Mass Spectrometry Letters

Korean Society for Mass Spectrometry (한국질량분석학회)
권호 표지
DOI QR코드

DOI QR Code

Estimation and Analysis Methods for Trastuzumab Deamidation Levels Using Mass Spectrometry

  • Received : 2024.05.23
  • Accepted : 2024.06.20
  • Published : 2024.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

We aimed to develop a suitable quantification method for detecting asparagine deamidation and aspartic acid isomerization in peptide mapping using LC-MS. Our assessment of its validity and suitability involved comparing its quantitative findings with those obtained from cation-exchange chromatography and capillary electrophoresis methods. By subjecting trastuzumab to rigorous conditions to induce these modifications, we validated the efficacy of this new analytical method in peptide mapping via LC-MS, evaluating both qualitative and quantitative aspects of asparagine deamidation and aspartic acid isomerization. Our investigation underscored the significance of enzyme selection and the presence of miss-cleaved or non-specific peptides in achieving accurate quantitative results. The experimental results demonstrated a strong correlation with results from cation-exchange chromatography and capillary electrophoresis analyses, confirming the reliability of the LC-MS based peptide mapping approach.

Keywords

Introduction

Various chemical modifications observed in protein drugs, including asparagine (Asn) deamidation, aspartate (Asp) isomerization, and methionine/tryptophan (Met/Trp) oxidation, have been extensively investigated.1-8 In particular, protein therapeutics are highly sensitive to these modifications in terms of their occurrence rate and extent, which may potentially influence the efficacy and safety of the drugs.9,10 Previous studies have highlighted the impact of Asn deamidation and Asp isomerization on various biological functions.11-19 Notably, instances such as the loss of activity in IgG1 monoclonal antibodies (mAbs) owing to deamidation or isomerization in complementarity-determining regions are prominent examples.20 Additionally, amino acid modifications in the Fc region have been linked to potential effects on Fcγ receptor binding, thereby impacting antibody-dependent cellular cytotoxicity (ADCC).1,4,21,22

After the deamidation of asparagine, aspartic acid can readily undergo isomerization. Figure 1 shows that Asn converts to Asp or isoaspartic acid (isoAsp) via a succinimide intermediate.23 Isoaspartic acid can subsequently compromise protein stability and activity, induce immune reactions, or shorten the biological half-life of proteins.24-26

E1MPSV_2024_v15n2_107_2_f0001.png 이미지

Figure 1. Pathway for spontaneous deamidation, isomerization, and racemization in Asn and Asp.

To control deamidation in protein pharmaceuticals, various analytical methods such as cation-exchange chromatography (CEX), capillary electrophoresis (CE), and reversed-phase chromatography (RP) are widely used. These methods are used to explore charge variants resulting from deamidation, a prevalent occurrence in protein therapeutics.1, 27-38

However, in analyses using chromatography methods like CEX or RP and gel electrophoresis techniques such as CE, it is challenging to identify the types of proteins within peaks generated by chromatography or gel electrophoresis without using mass spectrometry.

Mass spectrometry offers an alternative approach to characterize the deamidation status of proteins, including therapeutic monoclonal antibodies.39-47 Specifically, the integration of continuous mass analysis with RP provides significant advantages for deamidation analysis owing to its high sensitivity, speed, and specificity. Peptide mapping via LC-MS typically enables the prediction of quantitative ratios of deamidated peptides, facilitating relative quantification. Typically, the identification and quantification of deamidated peptides relative to non-deamidated peptides are conducted to assess the relative abundance ratio of deamidated peptides.30 However, various factors may influence peptide mapping during the analytical process. Indepth research is required to identify the factors impacting the determination of quantitative ratios. It is essential to estimate factors causing inevitable variations between experiments and compare them through experiments to explore and confirm conditions that minimize quantitative errors as much as possible. Furthermore, it is necessary to validate the quantitative results of ionized peptides obtained from mass spectrometry with conventional quantification methods.

Trastuzumab, a representative first-generation monoclonal antibody (mAb) widely used in treating HER2-positive breast cancer patients,48-50 has accumulated various studies on its chemical mechanisms, development of biological analysis methods, formulation stability, forced degradation, and biological impact assessment.48-50 The charge heterogeneity of trastuzumab is known to arise from Asn deamidation and Asp isomerization.27,41,50 Therefore, this study aims to investigate and confirm the validity and suitability of a method for quantifying the deamidation ratio of trastuzumab induced by stringent conditions, leading to asparagine deamidation and aspartic acid isomerization to isoaspartic acid. To achieve this, we will compare the results with the current CEX- and CE-based quantification methods.

