Accurate Quantification of N-Glycolylneuraminic ... - ACS Publications

Subscriber access provided by Kaohsiung Medical University

Article

Accurate Quantification of N-glycolylneuraminic Acid in Therapeutic Proteins Using Supramolecular Mass Spectrometry Hyun Hee Lucina Lee, Chae Eun Heo, Nari Seo, Seung Gyu Yun, Hyun Joo An, and Hugh I. Kim J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b07864 • Publication Date (Web): 28 Aug 2018 Downloaded from http://pubs.acs.org on August 30, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Accurate Quantification of N-glycolylneuraminic Acid in Therapeutic Proteins Using Supramolecular Mass Spectrometry. Hyun Hee L. Lee†, Chae Eun Heo†, Nari Seo‡, Seung Gyu Yun§, Hyun Joo An‡,*, and Hugh I. Kim†,* †

Department of Chemistry, Korea University, Seoul 02841, Republic of Korea Graduate School of Analytical Science & Technology, Chungnam National University, Daejon 34134, Republic of Korea § Department of Laboratory Medicine, Korea University College of Medicine, Seoul 02841, Republic of Korea ‡

Supporting Information Placeholder ABSTRACT: Practical applications of innovative host-guest systems are challenging because of unexpected guest competitors and/or subtle environmental differences. Herein, a supramolecular mass spectrometry (MS)-based method using a synthetic host cucurbit[7]uril (CB[7]) was developed for identifying and quantifying N-glycolylneuraminic acid (Neu5Gc) in therapeutic glycoproteins, which critically reduces drug efficacy. The development of a reliable derivatization-free analytical method for Neu5Gc is highly challenging because of the interference by the abundant Nacetylneuraminic acid (Neu5Ac). CB[7] recognized the subtle structural differences between Neu5Gc and Neu5Ac. Distinct host-guest interactions between CB[7] and the two sialic acids produced a highly linear relationship between the complexation and concentration proportions of the two sialic acids in MS. Furthermore, the developed method had sub-picomolar quantification limits and a wide range of applicability for diverse glycoproteins, demonstrating the potential utility of this method as a reliable assay of Neu5Gc in therapeutic glycoproteins.

lars, is important for both quality control and quality assessment during the manufacturing process.19 For their production, mammalian cell expression systems are generally utilized to maintain the characteristics of N-glycans such as number, type, and location, which influence biological activity, plasma clearance rate, solubility, and conformation of the glycoproteins.20-22 However, the glycoproteins expressed from the animal cells are usually less effective as pharmaceutics in human owing to the presence of Neu5Gc in the glycoproteins.20-22 Neu5Gc is the most abundant sialic acid found in non-human mammals; it cannot be synthesized by human beings because of the lack of the CMAH gene.20 Nevertheless, Neu5Gc is readily absorbed in human tissues from dietary sources (e.g., red meats) and then incorporated with pro-

INTRODUCTION Developing a highly selective host-guest system is important because of its potential applicability to diverse fields including molecular switches,1-2 molecular sensing,3-4 the construction of supramolecular architectures,5-6 and drug/gene delivery systems,78 among others.9-10 To develop a host with high selectivity toward the guest, it is essential that the host has the capability to recognize subtle differences in the guest structures in addition to high binding affinities.11 In aqueous solutions, for example, binding affinities can be increased by maximizing the hydrophobic interactions between the host and guest12-13 along with the stabilization of the complex system via electrostatic interactions.11, 14-15 A rigid structure is typically designed for the host for size complementarity with the guest molecule.11, 14 A number of host-guest systems with high selectivity have been developed with synthetic receptors such as cyclodextrins,16 cucurbit[n]urils (CB[n], n = 5–8, 10, 14),10 and molecular tweezers17 employing such strategies. However, developing practical applications of these innovative host-guest systems is challenging because of unexpected guest competitors and/or subtle environmental differences15, 18 that deviate from the optimized host-guest conditions. Accurately determining the amounts of N-glycolylneuraminic acid (Neu5Gc, Figure 1a) in therapeutic glycoproteins, whose annual market worth is of the order of tens of billions of US dol-

Figure 1. Host-guest strategy for sialic acid analysis. (a) Structures of CB[7] and the two sialic acids investigated herein. (b) Strategy the host-guest chemistry in this study. (c) Mass spectrum of a mixture containing CB[7], Neu5Gc, and Neu5Ac.

