave (of 1-2 V and 300 m/s) allowing the ions to keep their ion mobilities when they reach the TOF analyzer.
Traditional ion mobility spectrometers measure the time that ions take to travel through the buffer gas utilizing a static low electric field. Under this static electric field, the velocity of the ions is directly proportional to the electric field. The proportionality 35 constant K, is named the ion mobility constant and is directly proportional to the ion cross section.
½ ½ K = 3ze 2π M+m 1 Equation 2.13 16 N kT Mm Ω
In equation 2.13, N is the number density of the buffer gas, k is the Boltzmann constant,
T is the absolute temperature, m is the mass of the buffer gas atoms or molecules, M is the mass of the ion, and Ω is the cross section of the ion. The cross section of an ion can provide detailed information about its size and structure. As mentioned above, traditional
‗drift time‘ ion mobility spectrometers have the advantage that cross-sections can be deduced directly from the corresponding drift times. However that is not the case with the traveling wave (T-wave) ion mobililty spectrometers. Unlike with traditional IM spectrometry (IMS), where a static electric field is applied continously to the cell, T-wave
IMS instruments utlize an electric field that is varying so that the ions are carried through the cell with pulses of waves. In order to obtain ion cross-sections in T-wave IMS, the instrument must be first calibrated using the drift times of ions with known cross sections.
When two species with different structures but identical m/z ratios are analyzed via IMS, depending on their shape and size, they will move with different velocities through the IM cell. Ions with a smaller cross-section will travel faster than ions with a larger cross-section, because larger cross sectionscausemore collisions with the buffergas, which is flowing in the oppostite direction, and consequently slowthe ions down. Hence, ions with ashorter drift time will also have a smaller cross-section, and vice versa.
36 Sometimes ions of higher MW will have smaller drift times and cross-sections. This is a consequence of the charges present within the species analyzed. If the charges are close together in the structure of the molecule and experience repulsion, the molecule may fold around the charges to stabilize its structure; thus resulting in a smaller size and shorter drift time. Cross-sections of ions can provide valuble information about their geometry.
2.6 Liquid Chromatography Mass Spectrometry
In order to analyze a complex mixture a separation technique such as high performance liquid chromatography (HPLC) can be coupled to the mass spectrometer.
HPLC is a powerful analytical tool that has been utilized for separation and analysis of mixtures since the 1970s. The coupling of HPLC with MS has progressed greatly with the development of electrospray ionization in the 1980s.
An LC-MS system is composed of LC components that are interfaced with a mass spectrometer through an ionization source such as ESI (Figure 2.16). The LC components includethe solvent pumping system, injector, column, detector and data processing system.
37
Figure 2.16. Basic components of HPLC-MS system (Reprinted with permision from reference 96).
The dissolved mixture is introduced to the HPLC system by the means of the injector.
Solvent from the pumping system is forced through the HPLC column and the dissolved sample is injected into the flowing liquid, called mobile phase. For more complex mixtures, gradient chromatography is usually employed where two or more solvents in a continuously changing mobile-phase compostion allow for better resolution of the mixture components. The gradient system is achieved through the use of two solvent pumps and two solvent reservoirs to obtain a combination of the solvents. Alternatively, a gradient system can also be obtained by the use of a single pump but with multiple reservoirs and a switching valve. The switching valve gradient system is less expensive and is the one used for the work in this dissertation. The column is the heart of the HPLC where the actual separation occurs. It is packed with highly fine material that is referred to as the stationary phase. Each component in the mixture interacts differently with the packing material and separation occurs when the dissolved sample components interact differently with solvent or adhere with different affinities to the column material; in other words, separation is possibleif the distributions equilibria between mobile and stationary 38 phase are different for the sample components. HPLC interfaced with MS provides a highly sensitive tool for determining the number of components present in a mixture.96
2.6.1 Column and Separation Efficiency
The columns are usually made of stainless steel tubes packed with various fine materials depending on the intended applications. The most common packing material used is silica gel. Typically, the silica gel stationary phase is chemically modified by the attachment of ligands. The different types of ligands used to modify the silica determine the various types of chromatography that can be performed. The most widely used type is reverse-phase chromatography where the stationary phase is non-polar and the mobile phase polar. Typical ligands used in reverse-phase chromatography are alkyl groups such as octadecyl (C18), octyl (C8) and phenyl, which are bonded to the support surface through Si-O-Si linkage. In reverse-phase HPLC, separation occurs because of the strong interactions of the non-polar hydrophobic components with the stationary phase. Here, the mobile phase is composed of polar organic solvents, with the most common solvents used being methanol and acetonitrile.12,96 Reverse-phase high performance liquid chromatography (RP-HPLC) is the type of choromatography utilized in this dissertation; specifically silica-bonded C6-phenyl was used as stationary phase.
Today, HPLC columns are most commonly packed with spherical particleswith a diameter of 5 μm. The smaller size of the particles contributes to a tighter packing and a more efficient column. There are various column dimensions in use but the standard column size for most separations is 250 mm in length and 4.6 mm of internal diamter.In
39 addition, the stationary phase is made of porous material. Such material allows increased interactions of the sample with the stationary phase.96
Separation in chromatography depends on the distribution of the analyte between the stationary and the mobile phase. Components that exist more in the mobile phase will move through the column faster rather than those which reside more in the stationary phase. This phase preference can be expressed by the distribution coefficeint K, which is the ratio of the concentration of the component in the stationary and mobile phases. This distribution can also be described as analyte being in equilibrium between the two phases.
As the sample moves down the column it becomes dispersed and less concentrated. By measuring the amount of spreading that occurs, the column effieciency can be determined. A chromatographic peak is identified with the retention time of the analyte, tR, and the peak width wB, that represents the volume occupied by the analyte (VR). The efficiency of the column is measured by the quantitycalled number of theoretical plates
(N), which is described as:
2 N = 5.54( VR/w½ ) Equation 2.14
where w½ is the width of the peak at the half height. These plates are actually separate layers where separate equlibrations of the sample between the stationary and mobile phases occur. The longer the chromatographic bed the more theoretical plates are present. The plate number is dimensionless and, as it increases, the column becomes more efficient providing a greater level of separation.97
40 In order to achieve optimum separation,the size of the pores in the packing material of the column has to be appropriate. The analyte molecules can be trapped inside which can lead to slower elution and band broadening; the latter must be limited for efficient separation. There are three main factors or mechanisms that contribute to dispersion of solute as it travels through the column: eddy diffusion, longitudinal diffusion and mass transfer resistance. These should be carefully considered when performing experiments so that broadening iskept to a minimum.97, 98
Eddy diffusion is also known as multiple path effect (Figure 2.17 b). The mobile phase travels through the column that is packed with small stationary phase particles. Figure
2.20 (a) depicts the pore structure of a stationary phase particleshowing both the wide and narrow channels present in it. The solute molecules will randomly progress throught the column via different paths.
(a) (b)
Figure 2.17.(a) Pore structure of a stationary phase (b) Representation of eddy diffusion in a chromatographic column (Reprinted with permision from reference 98 and 97, respectively).
The path differences arise due to the unequal shape of the stationary particles or the imperfections in the packing material. Some molecules will pass down the column fast in
41 a more straight line while others will stay behind and undergo alternate ways as they elute. Because different paths are of different lengths, this causes broadening and decreases the number of theoretical plates. Eddy diffusion can be minimized by using the column with regulary sized particles and by decreaseing the mobile phase flow rate.
Longitudinal diffusion or flow distribution results becausethe flow of the mobile phase is faster in the center of the channel between the particles than at the edges (Figure 2.18).
The analyte diffuses from the center which also causes broadening. The longer the solute molecules spend in the column the more pronounced the longitudinal diffusion becomes.
Therefore, unlike with eddy diffusion, the longitudinal diffusion effect can be reduced by utilizing fast flow rates of mobile phase.
Figure 2.18. Flow distribution scheme for a longitudinal diffusion (Reprinted with permission from reference 98).
Resistance to mass transfer occurs due to the various sizes of channels in the pore (Figure
2.17). The time it takes for the analyte molecules to be equilibrated between the stationary and mobile phases can be longer compared to the flow rate of the mobile phase. When analyte molecules interact with the stationary phase particles, they might 42 spend longer or shorter time in the stationary phase than the time needed for actual equilibration between the two phases. Some analyte molecules will be swept away before equilibration and some will be left behind, causing band broadening. Moreover, the pores of the stationary phase particles are filled with mobile phase that does not move
(―stagnant‖), through which the molecules have to diffuse before they can reach the stationary phase. Molecules which diffuse longer inside the pores are left behind by those that diffused only shortly into the pore or bypassed the particle. As the flow rate of the mobile phase increases the broadening effect becomes larger. Generally, particles with small pores should be used for the stationary phase to decrease the effect of slow diffusion and band broadening. Usually, a standard 100 Å pore size is used for analytes of molecular weight of 3000 Da and less.
The three different mechanisms disscussed above are affected by the flow rate of the mobile phase. The longitudinal diffusion is reduced with a high flow rate while eddy diffusion and mass transfer effects are reduced with a low flow rate. Therefore, optimum conditions need to be established for the flow rate in order to keep the dispersion from all three causes to a minimum. The minimum dispersion can be predicted by plotting the number of theoretical plates (N), which measures dispearsion, against the flow rate of the mobile phase in the column. This is known as the Van Demeter plot which contains a hyperbolic function and predicts the optimum flow rate for the mobile phase that will deliver minimum dispersion and maximum separation efficacy.97, 98
43
CHAPTER III
MATERIALS AND INSTRUMENTATION
3.1 Materials
For the non-covalent complexes discussed in chapter 4, PEI 400 and PEI 800 were acquired from Sigma/Aldrich (Milwaukee, WI) and BASF (Ludwigshafen,
Germany), respectively. Aqueous ammonium hydroxide (28-30%) was obtained from
EMD Chemicals (Gibbstown, NJ). HPLC-grade water, ammonium acetate, and the single stranded pentadeoxynucleotides 5'-d(TTTTT)-3', 5'-d(CCCCC)-3', 5'-d(AAAAA)-
3', 5'-d(GGGGG)-3' and 5'-d(GCGAT)-3' were purchased from Fisher (Fair Lawn, NJ), as were all solvents. All chemicals were used as received without further purification.
For ternary complexes discussed in chapter 5, L-glutamic acid was purchased from Sigma-Aldrich (Milwaukee, WI). Glutamic acid dipeptide (EE) was a gift from
MPBiomedicals (Solon, OH). Single stranded oligodeoxynucleotides, HPLC-grade water, and ammonium acetate were purchased from Fisher (Fair Lawn, NJ). Ammonium hydroxide (28-30%) was obtained from EMD Chemicals (Gibbstown, NJ). All chemicals were used as received without further purification.
For the study reported in chapter 6, black raspberry fractions were provided from the Ohio Agricultural Research and Development Center, Wooser in collaboration with
Professor Peter L. Rinaldi (The University of Akron). The samples were stored at -80˚ C before analysis. All samples were dissolved in the appropriate solvent, and the solutions
44 were filtered through a 0.2 uM filter to remove remaining solids prior to their introduction to the LC-MS system.
The detailed sample preparation procedures, together with the corresponding instrumental conditions, are provided in each chapter.
3.2 Instrumentation
For the studies reported in chapters 4 and 5, the Waters Synapt HDMS quadrupole time-of-flight (Q/TOF) mass spectrometer and electrospray ionization (ESI) were utilized. Q/TOF and T-wave ion mobility components are interfaced in this instrument, as has been explained in sections 2.17 and 2.3.4, respectively. Specific instrumental conditions are given in individual chapters.
Figure 3.1.Background pressures in various segments of the Waters Synapt HDMS Q/TOF mass spectrometer Waters Corporation 2007 (Instrument picture reproduced with permision from reference 94).
45 Figure 3.1 displays the location in the instrument having backing pressure. It is the pressure in the extraction region between the ESI source and the quadrupole. For the work in chapter 5, optimization of the backing pressure was essential. This was done by manually adjusting the pump.
The study reported in chapter 6 employed an Esquire –LC (BrukerDaltonics,
Billerica, MA) quadrupole ion trap (QIT) equipped with electrospray ionization (ESI).
Figure 3.2.shows the pressures in the different sections of this instrument.
focusing drying gas skimmers octapole lenses
needle
plate capillary entrance
Figure 3.2. Diagram of the Bruker Daltonics Esquire-LC ESI-MS instrument (Reprinted with permision from reference 88).
In direct infusion mode, a solution of the sample was introduced to the ESI source at the rate of 250 uL/min via the syringe pump. The ESI chamber was operated at atmospheric pressure. The spray needle was grounded. The capillary potential was set to -3.5 kV or -
4.0 kV and the plate offset (potential difference between the plate and the capillary
46 entrance) at 0.5 kV; this potential difference creates the high electric field within the chamber. The pressure of the nebulizing gas (nitrogen) was set at 10 psi. The flow of the drying gas, which was also nitrogen, was 8 L/min and the drying gas temperature was set at 250 ˚C.
The electrostatic field inside of the ESI chamber allows the ions to move inside the glass capillary. Heat, provided by the heated drying gas flowing around the capillary, removes the extra solvent molecules before the ions enter the ion guide region. The ion guide region consists of two skimmers, an octapole and two lenses. These are controlled by voltages that allow the successful transfer of ions to the ion trap. The target mass,which controls the skimmer and lens voltages, was varied between m/z 300-700, depending on the sample analyzed.
Figure 3.3. Diagram of Agilent HPLC instrument (Reprinted with premission from reference 88).
47 For the work in chaper 6, an Agilent HP 1100 HPLC system was interfaced with the ESI-
QIT mass spectrometer (Figure 3.3). The main components of the system include solvent cabinet, degaser, binary pump, autosampler, column compartment and UV- detector. The UV detector was not used for the work in this dissertation. The role of the binary pump is to combine the solvents and allow analyte to elute through the column.
For sample introduction, a guard coulumn was placed before the LC column to remove any impurities present before LC analysis. The LC column was connected to the mass spectrometer using tubing that had a microsplitterattached. The role of a microspliter was to reduce the flow coming from the column before it entered into the ESI source.
48
CHAPTER IV
NON-COVALENT COMPLEXES BETWEEN POLY(ETHYLENE IMINE) AND
SINGLE STRANDED DEOXYNUCLEOTIDES
4.1Introduction
Gene therapy involves the delivery of appropriate genetic material to target cells, where it may replace defect genetic material, inhibit the production of a deleterious protein or cause the production of a therapeutic protein.99 Cationic polymers, such as poly(ethylene imine) (PEI), are increasingly explored as delivery vehicles (vectors) of genes or oligonucleotides to cells, as they are less cytotoxic, less costly, and more easily prepared than traditionally used viral vectors.99,100,101,102,103,104 Cationic polymers are positively charged at physiological pH and, thus, can develop attractive electrostatic interactions with nucleic acids and oligonucleotides, which generally carry negative charges.99 The resulting complexes have been termed polyplexes and usually have a net positive charge, which helps them to penetrate the negatively charged cell membrane and escape degradation until they reach their ultimate destination, the cell nucleus.99,100,103
Although a number of studies have been reported about the cytotoxicity, transfection efficiency, and degradability of polyplexes,99,100,103information about their compositions and binding interactions at the molecular level is scarce.102,105This issue is addressed here with the first characterization of PEI/oligodeoxynucleotide (ODN) polyplexes by electrospray ionization mass spectrometry (ESI-MS),106 tandem mass
49 spectrometry (MS2) and traveling wave ion mobility mass spectrometry (TWIM-MS).107
The stoichiometry and charge state distributions of the complexes formed by PEI 400 and
PEI 800 with a series of single-stranded pentadeoxynucleotides are examined, as are the corresponding complex stabilities in solution and the solvent-free environment.
4.2 Experimental Methods
A 10-mM ammonium acetate buffer was prepared in HPLC-grade water and its pH was adjusted to 7.2 by adding a few droplets of aqueous ammonium hydroxide. This buffer was used to prepare individual 0.5-mM solutions of the polymers and pentadeoxynucleotides. Polymer and nucleotide solutions were mixed in the ratios of
1:10, 1:5, 1:1, 5:1 or 10:1, and the resulting mixtures were introduced into the ESI source by direct infusion at a flow rate of 5 μL/min.
4.2.1 Mass Spectrometry Experiments
All experiments were performed with a Synapt HDMS quadrupole/time-of-flight
(Q/TOF) tandem mass spectrometer (Waters, Beverly, MA) equipped with ESI and
TWIM-MS capabilities.107,108 The instrument contains a triwave device between the Q and TOF mass analyzers, consisting of three cells arranged in the order trap cell, ion mobility cell and transfer cell. ESI mass spectra were acquired in positive mode by setting the spray voltage at 3.5 kV, the source temperature at 110 oC, the sampling cone voltage at 35 V, the extraction cone voltage at 3.2 V, the desolvation gas flow rate at 400
o L/h (N2), the desolvation temperature at 160 C, the quadrupole mass analyzer in RF-only
50 mode, the trap cell collision energy at 4.0 eV and the transfer cell collision energy at 6.0 eV. Under these conditions, all ions leaving the ESI source pass the quadrupole and triwave regions and are orthogonally accelerated into the TOF mass analyzer for m/z analysis. For the acquisition of MS2 spectra, a specific PEI-ODN complex (precursor ion) was mass-selected with Q using an isolation width of 4.5 and underwent collisionally activated dissociation (CAD) with Ar in the trap cell at an Ar gas flow of 1.5 mL/min.
MS2 (CAD) spectra were measured as a function of trap collision energy, which was varied within 6-45 eV to induce fragmentation. The product ions formed in this process were subsequently mass-analyzed by the ToF part.
Fragmentation efficiency curves were constructed from MS2 spectra of [M +
2H]2+ ions by plotting the relative abundance of the selected PEI-ODN complex versus
109 the corresponding center-of-mass collision energy (Ecm). Relative abundance was calculated by dividing the complex intensity, I(complex2+), by the sum I(complex2+) + ½
+ + 110,111 [I(PEI ) + I(ODN )]. Elab was calculated from the applied laboratory-frame collision energy (Elab) using the equation Ecm = Elab x (mAr / (mAr + mprecursor)), where mAr and mprecursor designate the masses of Ar atoms and the precursor ion, respectively. The points were fitted into sigmoid curves, constructed using the Origin 8.1 graphing software,112 in order to deduce the corresponding E50 values, which are the collision energies at which
110,111,113 50% of the PEI-ODN precursor ions were depleted due to CAD. The E50 value derived for each PEI-ODN complex is an average of three measurements for the doubly charged and two measurements for the singly charged complexes.
