MASS SPECTROMETRY METHODS FOR THE ANALYSIS OF
POLYMERS AND BIOCONJUGATES
A Dissertation
Presented to
The Graduate Faculty of the University of Akron
In Partial Fulfillment
of the Requirements for the Degree
Doctor of Philosophy
Sahar Sallam
December, 2017 MASS SPECTROMETRY METHODS FOR THE ANALYSIS OF
POLYMERS AND BIOCONJUGATES
Sahar Sallam
Dissertation
Approved: Accepted:
______Advisor Department Chair Dr. Chrys Wesdemiotis Dr. Christopher J. Ziegler
______Committee Member Dean of the Collage Dr. Wiley J. Youngs Dr. John C. Green
______Committee Member Dean of the Graduate School Dr. Yi Pang Dr. Chand K. Midha
______Committee Member Date Dr. Sailaja Paruchuri
______Committee Member Dr. Matthew L. Becker
ii ABSTRACT
This dissertation focuses on the characterization of complex biomaterials by mass spectrometry (MS), tandem mass spectrometry (MS/MS) and ion mobility mass spectrometry (IM-MS). Chapter I provides a brief summary of the analytical problems addressed in this dissertation. Chapter II discusses the basic principles of mass spectrometry, including ionization techniques, mass analyzers, tandem mass spectrometry, and ion mobility. Chapter III describes the materials and instruments used to accomplish this work.
Chapters IV, V, VI and VII are research project chapters, and each is briefly introduced below. Finally, Chapter VIII summarizes the conclusions drawn from this dissertation followed by an appendix and the copyright permissions obtained for this dissertation.
The analysis of isomeric biodegradable polyesters is discussed in chapter IV.
Matrix assisted laser desorption ionization (MALDI) and electrospray ionization (ESI) coupled with time-of-flight (ToF) mass analysis and tandem mass spectrometry (MS2) fragmentation were used to elucidate the composition, end groups, the chain sequence of poly(propylene maleate) (PPM) and poly(propylene fumarate) (PPF) copolymers.
Additionally, ion mobility mass spectrometry (IM-MS) was used to differentiate the isomeric PPM and PPF copolyesters and to probe the extent and efficiency of PPM to
PPF (i.e., all-cis to all-trans) isomerization.
The analysis of the alanine-rich peptides AQK18 and GpAQK18 and their poly(ethylene glycol)-conjugated (PEGylated) forms AQK18-PEG and GpAQK18-PEG,
iii is discussed in chapter V. In this work, ESI and MALDI interfaced with MS2 fragmentation and shape-sensitive separation via ion mobility mass spectrometry (IM-
MS), were applied to elucidate the sequence and conformation of alanine-rich polypeptides and their conjugates. IM-MS data revealed the existence of two conformers for both the polypeptides and the conjugates. These were assigned to a fast-drifting random coil and a more slowly drifting helical structure. The collision cross- sections of the random coil and helical conformers of the polypeptides and their PEG conjugates were simulated for comparison with the experimental values to confirm the identity of the observed architectures and understand the stabilizing effects of the polymer chain.
Chapter VI focuses on the structural characterization of a branched glycopolymer. Reversible addition-fragmentation chain transfer (RAFT) copolymerization of galactose acrylate and a polymerizable chain transfer agent branching unit was used to prepare this glycopolymer. Multistage mass spectrometry
(MSn) and ion mobility were employed in this work to unveil useful information about the glycopolymer composition, possible end groups and architecture.
Chapter VII discusses the analysis of a trehalose glycopolymer and its insulin conjugate. The goal of this study was to provide full characterization and detailed information about the chemical composition of the trehalose glycopolymer and its insulin conjugate and to determine the degree and the sites of conjugation by using multidimensional MS, including MALDI-MS, ESI-MS, and MS2.
iv DEDICATION
To my husband Ehab Zakri and my wonderful daughters (Lojain,
Faridah and Emadiah) who encouraged and supported me and made this possible.
To my lovely parents Awad Sallam and Faridah Abu Gazala for their love and unlimited support.
v ACKNOWLEDGEMENTS
First of all, I would like to thank Almighty Allah (God) for giving me the opportunity, strength, and knowledge to commence this research study. Without His blessings, this achievement would be impossible.
I would also like to express my sincere gratitude to my advisor Dr. Chrys
Wesdemiotis for his guidance, kindness, and assistance during my study at the
University of Akron. It was a pleasure to work under his supervision during my research work as he was always available and willing to help.
Thanks are also due to my committee members Dr. Sailaja Paruchuri, Dr. Yi
Pang, Dr. Willy Youngs and Dr. Matthew L. Becker for serving as my committee members, spending their valuable time to correct my dissertation, and giving me invaluable guidance and suggestions.
I would like to thank and acknowledge the Ministry of Higher Education and
Jazan University, Kingdom of Saudi Arabia for giving me the opportunity to pursue my graduate studies and supporting me through the past four years. Also, I would like to express my gratitude to the Department of Chemistry at the University of Akron for their extensive comments and assistance during my study.
In particular, I would like to thank my collaborators at Dr. Becker’s group (from the University of Akron), Dr. Jia’s group (from the University of Delaware), Dr.
vi Maynard’s group and Dr. Kasko’s group (from the University of California Los
Angeles) for their support and for providing me the samples used in this dissertation.
Special thanks go to the past group members Dr. Nadrah Alawani, Dr.
Ahlam Alalwiat, Dr. Xiumin Liu, Dr. Kai Guo and Dr. Lydia Cool for sharing their knowledge and expertise in mass spectrometry with me. Further thanks go to the current group members Dr. Ivan Dolog, Nicolas Alexander, Selim Gerislioglu,
Kevin Endres, Michelle Kushnir, Jialin Mao, Savannah Snyder, Jason O’Neill and
Chen Du for their friendship, help and support. I also wish to thank all my Akron friends for their friendship and support over the past years. They have always made me feel welcomed and never left my side.
My acknowledgement would be incomplete without thanking the main source of my strength, my family. The unlimited love and support of my parents
(Mr. Awad Sallam & Mrs. Faridah Abu Gazala), my sisters (Samar, Sumyyah and
Samah) and my brothers (Sallam, Mohamed and Ali) have made this project possible. Finally, no words can express how grateful and thankful I am to have my lovely husband Ehab Zakri and my adorable daughters (Lojain, Faridah, and
Emadiah) in my life. You’ve always believed in me and stood by my side during hard times.
vii TABLE OF CONTENTS
LIST OF TABLES ...... xii
LIST OF FIGURES ...... xiii
LIST OF SCHEMES ...... xx
ACRONYMNS AND ABBREVIATIONS ...... xxi
CHAPTER
I. INTRODUCTION ...... 1
II. MASS SPECTROMETRY BACKGROUND ...... 6
2.1. Mass Spectrometry ...... 6
2.1.1. Ionization methods ...... 7
2.1.2. Mass Analyzers...... 14
2.1.3 Detectors ...... 23
2.2. Tandem Mass Spectrometry ...... 25
2.2.1 Collisionally Activated Dissociation...... 28
2.2.2 Electron Transfer Dissociation (ETD) ...... 28
2.3 Ion mobility (IM-MS) ...... 28
III. MATERIALS AND INSTRUMENTATION ...... 34
3.1 Materials ...... 34
3.2 Instrumentation ...... 35
3.2.1 Ultraflex III ToF/ToF Mass Spectrometer ...... 35
3.2.2 Synapt HDMS Mass Spectrometer ...... 36
viii 3.2.3 HCT Ultra II ESI-QIT Mass Spectrometer ...... 38
IV. MULTIDIMENSIONAL MASS SPECTROMETRY CHARACTERIZATION OF ISOMERIC
BIODEGRADABLE POLYESTERS ...... 40
4.1 Introduction...... 40
4.2 Experimental ...... 42
4.2.1 Materials ...... 42
4.2.2 MALDI experiments ...... 43
4.2.3 ESI-MS and ESI-IM-MS experiments ...... 44
4.2.4 Molecular modeling ...... 45
4.3 Results and discussion ...... 46
4.3.1 MALDI characterization of PPM and PPF ...... 46
4.3.2 MS2 characterization of PPM and PPF ...... 47
4.3.3 IM-MS differentiation of PPF and PPM ...... 50
4.4 Conclusions ...... 59
V. SEQUENCE AND CONFORMATIONAL ANALYSIS OF POLYPEPTIDE-POLYMER
BIOCONJUGATES BY MULTIDIMENSIONAL MASS SPECTROMETRY ...... 60
5.1 Introduction ...... 60
5.2 Materials and methods ...... 63
5.2.1 Materials ...... 63
5.2.2 MALDI-MS and MS2 experiments ...... 64
5.2.3 ESI-MS, IM-MS, and MS2 experiments ...... 65
5.2.4 Molecular modeling ...... 66
5.3 Results and discussion ...... 66
5.3.1 MALDI-MS and MALDI-MS2 analysis of peptides AQK18 and GpAQK18
………………………………………………………..………………………………………..66
ix 5.3.2 ESI-MS (MS2) and ESI-IM-MS analysis of peptides AQK18
and GpAQK18...... 70
5.3.3 MALDI-MS and MALDI-MS2 analysis of bioconjugates AQK18-PEG
and GpAQK18-PEG ...... 76
5.3.4 ESI-IM-MS (MS2) analysis of bioconjugates AQK18-PEG and
GpAQK18-PEG ...... 80
5.3.5 Dependence of helical content and stability on bioconjugation and salt . 83
5.3.6 Structure and collision cross-section simulations ...... 86
5.4 Conclusion ...... 92
VI. MASS SPECTROMETRY CHARACTERIZATION OF GLYCOPOLYMERS WITH
CONTROLLED BRANCHING ...... 93
6.1 Introduction ...... 93
6.2 Experimental ...... 96
6.2.1 Materials ...... 96
6.2.2 Matrix-assisted laser desorption/ionization (MALDI) experiments ...... 96
6.2.3 Electrospray ionization (ESI) experiments ...... 97
6.2.4 Ion mobility mass spectrometry (IM-MS) experiments ...... 98
6.3 Results and disscussion ...... 98
6.3.1 MALDI-MS and tandem mass spectrometry analysis...... 98
6.3.2 ESI-MS and Ion mobility mass spectrometry analysis ...... 102
6.3.3 Tandem mass spectrometry analysis…………………………………….. 102
6.3.4 Structural Information from Collision Cross-Sections ...... 112
6.3.4.1 Derivation of Collision Cross-Sections from Traveling
Wave IM-MS Experiments ...... 112
6.4 Conclusion ...... 124
x VII. MASS SPECTROMETRY CHARACTERIZATION OF TREHALOSE GLYCOPOLYMER
AND ITS INSULIN CONJUGATE ...... 125
7.1 Introduction ...... 125
7.2 Experimental ...... 127
7.2.1 Materials ...... 128
7.2.2 MALDI experiments ...... 129
7.2.3 ESI-MS experiments ...... 130
7.3 Results and discussion ...... 130
7.3.1 MALDI-MS analysis of the trehalose and trehalose-insulin conjugate . 130
7.3.2 ESI-MS and MS2 characterization of the trehalose-insulin conjugate .. 135
7.4 Conclusions ...... 139
VIII. SUMMARY ...... 141
IX. REFRENCES ...... 143
xi LIST OF TABLES
Table Page
2. 1. Common MALDI matrices …………………………………………………………………………….13
4. 1. Measured and calculated collision cross-sections (CCS) of singly sodiated PPF and PPM
+ oligomers with the composition [Rn+C2H6O+Na] and CH3CH2O– and –H end groups………..53
+z 4. 2. Corrected collision cross-section of the polyalanine calibrant ions, [H(Ala)nOH+zH] (z=1-2), deduced from drift times measured at a traveling wave velocity of 250 m/s and a traveling wave height of 7.5 V………………………………………………………………………………………….58
5. 1. Experimental collision cross-sections (Ω)…………………………………………………………...83
5. 2. Simulated structures and collision cross-section (Ω)………………...…………………………….85
6. 1. Experimental collision cross-sections of singly charged glycopolymer ions with different end
groups derived using the calibration plot of Figure 6.6…………………………………………….119
6. 2. Experimental collision cross-sections of doubly and triply charged glycopolymer ions with different end groups derived using the calibration plot of Figure 6.6...... 121
xii TABLE OF FIGURES
Figur e Page 2. 1 General components of a mass spectrometer...... 7
2. 2 Droplet productions in the electrospray interface. Reproduced with permission from reference 1. (b) Droplet decomposing in an electrospray source according to the Rayleigh limit ...... 9
2. 3 Schematic of the Coulombic explosion of a charged droplet ...... 10
2. 4 Illustration of the MALDI process ...... 12
2. 5 Diagram illustrating the concept...... 15
2. 6 Schematic of a linear ToF tuned to detect positive ions produced by MALDI ...... 17
2. 7 Schematic of a reflectron ToF mass analyzer ...... 18
2. 8 Quadrupole instrument made up of the source, the focusing lenses, the quadruple rods and the detector ...... 19
2. 9 Stability areas as a function of U and V for three ions with different masses: m1, m2 and m3 20
2. 10 A 3D ion trap mass analyzer ...... 22
2. 11 Schematic diagram of a Daly detector...... 24
2. 12. (A) Cross-section of a microchannel plate and (B) electron multiplication within a channel. ... 25
2. 13 Schematic of a tandem mass spectrometry in-space and in-time ...... 27
2. 14 Traveling wave device inside the Synapt HDMSTM Q/ToF mass spectrometer ...... 29
2. 15 Stacked ring electrodes (top); schematic of the operation of a traveling wave ion guide...... 3
3. 1 Scheme of the Waters Synapt HDMSTM ESI-Q/ToF mass spectrometer………………… ...... 37
3. 2 The scheme of ion transmission and detection in the HCTultra II QIT mass ...... 39
4. 1 MALDI mass spectra of (a) poly(propylene maleate) (PPM) and (b) poly(propylene fumarate)
+ (PPF). All ions are sodiated species with the composition [Rn+EGs+Na] , where R and EGs
designate the PPM/PPF repeat unit (C7H8O4, 156 Da) and the corresponding end groups (in red color), respectively ...... 45
2 + 4. 2 MALDI-MS spectrum of the [M+Na] ion from the PPF 9-mer with CH3CH2O– and –H end groups (m/z 1473.4). The scheme on the top shows the fragment ions arising from 1,5-
hydrogen rearrangement over ester groups facing the CH3CH2O- ($, !) or -H (#, @) chain end. The consecutive dissociation of these fragments () leads to internal fragments (o).
xiii The Na+ ion has been omitted for brevity. An aesterisk fragment notation (*) indicates fragments ionized by H+ (Na+ is eliminated with the neutral fragment) ...... 47
+ 4. 3 ESI-IM-MS2 spectra of the [M+Na] ions of the 6-mers with CH3CH2O– and –H end groups (m/z 1005.2) from (a) PPF and (b) PPM. An asterisk above fragment notation (*) indicates fragments ioned by H+. These spectra were acquired on the singly charged component of m/z 1005.2 after the ion mobility separation from multiple charged isobars (vide infra) ...... 48
4. 4 (a) 2-D ESI-IM-MS plot (m/z vs. drift time) of PPM; the mobility regions of singly, doubly and triply charged ions are encased in ovals. (b) Mass spectrum extracted from the region of singly charged
ions, containing several ion distributions which include intact PPM ions with CH3CH2O– and –H end groups (46-Da end group mass) and degradation products with various end group masses (noted after the number of repeat units; see Figure 2 for plausible structures)...... 51
4. 5 (a) 2-D ESI-IM-MS plot (m/z vs. drift time) of PPF; the mobility regions of singly, doubly and triply charged ions are encased in ovals. (b) Mass spectrum extracted from the region of singly charged
ions, containing several ion distributions which include intact PPF ions with CH3CH2O– and –H end groups (46-Da end group mass) and degradation products with various end group masses (noted after the number of repeat units; see Figure 2 for plausible structures). The charge is
+ + + provided by the addition of H , Na or (C2H5)2NH2 (from residual PPM to PPF isomerization reagent). PPM leads to very similar ESI-IM-MS characteristics except for the absence of
+ (C2H5)2NH2 adducts...... 54
4. 6. ESI-IM-MS drift time distributions (IM-MS chromatograms) of [M+Na]+ ions form PPF (top) or
PPM (bottom) oligomers with CH3CH2O– and –H end groups and (a) 4, (b) 7 or (c) 10 repeat units...... 55
4. 7 ESI-IM-MS drift time distributions (IM-MS chromatograms) of mass selected m/z 1093.3 from (a) PPF and (b) PPM. These ions contain superimposed singly protonated 6-mers and doubly
protonated 13-mers; their end groups have the same composition as the repeat unit (C7H8O4). The collision cross-sections deduced from the shown drift times are (a) 382 Å2 for the PPF-13- mer and 275 Å2 for the PPF 6-mer; (b) 423 Å2 for the PPM 13-mer and 269 Å2 for the PPM 6-mer...... 56
4. 8 Drift time calibration curve is obtained by plotting the corrected drift times of singly and doubly protonated polyalanine oligomers against the corresponding normalized collision crosssections. Enlisted in Table 1 above are the data used to construct this curve...... 57
5. 1 Amino acid sequence of the peptides and PEGylated peptides investigated...... 62
xiv 5. 2 (a) MALDI-MS spectrum of AQK18 peptide. (b) MALDI-MS2 spectrum of protonated AQK18 (m/z 1680.9); bn and yn designate fragment ions that retain the N- or C-terminus, respectively, which are formed via peptide bond cleavages. CO loss from bn generates the an fragments; ions
labeled by single-letter code are internal fragments formed by consecutive fragmentation of bn or
2 yn (c) Peptide sequence corroborated by the MS fragments...... 68 5. 3 MALDI-MS spectrum of GpAQK18 peptide. (b) MALDI-MS2 spectrum of protonated GpAQK18
(m/z 1776.0); bn and yn designate fragment ions that retain the N- or C-terminus, respectively, which are formed via peptide bond cleavages. Ions labeled by single-letter code are internal
fragments formed by consecutive fragmentation of bn or yn. The fragments observed corroborate the sequence shown in Figure 1……………………………………………………………………...69 5. 4 (a) ESI-MS spectrum of AQK18 peptide. Peaks at m/z values labeled with a superscripted # arise from incomplete sequences (missing either one Lys or one Ala residue). Peaks without m/z labels are fragments from the doubly or triply charged peptide. (b) ESI-MS2 spectrum of doubly protonated AQK18 (m/z 840.952), acquired at a collision energy of 40 eV. The bn and yn
fragment series observed corroborate the sequence Ac-KAAAQAAAQAAAQAAAQK-NH2 (see Figure 1)...... 70
5. 5 (a) ESI-MS spectrum of GpAQK18 peptide (Gp = propargyl glycine). Peaks at m/z values labeled with a superscripted # or @ arise from sequences missing one Lys or one Ala residue (#) and sequences containing an extra Ala residue (@). Peaks without m/z labels are fragments from the doubly or triply charged peptide. (b) ESI-MS2 spectrum of doubly protonated GpAQK18 (m/z 888.470; Gp = propargyl glycine), acquired at a collision energy of 40 eV. The bn and yn
fragment series observed corroborate the sequence Ac-GpKAAAQAAAQAAAQAAAQK-NH2 (see Figure 1)...... 71
5. 6 (a) 2D ESI-IM-MS plot (m/z vs. drift time) of AQK18 peptide and (b,c) mass spectra extracted from the mobility regions of (b) doubly and (c) triply charged AQK18 ions...... 72
5.7 (a) 2D ESI-IM-MS plot (m/z vs. drift time) of GpAQK18 peptide and (b,c) mass spectra extracted from the mobility regions of (b) doubly and (c) triply charged GpAQK18ions………………..…..73
5. 8 IM-MS drift time distributions (mobilograms) of (a) doubly protonated vs. protonated-sodiated AQK18 and (b) quadruply protonated vs. triply protonated-sodiated AKQ18-PEG70. Drift times are marked next to the peaks; see Table 5.1 for the corresponding Ω values...... 74
5. 9 IM-MS drift time distributions (mobilograms) of (a) doubly protonated vs. protonated-sodiated GpAQK18 and (b) quadruply protonated vs. triply protonated-sodiated GpAKQ18-PEG69. Drift times are marked next to the peaks; see Table 5.1 for the corresponding Ω values...... 75
5. 10 MALDI-MS spectrum of C-terminally PEGylated AQK18; the inset shows an expanded view of the m/z 4600-4800 range. The minor peaks between those of the major distribution correspond to potassiated and protonated oligomers...... 77
xv 5. 11 MALDI-MS spectrum of C-terminally PEGylated GpAQK18; the inset shows an expanded view of the m/z 4400-4700 range. The minor peaks between those of the major distribution correspond to potassiated and protonated oligomers...... 78
5. 12 MALDI-MS2 spectra of the [M + Na]+ ions from (a) AQK18 (m/z 1702.9) and (b) AQK18-PEG63 (m/z 4476.5). The structure on top shows the bonds cleaved in the bioconjugate to form the N-
terminal an and C-terminal yn fragment series and the single fragments c17 and x17...... 79
5. 13 MALDI-MS2 spectra of the [M + Na]+ ions from (a) GpAKQ18 (m/z 1798.1) and (b) AKQ18- PEG64 (m/z 4615.4). The fragments observed confirm the peptide sequence and the C-terminal conjugation of the PEG chain...... 80
5. 14 (a) 2D ESI-IM-MS plots (m/z vs. drift time in IM cell) and (b) extracted mass spectra of charge state 4+ of bioconjugates AQK18-PEG (top) and GpAQK19-PEG (bottom)...... 81
5. 15 (a) ESI-IM-MS mobilogram of quadruply protonated AQK18-PEG71 (m/z 1202.45). (b) 2D ESI- IM-MS2 plot of [M + 4H]4+, acquired by IM separation followed by CAD at a collision energy of 80 eV. Two bands are observed for m/z 1202.45, corresponding to the random coil and α-helical conformer of AQK18-PEG71. Note that barely any fragments are formed from the α-helical conformer, consistent with a higher stability and dissociation after collapse to the random coil structure. (c) ESI-IM-MS2 spectrum extracted from the ions drifting at 9.30 ms (random coil
conformer). None of the N-terminal fragments (an and bn) but all of the C-terminal fragments (yn) contain the PEG chain, validating that the polymer is conjugated at the C-terminus...... 83
5. 16 [AQK18 + 2H]2+ conformers; (a) random coil structure protonated at lysines K1 and K18 and (b) α-helical structure protonated at glutamine Q17 and lysine K18...... 87
5. 17 Random coil (top) and α-helical (bottom) structures of doubly protonated AQK18. The proton attachment sites are given under the corresponding structures. See Table 5.2 for the sequence of AQK18...... 88
5. 18 Isomeric structures of protonated-sodiated AQK18 with partially helical or bent helical structure (top) or α-helical structure (bottom). The proton and sodium ion attachment sites are given under the corresponding structures. See Table 5.2 for the sequence of AQK18...... 89
5. 19 [AQK18-PEG70 + 4H]4+ conformers; (a) random coil structure protonated at lysines K1 and K18 and (b) α-helical structure protonated at glutamine Q17 and lysine K18. In both structures, two additional protons are attached on PEG………………………………………………….…………..89
5. 20 [AQK18-PEG70 + 4H]4+ tautomer with α-helical structure, having three protons on the PEG chain and one proton at lysine K18……………………………………………………………………90
xvi 5. 21 [AQK18-PEG70 + 3H + Na]4+ conformers; (a) random coil structure protonated at lysines K1 and K18 and (b) α-helical structure protonated at glutamine Q17 and lysine K18. In both structures, an additional proton and a sodium ion are attached on PEG…………………….…..90
6. 1 Synthesis of Branched poly(acryloyl-1,2:3,4-di-O-isopropylidene-α-D-galactose) by RAFT polymerization mechanism using galactose acrylate monomer and chain transfer agent in the present of AIBN initiator…………………………………………………….………………………….94
6. 2 MALDI mass spectrum of the 121 glycopolymer studied. Three ion distributions of oligomers
+ with the [Rn+EGs+Na] composition are observed; R is the galactose acrylate monomer and RAFT CTA is the branching unit. The observed oligomers have end groups (EGs) of 101 Da ($), 134 Da (%) and 204 Da (&). Figure 6.3 shows the structures of these oligomer…………………95
6. 3 MALDI-MS2 mass spectra of sodiated [R6+EGs+Na]+1 6-mers (a) EGs= 101 Da (m/z 2008.8), (b) EGs= 134 Da (m/z 2041.9) and (c) EGs= 204 (m/z 2111.8); the numbers on top of the peaks give the monoisotopic m/z ratio (in black) and the mass of the neutral loss(es) in Da (in color). The * sign indicates internal fragments Jna. The insets show the structures and end groups that agree well with these spectra………………………………...………………………..101
6. 4 Proposed structures of the oligomers observed in the MALDI and ESI spectra. All have the
+n composition [Rn+EGs+nNa] , where R and EGs designate the repeat unit of the protected
galactose acrylate (C15H22O7, 314.13 Da) and the corresponding end groups, respectively. .. 102
6. 5 2-D IM-MS plot (m/z vs. drift time) of the protected galactose acrylate dissolved in MeOH: THF (7:3, v/v %) at 0.01 mg/mL. b) Mass spectra extracted from the IM regions of (a) singly charged, c) doubly charged d) triply charged ions in the 2-D diagram……………………………………..109
6. 6 Q/ToF-ESI-MS2 spectra of sodiated singly charged glycopolymer ions with (a) 101-Da end groups (m/z 1380.48), (b) 134-Da end groups (m/z 1413.55) and (c) 204-Da end groups (m/z 1483.50); the numbers on top of the peaks give the monoisotopic m/z ratio (in black) and the mass of the neutral loss(es) in Da (in color)...... 111
6. 7 (a) QIT-ESI-MS2 mass spectrum of sodiated 121 glycopolymer with 134-Da end groups (m/z 1413.5); (b) MS3 mass spectrum of the fragment at m/z 1355.6 formed by acetone loss from 1413.5; (c) MS4 mass spectrum of the fragment at m/z 1297.5 formed by acetone loss from m/z 1355.5. The numbers on top of the peaks give the monoisotopic m/z ratio (in black) and the mass of the neutral loss(es) in Da (in color); −58, 28 and −242 indicate losses of acetone, CO and galactose pendant, respectively, see scheme 6.1 for more details………………………..113
xvii 6. 8 (a) QIT-ESI-MS2 mass spectrum of sodiated 121 glycopolymer 4-mer with 204-Da end groups (m/z 1483.6); (b) MS3 mass spectrum of m/z 1425.6, formed by acetone loss from 1483.6; (c) MS4 mass spectrum of m/z 1367.4, formed by acetone loss from m/z 1425.6. The numbers on top of the peaks give the monoisotopic m/z ratio (in black) and the mass of the neutral loss(es) in Da (in red)...... 115
6. 9 (a) QIT-ESI-MS2 mass spectrum of the sodiated 4-mer from glycopolymer 121 with 314-Da end groups (m/z 1593.7); (b) MS3 mass spectrum of m/z 1535.6, formed by acetone loss from 1593.7; (c) MS4 mass spectrum of m/z 1477.6, formed by acetone loss from m/z 1535.6. The numbers on top of the peaks give the monoisotopic m/z ratio (in black) and the mass of the neutral loss(es) in Da (in red); for example, −58 and −242 indicate losses of acetone and the galactose pendant, respectively. For more details see Scheme 6.1……116
6. 10 Plot of corrected drift times (arrival times) against corrected published cross- sections for the +1 and +2 ions formed by ESI of polyalanine. Drift times were measured at a wave velocity of 350 m/s and a wave height of 11 V...... 117
6. 11 Plot of experimental collision cross-section vs. m/z for sodiated 121 glycopolymer oligomers with 101 Da, 134 Da and 204 Da end groups (a) +3, (b) +2. ……………………123
7. 1 (a) MALDI spectrum of acetylated trehalose glycopolymer; all ions are sodiated species with
+ the composition [Rn + EGs + Na] , where R is the repeat unit (C33H40O17, 708.226 Da). (b) Expanded view of the highlighted m/z region of the trehalose glycopolymer spectrum. (c) Proposed structures of the polymer end groups observed. …………………………………132
7. 2 Comparison of the MALDI mass spectra of (a) insulin-trehalose glycopolymer conjugate after treatment by FA and before disulfide reduction and (b) native insulin using DHAP matrix………………………………………………………………………………………………133
7. 3 Comparison of the MALDI mass spectra of chain B of native insulin (top) and conjugated insulin (bottom). The spectrum in the top shows the protonated chain B while the spectrum in the bottom shows the protonated ion in addition to another ion that is 106 Da higher than chain B corresponding to the benzaldehyde linker ...... 134
7. 4 ESI-MS spectrum of the trehalose glycopolymer-insulin conjugate; the expanded area shows the cleaved glycopolymer with the remaining end of the RAFT chain transfer agent.