ACS Paragon Plus Environment

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

teins or lipids. This results in the expression of antibodies against the Neu5Gc glycoconjugates.20, 23-24 The presence of the antibodies is very important for the efficacy of Neu5Gc-containing glycoprotein pharmaceuticals because antibodies greatly reduce their clearance rate in the human body through antigen-antibody interactions.20-22 Thus, the amounts of non-human monosaccharide Neu5Gc in a therapeutic glycoprotein should be accurately measured to ensure product quality and regulate potential immunogenicity for process development and manufacturing procedures.20, 2425 However, existing analytical methods that identify and quantify Neu5Gc from other monosaccharides, especially from the most abundant sialic acid, N-acetylneuraminic acid (Neu5Ac, Figure 1a), require derivatization for separation and other additional identification processes.26-27 While electrospray ionization mass spectrometry (ESI-MS) combined with high performance liquid chromatography (HPLC) can be used to differentiate between Neu5Gc and Neu5Ac using the difference of a single oxygen atom, its low ionization efficiency and interference from Neu5Ac in the LC-MS spectrum hinder the accurate quantification of Neu5Gc without multiple preparation steps for prederivatization.26, 28-29 Therefore, it is desirable to develop a convenient and reliable method for accurate identification and quantification of Neu5Gc against Neu5Ac. In the present study, host-guest interactions of sialic acids and the synthetic receptor cucurbit[7]uril (CB[7], Figure 1a) in the gas phase were studied. Because of its adequate cavity size and two polar carbonyl portals, CB[7] can encapsulate monosaccharide isomers during the transfer from the solution to the gas phase via ESI with distinct structural orientations.30-31 CB[7] was able to perceive the differences between protonation sites of encapsulated sialic acids in the gas phase inducing similar complexation proportions of two sialic acids, Neu5Gc and Neu5Ac. Therefore, this unique host-guest system was practically developed into an analytical technique for simultaneously identifying and quantifying sialic acids in a glycoprotein using simple ESI-MS scanning. This highly sensitive and convenient method can identify and quantify Neu5Gc in diverse glycoproteins including glycoprotein therapeutics, without derivatization, pre-separation, or tandem MS analysis (Figure 1b).

The intensity of CB[7]-Neu5Ac complex ions was obtained by subtracting the tri-isotropic peak intensity of [CB[7]+Neu5Gc+NH4+H-H2O]2+ (observed: m/z 744.2398; calculated: m/z 744.2381) from the monoisotopic peak intensity of [CB[7]+Neu5Ac+NH4+H]2+ to exclude the increase in INeu5Ac as a result of overlap between the tri-isotopic and monoisotopic peaks (equation (2)). The intensity proportion of the tri-isotopic peak of the CB[7]-dehydrated Neu5Gc complex to its monoisotopic peak, 0.170, was calculated from the isotope pattern of the complex ion. It is noteworthy that there was no significant interaction between CB[7] and sialic acid in the solution of pH 2 and 7 (Figure S2). This implied that the abundances of CB[7]-sialic acid complex ion peaks in the mass spectra were a result of the gas-phase chemistry of CB[7] and sialic acids during ESI instead of a transfer from the solution.31 The relative complexation proportion for a sialic acid (Neu5Gc or Neu5Ac) was calculated using equation (3). Next, the complexation proportions of both Neu5Gc and Neu5Ac in their binary mixture during ESI were investigated. As seen in Figure 2a, a strong correlation was found between the complexation and the concentration proportions of the sialic acids in their binary solution. Diverse concentration proportions (10, 30, 50, 70, and 90%) of Neu5Gc to Neu5Ac in different total sialic acid concen-

RESULTS AND DISCUSSION Supramolecular quantification of Sialic Acids with CB[7] using ESI-MS. ESI-MS spectrum of the sialic acid mixture containing Neu5Gc and Neu5Ac with CB[7] showed that both Neu5Gc and Neu5Ac formed doubly-charged complex ions with CB[7] during ESI (Figure 1c).30-31 Several complex ion peaks of Neu5Gc and Neu5Ac could be observed overlapping each other in the spectrum (Figures 1c and S1, and Table S1). Two complex ion peaks, [CB[7]+Neu5Gc+NH4+H]2+ (observed: m/z 753.2409; calculated: m/z 753.2344) and [CB[7]+Neu5Ac+NH4+H]2+ (observed: m/z 745.2442; calculated: m/z 745.2458), had no overlap in their monoisotopic peaks, which allowed the calculation of the complexation proportion between Neu5Gc and Neu5Ac using the equations given below: / . 1 / . 0.170 / . 2 !"#$% &&#$% '& ( ⁄ ) * 3 where INeu5Gc and INeu5Ac represent the intensities of the monoisotopic complex ion peaks of Neu5Gc and Neu5Ac containing a proton and an ammonium cation, respectively. Each Im/z symbol denotes the intensity of a monoisotopic peak with the m/z described in the subscript. INeu5Gc was obtained from the monoisotopic peak intensity of [CB[7]+Neu5Gc+NH4+H]2+ (equation (1)).