The normalized intensities of the PEI-ODN complexes were used to determine the effect of solution phase composition on the stoichiometry of the resulting non-covalent
51 complexes and the relative solution stabilities of the complexes.111,114,115,116 Normalized intensities were obtained from peak heights by dividing the PEI-ODN complex intensity by that of uncomplexed ODN. The most abundant PEI-ODN oligomers (4 for PEI 400 and 12 for PEI 800) were used in this calculation. All detectable charge states of the PEI-
ODN complexes and uncomplexed ODNs were considered. The uncomplexed ODNs dissociated partly during ESI analysis. The ODN fragment intensities were added to those of intact ODN for the calculation of normalized intensities.
Two-dimensional TWIM-MS plots were acquired on all ions exiting the ESI source (Q in RF-only mode), using a traveling wave velocity of 675 m/s and a traveling wave height of 17.2 V. The IM gas (N2) flow was set at 22.7 mL/min and the trap and transfer cell potentials were kept at 6.0 and 4.0 V, respectively, during the IM measurement.
4.3 Polyplexes with PEI 400
The poly(ethylene imine)s used for polyplex formation were supplied as mixtures of linear and branched oligomers; both have a 43-Da repeat unit, amine end groups and
+ the nominal composition H2N(CH2CH2NH)nH. The [M + H] ions of such oligomers give rise to the main distribution in the ESI mass spectrum of PEI 400, which appears m/z
43n + 18 (labeled by * in Figure 4.1a). A second distribution results from oligomers that are missing NH3; these are observed at m/z 43n +1 (& in Figure 4.1a) and agree well with cyclized structures, which are typical byproducts in PEI syntheses. Two more distributions are detected, mainly in the low-mass range; these correspond to PEI fragments formed during ESI (+ and = in Figure 1a).117 The average molecular weight 52 (Mn) calculated from the ESI mass spectrum for the main polymer distribution, i.e. for
H2N(CH2CH2NH)nH, is 295; hence, the average PEI 400 oligomer contains 6-7 repeat units and 7-8 N atoms (potential protonation and hydrogen bonding sites).
Figure 4.1.ESI mass spectra of (a) PEI 400 and (b,c) the PEI-ODN polyplexes formed after mixing PEI 400 and 5'-d(TTTTT)-3' in the molar ratios 1:1 (b) and 5:1 (c). The signs * and & designate [M + H]+ ions of linear and cyclic PEI oligomers with the compositions H2N(CH2CH2NH)nH and (CH2CH2NH)n, respectively. The signs = and + designate PEI fragments with the compositions H N(CH CH NH) CH CH NH+=CHCH + 2 2 2 m 2 2 3 and CH3CH=N(CH2CH2NH)mCH2CH2NH =CHCH3, respectively. The signs # and ^ designate PEI-d(TTTTT) polyplexes containing linear (#) or cyclic (^) PEI and having the PEI-to-ODN stoichiometry 1:1 (PN). The signs ~and !designatepolyplexes with the stoichiometries PN2 (~) or P2N (!). The superscripted charges indicate degree of protonation. Monoisotopicm/z values are given on top of select peaks. The ions at m/z 1459.3 and 730.1 arise from singly and doubly protonated d(TTTTT), respectively, and + 2+ the ions at m/z 1235.2 and 616.1 are the corresponding w4 /w4 fragments.
PEI 400 was mixed with the pentadeoxynucleotides 5'-d(TTTTT)-3', 5'- d(CCCCC)-3', 5'-d(AAAAA)-3', 5'-d(GGGGG)-3' and 5'-d(GCGAT)-3' in the molar
53 ratios 10:1, 5:1, 1:1, 1:5 and 1:10. After the mixtures were allowed to equilibrate for 10 minutes, ESI mass spectra were acquired to identify the polyplexes formed. With all
ODNs and molar ratios examined, the PEI-ODN polyplex with 1:1 stoichiometry is the dominant product, as attested in Figures 4.1b and 4.1c for the PEI/d(TTTTT) complex formed from polymer-to-ODN ratios of 1:1 and 5:1, respectively. The 1:1 stoichiometry is observed in charge state +2, as [M + 2H]2+, as well as in charge state +1, as [M + H]+, with the latter increasing in relative abundance at higher polymer-to-ODN molar ratios
(cf. Figure4.1). For brevity, the notations PN2+ and PN+ will be used for the polyplexes, with P and N representing the polymer and oligodeoxynucleotide, respectively, and the superscript providing the corresponding charge state. At higher ODN concentrations, doubly and triply protonated complexes with the stoichiometry PN2 appear as minor product (Figure 4.1b). Inversely, a higher polymer concentration coproduces small
2+ + amounts of P2N /P2N complexes (Figure 4.1c).
Both the major PEI distribution with the nominal composition
H2N(CH2CH2NH)nH (* in Figure4.1a) as well as the minor, cyclized PEI distribution (& in Figure 4.1a) form complexes (# and ^, respectively, in Figures 4.1b and 4.1c). No complexes are observed with the PEI fragment series (+ and = in Figure 4.1a). The fraction of polyplexes that contain cyclic PEI (^) is somewhat smaller than the fraction of cyclic structures in PEI (&), pointing out that cyclic oligomers yield less stable complexes with ODNs.
54 Table 4.1 PEI-ODN (PN) Polyplexes Formed Between PEI 400 and Single-stranded Pentadeoxynucleotides
PEI/ODN mixing ratio 1:1 5:1
Complex stoichiometry 1:1 1:2 1:1 2:1
a b a c ODN (N) [PN]/[N] [PN2]/[N] [PN]/[N] [P2N]/[N]
5'-d(TTTTT)-3' 1.90 0.35 3.89 0.27
5'-d(CCCCC)-3' 0.41 0.09 1.06 0.25
5'-d(AAAAA)-3' 0.08 0.06 0.26 0.06
5'-d(GGGGG)-3' 0.55 0.18 1.17 0.63
5'-d(GCGAT)-3' 0.42 0.18 0.83 0.19
a 2+ + ([PN ] + [PN ]) / [N]free ([N]free is the sum of the intensities of all uncomplexed ODN species, viz. N2+, N+ and their 2+/1+ fragments); 10%. The complexation reaction → P + N ← PN is associated with an equilibrium constant (binding affinity) Kb = [PN] / ([N][P]), with [N] and [P] representing the molar concentrations of free ODN and PEI, respectively. Rewriting this equation as [PN]/[N] = Kb[P] shows that the [PN]/[N] concentration ratio is proportional to both the binding affinity Kb (a measure of the complex stability in solution) as well as the concentration of free polymer, [P], which increases with the total concentration of polymer (cf. Figure 4. 2). b 2+ 3+ ([PN2 + [PN2 ]) / [N]free; 20%. c + 2+ ([P2N ] + [P2N ]) / [N]free; 20%.
Table 4.1 summarizes the relative intensities of the polyplexes obtained from mixtures
with polymer-to-oligonucleotide molar ratios of 1:1 and 5:1. Only polyplexes containing
the major PEI distribution were considered in the calculation of these intensities. The
yield of the major product, viz. the PN complex with 1:1 polymer-to-oligonucleotide
stoichiometry, increases significantly with the molar PEI/ODN ratio of the reactants, as
expected. At all mixing ratios, the most abundant PN complex is observed for
d(TTTTT), with the relative intensities of the other complexes following the order PEI- 55 d(TTTTT) > PEI-d(GGGGG) ≈ PEI-d(CCCCC) ≈ PEI-d(GCGAT) > PEI-d(AAAAA).
The relative intensities of the polyplexes correlate linearly with the PEI concentration; this is shown in Figure 4.2 for the complex of d(TTTTT). Such relationship reveals that the ESI mass spectra provide snapshots of the PN/N solution concentrations and that the relative intensities of the PEI-ODN complexes reasonably approximate the corresponding relative solution stabilities (cf. footnote a of Table 1). Based on this fact, the relative PN intensities in Table 1 indicate that d(TTTTT) produces the most stable and d(AAAAA) the least stable 1:1 polyplex with PEI, while the other pentadeoxynucleotides yield polyplexes of intermediate stability.
The most stable PN2 complex is also formed with d(TTTTT), but the most stable
P2N complex is formed with d(GGGGG), cf. Table 1. Such a change in the order of polyplex stabilities could result from a change in the secondary ODN structure when a further constituent is added to the non-covalent complex.
Figure 4.2 Relative intensity of the PEI-d(TTTTT) polyplexes with polymer-to-ODN stoichiometry of 1:1 (PN) vs. the total PEI 400 concentration in the PEI/ODN solution.
56 It is noteworthy that the molecular weight distribution of the polyplexes is narrower than that of PEI 400 (cf. Figure 4.1), indicating that only a select range of PEI sizes can form stable complexes with pentadeoxynucleotides. The number of ethylene imine repeat units in the four most abundant PN2+ ions is 4-7 for PEI-d(TTTTT), 5-8 for
PEI-d(GGGGG), PEI-d(CCCCC) and PEI-d(GCGAT) and 6-9 for PEI-d(AAAAA), while the four most abundant ions within the corresponding PN+ series carry 4-7 repeat units in all cases. From these data, the N-atom/P-atom ratio of the polyplexes can be calculated, which is defined as the number the N atoms in the PEI component divided by the number of P atoms in the nucleic acid component. Considering that non-cyclic PEIs contain one more N atom than their number of repeat units and that the number of P atoms in the ODNs studied is four, the N-atom/P-atom ratios of the polyplexes (PN stoichiometry) fall within the range (5-9)/4 = 1.3-2.3. For the in vivo or in vitro delivery of nucleic acids and oligodeoxynucleotides, N-atom/P-atom ratios of 5-10 or larger are generally used, which maximize polyplex formation according to agarose gel electrophoresis analyses.100,101,103,118 The excess PEI, as compared to the complex stoichiometry, shifts the complexation equilibrium toward the polyplex and adds more positive charge which is believed to facilitate transfection through the cell membrane.99
4.4 Comparison of PEI 400 vs. PEI 800
The ESI mass spectrum of PEI 800 (Figure 4.3a) shows very similar characteristics to those described for PEI 400 (vide supra). The major distribution, appearing at m/z 43n + 18, originates again from [M + H]+ ions of amine-terminated oligomers which, as mentioned above, contain mixtures of linear and branched chains 57 (each branch converts one CH2CH2NH subunit to CH2CH2N and another one to
CH2CH2NH2, which does not alter the overall PEI composition).
Figure 4.3 ESI mass spectra of (a) PEI 800 and (b,c) the PEI-ODN polyplexes formed after mixing PEI 800 and 5'-d(TTTTT)-3' in the molar ratios 1:1 (b) and 5:1 (c). The signs * and ‡ designate [M + H]+ and [M + 2H]2+ ions, respectively, of linear PEI + oligomers with the composition H2N(CH2CH2NH)nH; the sign & designates [M + H] ions of cyclic PEI oligomers with the composition (CH2CH2NH)n. The signs = and + + designate PEI fragments with the compositions H2N(CH2CH2NH)mCH2CH2NH =CHCH3 + and CH3CH=N(CH2CH2NH)mCH2CH2NH =CHCH3, respectively. The sign # designates PEI-d(TTTTT) polyplexes with the linear PEI oligomers and the PEI-to-ODN stoichiometry 1:1 (PN). The sign ~ designates polyplexes with the stoichiometry PN2. The superscripted charges indicate degree of protonation. Monoisotopicm/z values are given on top of select peaks. The ions at m/z 1459.3 and 730.1 arise from singly and doubly protonated d(TTTTT), respectively, and the ion at m/z 1235.2 is the + corresponding w4 fragment. At the higher polymer concentration (bottom), heavier PEI oligomers (*) dominate in the low-mass end of the spectrum.
58 The proportion of cyclized oligomers, observed at m/z 43n + 1, is smaller with PEI 800 than PEI 400; on the other hand, the larger poly(ethylene imine) generates a sizable [M +
2H]2+ distribution (~40% of the [M + H]+ distribution), which was near noise level in the
ESI mass spectrum of PEI 400 (cf. Figure 4.1a). The average molecular weight (Mn) calculated from the singly and doubly protonated oligomers for the main polymer distribution, viz. H2N(CH2CH2NH)nH, is ~530, indicating that the average PEI 800 oligomer carries ~12 repeat units and ~13 N atoms.
It should be noted at this point that the Mn values obtained from the mass spectra are lower than those reported by the supplier (400 or 800), which were measured by gel permeation chromatography (GPC). Aggregation due to hydrogen bonding could cause the overestimation in the GPC results.119,120
The effect of increasing the molecular weight of the poly(ethylene imine) carrier was tested with d(TTTTT). Figures 4.3b and 4.3c show the ESI mass spectra obtained if polymer and ODN are mixed in molar ratios of 1:1 and 5:1, respectively. As with PEI
400, the polyplex with 1:1 stoichiometry (PN) is the predominant (Figure 4.3b) or solely detected (Figure 4.3c) product. The PN complexes of PEI 800 form mainly doubly protonated PN2+ ions (as with PEI 400) along with some singly charged PN+ and triply charged PN3+. The relative intensity of PN, calculated by adding the intensities of the PN complexes in all charge states and dividing the resulting sum by the intensity of free
ODN and its fragments in all charge states, increases from 4.39 to 9.78 when the polymer-to-oligodeoxynucleotide ratio is increased from 1: 1 to 5:1; the corresponding values with PEI 400 were 1.90 and 3.89, respectively. The larger PN/N intensity ratios with PEI 800 indicate a higher stability for the polyplexes with PEI 800, most likely
59 because the higher number of basic and hydrogen bonding sites in the larger polymer leads to stronger non-covalent interactions and higher binding affinities.
PN complexes with 4-25 ethylene imine repeat units are detected in the spectra of
Figure 4.3. This range is comparable with the range of poly(ethylene imine) n-mers observed in the ESI mass spectrum of PEI 800 (n = 2-27). Thus, essentially the entire linear/branched PEI distribution reacts to form polyplexes. The average molecular
2+ weight (Mn) of the PEI attached to d(TTTTT), calculated from the PN distributions in
Figure 4.3, is 480 when polymer and pentanucleotide are combined in the ratio of 1:1
(Figure 4.3b) and 570 when the combination ratio is 5:1 (Figure 4.3c). An increased PEI concentration (i.e. 5:1) favors disproportionally complexation with the heavier oligomers because of their higher binding affinities to d(TTTTT).
Interestingly, the cyclic components of PEI 800 do not react with d(TTTTT) to any appreciably degree (Figure 4.3), in contrast to the cyclic components of PEI 400 which reacted (Figure 4.1), albeit with an overall lower yield than the corresponding linear oligomers (vide supra). The increased binding affinities of larger linear/branched oligomers (more of them are present in PEI 800) and the poorer binding affinities of cyclic structures (less are present in PEI 800) reconcile these differences.
4.5 Intrinsic Stability of Polyplexes
The single-stage ESI mass spectra discussed thus far reveal information about the relative solution stabilities of the polyplexes which, in turn, depend on the corresponding intrinsic (i.e. gas-phase) stabilities and on solvation effects. Intrinsic stabilities were independently assessed by tandem mass spectrometry (MS2) experiments on PEI-ODN 60 complexes with PEI 400. Specifically, PN2+ ions containing the 5-mer (i.e. P =
H2N(CH2CH2NH)5H) were isolated and induced to decompose by collisionally activated dissociation (CAD). During this process, all PN2+ precursor ions examined undergo one major dissociation to yield their P+ and N+ constituents, as exemplified in Figure 4.4 by the MS2 (CAD) spectra of PEI-d(TTTTT) and PEI-d(GCGAT).
Figure 4.4MS2 (CAD) mass spectra of PN2+polyplex ions containing (a) d(TTTTTT) and (b) d(GCGAT). The oligomers selected (m/z 846.7 and 868.2, respectively) contain 5 ethylene imine repeat units. The laboratory-frame collision energies were (a) 14 eV and (b) 13 eV.
The MS2 (CAD) spectra of PN2+ were acquired as a function of the center-of-mass collision energy (Ecm), in order to construct fragmentation efficiency curves. Figure 4.5 shows three examples, referring to the polyplexes of d(TTTTT), d(GCGAT) and d(AAAAA). The collision energies causing 50% of the PN2+ precursor ions to dissociate, termed E50 energies, represent a measure of the corresponding intrinsic polyplex
61 stabilities.The E50 values of the five PEI-ODN complexes studied (Table 4.2) indicate the intrinsic stability order PEI-d(TTTTT) > PEI-d(GGGGG) ≈ PEI-d(CCCCC) ≈ PEI- d(GCGAT) > PEI-d(AAAAA), which is identical with the solution stability order of these polyplexes. Such agreement strongly suggests that the PEI-ODN complexes have similar structures in the gas phase and in solution (at physiological conditions).
Figure 4.5 Fragmentation efficiency curves (relative abundance of PN2+ precursor ion vs. center-of-mass collision energy) of polyplexes containing a PEI with 5 repeat units and d(TTTTT), d(AAAAA) or d(GCGAT).
62 2+ Table 4.2 E50 Values of PN Polyplexes Between PEI 400 (P) and Single-stranded Pentadeoxynucleotides (N)
a ODN (N) E50 (eV)
5'-d(TTTTT)-3' 0.35
5'-d(CCCCC)-3' 0.29
5'-d(AAAAA)-3' 0.21
5'-d(GGGGG)-3' 0.29
5'-d(GCGAT)-3' 0.29 a0.04.
From the singly charged PN+ distributions, only the polyplexes of d(TTTTT), d(AAAAA) and d(CCCCC) were sufficiently intense to produce MS2 spectra with useable signal-to-noise ratios (Figure 4.6). The complex of d(TTTTT) dissociates exclusively to P+, whereas the complexes of d(AAAAA) and d(CCCCC) lead to mixtures of P+ and N+. Because of the absence of charge repulsion in PN+, these ions require higher collision energies for dissociation into their constituents than PN2+ ions.
Additionally, competitive and consecutive reactions become possible within the ODN component of PEI-d(CCCCC), cf. Figure 4.6c. Due to these complications, energetics data were not derived from the singly protonated polyplexes.