-1 The sample was dissolved in H2O: ACN at 0.05 mg mL + 10% formic acid (v/v %)...... 136
7. 5 ESI-MS spectrum of the trehalose glycopolymer-insulin conjugate; the expanded area shows the cleaved glycopolymer with the remaining end of the RAFT chain transfer agent.
-1 The sample was dissolved in H2O: ACN at 0.01 mg mL + 0.1% formic acid (v/v %)...... 137
xviii
7. 6 (a) ESI-MS spectrum of the trehalose glycopolymer-insulin conjugate after acid treatment and disulfide reduction. (b) Expanded area of the m/z region of +3 chain A (m/z 795.0513) which shows one site modification. (c) Expanded area of the m/z region of +5 chain B (m/z
686.575) which shows also one site modification. The sample was dissolved in H2O: ACN at 0.05 mg/mL + 10 mM DTT +10% formic acid (v/v %)…………………………………………138 7. 7 ESI-MS2 spectrum of insulin-trehalose glycopolymer conjugate (top) and native insulin
(bottom). Fragmentation of chain B in charge state +5 from the conjugate (m/z 707.64) and insulin (m/z 686.45) mainly gives rise to bn and yn fragments...... 139
xix
LIST OF SCHEMES
Scheme Page
2.1 Two different fragmentation mechanisms…………………………………………………………….27
4.1 Synthesis of poly(propylene maleate) (PPM) via ring-opening polymerization of maleic anhyd- ride and propylene oxide and base-catalyzed post-polymerization isomerization of PPM to poly(propylene fumarate) (PPF). ………………………………………………………………………43
6.1 Fragmentation pathways of glycopolymer "121". Charge-remote H rearrangements in sodiated
oligomers lead to (a) 2x(CH3)2CO (acetone) losses from the galactose ring followed by CO loss to form a five-membered ring lactone. (b) Cross ring fragmentation leading to loss of 200 Da and followed by (c) CO elimination (overall loss of 228Da). (d) Expulsion of the galactose pendant (242 Da), resulting in a fragment with acid end group………..…………………………104
6.2 Backbiting in the acrylic radical ions emerging after random homolytic C–C bond cleavages in the Na+-cationized galactose acrylate glycopolymer chains. R abbreviates the galactose
• pendent (C13H19O7). α and ω designate the end groups; bn are the acrylic radical ions
• containing the α and zn the acrylic radical ions containing the ω end group. Backbiting gives
• • rise to terminal an and yn fragments from bn and zn , respectively, as well as to the internal
fragments Jn (m/z 323, 637, 951) and Jna (m/z 521, 835, 1149); the latter are generated after the
loss of 2 acetone moieties from Jn. The internal fragment ions Kn (m/z 337, 651, 965) are
observed as well. The terminal fragments (an and yn) are below noise level, suggesting that they dissociate consecutively via the same mechanism………………………………………………..107
7.1 (a) Synthesis of trehalose glycopolymer by RAFT polymerization using AIBN as initiator and a trithiocarbonate chain transfer agent (CTA). (b) Acetylation of the glycopolymer by using acetic anhydride and pyridine catalyst…………………………….…………………………………………126 7. 2 Conjugation procedure of trehalose glycopolymer to insulin by reductive amination…………..128
xx
ACRONYMNS AND ABBREVIATIONS
CAD Collisionally Activated Dissociation CCS Collision cross section
CE Capillary Electrophoresis
CE Collision Energy
CHCA α‐cyano‐4‐hydroxycinnamic acid
CID Collision‐Induced Dissociation
DCTB Trans‐2‐(3‐(4‐tert‐butylphenyl)‐2‐methyl‐2‐
propenyliedene)malononitrile
DHB 2,5‐dihydroxybenzoic acid
DIOS Desorption/Ionization on Silicon
DIT Dithranol
ECD Electron Capture Dissociation
ESI Electrospray Ionization
ETD Electron Transfer Dissociation
FT‐ICR Fourier Transform Ion Cyclotron Resonance
FWHM Full Width at Half Maximum
xxi
GC Gas Chromatography
GPC Gel Permeation Chromatography
HPLC High Performance Liquid Chromatography
IM‐MS Ion Mobility Mass Spectrometry
LC Liquid Chromatography
LIT Linear Ion Trap m/z Mass‐to‐Charge Ratio
MALDI Matrix‐Assisted Laser Desorption Ionization
MCP Multichannel plates
MS Mass spectrometry
MS/MS Tandem mass spectrometry
xxii
CHAPTER I
INTRODUCTION
Mass spectrometry (MS) is a powerful analytical method used to analyze and characterize a wide range of molecules, including synthetic polymers1,2 and biological samples such as proteins, peptides, saccharides, and bioconjugates.3-5 Mass spectrometry is a fast and sensitive technique and requires very little amount of the sample under investigation. Unlike other analytical methods, MS does not require high purity and relatively large amounts of sample. However, it has some limitations, especially when characterizing complex mixtures such as isomeric species, isobaric molecules, biomolecules, synthetic materials and bioconjugates. Such compounds might be challenging or even impossible to analyze by traditional, single-stage MS approaches due to overlapping signals at the same or very similar mass. To overcome this limitation, MS analysis can be combined with tandem mass spectrometry analysis
(MSn) or coupled with a separation methods such as chromatography or ion mobility
(IM)6 techniques. Size exclusion chromatography (SEC)7, gas chromatography (GC)8, liquid chromatography (LC)9 and capillary electrophoresis (CE)10 are chromatographic methods that have been successfully coupled to MS to improve the selectivity and sensitivity of this technique. The resulting hyphenated mass spectrometry techniques can be used to identify the chemical structures of complex molecules and determine architectures and sequences.
The first step in mass spectrometry analysis is converting the analyte to gas- phase ions, and the next step is to separate the ions with the mass analyzer based on
1 their mass to charge ratio (m/z) before they reach the detector.11 Nowadays, different ionization methods are used to characterize a variety of molecules ranging from small organic molecules to large biological samples. Electrospray ionization (ESI) and matrix assisted laser desorption /ionization (MALDI) are the most widely used soft ionization methods that form intact gas-phase ions from the analyte. ESI and MALDI mass spectrometry can be performed in single stage MS or multistage MSn mode. Single stage MS provides information about the composition and molecular mass of the sample, while multistage MSn offers information about primary structure and architecture.11-13
In ESI, the sample is normally dissolved in a volatile solvent, and a syringe is used to introduce the solution into the mass spectrometer. In the ion source, a heated desolvation gas is used to remove the solvent and assist the sample deposition. Highly charged droplets of the analyte experience charge repulsion until they reach the
Rayleigh limit where smaller droplets are formed. This process continues until individual ions remain with one or more charges. Analytes of high molecular weight such as proteins and biomolecules can form multiply charged ions in the ESI source which overcomes the mass limitation of the instrument.12,14
In MALDI, a small organic compound called matrix is used to absorb the laser power, prevent analyte degradation and assist the ionization process. A mixture of matrix and analyte solution is deposited on a sample plate and is allowed to dry before a laser beam irradiates it and ionizes the analyte.13
In tandem mass spectrometry experiments (MS2), the precursor ion is selected and accelerated to high kinetic energy to undergo fragmenting collisions with a neutral
2 gas like N2, He or Ar. This method is known as collisionally activated dissociation (CAD) or collision-induced dissociation (CID). The fragments produced are then analyzed using a second mass analyzer which separates them according to their mass to charge ratio before they are eventually transmitted to the detector. CAD is the commonly used method to produce fragments that can help to determine end groups and architectures of polymers and conjugates.11,12
Ion mobility spectrometry (IMS) is a characterization technique that utilizes a weak electric field to separate gas phase ions based on their mobility. Ion mobility devices can be attached to mass spectrometers to separate and distinguish isomers, isobars, complex mixtures and conjugates. Ions are separated in the ion mobility cell based on their size, shape and charge.15 In this method, the ions that enter the ion mobility cell are pushed by an electric field while colliding with a buffer gas. Small and more compact species drift faster through the buffer gas than the large elongated ones.
Ion mobility data can be used to calculate the collision cross section (CCS), which is a value characteristic for the analyte. Comparing an experimental CCS to the theoretical one for the same ion can help to determine the molecular architecture.16-18
Combining IM with mass analysis (MS) and tandem mass spectrometry (MS2) creates a variety of hyphenated techniques which have been termed multidimentional mass spectrometry. This dissertation focuses on applications of multidimensional mass spectrometry to the analysis of biodegradable isomeric polyesters, glycopolymers, and peptide-based bioconjugates. Chapter II will discuss the basic principles of mass spectrometry, including ionization techniques, mass analyzers, tandem mass spectrometry, and ion mobility. Chapter III will describe the materials and instruments
3 used to accomplish the aim of this dissertation. The following four chapters discuss the research projects completed in this dissertation, which are briefly introduced below.
Finally, Chapter VIII summarizes the conclusions drawn from this dissertation.
Chapter IV discusses the analysis of isomeric biodegradable polyesters, synthesized by Matthew L. Becker et al. (University of Akron). This study combined
MALDI and ESI with time-of-flight (ToF) mass analysis and tandem mass spectrometry
(MS2) fragmentation to elucidate the composition, end groups, and chain sequence of poly(propylene maleate) (PPM) and poly(propylene fumarate) (PPF) copolymers.
Furthermore, ion mobility mass spectrometry (IM-MS) was used to differentiate the isomeric PPM and PPF copolyesters and to probe the extent and efficiency of PPM to
PPF isomerization.19
Chapter V discusses the analysis of the alanine-rich peptides AQK18 and
GpAQK18 and their poly (ethylene glycol)-conjugated (PEGylated) forms AQK18-PEG and GpAQK18-PEG in which the polymer was attached at the C-terminus through an amide bond. The samples were synthesized by the Xinqiao Jia research group
(University of Delaware). ESI and MALDI, coupled with MS2 fragmentation and shape- sensitive separation via IM-MS, were applied to gain more structural insight into the sequence and conformation of alanine-rich peptides and their conjugates. IM-MS data revealed the existence of two conformers for both the peptides and the conjugates.
These were assigned to a fast-drifting random coil and a more slowly drifting helical structure. The collision cross-sections of the random coil and helical conformers of the polypeptides and their PEG conjugates were simulated for comparison with the
4 experimental values in order to confirm the identity of the observed architectures and understand the stabilizing effect of the polymer chain.
Chapter VI discusses the characterization of a glycopolymer with one controlling branch unit. The glycopolymer was synthesized by reversible addition-fragmentation chain transfer (RAFT) copolymerization of galactose acrylate and a polymerizable chain transfer agent (branching unit) by the Andrea M. Kasko group (UCLA). MALDI-MS, ESI-
MS, multistage tandem mass (MSn) and IM-MS were employed in this work to gain useful information about the glycopolymer composition, possible end groups and architecture.
Chapter VII discusses the analysis of a trehalose glycopolymer and its insulin conjugate. Both were synthesized by Heather Maynard and coworkers (UCLA). The goal of this study was to provide full characterization and detailed information about the chemical composition of the trehalose glycopolymer and its insulin conjugate and to determine the degree and sites of conjugation by using multidimensional MS, including
MALDI-MS, ESI-MS, and MS2.20
5
CHAPTER II
MASS SPECTROMETRY BACKGROUND
2.1. Mass Spectrometry
Mass spectrometry is a powerful analytical tool for identifying a wide range of molecules such as synthetic polymer,1,2 biomolecules,3-5 and bioconjugate materials.4, 21
It is widely used because of its high sensitivity, selectivity and speed of analysis. This method has been applied often in many fields, including pharmaceutical studies, polymer sciences, chemical engineering and environmental studies.
For any sample to be detected in mass spectrometry instrument, ions must be first generated in the positive or the negative ionization mode. This is can be done by creating gas phase ions of the analytes of interest which then can be separated by their mass-to-charge ratios (m/z). The basic components of any mass spectrometer are inlet system, ionization source, mass analyzer and ion detector. The inlet system is the first part used for introducing the samples into the source. The ion source is used to convert the analytes to gas phase ions; this device can be held at ambient pressure or under vacuum (10-5 - 10-10 torr). After that, the mass analyzers separate the ions according to their mass-to-charge (m/z) ratio. Then, the detector detects the ion current from each ion generated from the analyte and converts them into electric signals. The mass analyzers and detectors are held under vacuum to prevent collisions of the ions with other gaseous molecules. Finally, the data system records the generated signal and produces the data in the form of mass spectra (see Figure 2.1).11,12
6
Figure 2. 1. General components of a mass spectrometer.
Direct insertion, direct infusion and chromatography are the common methods used to interduce the analyte into the ion source. MALDI is an example of direct insertion method in which a probe or a target plate is placed in the instrument,13 while
ESI is an example of direct infusion in which the sample solution is injected into the instrument by a syringe.22 Finally, chromatography methods such as liquid chromatography or gas chromatograph can also be used to introduce samples into a mass spectrometer.14
Ionization sources, mass analyzers and detectors, which are used in this work to complete the dissertation, will be discussed in more details in this chapter.
2.1.1. Ionization methods
Ion generation is the first and the most important step in mass spectrometric analysis. The successful analysis always depends on the successful ionization of the analyte of interest. This step can be achieved by different methods such as ejection or the addition of an electron and the addition or subtraction of an ion. The appropriate ionization method should be selected based on the analyte of interest.11,23 The ionization methods can be classified as hard and soft. In the hard ionization, the analyte
7 is often fragmented; and the intact analyte ion are very hard to be detected in this method. Electron ionization (EI), chemical ionization (CI), fast atom bombardment
(FAB), and field desorption (FD) are examples of hard ionization methods. Although fragmentation can help determine the analyte structure, sometimes it complicates the spectrum and makes it difficult to interpret. To overcome this problem, soft ionization methods have been developed.11 The intact molecular ion is observed with minimal or no fragmentation in soft ionization methods. Methods such as electrospray ionization
(ESI) and matrix-assisted laser desorption/ionization (MALDI) represent examples of soft ionization sources. These two sources are extensively used to characterize and identify a wide range of molecules, including polymers and large biomolecules and synthetic bioconjugates.12 In this chapter, the ionization methods (MALDI and ESI), which were used to complete this dissertation, will be discussed in more details.
2.2.1.1 Electrospray ionization (ESI)
Electrospray ionization (ESI) is a soft ionization technique that was interfaced for the first time with MS in the late 1960s by Dole et al.13 After that, John Fenn successfully developed the modern ESI in the late 1980s,14 when he reported the first application of this method to characterize large biological molecules that mainly produced multiply charged ions. Biological samples such as peptides, proteins, carbohydrates, small oligonucleotides and lipids have been successfully analyzed and identified by ESI over the last decades.22 The sample is introduced in a solution into the ion source by means of either a syringe pump or the exit of an LC column. The sample solution passes through a steel capillary to the spray chamber which is filled by the nebulizing gas (normally N2) under a high electric field ranging from 2 to 6 kV. The high
8 voltage and the nebulizing gas induce a charge accumulation at the liquid surface at the end of the capillary. This step produces highly charged droplets of the sample solution.
The repulsion forces of the accumulated charges at the tip of the capillary become higher than the surface tension which causes the droplet to take the shape of a Taylor cone and break into smaller droplets Figure 2.2. Once the repulsive forces between charges become equal to the droplet surface tension, droplet decomposition take place as described by the Rayleigh equation (Equation 2.1), where q is the charge, ε0 the permittivity environment, γ the surface tension, and D the diameter of a spherical droplet.11,12,26
2 2 3 q =8π ε0γD Equation 2. 1
Figure 2. 2. (a) Droplet productions in the electrospray interface. (b) Droplet decomposing in an electrospray source according to the Rayleigh limit. Reproduced with permission from reference 11and 26.
9
The two main mechanisms that explain ion formation from the droplets are the charge-residue model and the ion-desorption model. In the charge-residue model, as the solvent evaporates, the droplets become smaller and the charge remains the same.
This process causes an increase in charge per volume unit and electrostatic tension at the surface of the droplets. At the Rayleigh limit, the surface tension force becomes close to the charge repulsion forces; the droplets deform into Taylor cones and eventually break down into smaller droplets. This process is called Coulombic explosion
(Figure 2.3). Replicating this process may eventually lead to single ions (Charge
Residue Model). As the droplet size decreases, the charge density and the electric field increases at the droplet surface. As the electric field at the droplet surface becomes large, it can cause the direct desorption of single ions out of the droplet (Ion Evaporation
Model).26-28
Figure 2. 3. Schematic of the Coulombic explosion of a charged droplet (adapted with permission from reference 11.
10
2.1.1.2 Matrix-Assisted Laser Desorption/Ionization (MALDI)
MALDI is another soft ionization method which allows the detection of intact large and labile molecules by mass spectrometry. MALDI was first introduced in 1988 by
Karas and Hillenkamp13,29 ; then, in the same year, Tanaka invented a soft laser desorption technique which is very similar to MALDI.30 MALDI is considered as the most common technique used to analyze a wide range of large and non‐volatile compounds such as proteins, oligonucleotides, large inorganic compounds and synthetic polymers.
This technique is more capable to salts, buffers, and other contaminants than ESI, where they may suppress signals and reduce the analysis efficiency. This method typically generates only singly‐charged ions except for high molecular weight proteins/ peptides where higher charge states can be seen.31
The MALDI preparation process involves mixing analyte with excess matrix, typically in the molar ratio of 1:1000-10000 (analyte: matrix). The matrix is a small organic compound that strongly absorbs the laser wavelength like a UV laser (e.g.,
Nd:YAG laser). Matrices are usually aromatic molecules with acidic COOH groups that easily donate a proton. The MALDI mechanism is shown in Figure 2.4. The first step of the MALDI procedure is desorption. The matrix is desorbed from the surface, and it facilitates the movement of the sample into the gas phase. The second step is ionization. The laser radiation causes the ionization of the matrix in the gas phase, which enhances the ionization of the analyte by transferring charges from the matrix ions to the analyte; biomolecular analytes are frequently ionized by either protonation or deprotonation. Analytes such as polymers do not protonate easily; therefore, a salt
11 solution is added to the analyte/matrix mixture (e.g., Na+ or Li+ salts for O‐containing polymers and Ag+ for double bonds or aromatic species such as polystyrene) prior to
MALDI analysis.
Figure 2. 4. Illustration of the MALDI process (adapted with permission from reference
11).
Sample preparation is the most important step in MALDI process, and any technique should give a homogeneous crystal formation with the sample of interest.12
There are different protocols for mixing the matrix and the analyte together, including the dried-droplet method, the sandwich matrix method and the solvent-free sample preparation. The dried-droplet method is the most commonly used sample preparation technique. The matrix, sample, and salt (if it is needed) are mixed together, and a small amount of the mixture is applied onto the steel MALDI plate and allowed to dry before it
12 is inserted in MALDI. In the sandwich method, a mixture of matrix and salt is applied on the bottom and then on a layer of the sample, followed by another layer of matrix and salt mixture. Insoluble analytes can be analyzed with MALDI using the solvent‐free approach in which both the sample matrix and the salt, if needed, are mixed; then, a small portion of the mixture is deposited on the target plate. Eventually, the matrix and analyte deposition on a target is followed by inserting the target into the vacuum after the solvents evaporate.
Table 2. 1 Common MALDI matrices (reproduced with permission from reference 11).
13
2.1.2. Mass Analyzers
The mass analyzer is the second part of any mass spectrometry instrument. Its purpose is to separate the gas‐phase ions generated in an ionization source according to their m/z. All the mass analyzers operate under a vacuum system to prevent ions collisions that can cause neutralize these ions before the detection step. Different types of analyzers have been developed such as quadrupole (Q), quadrupole ion trap (QIT), time of flight (ToF), orbitrap and Fourier transform ion cyclotron resonance (FTICR).
Each type has its advantages and limitations based on their characteristic features. The three main characteristics of any mass analyzer are mass range limit, resolution and mass accuracy; based on these parameters, the best option for the analysis can be selected.
The mass range determines the mass analyzer limitation to m/z values. For example, the mass limit of ToF linear mass analyzer is more than 1000 k Da, whereas it is around 4 k Da in the Quadrable mass analyzers. The mass resolution is an important parameter since it affects the mass analyzer accuracy directly. It is also known as resolving power reflecting the ability of a mass analyzer to distinguish between two ions with a small m/z difference. The definition of resolution R is shown in Equation 2.2 11
푚 푅 = Equation 2.2 ∆푚
Where ∆푚 is the mass difference of two peaks with masses at 푚 and 푚 + ∆푚. R can also be determined with an isolated peak; where ∆푚 is defined as the full width at half maximum (FWHM) of the peak with mass 푚. Figure 2.5 below illustrates the concept of resolving power, showing peaks m1 and m2 resolved at 10% and 80% valley.
14
Figure 2. 5. Diagram illustrating the concept of resolution. (reproduced with permission from reference 11).
The mass accuracy is related to the mass analyzer resolution as low-resolution mass analyzers cannot provide high accuracy. The mass accuracy measures the difference between experimental m/z values (m measured) and theoretical m/z values
(m theoretical), and it is often expressed in parts per million (ppm) as indication of m/z values accuracy provided by the mass analyzer.
Only time-of-flight (ToF), quadrupole (Q), and ion trap (IT) mass analyzers were used in this dissertation and will be covered in this chapter.
15
2.1.2.1 Time-of-Flight (ToF) Mass Analyzers
The concept of time-of-flight (ToF) is based on the physics of ion motion which was described in the 1940s by Stephens.32 ToF is usually coupled with laser desorption ionization methods like MALDI due to the pulsed nature of that ionization source. There are two types of ToF analyzers: linear and reflectron time-of-flight analyzers. Each of these will be discussed below.
The linear ToF mass spectrometer was the first commercially available ToF analyzer in 1955.33 The ToF mass analyzer becomes more popular after it was coupled with matrix-assisted laser desorption/ionization.34 Nowadays, the ToF analyzer has become one of the most valuable analyzers, and the development of MALDI-ToF has open the gate for MS applications to characterize samples such as biomolecules and synthetic polymers.5,35
After the ions have been formed in the MALDI source, they are accelerated and travel at constant velocity v inside a field-free tube that has a fixed length L until they arrive at the detector. The arrival time at the detector t is measured and is converted to m/z (see Figure 2.6).
The ToF analyzers separate ions based on their velocities when these ions drift in flight tube (field free region) as shown in Figure 2.6. The ions are accelerated with a potential (Vs) in the ionization source before they enter the flight tube. The kinetic energy Ek gained from the acceleration of an ion with total charge q and mass m can be described in Equation 2.3 and 2.4:
16
Figure 2. 6. Schematic of a linear ToF tuned to detect positive ions produced by MALDI
11.
1 퐸푘 = ze 푉푠= m휐2= 퐸푒푙 Equation 2.3 2
퐿 푡= Equation 2.4 υ
2푧푒푉푠 휐 = √ Equation2.5 푚
푚 퐿√ 푡 = 푧 Equation 2.6 √2푒푉푠
Where 휐 is the velocity of the ions leaving the source, m is the mass of the ion and ze is the charge of the ion.
17
The equation above shows that the large ions will travel more slowly and spend longer time in the flight tube than the small ions. However, linear ToF mass analyzers suffer from poor resolution as some ions with the same m/z value may have different initial kinetic energies and as a result different flight times, thus leading to poor resolution in this instrument. This problem was overcome by adding a reflectron device, which is an electrostatic mirror that reflects the ions and sends them back through the flight tube to increase their path length inside the tube (Figure 2.7).
Figure 2. 7. Schematic of a reflectron ToF mass analyzer, (adapted with permission from reference 11).
The ions with higher kinetic energies will enter deeper into the reflectron region due to their higher velocities. On the other hand, those with lower kinetic energies
(lower velocities) will not be able to travel very far. As a result, all ions of the same mass
18 to charge ratio will reach the detector at the same time, a movement that improves the overall resolution.
2.1.2.2 Quadrupole Mass Analyzers
The quadrupole was first introduced by Paul and Steinwegan in 1953. It consists of four parallel cylindrical rods (see Figure 2.8). The quadruple separates ions using electric fields where two opposite rods have the same applied potential, while the other two opposite rods have the same potential with opposite polarity. The rods change their polarity as a function of time.
Positively charged ions are attracted towards the negatively charged rod and repelled from the positive rods. The rods should change polarity before the ion touches a rod and becomes neutral. If the rod polarities change before the ion hits the rod, the ion will move towards the new negative rod and continue in a zigzag motion until it exits the quadrupole (see Figure 2.8).
Figure 2. 8. Quadrupole mass spectrometer made up of a source, focusing lenses, quadruple rods and a detector. Reproduced with permission from reference 11.
19
Figure 2.8 shows the distance between two opposite rods 2r0; when specific ions enter the space between the rods, they will be influenced by the electric field:
+ Φ0 = + (U – V cos ωt) Equation 2. 7
- Φ0 = - (U – V cos ωt) Equation 2. 8
Where Φ0 is the overall potential applied to the rods; U is the DC potential and V cos ωt is an RF potential with the amplitude V and the angular frequency ω.
By selecting proper U and V values, only ions with a specific m/z value will be allowed to pass through and get detected. The trapping parameters au and qu in
Equations 2.9 and 2.10, respectively, which are determined by the electric field applied to the rods, describe the ions’ trajectory in the x and y directions at any given time.11,12
Figure 2. 9. Stability areas as a function of U and V for three ions with different masses: m1, m2 and m3. Reproduced with permission from reference 11.
20
au = ax = -ay = 8zeU / mω2r0 Equation 2. 9
qu = qx = -qy = 4zeV/ mω2r0 Equation 2.10
Figure 2.9 shows the stability areas (areas of stable trajectories) of different masses as function of U and V. The stability areas determine the effective mass range, which can be obtained by plotting au vs. qu (or U vs. V). U and V are scanned at a constant ratio to bring m1, m2 and m3 in a stable trajectory and transmit them through the quadrupole. By adjusting the slope of the scan line, the resolution can be varied.
Resolution increases with increasing slope of the scan line.
2.1.2.3 Ion Trap Mass Analyzers
The ion trap analyzer was invented by Paul and Steinwedel in 1960 before it was modified by Stafford et al.36 It is an ion-storing device that uses the same principle as quadrupole analyzers. This mass analyzer can be classified as 2D ion trap or 3D ion trap. The 2D ion traps or linear ion traps consist of four rods ending in lenses that reflect the ions backwards and forwards. The 3D device consists of three electrodes; one ring electrode is located between two end cap electrodes (see Figure 2.10). Only the 3D ion trap will be described here as it was used in this work. This mass analyzer utilizes oscillating electric fields that allow different masses to be trapped inside the electrodes and stored in space by controlling their motions; at the same time, it can eject ions based on their masses.
21
Figure 2. 10. A 3D ion trap mass analyzer. Reproduced with permission from reference
11.
A radio frequency RF potential, 훷∘ = U-V cos 훚t, is applied to the ring electrode, while the two end caps are grounded. Due to the potential difference and the shape of the three electrodes, a three-dimensional field is generated such that it forces the ions to move in stable trajectories towards the trap center where they are stored and trapped.