Figure 2. Quantification of sialic acids using gas-phase host-guest chemistry. (a) Complexation proportion of Neu5Gc and Neu5Ac in different total sialic acid concentrations ([SA]t). (b) Quantitative results of binary sialic acid mixtures. (c) Quantitative results of single sialic acid solutions. Figures 2b and 2c show plots of the sialic acid concentrations in the solution ([SA]sol) on the x-axis versus the concentrations determined ([SA]det) by the method described herein.


Page 2 of 7

Page 3 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society trations (0.5, 1, 2, 5, 10, 20, 50, 75, 100 µM) were carefully investigated. The complexation and concentration proportions showed excellent linearity (R2 = 0.999) regardless of the total concentration of sialic acids and CB[7] (Figures 2a and S3). It was found that the complexation preferences of both sialic acids to CB[7] were highly similar. When the concentration ratio of two sialic acids was 1:1, the measured complexation percentages of Neu5Gc and Neu5Ac were 47.2 ± 0.9 and 52.8 ± 0.9%, respectively, using ESI-MS (Figure 2a). These unique host-guest properties of sialic acids with CB[7] during ESI provided the highly sensitive quantification of Neu5Gc by ESI-MS with a limit of detection (LOD) of 0.48 pmol and limit of quantification (LOQ) of 1.60 pmol.32 In addition, the LOD and LOQ for Neu5Ac were 0.20 and 0.65 pmol, respectively (Table S2).32 Figure 2b shows plots of the concentrations of the two sialic acids in the binary mixture determined using host-guest chemistry with CB[7] against their concentrations in solution. Very good agreement was observed between the determined concentrations of sialic acids and their concentrations in solution (Figure 2b and Table S3). This linearity was used to quantify single sialic acid solutions by adding known amounts of Neu5Gc and Neu5Ac as standards into the solutions. The quantification results showed that the determined concentrations of sialic acids were in good agreement with their concentrations in solution (Figure 2c and Table S4). Unique host-guest chemistry between CB[7] and sialic in the gas phase. Prior to developing the practical supramolecular MS method for identifying and quantifying Neu5Gc and Neu5Ac in glycoproteins, the unique complexation properties of the sialic acids with CB[7] were studied using ion mobility mass spec-

trometry (IM-MS). IM-MS can provide information regarding the structural features of ions based on their collision cross sections (CCSs) determined from the traveling velocities of ions under a weak electric field in neutral buffer gas (i.e., He).33 Figure 3a shows the ion arrival time distributions (ATDs) of [CB[7]+sialic acid+NH4+H]2+ and [CB[7]+2H]2+ along with their CCS values determined from the ATDs. The complex ions of both sialic acids had significantly high CCS (CCSCB[7]-Neu5Gc = 227.9.0 ± 0.9 Å2; CCSCB[7]-Neu5Ac = 226.0 ± 0.4 Å2) as compared with the uncomplexed CB[7] ion (CCSCB[7] = 207.0 ± 0.6 Å2). This implied that considerable portions of the sialic acid guests were exposed and lay outside of the CB[7] cavity. It has been previously shown that stable CB[7]-monosaccharide complex ions are formed with a proton and an ammonium cation using ESI with monosaccharide derivatives with an amino or acetamide group.30-31 Further investigation of the complex ions revealed that the protonation of the monosaccharide complex occurred via the N-containing functional group of the monosaccharide. Based on the previous studies, we examined two possible protonation structures34-36 of sialic acids in the complex of CB[7] with a proton and an ammonium cation (Figure 3b, O- and Nprotonations) of the protonated acetamide groups. The representative structures of CB[7]-sialic acid complex ions from molecular dynamics (MD) simulations and density functional theory (DFT) calculations are shown in Figures 3c and S4. The theoretical CCSs calculated using a trajectory (TJ) method,33 which employs a combination of MMFF94 parameters with the Exp-6 potential, were consistent with the experimental CCS values. This indicated that the calculated structures could effectively reflect the gas-

Figure 3. Gas-phase complexation of sialic acids with CB[7]. (a) IM spectra of CB[7]-sialic acid complex ions and the free CB[7] ion. The experimental collision cross section of each ion is indicated on the top of each IM peak. (b) Two protonation structures of the acetamide groups of Neu5Gc and Neu5Ac. (c) Representative structures of CB[7]-Neu5Gc (top) and CB[7]-Neu5Ac (bottom) complex ions. (d) Relative binding energies (∆BEs) of the sialic acid ions. The ∆BEs are the differences in the binding energy of O-protonated Neu5Gc ion. (e) Expanded structures of the representative structures.