63
Figure 4.6 MS2 (CAD) mass spectra of PN+ polyplex ions containing (a) d(TTTTTT), (b) d(AAAAA) and (c) d(GGGGG). The oligomers selected (m/z 1691.6, 1736.6 and 1616.5, respectively) contain 5 ethylene imine repeat units. The laboratory-frame collision energies were (a,b) 38 eV and (c) 39 eV. The top and, especially, the bottom spectrum show a distribution of polymer ions, indicating admixtures in the selected PN+ precursor ions (see ion mobility section).
4.6 Ion Mobility Separation of the Polyplexes
Ion mobility mass spectrometry (IM-MS) can separate ions according to their mass, charge and shape.121,122,123,124,125,126,127,128,129,130,131,132 Charge- and shape-sensitive dispersion permits the separation of isomers and isobars and deconvolutes the isotope patterns of overlapping charge states, enhancing the resolving power and sensitivity of mass spectrometry. Traveling wave ion mobility mass spectrometry (TWIM-MS), 107,108 a recent variant of IM-MS, was employed to corroborate the presence of the polyplexes observed in the ESI mass spectra and identify any additional complex stoichiometries hidden under the dominating products. These experiments were performed with the PEI
64 400/d(TTTTT) system (polymer-to-ODN molar ratio 5:1), which yields the highest absolute intensities with the best signal-to-noise ratio.
Figure 4.7 shows a two-dimensional TWIM-MS plot (m/z vs. drift time) for all ions exiting the ESI source. After IM separation through the traveling wave IM cell and m/z analysis by the orthogonal TOF analyzer, several families of ions could be identified, which have been marked 1-11. Figure 4.8 shows the spectra of the regions which are not readily detected in the ESI mass spectrum because they have very small intensities and/or overlap with more abundant ions or other charge states.
Figure 4.7.Two-dimensional ESI-TWIM-MS plot (m/z vs. ion drift time) of the polyplexes formed after mixing PEI 400 and 5'-d(TTTTT)-3' in the molar ratio 5:1. The regions marked by the encircled arrows correspond to: 1, protonated PEI (P+); 2, singly protonated d(TTTTT), N+, and w + fragment; 3, doubly protonated d(TTTTT), N2+, and 2+ + 4 2+ 2+ w4 fragment; 4, PN polyplexes; 5, PN polyplexes; 6, PN2 polyplexes; 7, PN 3+polyplexes; 8, P N+polyplexes; 9, P N 2+polyplexes; 10, PN3+polyplexes; 11, 2 4+ 2 2 2 P4N3 polyplexes. The four trend lines shown connect regions with identical charge states (+1 to +4 from right to left).
65 Regions 1-3 contain the reactants; PEI is observed in region 1 (distributions as in
Figure 4.1a) and the ODN and its w4 fragment in regions 2 (charge state +1) and 3
(charge state +2). All other areas contain non-covalent complexes, with the major PN polyplex appearing in region 4 (PN+ distribution) and region 5 (PN2+ distribution). The
2+ 3+ minor products PN2 and P2N are detected in regions 6/7 and 8 in the form of PN2 /PN2
+ and P2N distributions, respectively. All these ions were clearly discerned in the ESI mass spectrum of Figure 4.1c. IM separation enables the detection of three further minor products in regions 9-11. The corresponding mass spectra confirm the compositions
2+ 3+ 4+ P2N2 for region 9 (Figure 4.8a), PN3 for region 10 (Figure 4.8b) and P4N3 for region
11 (Figure 4.8c).
Figure 4.8.Mass spectra of regions 9 (a), 10 (b) and 11 (c) in the ESI-TWIM-MS diagram of Figure 4.7. The m/z ratios and isotope distributions agree well with the compositions 2+ 3+ 4+ P2N2 , PN3 and P4N3 , respectively. The distributions labeled by % contain one cyclic and one non-cyclic PEI oligomer; all other distributions contain only non-cyclic PEI oligomers. The abundances of m/z 1691.5 (Figure 4.8a), m/z 1536.7 (Figure 4.8b) and m/z 1563.7relative to that of m/z 846.3 (Figure 4.1c) are 2.0%, 0.5% and 0.3%, respectively.
66 Ion drift times through the IM cell decrease with the number of charges due to the greater mobilities of the higher charge states in the traveling wave field. Different charge states follow distinct trend lines that are clearly separated from each other. Note
+ 2+ that the PN distribution of the major polyplex is superimposed with the minor P2N2 distribution (cf. Figures 4.1c and 4.8c), which explains the appearance of several P+ fragments in the MS2 spectra of mass-selected PN+ oligomers (cf. Figure 4.6).
4.7 Conclusions
This study represents the first microstructure characterization of polyplexes composed of poly(ethylene imine) and oligodeoxynucleotides by ESI mass spectrometry.
For the polymer and ODN sizes investigated, the polyplex with 1:1 polymer-to-ODN stoichiometry, PN, is the principal product independent of the reactant mixing ratio.
Complexes with the stoichiometries PN2 and P2N are formed as minor products, their yields increasing with the concentration of oligonucleotide and polymer, respectively.
Other compositions (such as P2N2 and PN3) are formed in trace amounts that are detectable only after charge state deconvolution using TWIM-MS.
The relative intensities of the polyplex ions in ESI mass spectra are shown to reflect the corresponding polyplex solution stabilities; a similar relationship has been reported for the relative intensities of DNA duplexes111,115,116 and ODN-drug conjugates.114 The solution and gas-phase stabilities of the polyplexes studied follow identical orders, consistent with very similar structures in both phases. With the polymer systems studied (PEI 400 and 800), thymine-rich nucleotides give rise to the most stable and adenine-rich nucleotides to the least stable polyplexes. This selectivity suggests that 67 polymer structure and size may be tuned to favor complex formation with specific nucleic acid sequences.
68
CHAPTER V
TERNARY COMPLEXES OF POLY(ETHYLENE IMINE), SINGLE STRANDED
OLIGONUCLEOTIDES AND GLUTAMIC ACID MOIETIES
5.1 Introduction
Poly(ethylene imine), PEI, is a cationic polymer that has been extensively researched for non-viral gene transfection in vitro and in vivo.133,134 The complexes between PEI and nucleic acids, polyplexes, have shown a high degree of transfection efficiency in serum free medium. However, a drawback of this polycationic carrier is its significantly lower efficiency in the presence of serum, which downgrades the in vivo application of these cationic vectors for gene delivery.135,136,137 One of the reasons is that positively charged polyplexes bind non-specifically to negatively charged components in serum or blood, thus preventing successful gene therapy.138,139 In order to protect the stability of DNA/PEI complexes from non-specific interactions in the biological environment, several approaches have been reported in which these polymer-based systems are chemically modified to improve the biocompatibility for in vivo applications.137,139,140 One of the altervative ways to modify these cationic systems is to electrostatically bind anionic polymers to their surface. Polyglutamic acid (PGA) anionic polymer, with its biological properties of nontoxicity and biocompatibility, when added to polyplexes was found to significantly increase the transfection efficiency of polyplexes in the presence of serum.141,142
69 Various studies have reported on the benefits of high transfection efficiency and low cytotoxicity of the terplex and examined the interactions among its components by techniques such as circular dichroism spectroscopy or agarose gel electrophoresis.141,143
This study describes the first investigation of PEI/oligodeoxynucleotide terplexes with glutamic acid (Glu) or GluGlu dipeptide by electrospray ionization mass spectrometry
(ESI-MS), tandem mass spectrometry (MS2) and ion mobility mass spectrometry (IM-
MS). For brevity, these complexes will be symbolized by PNE or PN(EE), where
P=polymer, N=ODN, E=Glu and EE=GluGlu. The binding interaction between the components of the terplexe systems, as well as the preferred terplex stoichiometries and charge state distributions are examined for systems formed from PEI 400, a series of single-stranded pentadeoxynucleotides and either Glu of GluGlu (see below).
Glu GluGlu
O OH O O O OH HO HO OH N O H NH2 O NH2 C5H9NO4 C10H16N2O7 Exact Mass: 147.05 Exact Mass: 276.10
5.2 Materials and Methods
Glu or GluGlu, polymer and pentadeoxynucleotides were dissolved in 10-mM ammonium acetate buffer of pH 7.4. Individual 0.5-mM solutions were prepared for PEI and the oligodeoxynucleotides (ODNs), mixed in the ratio of 10:1(v/v) and incubated for
70 10 minutes. Solutions with different concentrations of glutamic acid (E) or glutamyl glutamic acid (EE) were then added to this mixture so that the final mixing ratio was
10:1:1 (v/v/v) PEI/ODN/E or PEI/ODN/EE. The resulting final mixtures were incubated for additional 10 minutes before introduction into the ESI source by direct infusion at a flow rate of 5 uL/min.
5.2.1 Mass Spectrometry Experiments
All experiments were carried out on a Synapt HDMS quadrupole/time-of-flight
(Q/ToF) tandem mass spectrometer (Waters, Beverly, MA) equipped with ESI and
TWIM-MS capabilites.107 ESI mass spectra were aquired in positive mode utilizing the following settings: spray voltage 3.5 kV, sampling cone voltage 35 V, extraction cone voltage 3.2 V, source temperature 100 °C and desolvation temperature 150 °C. The desolvation gas flow rate was 400L/h (N2), the trap collision energy was set at 4.0 eV and the transfer cell collision energy at 6.0 eV. A further instrument parameter optimized was the pressure in the interface region (Pi), which was set at 6mbar in order to maintain the non-covalent interactions. Experimental conditions were tuned to maximize the intensity of the +2 charged terplex composed of PEI 400, d(CCCCC) and Glu; the conditions were then applied to all other terplexes. For the MS2 spectra, a precursor terplex ion was isolated with the quadrupole using an isolation width of 4.5 and underwent collisions with Ar in the trap cell; the Ar gas flow was 1.5 mL/min. The collision energy in the MS2 measurements was varied within 1-12 eV to induce fragmentation.
71
The normalized intensities of the PEI/ODN/E or (EE) ternary complexes were obtained from peak heights by summing the intensities of all terplexes observed and dividing this sum by the sum of intensities of uncomplexed species (ODN, PEI, E or EE), and binary complexes (PEI-ODN, ODN-E or ODN-EE, PEI-E or PEI-EE). These normalized intensities were used to assess the solution phase stabilities for the ternary complexes.110,111,116 Reported intensity ratios are the average of two measurements in MS and three measurements in MS2 experiments.
Ion mobility plots were acquired using a traveling wave velocity of 350 m/s and traveling wave height of 11 V. The IM gas (N2) flow was 22.7 mL/min and the trap and transfer cell potentials were kept at 6.0 and 4.0 V during IM measurement.
5.3 Terplexes with PEI400
Self-assembled terplexes were prepared and investigated by ESI mass spectrometry. We previously reported the characterization of polyplexes between PEI and ODNs;144 here we extend this work from binary to ternary cationic systems to gain insight about the structural interactions among the constituents present in such systems and assess their solution and gas phase stabilities.
In order to preserve the non-covalent interactions within the ternary complexes during the ionization process, instrument parameters were optimized so as to maximize ion intensities while minimizing the dissociation of the non-covalent assemblies. For the
72 labile compounds investigated, variations in the acceleration potential (sampling cone voltage) and capillary temperature were found to be less critical if the pressure in the first pumping stage of the mass spectrometer (backing pressure) is kept high;145 thus the key parameter optimized in this study was the backing pressure, (see Figure 3.1) which is the pressure in the extraction region between the ESI source and the first quadrupole mass analyzer.146 A number of studies has shown that this parameter is of significant importance in improving the transmission of non-covalently bound species.146,147,148 The pressure in this part of the instrument affects the internal energy of the ions. Lower pressures increase the time between two successive collisions with ambient gas molecules, giving the ions more time to accumulate energy that later results in dissociation of weak non-covalent complexes. Conversely, increased pressure permits collisional cooling and better transmission of labile ions through the quadrupole and time-of-flight analyzers.149,150
The experimental conditions were tuned to maximize the intensity of the +2 charged terplex composed of PEI 400/5‘-d(CCCCC)-3‘/Glu (Figure 5.1); these conditions were applied to all other terplex ions studied. The mass spectra of the cationic 5‘- d(CCCCC)-3‘ terplex obtained at the lowest backing pressure (2 mbar) and at a high
(optimum) pressure (6 mbar) are compared in Figures 5.1a and 5.1b, respectively. With a low backing pressure, the mass spectrum shows only doubly charged binary complexes between PEI400 and 5‘-d(CCCCC)-3‘(# and ^) and no evidence of ternary complex ions. However, increasing the backing pressure, which was achieved by throttling the pump valve, resulted in the detection of doubly charged PEI 400/5‘-d(CCCCC)-3‘/Glu terplex ions ($in Figure 5.1b). 73 (a)
(b)
Figure 5.1. Expanded view of the m/z 800-940 region of the ESI mass spectra of the PEI 400/5'-d(CCCCC)-3'/glutamic acid terplex aquired at an accelerating voltage of 35 V and backing pressure of (a) 2 mbars and (b) 6 mbars. The signs # and ^designate PEI- d(CCCCC) binary complexes containing linear (#) or cyclic (^) PEI and having 1:1 stoichiometry. The sign $designates PEI-d(CCCCC)-Glu terplexes containing linear PEI oligomers and having a stoichiometry of 1:1:1 PEI-to-ODN-to-Glu.
PEI 400 was first mixed with the pentadeoxynucleotides 5‘-d(TTTTT)-3‘, 5‘- d(CCCCC)-3‘, 5‘-d(AAAAA)-3‘ or 5‘-d(GGGGG)-3‘ in the ratio of 10:1. After these mixtures equilibrated for 10 minutes, glutamic acid solutions of different concentrations
(0.5-5mM) were added. Polymer and oligonucleotide parameters were kept constant as the glutamic acid concentration was varied to optimize the terplexself-assembly. The final mixtures were equilibrated for an additional 10 minutes, before ESI mass spectra were acquired to identify the terplexes formed. The resulting spectra provided information about the types of complexes/terplexes formed, viz. PEI/ODN,
PEI/ODN/Glu, ODN/Glu and/or PEI/Glu, as well as their stoichiometries. The PEI-
74 ODN-Glu terplex with stoichiometry of 1:1:1 is the major ternary complex product, independent of the concentration of the glutamic acid solution added, as seen in Figure
5.2 for the PEI-d(TTTTT)-Glu terplex formed with 1 mM and 5 mM Glu, respectively.
The 1:1:1 ternary complex is observed primarily in +2 charge state, as [M+2H]2+, and with increasing relative abundance as the concentration of glutamic acid is amplified. A greater concentration of glutamic acid in the solution of the ternary complexes, coproduces terplexes in higher stoichiometries; more precisely, PEI/ODN/Glu complexes with 1:1:2, 1:1:3 and 1:1:4 stoichiometries appear as minor products (Figure 5.2c). The relative abundance of these minor terplex distributions decreases as the number of glutamic acid moieties within the assemblies increases. As with PN polyplexes, a portion
2+ 2+ of the terplexes contains cyclic PEI (˅ and <);these are observed as PNE and PNE2 ions (Figure 5.2c). These are always in a lower abundance than terplexes with branched/linear PEI chains, and decrease in intensity with increasing Glu content.
75
PN2+
(a) 846.3
#
867.7
#
730.1
1459.3 2+ 3+ PN2 +
618.1 PN PN
2
# ^ ^ 1235.1
529.1 # #
^ 1575.6
1481.3 1691.5 ^ 1050.7 ~ ~~ ~ ~ ~ # # # #
#
846.3 PNE2+ # (b)
NE+
+ PE2 919.7 2+
+ NE4 2+
NE
5
E #
4 $
# 1459.3 ^ $
^ 1606.2 527.3 # 993.2
730.1 " 589.2 ^ ^ $ 1024.2 ≡ 1575.6
≠ 1097.7 1235.1 # " " ~ ~~ # # 600 900 1200 1500 1800
PNE2+
919.7
$
941.2
$ (c) 910.7
˅
2+
PNE2
962.7 932.7
˅ $
993.2
2+ PNE3 2+ ≠ PNE 4 2+
PNE
923.5 5
984.3
950.7
˅
1014.7
$ " ≠
˅ 1024.2
1097.7 1140.3
" 1066.7 1171.2
≠ 1088.3 1213.8 ˅ $ < ≠ 1048.8 ∞ < 1072.7 ∞ " < ≠ ∞ 1103.8 "
950 1050 1150 m/z
Figure 5.2.ESI mass spectra of PEI 400-d(TTTTT)-Glu terplexesformed by mixing the corresponding solutions in the ratio 10:1:1,using(a) 1mM and (b) 5mM solutions of Glu;(c) expanded view (m/z 900–1200) of the spectrum in part (b). The signs # and ^ represent PN 2+ 3+ binary complexes with linear and cyclic PEI, respectively; binary PN2 and PN2 complexes are labeled with ~sign. The signs $ and ˅ designate terplexes with linear and cyclic PEI, respectively, having PNE stoichiometry, while the signs ≠ and < designate terplexes with linear and cyclic PEI with the stoichiometry PNE2. The signs ∞, ± and × designate terplexes with stoichiometries PNE3 (∞), PNE4 (±) and PNE5(×). The signs ≡ and " designate binary PE and NE complexes. (PEI complexes mainly appear in the low mass region of the spectrum, bellow m/z 600, where PEI oligomers dominate; the section is not shown in Figure 5.2). The superscripted charges indicate the degree of protonation. 76 A greater concentration of glutamic acid additionally leads to the formation of increasingly intense ODN/Glu and PEI/Glu("and ± ) binary complexes.
+ 2+ 2+ 2+ ODN/Glupolyplexes are observed as NE , NE2 , NE3 , and NE4 , while PEI/Glu complexes are seen as singly charged PE and PE2 ions.
Note that the increase in glutamic acid-based complexes and terplexesis accompanied by a decrease of PN polyplexes. At higher concentrations of PEI/ODN, doubly and triply charged PN2 complexes are detected in trace amounts. Contrariwise, as the amount of glutamic acid is increased higher order terplexes are coproduced. Remark
144 that P2N species, which were observed in binary PEI/ODN mixtures, are not detected in the ternary mixtures, suggesting that competitive binding is occurring. Introduction of
Glu into the complexes reduces the overall positive charge present, which is believed to help in balancing the cellular uptake and cytotoxic effect.141 Negatively charged carboxyl groups from glutamic acid can interact with positively charged amine groups of
PEI in the same way as negatively charged phosphate moieties of oligonucleotides interact with PEI. Expectedly, a large number of glutamic acid molecules in the environment lowers the stability and causes the disruption of PEI/ODN interactions, as also found by agarose gel electrophoresis and zeta potential measurements.141,143 This competition of N and E for P units reconciles the absence of P2N species in the ternary mixtures. Interestingly, the intensity of P/N/E terplexes is always higher than the intensity of P/E complexes even when the amount of glutamic acid added is high. This observation suggests that Glu prefers binding to PEI chains that are bound to ODN rather than free PEI chains, and that the PEI-Glu interaction is weak.