Ions inside the trap will carry the same charge and thus repel each other. Their trajectories would destabilize, leading to their ejection from the trap. A buffer gas such as helium is introduced in the trap to avoid ion losses by colliding with the ions to remove excess energy and pushing them back toward the center of the trap.
22
Ion trajectories in trap mass analyzers can be described by the Mathieu equation and trapping parameter 푞푧 :
8푧푒푉 푞푧 = 2 2 2 Equation 2.11 푚(푟0 +2푧0 ).휔
Where ze is the charge of the ions, m is the ion mass, V is the amplitude of the
RF potential, r0 is the inner radius of the ring electrode, z0 is the distance from the
2 center of the trap to an end cap (usually r0 = 2 z0 ), and 휔 is the frequency. To obtain stable trajectories, the ions should never reach or exceed the r0 and z0 coordinates.
11 Generally, ions remain in stable trajectory inside the trap, if qz <0.908.
2.1.3 Detectors
The function of the detector is to detect the ions passing through the mass analyzer and convert the ion current into an electrical signal that is proportional to the ion intensity. Usually, the number of ions coming out from the mass analyzer is small; hence, amplification is necessary to obtain observable signals.
Two detectors will be discussed in this section: the photo multiplier device (Daly detector) and the microchannel plate (MCP) detector.11
2.1.3.1 Photo multiplier (Daly) Detectors
The Daly detector is an electro-optical ion detector (EOID), which is a combination of ion and photon detection device. The Daly detector converts ions to electrons and then to photons (see Figure 2.11). When the ions from the mass analyzer strike a dynode, the dynode converts the input ion into electrons. Positive ions strike the
23 negatively charged dynode, while negative ions hit positive dynode. The secondary electrons generated in this process are accelerated towards a phosphorescent screen, which converts the electrons into photons. The photons are then sent to the photomultiplier which produces an electric current. Eventually, the electric current is amplified and sent to the computer. This type of detector amplifies the ion intensity by
104 to 105.
Figure 2. 11. Schematic diagram of a Daly detector. Reproduced with permission from reference 11.
2.1.3.2 Microchannel Plate (MCP) Detectors
The microchannel plate (MCP) detector is the most widely used detector in mass spectrometry. MCP is a type of electron multiplier detector that consists of a plate with parallel cylindrical channels. Each channel is a few millimeters long and few micrometers in diameter. The channels are coated with a semiconductor material which amplifies the signal intensity by emitting electrons when struck by ion beams. The electron cascade generated is then collected by a metal anode at the output side and
24 the current is measured (see Figure 2.12). The multiplication factor can reach 108 by using several parallel plates.
Figure 2. 12. (A) Cross-section of a microchannel plate and (B) electron multiplication within a channel. Reproduced with permission from reference 11.
2.2. Tandem Mass Spectrometry
Tandem mass spectrometry methods (MS2) can be employed either with ESI or with MALDI. Tandem mass spectrometry and multistage mass spectrometry refer to multi-dimensional methods that involve two or multiple mass analysis steps, (MS2 or
MSn, respectively).37
MS2 analysis is performed to gain more structural information about the analyte under investigation. In this method, a specific m/z (precursor or parent ion) is isolated and then fragmented to yield indicative product ions and neutral fragments.
In MS2, two mass analysis events take place,38 which are separated either in- space or in-time (Figure 2.13). In-space tandem mass spectrometry instruments utilize two physically distinct mass analyzers; ToF/ToF, Q/ToF and triple quadrupole instruments are examples of this type. In in-space MS2 experiments, the first mass
25 analyzer is used to select the precursor ion, which is guided to a collision cell and fragmented into product ions by collisions with a neutral gas like argon or helium. The collision cell is located between the two analyzers. Then, the second mass analyzer separates the fragment ions based on their mass-to-charge ratio (m/z).11
For in-time tandem mass spectrometry, the fragmentation procedure can be achieved through time separation events, which take place sequentially in a single ion storage device; ion trap analyzers are examples of this type. In this method, the precursor ion is isolated and stored in the storage device while other ions are ejected from the ion trap.11 During a certain period, the precursor ion is allowed to collide with a neutral gas and fragmented into product ions. These steps can be repeated to form fragments of the fragments (MSn, where n is the number of generations of ions being analyzed).
26
Figure 2. 13. Schematic of in-space and in-time tandem mass spectrometry (adapted with permission from reference 11).
Two different types of fragments can be formed based on how the electrons move during bond cleavage. Scheme 2.1 illustrates bond cleavages for a positive precursor ion.
Scheme 2.1. Two different fragmentation mechanisms.
A homolytic bond cleavage giving rise to radicals is illustrated in the first example, while a heterolytic bond cleavage is shown in the second example. The fragments that retain the charge represent the thermodynamically most stable product, which seen in the fragmentation spectrum.
27
2.2.1 Collisionally Activated Dissociation
Collisionally Activated Dissociation (CAD) is the most common fragmentation method used for tandem mass spectrometry. This method is also known in the literature as collision induced dissociation (CID).39,40 A neutral gas is used to induce fragmentation in this method. The selected precursor ion is usually accelerated to high kinetic energy and then allowed to collide with neutral atoms or molecules (He, N2, Ar).
During the collision, part of the ion’s kinetic energy is converted into internal energy which leads to fragmentation. The product ions are then analyzed by a mass analyzer as mentioned above.
2.2.2 Electron Transfer Dissociation (ETD)
ETD is a relatively new method, developed by Coon et al. in 2004.37 It requires the use of multiply charged precursor ions which are fragmented in the gas phase by electron transfer from reagent ions of opposite charge. Ion-ion reaction between positively charged precursor ions and negatively charged radical reagent ions produces radical ions.41,42 The radical ions (charge-reduced precursor ions that capture an electron from the reagent ions) dissociate via radical-induced fragmentation without randomization of their internal energy.43
2.3 Ion mobility (IM-MS)
Ion mobility spectrometry has been invented to overcome the limitation of conventional mass spectrometry which cannot separate isomers or isobars that have the same or very similar mass to charge ratio but different architectures or conformations.44,45 Ion mobility coupled with MS, i.e. IM-MS, can separate ions based
28 on their charge, size and shape. This technique has grown very rapidly, since it was shown that it can separate protein conformers.46
Ions in the ion mobility chamber travel under the influence of an electric field in the opposite direction of the flow of a buffer gas. This technique has been used in biological studies to simplify spectra, reduce chemical noise, and separate ionic isomers and conformers.
Four different types of ion mobility devices have been used for ion mobility based separation: drift time,47 aspiration,48,49 differential50 and traveling wave.15 Traveling wave ion mobility spectrometry (TWIMS) was the first commercially available ion mobility variant; it is interfaced with a Waters Q/ToF instrument (Synapt HDMSTM), which was used in this dissertation (see Figure 2.14).16
Figure 2. 14. Traveling wave (triwave) device inside the Synapt HDMSTM Q/ToF mass spectrometer.
Figure 2.14 shows the three main components of the traveling wave device which is located between the Q and ToF mass analyzers: trap, IMS and transfer cell. In the trap cell, the ions are trapped as packets before ion mobility separation. In the IMS cell, the ions in a packet drift against a stream of a buffer gas, such as helium or
29 nitrogen, under weak voltage pulses called traveling waves, and they are separated in this process according to their mobilities. The separated ions are transferred to the ToF mass analyzer after passing through the transfer cell. MS2 experiments on the separated ions can be performed in the transfer cell. On the other hand, fragmentation can be conducted in the trap cell if separation of the fragments by their mobilities is
51 -2 desired. Trap and transfer cells are filled with Ar at 10 mbar, whereas N2 flows through the IMS cell at ~ 1mbar.
Figure 2. 15. Stacked ring electrodes (top); schematic of the operation of a traveling wave ion guide (bottom). Reproduced with permission from reference 50.
The ion mobility chamber consists of a set of ring electrodes arranged in sequence. Opposite phases of RF voltage are applied to adjacent ring electrodes throughout the IMS cell so that ions introduced into this region maintain an axial motion.
30
In addition to the RF voltage, a DC voltage pulse (“traveling wave”) is also applied to sequential pairs of adjacent electrodes. The traveling waves move the ions down the
IMS cell (see Figure 2.15) which illustrates the stacked ring electrodes and the traveling wave process. Larger ions with large collision cross-sections will collide more frequently with the buffer gas that moves against the ions and their motion will slow down resulting in a longer drift time. On the other hand, ions with higher molecular weight carrying multiple charges can have short drift times because of their higher charge state.
Similarly, if the molecular structure allows the ion to fold around the charges to stabilize its structure through intramolecular solvation and minimize charge repulsion, a small size and shorter drift time would also be observed.15,17,52
The collision cross section of an ion can be obtained directly from the ion’s drift time through the ion mobility cell if a static field is applied to the cell.17,53 Under these conditions, the ion velocity 휐퐷 (see Equation 2.12) is proportional to the electric field E; the drift time (푡퐷) taken to cross the length (L) of the drift cell is shown in Equation
(2.13). The ion mobility constant (K) is directly proportional to the collision cross-section
(Ω) as shown in Equation 2.14.
휐퐷=퐾퐸 Equation 2.12
퐿 =퐾퐸 Equation 2.13 푡퐷
18휋 푧푒 1 1 1 1 퐾= √ √[ + ] Equation 2.14 16 √퐾푏푇 푚1 푚푁 푁 Ω
Where N is the number density (pressure) of the buffer gas, T is the effective temperature, 푘푏 is the Boltzmann constant, 푚1 is the mass of the ion, 푚푁 is the mass of
31 the buffer gas atom or molecule, z is the ion charge, e is elementary charge and Ω is the collision cross-section of the ion.54
In contrast, in traveling wave (T-wave) devices, there is no direct relationship between the electric field generated by the pulsed waves and the ion drift time through the mobility chamber. In this case, the drift time of the measured ions must be calibrated using the drift times of ions with already known collision cross-section before calculating the collision cross-section of the analyte ions.
As a result, Equation 2.15 (derived from 2.13 and 2.14) is converted to Equation
2.16, which contains correction parameters A and B: A is a correction parameter for the electric field and B is a correction factor for the non-linearity of the triwave device.
18π 푧푒 1 1 푡 퐸 760 푇 1 Ω = √ √[ + ] 퐷 Equation 2.15 16 √퐾푏푇 푚1 푚푁 퐿 푃 273.2 푁
√18π 푧푒 1 1 760 푇 1 퐵 Ω = √[ + ] 퐴푡퐷 Equation 2.16 16 √퐾푏 푚1 푚푁 푃 273.2 푁
Where the collision cross-section (CCS or Ω) depends on the masses of the ions and their charges. Dividing Equation 2.16 by ze and √[1/m1 + 1/mN] gives Equation
2.17 and a normalized CCS (Ω`) which is not dependent on masses and charges. The parameter A (Equation 2.17) can be combined with all constants in Equation 2.17 to give a single constant 퐴`, which is the correction factor for the temperature, pressure and electric field parameters (Equation 2.18). The relationship between CCS and normalized CCS is shown in Equations 2.19 and 2.20.
32
√18π 1 760 푇 1 퐵 Ω` = 퐴푡퐷 Equation 2.17 16 √퐾푏푇 푃 273.2 푁
퐵 Ω` = 퐴` 푡퐷 Equation 2.18
1 1 Ω = 푧푒 √[ + ] Ω` Equation 2.19 푚1 푚푁
1 1 Ω = 푧푒 √[ + ] 퐴` 푡 퐵 Equation 2.20 푚1 푚푁 퐷
퐶 푚 √ 푡` = 푡 -[ 푧 ] Equation 2.21 퐷 퐷 1000
A corrected (IM) drift time (푡퐷`) is obtained by Equation 2.21, where c is the enhanced duty cycle (EDC) delay coefficient.53,54 It is dependent on the specific instrument, and its value is between 1.4-1.6. Standards of known CCS, such as cytochrome c, myoglobin, bovine ubiquitin and/or ions from polymers like polyalanine,
` can be used to calibrate the Triwave drift time scale, by plotting 훺` vs. 푡퐷; the resulting calibration curve is then used to derive experimental CCS values from measured drift times for ions being analyzed in traveling wave instruments.55
33
CHAPTER III
MATERIALS AND INSTRUMENTATION
3.1 Materials
Methanol (MeOH), chloroform (CHCl3), water (H2O), tetrahydrofuran (THF), acetonitrile (ACN) and formic acid (FA), all HPLC grade, were purchased from Sigma-
Aldrich (St. Louis, MO). Ammonium acetate in LC-MS grade, was received from Fisher
Scientific (Pittsburgh, PA). Sodium trifloroacetate (NaTFA), which was used as a cationizing agent to improve ionization of the samples, was obtained from Fluka (Buchs,
Switzerland). α-Cyano-4-hydroxycinnamic acid (CHCA), 3,5-dimethoxy-4- hydroxycinnamic acid (sinapic acid, SA), trans-2-(3-(4-tert-butylphenyl)-2-methyl-2- propenyliedene) malononitrile (DCTB), and 1,8-dihydroxy-9,10-dihydroanthracen-9-one
(dithranol, DIT) were the MALDI matrices used to analyze the samples in this dissertation. CHCA and SA were purchased from Sigma‐Aldrich (St. Louis, MO), DCTB was bought from Santa Cruz Biotechnology (Santa Cruz, CA), and dithranol was purchased from Alfa Aesar (Ward Hill, MA).
The biodegradable materials poly(propylene maleate) (PPM) and poly(propylene fumarate) (PPF) were received from Dr. Matthew Becker (University of Akron).
The alanine-rich polypeptides AQK18 and GpAQK18 and their conjugates with polyethylene glycol (PEG) were received from Dr. Xinqiao Jia (University of Delaware).
The glycopolymer samples were received from Dr. Heather Maynard (trehalose glycopolymers and insulin-trehalose glycopolymer conjugate) and Dr. Andrea Kasko
34
(glycopolymer with controlled branching) (University of California Los Angeles, UCLA).
All materials were used in the condition received from their supplier without purification.
3.2 Instrumentation
The following sections describe the mass spectrometry instruments used in this dissertation. Specific instrument settings for each project are described in their corresponding chapters.
3.2.1 Ultraflex III ToF/ToF Mass Spectrometer
The MALDI-ToF/ToF mass spectrometer used to complete the work described in chapters IV, V and VI was a Bruker Ultraflex III (Bruker Daltonics, Billerica, MA). Bruker
Ultraflex III is equipped with a MALDI ion source operating with a Nd:YAG laser that emits light at a wavelength of 355 nm, two time-of-flight analyzers and a microchannel plate detector. The laser power of the MALDI source was adjusted for each compound to maximize the signal intensity without causing additional fragmentation. The two mass analyzers are a short linear time-of-flight tube (ToF-1) and a reflectron time-of-fight mass analyzer (ToF-2) interfaced with each other. The two ToF analyzers work as a single unit in either linear or reflectron MS mode (single-stage MS). The instrument also contains a LIFT device between ToF-1 and ToF-2, which is used for tandem mass analysis (MS2 mode). The principle of the MALDI ion source and ToF mass analyzers was discussed in Chapter II.
The MALDI parameters were adjusted as follows. The IS1 voltage was set to 25 kV, the IS2 voltage to 21.65 kV, the lens voltage to 9.65 kV, and the delay time to 150 ns. The reflectron lenses 1 and 2 were set at 26.30 kV and 17.30 kV, respectively.
35
The instrument was calibrated for each measurement using poly(methyl methacrylate) (PMMA); Mn=2kDa or 4kDaas a positive ion external standard (Sigma-
Aldarich, St.Louis, MO). FlexControl (version 3.4) was used for data acquisition and flexAnalysis (version 3.4) for data analysis. The laser power and the sample preparation conditions will be discussed in each respective chapter.
3.2.2 Synapt HDMS Mass Spectrometer
The Waters Synapt HDMSTM quadrupole/time-of-flight (Q/ToF) mass spectrometer (Milford, MA) was used to complete the work described in Chapter IV, V,
VI and VII. This instrument is equipped with an ESI ion source, two analyzers (a quadrupole and a time-of-flight tube, orthogonally to each other) and an ion mobility cell
(triwave) between the Q and ToF analyzers (see Figure 3.1). In MS mode, the quadrupole analyzer is set to RF-only mode that acts as an ion guide which transfers all ions coming from the source to the triwave chamber and ToF tube. The triwave device was described previously in Chapter II. In MS2 mode, the quadrupole is set to only transmit the selected ion. The precursor ion is then fragmented either in the trap cell or in the transfer cell.
36
Figure 3. 1. Scheme of the Waters Synapt HDMSTM ESI-Q/ToF mass spectrometer.
The ESI flow rate of the sample solutions was kept constant at 10 uL/min. The
ESI capillary voltage was set to 3.16 kV, the sampling cone voltage to 35 V and the extraction cone to 3.2 V. The source temperature and desolvation gas temperature were set to 80°C and 150°C, respectively, whereas the desolvation gas flow rate was set at 500 L/h. At single-stage MS analysis, the trap and transfer potentials were set to
6.0 V and 4.0 V, respectively. The gas (Ar) flow rate through these cells was kept constant at 1.5 mL/min, while the potential was changed for tandem mass experiments
MS2, as will be discussed in each chapter. The ion mobility conditions, viz. wave velocities and wave heights, will be discussed in each chapter. MassLynx (version 4.1) software package was used to acquire and analyze the mass spectral data, while the ion mobility data were analyzed using DriftScope (version 2.1).
37
3.2.3 HCT Ultra II ESI-QIT Mass Spectrometer
A Bruker HCT ultra II quadrupole ion trap (QIT) mass spectrometer (Bruker
Daltonics, Billerica, MA), shown in Figure 3.2, was used to perform the tandem mass experiments described in Chapter VI. The instrument is equipped with an ESI source, a quadrable ion trap mass analyzer and a Daly ion detector. Direct injection of the analyte solution into an ESI source through a syringe pump (KD Scientific, Holliston, MA) was utilized in this instrument. The solution was pumped into the ion source through a grounded needle surrounded by a nebulizer filled with N2 gas at 10 psi. The drying gas temperature was set at 300 ºC at a flow rate of 8 L/min. The nebulizer gas assists the spray formation and the drying gas promotes the solvent evaporation in the source. The ions formed in the ESI source are transmitted into a glass capillary that works as a transmission region that transfers the ions from atmospheric pressure to vacuum by the voltage difference between the grounded needle and the entrance of the glass capillary.
The entrance of the glass capillary was held between -4.5 kV and -1.5 kV relative to the spraying needle and at -0.5 kV relative to the end plate. The skimmers, octopoles and lenses in the transmission region are used to focus and transport the ions to the QIT mass analyzer as they arrive from the ion source. The ions are trapped and stored in the QIT, which is filled with helium gas (~10-3 mbar). All ESI spectra in Chapter VI were measured in positive mode.
38
Figure 3. 2. Scheme of ion transmission and detection in the HCTultra II QIT mass spectrometer, adapted with permission from reference 27.
The operation principles of the QIT, Daly detector and tandem mass spectrometry by CAD were discussed in Sections 2.1.2.3, 2.1.3.1 and 2.2.1, respectively.
39
CHAPTER IV
MULTIDIMENSIONAL MASS SPECTROMETRY CHARACTERIZATION OF ISOMERIC
BIODEGRADABLE POLYESTERS
4.1 Introduction
Biodegradable polymers undergo chemical decomposition in biological environments through enzymatic or non-enzymatic hydrolysis, without any prior thermal oxidation, photolysis or radiolysis.56-58 Such materials are utilized in a wide range of medical and ecological applications, including drug delivery, nanomedicine and tissue engineering.59,60 Their bone tissue engineering applications have been widely investigated in the last decades.61-63 Any material used for the latter purpose should be non-toxic, implantable and photo-cross-linkable once fabricated into the desired shape.64-66 These requirements are fulfilled by selected synthetic polyesters such as poly(lactide)s, poly(휀-caprolactone) and poly(propylene fumarate) (PPF).66-69 Temporary bone implants (scaffolds) made from these biodegradable polymers do not necessitate a second surgery for removal, in contrast to rigid, non-biodegradable implants which require a removal surgery that might cause unexpected complications.61,63 Polyesters based on fumaric acid are particularly suitable for bone tissue engineering because they are biodegradable, 3D printable into a variety of shapes and because they can be crosslinked at the double bonds on their backbone using photo-cross-linking 3D printing methodologies.66-70
40
The most extensively explored material in this field is the alternating copolyester poly(propylene fumarate) (PPF), which has superior degradation properties compared to alternative candidates like polylactides which undergo rapid bulk degradation causing localized inflammation or poly(휀-caprolactone), which needs years to degrade.59,71
Degradation of PPF through hydrolysis of the ester bonds proceeds within months and yields non-toxic products, viz. Fumaric acid, which is a naturally occurring substance (in the Krebs cycle), and 1,2-propanediol is a diluent commonly used in drug formulations.72 This material has met all important prerequisites for bone tissue engineering and orthopedic applications.
PPF is an unsaturated linear polyester copolymer; it is commonly synthesized by ring-opening polymerization of maleic anhydride and propylene oxide, to form poly
(propylene maleate) (PPM), which is subsequently isomerized to PPF using a base catalyst.70,73,74 The analytical methods generally employed to characterize polyesters include nuclear magnetic resonance (NMR)75 and infrared (IR) spectroscopy76 as well as mass spectrometry (MS).77-88 Whereas the first two methods provide definitive information about the functional groups and stereochemistry of the polymer chain, MS techniques are most powerful for determining the molecular weight of intermediate and final products, ascertaining their sequences and detecting the presence or absence of impurities or byproducts that might interfere with the synthesis and/or performance of the final material.
This study combines matrix-assisted laser desorption ionization (MALDI) and electrospray ionization (ESI) with time-of-flight (ToF) mass analysis and tandem mass spectrometry (MS2) fragmentation to elucidate the composition, end groups and the
41 chain sequence of PPM and PPF copolymers prepared according to the reaction sequence outlined in Scheme 1.15. In addition, ion mobility mass spectrometry (IM-
MS)89-96 is used to differentiate the isomeric PPM and PPF copolyesters and to probe the extent and efficiency of PPM to PPF (i.e. all-cis to all-trans) isomerization.
4.2 Experimental
4.2.1 Materials
All chemicals used for the synthesis or characterization were used as received.
Maleic anhydride (MA) (99%) and sodium trifluoroacetate (NaTFA) were purchased from Fluka (St. Louis, MO). Propylene oxide (PO) (99.5%), magnesium ethoxide
(Mg(OEt)2) (98%), diethylamine (99%, extra pure), toluene (anhydrous, 99.8%), tetrahydrofuran (THF) (ACS grade), chloroform (ACS grade) and methanol (MeOH)
(LC-MS grade) were purchased from Sigma-Aldrich (St. Louis, MO). The MALDI matrix trans-2-[3-(4-tert- Butylphenyl)-2-methyl-2-propenylidene] malononitrile (DCTB) (99+%) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA).
The synthesis of PPM via ring-opening polymerization of maleic anhydride and propylene oxide, using magnesium ethoxide as initiator, and the post-polymerization isomerization of PPM to PPF with diethylamine as catalyst have been described in details elsewhere.70 Size exclusion chromatography (SEC) analysis showed an average molecular weight (Mn) of approximately 2000 Da and a polydispersity (Mw/Mn) of <1.6 for both polymers.70
42
4.2.2 MALDI experiments
MALDI-MS and MS2 experiments were carried out on a Bruker UltraFlex III tandem time-of-flight (ToF/ ToF) mass spectrometer (Bruker Daltonics, Billerica, MA) equipped with a Nd:YAG laser emitting at a wavelength of 355 nm. Solutions of the polymers (10 mgmL•1), DCTB matrix (20 mgmL-1) and NaTFA cationizing salt (10mgmL-
1 ) were prepared in CHCl3. The matrix and salt solutions were mixed in the ratio 10:1
(v/v). Approximately 0.5–1 µL of matrix/salt solution was deposited in the MALDI sample plate and allowed to dry; 0.5–1 µL of sample solution was added on top of the dry matrix/salt spot, followed by another 0.5–1 µL of matrix/salt solution on the dried sample spot (three-layer sandwich method). MS2 experiments were performed using Bruker’s
LIFT mode with no additional collision gas.97 Bruker’s FlexAnalysis software was used for data analysis.
Scheme 4. 1. Synthesis of poly(propylene maleate) (PPM) via ring-opening polymerization of maleic anhydride and propylene oxide and base-catalyzed post- polymerization isomerization of PPM to poly(propylene fumarate) (PPF).70
43
4.2.3 ESI-MS and ESI-IM-MS experiments
-1 Stock solutions of the polymers were prepared in CHCl3 at 10 mgmL . The samples sprayed were diluted to 0.01 mgmL-1 by adding 1 µL of polymer solution to 300
µL CHCl3 and 700 µL MeOH. The final polymer solution was injected into a Waters
Synapt G1 HDMS quadrupole/time-of-flight (Q/ToF) mass spectrometer (Waters,
Milford, MA) equipped with the traveling-wave version of IM-MS.17,50 The traveling- wave region is located between the Q and ToF mass analyzers and comprises three confined cells in the order trap cell (closest to Q), IM cell and transfer cell (closest to
ToF). Instrument parameters were adjusted as follows: ESI capillary voltage, 3.5 kV; sample cone voltage, 30 V; extraction cone voltage, 3.2 V; desolvation gas flow, 550Lh-
1 (N2); trap collision energy (CE), 6.0 eV; transfer CE, 4.0 eV; trap and transfer gas flow,
1.5mLmin-1 (Ar); sample flow rate, 10 mL min-1; source temperature, 80 0C; and desolvation temperature, 150 0C.
IM-MS experiments were performed by setting the traveling wave velocity to 250
-1 -1 2 ms , the traveling-wave height to 7.5V and the IM gas flow to 22.7mLmin (N2). IM-MS experiments were acquired in the transfer cell, after IM separation, using a collision energy of 60 eV.
Collision cross-section (CCS) data for sodiated PPF and PPM oligomers were derived from the corresponding drift times, measured by IM-MS, after calibrating the drift time scale with ions of known CCS, as reported previously and briefly explained in section 2.3.53,98 Singly and doubly protonated polyalanine ions served as calibrants, which were analyzed at the same traveling-wave velocity, traveling wave height and IM gas flow conditions as the sodiated PPF and PPM ions.
44
4.2.4 Molecular modeling
Molecular dynamics simulations were performed on select sodiated PPM and
PPF oligomers using the Materials Studio software, version 7.0 (BIOVIA, San Diego,
CA). Energy and geometry optimizations were performed for 50 candidate structures from each polyester isomer. Theoretical collision cross-sections were calculated for the optimized structures by the projection approximation method99 available in the MOBCAL program.100
Figure 4.1 MALDI mass spectra of (a) poly(propylene maleate) (PPM) and (b) poly(propylene fumarate) (PPF). All ions are sodiated species with the composition
+ [Rn+EGs+Na] , where R and EGs designate the PPM/PPF repeat unit (C7H8O4, 156 Da) and the corresponding end groups (in red color), respectively.
45
4.3 Results and discussion
4.3.1 MALDI characterization of PPM and PPF
The MALDI mass spectra of poly(propylene maleate) (PPM) and its isomerization product, poly(propylene fumarate) (PPF), show one major and two minor distributions
+ differing in their end groups (EGs) and having the composition [Rn+EGs+Na] in which
R is the propylene maleate /fumarate copolyester repeat unit (C7H8O4, 156 Da). The m/z values of the major product (A in Figure 4.1) indicate a total end group mass of 46
Da (C2H6O), corresponding to an ethoxy moiety and a hydrogen atom. The C2H5O– chain end must arise from the Mg(OC2H5)2 catalyst (cf. Scheme 4.1) which evidently initiates the polymerization; conversely, the –H end group is introduced upon termination with aqueous hydrochloric acid.70 The minor products (B and C in Figure
4.1) appear 1 x 58 and 2 x 58 m/z units above the main product, and have total end group masses of 104 Da (C2H6O +1 x C3H6O) and 162 Da (C2H6O +2 x C3H6O), consistent with the incorporation of one or two additional propylene oxide comonomers, respectively (see structures in Figure 4.1); these byproducts point out that some oligomerization of propylene oxide may occur during copolymerization. It is noteworthy that the A, B and C product distributions in the MALDI mass spectra of PPM (Figure
4.1[a]) and PPF (Figure 4.1[b]) are very similar, verifying that no change in the ester linkages of the polymer chain (no transesterification events) takes place upon PPM to
PPF isomerization.