phase structures of the complex ions (Table S5). Representative structures showed that a major part of the sialic acid guest was encapsulated in the CB[7] cavity. The large CCSs of the complex ion as compared to that of the uncomplexed CB[7] ion were attributable to the exposure of the protonated acetamide group and a part of the hexose ring of sialic acid from the CB[7] cavity. The complex was stabilized via electrostatic interactions between the protonated acetamide group of the sialic acid and the carbonyl groups of the CB[7] portal. Investigation of the binding energies of the representative structures of the complex ions provided a different insight into the host-guest interactions of the two sialic acids with CB[7] in the gas phase. The binding energies were obtained by subtracting the energy of the optimized structure of each component from that of the CB[7]-sialic acid complex ion (details are provided in the Supporting Information).37 The calculated binding energies between the sialic acids and CB[7] are listed in Table S5. The binding energies of the sialic acids with CB[7] were considerably different depending on the protonation sites (Figure 3d and Table S5). For Neu5Gc, the N-protonated structures preferred complexation with CB[7] compared with the O-protonated structures (Figure 3d). The calculated difference in the binding energies (∆BEs) between the two structures in the CB[7] complexes was 70 kJ/mol. In contrast, in the case of Neu5Ac, the complexation of O-protonated Neu5Gc with CB[7] was energetically preferred to that of N-protonated Neu5Gc by 25 kJ/mol. Since O-protonated free sialic acid ions are more stable than their N-protonated counterparts by ~10–20 kJ/mol (Table S5), different ∆BEs of the sialic acids depending on the protonation sites were likely a result of distinct host-guest interactions between CB[7] and the sialic acids. Based on the ∆BEs, the complexation preference of Neu5Gc and Neu5Ac ions to CB[7] was estimated as N-protonated Neu5Gc > O-protonated Neu5Ac > N-protonated Neu5Ac > Oprotonated Neu5Gc (Figure 3d). The gas-phase host-guest interactions between CB[7] and sialic acid in the complex were also examined for further insights into the complexation preferences of the sialic acid ions. The protonated acetamide groups interact strongly with the carbonyl groups of CB[7] regardless of the protonation sites (Figures 3c and 3e). In the case of the O-protonated sialic acid ions, Ha and Hb (Figure 3b) can form two hydrogen bonds with the carbonyl groups of CB[7]. In addition, Hc and Hd of the N-protonated sialic acids (Figure 3b) can interact with the CB[7] portal. These interactions have a critical effect on the overall geometries of the CB[7]-sialic acid complex ions (Figures 3c and 3e). In the case of O-protonated Neu5Ac, the acetamide group was contiguous to the CB[7] portal, and both Ha and Hb were capable of H-bond interactions. In contrast, the N-protonated acetamide group of Neu5Ac, with the exception of Hc, Hd, and the positively charged nitrogen, was far from the CB[7] portal because of the partially negatively charged carbonyl oxygen of the acetamide group. Representative structures of CB[7]-Neu5Gc complex ions show that the structural features of Neu5Gc in the CB[7] cavity differ slightly from those of Neu5Ac because of the presence of the C5 hydroxymethyl group (Figures 3c and 3e). In the case of O-protonated Neu5Gc, the two hydrogen bonds, N– Ha⋯O and O–Hb⋯O, hindered the interaction between CB[7] and the C5 hydroxymethyl group, which enhanced the structural stability of the complex ion. This structure induced a deeper insertion of the O-protonated Neu5Gc ion into the CB[7] cavity, obstructing the effective electrostatic interactions of the Neu5Gc ion and CB[7] at the opposite CB[7] portal. In the case of Nprotonated Neu5Gc, the C5 hydroxymethyl group interacted with the CB[7] portal through the U-shaped structure of the C5 functional group. The carbonyl oxygen in the U-shaped structure was located farther than that in the N-protonated Neu5Ac complex ion, which explained the stronger complexation preference for N-