77 Since the PNE terplexes were produced by adding E to premixed P+N, it is reasonable to assume that their formation proceeds via the sequential equilibria
P+NPN and PN+EPNE, associated with the binding constant K1 and K2, respectively. Combining these reactions renders the equation [PNE]= K1K2[P][N][E] which relates the molar solution concentration of the terplex to the molar concentration of free PEI, ODN and Glu. If only the Glu concentration is varied, this relationship is simplified to [PNE]=K[E]. Hence, the terplex concentration is proportional to the terplex solution stability (K)110,111,116 as well as the concentration of free Glu, [E], which increases with the total solution concentration of Glu.144 The relative ESI intensities of the terplexesare found to correlate linearly with the total solution concentration of glutamic acid, as shown in Figure 5.3 for the P/d-(TTTTT)/E terplex. This finding confirms that the ESI spectra provide snapshots of the equilibrium concentration in solution and that the relative terplex intensities in these spectra can be used as approximate measures of the corresponding solution-phase stabilities. The relative intensities of the terplexes with P-to-N-to-E stoichiometries of 1:1:1(major products) and
1:1:2, obtained from mixtures with 5mM Glu, are reported in Figure 5.4.
78
0.15
)
rel y = 0.021x - 0.002 R² = 0.980 0.1
0.05
0
0 1 2 3 4 5 Relative Relative PNE Intensity ([PNE]
Glutamic Acid (E) concentration, mM -0.05
Figure 5.3. Relative intensity of PEI-d(TTTTT)-Glu terplexes with PEI/d(TTTTT)/Glu stoichiometry of 1:1:1 vs. the total Glu concetration in the solution. The glu concentration was varied from 1 mM to 5 mM. [PNE]rel was calculated by summing the intensities of PNE oligomers detected (see text) and dividing this sum by the sum of the intensities of P, N, E, PN, NE, and PE species.
The relative terplex intensities were derived by summing the intensities of the four most abundant terplex n-mers and dividing this total terplex intensity by the sum of the intensities of all unbounded species (free PEI and ODN) and binary complexes
(PEI/ODN, PEI/Glu and ODN/Glu). All charge states were considered as well as all discernable n-mers of the binary complexes.
79 0.1 0.092 a
PNE
0.08 b PNE2
0.06
0.04
PNE Abundance PNE 0.024 0.025 0.021 0.02 0.014 0.009 0.005 0.007 0 AAAAA CCCCC TTTTT GCGAT ODN Terplexes Figure 5.4. PEI-ODN-Glu ternary complexes formed from PEI 400, single-stranded pentadeoxynucleotides and glutamic acid solutions mixed in the ratio of 10:1:1.
a 2+ + 2+ + + + ([PNE ] + [PNE ]) / ([N]free + [P]free + [E]free + [PN ] + [PN ] + [PE ] + [NE ] + [NE2+]); ± 0.0014
b 2+ 2+ + + + ([PNE2 ]) / ([N]free + [P]free + [E]free + [PN ] + [PN ] + [PE ] + [NE ] + [NE2+]); ± 0.0003
As indicated in Figure 5.4, the most stable PNE 1:1:1 complex is produced with d(TTTTT) and the least stable with d(CCCCC); terplexes with d(AAAAA) and d(GCGAT) have intermediate and comparable stabilities while cationic terplexes withd(GGGGG) could not be detected at any experimental conditions (Figure 5.5). The solution stabilities of the observed terplexes follow the same order as the solution basicities of the phosphate groups connecting the nucleotide units, which decrease as the nucleobase is changed from T to A to C to G.151 The most stable ternary complex is formed with d(TTTTT) whose phosphate groups can most strongly interact with the protons of PEI. Furthermore, the difference in the stabilites of the terplexes may be due
80 to distinct secondary structures for the oligonucleotides studied or to a variance in the interactions among the constituents in the terplexes. The low stability of terplexes with d(GGGGG) which were not observed, cf. Figure 5.5, could result from the repulsion between the carboxylate groups of Glu and the phosphate groups of d(GGGGG) which, as mentioned above, have low basicity and tend to remain unprotonated.151 In addition, d(GGGGG) contains the most basic nucleobase, guanine,151,152 which after protonation could interact with the phosphate groups to yield a specific secondary structure152,153,154 that can not form stable ternary complexes. In support of this supposition, guanosine mononucleotides have been shown to favor a syn conformation which places the nucleobase above the ribose ring facilitating its interaction with the phosphate group.154,155 All these reasons made binding of PEI to Glu more competitive than binding of PEI to d(GGGGG), and the major product in ternary PEI/d(GGGGG)/Glu mixtures is a distinct distribution of binary PEn (n=1-5) complexes, cf. Figure 5.5. Finally, it is worth noting that the minor PNE2 terplexes with 1:1:2 stoichiometry of polymer to ODN to Glu show a very similar order of solution stabilities as the major PNE terplexes (cf. figure
5.4); again, the most stable and least stable complexes are formed with d(TTTTT) and d(CCCCC), respectively. Those with d(AAAAA) and d(GCGAT) have intermediate and comparable stabilities, and the PNE complex with d(GGGGG) is not observed.
81 +
z2
535.2 564.2 (a) b +
2 606.4
2+ + d(GGGGG)
w2 597.0
2+ PN
2+
649.4 P2N
792.5
632.4
660.9
908.6
#
930.1
758.1 952.1 1046.4
# 1024.9 # 893.2 # ! !
≡
+ PE2 ≡ +
527.3 (b) PE3
+
PE4
564.3
+
570.4 PE
674.4 5
ǂ
717.4
611.2
821.4
ǂ 760.5
≡
846.3 731.4
ǂ 864.3
696.4 968.5
907.5 1011.5 ≡ ǂ ≈ 1053.6 ≈ ≈ ≈
550 700 850 1000 m/z
Figure 5.5. ESI mass spectra of a mixture of PEI400 and d(GGGGG) oligonucleotide a) before and b) after the addition of glutamic acid (E). The labels on top of the peaks are explained in Figure 5.4 (PEI ions and PE+ complexes are observed below m/z 500.)
A tandem MS experiment on the PE+ complex containing the PEI 5-mer (Figure
5.6a) indicates that the major fragmentation involves loss of glutamic acid to yield the protonated polymer (P+), consistent with noncovalent binding of E to P+. The minor fragmentation channels observed mainly proceed by losses of small molecules (e.g., NH3,
H2O, CO2), pointing out that condensation reactions become possible in an energetically excited PE+ complex.
82
P+
233.3
(a)
PE+
380.3 -H O
2 +
E
296.9 319.4
363.3
202.1 278.9
130.1
173.2
148.1 345.4
233.3
(b)
P+ P(EE)+
190.2 509.4
100 200 300 400 500 m/z
Figure 5.6. MS2 (CAD) spectra of a) PE+ and b) P(EE)+ complexes. The ions selected m/z 380.3 and 509.4 respectively, contain 5 PEI oligomers. The top spectrum shows additional ions present, indicating admixtures in the selected precursor ion.
83 5.4 C/N Ratio Effect
The effect of increasing the number of carboxylic acid moieties in the terplex was
tested with d(TTTTT), d(AAAAA), d(CCCCC) and d(GCGAT) oligonucleotides.
846.3
#
730.3
(EE) +
2
# 553.2
P(EE)+ +
P(EE)2
859.3
# 2+
509.4 ^ PN(EE)
≡ #
785.5 2+
PN(EE) 552.4 984.4 4 2+
ǂ ^ 2+ PN2 +
≡ 575.2 $ PN(EE)2 PN
2+ +
^ PN(EE)3 PN(EE)
1459.3
694.3 ^ $ 618.2 ≡ 1122.5 # $ 1691.5 ˅ 1481.3
^ $ ˅ ≠ #
1967.8 1260.5
≠ 1398.7
≠ 1575.6 ˅ $ ≠ ∞∞ # # ∞ ∞ ~ ~ ~ # $ $ $ 500 800 1100 1400 1700 m/z Figure 5.7. ESI mass spectrum of PEI 400-d(TTTTT)-GluGlu terplexes formed by mixing the corresponding constituent solutions in the ratio 10:1:1,using a 5-mM GluGlu solution. The signs # and ^ represent PN binary complexes with linear and cyclic PEI, 2+ respectively; binary PN2 complexes are labeled with ~ sign. The signs $ and ˅ designate terplexes with linear and cyclic PEI, respectively, and PN(EE) stoichiometry, while≠ designates terplexes with linear PEI chains and stoichiometry PN(EE)2. The signs ∞ and ±designate higher-order terplexes with the stoichiometriesPN(EE)3 (∞) and PN(EE)4 (±). The signs ≡ and ǂ designate binary P(EE) and P(EE)2 complexes, respectively. The superscripted charges indicate the degree of protonation.
For this, the amino acid Glu (E) was replaced by the dipeptide GluGlu (EE) which has a molecular weight of 276.1 Da. Figure 5.7 displays the mass spectrum of the PEI400-
84 d(TTTTT)-GluGlu terplex, which is found to be slightly more stable than the terplexes with the other ODNs (cf. Figure 5.8). As with glutamic acid, the 1:1:1 stoichiometry,
PN(EE), predominates and is mainly observed as doubly protonated PN(EE)2+ along with
+ a trace amount of singly charged PN(EE) . Higher order PN(EE)n terplexes with up to 4
GluGlu molecules per terplex (Figure 5.7) are also observed, which decrease in intensity/stability as the number of EE moieties increases. The relative intensities of
PN(EE) and PN(EE)2 terplexes are summerized in Figure 5.8. Replacing E by EE reduces the relative intensities of the 1:1:1 terplexes (cf. Figures 5.4 and 5.8), except for the most weakly bound PEI/d(CCCCC)/GluGlu system which is not affected measurably by this change. The decreased yield of PN(EE) terplex could be caused by the higher tendency of GluGlu (vs. Glu) to form binary P(EE)n (n = 1-3) complexes. The MS/MS spectrum of P(EE)+, Figure 5.6b, shows exclusive dissociation to P+ and no condensation products (which were observed for PE+), confirming weak noncovalent bondig between the P and EE constituents. On the other hand, the higher number of acidic groups in
GluGlu vs. Glu lowers significantly the yieldof N(EE)n, as compared to that observed for
NEn complexes.
85 0.03 a PN(EE) b
PN(EE)2
0.019 0.02 0.018 0.018
0.014
0.01 0.008 0.008 PN(EE) Abundance PN(EE) 0.007 0.007
0 AAAAA CCCCC TTTTT GCGAT ODN Terplexes Figure 5.8. PEI-ODN-(Glu-Glu) ternary complexes formed from PEI 400, single- stranded pentadeoxynucleotides and glutamylglutamic acid solutions mixed in the ratio of 10:1:1.
a 2+ + 2+ + ([PN(EE) ] + [PN(EE) ]) / ([N]free + [P]free + [EE]free + [PN ] + [PN ] + [P(EE)+] + [N(EE)+] + [N(EE)2+]); ± 0.0014
b 2+ 2+ + + ([PN(EE)2 ]) / ([N]free + [P]free + [EE]free + [PN ] + [PN ] + [P(EE) ] + [N(EE)+] + [N(EE)2+]); ± 0.0007
The abundances of the ions at m/z 919.7 (PNE terplex with PEI 5-mer and Glu,
Figure 5.2b) and m/z 984.4 (PN(EE) terplex with Glu replaced by GluGlu, Figure 5.7)) relative to that of the PN binary polyplex at m/z 846.3 are 48 % and 22 %, respectively.
Furthermore, the relative intensity of the PN polyplexes themselves, calculated by adding the intensities of the PN ionsin all charge states and dividing the resulting sum by the intensities of free PEI and free ODN and its fragments in all charge states, decreases from
0.26 for PNE to 0.097 for PN(EE). The lower yield of PN polyplexes after the addition of EE indicates that the stabilities of the polyplexes decrease in the presence of a larger
86 number of carboxylic acid groups in the ternary system. The ESI mass spectrum of PEI
400 (Figure 4.1a) indicated an average molecular weight (Mn) of 295 Da;144 hence, the average PEI400, oligomer contains 6-7 repeat units and 7-8 potential N (amine protonation sites) atoms. The number of ethylene imine repeat units in the four most abundant PNE and PN(EE) terplexes is 4-7. Considering that E and EE contain 2 and 3
COOH groups, respectively (see introduction of chapter V), the carboxylic acid to amine
(C/N) ratio is 0.5-0.3 for PNE and 0.7-0.4 for PN(EE). For in vivo gene delivery, it was found that addition of poly(glutamic acid) (PGA) to PEI/DNA complexes at a C/N ratio of more than 0.8 caused terplex decomposition because PGA competed with DNA for
PEI; on the other hand DNA/PEI/PGA complexes could be formed at C/N ratios of less than 0.8.141 Our results are consistent with these findings, indicating a lower stability for terplexes with higher C/N ratio. The increased competition for PEI is further affirmed by the higher intensity of P(EE) complexes as compared to PE complexes.
Interestingly, the PN(EE) terplex with d(TTTTT) is not only less intense than the corresponding PNE terplex, but also of comparable intensityto all other ODN-terplexes
(Figure 5.8). This could result from a change in the secondary structure of d(TTTTT) when EE is added and form the competition between PEI-EE and PEI-d(TTTTT) interactions. In addition, PN(EE)2 species are higher in intensity than PNE2 species, in comparison to the stoichiometries PN(EE) and PNE respectively, which suggests an increased tendency to form higher order complexes as the binding interactions within
1:1:1 terplexes are weakened. It is further worth noting that the terplex (solution-phase) stabilities are more sensitive to changes in the ODN sequences for PNE than PN(EE) complexes, cf. Figures 5.4 and 5.8.141 87 5.5 Tandem Mass Spectrometry Studies on the Terplexes
Gas-phase stabilities were separately evaluated by fragmentation studies on the
P/N/E terplexes. Precisely, PNE2+ ions containing the 5-mer of PEI were isolated and subjected to collisionally activated dissociation (CAD). Relatively low collision energies were applied to disrupt the terplex ions. The major dissociation results in doubly charged binary PN complex and singly charged N, P and E constituents. Figure 5.9 demonstrates this by the MS2 spectra of d(TTTTT), d(CCCCC) or d(AAAAA) containing terplex ions.
PN2+
846.2
PNE2+
P+ N+
920.3
+ 233.2 190.2
E 825.7
276.3 1459.3
PN2+
808.7
P+
2+ +
E+ PNE N
233.2
882.7
787.7
190.2
148.2 1384.3
P+ PN2+
233.2 869.3 PNE2+ +
N
942.8
1504.3 190.2
100 600 1100 1600 m/z
Figure 5.9. MS2 spectra of PNE2+terplexes containing a) d(TTTTT), b) d(CCCCC) and c) d(AAAAA). The ions selected for collisionally activation dissociation (CAD) (m/z 920.3, 882.7 and 942.8, respectively) were dissociated at laboratory-frame collision energies of a) 10 eV, b) 6 eV and c) 5 eV. 88 1.0 d(CCCCC)
d(AAAAA) +
2 d(TTTTT)
PNE d(GCGAT)
0.5 E50 Relativeabundanceof
0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6
Ecm, eV
Figure 5.10. Fragmentation efficiency curves (relative precursor ion abundance versus center-of-mass collision energy) for PNE2+ ion from PEI-ODN-Gluterplexes containing d(TTTTT), d(AAAAA), d(CCCCC) and d(GCGAT). The Ecm values corresponding to the inflection point are 0.37, 0.27, 0.22 and 0.21 for ODN = d(TTTTT), d(GCGAT), d(CCCCC) and d(AAAAA), respectively.
In order to compare the gas-phase (i.e. intrinsic) stabilities of the different ternary complexes, fragmentation efficiency curves were constructed for each PNE2+ terplex system. MS2 (CAD) spectra were recorded as a funtion of collision energy, and the realtive abundance of the PNE2+ precursor ion in these spectra was plotted against the corresponding center-of-mass collsion energy (Ecm), as shown in Figure 5.10. The collision energy causing 50% of the precursor ions to decompose, called E50 energy, is customarily used as a measure of the intrinsic precursor ion stability.144 Because of the weak binding interactions between the terplex constituents, however, the ternary 89 complexes of d(AAAAA), d(GCGAT) and d(CCCCC) dissociate spontaneously (>50%) after isolation by the mass-selecting quadrupole even at the collision cell bias usually adjusted in regular MS mode (≤ 6V). For this reason, the collision energies corresponding to the inflection point of the fragmentation efficiency curves were used as a gauge for the gas-phase stability of the terplexes.151 The latter energies indicate that the intrinsic PNE terplex stabilities decrease as the ODN is changed in the following order: d(TTTTT) > d(GCGAT) > d(CCCCC) ≈ d(AAAAA). In solution, the PNE terplex stabilities have a slightly different dependence on ODN, viz.: d(TTTTT) > d(GCGAT) ≈ d(AAAAA) > d(CCCCC). The more strongly bound terplexes containing d(TTTTT) or d(GCGAT), have identical solution and gas-phase stability orders, consistent with very similar structures in both media. For the more weakly bound terplex, containing d(CCCCC) or d(AAAAA), solvent effects appear to play a significant role, reversing the intrinsic stability order.
5.6 Ion Mobility Separation of the Terplexes
Investigation of the ternary complexes by ion mobility mass spectrometry (IM-
MS) revealed the existence of multiply charged higher order complexes, which could not be detectedby classical ESI-MS analysis. IM-MS provides charge and shape-sensitive separations, allowing one to distinguish isobaric and isomeric ion structures, as well as to deconvolute ions in different charge states overlapping at the same m/z ratio. Ion mobility was applied to identify additional terplex stoichiometries that were present in trace amounts and thus hidden underneath the more abundant main products in regular
90 ESI mass spectra. These experiments were perfomed with the PEI-d(TTTTT)-GluGlu system, which provides the most abundant PN(EE) terplexes and a considerably more complex ESI mass spectrum than the corresponding PEI-d(TTTTT)-Glu system (cf.