46
Figure 4. 2. MALDI-MS2 spectrum of the [M+Na]+ ion from the PPF 9-mer with
CH3CH2O– and –H end groups (m/z 1473.4). The scheme on the top shows the fragment ions arising from 1,5-hydrogen rearrangement over ester groups facing the CH3CH2O- ($, !) or -H (#, @) chain end. The consecutive dissociation of these fragments (→) leads to internal fragments (o). The Na+ ion has been omitted for brevity.
An asterisk above the fragment notation (*) indicates fragments ionized by H+ (Na+ is eliminated with the neutral fragment).
4.3.2 MS2 characterization of PPM and PPF
Tandem mass spectrometry was utilized to ascertain the connectivity and end groups of the PPM and PPF polymer chains. Figure 4.2(a) shows the MALDI-MS2 spectrum of the sodiated 9-mer from the major PPF product capped with CH3CH2O– and –H end groups. Na+ cationized polyesters that have been energetically excited
47 mainly dissociate via charge-remote 1,5-hydrogen rearrangement over the ester group which cleaves the ester bond, creating a fragment with a carboxylic acid chain end and that with a vinyl (alkene) chain end, as shown in the scheme on top of Figure 4.2.24,84,101
2 + Figure 4. 3. ESI-IM-MS spectra of the [M+Na] ions of the 6-mers with CH3CH2O– and
–H end groups (m/z 1005.2) from (a) PPF and (b) PPM. An asterisk above the fragment notation (*) indicates fragments ionized by H+. These spectra were acquired on the singly charged component of m/z 1005.2 after the ion mobility separation from multiple charged isobars (vide infra).
Random dissociation of the different ester groups in the polyester chain gives rise to several fragment series with the expected repeating units of 156 Da, which were
48
AE HV EV AH labeled in the spectrum by the acronyms ln (#), ln (@), ln ($) and ln (!). In these notations, ln indicates the linear structure of the fragments (having n repeat units) and the superscripts give the corresponding end groups: A, E, V and H denote a carboxylic acid, ethoxy, vinyl (alkene) and hydroxy end group, respectively (cf. Figure 4. 2).
Because PPF is unsymmetrically substituted, different product combinations arise from the dissociation at a fumarate group on the side terminated by CH3CH2O– ($ and !) vs. the fumarate group on the side terminated by –H (# and @).
AE It is important to note that two of the mentioned fragment series, viz. ln (#) and
EV HV AH ln ($), contain the CH3CH2O– chain end, whereas the other two, viz. ln (@) and ln
(!), contain the –H chain end. All these fragments may dissociate consecutively, through the same mechanism, to form smaller fragments with the same structures and internal
AV fragments (i.e. fragments missing both end groups); the latter were labeled by ln (o) in
Figure 4.2. The fragment series observed conclusively confirm the alternating comonomer connectivity of the propylene oxide and fumarate comonomers as well as the CH3CH2O– and –H chain ends.
The PPM chain gives the same fragmentation pattern as the PPF chain with slight variations in the relative fragment abundances; further, the fragmentation extent
(total fragment ion intensity vs. precursor ion intensity) is higher for PPM than for PPF.
These differences are more pronounced in the corresponding ESI-MS2 spectra, in which the energy supplied for fragmentation can be more effectively controlled with the equipment used in this study (see Experimental).
49
Figure 4.3 compares the ESI-MS2 spectra of sodiated PPF and PPM 6-mers capped with CH3CH2O– and –H end groups, acquired at the same collision energy (60 eV). The same fragment series are present in both spectra and the fragmentation patterns are compared to those observed on MALDI-MS2. It is obvious, however, that
PPF fragments are less efficiently, presumably because the trans double bond configuration increases the critical energy needed for the 1,5-hydrogen rearrangements.
This finding also suggests a higher intrinsic stability and slower biodegradability for PPF than for PPM.
4.3.3 IM-MS differentiation of PPF and PPM
IM-MS can resolve constitutional, geometric and conformational isomers without the need for concentrated solutions and/or highly purified samples.89-92,94-96,102-106 Here, this technique was used to distinguish the isomeric polyesters. IM-MS analysis separates PPM and PPF ions according to their drift time through the IM region (IM dimension), which depends on the ions’ charge and collision cross-section (CCS or Ω), and by their m/z, which is determined by the ions’ composition and charge (MS dimension). Figure 4.4(a) shows the result of such 2D analysis for PPF ionized by ESI.
The ions are separated based on their charge state (+1 to +3) into unique 2D locations with specific drift times and m/z ratios. Most intense are the singly charged species; the mass spectrum extracted from their mobility region (+1 region in Figure 4.4[a]) is depicted in Figure 4.4(b) and clearly attests that ESI conditions partly degrade the polyesters, presumably because of electrochemically produced acid at the ESI electrode.
50
Figure 4. 4. (a) 2-D ESI-IM-MS plot (m/z vs. drift time) of PPM; the mobility regions of singly, doubly and triply charged ions are encased in ovals. (b) Mass spectrum extracted from the region of singly charged ions, containing several ion distributions which include intact PPM ions with CH3CH2O– and –H end groups (46-Da end group mass) and degradation products with various end group masses (noted after the number of repeat units; see Figure 4.2 for plausible structures).
51
Nevertheless, a significant portion of the PPF with the end groups instilled during polymerization (i.e. CH3CH2O– and –H) survives intact to permit IM-MS analysis of the
+ + corresponding [M + Na] ions, which have the composition [Rn+C2H6O+Na] . PPM gives rise to very similar ESI-IM-MS results, the only difference being the absence of ions with
+ (CH3CH2)2NH2 charges. With PPF, such ions are observed (cf. Figure 4.4[b]), originating from residual diethylamine which was used for the PPM to PPF isomerization
(see Experimental).
+ The [M+Na] ions of short PPM and PPF chains with the original CH3CH2O– and
–H end groups (four to five repeat units) show significant differences in their drift times through the IM region; however, as chain length increases, the drift times of PPM and
PPF isomers become more similar and ultimately (≥10 repeat units) indistinguishable.
This trend is documented in Figure 4.5 by the drift time distributions of the 4-mers,
7-mers and 10-mers. With shorter chains (<9 repeat units), the all-cis PPM oligomers drift faster, reflecting a higher degree of compactness. On the other hand, longer chains attain architectures of comparable compactness, irrespective of the double bond geometry in the diacid unit. Hence, IM-MS can differentiate PPF from PPM if relatively small oligomers with <9 repeat units are probed.
For a more quantitative assessment of the influence of cis vs. trans double bond geometry on the resulting macromolecular architecture, the collision cross-sections
(CCS or Ω) of a series of sodiated PPM and PPF n-mers were derived from the corresponding drift times and are listed in Table 4.1. The calibration curve and details used to derive these data are provided at the end of “Result and discussion” in Figure
4.8 and Table 4.2. In addition, theoretical values of the collision cross-sections expected
52 for these oligomers were obtained by molecular dynamics simulations; these are included in Table 4.1.
Table 4. 1. Measured and calculated collision cross-sections (CCS) of singly sodiated
+ PPF and PPM oligomers with the composition [Rn+C2H6O+Na] and CH3CH2O– and –H end groups.
aDeduced from the corresponding drift times after calibrating the drift time scale with singly and doubly protonated polyalanine oligomers, which were analyzed at the same conditions as PPM and PPF.98,107 The calibration curve is constructed for the drift time
0.5485 to Ω conversion is Ω=474.15(td’) , where td’ and Ω’ are the corrected drift times and normalized collision cross-sections of the calibrant ions, respectively (Figure 4.8); the derivation of td’ and Ω’ values (Table 4.2) from the measured drift times of the calibrant ions and their published collision cross-sections has been explained in detailed elsewhere.98,107
53
Figure 4. 5. (a) 2-D ESI-IM-MS plot (m/z vs. drift time) of PPF; the mobility regions of singly, doubly and triply charged ions are encased in ovals. (b) Mass spectrum extracted from the region of singly charged ions, containing several ion distributions which include intact PPF ions with CH3CH2O– and –H end groups (46-Da end group mass) and degradation products with various end group masses (noted after the number of repeat units; see Figure 2 for plausible structures). The charge is provided by
+ + + the addition of H , Na or (C2H5)2NH2 (from residual PPM to PPF isomerization reagent). PPM leads to very similar ESI-IM-MS characteristics except for the absence of
+ (C2H5)2NH2 adducts.
54
Measured and calculated collision cross-sections agree very well with PPF
(within ≤2%) and are similar with PPM (within ≤4% for 4-mer to 7-mer and ≤7% for 8- mer). The excellent match of theoretical and experimental Ω values for PPF further validates that the post-polymerization isomerization of PPM to PPF proceeds with quantitative yield, as was also indicated by the NMR analysis of these products.70
Mirroring the trend of drift times mentioned above, Ω values of the same PPM and PPF n-mers become more similar as n increases. Consequently, the best possible differentiation is achieved by focusing on the smallest oligomers detected. Note that
PPM and PPF differentiation based on the IM-MS characteristics of short chains does not require samples of high purity or samples devoid of unsaturated byproducts, which would be needed for NMR differentiation.
Figure 4. 6. ESI-IM-MS drift time distributions (IM-MS chromatograms) of [M+Na]+ ions form PPF (top) or PPM (bottom) oligomers with CH3CH2O– and –H end groups and (a)
4, (b) 7 or (c) 10 repeat units.
Since PPF and PPM are biodegradable polymers, only degradation products may be available for analysis. It is therefore important to establish whether the polymer chain stereochemistry of the original material can be elucidated from degradation
55 products. This issue was addressed by examining the IM-MS features of a prominent hydrolysate in the ESI mass spectra of PPF and PPM, viz. the oligomer at m/z 1093.3
+ (Figure 4.4), which has the composition [R6+C7H8O4+H] and end groups with a 156-Da mass (C7H8O4), cf. Figure 4.6; such end groups can arise by hydrolysis at two ester sites in the original PPF or PPM polyester chains. The IM-MS drift time distributions of mass selected m/z 1093.3 from PPF and PPM comprise two well-separated peaks,
+ 2+ corresponding to [R6+C7H8O4+H] and [R13+C7H8O4+2H] ions.
Figure 4. 7. ESI-IM-MS drift time distributions (IM-MS chromatograms) of mass selected m/z 1093.3 from (a) PPF and (b) PPM. These ions contain superimposed singly protonated 6-mers and doubly protonated 13-mers; their end groups have the same composition as the repeat unit (C7H8O4). The collision cross-sections deduced from the shown drift times are (a) 382 A° 2 for the PPF-13-mer and 275 Å2 for the PPF 6-mer; (b)
423 Å2 for the PPM 13-mer and 269 Å2 for the PPM 6-mer.
56
The singly charged 6-mers drift at 7.57 ms for PPF and 7.13 ms for PPM, whereas the doubly charged 13-mers drift at 3.98 ms for PPF and 4.78 ms for PPM.
Again, significant differences between the PPF and PPM connectivities are observed in both charge states. Note that the singly charged PPF ions drift more slowly than the isomeric PPM ions (as seen before). In sharp contrast, the drift time order reverses for the dications, with the PPM frame now attaining a more extended architecture with a longer drift time than the corresponding PPF isomer; evidently, the PPF geometry is capable of minimizing charge repulsion without reverting to a fully extended architecture, presumably because of better intramolecular stabilization of the two charges through ion-dipole interactions.
Figure 4. 8 Drift time calibration curve, obtained by plotting the corrected drift times of singly and doubly protonated polyalanine oligomers against the corresponding normalized collision cross sections. Table 4.2 lists the data used to construct this curve.
57
Table 4. 2. Corrected collision cross-section of the polyalanine calibrant ions,
+z [H(Ala)nOH+zH] (z=1-2), deduced from drift times measured at a traveling wave velocity of 250 m/s and a traveling wave height of 7.5 V. See section 2.3 for a brief description of the calibration procedure.
z n MW m/z tD tD' Ω reduced mass Ω' 1 2 160.0902 161.098 1.17 1.152104 81 4.882522961 395.4844 1 3 231.1222 232.13 1.2 1.178518 101 4.99823524 504.8218 1 4 302.1402 303.148 1.23 1.20545 115 5.062966181 582.2411 1 6 444.2042 445.212 1.96 1.930249 146 5.133083896 749.4302 1 10 728.3202 729.328 4.21 4.171921 199 5.19352242 1033.511 1 12 870.3812 871.389 5.45 5.408378 223 5.209290303 1161.672 1 16 1154.516 1155.524 8.06 8.01207 276 5.22939183 1443.312 1 20 1438.605 1439.613 10.83 10.7765 337 5.241665182 1766.441 1 22 1580.784 1581.792 12.27 12.21392 353 5.246172924 1851.899 2 17 1225.152 613.584 1.77 1.735073 265 5.232966251 693.368 2 18 1296.536 649.276 1.99 1.954072 276 5.236189107 722.5941 2 19 1367.455 684.7352 2.2 2.163104 287 5.239062762 751.8055 2 20 1438.757 720.3865 2.41 2.372156 297 5.241670476 778.3881 2 21 1509.514 755.7647 2.62 2.581237 308 5.24401801 807.5788 2 22 1580.543 791.2791 2.89 2.850337 317 5.246165942 831.5173 2 23 1651.552 826.7839 3.07 3.029457 327 5.248130924 858.0694 2 24 1722.586 862.3007 3.11 3.068595 337 5.249936429 884.6143 2 25 1793.614 897.8149 3.55 3.507751 348 5.251600458 913.7785 2 26 1864.637 933.3265 3.55 3.506924 358 5.253139011 940.3119
58
4.4 Conclusions
This study demonstrated the utility of combining MALDI-MS, ESI-MS, MS2 fragmentation and IMMS for the complete structural characterization of biodegradable polyester isomers. The MS dimension unveiled compositional heterogeneity, the MS2 dimension helped establish copolymer connectivity and end groups, and the IM dimension enabled the differentiation of all-cis from all-trans polymer chains.
Mass analysis coupled with ion mobility separation is a promising technique for the analysis of such isomeric species. It is fast, requires little material and can reveal information about both the compositional as well as architectural microstructure of the sample being analyzed without the need for crystallinity or high purity (and large quantity) as it is true for other analytical techniques such as X-ray diffraction and nuclear magnetic resonance spectroscopy. Mass spectrometry-based analyses like the one described in this study should be particularly useful for the characterization of multicomponent biomaterials that are difficult to elucidate by methods probing average structures or methods sensitive to harmless impurities or byproducts that do not need to be removed for the desired application.
59
CHAPTER V
SEQUENCE AND CONFORMATIONAL ANALYSIS OF PEPTIDE-POLYMER
BIOCONJUGATES BY MULTIDIMENSIONAL MASS SPECTROMETRY
5.1 Introduction
Bioconjugates are hybrid materials containing biomolecules covalently linked to synthetic polymers.108 The most widely used polymer for this purpose has been poly(ethylene glycol) (PEG), a water soluble and nontoxic material that has been approved by FDA for use as drug conjugates or in biomedical devices since 1990.109,110
Covalent attachment of PEG to other molecules, known as PEGylation, is primarily applied to therapeutic peptide and protein drugs. This process improves the bioavailability as well as the physiochemical and pharmacokinetic properties of such drugs110,111 by increasing their solubility and stability, decreasing their aggregation proclivity, and generally minimizing adverse immune system responses against them.109-116
PEGylated therapeutic drugs are employed for the treatment of several chronic diseases, including cancer, hepatitis C, kidney disease, and Crohn’s disease.112
PEGylated drugs approved by FDA include PEG-aspargase for acute lymphocytic leukemia treatment, methoxy polyethylene glycol-epoetin beta for kidney disease treatment, and Certolizumab pegol for rheumatoid arthritis and Crohn’s disease treatment.109,110,114 Such compounds and many similar bioconjugates cannot be generally prepared in crystalline or highly purified form for molecular structure
60 characterizations by spectroscopic methods that probe average structures like NMR and X-ray diffraction spectroscopy.117-119 This limitation can be overcome by mass spectrometry based methods, where the desired product can usually be separated from impurities, residual reactants, and/or byproducts by its unique mass so that its primary structure can be examined by tandem mass spectrometry (MS2) fragmentation.96,101
Further separation efficiency as well as shape/size selectivity is gained with ion mobility mass spectrometry (IM-MS), which interfaces dispersion according to mass-to-charge ratio (MS dimension) with dispersion by collision cross-section and charge (IM dimension).92,120 Combining a soft ionization method, such as matrix-assisted laser desorption ionization (MALDI) or electrospray ionization (ESI), with IM-MS and tandem mass spectrometry fragmentation (MS2 dimension) creates a multidimensional technique that can provide insights into the composition, structure, and architecture of bioconjugates and other complex biomacromolecules.96,121 Since the complete analysis takes place in the mass spectrometer without the need for offline degradation, derivatization, or fractionation, this approach qualifies as a top-down multidimensional methodology.122,123 Its applicability has been demonstrated with the comprehensive characterization of the primary structure and architecture of an acrylate-based branched glycopolymer,124 a poly(acrylic acid)-peptide biomaterial,125 and most recently a polyether dendron conjugated with two different bioactive peptides.126
61
Figure 5. 1. Amino acid sequence of the peptides and PEGylated peptides investigated.
A pertinent question that has remained unanswered in these studies is whether conjugation affects biomolecular conformation and folding, thus causing denaturation which could have a negative effect on the efficacy of bioconjugate drugs. This issue is explored here with an investigation of the alanine-rich peptides AQK18 and GpAQK18
(where Gp represents propargylglycine) and their PEGylated forms AQK18-PEG and
GpAQK18-PEG in which the polymer was attached at the C-terminus through an amide bond (Figure 5.1).116,127 The present study illustrates the utility of the MS2 and IM-MS /
MS2 techniques for the determination of the sequence, derivatization site, and conformation of bioconjugates which play an increasingly important role in polymer- based biopharmaceutics.128
62
5.2 Materials and methods
5.2.1 Materials
The materials and solvents used for the synthesis and mass spectrometry characterization of the peptides and their conjugates were purchased from Fisher
Scientific (Pittsburgh, PA), Sigma-Aldrich (St. Louis, MO), and ChemPep (Wellington,
FL); all were used as received.
Peptides with the sequences Ac-KAAAQAAAQAAAQAAAQK-NH2 (AQK18) and
Ac-GpKAAAQAAAQAAAQAAAQK-NH2 (GpAQK18), where Gp denotes L- propargylglycine, and their conjugates with PEG (Figure 5.1), were prepared by standard Fmoc-based solid-phase synthesis protocols using a PS3 peptide synthesizer
(Protein Technologies, Tucson, AZ). Details of the synthesis and characterization of
AQK18 have been reported elsewhere.116 These methods were also used for the synthesis and purification of GpAQK18. The expected molecular weight of GpAQK18
(1774.9 MW) was verified by electrospray ionization (ESI-MS) using a Thermo Finnigan
LCQ Advantage mass spectrometer with Surveyor MS pump.
TentaGel PAP Resin (Rapp Polymere GmbH, Tubingen, Germany) was used for solid-phase synthesis of the peptides directly onto a PEG chain. The conjugate was cleaved from resin in 95% trifluoracetic acid (TFA), 5% thioanisole for 12 hours. TFA was removed via evaporation, and the conjugates were precipitated twice into cold ethyl ether. Samples were redissolved in water, frozen in liquid nitrogen, and lyophilized.
Dried samples were then dissolved in water and dialyzed against water using a 1000
63
MWCO regenerated cellulose dialysis membrane (Spectrum Labs, Rancho Domingo,
CA); water was exchanged twice daily for three days. Samples were again frozen in liquid nitrogen and lyophilized. The successful synthesis of the conjugates was verified by 1H nuclear magnetic response (NMR) spectroscopy, using a Bruker AV600 (600
MHz) with samples dissolved in D2O. The synthetic procedures employed for the peptides and the peptide-PEG conjugates yield peptides that were acetylated at the N- terminus; the free peptides were amidated at the C-terminus. All peptides and conjugates contain four blocks of three alanine and one glutamine residue, i.e. (A3Q)4, which are equipped with lysine residues (K) at the C-terminus and either K or GpK residues at the N-terminus, cf. Figure 5.1. All peptides and conjugates were stored as dried solids at 4 °C.
5.2.2 MALDI-MS and MS2 experiments
MALDI-MS and MALDI-MS2 experiments were performed on a Bruker UltraFlex
III MALDI tandem time-of-flight (ToF/ToF) mass spectrometer equipped with a Nd:YAG laser emitting at 355 nm (Bruker Daltonics, Billerica, MA). The instrument was operated in positive ion mode. CHCA (α-cyano-4-hydroxycinnamic acid), DHB (2,5- dihyroxybenzoic acid), and SA (3,5-dimethoxy-4-hyroxycinnamic acid) matrices and sodium, potassium, and lithium trifluoroacetate (NaTFA, KTFA, and LiTFA, respectively) cationizing salts were examined; those which gave the least in-source fragmentation and highest signal-to-noise ratio were selected for the experiments described below.
Matrices CHCA and SA were dissolved in acetonitrile (ACN) / H2O (70:30, v/v), and matrix DHB was dissolved in THF, all at 20 mg mL-1. The peptides and the
64
-1 conjugates were dissolved in H2O at 10 mg mL . For the PEGylated samples, trifluoroacetate salts (see above) were also used to enhance cationization; these salts were dissolved in MeOH at 10 mg mL-1, and these solutions were mixed with the matrix solutions in the ratio 1:10 (v/v). The matrix and sample or matrix/salt and sample solutions were spotted onto a 384-well ground-steel MALDI plate by the three-layer sandwich method protocol.125,126 The spots were completely dry before inserting the plate into the vacuum system. MALDI-MS2 experiments were performed using Bruker’s
LIFT (laser-induced fragmentation) mode with no additional collision gas.97 MS and MS2 data analysis was conducted with Bruker’s flexAnalysis software.
5.2.3 ESI-MS, IM-MS, and MS2 experiments
ESI-MS and ESI-MS2 experiments were carried out on a Waters Synapt HDMS quadrupole/time-of-flight (Q/ToF) mass spectrometer (Waters, Beverly, MA), equipped with an ESI source and traveling-wave ion mobility mass spectrometry (IM-MS).17 The interface region between the Q and ToF mass analyzers comprises a Triwave region containing three confined cells in the following order: trap cell (closest to Q), ion mobility
(IM) cell (intermediate), and transfer cell (closest to ToF).
Stock solutions of the peptides and conjugates were prepared in H2O at 10 mg
-1 -1 mL . These solutions were diluted with MeOH / H2O (1:1, v/v) to 0.01 mg mL
(peptides) or 0.10 mg mL-1 (conjugates) before being introduced into the ESI source by direct infusion at a flow rate of 10 μL min-1. The instrument was operated in positive ion mode with a capillary voltage of 3.15 kV, cone voltage of 35 V, sampling cone voltage of
-1 3.2 V, desolvation gas flow rate of 550 L h (N2), trap cell collision energy (CE) of 6.0
65 eV, transfer cell CE of 4.0 eV, trap/transfer gas flow of 1.5 mL min-1 (Ar), source temperature of 80 °C, and desolvation temperature of 150 °C.
IM-MS experiments were carried out by applying a traveling-wave velocity of 350 m s-1 and a traveling-wave height of 7.5 V to the IM cell; all other parameters were set as summarized above. Experimental collision cross-sections (CCSs) were derived from measured drift times through the IM cell after calibrating the drift time scale with polyalanine standards.53,125,126
ESI-MS2 and ESI-IM-MS2 experiments were carried out via collisionally activated dissociation (CAD) in the trap cell or (for mobility-separated ions) transfer cell, respectively, with Ar as collision gas and a CE of 40-80 eV.
5.2.4 Molecular modeling
Molecular dynamics simulations were performed using the leap-frog algorithm to solve Newton’s equation of motion with the GROMACS 5.0 software129-134 and
CHARMM27 force field135 in 1-fs steps. The VMD program136 was used for the analysis and images. Structures were simulated in vacuum. Center-of-mass translation and rotation around the center-of-mass were removed to avoid artifacts. Before starting an annealing simulation, the system was energy-minimized, using the steepest descent algorithm in vacuum, in order to avoid unrealistic interactions.
The collision cross-sections of the computationally optimized structures were calculated using the trajectory method in the MOBCAL suite of programs.137
5.3 Results and discussion
5.3.1 MALDI-MS and MALDI-MS2 analysis of peptides AQK18 and GpAQK18 66
AQK18 is an alanine-rich peptide with 18 amino acid residues, including four blocks of three alanine units (Ala or A = C3H5NO, 71.037 Da) and one glutamine unit
(Gln or Q = C5H8N2O2, 128.059 Da). The peptide is terminated with lysine residues (Lys or K = C6H12N2O, 128.095) at both termini. GpAQK18 has the same sequence as
AQK18, with the addition of a single propargylglycine residue (Gp= C5H5NO, 95.037 Da) at the N-terminus. The addition of the propargylglycine residue enables the future production of higher-number block polymers of the conjugates. Both peptides were acetylated at the N-terminus and amidated at the C-terminus, giving rise to elemental compositions of C70H121N25O23 (1679.907 Da) for AQK18 and C75H126N26O24 (1774.944
Da) for GpAQK18.
Figure 5.2a shows the MALDI mass spectrum of AQK18 using SA as matrix. The most abundant ion arises from the protonated peptide, [AQK18 + H]+, which is observed at m/z 1680.911, in excellent agreement with the calculated m/z of 1680.915. Minor peaks are detected for the sodiated and potassiated peptide as well as for incomplete sequences, missing mainly one alanine or one lysine residue. Very similar results are obtained for peptide GpAQK18 (Figure 5.3a).
MALDI-MS revealed that peptides with missing or additional amino acid residues are produced during solid-state peptide synthesis (cf. Figures 5.2a and 5.3a). Therefore,
MALDI tandem mass spectrometry was used to ensure that AQK18 and GpAQK18 have the desired connectivities. The MALDI-MS2 spectrum of [AQK18 + H]+ (m/z
1680.9) contains contiguous series of N-terminal bn (n = 1-17) and C-terminal yn (n = 6-
17) fragment ions as well as a few an and internal fragments such as AQ, AQA, and
AQAA (Figure 5.2b). All bn ions are acetylated, while the yn ions were amidated, thus
67 providing strong evidence that AQK18 has the proposed primary structure, as expected.
Similarly, the MALDI-MS2 spectrum of [GpAQK18 + H]+ (m/z 1776.0) confirms the N- terminal location of Gp and the overall sequence of this peptide (cf. Figure 5.3b).
Figure 5. 2. (a) MALDI-MS spectrum of AQK18 peptide. (b) MALDI-MS2 spectrum of protonated AQK18 (m/z 1680.9); bn and yn designate fragment ions that retain the N- or
C-terminus, respectively, which are formed via peptide bond cleavages. CO loss from bn generates the an fragments; ions labeled by single-letter code are internal fragments formed by consecutive fragmentation of bn or yn. (c) Peptide sequence corroborated by the MS2 fragments.
68
( (a) a)
(b) (
b)
Figure 5. 3. (a) MALDI-MS spectrum of GpAQK18 peptide. (b) MALDI-MS2 spectrum of protonated GpAQK18 (m/z 1776.0); bn and yn designate fragment ions that retain the N- or C-terminus, respectively, which are formed via peptide bond cleavages. Ions labeled by single-letter code are internal fragments formed by consecutive fragmentation of bn or yn. The fragments observed corroborate the sequence shown in Figure 5.1.