Experimental Section

Materials and Reagents

Materials

Trastuzumab was obtained from Roche (Basel, Switzerland) through the Korea Food and Drug Administration.

Cation-Exchange Chromatography Reagents

Sodium phosphate monobasic monohydrate (Cat. No. S9638), sodium phosphate dibasic heptahydrate (Cat. No. S9390), sodium chloride (Cat. No. S6191), and Trizma® base (Cat. No. T6066) were obtained from Sigma-Aldrich (St. Louis, Missouri, USA). Water of HPLC grade (Cat. No. AH365-4) was sourced from Honeywell (North Carolina, USA). Hydrochloric acid fuming 37% (Cat. No. 1.00317) was procured from Merck (Darmstadt, Germany). The column used was a Shim-pack Bio IEX SP-NP (5 μm, 4.6 × 100 mm) (Part No. 227-31006-03) from Shimadzu (Kyoto, Japan). The filters used were Amicon® Ultra-0.5 mL centrifugal filters (10 kDa) (REF UFC501096) from Merck Millipore (Massachusetts, USA).

Protein Charge Variant Analysis Reagents

Anhydrous (99.8%) N,N-dimethylformamide (Cat. No. 227056-100mL) was obtained from Sigma-Aldrich (St. Louis, Missouri, USA). The DNA 5K/RNA/CZE 24 Chip (P/N: CLS138949) and HT Protein Charge Variant Kit (P/N: CLS760670) were procured from PerkinElmer (Massachusetts, USA). The 96-well PCR plate (REF PCR-96-FSC) was acquired from Axygen (California, USA). Amicon® Ultra-0.5mL Centrifugal Filters (30kDa, REF UFC503096) were purchased from Merck Millipore (Massachusetts, USA).

Peptide Mapping Reagents

Guanidine hydrochloride (Cat. No. G4505), Iodoacetic acid (IAA, Cat. No. I4386), Trizma base (Cat. No. T6066), and formic acid (Cat. No. 695076) were purchased from Sigma-Aldrich (St. Louis, Missouri, USA). Trypsin (Cat. No. 11418025001), chymotrypsin (Cat. No. 11418467001), and Glu-C (Cat. No. 11047817001) were obtained from Roche (Switzerland). PNGase F (Cat. No. P0705L) was purchased from NEB (Massachusetts, USA). Acetonitrile (Cat. No. AH015-4) and water (Cat. No. AH365-4) were acquired from Honeywell (North Carolina, USA).

Software

PMI-Byos (Ver 5.1.1) were purchased from Protein Matrics INC (California, USA).

Sample Preparation

To investigate charge variants derived from trastuzumab, harsh treatments were applied by adjusting pH and temperature. A 2.0 mg sample was subjected to buffer exchange into a 50 mM Tris-HCl (pH 8.5) solution using an Amicon filter. The exchanged sample was then concentrated to approximately 2.0 mg/mL in the same 50 mM Tris-HCl (pH 8.5) buffer. Subsequently, it was incubated in a convection oven at 37? for 0, 1, and 2 weeks to undergo the harsh treatments of pH and temperature.12,19

Methods

Cation-Exchange Chromatography

The sample prepared for cation-exchange chromatography (CEX-Ultra High Performance Liquid Chromatography) analysis was concentrated using an Amicon filter and subsequently adjusted to 1.0 mg/mL using mobile phase A (50 mM sodium phosphate, pH 6.5). To separate trastuzumab’s charge variants, ion gradient elution was performed using a Shim-pack Bio IEX SP-NP (5 μm, 4.6 × 100 mm) column coupled with an Agilent 1290 UHPLC system.

Mobile phase A was prepared as a 50 mM sodium phosphate solution by dissolving 5.47 g of sodium phosphate monobasic monohydrate and 2.79 g of sodium phosphate dibasic heptahydrate in 1 L of water and adjusting the pH to 6.5 ± 0.05. Mobile phase B consisted of a solution containing 50 mM sodium phosphate and 150 mM sodium chloride, with 58.44 g of sodium chloride dissolved in 1 L of water. For this solution, 150 mL of 1 M sodium chloride, 4.16 g of sodium phosphate monobasic monohydrate, and 5.33 g of sodium phosphate dibasic heptahydrate were dissolved in 850 mL of water, and the pH was adjusted to 6.5 ± 0.05.