Figures 5.2b
m/z
3000
2000 15 7 21 16
19 14 6 20 13 18 12 17 11 5 1000 10 8 4
3
2 9 1 2.0 4.0 6.0 8.0 10.0 12.0 14.0 Drift Time (ms)
Figure 5.11.2-D ESI-IM-MS plot (m/z vs. drift time) of the species in a 10:1:1 (molar) solution of PEI 400, d(TTTTT) and GluGlu (EE, 5 mM). Marked regions correspond to: + + + + 1, protonated PEI (P ); 2, P(EE) complexes; 3, P(EE)2 complexes; 4, P(EE)3 + + + complexes; 5, singly protonated d(TTTTT), N , and w4 fragment; 6, PN polyplexes; 7, + 2+ 2+ 2+ PN(EE) terplexes; 8, doubly protonated N and w4 ; 9, PN polyplexes; 10, 2+ 2+ 2+ PN(EE) terplexes; 11, PN(EE)2 terplexes; 12, PN(EE)3 terplexes; 13, 2+ 2+ 2+ PN(EE)4 terplexes; 14, PN2 polyplexes; 15, PN2(EE) terplexes; 16, 2+ 3+ 3+ P2N2(EE) terplexes; 17, P2N(EE)2 terplexes; 18, PN2 polyplexes; 19, 3+ 3+ 3+ PN2(EE) terplexes; 20, PN2(EE)2 terplexes; 21, PN2(EE)3 terplexes.
91 Figure 5.11 illustrates 2-D IM-MS plot with three distinctly separated bands, arising from species in three different charge states (+1, +2 and +3 from right to left). After IM separation additional non-covalent assemblies could be identified; a total of twenty-one species have been marked in Figure 5.11.
Region 1 contains the PEI distribution, while regions 5 and 8 correspond to singly and doubly charged d(TTTTT) respectively, with its singly and doubly charged w4 fragments. Regions 2, 3 and 4 designate singly charged P(EE), P(EE)2 and P(EE)3 complexes, respectively. Singly charged PN polyplexes are contained in region 6, while the doubly charged PN are found in region 9. Region 7 includes singly charged PN(EE) terplex ions and regions 10-13 include doubly charged PN(EE), PN(EE)2, PN(EE)3 and
PN(EE)4 ternary complex ions, respectively. Two additional minor products in the IM band for components in charge state +2, which could not be discerned by simple ESI-MS, appear in regions 15 and 16 and correspond to the ternary structuresPN2(EE) and
P2N2(EE), respectively. Four further minor PN(EE) terplexes that are detected in the +3
3+ 3+ 3+ charge IM band have the stoichiometries P2N(EE)2 , PN2(EE) , PN2(EE)2 and
3+ PN2(EE)3 (regions 17, 19, 20 and 21, respectively).
Compositions were assigned from the mass spectra obtained from the IM- separated regions. For example, integration of regions 16 and 17 gives rise to the mass
2+ spectra in Figure 5.12, which confirm the presence of composition of P2N2(EE) in
3+ region 16 and P2N(EE)2 in region 17. These multiply charged higher order terplexes mainly contain longer PEI oligomers which are present in very low concentration in PEI
400, reconciling their low intensities.
92
%
(b) 949.3
3+
P2N(EE)2
@ %
957.9
963.6
% %
@ 978.0
934.9
@ 972.3 943.6
2+ %
P2N2(EE) (a)
%
% 1864.3
1886.3 1872.7
% 1829.7
1851.3
1842.7
1808.2 1821.2
Figure 5.12. Mass spectra of regions a) 16 and b) 17in the ESI-IM-MS diagram of Figure 2+ 3+ 5.11, showing terplex compositions of P2N2(EE) and P2N(EE)2 , respectively. The distributions labeled by % contain one cyclic and one non-cyclic PEI oligomer, whereas the distributions labeled by @ contain two cyclic PEI oligomers. All other distributions contain only non-cyclic PEI oligomers.
The ability to separate by IM-MS superimposed isobars is illustrated in Figure 5.13.
There, region 6 of IM-MS plot is shown. The ions of m/z 1735 in this region can be extracted to obtain an IM chromatogram, which confirms the presence of two different species at this m/z ratio. Integration of the bands reveals that they arise from the isobaric complexes N(EE)+ and PN+ (Figure 5.13). The existence of N(EE)+ binary complexes
93 cannot be ascertained without IM-MS due to overlap of their isotope pattern with that of an isobaric PN+ oligomer (the one composed of the PEI 6-mer).
1735.5 1734.7
+ N(EE) PN+
10.01 10.74
6
Figure 5.13.Expanded plot of region 6 from the IM-MS diagram of Figure 5.11. The IM chromatogram extracted for singly charged m/z 1735 (top, center) contains two overlapping bands (centering at drift times 10.01 and 10.74 ms); integration of the bands (top, left and right) reveals the compositions N(EE)+ and PN+ for the faster (10.01 ms) and more slowly (10.74 ms) drifting species, respectively.
94 5.7 Conclusions
This study has reported the first characterization of ternary non-covalent complexes composed of poly(ethylene imine), single-stranded oligodeoxynucleotides and glutamic acid entities. Mass spectrometry confirmed the existence of self-assembled ternary associations, the preferred stoichiometry being 1:1:1 P-to-N-to-E in all cases studied. Additional compositions, involving the higher order terplexes with multiple glutamic acid moieties attached, were detected as minor products. Utilizing IM-MS other higher order ternary structures with multiple units of polymer and/or nucleotide present within the terplex could also be identified. Binary PEI-E or PEI-EE and ODN-E complexes were observed in higher intensities as the C/N ratio within the terplex was increased.
Solution and gas-phase stabilities, reflected by the relative ion abundances in the
ESI mass spectra, follow the same order for the more strongly bound PNE terplexes; however, solvent effects reverse the intrinsic stability order of the most weakly bound
PNE terplexes. The overall stability, in all cases studied, decreases with the rise of C/N ratio (i.e. by replacing E with EE). At the lower C/N ratio of the PNE stoichiometry, the terplex with d(TTTTT) is significantly more stable than the terplexes with the other
ODNs examined. This selectivity for binding T-rich sequences is lost, however, in the
PN(EE) terplexes, all of which show comparable stabilities. Hence, by proper adjustment of the C/N ratio, it is possible to promote terplex formation with a specific ODN sequence, or many different sequences.
95
CHAPTER VI
CHARACTERIZATION OF PHENOLIC COMPOUNDS IN BLACK RASPBERRIES
BY LIQUID CHROMATOGRAPHY MASS SPECTROMETRY (LC-MS)
6.1 Introduction
Black raspberry (Rubusoccidentalis L.) is a fruit rich in phenolic compounds which have shown to be chemoprotective against various types of cancers, such as esophageal, oral and colon.78,79,80 Anthocyanins establish a main group of phenolic components in berries and are important because they exhibit a wide range of antioxidant, anti-inflamatory, and anticarcinogenic properties.156157 Black raspberries are an especially rich source of anthocyanins and therefore of significant interest in chemoprevention research.81,76 Several studies examined the chemopreventative potential that includes freeze-dried (lyophilized) black raspberries.80,158 Harris et al. investigated the effect of a freeze-dried black raspberry diet on colon cancer in rats. They found that the growth of the tumor was inhibited by 71%.80 In another study, Krestyet al. reported that a freeze-dried black raspberry diet significantly reduced esophageal tumorgenesis in rats.78 In clinical chemopreventative studies, Kresty et al. examined how the addition of freeze-dried berries to a diet would affect patients with esophageal cancers. They observed reduction of two markers of oxidative stress in the urine of these patients, whose high levels are correlated with significant risk of cancer.159 In addition,
Stoner and colleagues conducted a clinical trial to determine the safety/tolerability of
96 freeze-dried black raspberries and found that their tolerability in healthy volunteers is as high as 45 g per day.160 These studies not only suggest that fruits in freeze-dried form may serve as an alternative natural chemopreventative option, but also point out the great benefits gained from components present in black raspberries.
Black raspberries contain numerous components that might be responsible for their antioxidant effects. To identify the contribution of individual compounds in black raspberries to their antioxidant properties, characterization of the individual components is essential. Although black raspberries are exceptionally rich in anthocyanins, other phenolic compounds present are believed to be significant. This study focuses on the identification of non-anthocyanin components in black raspberries. HPLC fractions were collected (based on their UV-Vis response) and then subjected to HPLC-ESI-MS and
MS2 analysis. The non-anthocyanin phenolics were identified by comparing their UV–
Vis profiles, ESI-MS and MS/MS spectra and, when available, the spectra and the retention times from standards. The HPLC-MS2 analysis produced poor quality spectra with little information for structural characterization. For better quality MS2 spectra, each sample was reanalyzed by direct injection ESI-MS and MS2.
This chapter provides a tentative characterization of the components identified in three representative fractions from black raspberry extracts; BRLC 1-H, BRLC-K, and
BRLC 1-E.
97 6.2 Sample Preparation
Dried HPLC fractions were kept at -80 °C until they were analyzed. Prior to analysis, all samples were dissolved in 30% acidic acetonitrile (acidified HPLC grade water (0.2% acetic acid) and acetornitrile in the ratio of 70:30, v/v) solutions at a concentration of ~0.5-1.0 mg/mL. After the solvent addition, the mixtures were vortexed for a few minutes and left sitting at room temperature for 30 minutes. Prior to HPLC-MS analysis, the dissolved samples were filtered through a 0.2 μM Whatman syringe filter.
6.3 Instrumental Conditions
HPLC-MS analyses were performed on a BrukerEquire –LC (Bruker, Daltonics,
Billerica, MA) Quadrupole Ion Trap (QIT) mass spectrometer, interfaced with Agilent
HP 1100 HPLC system. Chromatographic separation was effected with a Phenomenex
Gemini (C6-phenyl) column (250 x 4.6 mm, 5μm)held at 30 °C, usinga flow rate of 0.7 mL/min. The set up included a guard column(Phenomenex Security Guard), so that any additional impurities present within the solution analyzed were captured prior to entering the separation column. The injection volume was 20 μL. Chromatographic analysis was performed using a mobile phase consisting of acidified water (0.2 % of acetic acid was added to HPLC-grade water) as solvent A and acetonitrile, ACN, (100% CH3CN) as solvent B. Gradient elution was performed for 50 min, starting with water: ACN 91:9
(v/v) and ending with water: ACN 9:91 (v/v), followed by isocratic elution for additional
5 minutes. The gradient elution conditions are detailed in Table1. The LC-MS spectra
98 were acquired in negative mode, and extracted ion chromatogram plots were used for data analysis.
The mass spectrometry experiments were performed either on the LC system (as discussed above) or by direct infusion via a syringe pump. Nitrogen was used as nebulizing and drying gas. The drying gas temperature was set at 250 °C. The spray voltage was set at 3.5 or 4.0 kV in the negative mode. The flow rates of the nebulizing and drying gases were 10 L/min and 8L/min, respectively.
Table 6.1. Gradient Elution Program for the Samples Obtained by HPLC Fraction Using UV-Vis Detection
Time (min) Solvent A % Solvent B % Flow (mL/min) 0 91 9 0.7 10 91 9 0.7 20 78 22 0.7 35 70 30 0.7 40 40 60 0.7 45 40 60 0.7 50 91 9 0.7 55 91 9 0.7
6.4 Characterization of Phenolic Compounds from Black Raspberries
Using reversed-phase HPLC-UV, fractions of black raspberry samples were collected and each fraction was then separately analyzed by HPLC-ESI-MS and MS2.
Negative mode HPLC-ESI-MS spectra show deprotonated species, [M-H]¯.
99 6.4.1 BRLC 1-H
The total ion chromatogram of fraction BRLC 1-H is shown in Figure 6.1. Two major LC peaks are observed between 28.8 and 30.4, consistent with the HPLC-UV profile for this fraction (shown in Appendix, Figure A1). Additional poorly separated components are present within the two peaks; narrow integration within these regions did not produce significantly different mass spectra, due to overlap of the retention volumes.
Thus, the entire area of each of the two peaks was extracted for mass spectrometric analysis.
Figure 6.1. LC-MS total ion chromatogram (TIC) of sample BRLC 1-H (bottom), and expanded trace of the species eluting between 26-33 minutes (top).
100 The mass spectrum of the first LC peak that eluted between 28.8 -29.5 min is shown in Figure 6.2. The main three product anions are observed at m/z 299.9, m/z
463.0 and m/z 475.1. The mass difference between m/z 299.9 and m/z 463.0 is 163 Da which corresponds to one glucoside unit. Ion m/z 463 assigned to quercetin-O-glucoside and m/z 299.9 to the radical ion of its aglycone, formed after the loss of the glucoside moiety. For further investigation, ion m/z 463 was subjected to MS2 analysis, shown in
Figure 6.3. The major fragmentation resulted in the neutral loss of 162 Da, corresponding to the elimination of hexose glucoside unit to form the fragment ion at m/z
300.9. A closer look at the expanded region of the LC-MS base peak, m/z 299.9 in
Figure 6.2, suggests the existence of two species, the anion [M-H-162]¯ and the radical anion [M-H-163]¯● (m/z 299.9). Note that the ion at m/z 300.9 in the isotopic cluster is not likely to be entirely the 13C satellite of m/z 299.9 due to its significant intensity, but rather also contains the isobaric aglycone radical anion ion [M-H-163]¯●.
101 1
2
3
Figure 6.2.LC-MS spectrum of the fraction eluting between 29.0-29.5 min in the TIC of BRLC 1-H.
The existence of these two species is very useful in differentiation of structural isomers of flavonol-O-glucosides. As the structural characterization of isomers is challenging using
ESI-MS and MS2 alone, Geng at al. reported on their differentiation utilizing different instrumentation when performing collisionally activated dissociation (CAD).161 They reported that positional isomers of O-glycosides can be differentiated by monitoring their
[M-H-162]¯ and [M-H-163]¯● ions. In addition, the choice of instrumentation used for the MS2 (such as QIT or triple quadrupole) will affect the results for the same compound with respect to the abundance ratio of deprotonated to radical ion species.
102 Different positional isomers of quercetin-O-glucosidesproducedmarkedly different relative abundances of quercetin radical ions and deprotonated species upon CAD on a
QIT mass spectrometer, according to the study of Geng et al.161 The CAD spectrum of quercetin-O-glucoside (m/z 463) shown in Figure 6.3 reveals that the relative abundance of the radical aglycone ion (m/z 299.9) is significant, albeit lower than the abundance of the deprotonated ion (m/z 300.9). Only quercetin-3-O-glucoside produced an appreciable aglycon radical ion in the study of Geng et al.161 Hence, the m/z 462.9 ion in BRLC 1-H
(Figure 6.3) was assigned to quercetin-3-O-glucoside. The additional ions present in the low molecular weight region of the CAD spectrum of m/z 462.9 (m/z 106.9, 150.8, and
178.8 in Figure 6.3) most likely arise from fragmentations within the quercetin unit
(Scheme 6.1). The same ions are also seen in the MS2 spectrum of quercetin itself, presented in Figure 6.4(top).
Figure 6.3.MS2 spectrum of the ion at m/z 462.9 from fraction BRLC 1-H. 103 The major MS2 fragments of deprotonated quercetin-3-O-glucoside (i.e. m/z 299.9 and
300.9) are also observed in the LC-MS spectrum (Figure 6.2), where no additional energy is deposited (no CAD). The presence of aglycone fragments in the TIC of the fraction containing quercetin-3-O-glucoside (eluting at 29.0-29.5 min) is consistent with the facile fragmentation observed for this compound here (Figure 6.3) and by Geng et al.161
Quercetin cannot be present in this fraction; a quercetin standard analyzed by HPLC-MS under the same conditions, eluted at 42.9 min, as expected for the less polar aglycon (see
Figure 4A in the Appendix). This retention time difference confirms the absence of quercetin as single species in the fraction containing the corresponding 3-O-glucoside.
O H O H
H O O
O H O H O
Scheme 6.1 Plausible cleavages in quercetin
The ion at m/z 475.1(Figure 6.2)was also subjected to MS2analysis(Figure 6.4, bottom).
The resulting spectrum showed a main fragment ion at m/z 329 corresponding to a loss of
146 Da. Such loss is representative for the coumaroyl group and indicates that this phenolic compound contains a coumaroyl unit.162 No loss characteristic of a glucoside(162 Da) was observed.
104 -H OH OH HO O
OH m/z 301 OH O
-H -H HO -H HO O HO -CO2 O O HO OH OH O O m/z 107 m/z 151 m/z 179
Figure 6.4.MS2 spectra of theions at m/z 300.9 (top) and m/z 475.0 (bottom) from fraction BRLC 1-H (cf. Figure 6.2).
The LC-MS spectrum of the second chromatographic peak, eluting between 29.5-30.0 minutes, is shown in Figure 6.5. Although, several of the ions in this spectrum were also observed from the first chromatographic peak, due to partial LC overlap(Figure 6.1), an intense ion at m/z 226.9 together with m/z 389.0 were singled out as the new components. These were identified as resveratrol and resveratrol glucoside, respectively.
As with quercetin-glucoside and its aglycone fragment (Figure 6.2), in the same manner, ion m/z 226.9 was assigned to the aglycon fragment of glucosilated resveratrol. Indeed,
105 MS2 analysis of m/z 226.9 resulted in fragments diagnostic resveratrol (Figure 6.6).163
Resveratrol standard was not available for analysis by HPLC-MS and comparison. 4
5
Figure 6.5.LC-MS spectrum of the fraction eluting between 29.5-30.0 min in the TIC of BRLC 1-H. HO HO - CC O H- CC O HH HH HO HO
cleavage with H rearrangement -CHCOH
HO m/z 143
H2C. . C C O H H m/z 185
O - m/z 157
O
m/z 183
Figure 6.6.MS2spectrum of ion m/z 226.9 from fraction BRLC 1-H. The fragmentation pattern observed is identical to that reported for the [M-H]¯ ion of resveratrol.163
106 The structures of the phenolic non-anthocyanin components identified in BRLC 1-H by
LC-MS and MS2 are summarized in Figure 6.7. These compounds could be partially separated by LC due to subtle differences in their hydrophobicities and characterized by
MS and MS2 due to their unique masses and CAD fragmentation patterns.
