Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Reference‐free thio‐succinimide isomerization characterization by electron‐activated dissociation

Reference‐free thio‐succinimide isomerization characterization by electron‐activated dissociation INTRODUCTIONThiol‐maleimide conjugation is a highly popular reaction due to its specificity, selectivity, and fast kinetics.1 In recent decades, this reaction has been used in various fields, including biomolecule labelling2 and synthesis of novel polymeric materials.3 One of the most important applications of the thiol‐maleimide reaction is the bioconjugation between the linker‐payload and antibody backbone to generate antibody‐drug conjugates (ADCs).4,5 Comparing to well‐developed small molecule drugs, ADC shows promise as a new class of therapeutics, with the first ADC approved in 2001. As of October 2023, there are 15 different ADCs approved, and 10 of them utilize the thiol‐maleimide reaction during their conjugation.6There are two main competitive biotransformation processes occurring at the linker thio‐succinimide site. The reversible retro‐Michael reaction of thio‐succinimide leads to the pre‐mature linker‐payload deconjugation.7 Different approaches have been employed to avoid such problem.8–10 On the other hand, the thio‐succinimide is also subject to ring opening hydrolysis, and its product prevents the deconjugation, increasing the stability of ADCs.7 Notably, the hydrolysis of thio‐succinimide results in a pair of isomeric products, thio‐aspartyl (thio‐Asp) and thio‐isoaspartyl (thio‐isoAsp).11 This finding was confirmed by observations from several analytical techniques, including Fourier‐transform infrared spectroscopy (FTIR) and nuclear magnetic resonance (NMR).10 Nevertheless, to our knowledge, the differentiation of these isomers using liquid chromatography–mass spectrometry (LC–MS) has never been reported.Biotransformation of ADCs is predominantly characterized by intact protein mass analysis through liquid chromatography coupled with high resolution mass spectrometry (LC‐HRMS).12 Previously, we reported a workflow that uses hybrid ligand binding assay (LBA) coupled with intact protein mass LC‐HRMS to study ADC drug‐antibody ratio (DAR) changes and other biotransformations.13 However, the isomeric structures cannot be distinguished through the intact protein mass analysis due to their exact same mass. Furthermore, the reversed‐phase LC (RPLC) was not sufficient to separate large molecules with such subtle structural differences.12 Researchers also synthesized reference materials and compared their retention time (RT) to identify isomers.14 The complicated synthesis and purification processes compromise the efficiency of this analytical approach. Here, we report a bottom‐up LC–MS/MS approach with the help of electron‐activated dissociation (EAD) to distinguish thio‐succinimide hydrolysis isomeric products.Electron‐based dissociation (ExD) methods have a long history in aiding the characterization of biomolecules.15–17 The electron‐capture dissociation (ECD) method primarily cleaves peptide backbone,18 resulting in c and z ions,19 thus obtaining a higher sequence coverage compared with the traditional collision‐induced dissociation (CID).20,21 Besides, ECD leaves the side chain intact, making this dissociation method useful in identifying post‐translational modifications of proteins.22 Hot ECD (hECD), which employs higher kinetic energy (KE), enables secondary dissociation, which can, for instance, differentiate leucine and isoleucine.23 However, conventional ExD methods, such as ECD and electron‐transfer dissociation (ETD), mainly dissociate multiply charged ions,20 leaving singly or doubly charged ions less accessible to dissociation. Recently, EAD, which comprises of ECD and hECD, was shown to be flexible in dissociating ions with a wide range of charge states.24 The enzymatic digestion of ADCs yields peptides with various lengths and small molecule linker‐payloads, making EAD highly valuable for the characterization of ADCs, where the dissociation of differently protonated species may be achieved.Thio‐succinimide hydrolysis is analogous to protein deamidation. The asparagine deamidation generates a succinimide intermediate before the addition of water.25 The resulting aspartic acid leads to ~1 Da molecular weight increase compared with asparagine's original mass. Similar to thio‐succinimide hydrolysis, the asparagine deamidation also generates a pair of isomeric products, aspartic acid, and iso‐aspartic acid. We previously reported the identification of asparagine deamidation at the complementarity‐determining regions of another ADC, MEDI7247, employing LC–MS/MS.26 The diagnostic ions were c + 57 and z − 57 coming from the iso‐aspartic acid after dissociation between α‐ and β‐carbons through EAD.27 Here, we investigated the feasibility of employing a similar strategy (+57 u or −57 u related signature ions) to differentiate thio‐Asp and thio‐isoAsp, notwithstanding the slight structure difference compared with aspartic acid and iso‐aspartic acid. Moreover, we further explored if EAD could generate more unique diagnostic ions only for thio‐succinimide hydrolysis products. To our knowledge, we are the first to report an in‐depth characterization of thio‐succinimide hydrolysis isomerization using a bottom‐up LC–MS/MS approach without requiring reference materials.METHODSMaterialsADC1 and anti‐idiotype (anti‐ID) antibodies were generated in‐house by AstraZeneca (Gaithersburg, MD). The SMART IA magnetic beads, EZ‐LINK Sulfo‐NHS‐LC‐Biotin biotinylation kit, Zeba desalting spin columns, LC–MS grade formic acid (FA), and tris (hydroxymethyl)aminomethane hydrochloride (Tris–HCl) buffer (pH 7.5) were acquired from ThermoFisher Scientific (Waltham, MA). The Acquity UPLC BEH C18 columns and Oasis HLB solid phase extraction 96‐well plates were purchased from Waters (Milford, MA). LC–MS grade acetonitrile (ACN), water, methanol, and bovine serum albumin (BSA) were acquired from Sigma‐Aldrich (St. Louis, MO). Pooled human and CD1 mouse plasma were acquired from BioIVT (Hicksville, NY). Eppendorf Protein LoBind tubes, Deepwell 96‐well plates, Lonza deionized (DI) water, and all other general reagents and supplies were purchased from VWR Scientific (Radnor, PA).InstrumentationLC–MS/MS experiments were performed on ZenoTOF 7600 mass spectrometer (SCIEX, Toronto), coupled with Exion UHPLC system. The mass spectra were analyzed using a research version of PeakView (SCIEX, version number: 1.2.2.0).ADC1 incubation, immuno‐affinity enrichment, and sample clean‐upFigure S1 shows a brief description of the experiment steps of sample preparation for analyzing thio‐succinimide hydrolysis isomers. A total of 20 mg/mL ADC1 stock solution was diluted to 1 mg/mL working solution using DI water. A total of 2250 μL of human or CD1 mouse plasma was added into a 5 mL Protein LoBind tube containing 250 μL of 1 mg/mL ADC1 working solution. The solution was briefly vortexed before being split into 300 μL aliquots. The final concentration of ADC1 was 0.1 mg/mL. Three aliquots were incubated at 37°C for 168 h using digital HeatBlock (VWR) and then stored at −80°C until use, whereas three other aliquots were immediately stored at −80°C after solution preparation (denoted as 0 h).The biotinylation of the anti‐ID antibody was performed using an EZ‐LINK biotinylation kit following the manufacturer's protocol. The biotinylated anti‐ID antibody was purified using a Zeba 7 kDa molecular weight cut‐off desalting spin column before being stored at −80°C. The final concentration of biotinylated anti‐ID antibody was estimated to be 8.48 mg/mL. To ensure sufficient capture of ADC1 in plasma incubated samples, the ratio between biotinylated anti‐ID antibody and SMART IA magnetic beads was kept at 1:10 (the mass of biotinylated anti‐ID antibody in μg to the volume of SMART IA magnetic beads in mL). The tube that contains the solution mixture was rotated for 30 min at room temperature using a LabQuake shaker (Barnstead) before loading onto a magnetic tube rack. SMART IA magnetic beads were separated from the solution, and the supernatant was extracted and discarded. The SMART IA beads were washed three times with SN1 buffer (50 mM Tris–HCl in DI water containing 1 mg/mL BSA, pH 7.5) with the same volume as the original bead slurry. The anti‐ID conjugated SMART IA beads were stored at 4°C until use.A total of 1000 μL of SN1 buffer were added to a new 5 mL Protein LoBind tube, followed by mixing with 300 μL of plasma incubated ADC1 sample. The mixture was briefly vortexed before adding 600 μL of conjugated SMART IA beads to the tube. The tube was taped onto a Thermomixer C (Eppendorf) and shaken at 25°C and 1200 rpm for 2 h. After immuno‐affinity enrichment, the tube was put onto the magnetic tube rack, and the supernatant was discarded. The beads were washed three times using 1900 μL of SN2 buffer (50 mM Tris–HCl in DI water, pH 7.5) before 1500 μL SMART IA digestion buffer (provided in the kit) was added to the washed beads. To activate the pre‐immobilized trypsin on SMART IA beads, the mixture was shaken at 70°C and 1200 rpm using Thermomixer C (Eppendorf) for 2 h. After tryptic digestion, the tube was placed onto the magnetic tube rack, where the digested ADC1 solution was transferred into a new tube. The sample clean‐up was performed using an Oasis HLB cartridge (Waters) following the manufacturer's protocol. The post clean‐up sample was diluted to 120 μL using 0.1% FA in water, and 30 μL of sample was loaded into the LC–MS system.LC–MS/MS parametersThe separation of tryptic digested peptides was performed on a BEH C18 UPLC column (Waters, 1.7 μm, 2.1 × 50 mm) using 0.1% FA in water as mobile phase A (MPA) and 0.1% FA in ACN as mobile phase B (MPB). The ZenoTOF 7600 mass spectrometer was operated in full‐scan MS (m/z 100–1500) with collision energy (CE) set as 10 V. For MS/MS, CE = 40 V was applied to CID, whereas the KE for EAD was set as 11 eV. The LC gradient and other parameters are described in the Section S2 in detail.RESULTSCharacterization of thio‐succinimide hydrolysisThiol‐maleimide reaction is widely used for bioconjugation purposes. The resulted thio‐succinimide structure is vulnerable to ring‐opening hydrolysis. ADC1 is a thio‐succinimide‐linked ADC, produced from reducing interchain disulfide bonds, followed by covalently conjugating linker‐payload to the thiol groups from cysteine side chains using maleimide chemistry. To identify and characterize its hydrolysis isomers, we used a signature ADC1 tryptic digested product, which is composed of a tripeptide GEC, succinimide (with or without water adduct), and R2 to demonstrate the process. Figure 1 shows the hydrolysis reaction and potential diagnostic ions for differentiating hydrolysis isomers, using GEC + succinimide + R2 as an example.1FIGUREScheme of thio‐succinimide hydrolysis mechanism and the proposed structures of product ions after electron‐activated dissociation (EAD). Product ions in the red boxes are diagnostic ions with predominant charge state indicated. R1 is a tripeptide GEC, and R2 is linker‐payload. The linker‐payload is covalently conjugated to the antibody via a cysteine side chain thiol. Succ + H2O = hydrolyzed succinimide. [Color figure can be viewed at wileyonlinelibrary.com]Figure 2A shows the mass spectra of [GEC + succinimide + R2 + 3H]3+ of ADC1 with 0 h (black line) and 168 h (blue line) incubation. After 168 h incubation, the signal intensity of [GEC + succinimide + R2 + 3H]3+ (observed monoisotopic m/z 485.8744, −11 ppm difference) dropped significantly compared with the sample with 0 h incubation. To be specific, the absolute intensity of the monoisotopic peak of [GEC + succinimide + R2 + 3H]3+ dropped from 2.9 × 104 to 7.2 × 102 cps after 168 h of incubation. Instead, we noticed a notable increase in peaks representing GEC + succinimide + R2 with +18 Da shift (observed monoisotopic m/z 491.8779, z = 3, −11 ppm difference). Thus, the hydrolysis of succinimide over 168 h was confirmed. Figure 2B is the extracted ion chromatogram (XIC) of [GEC + succinimide + R2 + 3H]3+ (m/z 485.8799) in 0 and 168 h incubated samples. Figure 2C shows the XIC of [GEC + succinimide + H2O + R2 + 3H]3+ (m/z 491.8834) at different incubation times. Similarly, the XICs support the conclusion that most of the succinimide conjugated to GEC is hydrolyzed over 168 h of incubation. Table S2 gives theoretical and observed monoisotopic m/z values of the discussed ions.2FIGURE(A) Overlaid mass spectra of [GEC + succinimide + R2 + 3H]3+ and [GEC + succinimide + H2O + R2 + 3H]3+ from 0 h (black line) and 168 h (blue line) plasma incubated antibody‐drug conjugate 1 (ADC1) samples. Both species observed a charge state of 3. Overlaid extracted ion chromatograms (XICs) of (B) [GEC + succinimide + R2 + 3H]3+ or (C) [GEC + succinimide + H2O + R2 + 3H]3+. Black line represents the sample with no incubation, whereas the blue line is the sample incubated for 168 h. [Color figure can be viewed at wileyonlinelibrary.com]Identification of thio‐succinimide hydrolysis isomeric productsThe hydrolysis of thio‐succinimide creates a pair of constitutional isomeric products, as shown in Figure 2A and reported previously. In Figure 2C, two distinctive peaks with different RTs observed in XIC confirm the presence of GEC + succinimide + H2O + R2 isomers. The peak with RT = 45.9 min is denoted as Peak 1, whereas the peak with RT = 46.4 min is denoted as Peak 2. In order to elucidate the chemical structure corresponding to each peak, we employed two orthogonal dissociation methods, CID and EAD, on [GEC + succinimide + H2O + R2 + 3H]3+. Figure S2A,B show the CID MS/MS spectra of precursor ions from Peaks 1 and 2, respectively. However, we did not find any distinctive peaks by comparing these two spectra. Hence, CID does not provide useful information in distinguishing thio‐succinimide hydrolysis isomers.Differentiation between aspartic acid and iso‐aspartic acid from asparagine deamidation has been previously reported using ExD methods, such as ETD and ECD. c + 57 and z − 57 were confirmed to be a pair of diagnostic ions to identify iso‐aspartic acid generated from either ETD or ECD.26,28 Asparagine deamidation first generates a succinimide intermediate prior to ring‐opening hydrolysis, which is analogous to