69
Figure 5. 4 (a) ESI-MS spectrum of AQK18 peptide. Peaks at m/z values labeled with a superscripted # arise from incomplete sequences (missing either one Lys or one Ala residue). Peaks without m/z labels are fragments from the doubly or triply charged peptide. (b) ESI-MS2 spectrum of doubly protonated AQK18 (m/z 840.952), acquired at a collision energy of 40 eV. The bn and yn fragment series observed corroborate the sequence Ac-KAAAQAAAQAAAQAAAQK-NH2 (see Figure 5.1).
5.3.2 ESI-MS (MS2) and ESI-IM-MS analysis of peptides AQK18 and GpAQK18
The ESI mass spectra of AQK18 and GpAQK18 (Figures 5.4 and 5.5, respectively) mainly show triply and doubly charged ions of the intact peptides. Low- intensity signals from incomplete sequences and fragment ions from in-source fragmentation are also present in both spectra. Mass-selection and collisionally
70 activated dissociation (CAD) of the doubly protonated peptides ([M + 2H]2+) gives rise to
2 ESI-MS spectra with contiguous series of singly charged bn and yn fragments that once more affirm the corresponding peptide sequences (cf. Figures 5.4-5.5).
Figure 5. 5 (a) ESI-MS spectrum of GpAQK18 peptide (Gp = propargyl glycine). Peaks at m/z values labeled with a superscripted # or @ arise from sequences missing one
Lys or one Ala residue (#) and sequences containing an extra Ala residue (@). Peaks without m/z labels are fragments from the doubly or triply charged peptide. (b) ESI-MS2 spectrum of doubly protonated GpAQK18 (m/z 888.479; Gp = propargyl glycine), acquired at a collision energy of 40 eV. The bn and yn fragment series observed corroborate the sequence Ac-GpKAAAQAAAQAAAQAAAQK-NH2 (see Figure 5.1).
71
A more precise molecular characterization is achieved by interfacing ESI-MS with ion mobility (IM) separation. With the IM cell turned on, the ions travel through an electric field against the flow of a bath gas, and they are separated in this process by their sizes, shapes, and charge states before mass analysis.92,96,120,121 In the overall IM-
MS experiment, 2D dispersion occurs, according to the drift time of the ions in the IM cell (IM dimension) and their m/z value in the mass analyzer (MS dimension).
Figure 5. 6. (a) 2D ESI-IM-MS plot (m/z vs. drift time) of AQK18 peptide and (b,c) mass spectra extracted from the mobility regions of (b) doubly and (c) triply charged AQK18 ions.
72
Scheme 5. 7 (a) 2D ESI-IM-MS plot (m/z vs. drift time) of GpAQK18 peptide and (b,c) mass spectra extracted from the mobility regions of (b) doubly and (c) triply charged
GpAQK18 ions.
This procedure separates chemical noise and fragments from the intact peptides in various charge states, as illustrated by the 2D IM-MS plot of AQK18 in Figure 5.6a.
The most intense regions in this plot contain intact peptide with 2+ or 3+ charges
(Figures 5.6b and 5.6c, respectively). 2D dispersion enables the observation of minor products such as peptide dimer in 3+ charge state and intact peptide with 4+ charges, which cannot be detected conclusively without the IM dimension (cf. Figure 5.4a vs.
5.6a). Similar compositional insight is gained for GpAQK18 by IM-MS analysis (cf.
Figure 5.7).
73
Figure 5. 8. IM-MS drift time distributions (mobilograms) of (a) doubly protonated vs. protonated-sodiated AQK18 and (b) quadruply protonated vs. triply protonated-sodiated
AKQ18-PEG70. Drift times are marked next to the peaks; see Table 5.1 for the corresponding Ω values.
Mass-selected ions can also be sent through the IM region to examine their isomeric or conformational purity and to detect any overlap of isobaric charge states based on the ensuing drift time distributions (“mobilograms”).96,120,138,139 The mobilogram of [AQK18 + 2H]2+ shows two peaks centered at 5.42 ms and 6.77 ms, while only a single peak centered at 6.77 ms is observed for [AQK18 + H + Na]2+ (see Figure 5.8a).
These data provide evidence for the existence of two distinct conformations for doubly charged AQK18, which are assigned to a compact random coil structure (drifting faster) and a more elongated α-helical conformer (drifting more slowly).140 The same conformational distribution is found for [GpAQK18 + 2H]2+ vs. [GpAQK18 + H + Na]2+, cf. Figure 5.9a.
74
Scheme 5. 9 IM-MS drift time distributions (mobilograms) of (a) doubly protonated vs. protonated-sodiated GpAQK18 and (b) quadruply protonated vs. triply protonated- sodiated GpAKQ18-PEG69. Drift times are marked next to the peaks; see Table 1.5 for the corresponding Ω values.
The drift time of an ion through the IM cell is proportional to its collision cross- section (CCS or Ω) which is the rotationally averaged forward-moving area of the ion as it drifts within the IM bath gas. CCSs derived from drift times measured by IM-MS provide information about the architecture and conformation of the macromolecular ion under study, analogous to that revealed for the neutral macromolecule by hydrodynamic volumers derived from retention times in size exclusion chromatography experiments.
The CCS values of the random coil and α-helical components of doubly protonated and protonated-sodiated AQK18 and GpAQK18 are listed in Table 5.1 and fall within the ranges found for other random coils and α-helices of alanine-rich peptides.140
75
Table 5. 1. Experimental collision cross-sections (Ω)
Ion m/z drift time (ms) Ω (Å2) a
random α-helix random α-helix
coil coil
[AQK18 + 2H]2+ 840.965 5.42 6.77 335 380
[AQK18 + H + Na]2+ 851.957 6.77 380
[GpAQK18 + 2H]2+ 888.478 5.78 7.40 347 400
[GpAQK18 + H + Na]2+ 899.478 7.40 400
4+ [AQK18-PEG70 + 4H] 1191.434 9.21 11.50 900 1020
4+ [AQK18-PEG70 + 3H + Na] 1196.938 9.30 11.50 905 1020
4+ [GpAQK18-PEG70 + 4H] 1204.191 9.48 12.00 915 1045
4+ [GpAQK18-PEG70 + 3H + Na] 1209.699 9.57 12.00 919 1045
a ±3%.
5.3.3 MALDI-MS and MALDI-MS2 analysis of bioconjugates AQK18-PEG and GpAQK18-PEG
The MALDI mass spectrum of PEGylated AQK18 (Figure 5.) includes a major distribution of singly sodiated oligomers with the PEG repeat unit (C2H4O; 44.026 Da), which are observed at m/z values that match those calculated for a (C2H4O)n chain capped with the AQK18 peptide (C70H121N25O23; 1679.907 Da). For example, the 66- mer of the bioconjugate would be expected at m/z 66 x 44.02621 (PEG) + 1679.907
76
(AQK18) + 22.990 (Na+) = 4608.627, in excellent agreement with the measured m/z value of 4608.601. Two minor distributions arising from potassiated and protonated oligomers are also formed upon MALDI. Very similar MALDI-MS characteristics are observed for GpAQK18-PEG (Figure 5.11).
Figure 5.10. MALDI-MS spectrum of C-terminally PEGylated AQK18; the inset shows an expanded view of the m/z 4600-4800 range.
Under MALDI-MS2 conditions, [M + Na]+ ions from AQK18-PEG dissociate to form N-terminal an and C-terminal yn fragments, as demonstrated for the 63-mer (m/z
4476.5) in Figure 5.12. All an ions are the same as those formed from the unconjugated peptide, whereas all yn ions appear at much higher m/z values relative to those from the unconjugated peptide, thus confirming the C-terminal attachment of the PEG chain. The combined MALDI-MS and MS2 data clearly show that PEG-peptide conjugation took
77 place with the expected linkage and end groups. This conclusion is further substantiated by the MALDI-MS and MALDI-MS2 spectra of GpAQK18-PEG (Figures 5.11 and 5.13).
Figure 5.11 MALDI-MS spectrum of C-terminally PEGylated GpAQK18; the inset shows an expanded view of the m/z 4400-4700 range. The minor peaks between those of the major distribution correspond to potassiated and protonated oligomers.
78
Figure 5.12. MALDI-MS2 spectra of the [M + Na]+ ions from (a) AQK18 (m/z 1702.9) and
(b) AQK18-PEG63 (m/z 4476.5). The structure on top shows the bonds cleaved in the bioconjugate to form the N-terminal an and C-terminal yn fragment series and the single fragments c17 and x17.
79
Figure 5.13. MALDI-MS2 spectra of the [M + Na]+ ions from (a) GpAKQ18 (m/z 1798.1) and (b) GpAKQ18-PEG64 (m/z 4615.4). The fragments observed confirm the peptide sequence and the C-terminal conjugation of the PEG chain.
5.3.4 ESI-IM-MS (MS2) analysis of bioconjugates AQK18-PEG and GpAQK18-PEG
The ESI mass spectra of the PEGylated peptides are very complex due to superimposed charges and molecular weight distributions. The use of the IM dimension is imperative to overcome this problem and separate the ion mixture based on its charges, so that interpretable mass spectra can be extracted. The 2D IM-MS plots of the bioconjugates (see Figure 5.14 a) contain five well-separated regions for charge states 3+ to 7+. The least convoluted mass spectra are extracted from the group of ions with 4+ charges (see Figure 5.14 b), which are generated by adding mainly four protons and, to a lesser extent, three protons plus a sodium cation. More combinations of
H+/Na+ charges and/or poorer signal-to-noise ratio are observed in the spectra of the
80 other charge states. Again, measured m/z ratios are in excellent agreement with those calculated for the desired bioconjugates. For example, the [M+ 4H]4+ ions of AQK18-
PEG70 and GpAQK18-PEG70 are observed at m/z 1191.434 and 1215.207, respectively, which match within 7 ppm the corresponding calculated m/z values of 1196.443 and
1215.203, respectively. Furthermore, ESI-IM-MS2 experiments on select (mobility- separated) n-mers affirm the C-terminal conjugation site of the polymer, in keeping with
2 the MALDI-MS data; this is illustrated for AQK18-PEG71 in Figure 5.15.
Figure 5.14. (a) 2D ESI-IM-MS plots (m/z vs. drift time in IM cell) and (b) extracted mass spectra of charge state 4+ of bioconjugates AQK18-PEG (top) and GpAQK18-PEG
(bottom).
The conformational composition of the PEGylated peptides was examined by the drift time analysis of the 4+ ions. Generally, native solution conformations are retained
81 in lower charge states, where charge repulsion in the gas-phase ions formed by ESI is minimized. The lowest charge state with adequate intensity to permit the acquisition of drift time distributions was 4+ (cf. Figure 14a); those resulting from mass-selected
4+ + [AQK18-PEG70 + 4H] (m/z 1191.434) and [AQK18-PEG70 + 3H + Na] (m/z 1196.938) are included in Figure 5.8b, whereas the mobilograms of the corresponding congeners from GpAQK18-PEG70 are depicted in Figure 5.9b. Interestingly, the bioconjugates show a bimodal drift time distribution both when quadruply protonated and when charged by 3H+ + Na+. In both cases, the faster drifting random coil conformer is more abundant than the more slowly drifting α-helical conformer; the proportion of the α- helical structure slightly increases in the presence of the Na+ cation. The CCS values of the random coil and 훼-helical components of the bioconjugates are summarized in
Table 5.1.
82
Figure 5.15. (a) ESI-IM-MS mobilogram of quadruply protonated AQK18-PEG71 (m/z
1202.45). (b) 2D ESI-IM-MS2 plot of [M + 4H]4+, acquired by IM separation followed by
CAD at a collision energy of 80 eV. Two bands are observed for m/z 1202.45, corresponding to the random coil and α-helical conformer of AQK18-PEG71. Note that barely any fragments are formed from the α-helical conformer, consistent with a higher stability and dissociation after collapse to the random coil structure. (c) ESI-IM-MS2 spectrum extracted from the ions drifting at 9.30 ms (random coil conformer). None of the N-terminal fragments (an and bn) but all of the C-terminal fragments (yn) contain the
PEG chain, validating that the polymer is conjugated at the C-terminus.
5.3.5 Dependence of helical content and stability on bioconjugation and salt
Our results for the [M + H + Na]2+ ions of the peptides clearly demonstrate that sodiation promotes the formation of α-helical structures. A similar trend was found by
83
IM-MS studies on singly sodiated polyalanines or alanine-rich peptides.141,142 It is worth noting that IM-MS measurements probe gas-phase (viz. unsolvated) structures. Circular dichroism (CD) spectroscopy studies have, however, reported that α-helical contents also increase in solution when sodium salts are added.143,144
Our study here employed water/methanol solutions, which produced the ion intensities needed for ESI-IM-MS experiments; using 100% aqueous solutions or solutions in ammonium acetate buffer led to the same mobilograms (coil/helix proportions), but at lower sensitivity. The helical content, determined via CD analysis, of
AQK18 and AQK18-PEG in aqueous buffer are similar to one another and despite the changes in solvent between the CD and MS studies, the helicity values determined via
CD are similar with the helical content of AQK18-PEG with 4 proton charges (cf. Figure
5.8), strongly suggesting that PEGylation preserves the native conformation. CD confirms that helical content is present for the GpAQK18 and GpAQK18-PEG as well.
The lower helical content of AQK18 in the MS analysis could be due to partial unfolding during ESI of the unprotected peptide. Conversely, the higher helical content of AQK18-
PEG with 3H+ + 1Na+ charges is ascribed to the stabilization of α-helices by sodiation
(vide supra).
Previous studies have documented that helical structures are preferred if peptides are positively charged at or near the C-terminus.140,141 PEG conjugation at the
C-terminus could stabilize cationic charges near this location through noncovalent interactions with the polyether chain; PEG can also bind cations, thereby increasing the positive charge density at the C-terminus. Both of these events would help preserve the helical conformation and prevent unfolding. This premise was investigated further by
84 molecular mechanics/dynamics simulations on the AQK18 peptide and AQK18-PEG bioconjugate structures and collision cross-sections.
Table 5. 2. Simulated structures and collision cross-sections (Ω)
Ion Conformer Charge sites Ω (Å2) a [AQK18 + 2H]2+ Random coil K1 (H), K18 (H) 343 (3)
Random coil K1 (H), Q5 (H) 365 (9) or K1 (H), Q9 (H) or K1 (H), Q13 (H) or K1 (H), Q17 (H)
α-helix Q5 (H), K18 (H) 389 (17) or Q9 (H), K18 (H) or Q13 (H), K18 (H) or Q17 (H), K18 (H)
[AQK18 + H + Na]2+ α-helix Q5 (Na), K18 (H) 407 (8) or Q9 (Na), K18 (H) or Q13 (Na), K18 (H) or Q17 (Na), K18 (H)
partial or bent K1 (H), Q5 (Na) 371 (12) helix or K1 (H), Q9 (Na) or K1 (H), Q13 (Na) or K1 (H), Q17 (Na)
4+ [AQK18-PEG70 + 4H] Random coil K1 (H), K18 (H), PEG (2H) 968 (17)
α-helix Q17 (H), K18 (H), PEG (2H) 1040 (98) or K18 (H), PEG (3H) or K1 (H), PEG (3H)
4+ [AQK18-PEG70 + 3H + Na] Random coil K1 (H), K18 (H), PEG (H + Na) 902 (52)
α-helix Q17 (H), K18 (H), PEG (H + Na) 1038 (15) or Q17 (Na), K18 (H), PEG (2H) a Average value for the listed charge sites. The number in parenthesis is the standard deviation of the individual CCSs used to calculate the average value.
85
5.3.6 Structure and collision cross-section simulations
Two initial confirmations were used for the doubly protonated AQK18 peptide: an extended conformer and a right hand α-helix. The amine groups on the N- and C- terminal lysines and the amide groups of the glutamines were considered as protonation sites. The simulation results show that adding protons to both lysines (K1 and K18, see numbering in Table 5.2) consistently results in a random coil conformation (see Figure
5.16a) with a CCS of 343 Å2. Random coil structures persist if one proton is retained on the N-terminal lysine (K1), while the other is moved to one of the glutamines (Q5, Q9,
Q13, or Q17), cf. Figure 5.17; their larger average CCS of 365 Å2 (Table 5.2) suggests a less dense hydrogen bonding network than in the random coil generated by the protonation of the two lysines. On the other hand, retaining one proton on the C-terminal lysine (K18) and placing the other on one of the glutamines gives rise to helical conformations (see Figures 5.16b and 5.17), having an average CCS of 388 Å2 (Table
5.2). Our IM-MS mobilogram of [AQK18 + 2H]2+ (Figure 5.8a) and experimental CCS values (Table 5.1) are best reconciled by the presence of a major random coil conformer protonated at the two lysines and a minor helical conformer protonated at the
C-terminal lysine and one of the glutamine residues. The lower proton affinity of glutamine (938 kJ mol-1)193, compared to lysine (996 kJ mol-1)193, justifies the predominance of the random coil structure.
86
Figure 5.16. [AQK18 + 2H]2+ conformers; (a) random coil structure protonated at lysines
K1 and K18 and (b) α-helical structure protonated at glutamine Q17 and lysine K18.
For protonated-sodiated AQK18, the proton was attached to one of the lysines and the sodium ion to one of the glutamines because of the high proton or sodium ion affinity of these residues, respectively.147, 193 The protonation at K18 consistently yields an α-helix, independent of which Q residue binds Na+ (see Table 5.2 and Figure 5.18); the average CCS predicted for these conformers is 407 Å2. Interestingly, moving the proton to K1 does not produce random coil structures but partial or bent helices with smaller CCS (372 Å2). The measured CCS of 380 Å2 strongly suggests that all these helical structures are sampled in the IM-MS experiment.
With quadruply protonated and triply protonated-sodiated AQK18-PEG70, the starting point in the simulations was either an extended or a helical conformation of the peptide attached to an extended PEG chain. The charge distribution on the PEG block was varied: charges were distributed evenly along the PEG chain or placed close to the free PEG chain end to minimize repulsion by the charges on the peptide.
87
4+ For [AQK18-PEG70 + 4H] , either two or one proton was placed on the peptide and the rest on the long PEG tail. The protonation sites considered for the AQK18 segment were the same as for the bare peptide (its K and Q residues). If both lysines are protonated, the simulations predict a random coil structure for the peptide, a random coil conformation for the PEG tail, and some interactions between the PEG and peptide blocks (cf. Figure 5.19a). The average structure has a CCS of 968 Å2 (Table 5.2), a value that reasonable agrees with the experimental value of 900 Å2 (Table 5.1)
Figure 5.17. Random coil (top) and α-helical (bottom) structures of doubly protonated
AQK18. The proton attachment sites are given under the corresponding structures. See
Table 5.2 for the sequence of AQK18.
88
Figure 5.18. Isomeric structures of protonated-sodiated AQK18 with partially helical or bent helical structure (top) or α-helical structure (bottom). The proton and sodium ion attachment sites are given under the corresponding structures. See Table 5.2 for the sequence of AQK18.
Figure 5.19. [AQK18-PEG70 + 4H]4+ conformers; (a) random coil structure protonated at lysines K1 and K18 and (b) α-helical structure protonated at glutamine Q17 and lysine
K18. In both structures, two additional protons are attached on PEG.
89
4+ Figure 5.20. [AQK18-PEG70 + 4H] tautomer with α-helical structure, having three protons on the PEG chain and one proton at lysine K18.
4+ Figure 5.21. [AQK18-PEG70 + 3H + Na] conformers; (a) random coil structure protonated at lysines K1 and K18 and (b) α-helical structure protonated at glutamine
Q17 and lysine K18. In both structures, an additional proton and a sodium ion are attached on PEG.
90
Conversely, helical conformations are observed if the peptide is protonated at
K18 and one glutamine and two protons are placed on the PEG chain (see Figure
5.19b), or if only lysine K18 is protonated and three protons are bound to PEG (see
Figure 5.20). In both cases, hydrogen bonding interactions between the PEG chain and the C-terminal peptide segment are evident; such intramolecular solvation stabilizes positive charges at the C-terminus which also enhances the stability of α-helical structures (vide supra). It is also noticed that the PEG chain does not interact with the
N-terminal segment of the peptide in the helical conformations (cf. Figures 5.19b and
5.20 vs. Figure 19a. Thus, the PEG charges only increase the charge density at the C- terminus, further favoring the helical structure. The set of possible helical conformers of the quadruply protonated bioconjugate has an average CCS of 1040 Å2 and a considerable spread in compactness (see Table 5.2), which explains satisfactorily the measured CCS of 1020 Å2 (see Table 5.1) and the experimentally observed broad tail in the mobilogram of Figure 5.8b.
Similar results are obtained for the triply protonated-sodiated bioconjugate when the Na+ is attached on the PEG chain which is the most probable Na+ binding site due to the high Na affinity of polyethers145,146 (relative to peptides of similar size)147 and their ability to coordinate metal ions in multidentate fashion.145,146 With Na+ and one H+ on the PEG chain, a random coil structure results if both K1 and K18 are protonated, whereas helices are formed if the K1 proton is transferred to one glutamine residue (cf.
Table 5.2 and Figure 5.21).
91
5.4 Conclusion
The sequence and conformation of the alanine-rich polypeptides AQK18 and
GpAQK18 and their conjugates with PEG were elucidated by MALDI-MS, ESI-MS, MS2 fragmentation, and shape-specific dispersion by IM-MS. Fragmentations by either
MALDI-MS2 or ESI-MS2 affirmed that the PEG chain is attached to the C-terminus of the peptides. The IM-MS experiments revealed the existence of random coil and helical conformers in both the peptides and the bioconjugates. The IM-MS data also provided evidence that the helical structure is stabilized by PEG attachment at the C-terminus.
The collision cross-sections of the random coil and helical conformers of the simulated polypeptides in charge state 2+ and their PEG conjugates in charge state 4+ agreed well with the experimental values and provided snapshots of the corresponding architectures, which explain adequately the stabilizing effect of the polymer chain.
Overall, this study documented that multidimensional MS offers a fast and sensitive analytical tool for the characterization of the compositions, sequences, and secondary structures of biomacromolecules and bioconjugates.
92
CHAPTER VI
MASS SPECTROMETRY CHARACTERIZATION OF GLYCOPOLYMERS WITH
CONTROLLED BRANCHING
6.1 Introduction
Synthetic glycopolymers containing pendant carbohydrate moieties have recently gained particular interest due to potential applications in many biomedical fields such as drug delivery, molecular recognition interactions and biotherapeutics.124,148-150
Synthetic glycopolymers can be used as models to study the biological functions of carbohydrates because of their ability to mimic the biological actions of sugars which are very difficult to isolate from natural sources.151 Polysaccharides play a key role in numerous biological processes including cellular recognition, signal transmission, inflammation, and cancer cell metastasis, all of which are mediated by noncovalent carbohydrates-protein interactions.151-154 Carbohydrate-protein interactions often involve clustered glycans to enhance the binding affinity (“multivalent effect”).153,154
Glycoproteins and glycolipids are common examples of such behavior in the living cells.155 Synthetic glycopolymers, which can be prepared with varying density of glycan units, offer a means to gain a better understanding of the multivalent effect and the roles of carbohydrates in biological functions.
93
Figure 6.1 Synthesis of branched poly(acryloyl-1,2:3,4-di-O-isopropylidene-α-D- galactose) by RAFT polymerization using galactose acrylate monomer and chain transfer agent in the present of AIBN initiator.
Different polymerization methods have been reported for the synthesis of well- defined glycopolymers.156 Controlled living radical polymerization techniques, such as atom transfer radical polymerization (ATRP),157,158 ring opening metathesis polymer- ization (ROMP),159,160 nitroxide-mediated polymerization (NMP)161 and reversible addition–fragmentation chain transfer (RAFT)162-164 have been used to synthesize glycopolymers with controlled structures. RAFT techniques have been most widely used to synthesize glycopolymers because they tolerate a variety of solvents and monomers and do not need a metal catalyst which is important for biomedical uses.165
Furthermore, this polymerization method is applicable of both protected and unprotected glycomonomers either in aqueous or organic media which is simplifying the synthetic process. In RAFT polymerization, a chain transfer agent (CTA) with dithio- or trithio-terminal groups is utilized, which usually results in polymers with thiol end
94 moieties that enable a variety of bioconjugation and postmodification reactions by thiol chemistry.151
The glycopolymers studied in this work were synthesized by RAFT copolymerization of galactose acrylate and a polymerizable chain transfer agent branching unit that allowed for systematic variation of the number of branches.166
Figure 6. 2 MALDI mass spectrum of the 121 glycopolymer studied. Three ion
+ distributions of oligomers with the [Rn+EGs+Na] composition are observed; R is the galactose acrylate monomer and RAFT CTA is the branching unit. The observed oligomers have end groups (EGs) of 101 Da ($), 134 Da (%) and 204 Da (&). Figure 6.3 shows the structures of these oligomers.
95
Matrix-assisted laser desorption ionization and electrospray ionization mass spectrometry (MALDI-MS and ESI-MS, respectively), ion mobility mass spectrometry
(IM-MS), as well as tandem (MS2) and multistage mass spectrometry (MSn) were used to characterize the composition, possible end groups and architecture of such glycopolymer.
6.2 Experimental
6.2.1 Materials
The glycopolymer studied was synthesized by RAFT copolymerization of galactose acrylate and a polymerizable chain transfer agent (CTA) branching unit as has been reported in the literature.166 The sample analyzed contained poly(acryloyl-
1,2:3,4-di-O-isopropylidene-α-D-galactose) with an average degree of polymerization of
12 and one 2-acryloylethyl (phenylcarbonothioylthio) propionate branching unit
(acronym 121).
6.2.2 Matrix-assisted laser desorption/ionization (MALDI) experiments
MALDI-MS and MS2 experiments were carried out on a Bruker UltraFlex III tandem time-of-flight (ToF/ToF) mass spectrometer (Bruker Daltonics, Billerica, MA), equipped with a Nd: YAG laser emitting at a wavelength of 355 nm. The sample was
-1 dissolved in either THF or CHCl3 at a final concentration of 10 mg mL and deposited onto the sample plate using the sandwich method, with trans-2-[3-(4-tert-butylphenyl)-2- methyl-2-propenylidene] malononitrile (DCTB) and sodium trifluoroacetate (NaTFA) serving as matrix and cationizing agent, respectively. Solutions of the matrix (20 mg
96 mL−1) and cationizing salt (10 mg mL−1) were prepared and mixed in the ratio 10 :1, respectively. Approximately 0.5 μL of the resulted mixture was spotted on the MALDI plate (384-well ground-steel), then the sample was applied onto the dried matrix/salt spot, and a second layer of the matrix/salt mixture was applied again on the top of the dried sample. MS2 experiments were performed using Bruker’s LIFT mode with no additional collision gas.97 FlexAnalysis software was used to analyze the MALDI data.
6.2.3 Electrospray ionization (ESI) experiments
A stock solution of the polymer was prepared in THF at 10 mg mL−1. The sample sprayed was prepared by adding 1 μL of the polymer solution to a mixture of 700 μL
MeOH and 300 μL THF to obtain a final polymer concentration of 0.01 mg mL−1 (in
MeOH: THF 7:3, v/v %).
For MSn experiments, the final polymer solution was injected into a Bruker
HCTultra II quadrupole ion trap (QIT) mass spectrometer (Bruker Daltonics, Billerica,
MA) by direct infusion with a syringe pump at a flow rate of 250 μL h−1. The temperature
−1 and flow rate of the drying gas (N2) were 300 °C and 8 L min , respectively; the pressure of the nebulizing gas (N2) was set at 10 psi.
MS2 spectra were acquired by isolating the selected precursor ion and accelerating it with an RF field to induce collisionally activated dissociation (CAD) with the He gas in the QIT; for MSn spectra (n = 3-4), the isolation and activation procedures were repeated on a specific fragment in the MS2 or MS3 spectra.