Figure 6.7. Chemical structures of the phenolic compounds identified in fraction BRLC 1-H.
107 Expectedly, the more polar phenolic compounds containing glucosides (quercetin- glucoside and resveratrol-glucoside, 2 and 5, respectively) eluted at earlier retention times. Compound 3, which was shown to contain a coumaroyl group but no sugar part eluted at a later retention time. This order is consistent with the increasing hydrophobicity at the longer retention times in reversed-phase HPLC.
6.4.2 BRLC 1-E
HPLC-MS analysis of fraction BRLC 1-Eresulted in one major peak in the chromatogram (Figure 6.8, top), which is consistent with the previous analysis by HPLC-
UV (see Figure 2A in the Appendix). The observed peak eluted between 24.0-25.0 minutes and is indicative of a single component being present in this fraction. The mass spectrum extracted from the HPLC peak (Figure 6.8, bottom) contains a major ion at m/z
289.0, which was identified as deprotonated epicatechin. HPLC-MS also produced an ion at m/z 579.1 which was assigned to the deprotonated dimeric structure of epicatechin.
For further structure information, ion m/z 289 was subjected to MS2 via CAD (Figure
6.9).
108 Figure 6.8. LC-MS total ion chromatogram (TIC) of sample BRLC 1-E (top) and extracted mass spectrum for the peak eluting between 24.0-25.0 minutes (bottom).
Figure 6.9.MS2 spectrum of the ion at m/z 288.9 from fraction BRLC 1-E.
109 The fragment at m/z 270.9 corresponds to loss of water, indicative of an aliphatic alcohol.
The fragment at m/z 244.9 is the base peak in the CAD spectrum and is characteristic of
164 catechin; it is more likely produced by the loss of CH2=CH-OH (vinyl alcohol) from the benzopyran ring. The ion at m/z 179 is proposed to result from the loss of the entire aromatic ring of the benzopyran unit,164 while m/z 205 is possibly due to the loss of 84
164 2 Da from the same ring (C4H4O2). Overall, the HPLC-MS and MS analysis of the fraction BRLC 1-E confirmed the presence of epicatechin in black raspberries.
6.4.3 BRLC 1-K
Analysis of fraction BRLC 1-K by HPLC-MS resulted in two main peaks in the chromatogram (Figure 6.10). These peaks eluted between 34.4 and 36.4 minutes. Each peak appeared to contain two different components that were not well separated. The mass spectra extracted from the narrow fractions eluting between 34.4-34.5min, 34.7-
35.0 min and 35.1-35.2 min (all within the first main peak)are compared in Figure 6.11.
110
Figure 6.10. LC-MS total ion chromatogram (TIC) of sample BRLC 1-K, (bottom) and expanded trace of the component eluting between 34-38 minutes (top).
111
Figure 6.11.LC-MS spectra of the fractions eluting between 34.4-34.5min, 34.7-35.0 min and 35.1-35.2 min in the TIC of sample BRLC 1-K.
All three spectra include two major ions at m/z 503.3 and m/z 725.3 and a less abundant m/z 665.3. There are no significant differences in the contents of the three extracted regions. On the other hand, the characteristic mass difference of 162 Da is observed which corresponds to the glucoside sugar unit.
Additional signals are present in all three spectra; these have been assigned as ―spikes‖ since they do not contain 13C isotope peaks. Spikes are due to the noise.165 They usually
112 appear in single scans or in the spectra obtained by averaging a small number of noisy scans. The extraction of chromatograms from spikes resulted in sharp narrow peaks, such as those between 5 and 30 minutes in Figure 6.10, whose shape is not representative of real chromatographic peaks (vide infra). Because of these arguments, and the similarity of the three spectra in Figure 6.11, the splitting of the main TIC peak centered at 34.4 min is attributed to noise. The ions at m/z 503, 665 and 725 probably originate from only one component present in the entire first chromatography peak. For more information on this component, these ions were further analyzed via MS2.
During its MS2analysis, the ion at m/z 725.3produces fragments at m/z 665 and m/z 503.
The first neutral loss of 60 Da corresponds to CH3COOH and is characteristic of acetylated species. It is followed by a sequential loss of a sugar unit (162 Da). The resulting fragment of m/z 503 undergoes sequential losses of H2O, CH3OH, and CO2(for
CH2CHOH). Additional ions, differentiating by 12 Da, are observed in the low molecular weight region, suggesting the presence of aromatic entities. Thus, this component is tentatively assigned as acetylated structure containing one sugar unit and aromatic rings.
113
Figure 6.12.MS2spectra of the ions at m/z 725 and m/z 503 (top and bottom, respectively) from fraction BRLC 1-K.
From the second LC-MS peak of the BRLC 1-K sample, three mass spectra were extracted from narrow segments eluting between 35.5-35.6min, 35.6-35.8 min and 35.8-
36.0 min, cf. Figure 6.13. There are at least six common ions present in all three mass spectra, due to the overlap of the analyzed chromatographic peak segments. Unlike with the first chromatographic peak of BRLC 1-K, here the abundances of the common ions vary, as the elution time of the segment increases. The LC-MS section between 35.5-
35.6 min produces major ions at m/z 462.9 and m/z 300.9. They are assigned to quercetin and its glucosylated derivative, respectively, as they have already been characterized in sample BRLC 1-H.
114 1 2
3 4
Figure 6.13.LC-MS spectra of the fractions eluting between 35.5-35.6min, 35.6-35.8 min and 35.8-36.0 min in the TIC of sample BRLC 1-K (Figure 6.10).
In the following fraction, eluting at 35.6 -35.8min, the abundances of two other common ions increase, viz. of m/z 272.9 and m/z 435.0. These are identified as phloretin and phloretin-glucoside, respectively (vide infra). Lastly, the fraction eluting between 35.8 min and 36.0 min displayed increased abundances for m/z 316.9 and m/z 478.9, which are characteristic of myrcetin and myrcetin-O-glucoside, respectively (vide infra).
115
Figure 6.14.MS2spectra of the ions at m/z 435.0(top) and m/z 272.9(bottom) from sample BRLC 1-K.
The ion of m/z 435 fragments by loss of 162 Da, as expected for a glucoside, to produce deprotonated phloretin, m/z 272.9.166Ion m/z 272.9 dissociates consequently by loss of
106 Da,166 as confirmed by its own MS2 spectrum (Figure 6.14, bottom). Ion m/z 216.6 was absent in the reported MS2 spectrum of phloretin-glucoside.166 It is attributed to contamination of the glucoside (436 Da) by the dimeric glucoside+HCl which forms the
37 37 isobaric anion [(glucose)2+H Cl+ Cl]¯ (m/z 435). Loss of glucose+HCl from the latter ion leads to m/z 216.6.
116 Ion m/z 316.9 is a characteristic for myricetin for phenolics in the negative mode.167
Thus, it is reasonable to assume that m/z 316.9 is the aglycone fragment from myricetin- glucoside (m/z 479.0) as was the case with quercetin-glucoside and its aglycone fragment.
Figure 6.15. MS2 spectrum of m/z 479.0 form sample BRLC 1-K.
MS2examination of m/z 479 revealed a characteristic sugar unit loss (162 Da) resulting in the deprotonated myricetin component (Figure 6.15). When m/z 316.9 was subjected to the MS2 it did not produce detectable fragments. Myricetin standard was then analyzed via HPLC-MS (see Figure A5 in the Appendix) and resulted in a chromatographic peak at about the same retention time as that for sample BRLC 1-K, 36.5 minutes. Thus, glucosylated myricetin and myricetin as an individual species could be present together.
It is reminded that quercetin was found to have a significantly longer retention time than the glucosylated analog. The very similar retention times of myricetin and glucosylated myricetin are therefore puzzling and might be due to increased polarity of myricetin 117 versus quercetin (one more OH group). Additional experiments are needed to corroborate this supposition.
Figure 6.16.Chemical structures of the phenolic compounds tentatively identified in fraction BRLC 1-K.
Figure 6.16 summarizes the tentative structural assignments made for the components observed in fraction BRLC 1-K. Due to their less polar structures, phloretin and its glucosylated form eluted at a later time than myricetin and its glucoside.
118
Figure 6.17. Comparison of extracted chromatogram peaks for m/z 579.9 and m/z 462.9 from fraction BRLC 1-K and distinguishing between real and spike peaks.
Finally, Figure 6.17 exemplifies how noise spikes in LC-MS spectra can be distinguished from normal chromatographic peaks. For this, the putative ions at m/z
462.9 and m/z 579.9 in Figure 6.13 (top) were selected. Note that the latter shows no isotopes. Moreover, the single-ion chromatogram extracted from 579.9 is too narrow to represent a real signal from elution. In contrast, m/z 462.9 shows isotopes and leads to a chromatographic signal with a realistic width. Hence, m/z 462.9 arises from real ions, whereas the peak appearing at m/z 579.9 must be a noise spike. The peaks > m/z 800 in
Figure 6.11 are mainly noise spikes.
6.5 Conclusions
HPLC-MS was applied to the characterization of phenolic components in three representative fractions from black raspberries, BRLC 1-H, BRLC 1-K, and BRLC 1-E. 119 In combination with ESI, LC-MS and MS2 techniques were utilized to gain structural information about these compounds. In addition, standards of quercetin and myricetin were used to obtain additional information about LC retention times. Identified components in the given fractions include quercetin, myricetin, phloretin with their glucosides. In addition, compounds with coumaroyl entities as well as species containing not yet ascertained aromatic moieties were also observed.
This work demonstrated that MS interfaced with HPLC can be used to characterize various phenolics in black raspberry extracts. As a cautionary note, it must be stated that the combination of liquid chromatography with ESI-MS can produce chromatograms and spectra with high level of background, noise and spike peaks.
Chromatogram extraction for certain ―ions‖ in the mass spectra allowed for differentiation between real and spike peaks.
120
CHAPTER VII
SUMMARY
Mass spectrometry is utilized in a variety of areas due to its applicability to a wide range of samples. It allows for the analysis of labile compounds and, as a sensitive method, it provides information about species present in low concentrations. To enhance analysis and gain more structural information, mass spectrometry was interfaced with the separation methods of liquid chromatography (LC) and ion mobility (IM). Separation prior to mass measurement provides an additional dimension for investigation and makes mass spectrometry a powerful tool in elucidating structures.
Chapter 4 focused on the characterization of polyplexes formed between poly(ethylene imine) (PEI) and oligodeoxynucleotides (ODNs) by ESI mass spectrometry. PEI 400 and PEI 800 were selected as hosts for five different pentadeooxynucleotides, added in different molar ratios. The major product in all cases studied had 1:1 stoichiometry of polymer-to-ODN (PN). PN2 and P2N stoichiometries were also detected in trace amounts, while compositions of P2N2 and PN3could only be detected after separation by ion mobility (IM) prior to analysis by ESI-MS. Relative intensities of the polyplex ions in mass and tandem mass spectra were used to assess their stabilities in solution and the gas phase, respectively. Identical stability orders were found in both solution and the gas phase for the different ODN polyplexes examined, suggesting that similar structures are present in both phases. With the polymer carriers
121 studied (PEI 400 and PEI 800), the most stable polyplex was formed with thymine-rich nucleotides and the least stable was adenine-rich nucleotide.
In chapter 5 of this dissertation, ternary complexes among PEI 400, penta- dexoynucleotides and glutamic acid moieties (specifically glutamic acid monomer and dipeptide, E and EE respectively) were investigated. Various concentrations of E and
EE were added to constant molar ratio (10:1) mixtures of PEI and ODN in order to create self-assembled terplexes (PNE). Due to their labile nature, instrument parameters were optimized to preserve the terplexes, with special attention given to the pressure in the first pumping region (the extraction region) of the mass spectrometer. The preferred stoichiometry of the ternary complexes was 1:1:1 P-to-N-to-E in all cases studied.
Higher order complexes with multiple glutamic acid moieties were also observed as minor products. Separation via ion mobility (IM) prior to mass analysis enabled the detection of additional higher order complexes containing multiple polymer and/or nucleotide moieties within the terplex. Binary PEI-E and PEI-EE complexes were also detected, their yield increasing with the C/N (carboxyl to amine) ratio. As with the binary polyplexes, the solution and gas phase stabilities of the ternary were highest with thymine-rich nucleotides. The overall stability for all cases studied decreased as the C/N ratio was increased
In chapter 7, the HPLC-ESI-MS and ESI-MS2 methods were applied to the characterization of phenolic components in the extracts from black raspberries. The major components tentatively identified include quercetin, myricetin, phloretin with their glucosides. For additional information on the compositions present, standard samples of
122 a few available phenolics were studied by HPLC-MS for comparison. Furthermore, manipulation of the chromatogram and mass spectral features allowed for the gain of additional information. The extracted chromatograph for specific ions enabled the differentiation between real and spike peaks in the spectra. This reduced the complexity of composition assignments.
123
REFERENCES
1 Schalley, C.A.; Springer, A. Mass Spectrometry and Gas-Phase Chemistry of Non- covalent Complexes, John Wiley & Sons: 2009.
2 Veenestra, T. Electrosparay ionization mass spectrometry in the study of bimolecular non-covalent interactions, Biophysical Chemistry, 1999, 79, 63-79.
3 Uhlenheuer, D.A.; Petkau, K.; Brunsveld, L. Combining supramolecular chemistry with biology, Chemical Society Reviews, 2010, 39, 2817-2826.
4 Lehn, J. M. Toward complex matter: Supramolecular chemistry and self-organization, Proceedings of the National Academy of Sciences, 2002, 99, 4763-4768.
5 Schalley, C.A. Analytical Methods in Supramolecular Chemistry, WILEY-VCH Verlag GmbH & Co.: Weinheim, 2007.
6 Albrecht, M. Supramolecular chemistry-general principles and selected examples from anion recognition and metallosupramolecular chemistry, Naturwissenschaften, 2007, 94, 951-966.
7 Baytekin, B.; Baytekin, H. T.; Schalley, C. A. Mass spectrometric studies of non- covalent compounds: why supramolecular chemistry in the gas phase?, Organic and Biomolecular Chemistry, 2006, 4, 2825-2841.
8 Lehn, J. M. Cryptates: Inclusion Complexes of Macropolycyclic Receptor Molecules, Pure and Applied chemistry, 1978, 50, 871-892.
9 Voet, D.; Voet, J. D.; Pratt, C. W. Fundamentals of Biochemistry, Life at the Molecular Level, 2nd Ed., John Wiley & Sons: New Jersey, 2006.
10 Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Electrospray ionization-principles and practice, Mass Spectrometry Reviews, 1990, 9, 37-70
11 Karas, M.; Bahr, U.; Giebmann, U. Matrix assisted laser desorption ionization mass spectrometry, Mass Spectrometry Reviews, 1991, 10, 335-357
12 Dass, C. Fundamentals of Contemporary Mass Spectrometry, Wiley & Sons: New Jersey, 2007
124
13 Loo, J. A. Studying Noncovalent Protein Complexes by Electrospray Ionization Mass Spectrometry, Mass Spectrometry Reviews, 1997, 16, 1-23
14 McLafferty, F. W. Tandem Mass Spectrometry, Science, 1981, 214, 280-287
15 Loo, J.A.; Hu, P.; Mconnell, P.; Mueller, W. T.; Swayer, T. K.; Thanabal, V. A Study of Src SH2 domain protein-phosphopeptide binding interactions by electrospray ionization mass spectrometry, Journal of American Society for Mass Spectrometry, 1997, 8, 234-243
16 Cheng, X.; Chen, R.; Bruce, J. E.; Schwartz, B. L.; Anderson, G. A.; Hofstadler, S. A.; Gale, D. C.; Smith, R. D.; Gao, J.; Sigal, G. B.; Mammern, M.; Whitesides, G M. Using Electrospray ionization FTICR mass spectrometry to study competitive binding of inhibitors to carbonic anhydrase, Journal of American Chemical Society, 1995a, 117, 8859-8860
17 Fitzgerald, M. C.; Chernushevich, I.; Standing, K. G.; Whitman, C. P.; Kent, S. B. H. Probing the oligomeric structure of an enzyme by electrospray ionization time-of- flight mass spectrometry, Proceedings of the National Academy of Sciences, USA, 1996, 93, 6851-6856
18 Sanglier, S.; Leize, E.; Van Dorsselaer, A. Comparitive ESI-MS Study of ~ 2.2 MDa Native HemocyaninsfromDeep-Sea and Shore Crabs: From Protein Oligomeric State to Biotope, Journal of American Society for Mass Spectrometry, 2003, 14, 419-429.
19 Rogniaux, H.; Van Dorsselaer, A. Binding of Aldose Reductase Inhibitors: Correlation of Crystallographic and Mass Spectrometric Studies, Journal of American Society for Mass Spectrometry, 1998, 10, 635-647.
20 Kiselar J. G.; Downrad, K. M. Preservation and Detection of Specific Antibody- peptide Complexes by Matrix-assisted Laser Desorption Ionization Mass spectrometry, Journal of American Society for Mass Spectrometry, 2000, 11, 746- 750.
21 Song, G. A Study of Noncovalent Protein Complexes by Matrix-Assisted Laser Desorption/Ionization, Journal of American Society for Mass Spectrometry, 2007, 18, 1286-1290.
22 Woods, A. S.; Huestis, M. A. A Study of Peptide-Peptide Interaction by Matrix Assisted Laser Desorption/ionization, Journal of American Society for Mass Spectrometry, 2000, 12, 88-96.
125
23 Woods, A. S.; Buchsbaum, T. A.; Worrall, J. M.; Berg, J. M.; Cotter, R. J. Matrix- Assisted Laser Desorption/Ionization of Noncovalently Bound Compounds, Analytical Chemistry, 1995, 67, 4462-4465.
24 Goodlett, D. R.; Ogorzalek, Loo, R. R.; Loo, J.A, Wahl, J. H.; Udseth, H. R.; Smith, R. D. A study of the thermal denaturation of ribonuclease S by electrospray ionization mass spectrometry, Journal of American Society for Mass Spectrometry, 1994, 5, 614-622
25 Veenestra, T. D.; Johnson, K. T.; Tomlinson, A. J.; Naylor, S.; Kumar, R. Determination of calcium-binding sites in rat brain calbindin D28K by electrospray ionization mass spectrometry, Biochemistry, 1997, 36, 3535-3542.