protonated Neu5Gc. In summary, the abovementioned results show that there are distinct host-guest interactions between CB[7] and sialic acid ions that induce different complexation preferences and lead to the rigid complexation proportions having a linear relationship with their concentration proportions. Quantitative Analysis of Neu5Gc and Neu5Ac Using Gasphase Host-guest Chemistry. Based on our study of the unique host-guest chemistry between CB[7] host and sialic acid guests in the gas phase, which allowed us to facilitate a quantitative analysis of the guest by MS scanning, we developed a practical supramolecular MS method for identifying and quantifying Neu5Gc and Neu5Ac in glycoproteins. Herein, the utility of the method is demonstrated with three bovine glycoproteins, i.e., submaxillary mucin I-S (mucin), fetuin (bF), and transferrin (bT), and two human glycoproteins (α1-acid glycoprotein, hAGP; transferrin, hT). Sialic acids in cetuximab (Erbitux®, cet) and human erythropoietin expressed in HEK293 cells (hEPO) were also analyzed to expand the applicability of the method to glycoprotein therapeutics. Sialic acids in glycoproteins were extracted by acid hydrolysis and two separation steps using the solid phase extraction cartridge and reverse-phase chromatography (details available in the Experimental Section). Next, the supramolecular MS scanning of the sample was performed using CB[7] for analysis of the sialic acid. Figure 4 and Table S6 show the determined concentrations of the sialic acids in each glycoprotein investigated in the present study. It was found that the developed method could accurately determine the concentration of glycoprotein-bound Neu5Gc over a wide range of molar ratios with Neu5Ac (from 0.07:1 to 21.25:1 of Neu5Gc:Neu5Ac). The experiments with human glycoproteins showed that a single sialic acid in each glycoprotein could also be quantified using this unique host-guest chemistry. Overall, there was good agreement between the quantitation of the sialic acids in glycoproteins and the values reported in literature, which indicates

Figure 4. Quantification of sialic acids in glycoproteins using gas-phase host-guest chemistry. All proteins are abbreviated in this Figure. The µgsialic acid/mgprotein value of cetuximab (cet) and molsialic acid/molprotein value of bovine mucin I-S (mucin) were excluded because the mass of cet and molar concentration of mucin could not be obtained as a result of the solution state of the cet specimen and the unknown sequence of the mucin, respectively. The reference concentrations were obtained by averaging the previously reported concentrations.


Page 4 of 7

Page 5 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society excellent accuracy and precision of the analytical method (the details are available in the Supporting Information, Figure 4, and Table S6).20, 38-43 All investigations in the present study revealed that the developed supramolecular MS method could be successfully applied for the identification and quantification of Neu5Gc and Neu5Ac in their mixture simultaneously without any pre-derivatization step, separation routine, or tandem MS technique. The reverse phase LC-ESI-MS with multiple reactions monitoring (MRM) is a widely used simple MS technique for analysis of Neu5Gc in a glycoprotein without derivatization.27, 44 However, its false positives and negatives of Neu5Gc and Neu5Ac intensities in the LCMS spectra due to competitive ionization and fragmentation of Neu5Gc during the ESI process limit its practical application (the details are given in the Supporting Information, Figures S5–S9). The supramolecular MS method achieved sub-picomolar LOQs and a wide linear range from 5 to 9000 pmol without any bias of Neu5Gc during the ionization process. This is comparable to the quantification results of high performance anion exchange chromatography with pulsed amperometric detection (HPAECPAD),38-39 HPLC with fluorescence/UV-Vis spectroscopy,40-41 and cation exchange chromatography with UV-Vis spectroscopy (Table S2).42 Furthermore, this method could be practically used to assay Neu5Gc in glycoproteins with diverse molar ratios of Neu5Gc with Neu5Ac. Thus, the findings suggest that the supramolecular MS method is a highly convenient and accurate method for the analysis of Neu5Gc in diverse therapeutic glycoproteins.

CONCLUSION In this study, a practical application of the host-guest principle is demonstrated with the simple and accurate identification and quantification of Neu5Gc and Neu5Ac using the CB[7] host. The unique host-guest chemistry between CB[7] and sialic acids in the gas phase was utilized to analyze Neu5Gc in a glycoprotein with simple ESI-MS scanning. The distinct host-guest interactions of the sialic acids with CB[7] led to their rigid complexation proportions with a highly linear relationship with their concentration proportions in solution. This linear relationship was employed for developing a practical MS method for identifying and quantifying Neu5Gc and Neu5Ac in diverse glycoproteins with high sensitivity and simplicity. This approach provides excellent separation and quantitation of non-human glycan and can be applied for controlling immunogenicity-related risks in the development of biotherapeutics. Ultimately, it is anticipated that this study will broaden the scope of host-guest chemistry.

EXPERIMENTAL SECTION Materials. Standards of N-glycolylneuraminic acid (Neu5Gc) and N-acetylneuraminic acid (Neu5Ac) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Cucurbit[7]uril (CB[7]) and ammonium acetate (NH4OAc) were also purchased from Sigma-Aldrich (St. Louis, MO, USA). Formic acid (FA) and acetic acid (≥99%) were also purchased from Sigma-Aldrich (St. Louis, MO, USA). Hydrogen Chloride standard solution (6.0 M) were purchased from Honeywell Fluka (Morris Plains, NJ, USA). Monobasic and dibasic sodium phosphates were purchased from Biosesang (Seongnam, Gyeonggi, South Korea). Bovine mucin Ⅰ-S, bovine transferrin, human transferrin, human α1-acid glycoprotein, and erythropoietin expressed in HEK293 cells were purchased from Sigma-Aldrich (St. Louis, MO, USA). Cetuximab (Erbitux®) was donated from Korea University College of Medicine. Solid-phase extraction cartridge packed with porous graphitic carbon (HyperSep™ Hypercarb™ SPE Cartridge) was pur-