thio‐succinimide hydrolysis as shown in Figure 2A. Figure S3 shows the mechanism of asparagine deamidation and structures of diagnostic c‐ and z‐related ions. Figure S4 shows the dissociation sites and the structures of c‐ and z‐related ions in both deamidation and thio‐succinimide hydrolysis isomers. Figure 3A,B are the MS/MS spectra (m/z range: 100–2900) of [GEC + succinimide + H2O + R2 + 3H]3+ precursor ions from Peaks 1 and 2, respectively, using EAD. The product ions in both figures show high agreement with each other. (I), (II), and (III) in Figure 3A,B are zoomed‐in MS/MS spectra with different m/z range. Comparing Figures 3A (I) and 3B (I), we found a singly charged product ion with monoisotopic m/z of 365.0839, which is in accordance with [R1 + Thio + 57 + H]+ diagnostic ion with a −15 ppm mass difference. Similarly, by comparing Figures 3A (III) and 3B (III), we found a singly charged product ion with a monoisotopic m/z of 1109.5403. This peak suggests the presence of [R2 + Succ + H2O‐57 + H]+ diagnostic ion with −12 ppm mass difference. Thus, we can confirm Figure 3A represents the MS/MS spectrum of thio‐Asp from Peak 1, whereas Figure 3B is the MS/MS spectrum of thio‐isoAsp from Peak 2. Observed m/z values of discussed ions all show a similar extent of deviation to their theoretical values. A recalibration performed after acquisition corrects the m/z difference to within ±3 ppm (see Section S3). Hence, the identification of ions of interest is accurate, despite the mass errors being close to the high end of tolerance.3FIGUREElectron‐activated dissociation (EAD) MS/MS spectra (m/z 100–1500) of (A) Peak 1 and (B) Peak 2 observed in Figure 2C. Signature product ions are labelled on both spectra. Pink arrows point to a doubly charged diagnostic ion, [R2 + Succ + H2O − 44 + 2H]2+. Dashed arrows point to zoomed‐in EAD MS/MS spectra with different m/z ranges. (i) m/z 363–369, (ii) m/z 572–582, and (iii) m/z 1105–1115. Blue boxes are the zoomed‐in spectra of m/z 1410–1440 (see Figure S7). The stars with 4, 5, and 6 arms in inset figures correspond to the monoisotopic peak of each diagnostic product ion, respectively. Their observed m/z values are 365.0839, 577.7643, and 1109.5403, respectively. [R2 + Succ + H2O − 44 + 2H]2+ ion was also observed in Figure 3A (II) with −12 ppm difference to its theoretical value, although the intensity is low. [Color figure can be viewed at wileyonlinelibrary.com]Another notable difference is that [R2 + Succ + H2O − 44 + 2H]2+ (indicated by the pink arrow) shows significantly higher intensity in Figure 3A, whereas the intensity of the same product ion in Figure 3B is low. Figure S5A,B show zoomed‐in MS/MS spectra from both Peaks 1 and 2 at m/z 555–570 range, respectively. The higher signal intensity of [R2 + Succ + H2O − 44 + 2H]2+ in Figure S5A suggests the CO2 loss after forming the [R2 + Succ + H2O + 2H]2+ ion of thio‐Asp is significantly enhanced. This finding aligns with the ExD dissociation results of aspartic acid in deamidation.We also observed a doubly‐charged product ion with a monoisotopic m/z of 561.2728 with enhanced intensity in Figure 3A (II), but low intensity in Figure 3B (II). We propose the structure of thio + succinimide + H2O + R2 − CO2 (Figure 1). We denote this product ion as [R2 + ThioSucc + H2O − 44 + 2H]2+. Although both thio‐Asp and thio‐isoAsp are capable of CO2 loss after forming the [R2 + ThioSucc + H2O + 2H]2+ product ion, clearly, thio‐isoAsp is more prone to this neutral loss. We also found the [R2 + ThioSucc + H2O + 2H]2+ product ion in both thio‐Asp and thio‐isoAsp with similar intensities, as shown in Figure S6A,B. Thus, [R2 + ThioSucc + H2O − 44 + 2H]2+ product ion can serve as an additional diagnostic ion for differentiating thio‐Asp and thio‐isoAsp. To our knowledge, this product ion has never been reported in the context of differentiation of deamidation isomers. Hence, this diagnostic ion is distinctive of the thio‐succinimide system. We infer that the generation of this unique product ion is related to the thiol group linked with the succinimide ring and R2.Neutral losses are not limited to the product ions generated from ExD dissociation. [M − 60 + H]+ was reported to be a diagnostic ion for Asp in deamidation.29 Figure S7A,B show the zoomed‐in MS/MS spectra of Figure 3A,B at m/z 1410–1440. To our surprise, the [M − 60 + H]+ product ions (observed monoisotopic m/z 1414.6355, 10 ppm difference to its theoretical value) in both Figure S7A,B show similar intensities. Thus, [M − 60 + H]+ cannot differentiate thio‐Asp and thio‐isoAsp, and we think the 60 Da neutral loss might come from the contribution of R2. We also observed peaks with monoisotopic m/z 1430.6428 (−7 ppm difference to theoretical m/z of [M − 44 + H]+) in both isomers. The similar intensities of [M − 44 + H]+ and [M − 60 + H]+ suggest that neutral losses to the whole precursor ion cannot distinguish thio‐Asp and thio‐isoAsp.Furthermore, we also investigated another tryptic peptide, SCDK, that contains succinimide + R2 from ADC1, followed by the same sample preparation steps. The SCDK + succinimide + R2 is mostly hydrolyzed after 168 h of incubation, as shown in Figure S8A. Both XICs of non‐hydrolyzed and hydrolyzed at 0 and 168 h in Figure S8B,C support the same conclusion. Additionally, Figure S8C shows two distinct peaks that were well separated under the current LC gradient. The presence of two peaks suggests the formation of isomers after thio‐succinimide hydrolysis. This finding aligns with our observations of GEC‐linked succinimide structure.CONCLUSIONThe formation of thio‐succinimide hydrolysis isomeric products is well known, however, the differentiation between them has not been sufficiently studied so far. This work demonstrated a reference material‐free characterization of hydrolyzed thio‐succinimide isomers using EAD. Similar to the differentiation between aspartic acid and iso‐aspartic acid generated from asparagine deamidation, we observed distinctive [R1 + Thio + 57 + H]+, [R2 + Succ + H2O − 57 + H]+ for thio‐isoAsp and enhanced [R2 + Succ + H2O − 44 + 2H]2+ for thio‐Asp in EAD MS/MS spectra. However, [M − 44 + H]+ and [M − 60 + H]+ failed to serve as diagnostic ions because of their similar intensities in EAD MS/MS spectra for both thio‐Asp and thio‐isoAsp. To our surprise, we observed a significantly enhanced CO2 neutral loss after formation of the [R2 + ThioSucc + H2O + 2H]2+ product ion for thio‐isoAsp. This product ion is denoted as [R2 + ThioSucc + H2O − 44 + 2H]2+, and for the very first time, we report it as a unique diagnostic ion for differentiating thio‐Asp and thio‐isoAsp. To our knowledge, an analogous product ion has not been identified for asparagine deamidation. The formation of this product ion might relate to the thiol group linked to the succinimide ring and the structures of R2. For instance, in the case of deamidation, the succinimide ring is linked with amino acids at both ends. Compared with the conventional approach, which matches RT of purified reference material to distinguish isomers, EAD serves as a novel tool to identify isomeric structures without reference standards. The differentiation between thio‐Asp