Sample preparation: A sample of trastuzumab was denaturated with 7.2 M guanidine hydrochloride, 100 mM Tris buffer pH 7.2, followed by reduction with 10 mM DL-dithiothreitol and alkylation with 30 mM iodoacetamide. Digestion was performed with trypsin/Lys-C enzyme at 37 °C for 16 h.
Chromatography: 10 µl (4 µg) of the trypsin/Lys-C digest were separated with a CSH C18 column (1.7 μm particle size, 130 Å, 2.1×100 mm, Waters) using an ExionLC system. The mobile phase A consisted of water with 0.1% formic acid, while the organic phase B was acetonitrile 0.1% formic acid. A gradient profile was used at a flow rate of 300 μL/min (Table 1). The column temperature was maintained at 50°C.
Mass spectrometry: Data were acquired with an information dependent acquisition (IDA) method using the SCIEX ZenoTOF 7600 system. General method parameters were kept the same and are summarized in Table 2. Parameters specific for EAD or CID can be found in Table 3.
Data processing: Data were processed in Byos software (Protein Metrics Inc.). To achieve side by side comparison, the standard PTM workflow was modified to include two MS/MS Id Byonic processing nodes, one for CID data processing, one for EAD data processing. All other processing parameters were kept the same. Peptide identification and fragments mass tolerance were set 6 ppm and 20 ppm, respectively. The processed results were filtered to eliminate results with MS/MS scores lower than 100.
In biotherapeutics characterization, glycosylations are usually being classified as a critical quality attribute and therefore closely monitored. Liquid chromatography mass spectrometry (LC-MS) based peptide mapping is considered to be a versatile tool for characterization of protein glycosylation, since it eliminates the need to remove the glycan from the protein, while providing very comprehensive information about the molecule sequence and other PTMs.3,6 However, traditional CID approaches can either provide fragment information of the fragile glycans when applying low collision energies or of the peptide backbone when higher collision energies are used. Achieving both at the same time and at high quality, along with a general peptide mapping approach, remains a challenge with CID. In addition, the high energies used for CID usually result in the dissociation of the glycan structures from the peptide backbone. Therefore, identification of the peptide and exact localization of the glycan is limited, especially in the case of multiple potential modification sites in a given peptide. On the other hand, in addition to diagnostic oxonium ions of the glycans, the tunable electron energy in the SCIEX ZenoTOF 7600 system produces rich peptide backbone fragment ions and fragments with the intact glycan attached, simultaneously. These data allow for confidence in the correct identification of peptides and the accurate localization and identification of the attached glycans. Using the Zeno trap in combination with EAD allows for accurate and detailed identification of even low abundant glycopeptides due to a boost in the sensitivity of the fragments.
The most intense glycopeptides for each glycan type found in the trastuzumab digest contained one miscleavage site after R304 due to the steric hindrance introduced by glycans at N300 (Figures 2, 3 and 5). Figure 2 shows an example of a glycopeptide carrying G0F. The precursor ion and fragment ion spectra from both CID and EAD were compared side-by-side. MS data were matched with a tolerance of maximal 6 ppm. Subsequent data interpretation of MS/MS spectra included the identification of peptide fragments, oxonium ions and peptideglycan fragments (Figure 2, top right). As seen in the spectrum, the dominant ions were oxonium ions, in the case of CID, and low abundant b- and y-ions. It should be noted that the default parameters for the rolling collision energy (CID) can be adjusted to increase the coverage of the peptide backbone as shown previously9 , however this approach usually also limits the overall MS/MS sequence coverage for other peptides. Furthermore no peptide fragments with intact glycans were detected in the case of CID. On the contrary, EAD did not only provide very comprehensive fragmentation of the peptide backbone with 100% MS/MS sequence coverage being superior to CID, but the doubly charged c9 and c12 ion with intact glycan also provided accurate localization of the site of modification (encircled ions in Figure 2, top right).
For the peptide with a G1F modification, a similar behavior was observed (Figure 3). Comprehensive fragmentation coverage was achieved with EAD compared to CID and ions proving the localization of the fragile modification could be detected (c9+++ and c10++ etc.). Apart from the high abundant glycosylation forms of G0F and G1F, lower abundant forms were reproducibly identified (Figure 4). One example is a high mannose species (Man5 at ~ 3%) in Figure 5. Despite its low abundance (more than 10x lower in relative abundance than the G0F-containing peptide, see Figure 4), high-quality fragment ion spectra were achieved, demonstrating the high sensitivity of Zeno EAD and Zeno CID. In addition, a full series of z1- z21 ions together with a series of c ions allowed for 96-100% fragment coverage for EAD, while CID only achieved 61-68% fragment coverage depending on the glycopeptide (Figures 2, 3 and 5).
For an easy review of the data, a glycan profiling report was generated. The template was formatted to sum and report the glycol forms detected in different peptide sequences (including tryptic cleaved peptides and missed cleavages) and filtered to show peptides with N-linked glycan as a single modification. The automated color coding heat map facilitates a quick understanding of which glycoforms are present in relative high, medium or low abundance, ranging from 44% to 0.2%. All Nlinked glycosylations found to be present were in alignment with those previously reported (Figure 4).10 The results demonstrate great repeatability of the Zeno EAD technology for glycopeptide analysis across different abundancies.