The UHPLC system was configured with an automatic sample injector temperature of 4ºC and a column oven temperature of 25ºC. The mobile phase flow rate was set at 0.6 mL/min, and elution was carried out using a linear gradient of mobile phase B, increasing from 0% to 20% over 16 min. UV absorbance was measured at 214 nm, with a reference wavelength of 380 nm.

Protein Charge Variant Analysis

The samples obtained from forced degradation studies and the untreated samples were buffer-exchanged into distilled water (DW) using a 30K Amicon filter and prepared at a 2.0 mg/mL concentration. Subsequently, for fluorescence labeling, 25 μL of the sample at a concentration of 2.0 mg/mL was taken, followed by the addition of 5 μL of labeling buffer and 5 μL of dye solution (N,N-dimethylformamide 145 μL + dye concentrate 5 μL). The mixture was thoroughly mixed and allowed to react at room temperature for 10 min. Upon completion of the reaction, 60 μL of DW was added to the samples, mixed, and aliquoted into a 96-well PCR plate. The samples in the 96-well PCR plate were prepared by centrifuging at 1000 rpm for 1 min.

The DNA 5K/RNA/CZE 24 Chip and LabChip GXII Touch 24 systems were employed for charge variant analysis. The analysis was performed using a running buffer with pH 6.1 and the HT Protein Charge Variant 90s method.51

Peptide Mapping

Approximately 1 mg of each sample from forced degradation studies and untreated controls were taken. They were denatured and reduced by heating at 60ºC for 30 min in denaturing buffer (7 M Guanidine-HCl, 0.25 M Tris-HCl, pH 8.4) containing 5 mM DTT. Subsequently, alkylation was performed by adding iodoacetic acid (IAA) to a final concentration of 7.5 mM for 60 min at 25ºC. The alkylated and denatured samples were fractionated using a PD-10 column (GE Healthcare) with a buffer of 50 mM Tris-HCl, pH 7.5, and 50 μL of the sample was collected from the fraction exhibiting the highest absorbance at A280. Next, 2 μL of PNGase F was added to the samples, and deglycosylation was performed at 37ºC for 99 min in a rapid enzyme reactor. Following this, peptide mapping was performed for each enzyme. Trypsin and chymotrypsin were added to the samples at a ratio of 25:1 (protein to enzyme), and digestion was carried out at 37ºC for 18 h. Glu-C was added to the samples at the same ratio of 25:1 (protein to an enzyme), and digestion was performed at 25ºC for 18 h to yield peptides digested by each enzyme.

The peptides digested by each enzyme were separated using an ACQUITY CSH C18 column (1.7 μm, 2.1 × 100 mm, Waters). Mobile phase A comprised 0.1% formic acid in water, while mobile phase B comprised 0.1% formic acid dissolved in acetonitrile (ACN). The flow rate was maintained at 0.3 mL/min, and the column temperature was held at 60ºC, with a gradient changing from 2% B to 45% B over 115 min. Samples were stored at 6? during analysis.

The LC-MS analysis was performed using a Thermo Ultimate 3000 system coupled with an Orbitrap Fusion™ Tribrid™ mass spectrometer equipped with an ESI source (Thermo Scientific). Source settings for the mass spectrometer were optimized for CID mode, including Sheath Gas (Arb): 50, Aux Gas (Arb): 10, Ion Transfer Temp (º C): 325, Vaporizer Temp (ºC): 350, and static Spray Voltage (Positive Ion (V): 3500). MS/MS analysis was conducted in MS mode with a resolution of 60000 in the Orbitrap detector. The scan range spanned from 220 to 3000, with RF Lens (%): 60, AGC Target: 2.0e5, Maximum Injection time (ms): 50, and Microscans: 1. MS/MS fragmentation targeted the top m/z intensities during the initial 3 s of MS analysis. Precursor ions with a minimum threshold of 1.0e4 were identified, and those with charge states ranging from 1 to 6 were selected for MS/MS fragmentation, employing a 5 s exclusion window following a single occurrence. For CID mode analysis concerning deamidation, a collision energy of 38% was determined based on the m/z and charge state of precursor ions. For electron transfer dissociation (ETD) mode analysis addressing isomerization, fragmentation was carried out with a reaction time of 150 ms between precursor ions and radicals.