97
6.2.4 Ion mobility mass spectrometry (IM-MS) experiments
The glycopolymer architecture was examined by ion mobility mass spectrometry
IM-MS on a Waters Synapt HDMS quadrupole/time-of-flight (Q/ToF) mass spectrometer
(Waters, Milford, MA), equipped with the traveling wave version of IM-MS. Instrument parameters were adjusted as follows: ESI capillary voltage, 3.5 kV; sample cone
−1 voltage, 30 V; extraction cone voltage, 3.2 V; desolvation gas flow, 550 L h (N2); trap collision energy (CE), 6.0 eV; transfer CE, 4.0 eV; trap gas flow, 1.5 mL min−1 (Ar); IM
−1 −1 gas flow, 22.7 mLmin (N2); sample flow rate, 10 μL min ; source temperature, 80 °C; desolvation temperature, 150 °C; IM traveling wave velocity, 350 m s−1; and IM traveling wave height, 11 V.
The collision cross-section (CCS) of glycopolymer oligomers with the
+x composition [Rn+EGs+ xNa] was deduced from the corresponding drift times after calibration of the drift time scale with ions of known CCS. Polyalanine ions served as calibrants. The calibration curve was obtained by plotting the corrected CCSs of the calibrant ions against their corrected drift times measured by IM-MS at the same instrument parameters used for the glycopolymer (see table 6.1).
6.3 Results and disscussion
6.3.1 MALDI-MS and MS2 analysis
Figure 6.2 shows MALDI mass spectrum of the protected glycopolymer, which had a degree of polymerization of 12 and one branching unit. Several distributions with
+ different end groups (EGs) and the composition [Rn+EGs+Na] are observed. The mass
98 difference between two adjacent peaks within each distribution is 314 Da, which matches the mass of the protected galactose monomer (C15H22O7). End groups with masses of 101 Da, 134 Da and 204 Da corresponding to C4H7NS, C5H10O2S and
C8H12O4S moieties, respectively, detected; plausible structures are illustrated in Figure
6.4. The product distributions in the MALDI-MS spectrum confirm the successful synthesis of protected glycopolymer with one branch unit.
MALDI-MS2 analysis of the protected glycopolymer (Figure 6.3) shows that the galactose acrylate chains mainly undergo consecutive and competitive decompositions via H rearrangement at either the di-O-isopropylidene protecting groups or the galactose side chain ester groups, giving rise to neutral losses of acetone (58-Da) and a
242-Da moiety which corresponds to the dehydrated, protected galactose side chain with an exocyclic double bond, cf. Scheme 6.1a and 6.1d. Fragments arising by a 200-
Da loss from cross-ring cleavage at the protected galactose pendant are observed as well (Scheme 6.1b). The loss of 200 Da may be followed by CO loss which leads to ions missing 228-Da (Scheme 6.1c). Fragmentation can also occur through random homolytic C–C bond cleavages in the Na+ cationized main chain of the glycopolymer,
. . resulting in radicals (bn or zn in Scheme 6.2), which then undergo monomer losses and
backbiting H rearrangements that give rise to the internal fragments Kn and Jn (Scheme
6.2); the later ions may decompose further by losses of acetone moieties to form Kna and Jna fragments (Scheme 6.2). As is evident from Figure 6.3 a-c, the same types of fragmentation pathways are observed for all glycopolymer distributions, irrespective of whether they have end groups of (a) 101 Da, (b) 134 Da or (c) 204 Da.
99
100
2 +1 Figure 6. 3 MALDI-MS mass spectra of sodiated [R6+EGs+Na] 6-mers (a) EGs= 101
Da (m/z 2008.8), (b) EGs= 134 Da (m/z 2041.9) and (c) EGs= 204 (m/z 2111.8); the numbers on top of the peaks give the monoisotopic m/z ratio (in black) and the mass of the neutral loss(es) in Da (in color). The * sign indicates internal fragments Jna. The insets show the structures and end groups that agree well with these spectra.
101
Figure 6. 4 Proposed structures of the oligomers observed in the MALDI and ESI
+n spectra. All have the composition [Rn+EGs+nNa] , where R and EGs designate the repeat unit of the protected galactose acrylate (C15H22O7, 314.13 Da) and the corresponding end groups, respectively.
6.3.2 ESI-MS and IM-MS analysis
ESI-MS of the protected glycopolymer 121 shows a very complex spectrum in positive mode, consisting of multiple singly, doubly and triply charged ion distributions.
102
Ion mobility mass spectrometry was used to separate the glycopolymer according to its charge state and from contaminants and noise.
ESI-IM-MS (Figure 6.5) separates the glycopolymers into distinct singly, doubly and triply charged bands, each containing several ion series with the galactose acrylate repeat unit (C15H22O7, 314.137/z). The distance between adjacent oligomers of the same type is 314.1 m/z units for singly charged, 314.1/ 2=157.1 m/z units for doubly charged, and 314.1/3=104.6 m/z units for the triply charged ions (Figure 6.5b-d, respectively).
103
Scheme 6.1 Fragmentation pathways of glycopolymer “121”. Charge-remote H rearrangements in sodiated oligomers lead to (a) 2x(CH3)2CO (acetone) losses from the galactose ring followed by CO loss to form a five-membered ring lactone. (b) Cross ring fragmentation leading to loss of 200 Da and followed by (c) CO elimination (overall loss of 228 Da). (d) Expulsion of the galactose pendant (242 Da), resulting in a fragment with acid end group.
104
In analogy to the MALDI results, each ESI-IM-MS band contains mainly four major distributions corresponding to ions with end groups of 101 Da, 134 Da, 204 Da and 314 Da; the later distribution arises from oligomers that lost both ends, as shown in
Figure 6.4.
6.3.3 ESI-MS2 analysis
The MS2 characteristics of oligomers with 134-Da and 101-Da end groups were examined using either Q/ToF (MS2) or QIT multistage (MSn) mass spectrometry in order to gain more information on the fragmentation pathways of the linear chains. Figure 6.6 shows the Q/ToF MS2 spectra of the singly charged 4-mers with101-Da (m/z 1380.5 Da) and 134-Da (m/z 1413.5 Da) end groups. The ions undergo competitive and consecutive rearrangement decompositions at either the isopropylidene protecting groups or the glycopolymer side chain ester groups, giving rise to neutral losses of acetone (58 Da), carbon monoxide (28Da), twice C5H8O2 (200 Da) and the dehydrated, fully protected galactose pendant (C12H18O5, 242 Da), Scheme 6.1.
The ESI-MS2 fragmentation pathways of the sodiated ions at m/z 1380.5 and
1413.5 are similar to those generated by MALDI-MS2; in both cases, 1,5-H rearrangements over the ester groups predominate. Furthermore, in both cases random backbone cleavages within the acrylate connectivity occur as well, similarity to sodiated poly(methyl acrylate) or silverated polystyrene oligomers. These dissociations proceed through homolytic backbone C–C bond cleavages followed by backbiting rearrangements, which ultimately lead to internal J and K fragments, Scheme 6.2.167
Chains with one branching unit (C8H12O4S, 204 Da, see Figure 6.4), were also examined by Q/ToF-MS2 (Figure 6.7c). They show a very similar fragmentation pattern
105 as the linear architectures with 101-Da and 134-Da end groups. The same internal fragments are observed independent of the end group identity, strongly suggesting that the branching unit is attached at the end of the polyacrylate chain, as shown in Figure
6.4.
Under QIT-MSn conditions, the chains with 101-Da and 134-Da end groups form minor fragments arising by elimination of repeat units (monomer) as well as one or both
+ end groups. This is exemplified for [R4+134+Na] (m/z 1413.5) in Figure 6.7. The predominant fragments are the rearrangement elimination products from the sugar pendants (Scheme 6.1), as was also observed upon Q/ToF-MS2. Sequential CAD (MS3) and (MS4) on the major products generated by loss of one or two acetone units does not change significantly the fragmentation pattern.
106
Scheme 6.2 Backbiting in the acrylic radical ions emerging after random homolytic C–C bond cleavages in the Na+-cationized galactose acrylate glycopolymer chains. R
• abbreviates the galactose pendent (C13H19O7). α and ω designate the end groups; bn
• are the acrylic radical ions containing the α and zn the acrylic radical ions containing the
• • ω end group. Backbiting gives rise to terminal an and yn fragments from bn and zn , respectively, as well as to the internal fragments Jn (m/z 323, 637, 951) and Jna (m/z
521, 835, 1149); the latter are generated after the loss of 2 acetone moieties from Jn.
The internal fragment ions Kn (m/z 337, 651, 965) are observed as well. The terminal fragments (an and yn) are below noise level, suggesting that they dissociate consecutively via the same mechanism.
107
108
Figure 6. 5 a) 2-D IM-MS plot (m/z vs. drift time) of the protected galactose acrylate dissolved in MeOH: THF (7:3, v/v %) at 0.01 mg/mL. b) Mass spectra extracted from the
IM regions of (a) singly charged, c) doubly charged d) triply charged ions in the 2-D diagram.
Multistage tandem mass spectra (MSn) were also acquired for the same 4-mer with a 204-Da end group (m/z =1483.5 Da), c.f. Figure 6.8. Unfortunately, they are less informative than the MS2 spectrum acquired on the Q/ToF instrument, c.f. Figure 6.6 c; on the QIT, the fragments produced by neutral losses from the pendants suppress the backbone cleavages required to characterize the chain connectivity and locate the branching unit. Furthermore, the low-mass cutoff of QIT mass analyzers makes it impossible to detect and analyze the low-mass region, where internal fragments that reveal connectivity are present.
109
Finally, the chains with 314-Da end groups were studied by multistage MSn
(Figure 6.9). These chains contain a linear polyacrylate backbone with protected galactose ester pendants, c.f. Figure 6.4. Simpler polyacrylates with such inert end groups undergo extensively backbone bond cleavages.167 Here, these are suppressed by the neutral losses from the glycan side chains,190 which also dominated the MSn spectra of oligomers with 101-Da and 134-Da end groups (vide supra).
Overall, the MS2 and MSn fragmentation results reveal useful information about the side chain and end group functionalities, based on the neutral losses observed, but do not reveal any information about the chain connectivity, especially for the ions with the branching unit.
110
Figure 6. 6 Q/ToF-ESI-MS2 spectra of sodiated singly charged glycopolymer ions with
(a) 101-Da end groups (m/z 1380.48), (b) 134-Da end groups (m/z 1413.55) and (c)
204-Da end groups (m/z 1483.50); the numbers on top of the peaks give the monoisotopic m/z ratio (in black) and the mass of the neutral loss(es) in Da (in color).
111
6.3.4 Structural information from collision cross-sections
6.3.4.1 Derivation of collision cross-sections from traveling wave IM-MS experiments
Oligomers with 134-Da and 101-Da end groups can only have a linear structure
(cf. Figure 6.4), in contrast, oligomers with 204-Da end groups contain one branching unit which can result in a linear or branched architecture, depending on whether the branch is incorporated at a terminal or internal position, respectively, within the polyacrylate chain. Comparing the CCSs of these differently functionalized chains could shed more light on this issue.
The collision cross-section values of the glycopolymer oligomers with 101-,134- and 204-Da end groups can be derived from the corresponding drift times in the IM region. The CCS is an important physical property that can provide information about the ion’s size and shape. With traveling-wave IM-MS there is no direct relationship between the measured drift time, 푡퐷, and the collision cross-section, Ω, as mentioned in chapter 2.3. As a result, the collision cross-sections are derived by using standards of known Ω such as polyalanine (PA) following a procedure that has been described in the literature and is briefly outlined in chapter 2.3.53,55
112
Figure 6. 7 (a) QIT-ESI-MS2 mass spectrum of sodiated 121 glycopolymer with 134-Da end groups (m/z 1413.5); (b) MS3 mass spectrum of the fragment at m/z 1355.6 formed by acetone loss from 1413.5; (c) MS4 mass spectrum of the fragment at m/z 1297.5 formed by acetone loss from m/z 1355.5. The numbers on top of the peaks give the monoisotopic m/z ratio (in black) and the mass of the neutral loss(es) in Da (in color);
−58, 28 and −242 indicate losses of acetone, CO and galactose pendant, respectively, see scheme 6.1 for more details.
113
114
Figure 6. 8 (a) QIT-ESI-MS2 mass spectrum of sodiated 121 glycopolymer 4-mer with
204-Da end groups (m/z 1483.6); (b) MS3 mass spectrum of m/z 1425.6, formed by acetone loss from 1483.6; (c) MS4 mass spectrum of m/z 1367.4, formed by acetone loss from m/z 1425.6. The numbers on top of the peaks give the monoisotopic m/z ratio
(in black) and the mass of the neutral loss(es) in Da (in red).
115
Figure 6. 9 (a) QIT-ESI-MS2 mass spectrum of the sodiated 4-mer from glycopolymer
121 with 314-Da end groups (m/z 1593.7); (b) MS3 mass spectrum of m/z 1535.6, formed by acetone loss from 1593.7; (c) MS4 mass spectrum of m/z 1477.6, formed by acetone loss from m/z 1535.6. The numbers on top of the peaks give the monoisotopic m/z ratio (in black) and the mass of the neutral loss(es) in Da (in red); for example, −58 and −242 indicate losses of acetone and the galactose pendant, respectively. For more details see Scheme 6.1.
116
The collision cross-sections of polyalanine ions (calibrants used in this dissertation) have been measured by drift -tube IM-MS, where a constant electric field is applied to the ion mobility cell, thus allowing for a direct correction of tD and Ω. The calibrant values were determined using He buffered gas in the IM cell. Experimental collision cross-sections can be acquired by traveling-wave IM-MS by calibrating the drift time scale with these known Ω ions. Singly and doubly charged polyalanine oligomers were used to construct the calibration carve shown in Figure 6.10, which was obtained by plotting the corrected collision cross-sections of these calibrant ions (Ω`) against their corrected drift times (tD`), measured at the same traveling wave velocity, traveling wave height, and ion mobility gas flow setting used for the glycopolymer ions.
Figure 6.10 Plot of corrected drift times (arrival times) against corrected published cross- sections for the +1 and +2 ions formed by ESI of polyalanine. Drift times were measured at a wave velocity of 350 m/s and a wave height of 11 V.
117
This calibration curve (Figure 6.10) was used to determine the CCS of several charge states of the protected glycopolymer ions with different end groups. Table 6.1 shows the drift times measured in this study and experimental CCSs deduced for the singly charged glycopolymer ions with different end groups.
The experimental CCS of the doubly and triply sodiated oligomers with 101-Da,
134-Da and 204-Da end groups were also derived using the calibration carve in Figure
6.10, Table 6.2 shows the drift times and the experimental CCSs of these oligomers.
The experimental CCS values of the oligomers with one branching unit (EG=204
Da) (linear or branched architecture) were compared with the CCS values of the oligomers known to have a linear architecture (EG=101 Da or 134 Da) in order to reveal some information about the possible architecture of the former (EG=204 Da) oligomers.
118
Table 6. 1. Experimental collision cross-sections of singly charged glycopolymer ions
with different end groups derived using the calibration plot of Figure 6.10. z n MW m/z tD tD' Ω` reduced Ω
(ms) (ms) (Å2) mass (Å2)
EG=314 Da
1 2 628.2792 651.269 2.44 2.404 898.7459 5.178 173.56
1 3 942.4232 965.413 4.24 4.196 1203.9135 5.216 230.83
1 4 1256.5942 1279.584 6.32 6.269 1486.3155 5.234 283.95
1 5 1570.7262 1593.716 8.75 8.694 1764.4762 5.246 336.35
EG=69 Da
1 2 697.3212 720.311 2.89 2.852 983.0961 5.189 189.45
1 3 1011.4792 1034.469 4.96 4.915 1308.0255 5.221 250.55
1 4 1325.5892 1348.579 6.95 6.898 1562.7522 5.237 298.38
1 5 1639.7522 1662.742 9.39 9.333 1831.3695 5.248 348.98
EG=101 Da
1 3 1043.4632 1066.453 4.24 4.1939 1203.5768 5.223 230.45
1 4 1357.5292 1380.519 6.32 6.268 1486.0727 5.239 283.67
1 5 1671.7612 1694.751 8.75 8.692 1764.289 5.2487 336.14
EG=134 Da
1 2 762.3182 785.308 3.16 3.121 1030.5969 5.198 198.27
1 3 1076.4702 1099.46 5.23 5.183 1345.0672 5.225 257.43
1 4 1390.6092 1413.599 7.4 7.345 1615.3069 5.239 308.27
1 5 1704.7942 1727.784 9.93 9.87139 1886.1259 5.249 359.29
119
EG=204 Da
1 2 832.3272 855.317 3.52 3.478 1091.0903 5.206 209.6
1 3 1146.4582 1169.448 5.51 5.462 1382.5281 5.229 264.39
1 4 1460.6152 1483.605 7.76 7.706 1656.226 5.242 315.93
The CCSs for the singly charged ions with EG= 134 Da (linear architecture) vary
between 198.3 Ų and 359.3 Ų for oligomers with 2-5 repeat units; for the triply charged
ions with the same EGs, the CCS range is between 548.5 to 635.5 Ų for oligomers with
12-15 repeat units. On the other hand, the CCSs for the singly charged ions with EG=
204 Da lie between 209.6 Ų and 316.0 Ų for oligomers with 2-4 repeat units, while the
CCS range for the triply charged ions with the same EGs is between 548.4 to 635.4 Ų
for oligomers with 12-15 repeat units, see Table 6.1 and 6.2.
Surprisingly, the drift times and experimental CCSs of the linear oligomers, for
example the triply charged oligomers with EG=134 Da are very close to the drift times
and experimental CCSs of the oligomers with EG=204 Da and the same number of
repeat units (Figure 6.11 and Table 6.1 and 6.2). It can be concluded based on this
CCS comparison that the oligomers with one branching unit and EG=204 Da have most
likely a linear architecture and that the branching unit is attached at the polyacrylate
chain end.
120
Table 6. 2. Experimental collision cross-sections of doubly and triply charged
glycopolymer ions with different end groups derived using the calibration plot of Figure
6.10. z n MW m/z tD tD' Ω` reduced Ω
(ms) (ms) (Å2) mass (Å2)
EG=101 Da
2 7 2300.0164 1172.998 3.25 3.203 1044.5886 5.261 397.14
2 8 2614.1604 1330.07 3.79 3.739 1133.1234 5.264 430.49
2 9 2928.3464 1487.163 4.51 4.456 1242.41923 5.267 471.75
2 10 3242.5024 1644.241 5.14 5.083 1331.3276 5.269 505.27
2 11 3556.6424 1801.311 5.87 5.81 1428.1266 5.272 541.81
EG=134 Da
2 7 2333.0864 1189.533 3.25 3.201 1044.5305 5.261 397.09
2 8 2647.2084 1346.594 3.88 3.828 1147.3079 5.265 435.85
2 9 2961.3464 1503.663 4.51 4.455 1242.3752 5.268 471.7
2 10 3275.5164 1660.748 5.41 5.353 1367.9465 5.269 519.15
2 11 3589.6564 1817.818 5.87 5.809 1428.0913 5.271 541.77
EG=204 Da
2 7 2403.0624 1224.521 3.43 3.3807 1074.8327 5.262 408.54
2 8 2717.1724 1381.576 3.97 3.9176 1161.2810 5.265 441.1
2 9 3031.3884 1538.684 4.69 4.6347 1268.3778 5.2682 481.53
2 10 3345.5964 1695.788 5.23 5.1712 1343.5260 5.27 509.84
2 11 3659.6184 1852.799 6.05 5.9893 1451.0690 5.272 550.45
121
EG=101 Da
3 12 3870.6966 1313.222 2.71 2.6589 947.5559 5.273 539.06
3 13 4184.9946 1417.988 3.07 3.0169 1012.4995 5.275 575.85
3 14 4499.1186 1522.696 3.34 3.2849 1058.7593 5.276 602.02
3 15 4813.2936 1627.421 3.61 3.5531 1103.2675 5.277 627.19
EG=134 Da
3 12 3903.9186 1324.296 2.8 2.7487 964.2153 5.273 548.52
3 13 4218.0966 1429.022 3.07 3.0166 1012.4632 5.275 575.81
3 14 4532.1186 1533.696 3.34 3.2848 1058.7257 5.276 601.98
3 15 4846.3776 1638.449 3.7 3.6429 1117.8154 5.277 635.46
EG=204 Da
3 12 3973.7736 1347.581 2.8 2.7482 964.1326 5.274 548.44
3 13 4288.0836 1452.351 3.16 3.1063 1028.1292 5.275 584.69
3 14 4602.1746 1557.048 3.43 3.3744 1073.7815 5.276 610.52
3 15 4916.3046 1661.758 3.7 3.6425 1117.7502 5.27 635.39
122
(a)
(b)
Figure 6. 11. Plot of experimental collision cross-section vs. m/z for sodiated 121 glycopolymer oligomers with 101-Da, 134-Da and 204-Da end groups (a) +3, (b) +2.
123
6.4 Conclusion
MALDI and ESI mass spectrometry coupled with tandem and multistage MS
(MS2 and MSn) experiments confirmed the successful synthesis of the studied glycol- polymer via the RAFT mechanism and revealed information about the glycopolymer end groups. The MS2 and MSn experiments show that polyacrylate based glycopolymer ions dissociate mainly at the galactose pendant that confirm the identity and protected nature of the glycan side chains. No backbone (main chain) cleavages occur, however, to ascertain branching architecture. Fortunately, valuable architectural information could be obtained by ion mobility MS, which strongly suggested that the branching unit is incorporated at the end of the chain, giving rise to a linear-like structure.
124
CHAPTER VII
MASS SPECTROMETRY CHARACTERIZATION OF TREHALOSE GLYCOPOLYMER
AND ITS INSULIN CONJUGATE
7.1 Introduction
Peptide and protein therapeutics are extensively used for treatment of many diseases.168,169 However, their instability during storage and delivery might affect their effectiveness and reduces in vivo half-life as well as increases costs due to the need of refrigeration of such protein drugs.170,171 Therefore, many excipients such as amino acids,172 sugars,173 salts,174 surfactants175 and polymers115,176 have been developed and used to improve protein drugs’ stability and prevent their degradation.
Protein−polymer conjugates are biomaterials showing a unique combination of properties derived from the biological compounds and the synthetic materials, that can be tuned to produce the desired properties.177-179 Polymers such as polyoxazolines,180 poly(N-(2-hydroxypropyl) methacrylamide) (pHPMA),181 hydroxyethyl starch (HES)182 and polyethylene glycol (PEG)183 have been successfully conjugated to proteins drugs to increase their vivo half-life. PEGylation or covalent attachment of PEG to protein drugs is the most widely used method to solve pharmacokinetic challenges. PEG has high water solubility and non-toxic properties.110,109 PEGylation shows improvement in the pharmacokinetic behavior of the drug including stability, solubility and decreased immunogenicity.111
125
PEGylated proteins have been commercially available since the first approval of such drugs by FDA in 1990.184,185 However, PEGylation does not necessarily improve storage stability of protein drugs toward environmental stressors.168 As a result, protein drugs are still required to be refrigerated even in the presence of PEG or other stabilizing additives.186 Denaturation of protein drugs that result from storage environment may cause life-threatening events; therefore, there is still an enormous need in developing new nontoxic protein polymer conjugates to stabilize proteins during the storage stage and improve their in vivo half-life. Recently, it has been reported by
Maynard and co-workers150,170 that glycopolymers with trehalose side chains187,188 can provide both storage stability as well as increasing in vivo availability.150,170,187,188
Scheme 7. 1(a) Synthesis of a trehalose glycopolymer by RAFT polymerization using
AIBN as initiator and a trithiocarbonate chain transfer agent (CTA). (b) Acetylation of the glycopolymer by using acetic anhydride and pyridine catalyst.
126
Trehalose is a non-reducing disaccharide consisting of two glucose units in a
α,α-1,1-glycosidic linkage, it is a stable sugar with low reactivity toward most proteins.150,188 FDA has classified trehalose as “generally regarded as safe” (GRAS) and approved it as an excipient in four pharmaceutical products.2,187 Conjugating a trehalose polymer to a therapeutic protein enhances both the thermal stability and pharmacokinetic properties of the protein. Trehalose glycopolymers can be used as excipients to stabilize proteins such as (glucose oxidase (GOx), horseradish peroxidase
(HRP), and β-galactosidase against environmental stressors such as lyophilization and heat.150,189 The characterization of synthetic glycopolymers and glycoconjugates is a very challenging procedure; however, the development of new analytical methods over the last decades have made it possible to identify such materials. Mass spectrometry
(MS) is a powerful and promising method for the analysis of glycopolymers and glycoconjugates, as it offers molecular structure and connectivity information with high sensitivity in relatively short time compared to other methods. Tandem mass spectrometry (MS2) and multistage mass spectrometry (MSn) via collisionally activated dissociation (CAD) have been reported to significantly augment the tools available for the identification of glycopolymers by enabling the identification of primary structure and/ or substructures such as side chains and end groups.190
This chapter reports the characterization of a trehalose glycopolymer (Scheme
7.1) and its conjugated form with insulin by MALDI-MS, ESI-MS and tandem mass spectrometry (MS2).
7.2 Experimental
127
7.2.1 Materials
All chemicals were purchased from Sigma-Aldrich and Fisher Scientific and were used without any additional purification unless noted elsewhere.170 Trehalose was purchased from The Healthy Essential Management Corporation (Houston, TX).
Recombinant human insulin was purchased from Sigma-Aldrich. Styrenyl acetal trehalose monomer and trehalose glycopolymer were prepared using previously reported techniques.150 Briefly, a polystyrene backbone with an acetal linkage to trehalose was synthesized by reversible addition− fragmentation chain transfer (RAFT) polymerization using an aldehyde functionalized chain transfer agent and 2,2-azobis(2- methylpropionitrile) (AIBN) as an initiator, as described in Scheme 7.1a. The polymer had an average molecular weight (Mn) of 9.9 kDa and a polydispersity of 1.10 according to gel permeation chromatography (GPC) analysis. The insulin-trehalose glycopolymer conjugate was prepared by linking the aldehyde-functionalized glycopolymer to the lysine free amine group of the protein by reductive amination (Scheme 7.2).
Scheme 7. 2 Conjugation procedure of trehalose glycopolymer to insulin by reductive amination.
128
7.2.2 MALDI-MS experiments
Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) experiments were carried out on a Bruker UltraFlex III tandem time-of-flight (ToF/ToF) mass spectrometer (Bruker Daltonics, Billerica, MA), equipped with a Nd: YAG laser emitting at a wavelength of 355 nm. The unprotected trehalose glycopolmer was hard to analyze by MALDI. Therefore, the glycopolymer was derivatized by acetylation of the free OH groups in order to increase volatility and thermal stability by reducing intermolecular forces. Although derivatization increases the molecular weight, it also reduces intermolecular hydrogen bonding by converting the polar hydroxy into less polar acetoxy groups.191 Acetylation was performed as indicated in Scheme 7.1b. The protected glycopolymer was used to confirm the successful synthesis. The acetylated glycopolymer was dissolved in THF at a final concentration of 10 mg mL-1; trans-2-[3-(4- tert-Butylphenyl)-2-methyl-2-propenylidene] malononitrile (DCTB) (20 mg mL-1) served as matrix and NaTFA (10 mg mL-1) as the cationizing salt. The matrix and salt solutions were mixed in the ratio 10:1 (v/v) and the three-layer sandwich method was used to deposit the glycopolymer onto the MALDI plate. The conjugate could only be analyzed after treatment with formic acid (FA), which cleaves the polymer, leaving a 106-Da linker attached to insulin at the conjugation site(s). For this, the conjugate was dissolved in
-1 H2O: ACN 1:1 (v/v) at 10 mg mL and 10 % FA was added to this solution, which was subsequently mixed with a solution of either DTT or TCEP for disulfide bond cleavages.
2,5-Dihydroxyacetophenone (DHAP) was used as a matrix for the conjugate.
129
7.2.3 ESI-MS experiments
For ESI-MS analysis, the conjugate was diluted with H2O: ACN 1:1 (v/v) at a final concentration of 0.01-0.05 mg mL-1, and the solution was injected into the Waters
Synapt G1 HDMS quadrupole/time-of-flight (Q/ToF) mass spectrometer (Waters,
Milford, MA), after treating it with FA and DTT 10 mM, DTT to reduce the disulfide bonds and release chain A and chain B of insulin including the 106-Da linker.