and thio‐isoAsp may eventually benefit the drug development where thiol‐maleimide conjugation is employed.AUTHOR CONTRIBUTIONSJunyan Yang: Conceptualization; methodology; investigation; formal analysis; writing—original draft; visualization. Jiaqi Yuan: Conceptualization; methodology; investigation; draft review. Yue Huang: Conceptualization; methodology; draft review. Anton I. Rosenbaum: Conceptualization; methodology; investigation; draft review; supervision.ACKNOWLEDGMENTSThis study was funded by AstraZeneca. We are deeply grateful to Haichuan Liu and Zoe Zhang from SCIEX for providing help with LC–MS data analysis, insightful discussions, and advice.CONFLICT OF INTEREST STATEMENTJ. Yang, J. Yuan, Y. H., and A. I. R. are or were employees of AstraZeneca at the time this work was conducted and may hold stock and/or stock options or interests in the company.DATA AVAILABILITY STATEMENTThe data that supports the findings of this study are available in the supplementary material of this article.REFERENCESNorthrop BH, Frayne SH, Choudhary U. Thiol–maleimide “click” chemistry: evaluating the influence of solvent, initiator, and thiol on the reaction mechanism, kinetics, and selectivity. Polym Chem. 2015;6(18):3415‐3430. doi:10.1039/c5py00168dRenault K, Fredy JW, Renard PY, Sabot C. Covalent modification of biomolecules through maleimide‐based labeling strategies. Bioconjug Chem. 2018;29(8):2497‐2513. doi:10.1021/acs.bioconjchem.8b00252Summonte S, Racaniello GF, Lopedota A, Denora N, Bernkop‐Schnurch A. Thiolated polymeric hydrogels for biomedical application: cross‐linking mechanisms. J Control Release. 2021;330:470‐482. doi:10.1016/j.jconrel.2020.12.037Ferhati X, Jimenez‐Moreno E, Hoyt EA, et al. Single mutation on trastuzumab modulates the stability of antibody‐drug conjugates built using acetal‐based linkers and thiol‐maleimide chemistry. J am Chem Soc. 2022;144(12):5284‐5294. doi:10.1021/jacs.1c07675Su Z, Xiao D, Xie F, et al. Antibody‐drug conjugates: recent advances in linker chemistry. Acta Pharm Sin B. 2021;11(12):3889‐3907. doi:10.1016/j.apsb.2021.03.042BroadPharm. ADC Approval up to 2023. https://broadpharm.com/blog/ADC-Approval-up-to-2023 (Accessed May 3, 2024)Wei C, Zhang G, Clark T, et al. Where did the linker‐payload go? A quantitative investigation on the destination of the released linker‐payload from an antibody‐drug conjugate with a maleimide linker in plasma. Anal Chem. 2016;88(9):4979‐4986. doi:10.1021/acs.analchem.6b00976Lahnsteiner M, Kastner A, Mayr J, Roller A, Keppler BK, Kowol CR. Improving the stability of maleimide‐thiol conjugation for drug targeting. Chemistry. 2020;26(68):15867‐15870. doi:10.1002/chem.202003951Christie RJ, Fleming R, Bezabeh B, et al. Stabilization of cysteine‐linked antibody drug conjugates with N‐aryl maleimides. J Control Release. 2015;220(Pt B):660‐670. doi:10.1016/j.jconrel.2015.09.032Huang W, Wu X, Gao X, et al. Maleimide‐thiol adducts stabilized through stretching. Nat Chem. 2019;11(4):310‐319. doi:10.1038/s41557‐018‐0209‐2Fontaine SD, Reid R, Robinson L, Ashley GW, Santi DV. Long‐term stabilization of maleimide‐thiol conjugates. Bioconjug Chem. 2015;26(1):145‐152. doi:10.1021/bc5005262Huang Y, Tan HY, Yuan J, et al. Extensive biotransformation profiling of AZD8205, an anti‐B7‐H4 antibody‐drug conjugate, elucidates pathways underlying its stability in vivo. Anal Chem. 2024. doi:10.1021/acs.analchem.4c02309Huang Y, Mou S, Wang Y, Mu R, Liang M, Rosenbaum AI. Characterization of antibody‐drug conjugate pharmacokinetics and in vivo biotransformation using quantitative intact LC‐HRMS and surrogate analyte LC‐MRM. Anal Chem. 2021;93(15):6135‐6144. doi:10.1021/acs.analchem.0c05376Stalcup AM. Chiral separations. Annu Rev Anal Chem. 2010;3(1):341‐363. doi:10.1146/annurev.anchem.111808.073635Xie Y, Zhang J, Yin S, Loo JA. Top‐down ESI‐ECD‐FT‐ICR mass spectrometry localizes noncovalent protein‐ligand binding sites. J am Chem Soc. 2006;128(45):14432‐14433. doi:10.1021/ja063197pZubarev RA, Horn DM, Fridriksson EK, et al. Electron capture dissociation for structural characterization of multiply charged protein cations. Anal Chem. 2000;72(3):563‐573. doi:10.1021/ac990811pChi A, Huttenhower C, Geer LY, et al. Analysis of phosphorylation sites on proteins from saccharomyces cerevisiae by electron transfer dissociation (ETD) mass spectrometry. Proc Natl Acad Sci U S a. 2007;104(7):2193‐2198. doi:10.1073/pnas.0607084104Sze SK, Ge Y, Oh H, McLafferty FW. Plasma electron capture dissociation for the characterization of large proteins by top down mass spectrometry. Anal Chem. 2003;75(7):1599‐1603. doi:10.1021/ac020446tRoepstorff P, Fohlman J. Proposal for a common nomenclature for sequence ions in mass spectra of peptides. Biomed Mass Spectrom. 1984;11(11):601. doi:10.1002/bms.1200111109Cooper HJ, Hakansson K, Marshall AG. The role of electron capture dissociation in biomolecular analysis. Mass Spectrom Rev. 2005;24(2):201‐222. doi:10.1002/mas.20014Fort KL, Cramer CN, Voinov VG, et al. Exploring ECD on a benchtop Q exactive orbitrap mass spectrometer. J Proteome Res. 2018;17(2):926‐933. doi:10.1021/acs.jproteome.7b00622Kelleher NL, Zubarev RA, Bush K, et al. Localization of labile posttranslational modifications by electron capture dissociation: the case of γ‐carboxyglutamic acid. Anal Chem. 1999;71(19):4250‐4253. doi:10.1021/ac990684xKjeldsen F, Haselmann KM, Budnik BA, Jensen F, Zubarev RA. Dissociative capture of hot (3‐13 eV) electrons by polypeptide polycations: An efficient process accompanied by secondary fragmentation. Chem Phys Lett. 2002;356(3‐4):201‐206. doi:10.1016/S0009‐2614(02)00149‐5SCIEX. Electron Activated Dissociation (EAD). https://sciex.com/content/dam/SCIEX/pdf/brochures/zenotof-7600-whitepaper-ead.pdf (Accessed July 18, 2024)Irudayanathan FJ, Zarzar J, Lin J, Izadi S. Deciphering deamidation and isomerization in therapeutic proteins: effect of neighboring residue. MAbs. 2022;14(1):2143006. doi:10.1080/19420862.2022.2143006Huang Y, Yuan J, Mu R, et al. Multiplex bioanalytical methods for comprehensive characterization and quantification of the unique complementarity‐determining‐region deamidation of MEDI7247, an anti‐ASCT2 pyrrolobenzodiazepine antibody‐drug conjugate. Antibodies (Basel). 2023;12(4). doi:10.3390/antib12040066Liu H, Zhang Z. Comprehensive differentiation of deamidation isomers from forced degradation by electron activation dissociation (EAD); SCIEX. https://sciex.com/tech-notes/biopharma/comprehensive-differentiation-of-deamidation-isomers-from-forced-degradation-by-electron-activation-dissociation-ead (Accessed May 3, 2024)Yang H, Zubarev RA. Mass spectrometric analysis of asparagine deamidation and aspartate isomerization in polypeptides. Electrophoresis. 2010;31(11):1764‐1772. doi:10.1002/elps.201000027Sargaeva NP, Lin C, O'Connor PB. Differentiating N‐terminal aspartic and Isoaspartic acid residues in peptides. Anal Chem. 2011;83(17):6675‐6682. doi:10.1021/ac201223d http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Rapid Communications in Mass Spectrometry Wiley

Reference‐free thio‐succinimide isomerization characterization by electron‐activated dissociation

Download PDF
7 pages

Loading next page...