chased from ThermoFisher Scientific (Waltham, MA, USA). HPLC-grade water and acetonitrile (ACN) (Avantor Performance Materials, Center Valley, PA, USA) were used as the solvents. Electrospray Ionization Mass Spectrometry (ESI-MS). All MS experiments for determining sialic acids in their mixtures using host-guest chemistry were performed in positive ion mode using an Agilent 6560 ion mobility quadrupole time-of-flight (IMqTOF) mass spectrometer with an electrospray ionization (ESI) source. To ionize CB[7]-sialic acid complexes intact for the quantification of Neu5Gc and Neu5Ac, the gas temperature, drying gas flow rate, and nebulizer pressure were set to 325 Ⅰ, 3 L/min, and 20 psig, respectively. The sheath gas flow rate and temperature were set as 8 L/min and 100 Ⅰ, respectively. The VCap, nozzle voltage, fragmentor voltage, and Oct 1 RF Vpp were also set to 3500, 2000, 350, and 700 V, respectively. The IM experiments were performed under the drift cell filled with helium, and the drift cell pressure was set to 3.950 Torr. Computational Calculations and Calculation of Theoretical CCSs of Calculated Structures. All MD simulation results were obtained by using GROMACS 4.5.5. package.45-46 The forcefield parameters of CB[7], two sialic acids, and cations were employed from the CHARMM general force field (CGENFF) parameters.47 The initial structures of CB[7]-sialic acid complex ions were constructed by using Hyperchem 7.5 (Hypercube Inc., Gainesville, FL, USA). The molecular dynamics (MD) simulations were performed to obtain the lowest-energy conformations of the complex ions in the gas phase as candidate structures. Simulated annealing was performed between 300 and 600 K to prevent the confinement to a specific structure. The last frames of the 300 annealing cycles (90 ns) were extracted, and five structures for each complex ion were selected based on its potential energy. The selected structures were further optimized using density functional theory (DFT) calculations with the Q-Chem 4.1 computational package (Q-Chem Inc., Pittsburgh, PA, USA),48 utilizing Becke threeparameter functional (B3) combined with the correlation function of Lee, Yang, and Parr (LY), as well as the 6-31G(d) basis set.49-50 The theoretical collision cross sections (CCSs) of the complex ions were calculated using a modified trajectory (TJ) method developed by Lee et al. which has been suggested to be accurate regardless of the ion size.33 Briefly, this CCS calculation method combines an Exp-6 vdW interaction potential with the vdW interaction parameters of a molecular mechanics force field (MMFF94). The vdW interaction parameters are linearly scaled for accurate CCS predictions of various ions. Separation of Sialic Acids in Glycoproteins. We dissolved 0.2-1.0 mg glycoproteins in 120 µl of HPLC water. After adding 100 µl of 4M acetic acid solution into 100 µl of the protein solution, the solution was incubated at 80 Ⅰ for 3.5 h. The remaining protein solution (20 µl) was utilized for investigating the UV-Vis absorbance of each glycoprotein for the determination of the protein concentrations. The purification of the incubated solution was performed using PGC SPE cartridge. Before the injection of the protein solution, the cartridge was washed with 3 ml of 80% ACN containing 0.05% TFA and conditioned with 2 ml of HPLC water. After the sample injection, 1 ml of HPLC water, 10% ACN with 0.05% TFA, and 20% ACN with 0.05% TFA were sequentially injected into the cartridge. Each eluent was lyophilized, before dissolving in 100 µl of HPLC water. A portion of the solution was injected into LCMS-8580 equipped with a Shim-pack GIS C18 column (2 µm, 2.1 × 100 mm), and the sialic acids in the sample were fractionated. The injected volumes of sample solutions were varied depending on the concentrations of sialic acids in target proteins. 0.1% FA water (solvent A) and 0.1% FA ACN (solvent B) were used at a flow rate of 0.1 ml/min. A gradient of the two solvents were set as follows: 0–3.8 min, 0% B; 3.8–8.6 min, linear increase from 0 to 60% B; 8.6–10.6 min, 60% B; 10.6–11.6 min,



linear decrease from 60 to 0% B; 11.6–25.0 min, 0% B. Finally, the fractionated solutions were lyophilized, and were utilized for quantifying sialic acids in the target glycoprotein using our method.