26 Greig, M. J.; Gaus, H.; Cummins, L. L.; Sasmor, H.; Griffey, R. H. Measurement of macromolecular binding using electrospray mass spectrometry. Determination of dissociation constants for oligonucleotides-serum albumin complexes, Journal of American Chemical Society, 1995, 117, 10765-10766.
27 Luo, S.; Li, Y.; Qiang,W.; Zhao, Y.; Abe, H.; Nemoto, T.; Qin, X.; Nakanishi, N. Detection of specific noncovalent interaction of peptide with DNA by MALDI-TOF, Journal of American Society for Mass Spectrometry, 2004, 15, 28-31.
28 Park, H.; Park, K. Polymers in Pharmaceutical Products, Polymers of Biological and Biomedical Significance, 1994, vol. 540, Chapter 1, 2-15.
29 Sasmal, A.; Nayak, P. L.; Sasmal, S. Desing of some novel polymers for gene delivery systems, DNA Delivery Systems, Popular Plastics &Packagins, 2009, 54, 39-46.
30 Pillai, O.; Panchagnula, R. Polymers in Drug Delivery, Current Opinion in Chemical Biology, 2001, 5, 447-451.
31 Gunatillake, P. A.; Adhikari, R. Biodigradable Synthetic Polymers for Tissue Engineering, European Cells and Materials, 2003, 5, 1-16.
32 Itaka, K.; Ishii, T,; Hasegawa, Y.; Kataoka, K. Biodegradable polyamino acid-based polycations as safe and effective gene carrier minimizing cumulative toxicity, Biomaterials, 2010, 31, 3707-3714.
33 Pack, D. W.; Hoffman, A. S.; Pun, S.; Stayton, P. S. Design and Development of Polymers for Gene Delivery, Nature Reviews, 2005, 4, 581-593.
34 De Smedt, S. C.; Demeester, J.; Hennink, W. E. Cationic Polymer Based Gene Delivery Systems, Pharmaceutical Research, 2000, 17, 113-126.
126
35 Zhang, S.; Xu, Y.; Wang, B.; Qiao, W.; Liu, D.; Li, Z. Cationic compounds used in lipoplexes and polyplexes for gene delivery, Journal of Controlled Release, 2004, 100, 165-180.
36 Boussif, O.; Lezoualc‘h, F.; Zanta, M. A.; Mergny, M. D.; Scherman, D.; Demeneix, B.; Behr, J. P. A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: Polyethylenimine, Proceedings of the National Academy of Sciences, 1995, 92, 7297-7301.
37 Godbey, W. T.; Barry, M. A.; Sagaus, P.; Wu, K. K; Mikos, A. G. Poly(ethylenimine)- mediated transfection: A new paradigm for gene delivery, Journal of Biomedical Material Research, 2000, 51, 321-328.
38 Zhang, C.; Yadava, P.; Hughes, J. Polyethylenimine strategies for plasmid delivery to brain-derived cells, Methods, 2004, 33, 144-150.
39 Ko, Y. T.; Bhattacharya, R.; Bickel, U. Liposome encapsulated polyethylenimine/ODN polyplexes for brain targeting, Journal of Controlled Release, 2009, 133, 230-237.
40 Peng, M.; Chen, Q. Interaction and morphology of polyethylenimine/DNA complexes as studies by solid-state NMR spectroscopy, Chemical Physics Letters, 2008, 456, 189-192.
41 Pun, S. H.; Bellocq, N. C.; Liu, A.; Jensen, G.; Machemer, T.; Quijano, E.; Schluep, T.; Wen, S.; Engler, H.; Heidel, J.; Davis, M. E. Cyclodextrin-Modified Polyethylenimine Polymer for Gene Delivery, Bioconjugate Chemistry, 2004, 15, 831-840.
42 Luo, X.; Pan, S.; Feng, M.; Wen, Y.; Zhang, W. Stability of poly(ethylene glycol)- graft-polyethylenimine copolymer/DNA complexes: influences of PEG molecular weight and PEGylation degree, Journal of Material Science: Materials in Medicine, 2010, 21, 597-607.
43 Sun, J.; Paik, H. J.; Hwang, B. K. Ionization of poly(ethylenimine) and poly(allylamine) at various pH‘s, Bioorganic Chemistry, 1994, 22, 318-327.
44 Godbey, W. T.; Wu, K.; Mikos, A. G. Poly(ethylenimine) and its role in gene delivery, Journal of Controlled Release, 1999, 60, 149-160.
45 Ehrhardt, C.; Schmolke, M.; Matzke, A.; Knoblauch, A.; Will, C.; Wixler, V.; Ludwig, S. Polyethylenimine, a cost-effective transfection reagent, Signal Transduction, 2006, 6, 179-184.
127
46 Lungwitz, U.; Breunig, M.; Blunk, T, Gopferich, A. Polyethylenimine-based non-viral gene delivery systems, European Journal of Pharmaceutics and Biopharmaceutics, 2005, 60, 247-266.
47 Wightman, L.; Kircheis, R.; Rossler, V.; Carotta, S.; Ruzicka, R.; Kursa, M.; Wagner, E. Different behavior of branched and linear polyethylenimine for gene delivery in vitro and in vivo, Journal of Gene Medicine, 2001, 3, 362-372.
48 Kunath, K.; von Harpe, A.; Fischer, D.; Petersen, H.; Bickel, U.; Voigt, K.; Kissel, T. Low-molecular-weight polyethylenimine as a non-viral vector for DNA delivery: comparison of physicochemical properties, transfection efficiency and in vivo distribution with high-molecular-weight polyethylenimine, Journal of Controlled Release, 2003, 89, 113-125.
49 Godbey, W. T.; Wu, K. K.; Mikos, A. G. Size matters: Molecular weight affects the efficiency of poly(ethylenimine) as a gene delivery vehicle, Journal of Biomedical Material Research, 1999, 45, 268-275.
50 Abdallah B.; Hassan, A.; Benoist, C.; Goula, D.; Behr, J. P.; Demeneix, B. A powerful Nonviral Vector for In Vivo Gene Transfer into the Adult Mammalian Brain: Polyethylenimine, Human Gene Therapy, 1996, 7, 1947-1954.
51 Fischer, D.; Bieber, T.; Li, Y.; Elsasser, H. P.; Kissel, T. A Novel Non-Viral Vector for DNA Delivery Based on Low Molecular Weight, Branched Polyethylenimine: Effect of Molecular Weight on Transfection Efficiency and Cytotoxicity, Pharmaceutical Research, 1999, 16, 1273-1279.
52 Oh, Y.K.; Suh, D.; Kim, J.M.; Choi, H. G.; Shin, K.; Ko, J.J. Polyethylenimine- meidated cellular uptake, nucleus trafficking and expression of cytokine plasmid DNA, Gene Therapy, 2002, 9, 1627-1632.
53 Choi, S.; Lee, K. D. Enhanced gene delivery using disulfide-crosslinked low molecular weight polyethylenimine with listeriolysin o-polyethylenimine disulfide conjugate, Journal of Controlled Release, 2008, 131, 70-76.
54 Guo, W.; Lee, R. J. Efficient gene delivery via non-covalent complexes of folic acid and polyethyleneimine, Journal of Controlled Release, 2001, 77, 131-138.
55 Tag Ko, Y.; Bhattacharya, R.; Bickel, U. Liposome encapsulatedpolyethyleneimine/ODN polyplexes for brain targeting, Journal of Controlled Release, 2009, 133, 230-237.
128
56 Ogris, M.; Brunner, S.; Schuller, S.; Kircheis, R.; Wagner, E. PEGylated DNA/transferrin-PEI complexes: reduced interaction with blood components, extended circulation in blood and potential for systemic gene delivery, Gene Therapy, 1999, 6, 595-605.
57 Orgis, M.; Wagner, E. Targeting tumors with non-viral gene delivery systems, Drug Discovery Today, 2002, 8, 479-485.
58 Kircheis, R.; Wightman, L.; Wagner, E. Design and gene delivery activity of modified polyethylenimines, Adv. Drug Deliv. Rev., 2001, 53, 341-358.
59 Peterson, H.; Fechner, P. M.; Martin, A. L.; Kunath, K.; Stolnik, S.; Roberts, C. J.; Fischer, D.; Davies, M. C.; Kissel, T. Polyethylenimine-graft-Poly(ethylene glycol) Copolymers: Influence of Copolymer Block Structure on DNA Complexation and Biological Activities as Gene Delivery System, Bioconjugate Chemistry, 2002, 13, 845-854.
60 Li, C. Poly(L-glutamic acid)-Anticancer Drug Conjugates, Advanced Drug Delivery Reviews, 2002, 54, 695-713.
61 Wang, C.; Feng, M.; Deng, J.; Zhao, Y.; Zeng, X.; Han, L.; Pan, S.; Wu, C. Poly(α- glutamic acid) combined with polycation as serum-resistant carriers for gene delivery, International Journal of Pharmaceutics, 2010, 398, 237-245.
62 Kurosaki, T.; Kitahara, T.; Fumoto, S.; Nishida, K.; Nakamura, J.; Niidome, T.; Yukinobu, K.; Nakagawa, H.; To, H.; Sasaki, H. Ternary complexes of pDNA, polyethylenimine, and γ-polyglutamic acid for gene delivery systems, Biomaterials, 2009, 30, 2846-2853.
63 Manocha, B.; Margaritis, A. Production and Characterization of γ-Polyglutamic Acid Nanoparticles for Controlled Anticancer Drug Release, Critical Reviews in Biotechnology, 2008, 28, 83-99.
64 Paredes-Lopez, O.; Cervantes-Ceja, M. L.; Vigna-Perez, M.; Hernandez-Perez, T. Berries: Improving Human Health and Healthy Aging, and Promoting Quality Life- A Review, Plant Foods for Human Nutrition, 2010, 65, 299-308.
65 Visioli, F.; Borsani, L.; Galli, C. Diet and prevention of coronary heart disease: the potential role of phytochemicals, Cardiovascular Research, 2000, 47, 419 – 425.
66 Nothlings, U., et al. Intake of Vegetables, Legumes, and Fruit, and Risk for All-Cause, Cardiovascular, and Cancer Mortality in a European Diabetic Population, The Journal of Nutrition, 2008, 138, 775-781.
129
67 La Vecchia, C; Altieri, A.; Tavani, A. Vegetables, fruit, antioxidants and cancer: a review of Italian studies, European Journal of Nutrition, 2001, 40, 261-267.
68 Szajdek A.; Borowska, E. J. Bioactive Compounds and Health-Promoting Properties of Berry Frutis: A review, Plant Foods of Human Nutrition, 2008, 63, 147-156.
69 Li, W.; Hydamaka, A. W.; Lowry, L.; Beta, T. Comparison of antioxidant capacity and phenolic compounds of berries, chokecherryandseabuckthorn, Central Europe Journal of Biology, 2009, 4, 499-506.
70 Seeram, N. P. Berry Frutis: Compositional Elements, Biochemical Activities, and the Impact of Their Intake on Human Health, Performance, and Disease, Journal of Agricultural and Food Chemistry, 2008, 56, 627-629.
71 Nohynek, L. J.; Alakomi, H. L.; Kahkonen, M. P.; Heinonen, M.; Helander, I. M.; Oksman-Caldentey, K. M.; Puupponen-Pimia, R. H. Berry Phenolics: Antimicrobial Properties and Mechanisms of Action Against Severe Human Pathogens, Nutrition and Cancer, 2006, 54, 18-32.
72 Heim, K. E.; Tagliaferro, A. R.; Bobilya, D. J. Flavonoid antioxidants: chemistry, metabolism, and structure-activity relationships, Journal of Nutritional Biochemistry, 2002, 13, 572-584.
73 Hakkinen, S. H.; Torronen, A. R. Content of flavonols and selected phenolic acids in strawberries and Vaccinium species: influence of cultivar, cultivation site and technique, Food Research International, 2000, 33, 517—524.
74 Kellogg, J.; Wang, J.; Flint, C.; Ribnicky, D.; Kuhn, P.; De Mejia, E. G.; Raskin, I.; Lila, M. A. Alaskan Wild Berry Resources and Human Health under the Could of Climate Change, Journal of Agricultural Food and Chemistry, 2010, 58, 3884-3900.
75 Castrejon, A. D. R.; Eichhols, I, Rohn, S.; Kroh, L. W.; Huyskens-Keli, S. Phenolic profile and antioxidant activity of highbush blueberry (Vacciniumcorymbosum L.) during fruit maturation and ripening, Food Chemistry, 2008, 109, 564-572.
76 Wang, S. Y.; Lin, H. S. Antioxidant Activity in fruits and Leaves of Blackberry, Raspberry, and Strawberry Varries with Cultivar and Development Stage, Journal of Agricultural Food and Chemistry, 2000, 48, 140-146.
77 Dosset, M.; Lee, J.; Finn, C. E. Inheritance of Phenological, Vegetative, and Fruit Chemistry Traits in Black Raspberry, Journal of American Society for Horticultural Science, 2008, 133, 408-417.
130
78 Kresty, L. A.; Morse, M. A.; Morgan, C.; Carlton, P. S.; Lu, J.; Gupta, A.; Blackwood, M.; Stoner, G. D. Chemoprevention of Esophageal Tumorigenesis by Dietary Administration of Lyophilized Black Raspberries, Cancer Research, 2001, 61, 6112- 6119.
79 Casto, B. C.; Kresty, L. A.; Kraly, C. L.; Pearl, D. K.; Stoner, G. D.; D‘Ambrosio, S. M.; Mallery, S. R.; Pearl, D. K.; Schut, H. A. J. Chemoprevention of oral cancer by black raspberries, Anticancer Research, 2002, 22, 4005-4015.
80 Harris, G. K.; Gupta, A.; Nines, R. G.;Kresty, L. A.;Habib, S. G.;Frankel, W. L.; LaPerle, K.; Gallaher, D. D.; Schwartz, S. J.; Stoner, G. D. Effects of Lyophilized Black Raspberries on Azoxymethane-Induced Colon Cancer and 8-Hydroxy- 2'Deoxyguanosine Levels in the Fischer 344 Rat, Nutrition and Cancer, 2001, 40, (2),125-33.
81 Dosset, M.; Lee, J.; Finn, C. E. Variations in anthocyanins and total phenolics of black raspberry populations, Journal of Functional Foods, 2010, 2, 292-291.
82 Zettersten, C.; Co, M.; Wende, S.; Turner, C.; Nyholm, L.; Sjoberg, P. J. R. Identification and Characterization of Polyphenolic Antioxidants Using On-Line Liquid Chromatography, Electrochemistry, and Electrospray Ionization tandem Mass Spectrometry, Analytical Chemistry, 2009, 81, 8968-8977.
83 Sanchez-Rabaneda, F.; Jauregui, O.; Lmuela-Raventos, R. M.; Viladomat, F.; Bastida, J.; Codina, C. Qualitative analysis of phenolic compounds in apple pomace using liquid chromatography coupled to mass spectrometry in tandem mode, Rapid Communications in Mass Spectrometry, 2004, 18, 553-563.
84 De Hoffmann, E.; Stroobant, V. Mass Spectrometry Principles and Applications, 3rd ed.; John Wiley & Sons: England, 2007.
85 Seeram, N.; Lee, R.; Scheuller, H. S.; Heber, D. Identification of phenolic compounds in strawberries by liquid chromatography electrospray ionization mass spectroscopy, Food Chemistry, 2006, 97, 1–11.
86 Fabre, N.; Rustan, I.; de Hoffmann, E.; Quetin-Leclercq, J. Determination of Flavone, Flavonol, and FlavanoneAglycones by Negative Ion Liquid Chromatography Electrospray Ion Trap Mass Spectrometry, Journal of American Society for Mass Spectrometry, 2001, 12, 707-715.
87 Dole, M.; Mack, L. L.; Hines, R. L. Molecular Beams of Macroions, The Journal of Chemical Physics, 1968, 49, 2240-2249.
88 Esquire ESI 3000 Operator‘s Manual, Volume 1, Version 4.0, BrukerDaltonics
131
89 Gomez, A.; Tang, K. Charge and fission of droplets in electrostatic sprays, Physics of Fluids, 1994, 6, 404-414.
90 Cole, R. B. Some tenets pertaining to electrospray ionization mass spectrometry, Journal of Mass Spectrometry, 2000, 35, 763-772.
91 Skoog, D. A.; Holler, F. J.; Crouch, S. R. Principles of Instrumental Analysis, 6th ed., Thomson Brooks/Cole, Belmont: CA, 2007.
92 March, R. E. Quadrupole Ion Traps, Mass Spectrometry Reviews, 2009, 28, 961-989.
93 Kanu, A. B.; Dwivedi, P.; Tam, M.; Matz, L.; Hill Jr., H. H. Ion mobility – mass spectrometry, Journal of Mass Spectrometry, 2008, 43, 1-22.
94 Waters Synpat HDMS System Operator‘s Guide, Revision A, Waters Co., 2007.
95 Uetrecht, C.;Rose, R. J.; van Dujin, E.; Lorenzen, K.; Heck, A. J. R. Ion mobility mass spectrometry for proteins and protein assemblies, Chemical Society Reviews, 2010, 39,1633-1655. 96 McMaster, M. C. LC/MS A Practical User‘s Guide, John Wiley and Sons, Hoboken: New Jersey, 2005.
97 Lindsay S.; Kealy D. High Performance Liquid Chromatography, ACOL, John Wiley & Sons: London, England, 1987.
98 Meyer, R. V. Practical High-Performance Liquid Chromatography, John Wiley & Sons: Germany, 1988.
99 Kan, P.L.; Schatzlein, A.G.; Uchegbu, I.F. Polymers in Drug Delivery; I.F. Uchegbu, A.G. Schätzlein, (Eds.), CRC Press: Boca Raton, 2006.
100 Zhang, S.; Xu, Y.; Wang, B.; Qiao, W.; Liu, D.; Li, Z. Cationic compounds used in lipoplexes and polyplexes for gene delivery, Journal of Controlled Release, 2004, 100, 165-180.
101 Remaut, K.; Lucas, B.; Raemdonck, K.; Braeckmans, K.; Demeester, J.; De Smedt, S.C. Protection of Oligonucleotiedes against enzymatic degradation by Pegylated and Nonpegylated Branched Polyethyleneimine, Biomacromolecules, 2007, 8, 1333- 1340.