 
/lp/wiley/reference-free-thio-succinimide-isomerization-characterization-by-01bowc0oma

References (33)

Publisher
Wiley
Copyright
© 2024 John Wiley & Sons, Ltd.
ISSN
0951-4198
eISSN
1097-0231
DOI
10.1002/rcm.9910
Publisher site
See Article on Publisher Site

Abstract

INTRODUCTIONThiol‐maleimide conjugation is a highly popular reaction due to its specificity, selectivity, and fast kinetics.1 In recent decades, this reaction has been used in various fields, including biomolecule labelling2 and synthesis of novel polymeric materials.3 One of the most important applications of the thiol‐maleimide reaction is the bioconjugation between the linker‐payload and antibody backbone to generate antibody‐drug conjugates (ADCs).4,5 Comparing to well‐developed small molecule drugs, ADC shows promise as a new class of therapeutics, with the first ADC approved in 2001. As of October 2023, there are 15 different ADCs approved, and 10 of them utilize the thiol‐maleimide reaction during their conjugation.6There are two main competitive biotransformation processes occurring at the linker thio‐succinimide site. The reversible retro‐Michael reaction of thio‐succinimide leads to the pre‐mature linker‐payload deconjugation.7 Different approaches have been employed to avoid such problem.8–10 On the other hand, the thio‐succinimide is also subject to ring opening hydrolysis, and its product prevents the deconjugation, increasing the stability of ADCs.7 Notably, the hydrolysis of thio‐succinimide results in a pair of isomeric products, thio‐aspartyl (thio‐Asp) and thio‐isoaspartyl (thio‐isoAsp).11 This finding was confirmed by observations from several analytical techniques, including Fourier‐transform infrared spectroscopy (FTIR) and nuclear magnetic resonance (NMR).10 Nevertheless, to our knowledge, the differentiation of these isomers using liquid chromatography–mass spectrometry (LC–MS) has never been reported.Biotransformation of ADCs is predominantly characterized by intact protein mass analysis through liquid chromatography coupled with high resolution mass spectrometry (LC‐HRMS).12 Previously, we reported a workflow that uses hybrid ligand binding assay (LBA) coupled with intact protein mass LC‐HRMS to study ADC drug‐antibody ratio (DAR) changes and other biotransformations.13 However, the isomeric structures cannot be distinguished through the intact protein mass analysis due to their exact same mass. Furthermore, the reversed‐phase LC (RPLC) was not sufficient to separate large molecules with such subtle structural differences.12 Researchers also synthesized reference materials and compared their retention time (RT) to identify isomers.14 The complicated synthesis and purification processes compromise the efficiency of this analytical approach. Here, we report a bottom‐up LC–MS/MS approach with the help of electron‐activated dissociation (EAD) to distinguish thio‐succinimide hydrolysis isomeric products.Electron‐based dissociation (ExD) methods have a long history in aiding the characterization of biomolecules.15–17 The electron‐capture dissociation (ECD) method primarily cleaves peptide backbone,18 resulting in c and z ions,19 thus obtaining a higher sequence coverage compared with the traditional collision‐induced dissociation (CID).20,21 Besides, ECD leaves the side chain intact, making this dissociation method useful in identifying post‐translational modifications of proteins.22 Hot ECD (hECD), which employs higher kinetic energy (KE), enables secondary dissociation, which can, for instance, differentiate leucine and isoleucine.23 However, conventional ExD methods, such as ECD and electron‐transfer dissociation (ETD), mainly dissociate multiply charged ions,20 leaving singly or doubly charged ions less accessible to dissociation. Recently, EAD, which comprises of ECD and hECD, was shown to be flexible in dissociating ions with a wide range of charge states.24 The enzymatic digestion of ADCs yields peptides with various lengths and small molecule linker‐payloads, making EAD highly valuable for the characterization of ADCs, where the dissociation of differently protonated species may be achieved.Thio‐succinimide hydrolysis is analogous to protein deamidation. The asparagine deamidation generates a succinimide intermediate before the addition of water.25 The resulting aspartic acid leads to ~1 Da molecular weight increase compared with asparagine's original mass. Similar to thio‐succinimide hydrolysis, the asparagine deamidation also generates a pair of isomeric products, aspartic acid, and iso‐aspartic acid. We previously reported the identification of asparagine deamidation at the complementarity‐determining regions of another ADC, MEDI7247, employing LC–MS/MS.26 The diagnostic ions were c + 57 and z − 57 coming from the iso‐aspartic acid after dissociation between α‐ and β‐carbons through EAD.27 Here, we investigated the feasibility of employing a similar strategy (+57 u or −57 u related signature ions) to differentiate thio‐Asp and thio‐isoAsp, notwithstanding the slight structure difference compared with aspartic acid and iso‐aspartic acid. Moreover, we further explored if EAD could generate more unique diagnostic ions only for thio‐succinimide hydrolysis products. To our knowledge, we are the first to report an in‐depth characterization of thio‐succinimide hydrolysis isomerization using a bottom‐up LC–MS/MS approach without requiring reference materials.METHODSMaterialsADC1 and anti‐idiotype (anti‐ID) antibodies were generated in‐house by AstraZeneca (Gaithersburg, MD). The SMART IA magnetic beads, EZ‐LINK Sulfo‐NHS‐LC‐Biotin biotinylation kit, Zeba desalting spin columns, LC–MS grade formic acid (FA), and tris (hydroxymethyl)aminomethane hydrochloride (Tris–HCl) buffer (pH 7.5) were acquired from ThermoFisher Scientific (Waltham, MA). The Acquity UPLC BEH C18 columns and Oasis HLB solid phase extraction 96‐well plates were purchased from Waters (Milford, MA). LC–MS grade acetonitrile (ACN), water, methanol, and bovine serum albumin (BSA) were acquired from Sigma‐Aldrich (St. Louis, MO). Pooled human and CD1 mouse plasma were acquired from BioIVT (Hicksville, NY). Eppendorf Protein LoBind tubes, Deepwell 96‐well plates, Lonza deionized (DI) water, and all other general reagents and supplies were purchased from VWR Scientific (Radnor, PA).InstrumentationLC–MS/MS experiments were performed on ZenoTOF 7600 mass spectrometer (SCIEX, Toronto), coupled with Exion UHPLC system. The mass spectra were analyzed using a research version of PeakView (SCIEX, version number: 1.2.2.0).ADC1 incubation, immuno‐affinity enrichment, and sample clean‐upFigure S1 shows a brief description of the experiment steps of sample preparation for analyzing thio‐succinimide hydrolysis isomers. A total of 20 mg/mL ADC1 stock solution was diluted to 1 mg/mL working solution using DI water. A total of 2250 μL of human or CD1 mouse plasma was added into a 5 mL Protein LoBind tube containing 250 μL of 1 mg/mL ADC1 working solution. The solution was briefly vortexed before being split into 300 μL aliquots. The final concentration of ADC1 was 0.1 mg/mL. Three aliquots were incubated at 37°C for 168 h