Monoclonal antibodies (mAbs) form a major class of biologics and recently biosimilars and biobetters are being added to the
growing inventory of therapeutics. In-depth characterization of mAbs at various stages of development and manufacturing is
essential to maintain product safety and efficacy. However, analysis of mAbs is challenging due to their high molecular weight,
the microheterogeneity presented by the glycans, and degradative modifications that occur during production. Any analytical
technique that provides greater depth of information without a time penalty is an advantage. A recent advancement to meet
this need was the introduction of CESI–MS. CESI is the integration of capillary electrophoresis (CE) and electrospray ionization
(ESI) in a dynamic process, within the same device. In this technology, the analytes are separated inside an open nontapered
capillary based on their electrophoretic mobility, and electrosprayed directly into the MS (2). At operating flow rates less
than 30 nL/min, very efficient desolvation and, thus, ionization is achieved.
Though high speed CESI separations reduce analysis time, it also necessitates the use of high speed MS to preserve the separation
efficiency. The TripleTOF® 5600+ system (AB Sciex) offers the necessary high acquisition speed, high resolution, and high
mass accuracy, in both MS and MS-MS modes. CESI performance was evaluated by analyzing a tryptic digest of trastuzumab using
the SCIEX CESI 8000 - TripleTOF® 5600+ MS platform.
Trastuzumab was reduced, alkylated, and digested with trypsin. After drying, it was resuspended in the leading electrolyte
(100 mM ammonium acetate at pH 4) and 50 nL (100 fmol) was injected into the separation capillary. The background electrolyte
used was 10% acetic acid and a separation voltage of 20 kV (normal polarity) was applied for 60 min. Information dependent
acquisition (IDA) was utilized to trigger MS-MS. IDA parameters were optimized so that the duty cycle of the MS was matched
to the high speed CE separation. Data analysis was performed using BioPharmaView™ software (AB Sciex, Massachusetts).
Results and Discussion
Primary Sequence Coverage: 100% primary sequence coverage of both the light and heavy chains of the antibody were obtained. Peptides ranging from 4 to
63 amino acids in length without any missed cleavages were detected. Electrophoretic separation is based on the charge-to-mass
ratio of the peptides and is not dependent on relative hydrophobicity. Thus, small hydrophilic peptides often lost in the
LC void volume, and large hydrophobic peptides, which tend to be retained on the column, can be identified by CESI–MS, resulting
in the high sequence coverage observed.
Figure 1: (a) Extracted ion electropherogram of N-terminal peptide with Glu and pyroGlu separated by CESI and (b) MS-MS identification
of N-terminal peptide with pyroGlu.
PTM Characterization: Data from CESI-MS analysis showed the presence of several PTM hotspots such as N-terminal pyroGlu formation, methionine oxidation,
and asparagine deamidation. Pyroglutamination leads to loss of a positive charge which results in the electrophoretic mobility
of the modified peptide being lower than the unmodified one. This is advantageous since the modified and unmodified forms
can be separated by CESI–MS and the MS-MS spectra confirmed the presence of the pyroGlutamate residue (Figure 1). Oxidative
degradations at Met255 and Met431 and deamidations at Asn55 and Asn387 in the heavy chain and at Asn30 in the light chain were also identified. A typical identification from the CESI–MS data is shown in Figure 2.
Figure 2: MS-MS identification of asparagine deamidation with (a) showing unmodified peptide and (b) deamidation at Asn55 in the heavy chain of trastuzumab.
Glycosylation Heterogeneity: Trastuzumab possesses one N-glycosylation site at Asn300 in the HC where different glycoforms such as a-fucosylated or fucosylated glycans can be present (3). By using CESI-MS,
the G0F, G1F, and G2F forms of the peptide TKPREEQYNSTYR were separated well as shown in Figure 3. Furthermore, the identification
of G0F, G1F, and G2F forms of peptide EEQYNSTYR without the missed cleavage (at the arginine residue in the peptide TKPREEQYNSTYR)
also confirmed the presence of these glycoforms. In addition, the a-fucosylated forms of this peptide, such as G0 and G1,
were identified, but it has to be further confirmed that the a-fucosylated forms were not generated due to source fragmentation
of fucosylated counterparts.
Figure 3: Extracted ion electropherograms of peptide TKPREEQYNSTYR with G0F, G1F, and G2F modifications.