C146-E340 LCMS Bioanalysis of Antibody Drugs Using Technical Fab-selective Proteolysis “nSMOL Method” Report — Selection of Signature Peptide — Noriko Iwamoto1, Takashi Shimada1 Abstract: nSMOL (nano-surface and molecular-orientation limited proteolysis) is Shimadzu's completely new, proprietary, and innovative technique for selective proteolysis of Fab region of monoclonal antibodies. nSMOL allows analytical method development of antibody drugs indepen- dent of a variety of antibody drugs. Fab-derived peptide fragments produced by nSMOL can be precisely quantified by multiple reaction monitoring (MRM) using the Shimadzu LCMS-8050/8060 triple quadrupole liquid chromatograph mass spectrometer (TQ-LCMS). This report describes a selection protocol of signature peptides suitable for pharmacokinetic studies. Keywords: nano-surface and molecular-orientation limited proteolysis, antibody drug, bioanalysis, LC/MS/MS 1. Introduction A peptide for quantitation (signature peptide) is selected from the tryptic peptides containing a complementarity-determining region Pharmacokinetic information provides some of the most fundamental (CDR), which defines the specificity of the antibody. However, it is not indicators. The effective drug discovery is supported by the overall possible that the CDR-containing peptide does not have the same pharmacokinetic profile such as for drug efficacy and toxicity. amino acid sequence as that in the endogenous IgGs. At this point, it must be confirmed that there is no competition with the signature The current method used for measuring drug concentration in blood peptide in the biological matrix. is enzyme-linked immunosorbent assay (ELISA). However, there are critical issues with ELISA, including influences from cross-reaction and inhibitory materials. In contrast, by MS, analysis is performed based on the structural information; thus, the aforementioned issues can 2. Structure and Specificity of potentially be resolved. Antibodies The LCMS analysis of high-molecular-weight proteins, such as antibod- Antibodies are large biological molecules with a heterotetrameric ies, is normally performed after fragmentation of the protein into small- protein containing two heavy and light chains. Antibody structures er peptides using a protease, such as trypsin or lysyl endopeptidase. are extraordinarily well conserved, and antibodies have a flexible However, this process also generates a large number of peptides includ- hinge region between the rigid Fc and the variable Fab. A variable ing the signature peptides. These peptides increase the background region, Fv, is positioned at the top of the Fab. The antibody specific- noise and ionization suppression, and become a major cause of instabil- ity in this region is defined as amino acid substitution by somatic ity in the LCMS system. nSMOL can decrease these issues by selective mutation. Portions of the Fv with a particularly high frequency mu- proteolysis on the analytical target region of the antibody Fab. There- tation are called CDRs, which play a major role in antigen binding fore, the use of this approach can improve the reproducibility and ro- (see Fig. 1). bustness of the analytical system with maintaining antibody specificity. Fab Hinge Variable region Variable region Complementarity-determining region (CDR) Light chains Antigen-binding site CDR3 has the highest affinity Sugar chain to antigen. Fc Constant region Heavy chains Fig. 1 Basic Antibody Structure 1 Technology Research Laboratory 1 3. Frequency of Amino Acid 160 Substitution in the Fv 12 3 The Fv contains three regions of high-frequency amino acid substitu- 120 tion, namely CDR1, CDR2, and CDR3 (see Fig. 2). Almost all amino acid residues in these regions can undergo substitution; therefore, an extremely large number of possible sequences can be generated. 80 Consequently, predicting amino acid sequences is difficult. Therefore, nSMOL bioanalysis requires the setting and selection of signature peptides based on actual analysis data. 40 4. Amino Acid Sequence Alignment Frequency of amino acid substitution Frequency Using ClustalW 0 0 20406080100 An example is shown the prediction of candidate signature peptides Amino acid sequence in four chimeric antibodies (Rituximab, Brentuximab vedotin, Cetux- TT Wu, J Exp Med. 1970, modified imab, and Infliximab) by ClustalW alignment (Fig. 3, black: common residues, gray: similar residues, white: unique residues). Fig. 2 Frequency of Amino Acid Substitution in Fv Amino acid sequence alignment of heavy chains RTX H 1 QVQLQQPGAEAELVKPGAGASVKMKMSCKASGYTYTFTSTSYNMHYNMHWVKQTPGRGLEWIGAGAIYPYPGN--GN--GDGDT 58 BRX H 1 QIQLQQSGPEVVKPGASVKISCKASGYTFTDYYITWVKQKPGQGLEWIGWIYPGS--GNT 58 CTX H 1 QVQLKQSGPGLVQPSQSLSITCTVSGFSLTNYGVHWVRQSPGKGLEWLGVIWSG---GNT 57 IFX H 1 EVKLEESGGGLVQPGGSMKLSCVASGFIFSNHWMNWVRQSPEKGLEWVAEAEIRSR