using digital HeatBlock (VWR) and then stored at −80°C until use, whereas three other aliquots were immediately stored at −80°C after solution preparation (denoted as 0 h).The biotinylation of the anti‐ID antibody was performed using an EZ‐LINK biotinylation kit following the manufacturer's protocol. The biotinylated anti‐ID antibody was purified using a Zeba 7 kDa molecular weight cut‐off desalting spin column before being stored at −80°C. The final concentration of biotinylated anti‐ID antibody was estimated to be 8.48 mg/mL. To ensure sufficient capture of ADC1 in plasma incubated samples, the ratio between biotinylated anti‐ID antibody and SMART IA magnetic beads was kept at 1:10 (the mass of biotinylated anti‐ID antibody in μg to the volume of SMART IA magnetic beads in mL). The tube that contains the solution mixture was rotated for 30 min at room temperature using a LabQuake shaker (Barnstead) before loading onto a magnetic tube rack. SMART IA magnetic beads were separated from the solution, and the supernatant was extracted and discarded. The SMART IA beads were washed three times with SN1 buffer (50 mM Tris–HCl in DI water containing 1 mg/mL BSA, pH 7.5) with the same volume as the original bead slurry. The anti‐ID conjugated SMART IA beads were stored at 4°C until use.A total of 1000 μL of SN1 buffer were added to a new 5 mL Protein LoBind tube, followed by mixing with 300 μL of plasma incubated ADC1 sample. The mixture was briefly vortexed before adding 600 μL of conjugated SMART IA beads to the tube. The tube was taped onto a Thermomixer C (Eppendorf) and shaken at 25°C and 1200 rpm for 2 h. After immuno‐affinity enrichment, the tube was put onto the magnetic tube rack, and the supernatant was discarded. The beads were washed three times using 1900 μL of SN2 buffer (50 mM Tris–HCl in DI water, pH 7.5) before 1500 μL SMART IA digestion buffer (provided in the kit) was added to the washed beads. To activate the pre‐immobilized trypsin on SMART IA beads, the mixture was shaken at 70°C and 1200 rpm using Thermomixer C (Eppendorf) for 2 h. After tryptic digestion, the tube was placed onto the magnetic tube rack, where the digested ADC1 solution was transferred into a new tube. The sample clean‐up was performed using an Oasis HLB cartridge (Waters) following the manufacturer's protocol. The post clean‐up sample was diluted to 120 μL using 0.1% FA in water, and 30 μL of sample was loaded into the LC–MS system.LC–MS/MS parametersThe separation of tryptic digested peptides was performed on a BEH C18 UPLC column (Waters, 1.7 μm, 2.1 × 50 mm) using 0.1% FA in water as mobile phase A (MPA) and 0.1% FA in ACN as mobile phase B (MPB). The ZenoTOF 7600 mass spectrometer was operated in full‐scan MS (m/z 100–1500) with collision energy (CE) set as 10 V. For MS/MS, CE = 40 V was applied to CID, whereas the KE for EAD was set as 11 eV. The LC gradient and other parameters are described in the Section S2 in detail.RESULTSCharacterization of thio‐succinimide hydrolysisThiol‐maleimide reaction is widely used for bioconjugation purposes. The resulted thio‐succinimide structure is vulnerable to ring‐opening hydrolysis. ADC1 is a thio‐succinimide‐linked ADC, produced from reducing interchain disulfide bonds, followed by covalently conjugating linker‐payload to the thiol groups from cysteine side chains using maleimide chemistry. To identify and characterize its hydrolysis isomers, we used a signature ADC1 tryptic digested product, which is composed of a tripeptide GEC, succinimide (with or without water adduct), and R2 to demonstrate the process. Figure 1 shows the hydrolysis reaction and potential diagnostic ions for differentiating hydrolysis isomers, using GEC + succinimide + R2 as an example.1FIGUREScheme of thio‐succinimide hydrolysis mechanism and the proposed structures of product ions after electron‐activated dissociation (EAD). Product ions in the red boxes are diagnostic ions with predominant charge state indicated. R1 is a tripeptide GEC, and R2 is linker‐payload. The linker‐payload is covalently conjugated to the antibody via a cysteine side chain thiol. Succ + H2O = hydrolyzed succinimide. [Color figure can be viewed at wileyonlinelibrary.com]Figure 2A shows the mass spectra of [GEC + succinimide + R2 + 3H]3+ of ADC1 with 0 h (black line) and 168 h (blue line) incubation. After 168 h incubation, the signal intensity of [GEC + succinimide + R2 + 3H]3+ (observed monoisotopic m/z 485.8744, −11 ppm difference) dropped significantly compared with the sample with 0 h incubation. To be specific, the absolute intensity of the monoisotopic peak of [GEC + succinimide + R2 + 3H]3+ dropped from 2.9 × 104 to 7.2 × 102 cps after 168 h of incubation. Instead, we noticed a notable increase in peaks representing GEC + succinimide + R2 with +18 Da shift (observed monoisotopic m/z 491.8779, z = 3, −11 ppm difference). Thus, the hydrolysis of succinimide over 168 h was confirmed. Figure 2B is the extracted ion chromatogram (XIC) of [GEC + succinimide + R2 + 3H]3+ (m/z 485.8799) in 0 and 168 h incubated samples. Figure 2C shows the XIC of [GEC + succinimide + H2O + R2 + 3H]3+ (m/z 491.8834) at different incubation times. Similarly, the XICs support the conclusion that most of the succinimide conjugated to GEC is hydrolyzed over 168 h of incubation. Table S2 gives theoretical and observed monoisotopic m/z values of the discussed ions.2FIGURE(A) Overlaid mass spectra of [GEC + succinimide + R2 + 3H]3+ and [GEC + succinimide + H2O + R2 + 3H]3+ from 0 h (black line) and 168 h (blue line) plasma incubated antibody‐drug conjugate 1 (ADC1) samples. Both species observed a charge state of 3. Overlaid extracted ion chromatograms (XICs) of (B) [GEC + succinimide + R2 + 3H]3+ or (C) [GEC + succinimide + H2O + R2 + 3H]3+. Black line represents the sample with no incubation, whereas the blue line is the sample incubated for 168 h. [Color figure can be viewed at wileyonlinelibrary.com]Identification of thio‐succinimide hydrolysis isomeric productsThe hydrolysis of thio‐succinimide creates a pair of constitutional isomeric products, as shown in Figure 2A and reported previously. In Figure 2C, two distinctive peaks with different RTs observed in XIC confirm the presence of GEC + succinimide + H2O + R2 isomers. The peak with RT = 45.9 min is denoted as Peak 1, whereas the peak with RT = 46.4 min is denoted as Peak 2. In order to elucidate the chemical structure corresponding to each peak, we employed two orthogonal dissociation methods, CID and EAD, on [GEC + succinimide + H2O + R2 + 3H]3+. Figure S2A,B show the CID MS/MS spectra of precursor ions from Peaks 1 and 2, respectively. However, we did not find any distinctive peaks by comparing these two spectra. Hence, CID does not provide useful information in distinguishing thio‐succinimide hydrolysis isomers.Differentiation between aspartic acid and iso‐aspartic acid from asparagine deamidation has been previously reported using ExD methods, such as ETD and ECD. c + 57 and z − 57 were confirmed to be a pair of diagnostic ions to identify iso‐aspartic acid generated from either ETD or ECD.26,28 Asparagine deamidation first generates a succinimide