102 Nyunt, M.T.; Dicus, C.W.; Cui, Y.-Y.; Yappert, M.C.; Huser, T.R.; Nantz, M.H.; Wu, J. Physico-Chemical Characterization of Polylipid Nanoparticles for Gene Delivery to the Liver, Bioconjugate Chemistry, 2009, 20, 2047-2054.
132
103 He, C.-X.; Tabata, Y.; Gao, J.-Q.; Non-viral gene delivery carrier and its three- dimensional transfection system, International Journal of Pharmaceutics, 2010, 386, 232-242.
104 Itaka, K.; Ishii, T.; Hasegawa, Y.; Kataoka, K. Biodegradable polyamino acid-based polycations as safe and effective gene carrier minimizing cumulative toxicity, Biomaterials, 2010, 31, 3707-3714.
105 Terrier, P.; Tortajada, J.; Zin, G.; Buchmann, W. Noncovalent Complexes between DNA and basic polypeptides or polyamines by MALDI-TOF, Journal of American Society for Mass Spectrometry, 2007, 18, 1977-1989.
106 Fenn, J.B.; Mann, M.; Meng, C.K.; Wong, S.F.; Whitehouse, C.M. Electrospray ionization for mass spectrometry of large biomolecules, Science, 1989, 246, 64-71.
107 Pringle, S.D.; Giles, K.; Wildgoose, J.L.; Williams, J.P.; Slade, S.E.; Thalassinos, K.; Bateman, R.H.; Bowers, M.T.; Scrivens, J.H. An investigation of the mobility separation of some peptide and protein ions using a new hybrid quadrupole/traveling wave IMS/oa-ToF instrument, International Journal of Mass Spectrometry, 2007, 261, 1-12.
108 Giles, K.; Pringle, S.D.; Worthington, K.R.; Little, D.; Wildgoose, J.L.; Bateman, R.H. Application of travelling wave-based radio-frequency-only stacked ring ion guide, Rapid Communication in Mass Spectrometry, 2004, 18, 2401-2414.
109 Laskin, J.; Denisov, E.; Futrell, J. A Comparative Study of Collision-Induced and Surface-Induced Dissociation: Fragmentation of Protonated Dialanine, Journal of American Chemical Society, 2000, 122, 9703-9714.
110 Gabelica, V.; De Pauw, E. Comparison between solution-phase stability and gas-phase kinetic stability of oligodeoxynucleotide duplexes, Journal of Mass Spectrometry, 2001, 36, 397-402.
111 Pan, S.; Sun, X.; Lee, J.K. DNA Stability in the Gas Versus Solution Phases: A Systematic Study of Thirty-One Duplexes with Varying Length, Sequence, and Charge Level, Journal of American Society for Mass Spectrometry, 2006, 17, 1383- 1395.
112 Deschenes, L.A.; Vanden Bout, D.A. Origin 6.0: Scientific Data Analysis and Graphing Software Origin, Journal of American Chemical Society, 2000, 122, 9567.
133
113 Wan, K.X.; Gross, M.L.; Shibue, T. Gas-Phase Stability of Double-Stranded Oligodeoxynucleotides and Their Noncovalent Complexes with DNA-Binding Drug As Revealed by Collisional Activation in an Ion Trap, Journal of American Society for Mass Spectrometry, 2000, 11, 450-457.
114 Wan, K.X.; Shibue, T.; Gross, M.L. Non-Covalent Complex between DNA-Binding Drugs and Double-Stranded Oligodeoxynucleotides: A Study by ESI Ion-Trap Mass Spectrometry, Journal of American Chemical Society, 2000, 122, 300-307.
115 Gabelica, V.; Rosu, F.; Houssier, C.; De Pauw, E. Gas-phase thermal denaturation of an oligonucleotide duplex and its complexes with minor groove binders, Rapid Communication in Mass Spectrometry, 2000, 14, 464-467.
116 Sun, X.; Lee, J.K. Stability of DNA Duplexes Containing Hypoxanthine (Inosine): Gas versus Solution Phase and Biological Impacts, Journal of Organic Chemistry, 2010, 75, 1848-1854.
117 E. Rivera-Tirado, Association and Fragmentation Characteristics of Biomolecules and Polymers Studied by Mass Spectrometry, Ph.D. Dissertation, The University of Akron, 2007
118 Lee, C.-H.; Ni, Y.-H.; Chen, C.-C.; Chou, C.-K.; Chang, F.-H. Synergistic effect of polyethylenimine and cationic liposomes in nucleic acid delivery to human cells, BiochimicaetBiophysicaActa, 2003, 1611, 55-62.
119 Puglisi, C.; Samperi, F.; Carroccio, S.; Montaudo, G. Analysis of Poly(bisophenol A carbonate) by Size Exclusion Chromatography/Matrix-assisted Laser Desorption/Ionization. 2. Self-association Due to Phenol End Groups, Rapid Communication in Mass Spectrometry, 1999, 13, 2268-2277.
120 Puglisi, C.; Samperi, F.; Carroccio, S.; Montaudo, G. SEC/MALDI analysis of poly(bisphenol a carbonate) self association due to phenol end groups, Polymer Preprints, 2000, 41(1), 653-654.
121 vonHelden, G.; Hsu, M.-T.; Kemper, P.R.; Bowers, M.T. Structures of carbon cluster ions from 3 to 60 atoms: linear to rings to fullerenes, Journal of Chemical Physics, 1991, 95, 3835-3837.
122 Bowers, M.T.; Kemper, P.R.; von Helden, G.; van Koppen, P.A.M. Gas-Phase Ion Chromatography Transition Metal State Selection and Carbon Cluster Formation, Science,1993, 260, 1446-1451.
123 Clemmer, D.E.; Jarrold, M.F Ion Mobility Measurements and their Applications to Cluster and Biomolecules, Journal of Mass Spectrometry, 1997, 32, 577-592.
134
124 McLean, J.A.; Ruotolo, B.T.; Gillig, K.J.; Russell, D.H. Ion mobility mass- spectrometry: a new paradigm for proteomics, International Journal of Mass Spectrometry, 2005, 240, 301-315.
125 Wyttenbach, T.; Bowers, M.T. Intermolecular Interactions in Biomolecular Systems Examined by Mass Spectrometry, Annual Review of Physical Chemistry, 2007, 58, 511-533.
126 Kanu, A.B.; Dwivedi, P.; Tam, M.; Matz, L.; Hill Jr., H.H. Ion mobility-mass spectrometry, Journal of Mass Spectrometry, 2008, 43, 1-22.
127 Fenn, L.S.; McLean, J.A. Biomolecular structural separation by ion mobility-mass spectrometry, Analytical and Bioanalytical Chemistry, 2008, 391, 905-909.
128 Bohrer, B.C.; Merenbloom, S.I.; Koeniger, S.L.; Hilderbrand, A.E.; Clemmer, D.E. Biomolecule Analysis by Ion Mobility Spectrometry, Annual Review of Analytical Chemistry, 2008, (1), 293-327.
129 Trimpin, S.; Clemmer, D. E. Ion Mobility Spectrometry/ Mass Spectrometry Snapshots for Assessing the Molecular Composition of Complex Polymeric Systems, Analytical Chemistry, 2008, 80, 9073-9083.
130 Clemmer, D.E.; Valentine, S.J. Bioanalytical chemistry: Protein oligomers frozen in time, Nature Chemistry, 2009, (1), 257-258.
131 Chan, Y.-T.; Li, X.; Soler, M. Wang, J.-L.; Wesdemiotis, C.; Newkome, G.R. Self- Assembly and Traveling Wave Ion Mobility Mass Spectrometry Analysis of HexacadmiumMacrocycles, Journal of American Chemical Society, 2009, 131, 16395-16397.
132 McLean, J. A. The Mass-Mobility Correlation Redux: The Conformational Landscape of Anhydrous Biomolecules, Journal of American Society for Mass Spectrometry, 2009, 20, 1775-1781.
133 Lungwitz, U.; Breuing, M.; Blunk, T.; Gopferich, A. Polyethyleneimine-based non- viral gene delivery systems, European Journal of Pharmaceutics and Biopharmaceutics, 2005, 60, 247-266.
134 Kircheis, R.; Wighhtman, L.; Wagner, E. Design and gene delivery activity of modified polyethylenimines, Adv. Drug Deliv. Rev., 2001, 53, 341-358.
135 Guo, W.; Lee, R. J. Efficient gene delivery via non-covalent complexes of folic acid and polyethyleneimine, Journal of Controlled Release, 2001, 77, 131-138.
135
136 Tag Ko, Y.; Bhattacharya, R.; Bickel, U. Liposome encapsulatedpolyethyleneimine/ODN polyplexes for brain targeting, Journal of Controlled Release, 2009, 133, 230-237.
137 Yang, J.P.; Huang, L. Overcoming the inhibitory effect of serum on lipofection by increasing the charge ratio of cationic liposome to DNA, Gene Therapy, 1997, 4, 950-960.
138 Ogris, M.; Brunner, S.; Schuller, S.; Kircheis, R.; Wagner, E. PEGylated DNA/transferrin-PEI complexes: reduced interaction with blood components, extended circulation in blood and potential for systemic gene delivery, Gene Therapy, 1999, 6, 595-605.
139 Haberland, A.; Knaus, T.; Zaitsev, S.V.; Buchberger, A.L.; Haller, H.; Bottger, M. Histone H1-Meidated Trasfection: Serum Inhibition Can Be Overcome by Ca2+ ions, Pharmaceutical Research, 2000, 17, 229-235.
140 Shen, Y.; Deng, J.; Luo, X.; Zhang, X.; Zeng, X.; Feng, M.; Pan, S.R. Synthesis and characterization of sterically stabilized polyelectrolyte using isophoronediisocyanate as the coupling reagent, Journal of Biomaterial Science, Polymer Edition, 2009, 20, 1217-1233.
141 Wang, C.; Feng, M.; Deng, J.; Zhao, Y.; Zeng, X.; Han, L.; Pan, S.; Wu, C. Poly(α- glutamic acid) combined with polycation as serum-resistant carriers for gene delivery, International Journal of Pharmaceutics, 2010, 398, 237-245.
142 Manocha, B.; Margaritis, A. Production and Characterization of γ-Polyglutamic Acid Nanoparticles for Controlled Anticancer Drug Release, Critical Reviews in Biotechnology, 2008, 28, 83-99.
143 Kurosaki, T.; Kitahara, T.; Fumoto, S.; Nishida, K.; Nakamura, J.; Niidome, T.; Yukinobu, K.; Nakagawa, H.; To, H.; Sasaki, H. Ternary complexes of pDNA, polyethylenimine, and γ-polyglutamic acid for gene delivery systems, Biomaterials, 2009, 30, 2846-2853.
144 Smiljanic, D.; Wesdemiotis, C. Non-covalent complexes between single-stranded oligodeoxynucleotides and poly(ethylene imine), International Journal of Mass Spectrometry, 2011, DOI: 10.1016/j.ijms.2010.06.035.
145 Schmidt, A.; Bahr, U.; Karas, M. Influence of Pressure in the First Pumping Stage on Analyte Desolvation and Fragmentation in Nano-ESI MS, Analytical Chemistry, 2001, 73 6040-6046.
136
146 Smith, D.P.; Knapman, T.W.; Campuzano, I.; Malham, R.W.; Berryman, J.T.; Radford, S.E.; Ashcroft, A.E. Deciphering drift time measurments from travelling wave ion mobility spectrometry-mass spectrometry studies, European Journal of Mass Spectrometry, 2009, 15, 113-130.
147 Tahallah, N.; Pinkse, M.; Maier, C.S.; Heck, A.J.R. The effect of the source pressre on the abundace of ions of noncovalent protein assemblies in an electrospray ionization orthogonal time-of-flight instrument, Rapid Communications in Mass Spectrometry, 2001, 15, 596-601.
148 Chernushevich, I.V.; Thomson, B.A. Collisional Cooling of Large Ions in Electrospray Mass Spectrometry, Analytical Chemistry, 2004, 76, 1754-1760.
149 Sanglier, S.; Atmanene, C.; Chevreux, G.; Dorsselaer, A.V. Nondenaturing Mass Spectrometry to Study Noncovalent Protein/Protein and Protein/Ligand Complexes: Technical Aspects and Application to the Determination of Binding Stoichiometries, Thompson et al., (eds.), Functional Proteomics: Methods in Molecular Biology, vol.484: Methods and Protocols, Humana Pres: Totowa, 2008, p. 217-243.
150 Sanglier, S.; Ramstrom, H.; Haiech, J.; Leize, E.; Dorsselaer, A.V. Electrospray ionization mass spectrometry analysis revealed a ~310 kDanoncovalentheamer of HPr kinase/phosphatase from Bacillus subtilis, International Journal of Mass Spectrometry, 2002, 219, 681-696.
151 Zubay, J. Biochemistry, 2nd edition, McMillan Publishing Company: 1988 (p. 219).
152 Greco, F.; Liguori, A.;Sindona, G.; Uccella U. Gas-phase proton affinity of deoxyribonucleosides and related nucleobases by fast atom bombardment tandem mass spectrometry, Journal of American Chemical Society, 1990, 112, 9092.
153 Alves, S.; Woods, A; Delvolve, A.; Tabet, J.C. Influence of salt bridge interactions on the gas phase stability of DNA/peptide complexes, International Journal of Mass Spectrometry, 2008, 278, 122-128.
154 Nir, E.; Imhof, P.; Kleinermanns, K.; S. de Vriers, M. REMPI Spectroscopy of Laser Desorbed Guanosines, Journal of American Chemical Society, 2000, 122, 8091- 8092.
155 Chipuk, J.E.; Brodbelt, J.S. Gas-Phase Hydrogen/Deuterium Exchange of 5‘- and 3‘- Mononucleotides in a Quadrupole Ion Trap: Exploring the Role of Conformation and System Energy, Journal of American Society for Mass Spectrometry, 2007, 18, 724-736.
137
156 Vekey, K.; Somoqyi, A.; Wysocki, V.H. Average Activation Energies of Low-energy Fragmentation Processes of Protonated Peptides Determined by a New Approach, Rapid Communications in Mass Spectrometry, 1996, 1, 911-918.
157 Zafra-Stone, S.; Yasmin, T.;Bagchi, M.; Chatterjee, A.; Vinson, J. A.; Bagch, D. Berry anthocyanins as novel antioxidants in human health and disease prevention, Molecular Nutrition and Food Research, 2007, 51, 675-683.
158 Stoner, G. D.; Wanga, L.-S.; Zikrib, N.; Chenc, T.; Hechtd, S. S.; Huange, C.; Sardoc, C.; Lechner, J.F. Cancer prevention with freeze-dried berries and berry components, Seminars in Cancer Biology, 2007, 17, 403-410.
159 Kresty, L. A.; Frankel, W. L.; Hammond, C. D.; Barid, M. E.; Mele, J. M.; Stoner, G. D.; Fromkes, J. J. Transitioning From Preclinical to Clinical Cchemopreventative Assessments of Lyophilized Black Raspberries: Interin Results Show Berries Modulate Markers of Oxidative Stress in Barrett‘s Esophagus Patients, Nutrition and Cancer, 2006, 54, (1), 148-156.
160 Stoner, G. D.; Sardo, C.; Apseloff, G.; Mullet, D.; Wargo, W.; Pound, V.; Singh, A.; Sanders, J.; Aziz, R.; Casto, B.; XiaoLi, S. Pharmacokinetics of ANthocyanins and Ellagic Acid in Helathy Volunteers Fed Freeze-Dried Black Raspberries Dailty for 7 Days, Journal of Clinical Pharmacology, 2005, 45, 1153-1164.
161 Geng, P.; Sun, J.; Zhang, R.; He, J.; Abliz, Z. An investigation of the fragmentation differences of isomeric flavonol-O-glycosides under different collision-induced dissociation based mass spectrometry, Rapid Communications in Mass Spectrometry, 2009, 23, 1519-1524.
162 Dou, J.; Lee V. Y.; Tzen, J. T. C.; Lee, M.-R. Identification and Comparison of Phenolic Compounds in the Preparation of Oolong Tea Manufactured by Semifermentation and Drying Processes, Journal of Agricultural Food and Chemistry, 2007, 55, (18), 7462-7468.
163 Stella, L.; De Rosso, M.; Panighel, A.; Vedova, A. D.; Riccardo, F.; Traldi, P. Collisionally induced fragmentation of [M-H]- species of resveratrol and piceatannol investigated by deuterium labeling and accurate mass measurements, Rapid Communications in Mass Spectrometry, 2008, 22, 3867-3872.
164 Lui, S.-Y.; Wen, J.-F.; Kim, S.-J.; Lee, Y.-I. Investigation of the Interaction Between Sodium (meta) Arsenite and Catechin via ESI Tandem Mass Spectrometry, Chemistry Research in Chinese Universities, 2007, 23, 524-529.
138
165 Winding, W.; Phalp, J. M.; Payne, A. W. A Noise and Background Reduction Method for Component Detection in Liquid Chromatography/Mass Spectrometry, Analytical Chemistry, 1996, 68, 3602-3606.
166 Jokopic, J.; MikulivPetrovsek, M.; Solar, A.; Stampar, F.; Veberic, R. HPLC-MS identification of phenols in hazelnut (Corylusavellana L.) kernels, Food Chemistry, 2011, 124, 1100-1106.
167 Bonat Celli, G.; BellarninoPerreira-Netto, A.; Beta, T. Comparative analysis of total phenolic content, antioxidant activity, and flavonoids profile of fruits from two varieties of Brazilian cherry (Eugenia uniflora L.) throughout the fruit developmental stages, Food research International, 2011, 44, 2245-2251.
139
APPENDICES:
140 APPENDIX A
ADDITIONAL DATA
Figure A1. Chromatogram for fraction 1-H utilizing HPLC-UV.
141
Figure A2. Chromatogram for fraction 1-E utilizing HPLC-UV.
Figure A3. Chromatogram for fraction 1-K utilizing HPLC-UV.
142
Figure A4. Total ion chromatogram (TIC) for the standard of quercetin (top) and the extracted mass spectrum (bottom) for the peak eluting at 42.9 min.
143
Figure A5. Total ion chromatogram (TIC) for the standard of myricetin (top) and the extracted mass spectrum (bottom) for the peak eluting at 36.5 min.
144
APPENDIX B
COPYRIGHT PERMISSIONS
145
146
147
148
149
150
151
152
153
154