Instrument parameters were adjusted as follows: ESI capillary voltage, 3.5 kV; sample
-1 cone voltage, 30 V; extraction cone voltage, 3.2 V; desolvation gas flow, 550 L h (N2); trap collision energy (CE), 6.0 eV; transfer CE, 4.0 eV; trap gas flow, 1.5 mL min-1 (Ar); sample flow rate, 10 μL min-1; source temperature, 90 °C; desolvation temperature, 150
°C; for tandem mass analysis trap CE was adjusted to 20 V.
7.3 Results and discussion
7.3.1 MALDI-MS analysis of the trehalose glycopolymer and its insulin conjugate
The MALDI-MS spectrum of the glycopolymer shows several distributions with different end groups (EGs), having the composition [Mn+EGs+Na]+. The mass difference between two consecutive peaks within each distribution is 708 Da, which matches the mass of the acetylated trehalose PS unit (C33H40O17). End groups with masses of 205 Da (C7H11NS3), 708 Da (C33H40O17), 178 Da (C10H10O3), 69 Da (C4H7N),
138 Da (C3H6S3) and 354 Da (C20H18O6) are observed, in accord with the RAFT mechanism, cf. Figure 7.1. The most abundant distribution contains a trithio-carbonate end from the RAFT CTA and an isobutyronitrile end from the AIBN initiator.
130
The observation of a trithiocarbonate end group in the MALDI spectrum confirms that this substituent was not reduced under the reductive amination conditions used in the conjugation step, as had been suggested by UV−vis spectroscopy previously.170
The detection of the intact insulin-trehalose glycopolymer conjugate was very challenging by MALDI. The MALDI mass spectrum of the conjugated sample exhibited low-intensity peaks at high mass range, however, it was hard to confirm their identity.
Therefore, the conjugate was treated with formic acid (FA) in order to cleave off the phenyl ester bond between the glycopolymer and protein (Scheme 7.2), leaving a small,
106-Da linker. The MALDI mass spectrum of the conjugate after treatment with FA shows ions with 106 and 212 Da greater mass than insulin, which were not observed for native insulin, Figure 7.2d, consistent with one or two glycopolymers chains being attached to insulin. Furthermore, the cleaved glycopolymer is successfully observed in the conjugate spectrum with end groups equal to the remaining moieties after bond cleavages with FA, having a mass of 210 Da (C6H10O2S3), cf. Figure 7.2c. These findings corroborate the successful conjugation by reductive amination (Scheme 7.2) and indicate the formation of mono- and di-substituted conjugates.
131
Figure 7. 1 (a) MALDI spectrum of acetylated trehalose glycopolymer; all ions are
+ sodiated species with the composition [Rn + EGs + Na] , where R is the repeat unit
(C33H40O17, 708.226 Da). (b) Expanded view of the highlighted m/z region of the trehalose glycopolymer spectrum. (c) Proposed structures of the polymer end groups observed.
132
Figure 7. 2 Comparison of the MALDI mass spectra of (a) insulin-trehalose glycopolymer conjugate after treatment by FA and before disulfide reduction and (b) native insulin using DHAP matrix.
133
Figure 7. 3. Comparison of the MALDI mass spectra of chain B of native insulin (top) and conjugated insulin (bottom). The spectrum in the top shows the protonated chain B while the spectrum in the bottom shows the protonated ion in addition to another ion that is 106 Da higher than chain B corresponding to the benzaldehyde linker.
The MALDI mass spectra in Figure 7.3 compare native insulin (top) and the conjugated insulin after treatment with both formic acid and TCEP (bottom). The protonated chain B (3428.6 Da) is observed in both spectra, however, the spectrum of the conjugate shows one additional peak at (3534.7 Da) corresponding to chain B plus the 106 Da benzaldehyde linker. This result suggests that the lysine residue of chain B is a major conjugation site (vide infra).
134
7.3.2 ESI-MS and MS2 characterization of the insulin-trehalose glycopolymer conjugate
The ESI-MS spectra of the conjugate (Figure 7.4 and 7.5) show multiply charged ions (+6, +5, +4 and +3) that corroborate the MALDI results. Using a rather concentrated solution of the conjugate and 10% formic acid yields a spectrum showing mainly insulin ions and ions of insulin plus the linker (106 Da). Again, two repeat units of
106 Da are detected (Figure 7.4) confirming that up to two glycopolymers chains are attached to insulin. If a dilute solution of the conjugate and 0.1% formic acid are used, the cleaved glycopolymer with EG mass of 210 Da (C6H10O2S3) is observed (Figure
7.5). After disulfide bond cleavages with DTT, both chain A and chain B show one additional peak corresponding to species with a 106-Da greater mass (Figure 7.6). Both chain A and chain B are detected in modified form, indicating that the glycopolymer is attached to either (or both) of these chains.
These results reveal that the glycopolmer was conjugated to the N-terminal of chain A (GlyA1) and to either the N-terminal (PheB1) or lysine (LysB29) of chain B, which are the possible attachment sites based on the conjugation chemistry used. MS2 experiments were performed on the +5 -charge state of native and conjugated chain B of insulin to determine which amino acid residue was attached to the glycopolymer
2 chain. The MS spectrum of the conjugate exhibits a y3 +106 (m/z 451.20) fragment, which conclusively confirms that the linker (and, hence, the polymer) was attached to
LysB29 adjacent to the C-terminus (Figure 7.7).
135
Figure 7. 4. ESI-MS spectrum of the trehalose glycopolymer-insulin conjugate; the expanded area shows the cleaved glycopolymer with the remaining end of the RAFT
-1 chain transfer agent. The sample was dissolved in H2O: ACN at 0.05 mg mL + 10% formic acid (v/v %).
136
Figure 7. 5. ESI-MS spectrum of the trehalose glycopolymer-insulin conjugate; the expanded area shows the cleaved glycopolymer with the remaining end of the RAFT
-1 chain transfer agent. The sample was dissolved in H2O: ACN at 0.01 mg mL + 0.1% formic acid (v/v %).
137
Figure 7. 6 (a) ESI-MS spectrum of the trehalose glycopolymer-insulin conjugate after acid treatment and disulfide reduction. (b) Expanded area of the m/z region of +3 chain
A (m/z 795.0513) which shows one site modification. (c) Expanded area of the m/z region of +5 chain B (m/z 686.575) which shows also one site modification. The sample was dissolved in H2O: ACN at 0.05 mg/mL + 10 mM DTT +10% formic acid (v/v %).
138
Figure 7. 7. ESI-MS2 spectrum of insulin-trehalose glycopolymer conjugate (top) and native insulin (bottom). Fragmentation of chain B in charge state +5 from the conjugate
(m/z 707.64) and insulin (m/z 686.45) mainly gives rise to bn and yn fragments.
7.4 Conclusions
The trehalose glycopolymer is a promising alternative to PEG because of its ability to enhance in vivo bioavailability and the environmental stability of the therapeutic protein, as recently shown by Maynard and co-workers.170 The same group also reported that trehalose glycopolymers can be used as excipients to stabilize different proteins against stressors such as freeze, heat and acidic conditions.
139
The analytical research reported in this dissertation highlights the effectiveness of mass spectrometry methodologies (MALDI-MS, ESI-MS and MS2) for the complete characterization of trehalose glycopolymers. The obtained results confirmed the successful synthesis of the glycopolymer and its conjugate. The MS2 analysis further indicated that the trehalose glycopolymer was conjugated to GlyA1 and LysB29, consistent with earlier literature reports that these two amine sites are much more reactive toward conjugation than PheB1.192 Overall, MS analysis provided rapid structural information about the trehalose glycopolymer and its insulin conjugate and revealed both the degree as well as sites of conjugation.
140
CHAPTER VII
SUMMARY
Mass spectrometry (MS) was used in this dissertation to characterize synthetic polymers as well as complex biomolecules including peptides, saccharides, and bioconjugates. MS is a fast and sensitive analytical method that requires very little amount of the sample under investigation. This work has shown the capability of this method for the characterization of the complex samples by using different ionization methods, such as MALDI and ESI, different mass analyzers, such as ToF/ToF, Q/ToF and QIT, and different activation methods to cause ion fragmentation. Furthermore, combining mass spectrometry analysis with ion mobility separation provided an additional dimension to gain information about a sample’s shape and possible architecture.
Chapter IV reported the successful characterization of isomeric biodegradable polyesters by MALDI-MS, ESI-MS and tandem mass spectrometry (MS2) fragmentation.
With these methods, it was possible to elucidate the composition, end groups, and chain sequence of poly(propylene maleate) PPM and poly(propylene fumarate) PPF copolymers. Meanwhile, ion mobility mass spectrometry (IM-MS) differentiated between the isomeric PPM and PPF copolyesters and probed the extent and efficiency of PPM to
PPF isomerization.
141
Chapter V described the full characterization of the alanine-rich peptides AQK18 and GpAQK18 and their PEGylated forms AQK18-PEG and GpAQK18-PEG. ESI and
MALDI with tandem mass spectrometry confirmed the peptides sequences and the successful conjugation of the PEG at the C-terminus. IM-MS data revealed the existence of two conformers (random coil and helical structure) for both the poly- peptides and the conjugates. Simulations confirmed the identity of the observed architectures and highlighted the stabilization of helical conformations by the polymer chain.
Chapter VI reported the structural characterization of a glycopolymer containing a controlling branch unit. Tandem MS, multistage MS and ion mobility MS provided valuable information about the glycopolymer composition, possible end groups and architecture.
Chapter VII discussed the analysis of a trehalose glycopolymer and its insulin conjugate and described their full characterization before and after conjugation. Further,
ESI-MS2 allowed to determine the degree and sites of conjugation after acid treatment and disulfide bond cleavages.
In conclusion, this dissertation demonstrated the strength of multidimensional mass spectrometry methods for the full characterization of complex molecules including isomeric polymers, glycopolymers and bioconjugates.
142
REFRENCES
1. Koenig JL. Mass spectrometry of polymers. In: Spectroscopy of Polymers. ;
1999:441-480. doi:10.1016/B978-044410031-3/50010-4.
2. Hanton SD, Maldi F. Mass Spectrometry of Polymers and Polymer Surfaces.
2001.
3. Van Riper SK, de Jong EP, Carlis J V., Griffin TJ. Mass spectrometry-based
proteomics. Adv Exp Med Biol. 2013;990(March):1-35. doi:10.1007/978-94-007-
5896-4_1.
4. Finehout EJ, Lee KH. An Introduction to Mass Spectrometry Applications in
Biological Research. Biochem Mol Biol Educ. 2004;32(2):93-100.
doi:10.1002/bmb.2004.494032020331.
5. El-Aneed A, Cohen A, Banoub J. Mass Spectrometry, Review of the Basics:
Electrospray, MALDI, and Commonly Used Mass Analyzers. Appl Spectrosc Rev.
2009;44(3):210-230. doi:10.1080/05704920902717872.
6. Lanucara F, Holman SW, Gray CJ, Eyers CE. The power of ion mobility-mass
spectrometry for structural characterization and the study of conformational
dynamics. Nat Chem. 2014;6(4):281-294. doi:10.1038/nchem.1889.
7. Deery MJ, Stimson E, Chappell CG. Size exclusion chromatography/mass
spectrometry applied to the analysis of polysaccharides. Rapid Commun Mass
143
Spectrom. 2001;15(23):2273-2283. doi:10.1002/rcm.458.
8. Honour JW. Gas chromatography-mass spectrometry. Methods Mol Biol.
2006;324:53-74. doi:10.1385/1-59259-986-9:53.
9. Makarov A, Scigelova M. Coupling liquid chromatography to Orbitrap mass
spectrometry. J Chromatogr A. 2010;1217(25):3938-3945.
doi:10.1016/j.chroma.2010.02.022.
10. Monton MRN, Soga T. Metabolome analysis by capillary electrophoresis-mass
spectrometry. J Chromatogr A. 2007;1168(1-2):237-246.
doi:10.1016/j.chroma.2007.02.065.
11. Hoffmann E De, Stroobant V. Mass Spectrometry Principles and Applications.;
2007. doi:10.1002/9783527654703.ch11.
12. Dass C. Fundamentals of Contemporary Mass Spectrometry.; 2006.
doi:10.1002/9780470118498.
13. Karas M, Hillenkamp F. Laser desorption ionization of proteins with molecular
masses exceeding 10,000 daltons. Anal Chem. 1988;60(20):2299-2301.
doi:10.1021/ac00171a028.
14. Dewald HD. Electrospray Ionization Mass Spectrometry: Fundamentals,
Instrumentation and Applications (ed. Cole, Richard B.). J Chem Educ.
1999;76(1):33. doi:10.1021/ed076p33.1.
15. Shvartsburg AA, Smith RD. Fundamentals of traveling wave ion mobility
spectrometry. Anal Chem. 2008;80(24):9689-9699. doi:10.1021/ac8016295.
144
16. Waters synapt high definition mass spectrometry system operator’s guide, Waters
Corporation, 2007. 2007;18(6):1146-1159. doi:10.1016/j.jasms.2007.03.004.
17. Pringle SD, Giles K, Wildgoose JL, et al. An investigation of the mobility
separation of some peptide and protein ions using a new hybrid
quadrupole/travelling wave IMS/oa-ToF instrument. Int J Mass Spectrom.
2007;261(1):1-12. doi:10.1016/j.ijms.2006.07.021.
18. Smith DP, Knapman TW, Campuzano I, et al. Deciphering drift time
measurements from travelling wave ion mobility spectrometry-mass spectrometry
studies. Eur J Mass Spectrom (Chichester, Eng). 2009;15(2):113-130.
doi:10.1255/ejms.947.
19. Sallam S, Luo Y, Becker ML, Wesdemotis C. Multidimensional mass spectrometry
characterization of isomeric biodegradable polyesters. 2017.
doi:10.1177/1469066717711401.
20. Liu Y, Lee J, Mansfield KM, et al. Trehalose Glycopolymer Enhances Both
Solution Stability and Pharmacokinetics of a Therapeutic Protein. Bioconjug
Chem. 2017;28(3). doi:10.1021/acs.bioconjchem.6b00659.
21. Alalwiat AA. Mass Spectrometry Methods for the Analysis of Biodegradable
Hybrid Materials. 2015.
http://rave.ohiolink.edu/etdc/view?acc_num=akron1434894354. Accessed June
19, 2017.
22. Fenn JB, Mann M, Meng CK, Wong SF, Whitehouse CM. Electrospray ionization
for mass spectrometry of large biomolecules. Science (80- ). 1989;246(4926):64-
145
71. doi:10.1126/science.2675315.
23. Gruendling T, Weidner S, Falkenhagen J, Barner-Kowollik C. Mass spectrometry
in polymer chemistry: a state-of-the-art up-date. Polym Chem. 2010;1(5):599.
doi:10.1039/b9py00347a.
24. Dole M, Mack LL, Hines RL, Mobley RC, Ferguson LD, Alice MB. Molecular
Beams of Macroions. J Chem Phys. 1968;49(5):2240-2249.
doi:10.1063/1.1670391.
25. Grayson MA. John Bennett Fenn: A Curious Road to the Prize. J Am Soc Mass
Spectrom. 2011;22(8):1301-1308. doi:10.1007/s13361-011-0136-6.
26. Gaskell SJ. Electrospray: Principles and practice. J Mass Spectrom.
1997;32(7):677-688. doi:10.1002/(SICI)1096-9888(199707)32:7<677::AID-
JMS536>3.0.CO;2-G.
27. Esquire ESI Operator‘s Manual, Volume 1, Version 4.0, BrukerDaltonics.
http://www.visionaustralia.org/living-with-low-vision/learning-to-live-
independently/using-technology-and-computers/computer-training/training-
courses#usingcomputers.
28. Gomez A, Tang K. Charge and fission of droplets in electrostatic sprays. Phys
Fluids. 1994;6(1):404-414. doi:10.1063/1.868037.
29. Karas M, Bachmann D, Bahr U, Hillenkamp F. Matrix-assisted ultraviolet laser
desorption of non-volatile compounds. Int J Mass Spectrom Ion Process.
1987;78(C):53-68. doi:10.1016/0168-1176(87)87041-6.
146
30. Tanaka K, Waki H, Ido Y, Akita S, Yoshida Y, Yohida T. Protein and polymer
analyses up to m/z 100,000 by laser ionization time-of-flight mass spectrometry.
Rapid Commun Mass Spectrom. 1988;2(8):151-153.
31. Trimpin S, Inutan ED, Herath TN, McEwen CN. Matrix-assisted laser
desorption/ionization mass spectrometry method for selectively producing either
singly or multiply charged molecular ions. Anal Chem. 2010;82(1):11-15.
doi:10.1021/ac902066s.
32. Wolff MM, Stephens WE. A pulsed mass spectrometer with time dispersion. Rev
Sci Instrum. 1953;24(8):616-617. doi:10.1063/1.1770801.
33. Wiley WC, McLaren IH. Time-of-flight mass spectrometer with improved
resolution. Rev Sci Instrum. 1955;26(12):1150-1157. doi:10.1063/1.1715212.
34. Sturiale L, Naggi A, Torri G. MALDI mass spectrometry as a tool for
characterizing glycosaminoglycan oligosaccharides and their interaction with
proteins. Semin Thromb Hemost. 2001;27(5):465-472. doi:10.1055/s-2001-17957.
35. Kaufmann R. Matrix-assisted laser desorption ionozation (MALDI) mass
spectrometry: a novel analytical tool in molecular biology and biotechnology. J
Biotechnol. 1995;41:155-175. doi:10.1016/0168-1656(95)00009-F.
36. Stafford G. Ion trap mass spectrometry: A personal perspective. J Am Soc Mass
Spectrom. 2002;13(6):589-596. doi:10.1016/S1044-0305(02)00385-9.
37. Syka JEP, Coon JJ, Schroeder MJ, Shabanowitz J, Hunt DF. Peptide and protein
sequence analysis by electron transfer dissociation mass spectrometry. Proc Natl
147
Acad Sci U S A. 2004;101(26):9528-9533. doi:10.1073/pnas.0402700101.
38. Johnson J V, Yost R a, Kelley PE, Bradford DC. Tandem-in-space and tandem-in-
time mass spectrometry: triple quadrupoles and quadrupole ion traps. AnalChem.
1990;62(20):2162-2172. doi:10.1021/ac00219a003.
39. Cooks RG. Special feature: Historical. Collision‐induced dissociation: Readings
and commentary. J Mass Spectrom. 1995;30(9):1215-1221.
doi:10.1002/jms.1190300902.
40. Hunt DF, Buko AM, Ballard JM, Shabanowitz J, Giordani AB. Sequence analysis
of polypeptides by collision activated dissociation on a triple quadrupole mass
spectrometer. Biol Mass Spectrom. 1981;8(9):397-408.
doi:10.1002/bms.1200080909.
41. McLuckey SA. Principles of collisional activation in analytical mass spectrometry.
J Am Soc Mass Spectrom. 1992;3(6):599-614. doi:10.1016/1044-0305(92)85001-
Z.
42. Zhang Z. Prediction of electron-transfer/capture dissociation spectra of peptides.
Anal Chem. 2010;82(5):1990-2005. doi:10.1021/ac902733z.
43. Hart-Smith G. A review of electron-capture and electron-transfer dissociation
tandem mass spectrometry in polymer chemistry. Anal Chim Acta. 2014;808:44-
55. doi:10.1016/j.aca.2013.09.033.
44. Kanu AB, Dwivedi P, Tam M, Matz L, Hill HH. Ion mobility-mass spectrometry. J
Mass Spectrom. 2008;43(1):1-22. doi:10.1002/jms.1383.
148
45. Hudgins RR, Woenckhaus J, Jarrold MF. High resolution ion mobility
measurements for gas phase proteins: correlation between solution phase and
gas phase conformations. Int J Mass Spectrom. 1997;165:497-507.
46. Clemmer DE, Hudgins RR, Jarrold MF. Naked Protein Conformations:
Cytochrome c in the Gas Phase. J Am Chem Soc. 1995;117(40):10141-10142.
doi:10.1021/ja00145a037.
47. Harris GA, Kwasnik M, Fernández FM. Direct analysis in real time coupled to
multiplexed drift tube ion mobility spectrometry for detecting toxic chemicals. Anal
Chem. 2011;83(6):1908-1915. doi:10.1021/ac102246h.
48. Sacristan E, Soils AA. A swept-field aspiration condenser as an ion-mobility
spectrometer. IEEE Trans Instrum Meas. 1998;47(3):769-775.
doi:10.1109/19.744345.
49. Solis AA, Sacristán E. Designing the measurement cell of a swept-field differential
aspiration condenser. Rev Mex Fis. 2006;52(4):322-328.
50. Giles K, Pringle SD, Worthington KR, Little D, Wildgoose JL, Bateman RH.
Applications of a travelling wave-based radio-frequency-only stacked ring ion
guide. Rapid Commun Mass Spectrom. 2004;18(20):2401-2414.
doi:10.1002/rcm.1641.
51. Kanu AB, Dwivedi P, Tam M, Matz L, Hill HHJ. Ion mobility – mass spectrometry.
J Mass Spectrom. 2008;43:1-22. doi:10.1002/jms.
52. Lapthorn C, Pullen F, Chowdhry BZ. Ion mobility spectrometry-mass spectrometry
149
(IMS-MS) of small molecules: Separating and assigning structures to ions. Mass
Spectrom Rev. 2013;32(1):43-71. doi:10.1002/mas.21349.
53. Ruotolo BT, Benesch JLP, Sandercock AM, Hyung S-J, Robinson C V. Ion
mobility–mass spectrometry analysis of large protein complexes. Nat Protoc.
2008;3(7):1139-1152. doi:10.1038/nprot.2008.78.
54. Smith D, Knapman T, Campuzano I, et al. Deciphering drift time measurements
from travelling wave ion mobility spectrometry-mass spectrometry studies. Eur J
Mass Spectrom. 2009;15(5):113. doi:10.1255/ejms.947.
55. Michaelevski I, Kirshenbaum N, Sharon M. T-wave Ion Mobility-mass
Spectrometry: Basic Experimental Procedures for Protein Complex Analysis. J Vis
Exp. 2010;(41). doi:10.3791/1985.
56. Amass W, Amass A, Tighe B. A review of biodegradable polymers: uses, current
developments in the synthesis and characterization of biodegradable polyesters,
blends of biodegradable polymers and recent advances in biodegradation studies.
Polym Int. 1998;47(2):89-144. doi:10.1002/(SICI)1097-
0126(1998100)47:2<89::AID-PI86>3.0.CO;2-F.
57. Ikada Y, Tsuji H. Biodegradable polyesters for medical and ecological
applications. Macromol Rapid Commun. 2000;21(3):117-132.
doi:10.1002/(SICI)1521-3927(20000201)21:3<117::AID-MARC117>3.3.CO;2-O.
58. Nair LS, Laurencin CT. Biodegradable polymers as biomaterials. Prog Polym Sci.
2007;32(8-9):762-798. doi:10.1016/j.progpolymsci.2007.05.017.
150
59. Gunatillake PA, Adhikari R, Gadegaard N. Biodegradable synthetic polymers for
tissue engineering. Eur Cells Mater. 2003;5:1-16. doi:10.22203/eCM.v005a01.
60. Luckachan GE, Pillai CKS. Biodegradable Polymers- A Review on Recent Trends
and Emerging Perspectives. J Polym Environ. 2011;19(3):637-676.
doi:10.1007/s10924-011-0317-1.
61. Temenoff JS, Mikos a G. Injectable biodegradable materials for orthopedic tissue
engineering. Biomaterials. 2000;21(23):2405-2412. doi:10.1016/S0142-
9612(00)00108-3.
62. Wang S, Lu L, Yaszemski MJ. Bone-tissue-engineering material poly(propylene
fumarate): Correlation between molecular weight, chain dimension, and physical
properties. Biomacromolecules. 2006;7(6):1976-1982. doi:10.1021/bm060096a.
63. Bose S, Roy M, Bandyopadhyay A. Recent advances in bone tissue engineering
scaffolds. Trends Biotechnol. 2012;30(10):546-554.
doi:10.1016/j.tibtech.2012.07.005.
64. Middleton JC, Tipton AJ. Synthetic biodegradable polymers as orthopedic
devices. Biomaterials. 2000;21(23):2335-2346. doi:10.1016/S0142-
9612(00)00101-0.
65. Horch RA, Shahid N, Mistry AS, Timmer MD, Mikos AG, Barron AR.
Nanoreinforcement of poly(propylene fumarate)-based networks with surface
modified alumoxane nanoparticles for bone tissue engineering.
Biomacromolecules. 2004;5(5):1990-1998. doi:10.1021/bm049768s.
151
66. Wang S, Lu L, Gruetzmacher J a, Currier BL, Yaszemski MJ. A Biodegradable
and Cross-Linkable Multiblock Copolymer Consisting of Poly ( propylene fumarate
) and Poly ( -caprolactone ): Synthesis , Characterization , and Physical
Properties. Macromolecules. 2005;38(17):7358-7370.
http://pubs.acs.org/doi/abs/10.1021/ma050884c.
67. Ronca A, Ambrosio L, Grijpma DW. Design of porous three-dimensional
PDLLA/nano-hap composite scaffolds using stereolithography. J Appl Biomater
Funct Mater. 2012;10(3):249-258. doi:10.5301/JABFM.2012.10211.
68. Skoog SA, Goering PL, Narayan RJ. Stereolithography in tissue engineering. J
Mater Sci Mater Med. 2014;25(3):845-856. doi:10.1007/s10856-013-5107-y.
69. Childers EP, Wang MO, Becker ML, Fisher JP, Dean D. 3D printing of resorbable
poly(propylene fumarate) tissue engineering scaffolds. MRS Bull. 2015;40(2):119-
126. doi:10.1557/mrs.2015.2.
70. Luo Y, Dolder CK, Walker JM, Mishra R, Dean D, Becker ML. Synthesis and
Biological Evaluation of Well-Defined Poly(propylene fumarate) Oligomers and
Their Use in 3D Printed Scaffolds. Biomacromolecules. 2016;17(2):690-697.
doi:10.1021/acs.biomac.6b00014.
71. Sabir MI, Xu X, Li L. A review on biodegradable polymeric materials for bone
tissue engineering applications. J Mater Sci. 2009;44(21):5713-5724.
doi:10.1007/s10853-009-3770-7.
72. Kasper FK, Tanahashi K, Fisher JP, Mikos AG. Synthesis of poly(propylene
fumarate). Nat Protoc. 2009;4(4):518-525. doi:10.1038/nprot.2009.24.
152
73. Jeske RC, DiCiccio AM, Coates GW. Alternating copolymerization of epoxides
and cyclic anhydrides: An improved route to aliphatic polyesters. J Am Chem Soc.
2007;129(37):11330-11331. doi:10.1021/ja0737568.
74. Diciccio AM, Coates GW. Ring-opening copolymerization of maleic anhydride with
epoxides: A chain-growth approach to unsaturated polyesters. J Am Chem Soc.
2011;133(28):10724-10727. doi:10.1021/ja203520p.
75. Spyros A. Characterization of unsaturated polyester and alkyd resins using one-
and two-dimensional NMR spectroscopy. J Appl Polym Sci. 2003;88(7):1881-
1888. doi:10.1002/app.11950.
76. Chen R, Huang X, Wang M, Jin G. Evaluation and comparison of mid-infrared,
Raman and near-infrared spectroscopies for characterization and determination of
the compositions in fully biodegradable poly(lactic acid)/poly(propylene
carbonate)/poly(butylene adipate-co-terephthalate) blends . J Macromol Sci Part
A. 2016;53(6):354-361. doi:10.1080/10601325.2016.1166001.
77. Carroccio S, Rizzarelli P, Puglisi C. Matrix-assisted laser desorption/ionisation
time-of-flight characterisation of biodegradable aliphatic copolyesters. Rapid
Commun Mass Spectrom. 2000;14(16):1513-1522. doi:10.1002/1097-
0231(20000830)14:16<1513::AID-RCM57>3.0.CO;2-C.