intermediate prior to ring‐opening hydrolysis, which is analogous to thio‐succinimide hydrolysis as shown in Figure 2A. Figure S3 shows the mechanism of asparagine deamidation and structures of diagnostic c‐ and z‐related ions. Figure S4 shows the dissociation sites and the structures of c‐ and z‐related ions in both deamidation and thio‐succinimide hydrolysis isomers. Figure 3A,B are the MS/MS spectra (m/z range: 100–2900) of [GEC + succinimide + H2O + R2 + 3H]3+ precursor ions from Peaks 1 and 2, respectively, using EAD. The product ions in both figures show high agreement with each other. (I), (II), and (III) in Figure 3A,B are zoomed‐in MS/MS spectra with different m/z range. Comparing Figures 3A (I) and 3B (I), we found a singly charged product ion with monoisotopic m/z of 365.0839, which is in accordance with [R1 + Thio + 57 + H]+ diagnostic ion with a −15 ppm mass difference. Similarly, by comparing Figures 3A (III) and 3B (III), we found a singly charged product ion with a monoisotopic m/z of 1109.5403. This peak suggests the presence of [R2 + Succ + H2O‐57 + H]+ diagnostic ion with −12 ppm mass difference. Thus, we can confirm Figure 3A represents the MS/MS spectrum of thio‐Asp from Peak 1, whereas Figure 3B is the MS/MS spectrum of thio‐isoAsp from Peak 2. Observed m/z values of discussed ions all show a similar extent of deviation to their theoretical values. A recalibration performed after acquisition corrects the m/z difference to within ±3 ppm (see Section S3). Hence, the identification of ions of interest is accurate, despite the mass errors being close to the high end of tolerance.3FIGUREElectron‐activated dissociation (EAD) MS/MS spectra (m/z 100–1500) of (A) Peak 1 and (B) Peak 2 observed in Figure 2C. Signature product ions are labelled on both spectra. Pink arrows point to a doubly charged diagnostic ion, [R2 + Succ + H2O − 44 + 2H]2+. Dashed arrows point to zoomed‐in EAD MS/MS spectra with different m/z ranges. (i) m/z 363–369, (ii) m/z 572–582, and (iii) m/z 1105–1115. Blue boxes are the zoomed‐in spectra of m/z 1410–1440 (see Figure S7). The stars with 4, 5, and 6 arms in inset figures correspond to the monoisotopic peak of each diagnostic product ion, respectively. Their observed m/z values are 365.0839, 577.7643, and 1109.5403, respectively. [R2 + Succ + H2O − 44 + 2H]2+ ion was also observed in Figure 3A (II) with −12 ppm difference to its theoretical value, although the intensity is low. [Color figure can be viewed at wileyonlinelibrary.com]Another notable difference is that [R2 + Succ + H2O − 44 + 2H]2+ (indicated by the pink arrow) shows significantly higher intensity in Figure 3A, whereas the intensity of the same product ion in Figure 3B is low. Figure S5A,B show zoomed‐in MS/MS spectra from both Peaks 1 and 2 at m/z 555–570 range, respectively. The higher signal intensity of [R2 + Succ + H2O − 44 + 2H]2+ in Figure S5A suggests the CO2 loss after forming the [R2 + Succ + H2O + 2H]2+ ion of thio‐Asp is significantly enhanced. This finding aligns with the ExD dissociation results of aspartic acid in deamidation.We also observed a doubly‐charged product ion with a monoisotopic m/z of 561.2728 with enhanced intensity in Figure 3A (II), but low intensity in Figure 3B (II). We propose the structure of thio + succinimide + H2O + R2 − CO2 (Figure 1). We denote this product ion as [R2 + ThioSucc + H2O − 44 + 2H]2+. Although both thio‐Asp and thio‐isoAsp are capable of CO2 loss after forming the [R2 + ThioSucc + H2O + 2H]2+ product ion, clearly, thio‐isoAsp is more prone to this neutral loss. We also found the [R2 + ThioSucc + H2O + 2H]2+ product ion in both thio‐Asp and thio‐isoAsp with similar intensities, as shown in Figure S6A,B. Thus, [R2 + ThioSucc + H2O − 44 + 2H]2+ product ion can serve as an additional diagnostic ion for differentiating thio‐Asp and thio‐isoAsp. To our knowledge, this product ion has never been reported in the context of differentiation of deamidation isomers. Hence, this diagnostic ion is distinctive of the thio‐succinimide system. We infer that the generation of this unique product ion is related to the thiol group linked with the succinimide ring and R2.Neutral losses are not limited to the product ions generated from ExD dissociation. [M − 60 + H]+ was reported to be a diagnostic ion for Asp in deamidation.29 Figure S7A,B show the zoomed‐in MS/MS spectra of Figure 3A,B at m/z 1410–1440. To our surprise, the [M − 60 + H]+ product ions (observed monoisotopic m/z 1414.6355, 10 ppm difference to its theoretical value) in both Figure S7A,B show similar intensities. Thus, [M − 60 + H]+ cannot differentiate thio‐Asp and thio‐isoAsp, and we think the 60 Da neutral loss might come from the contribution of R2. We also observed peaks with monoisotopic m/z 1430.6428 (−7 ppm difference to theoretical m/z of [M − 44 + H]+) in both isomers. The similar intensities of [M − 44 + H]+ and [M − 60 + H]+ suggest that neutral losses to the whole precursor ion cannot distinguish thio‐Asp and thio‐isoAsp.Furthermore, we also investigated another tryptic peptide, SCDK, that contains succinimide + R2 from ADC1, followed by the same sample preparation steps. The SCDK + succinimide + R2 is mostly hydrolyzed after 168 h of incubation, as shown in Figure S8A. Both XICs of non‐hydrolyzed and hydrolyzed at 0 and 168 h in Figure S8B,C support the same conclusion. Additionally, Figure S8C shows two distinct peaks that were well separated under the current LC gradient. The presence of two peaks suggests the formation of isomers after thio‐succinimide hydrolysis. This finding aligns with our observations of GEC‐linked succinimide structure.CONCLUSIONThe formation of thio‐succinimide hydrolysis isomeric products is well known, however, the differentiation between them has not been sufficiently studied so far. This work demonstrated a reference material‐free characterization of hydrolyzed thio‐succinimide isomers using EAD. Similar to the differentiation between aspartic acid and iso‐aspartic acid generated from asparagine deamidation, we observed distinctive [R1 + Thio + 57 + H]+, [R2 + Succ + H2O − 57 + H]+ for thio‐isoAsp and enhanced [R2 + Succ + H2O − 44 + 2H]2+ for thio‐Asp in EAD MS/MS spectra. However, [M − 44 + H]+ and [M − 60 + H]+ failed to serve as diagnostic ions because of their similar intensities in EAD MS/MS spectra for both thio‐Asp and thio‐isoAsp. To our surprise, we observed a significantly enhanced CO2 neutral loss after formation of the [R2 + ThioSucc + H2O + 2H]2+ product ion for thio‐isoAsp. This product ion is denoted as [R2 + ThioSucc + H2O − 44 + 2H]2+, and for the very first time, we report it as a unique diagnostic ion for differentiating thio‐Asp and thio‐isoAsp. To our knowledge, an analogous product ion has not been identified for asparagine deamidation. The formation of this product ion might relate to the thiol group linked to the succinimide ring and the structures of R2. For instance, in the case of deamidation, the succinimide ring is linked with amino acids at both ends. Compared with the conventional approach, which matches RT of purified reference material to distinguish isomers, EAD serves as a novel tool