KSINSINSASAT 60 RTX H 59 SYNQKNQKFKGKATLTADKSSSTAYMQLSSLTSEDSAVYYCARSTYYGGDWYFNVWGAGTTVT 118 BRX H 59 KYNEKFKGKATLTVDTSSSTAFMQLSSLTSEDTAVYFCAN---YG-NYWFAYWGQGTQVT 114 CTX H 58 DYNTPFTSRLSINKDNSKSQVFFQVFFKMNKM SLQSNDTAIYYCARALTYY-DYEFAYWGQGTLVT 116 IFX H 61 HYAESAESVKGRFTISRDDSKSAVYLAVYLQMTQMTDLRTEDTGVYYCSRN-YYG--STYDYWGQGTTLT 117 RTX H 119 VSAASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVL 178 BRX H 115 VSAASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVL 174 CTX H 117 VSAASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVL 176 IFX H 118 VSSASTKGPSVFPLAPSSKSTSGGTAALGCLVKDYFPEPVTVSWNSGALTSGVHTFPAVL 177 RTX H 179 QSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKS---CDKTHTCPPCPA 235 BRX H 175 QSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKS---CDKTHTCPPCPA 231 CTX H 177 QSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKRVEPKSPKSCDKTHTCPPCPA 236 IFX H 178 QSSGLYSLSSVVTVPSSSLGTQTYICNVNHKPSNTKVDKKVEPKS---CDKTHTCPPCPA 234 RTX H 236 PELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKP 295 BRX H 232 PELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKP 291 CTX H 237 PELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKP 296 IFX H 235 PELLGGPSVFLFPPKPKDTLMISRTPEVTCVVVDVSHEDPEVKFNWYVDGVEVHNAKTKP 294 RTX H 296 REEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTL 355 BRX H 292 REEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTL 351 CTX H 297 REEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTL 356 IFX H 295 REEQYNSTYRVVSVLTVLHQDWLNGKEYKCKVSNKALPAPIEKTISKAKGQPREPQVYTL 354 RTX H 356 PPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLT 415 BRX H 352 PPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLT 411 CTX H 357 PPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLT 416 IFX H 355 PPSRDELTKNQVSLTCLVKGFYPSDIAVEWESNGQPENNYKTTPPVLDSDGSFFLYSKLT 414 RTX H 416 VDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK 451 BRX H 412 VDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK 447 CTX H 417 VDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK 452 IFX H 415 VDKSRWQQGNVFSCSVMHEALHNHYTQKSLSLSPGK 450 Amino acid sequence alignment of light chains RTX L 1 QIVLSQSPAILSASPGEKVTMTCRASSSV-----SYIHWFQQKPGSSPKPWIYATSNLAS 55 BRX L 1 DIVLTQSPASLAVSLGQRATISCKASQSVDFDGDSYMNWYQQKPGQPPKVVLIYAYAASNLENLES 60 CTX L 1 DILLTQSPVILSVSPGERVSFSCRASQSIG----IG----TNTNIHWYQQRTNGR SPRLLIKYYASESIESIS 56 IFX L 1 DILLTQSPAILSVSPGERVSFSCRASQFVG----VG----SSSSIHWYQQRTNGR SPRLLIKYYASESMESMS 56 RTX L 56 GVPVRFSGSGSGTSYSLTISRSRVEAEDAATYYCQQWTSNPPTFGGGTKLEIKRTVAAPSVF 115 BRX L 61 GIPARFSGSGSGTDFTLNIHPVEEEDAATYYCQQSNEDPWTFGGGTKLEIKRTVAAPSVF 120 CTX L 57 GIPSRFSGSGSGTDFTLSINSVESEDIADYYCQQNNNWPTTFGAGTKLELKRTVAAPSVF 116 IFX L 57 GIPSRFSGSGSGTDFTLSINTVESEDIADYYCQQSHSWPFTFGSGTNLEVKRTVAAPSVF 116 RTX L 116 IFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLS 175 BRX L 121 IFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLS 180 CTX L 117 IFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLS 176 IFX L 117 IFPPSDEQLKSGTASVVCLLNNFYPREAKVQWKVDNALQSGNSQESVTEQDSKDSTYSLS 176 RTX L 176 STLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC 213 BRX L 181 STLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC 218 CTX L 177 STLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGA- 213 IFX L 177 STLTLSKADYEKHKVYACEVTHQGLSSPVTKSFNRGEC 214 Fig. 3 Results of Amino Acid Sequence Alignment 2 5. Discussion of Alignment Analysis ClustalW sequence alignment From the sequence alignment analysis, it can be assumed that areas with a predominance of unique residues (white) are CDRs. The re- CDR estimation sults also demonstrated the presence of unique residues through- out the Fv. The prediction of signature peptides is simplified by amino acid sequences of several representative antibodies accord- ing to antibody classification (chimeric, humanized, or fully human antibody). During quantitative analysis, identified signature peptides are high- 1. Sequence information 2. TOF-MS analysis lighted in Ű for Rituximab, Ű for Brentuximab vedotin, Ű for analysis Cetuximab, and Ű for Infliximab. Some of the signature peptides High-resolution ESI-MS include N-terminal sequences that may become heterogeneity Skyline software during post-translational modification or may interfere with another biological matrix during preclinical studies. Therefore, N-terminal sequences are not used during an actual nSMOL bioanalysis. Building of peptide from Trypsin digestion of MRM data using target antibody 6. Conclusions sequence information nSMOL reaction A reliable selection of signature peptides is essential for clinical pharmacokinetic studies. Fig. 4 shows a procedure from a structural analysis technique by MS/MS analysis Import into LabSolutions mass spectrometry and a recent information-based approach. Iden- Identification of the tification of the peptide structure