78. Arnould MA, Wesdemiotis C, Geiger RJ, Park ME, Buehner RW, Vanderorst D.
Structural characterization of polyester copolymers by MALDI mass spectrometry.
In: Progress in Organic Coatings. Vol 45. ; 2002:305-312. doi:10.1016/S0300-
9440(02)00114-5.
153
79. Arnould M, Buehner R, Wesdemiotis C. Tandem mass spectrometry
characteristics of polyester anions and cations formed by electrospray ionization.
Eur J Mass Spectrom. 2005;11(1):243. doi:10.1255/ejms.746.
80. Adamus G, Rizzarelli P, Montaudo MS, Kowalczuk M, Montaudo G. Matrix-
assisted laser desorption/ionization time-of-flight mass spectrometry with size-
exclusion chromatographic fractionation for structural characterization of synthetic
aliphatic copolyesters. Rapid Commun Mass Spectrom. 2006;20(5):804-814.
doi:10.1002/rcm.2365.
81. Kricheldorf HR, Behnken G. Biodegradable hyperbranched aliphatic polyesters
derived from pentaerythritol. Macromolecules. 2008;41(15):5651-5657.
doi:10.1021/ma800201f.
82. Kricheldorf HR, Yashiro T, Weidner S. Isomerization-free polycondensations of
maleic anhydride with α,ω-alkanediols. Macromolecules. 2009;42(17):6433-6439.
doi:10.1021/ma9009356.
83. Kricheldorf HR, Weidner SM, Scheliga F. Poly(ε-caprolactone) by combined ring-
opening polymerization and azeotropic polycondensationa - The role of cyclization
and equilibration. J Polym Sci Part A Polym Chem. 2012;50(20):4206-4214.
doi:10.1002/pola.26222.
84. Scionti V, Wesdemiotis C. Electron transfer dissociation versus collisionally
activated dissociation of cationized biodegradable polyesters. J Mass Spectrom.
2012;47(11):1442-1449. doi:10.1002/jms.3097.
85. Rizzarelli P. Matrix-assisted laser desorption ionization time-of-flight/time-of-flight
154
tandem mass spectra of biodegradable polybutylenesuccinate. Rapid Commun
Mass Spectrom. 2013;27(19):2213-2225. doi:Doi 10.1002/Rcm.6669.
86. Adamus G, Kwiecien I, Maksymiak M, Balakier T, Jurczak J, Kowalczuk M.
Molecular level structure of novel synthetic analogues of aliphatic biopolyesters as
revealed by multistage mass spectrometry. Anal Chim Acta. 2014;808:104-114.
doi:DOI 10.1016/j.aca.2013.09.001.
87. Katzenmeyer BC, Cool LR, Williams JP, Craven K, Brown JM, Wesdemiotis C.
Electron transfer dissociation of sodium cationized polyesters: Reaction time
effects and combination with collisional activation and ion mobility separation. Int
J Mass Spectrom. 2015;378:303-311. doi:10.1016/j.ijms.2014.09.021.
88. Kowalczuk M, Adamus G. Mass spectrometry for the elucidation of the subtle
molecular structure of biodegradable polymers and their degradation products.
Mass Spectrom Rev. 2016;35(1):188-198. doi:10.1002/mas.21474.
89. Wyttenbach T, Bowers MT. Gas-Phase Conformations : The Ion Mobility / Ion
Chromatography Method. Top Curr Chem. 2003;225:207-232.
doi:10.1007/b10470.
90. Wyttenbach T, Bowers MT. Intermolecular Interactions in Biomolecular Systems
Examined by Mass Spectrometry. Annu Rev Phys Chem. 2007;58(1):511-533.
doi:10.1146/annurev.physchem.58.032806.104515.
91. Kanu AB, Dwivedi P, Tam M, Matz L, Jr HHH. Ion mobility – mass spectrometry. J
Mass Spectrom. 2008;43:1-22. doi:10.1002/jms.
155
92. Jurneczko E, Barran PE. How useful is ion mobility mass spectrometry for
structural biology? The relationship between protein crystal structures and their
collision cross sections in the gas phase. Analyst. 2011;136(1):20-28.
doi:10.1039/C0AN00373E.
93. Trimpin S, Clemmer DE. Ion mobility spectrometry/mass spectrometry snapshots
for assessing the molecular compositions of complex polymeric systems. Anal
Chem. 2008;80(23):9073-9083. doi:10.1021/ac801573n.
94. May JC, McLean JA. Ion mobility-mass spectrometry: Time-dispersive
instrumentation. Anal Chem. 2015;87(3):1422-1436. doi:10.1021/ac504720m.
95. Ewing MA, Glover MS, Clemmer DE. Hybrid ion mobility and mass spectrometry
as a separation tool. J Chromatogr A. 2016;1439:3-25.
doi:10.1016/j.chroma.2015.10.080.
96. Wesdemiotis C. Multidimensional Mass Spectrometry of Synthetic Polymers and
Advanced Materials. Angew Chemie Int Ed. 2017. doi:10.1002/ange.201607003
Link.
97. Suckau D, Resemann A, Schuerenberg M, Hufnagel P, Franzen J, Holle A. A
novel MALDI LIFT-TOF/TOF mass spectrometer for proteomics. Anal Bioanal
Chem. 2003;376(7):952-965. doi:10.1007/s00216-003-2057-0.
98. Alalwiat A, Grieshaber SE, Paik BA, Kiick KL, Jia X, Wesdemiotis C. Top-down
mass spectrometry of hybrid materials with hydrophobic peptide and hydrophilic
or hydrophobic polymer blocks. Analyst. 2015;140(22):7550-7564.
doi:10.1039/C5AN01600B.
156
99. Shvartsburg AA, Mashkevich S V., Baker ES, Smith RD. Optimization of
algorithms for ion mobility calculations. J Phys Chem A. 2007;111(10):2002-2010.
doi:10.1021/jp066953m.
100. MOBCAL http://nano. chem. indiana. ed. (accessed 11, 2017). M. No Title.
101. Wesdemiotis C, Solak N, Polce MJ, Dabney DE, Chaicharoen K, Katzenmeyer
BC. Fragmentation pathways of polymer ions. Mass Spectrom Rev.
2011;30(4):523-529. doi:10.1002/mas.20282.
102. Hoskins JN, Trimpin S, Grayson SM. Architectural differentiation of linear and
cyclic polymeric isomers by ion mobility spectrometry-mass spectrometry.
Macromolecules. 2011;44(17):6915-6918. doi:10.1021/ma2012046.
103. Zhong Y, Hyung S-J, Ruotolo BT. Ion mobility–mass spectrometry for structural
proteomics. Expert Rev Proteomics. 2012;9(1):47-58. doi:10.1586/epr.11.75.
104. López A, Tarragó T, Vilaseca M, Giralt E. Applications and future of ion mobility
mass spectrometry in structural biology. New J Chem. 2013;37(5):1283.
doi:10.1039/c3nj41051j.
105. Lanucara F, Holman SW, Gray CJ, Eyers CE. The power of ion mobility-mass
spectrometry for structural characterization and the study of conformational
dynamics. Nat Chem. 2014;6(4):281-294. doi:10.1038/nchem.1889.
106. Liu X, Cool LR, Lin K, Kasko AM, Wesdemiotis C. Tandem mass spectrometry
and ion mobility mass spectrometry for the analysis of molecular sequence and
architecture of hyperbranched glycopolymers. Analyst. 2015;140(4):1182-1191.
157
doi:10.1039/C4AN01599A.
107. Ruotolo BT, Benesch JLP, Sandercock AM, Hyung S-J, Robinson C V. Ion
mobility-mass spectrometry analysis of large protein complexes. Nat Protoc.
2008;3(7):1139-1152. doi:10.1038/nprot.2008.78.
108. J. Kalia RTR. Advances in Bioconjugation. Curr Org Chem. 2010;14(2):138-147.
doi:10.3816/CLM.2009.n.003.Novel.
109. Jevševar S, Kunstelj M, Porekar VG. PEGylation of therapeutic proteins.
Biotechnol J. 2010;5(1):113-128. doi:10.1002/biot.200900218.
110. Payne RW, Murphy BM, Manning MC. Product development issues for PEGylated
proteins. Pharm Dev Technol. 2011;16(5):423-440.
doi:10.3109/10837450.2010.513990.
111. Yoshimoto N, Isakari Y, Itoh D, Yamamoto S. PEG chain length impacts yield of
solid-phase protein PEGylation and efficiency of PEGylated protein separation by
ion-exchange chromatography: Insights of mechanistic models. Biotechnol J.
2013;8(7):801-810. doi:10.1002/biot.201200325.
112. Veronese FM. Peptide and protein PEGylation: a review of problems and
solutions. Biomaterials. 2001;22(5):405-417. doi:10.1016/S0142-9612(00)00193-
9.
113. Top A, Kiick KL, Roberts CJ. Modulation of Self-Association and Subsequent
Fibril Formation in an Alanine-Rich Helical Polypeptide Ayben.
Biomacromolecules. 2008;9(6):1595-1603. doi:10.1021/bm800056r.
158
114. Banerjee SS, Aher N, Patil R, et al. Poly(ethylene glycol)-Prodrug Conjugates:
Concept, Design, and Applications. J Drug Deliv. 2012;2012:1-17.
doi:10.1155/2012/103973, 10.1155/2012/103973.
115. Hamley IW. PEG − Peptide Conjugates. Biomacromolecules. 2014;15(5):1543-
1559. doi:doi.org/10.1021/bm500246w |.
116. Calero-Rubio C, Paik B, Jia X, Kiick KL, Roberts CJ. Predicting unfolding
thermodynamics and stable intermediates for alanine-rich helical peptides with the
aid of coarse-grained molecular simulation. Biophys Chem. 2016;217:8-19.
doi:10.1016/j.bpc.2016.07.002.
117. Phelps EA, Enemchukwu NO, Fiore VF, et al. Maleimide cross-linked bioactive
PEG hydrogel exhibits improved reaction kinetics and cross-linking for cell
encapsulation and in situ delivery. Adv Mater. 2012;24(1):64-70.
doi:10.1002/adma.201103574.
118. Grieshaber SE, Farran AJE, Bai S, Kiick KL, Jia X. Tuning the properties of
elastin mimetic hybrid copolymers via a modular polymerization method.
Biomacromolecules. 2012;13(6):1774-1786. doi:10.1021/bm3002705.
119. Dehn S, Castelletto V, Hamley IW, Perrier S. Altering peptide fibrillization by
polymer conjugation. Biomacromolecules. 2012;13(9):2739-2747.
doi:10.1021/bm3007117.
120. Thomas Wyttenbach, Jennifer Gidden and MTB. Developments in Ion Mobility:
Theory, Instrumentation, and Applications. In Ion Mobility Spectrometry–Mass
Spectrometry, ed. CL Wilkins, S Trimpin,. In: ; 2011:3-30. doi:Boca Raton, FL:
159
CRC.
121. Li X, Guo L, Casiano-Maldonado M, Zhang D, Wesdemiotis C. Top-down
multidimensional mass spectrometry methods for synthetic polymer analysis.
Macromolecules. 2011;44(12):4555-4564. doi:10.1021/ma200542p.
122. McLafferty FW, Breuker K, Jin M, et al. Top-down MS, a powerful complement to
the high capabilities of proteolysis proteomics. FEBS J. 2007;274(24):6256-6268.
doi:10.1111/j.1742-4658.2007.06147.x.
123. Kelleher NS and NL. Decoding protein modifications using top-down mass
spectrometry. Nat Methods. 2007;4(10):817-821. doi:10.1038/nmeth1097.
124. Liu X, Cool LR, Lin K, Kasko AM, Wesdemiotis C. Tandem mass spectrometry
and ion mobility mass spectrometry for the analysis of molecular sequence and
architecture of hyperbranched glycopolymers. Analyst. 2015;140(4):1182-1191.
doi:10.1039/c4an01599a.
125. Alalwiat A, Grieshaber SE, Paik BA, Kiick KL, Jia X, Wesdemiotis C. Top-down
mass spectrometry of hybrid materials with hydrophobic peptide and hydrophilic
or hydrophobic polymer blocks. Analyst. 2015;140(22):7550-7564.
doi:10.1039/c5an01600b.
126. Alalwiat A, Tang W, Gerişlioğlu S, Becker ML, Wesdemiotis C. Mass
Spectrometry and Ion Mobility Characterization of Bioactive Peptide–Synthetic
Polymer Conjugates. Anal Chem. 2016:acs.analchem.6b03553.
doi:10.1021/acs.analchem.6b03553.
160
127. Grun J, Revell JD, Conza M, Wennemers H. Peptide-polyethylene glycol
conjugates: Synthesis and properties of peptides bearing a C-terminal
polyethylene glycol chain. Bioorganic Med Chem. 2006;14(18):6197-6201.
doi:10.1016/j.bmc.2006.05.079.
128. Wilson P. Synthesis and Applications of Protein / Peptide-Polymer Conjugates.
2017;201600595:1-15. doi:10.1002/macp.201600595.
129. Berendsen HJC, van der Spoel D, van Drunen R. GROMACS: A message-
passing parallel molecular dynamics implementation. Comput Phys Commun.
1995;91(1-3):43-56. doi:10.1016/0010-4655(95)00042-E.
130. Van Der Spoel D, Lindahl E, Hess B, Groenhof G, Mark AE, Berendsen HJC.
GROMACS: Fast, flexible, and free. J Comput Chem. 2005;26(16):1701-1718.
doi:10.1002/jcc.20291.
131. Hess B, Kutzner C, Van Der Spoel D, Lindahl E. GRGMACS 4: Algorithms for
highly efficient, load-balanced, and scalable molecular simulation. J Chem Theory
Comput. 2008;4(3):435-447. doi:10.1021/ct700301q.
132. Pronk S, P??ll S, Schulz R, et al. GROMACS 4.5: A high-throughput and highly
parallel open source molecular simulation toolkit. Bioinformatics. 2013;29(7):845-
854. doi:10.1093/bioinformatics/btt055.
133. GROMACS. http://www.gromacs.org/About_Gromacs.
134. Abraham M, Hess B, Spoel D van der, Lindahl E. GROMACS User Manual
version 5.0.7. www.gromacs.org. 2015. doi:10.1007/SpringerReference_28001.
161
135. Foloppe N, MacKerell ADJ. All-Atom Empirical Force Field for Nucleic Acids : I .
Parameter Optimization Based on Small Molecule and Condensed Phase
Macromolecular Target Data. J Comput Chem. 2000;21(2):86-104.
doi:10.1002/(SICI)1096-987X(20000130)21:2<86::AID-JCC2>3.0.CO;2-G.
136. Humphrey W, Dalke A, Schulten K. VMD: Visual molecular dynamics. J Mol
Graph. 1996;14(1):33-38. doi:10.1016/0263-7855(96)00018-5.
137. Http://nano.chem.indiana.edu/software.html. No Title.
138. Bowers MT, Kemper PR, von Helden G, van Koppen P a. Gas-phase ion
chromatography: transition metal state selection and carbon cluster formation.
Science. 1993;260(5113):1446-1451. doi:10.1126/science.260.5113.1446.
139. Wyttenbach T, Von Helden G, Bowers MT. Gas-phase conformation of biological
molecules: Bradykinin. J Am Chem Soc. 1996;118(35):8355-8364.
doi:10.1021/ja9535928.
140. Jarrold MF. Helices and Sheets in vacuo. Phys Chem Chem Phys.
2007;9(14):1659-1671. doi:10.1039/b612615d.
141. Kohtani M, Jarrold M, Wee S, O’Hair R. Metal Ion Interactions with Polyalanine
Peptides. J Phys Chem B. 2004;108:6093-6097. doi:10.1021/jp049708g.
142. McLean JR, McLean JA, Wu Z, et al. Factors that influence helical preferences for
singly charged gas-phase peptide ions: The effects of multiple potential charge-
carrying sites. J Phys Chem B. 2010;114(2):809-816. doi:10.1021/jp9105103.
143. Xiong K, Asciutto EK, Madura JD, Asher SA. Salt dependence of an α-helical
162
peptide folding energy landscapes. Biochemistry. 2009;48(45):10818-10826.
doi:10.1021/bi9014709.
144. Asciutto EK, General IJ, Xiong K, Asher SA, Madura JD. Sodium perchlorate
effects on the helical stability of a mainly alanine peptide. Biophys J.
2010;98(2):186-196. doi:10.1016/j.bpj.2009.10.013.
145. More MB, Ray D, Armentrout PB. Cation−Ether Complexes in the Gas Phase:
Bond Dissociation Energies of Na+(dimethyl ether)x, x = 1−4; Na+(1,2-
dimethoxyethane)x, x = 1 and 2; and Na+(12-crown-4). J Phys Chem A.
1997;101(5):831-839. doi:10.1021/jp962851s.
146. Memboeuf A, Vekey K, Lendvay G. Structure and energetics of poly(ethylene
glycol) cationized by Li+, Na+, K+ and Cs+: a first-principles study. Eur J Mass
Spectrom. 2011;17(1):33-46. doi:10.1255/ejms.1107.
147. Wang P, Wesdemiotis C, Kapota C, Ohanessian G. The Sodium Ion Affinities of
Simple Di-, Tri-, and Tetrapeptides. J Am Soc Mass Spectrom. 2007;18(3):541-
552. doi:10.1016/j.jasms.2006.10.024.
148. Thakur VK, Thakur MK. Recent trends in hydrogels based on psyllium
polysaccharide: A review. J Clean Prod. 2014;82:1-15.
doi:10.1016/j.jclepro.2014.06.066 Review.
149. Spain SG, Cameron NR. A Spoonful of Sugar: The Application of Glycopolymers
in Therapeutics. Polym Chem. 2011;2(1):60-68. doi:10.1039/c0py00149j.
150. Lee J, Lin EW, Lau UY, Hedrick JL, Bat E, Maynard HD. Trehalose glycopolymers
163
as excipients for protein stabilization. Biomacromolecules. 2013;14(8):2561-2569.
doi:10.1021/bm4003046
151. Miura Y. Design and synthesis of well-defined glycopolymers for the control of
biological functionalities. Polym J. 2012;44(7):679-689. doi:10.1038/pj.2012.4.
152. Zeng X, Andrade CAS, Oliveira MDL, Sun XL. Carbohydrate-protein interactions
and their biosensing applications. Anal Bioanal Chem. 2012;402(10):3161-3176.
doi:10.1007/s00216-011-5594-y.
153. Lee YC, Lee RT. Carbohydrate-Protein Interactions: Basis of Glycobiology. Acc
Chem Res. 1995;28(8):321-327. doi:10.1021/ar00056a001.
154. Miura Y. Synthesis and biological application of glycopolymers. J Polym Sci Part
A Polym Chem. 2007;45(22):5031-5036. doi:10.1002/pola.22369.
155. Álvarez-Paino M, Bordegé V, Cuervo-Rodríguez R, Muñoz-Bonilla A, Fernández-
García M. Well-defined glycopolymers via raft polymerization: Stabilization of gold
nanoparticles. Macromol Chem Phys. 2014;215(19):1915-1924.
doi:10.1002/macp.201400306.
156. Moad G, Chong YK, Postma A, Rizzardo E, Thang SH. Advances in RAFT
polymerization: The synthesis of polymers with defined end-groups. Polymer
(Guildf). 2005;46(19 SPEC. ISS.):8458-8468. doi:10.1016/j.polymer.2004.12.061.
157. Matyjaszewski K, Xia J. Atom transfer radical polymerization. Chem Rev.
2001;101(9):2921-2990. doi:10.1021/cr940534g.
158. Matyjaszewski K. Atom Transfer Radical Polymerization (ATRP): Current Status
164
and Future Perspectives. Macromolecules. 2012;45(10):4015-4039.
doi:10.1021/ma3001719.
159. Bielawski CW, Grubbs RH. Living ring-opening metathesis polymerization. Prog
Polym Sci. 2007;32(1):1-29. doi:10.1016/j.progpolymsci.2006.08.006.
160. Elling BR, Xia Y. Living Alternating Ring-Opening Metathesis Polymerization
Based on Single Monomer Additions. J Am Chem Soc. 2015;137(31):9922-9926.
doi:10.1021/jacs.5b05497.
161. Kakuchi T, Narumi A, Matsuda T, et al. Glycoconjugated polymer. 5. Synthesis
and characterization of a seven-arm star polystyrene with a β-cyclodextrin core
based on TEMPO-mediated living radical polymerization. Macromolecules.
2003;36(11):3914-3920. doi:10.1021/ma021296r.
162. Rizzardo E, Chen M, Chong B, Moad G, Skidmore M, Thang SH. RAFT
Polymerization: Adding to the picture. In: Radical Polymerization: Kinetics and
Mechanism. ; 2007:104-116. doi:10.1002/9783527610860.ch11.
163. Bulmus V. RAFT polymerization mediated bioconjugation strategies. Polym
Chem. 2011;2(7):1463-1472. doi:10.1039/c1py00039j.
164. Mayadunne RTA, Rizzardo E, Chiefari J, et al. Living polymers by the use of
trithiocarbonates as reversible addition-fragmentation chain transfer (RAFT)
agents: ABA triblock copolymers by radical polymerization in two steps.
Macromolecules. 2000;33(2):243-245. doi:10.1021/ma991451a.
165. An Z, Qiu Q, Liu G. Synthesis of architecturally well-defined nanogels via RAFT
165
polymerization for potential bioapplications. Chem Commun (Camb).
2011;47(46):12424-12440. doi:10.1039/c1cc13955j.
166. Liau WT, Bonduelle C, Brochet M, Lecommandoux S, Kasko AM. Synthesis,
characterization, and biological interaction of glyconanoparticles with controlled
branching. Biomacromolecules. 2015;16(1):284-294. doi:10.1021/bm501482q.
167. Chaicharoen K, Polce MJ, Singh A, Pugh C, Wesdemiotis C. Characterization of
linear and branched polyacrylates by tandem mass spectrometry. Anal Bioanal
Chem. 2008;392(4):595-607. doi:10.1007/s00216-008-1969-0.
168. Leader B, Baca QJ, Golan DE. Protein therapeutics: a summary and
pharmacological classification. Nat Rev Drug Discov. 2008;7(1):21-39.
doi:10.1038/nrd2399.
169. Walsh G. Biopharmaceutical benchmarks 2014. Nat Biotechnol. 2014;32(10):992-
1000. doi:10.1038/nbt.3040.
170. Liu Y, Lee J, Mansfield KM, et al. Trehalose Glycopolymer Enhances Both
Solution Stability and Pharmacokinetics of a Therapeutic Protein. Bioconjug
Chem. 2017:acs.bioconjchem.6b00659. doi:10.1021/acs.bioconjchem.6b00659.
171. Frokjaer S, Otzen DE. Protein drug stability: a formulation challenge. Nat Rev
Drug Discov. 2005;4(4):298-306. doi:10.1038/nrd1695.
172. Shiraki K, Kudou M, Nishikori S, Kitagawa H, Imanaka T, Takagi M. Arginine
ethylester prevents thermal inactivation and aggregation of lysozyme. Eur J
Biochem. 2004;271(15):3242-3247. doi:10.1111/j.1432-1033.2004.04257.x.
166
173. Kaushik JK, Bhat R. Why is trehalose an exceptional protein stabilizer? An
analysis of the thermal stability of proteins in the presence of the compatible
osmolyte trehalose. J Biol Chem. 2003;278(29):26458-26465.
doi:10.1074/jbc.M300815200.
174. Zhang Y, Cremer PS. Chemistry of Hofmeister Anions and Osmolytes. Annu Rev
Phys Chem. 2010;61(1):63-83.
doi:10.1146/annurev.physchem.59.032607.093635.
175. Randolph TW, Jones LS. Surfactant-protein interactions. Pharm Biotechnol.
2002;13(Rational Design of Stable Protein Formulations):159-175.
doi:10.1007/978-1-4615-0557-0_7.
176. Klok HA. Peptide/protein-synthetic polymer conjugates: Quo vadis.
Macromolecules. 2009;42(21):7990-8000. doi:10.1021/ma901561t.
177. Pelegri-Oday EM, Lin EW, Maynard HD. Therapeutic protein-polymer conjugates:
Advancing beyond pegylation. J Am Chem Soc. 2014;136(41):14323-14332.
doi:10.1021/ja504390x.
178. Elvira C, Gallardo A, San Roman J, Cifuentes A. Covalent polymer-drug
conjugates. Molecules. 2005;10(1):114-125. doi:10.3390/10010114.
179. Naipu H, Yufeng H, Rongmin W, Pengfei S, Yun Z, Gang L. Protein polymer
conjugates. Prog Chem. 2010;22(12):2388-2396. doi:10.1016/B978-0-444-53349-
4.00250-8.
180. Mero A, Fang Z, Pasut G, Veronese FM, Viegas TX. Selective conjugation of
167
poly(2-ethyl 2-oxazoline) to granulocyte colony stimulating factor. J Control
Release. 2012;159(3):353-361. doi:10.1016/j.jconrel.2012.02.025.
181. Tao L, Liu J, Davis TP. Branched polymer-protein conjugates made from mid-
chain-functional P(HPMA). Biomacromolecules. 2009;10(10):2847-2851.
doi:10.1021/bm900678r.
182. Liebner R, Mathaes R, Meyer M, Hey T, Winter G, Besheer A. Protein HESylation
for half-life extension: Synthesis, characterization and pharmacokinetics of
HESylated anakinra. Eur J Pharm Biopharm. 2014;87(2):378-385.
doi:10.1016/j.ejpb.2014.03.010.
183. Veronese FM, Pasut G. PEGylation, successful approach to drug delivery
REVIEWS. Drug Discov Today. 2005;10(21):1451-1458. doi:10.1016/s1359-
6446(05)03575-0.
184. Alconcel SNS, Baas AS, Maynard HD. FDA-approved poly(ethylene glycol)–
protein conjugate drugs. Polym Chem. 2011;2(7):1442. doi:10.1039/c1py00034a.
185. Pfister D, Morbidelli M. Process for protein PEGylation. J Control Release.
2014;180(1):134-149. doi:10.1016/j.jconrel.2014.02.002.
186. Manning MC, Chou DK, Murphy BM, Payne RW, Katayama DS. Stability of
protein pharmaceuticals: An update. Pharm Res. 2010;27(4):544-575.
doi:10.1007/s11095-009-0045-6.
187. Teramoto N, Sachinvala ND, Shibata M. Trehalose and trehalose-based polymers
for environmentally benign, biocompatible and bioactive materials. Molecules.
168
2008;13(8):1773-1816. doi:10.3390/molecules13081773.
188. Jain NK, Roy I. Effect of trehalose on protein structure. Protein Sci.
2008;18(1):24-36. doi:10.1002/pro.3.
189. Mancini RJ, Lee J, Maynard HD. Trehalose glycopolymers for stabilization of
protein conjugates to environmental stressors. J Am Chem Soc.
2012;134(20):8474-8479. doi:10.1021/ja2120234.
190. Liu X, Cool LR, Lin K, Kasko AM, Wesdemiotis C. Tandem mass spectrometry
and ion mobility mass spectrometry for the analysis of molecular sequence and
architecture of hyperbranched glycopolymers. Analyst. 2015;140(4):1182-1191.
doi:10.1039/c4an01599a.
191. Siuzdak G. Mass Spectrometry for Biotechnology. Academic Press; 1996.
http://www.sciencedirect.com/science/book/9780126474718. Accessed June 2,
2017.
192. Uchio T, Baudyš M, Liu F, Song SC, Kim SW. Site-specific insulin conjugates with
enhanced stability and extended action profile. Adv Drug Deliv Rev. 1999;35(2-
3):289-306. doi:10.1016/S0169-409X(98)00078-7.
193. http://webbook.nist.gov/chemistry/
169
APPENDIX
COPYRIGHT PERMISSION
170
171
172
173
174
175
176