to identify isomeric structures without reference standards. The differentiation between thio‐Asp and thio‐isoAsp may eventually benefit the drug development where thiol‐maleimide conjugation is employed.AUTHOR CONTRIBUTIONSJunyan Yang: Conceptualization; methodology; investigation; formal analysis; writing—original draft; visualization. Jiaqi Yuan: Conceptualization; methodology; investigation; draft review. Yue Huang: Conceptualization; methodology; draft review. Anton I. Rosenbaum: Conceptualization; methodology; investigation; draft review; supervision.ACKNOWLEDGMENTSThis study was funded by AstraZeneca. We are deeply grateful to Haichuan Liu and Zoe Zhang from SCIEX for providing help with LC–MS data analysis, insightful discussions, and advice.CONFLICT OF INTEREST STATEMENTJ. Yang, J. Yuan, Y. H., and A. I. R. are or were employees of AstraZeneca at the time this work was conducted and may hold stock and/or stock options or interests in the company.DATA AVAILABILITY STATEMENTThe data that supports the findings of this study are available in the supplementary material of this article.REFERENCESNorthrop BH, Frayne SH, Choudhary U. Thiol–maleimide “click” chemistry: evaluating the influence of solvent, initiator, and thiol on the reaction mechanism, kinetics, and selectivity. Polym Chem. 2015;6(18):3415‐3430. doi:10.1039/c5py00168dRenault K, Fredy JW, Renard PY, Sabot C. Covalent modification of biomolecules through maleimide‐based labeling strategies. Bioconjug Chem. 2018;29(8):2497‐2513. doi:10.1021/acs.bioconjchem.8b00252Summonte S, Racaniello GF, Lopedota A, Denora N, Bernkop‐Schnurch A. Thiolated polymeric hydrogels for biomedical application: cross‐linking mechanisms. J Control Release. 2021;330:470‐482. doi:10.1016/j.jconrel.2020.12.037Ferhati X, Jimenez‐Moreno E, Hoyt EA, et al. Single mutation on trastuzumab modulates the stability of antibody‐drug conjugates built using acetal‐based linkers and thiol‐maleimide chemistry. J am Chem Soc. 2022;144(12):5284‐5294. doi:10.1021/jacs.1c07675Su Z, Xiao D, Xie F, et al. Antibody‐drug conjugates: recent advances in linker chemistry. Acta Pharm Sin B. 2021;11(12):3889‐3907. doi:10.1016/j.apsb.2021.03.042BroadPharm. ADC Approval up to 2023. https://broadpharm.com/blog/ADC-Approval-up-to-2023 (Accessed May 3, 2024)Wei C, Zhang G, Clark T, et al. Where did the linker‐payload go? A quantitative investigation on the destination of the released linker‐payload from an antibody‐drug conjugate with a maleimide linker in plasma. Anal Chem. 2016;88(9):4979‐4986. doi:10.1021/acs.analchem.6b00976Lahnsteiner M, Kastner A, Mayr J, Roller A, Keppler BK, Kowol CR. Improving the stability of maleimide‐thiol conjugation for drug targeting. Chemistry. 2020;26(68):15867‐15870. doi:10.1002/chem.202003951Christie RJ, Fleming R, Bezabeh B, et al. Stabilization of cysteine‐linked antibody drug conjugates with N‐aryl maleimides. J Control Release. 2015;220(Pt B):660‐670. doi:10.1016/j.jconrel.2015.09.032Huang W, Wu X, Gao X, et al. Maleimide‐thiol adducts stabilized through stretching. Nat Chem. 2019;11(4):310‐319. doi:10.1038/s41557‐018‐0209‐2Fontaine SD, Reid R, Robinson L, Ashley GW, Santi DV. Long‐term stabilization of maleimide‐thiol conjugates. Bioconjug Chem. 2015;26(1):145‐152. doi:10.1021/bc5005262Huang Y, Tan HY, Yuan J, et al. Extensive biotransformation profiling of AZD8205, an anti‐B7‐H4 antibody‐drug conjugate, elucidates pathways underlying its stability in vivo. Anal Chem. 2024. doi:10.1021/acs.analchem.4c02309Huang Y, Mou S, Wang Y, Mu R, Liang M, Rosenbaum AI. Characterization of antibody‐drug conjugate pharmacokinetics and in vivo biotransformation using quantitative intact LC‐HRMS and surrogate analyte LC‐MRM. Anal Chem. 2021;93(15):6135‐6144. doi:10.1021/acs.analchem.0c05376Stalcup AM. Chiral separations. Annu Rev Anal Chem. 2010;3(1):341‐363. doi:10.1146/annurev.anchem.111808.073635Xie Y, Zhang J, Yin S, Loo JA. Top‐down ESI‐ECD‐FT‐ICR mass spectrometry localizes noncovalent protein‐ligand binding sites. J am Chem Soc. 2006;128(45):14432‐14433. doi:10.1021/ja063197pZubarev RA, Horn DM, Fridriksson EK, et al. Electron capture dissociation for structural characterization of multiply charged protein cations. Anal Chem. 2000;72(3):563‐573. doi:10.1021/ac990811pChi A, Huttenhower C, Geer LY, et al. Analysis of phosphorylation sites on proteins from saccharomyces cerevisiae by electron transfer dissociation (ETD) mass spectrometry. Proc Natl Acad Sci U S a. 2007;104(7):2193‐2198. doi:10.1073/pnas.0607084104Sze SK, Ge Y, Oh H, McLafferty FW. Plasma electron capture dissociation for the characterization of large proteins by top down mass spectrometry. Anal Chem. 2003;75(7):1599‐1603. doi:10.1021/ac020446tRoepstorff P, Fohlman J. Proposal for a common nomenclature for sequence ions in mass spectra of peptides. Biomed Mass Spectrom. 1984;11(11):601. doi:10.1002/bms.1200111109Cooper HJ, Hakansson K, Marshall AG. The role of electron capture dissociation in biomolecular analysis. Mass Spectrom Rev. 2005;24(2):201‐222. doi:10.1002/mas.20014Fort KL, Cramer CN, Voinov VG, et al. Exploring ECD on a benchtop Q exactive orbitrap mass spectrometer. J Proteome Res. 2018;17(2):926‐933. doi:10.1021/acs.jproteome.7b00622Kelleher NL, Zubarev RA, Bush K, et al. Localization of labile posttranslational modifications by electron capture dissociation: the case of γ‐carboxyglutamic acid. Anal Chem. 1999;71(19):4250‐4253. doi:10.1021/ac990684xKjeldsen F, Haselmann KM, Budnik BA, Jensen F, Zubarev RA. Dissociative capture of hot (3‐13 eV) electrons by polypeptide polycations: An efficient process accompanied by secondary fragmentation. Chem Phys Lett. 2002;356(3‐4):201‐206. doi:10.1016/S0009‐2614(02)00149‐5SCIEX. Electron Activated Dissociation (EAD). https://sciex.com/content/dam/SCIEX/pdf/brochures/zenotof-7600-whitepaper-ead.pdf (Accessed July 18, 2024)Irudayanathan FJ, Zarzar J, Lin J, Izadi S. Deciphering deamidation and isomerization in therapeutic proteins: effect of neighboring residue. MAbs. 2022;14(1):2143006. doi:10.1080/19420862.2022.2143006Huang Y, Yuan J, Mu R, et al. Multiplex bioanalytical methods for comprehensive characterization and quantification of the unique complementarity‐determining‐region deamidation of MEDI7247, an anti‐ASCT2 pyrrolobenzodiazepine antibody‐drug conjugate. Antibodies (Basel). 2023;12(4). doi:10.3390/antib12040066Liu H, Zhang Z. Comprehensive differentiation of deamidation isomers from forced degradation by electron activation dissociation (EAD); SCIEX. https://sciex.com/tech-notes/biopharma/comprehensive-differentiation-of-deamidation-isomers-from-forced-degradation-by-electron-activation-dissociation-ead (Accessed May 3, 2024)Yang H, Zubarev RA. Mass spectrometric analysis of asparagine deamidation and aspartate isomerization in polypeptides. Electrophoresis. 2010;31(11):1764‐1772. doi:10.1002/elps.201000027Sargaeva NP, Lin C, O'Connor PB. Differentiating N‐terminal aspartic and Isoaspartic acid residues in peptides. Anal Chem. 2011;83(17):6675‐6682. doi:10.1021/ac201223d

Journal

Rapid Communications in Mass Spectrometry – Wiley

Published: Dec 15, 2024

There are no references for this article.