ASSOCIATED CONTENT Supporting Information The supporting Information is available free of charge on the ACS Publications website. Additional discussions and additional experimental text with six tables and nine figures. (PDF)

AUTHOR INFORMATION Corresponding Author * [email protected] (H. J. A) * [email protected] (H. I. K)

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Notes The authors declare no competing financial interests.

ACKNOWLEDGMENT This work was supported by Basic Research Program No. 2016R1A2B4013089 through the National Research Foundation (NRF) of Korea, funded by the Ministry of Science, ICT, and Future Planning (MSIP), and Basic Science Research Program No. 20100020209 through the NRF of Korea funded by the Ministry of Education. This research was also supported by the National Research Council of Science & Technology (NST) grant by the Korea government (MSIP) (No. CAP-15-10-KRICT) and by a grant of Korea University Anam Hospital, Seoul, Republic of Korea (Grant No. K1620111). The authors acknowledge Agilent Technologies Inc. for 6560 LC-IMS QTOFMS instrument support and technical/scientific advisories.

REFERENCES (1)

Yang, Y.-W.; Sun, Y.-L.; Song, N. Accounts Chem. Res. 2014, 47 (7), 1950-1960. (2) Hsueh, S. Y.; Kuo, C. T.; Lu, T. W.; Lai, C. C.; Liu, Y. H.; Hsu, H. F.; Peng, S. M.; Chen, C. h.; Chiu, S. H. Angew. Chem. Int. Ed. 2010, 49 (48), 9170-9173. (3) Gomez-Casado, A.; Jonkheijm, P.; Huskens, J. Langmuir 2011, 27 (18), 11508-11513. (4) Minami, T.; Esipenko, N. A.; Zhang, B.; Kozelkova, M. E.; Isaacs, L.; Nishiyabu, R.; Kubo, Y.; Anzenbacher Jr, P. J. Am. Chem. Soc. 2012, 134 (49), 20021-20024. (5) Ko, Y. H.; Kim, E.; Hwang, I.; Kim, K. Chem. Commun. 2007, (13), 1305-1315. (6) Chen, Y.; Liu, Y. Chem. Soc. Rev. 2010, 39 (2), 495-505. (7) Park, K. M.; Lee, D. W.; Sarkar, B.; Jung, H.; Kim, J.; Ko, Y. H.; Lee, K. E.; Jeon, H.; Kim, K. Small 2010, 6 (13), 1430-1441. (8) Zhang, J.; Ma, P. X. Adv. Drug Deliv. Rev. 2013, 65 (9), 12151233. (9) Harada, A.; Kobayashi, R.; Takashima, Y.; Hashidzume, A.; Yamaguchi, H. Nat. Chem. 2011, 3 (1), 34. (10) Lee, D.-W.; Park, K. M.; Banerjee, M.; Ha, S. H.; Lee, T.; Suh, K.; Paul, S.; Jung, H.; Kim, J.; Selvapalam, N. Nat. Chem. 2011, 3 (2), 154.

(11) Schneider, H.-J.; Yatsimirsky, A. K. Chem. Soc. Rev. 2008, 37 (2), 263-277. (12) Biedermann, F.; Nau, W. M.; Schneider, H. J. Angew. Chem. Int. Ed. 2014, 53 (42), 11158-11171. (13) Hillyer, M. B.; Gibb, B. C. Annu. Rev. Phys. Chem. 2016, 67, 307329. (14) Lee, S. J. C.; Lee, J. W.; Lee, H. H.; Seo, J.; Noh, D. H.; Ko, Y. H.; Kim, K.; Kim, H. I. J. Phys. Chem. B 2013, 117 (29), 8855-8864. (15) Lee, J. W.; Lee, H. H. L.; Ko, Y. H.; Kim, K.; Kim, H. I. J. Phys. Chem. B 2015, 119 (13), 4628-4636. (16) Irie, T.; Uekama, K. Adv. Drug Deliv. Rev. 1999, 36 (1), 101-123. (17) Fokkens, M.; Schrader, T.; Klärner, F.-G. J. Am. Chem. Soc. 2005, 127 (41), 14415-14421. (18) Gibb, C. L. D.; Oertling, E. E.; Velaga, S.; Gibb, B. C. J. Phys. Chem. B 2015, 119 (17), 5624-5638. (19) Aggarwal, S. Nat. Biotechnol. 2014, 32, 32. (20) Ghaderi, D.; Taylor, R. E.; Padler-Karavani, V.; Diaz, S.; Varki, A. Nat. Biotechnol. 2010, 28 (8), 863-867. (21) Durocher, Y.; Butler, M. Curr. Opin. Biotechnol. 2009, 20 (6), 700707. (22) Higgins, E. Glycoconj. J. 2010, 27 (2), 211-225. (23) Tangvoranuntakul, P.; Gagneux, P.; Diaz, S.; Bardor, M.; Varki, N.; Varki, A.; Muchmore, E. Proc. Natl. Acad. Sci. U.S.A. 2003, 100 (21), 12045-12050. (24) Varki, N. M.; Varki, A. Lab. Investig. 2007, 87 (9), 851-857. (25) Borys, M. C.; Dalal, N. G.; AbuⅠAbsi, N. R.; Khattak, S. F.; Jing, Y.; Xing, Z.; Li, Z. J. Biotechnol. Bioeng. 2010, 105 (6), 10481057. (26) Bruggink, C.; Maurer, R.; Herrmann, H.; Cavalli, S.; Hoefler, F. J. Chromatogr. A 2005, 1085 (1), 104-109. (27) Hammad, L. A.; Saleh, M. M.; Novotny, M. V.; Mechref, Y. J. Am. Soc. Mass Spectrom. 2009, 20 (6), 1224-1234. (28) Mechref, Y. Electrophoresis 2011, 32 (24), 3467-3481. (29) Dong, X.; Xu, X.; Han, F.; Ping, X.; Yuang, X.; Lin, B. Electrophoresis 2001, 22 (11), 2231-2235. (30) Lee, H. H. L.; Kim, H. I. Isr. J. Chem. 2018. (31) Lee, H. H. L.; Lee, J. W.; Jang, Y.; Ko, Y. H.; Kim, K.; Kim, H. I. Angew. Chem. Int. Ed. 2016, 55 (29), 8249-8253. (32) Long, G. L.; Winefordner, J. D. Anal. Chem. 1983, 55 (7), 712A724A. (33) Lee, J. W.; Davidson, K. L.; Bush, M. F.; Kim, H. I. Analyst 2017, 142 (22), 4289-4298. (34) Gillespie, R.; Birchall, T. Can. J. Chem. 1963, 41 (1), 148-155. (35) Szostak, M.; Yao, L.; Day, V. W.; Powell, D. R.; Aubé, J. J. Am. Chem. Soc. 2010, 132 (26), 8836-8837. (36) Nanubolu, J. B.; Sridhar, B.; Ravikumar, K. CrystEngComm 2012, 14 (7), 2571-2578. (37) Hagemeister, F. C.; Gruenloh, C. J.; Zwier, T. S. J. Phys. Chem. A 1998, 102 (1), 82-94. (38) Rohrer, J. S.; Thayer, J.; Weitzhandler, M.; Avdalovic, N. Glycobiology 1998, 8 (1), 35-43. (39) Hurum, D. C.; Rohrer, J. S. Anal. Biochem. 2011, 419 (1), 67-69. (40) Hara, S.; Takemori, Y.; Yamaguchi, M.; Nakamura, M.; Ohkura, Y. Anal. Biochem. 1987, 164 (1), 138-145. (41) Karamanos, N. K.; Wikström, B.; Antonopoulos, C. A.; Hjerpe, A. J. Chromatogr. A 1990, 503, 421-429. (42) Honda, S.; Suzuki, S. Anal. Biochem. 1984, 142 (1), 167-174. (43) Suzuki, S. Chromatography 2014, 35 (1), 1-22. (44) Hammad, L. A.; Derryberry, D. Z.; Jmeian, Y. R.; Mechref, Y. Rapid Commun. Mass Spectrom. 2010, 24 (11), 1565-1574. (45) Hess, B.; Kutzner, C.; Van Der Spoel, D.; Lindahl, E. J. Chem. Theory Comput. 2008, 4 (3), 435-447. (46) Pronk, S.; Páll, S.; Schulz, R.; Larsson, P.; Bjelkmar, P.; Apostolov, R.; Shirts, M. R.; Smith, J. C.; Kasson, P. M.; van der Spoel, D. Bioinformatics 2013, 29 (7), 845-854. (47) Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.; Kundu, S.; Zhong, S.; Shim, J.; Darian, E.; Guvench, O.; Lopes, P.; Vorobyov, I. J. Comput. Chem. 2010, 31 (4), 671-690. (48) Shao, Y.; Molnar, L. F.; Jung, Y.; Kussmann, J.; Ochsenfeld, C.; Brown, S. T.; Gilbert, A. T.; Slipchenko, L. V.; Levchenko, S. V.; O’Neill, D. P. Phys. Chem. Chem. Phys. 2006, 8 (27), 3172-3191. (49) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37 (2), 785. (50) Becke, A. D. J. Chem. Phys 1993, 98, 5648-5652.


Page 6 of 7

Page 7 of 7 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society For Table of Contents,


7

Accurate Quantification of N-Glycolylneuraminic ... - ACS Publications

Recommend Documents