Data Analysis

We used PMI-Byos software (Ver 5.1.1, Protein Metrics Inc.) for data analysis. The precursor mass tolerance was set to 20 ppm, and the fragment mass tolerance was 0.3 Da. With these parameters, we processed the raw data obtained from peptide mapping through mass analysis. During this process, we identified modified and unmodified peptides using MS and MS/MS, extracted their respective XICs corresponding to the molecular weights, and obtained area values for quantification. The deamidation ratio is calculated using XIC by comparing the area values of deamidated peptides with wildtype peptides, as shown in the following equation.52

Peptides with various charges are generated during the ionization process owing to the unique characteristics of each peptide. Because the ionization efficiency differs among peptides with different charges, an extracted ion chromatogram (XIC) is generated for each charged peptide and subsequently summed for calculation. Moreover, peptides generated by protein digestion enzymes result in missed cleavage and non-specific peptides containing sites of Asn deamidation in varying proportions depending on enzyme efficiency. Therefore, the XIC for each charged missed cleavage peptide and non-specific peptide containing sites of Asn deamidation are summed for calculation. The modification percentage is quantified by determining the ratio of the XIC area for modified peptides obtained through this process to the sum of XIC areas for modified and unmodified peptides, thus providing the final quantification.

In the case of Asp isomerization, where two peptides have the same molecular weight, analysis was based on the difference in retention time observed in LC because the peptides exhibit identical MS and MS/MS spectra.

Results

Cation-Exchange Chromatography

In cation-exchange chromatograms, charge variants are differentiated as acidic peaks, which demonstrate faster retention times compared to the main peak, and basic peaks, which exhibit slower retention times (Figure 2).53 However, current separation system technologies do not achieve complete separation of charge variants on cationexchange columns, leading to their detection as overlapping peaks.54 Consequently, precise quantification of specific modifications becomes challenging, though the variant characteristics remain distinguishable.

E1MPSV_2024_v15n2_107_4_f0001.png 이미지

Figure 2. Cation-exchange chromatograms of stressed trastuzumab.

To investigate the charge variants of trastuzumab induced under harsh conditions, we analyzed them using CEX and assigned each peak based on mass analysis results from previous research (Table 1).50 Stress-induced trastuzumab (1, 2 weeks) exhibited 7 acidic variants (A1–A7) and 2 basic variants (B1–B2) relative to the main peak. However, owing to the technical limitations of CEX, deamidation levels were calculated by aggregating the expected variations of each amino acid across all peaks. Consequently, significant changes were observed in Asn30 and Asn55 over the stress periods. The deamidation level of Asn30 increased from 25.97% to 76.24% over the stress periods of 0, 1, and 2 weeks, respectively, while that of Asn55 increased from 1.59% to 23.86%.

Table 1. Expected modifications and ratios of stressed trastuzumab.

E1MPSV_2024_v15n2_107_4_t0001.png 이미지

*N.D: Not detected

Protein Charge Variant Analysis

In protein charge variant analysis, observed charge variants move at different rates relative to the wild type (main peak) as a reference. Basic variants move fast toward the cathode owing to stronger electrophoretic forces, while acidic variants move more slowly, eluting later compared to the wild type. This allows for the differentiation of basic variants (B1, B2), wild type (Main), and acidic variants (A1–A7) (Figure 3).55,56 Analytical findings reveal that with increasing stress duration over 0, 1, and 2 weeks, the wild type (main peak) detected at 35 s under the 0 time stress condition disappears, and acidic variants emerge in the acidic region. The proportions of each basic variant (B1, B2), wild type (Main), and acidic variants (A1–A7) identified in the electropherogram under various stress conditions are summarized in Table 2.

E1MPSV_2024_v15n2_107_5_f0001.png 이미지

Figure 3. (a) Bands and (b) electropherogram of charge variant for stressed trastuzumab observed on the LabChip® high pI charge variant assay

Table 2. Expected modification ratio of stressed trastuzumab

E1MPSV_2024_v15n2_107_5_t0001.png 이미지

*N.D: Not detected

Peptide Mapping

Peptide mapping analysis of stressed trastuzumab samples aimed to identify major modification sites and enable comparative quantification. This involved using mass spectrometry to measure raw data, verifying peptide masses with the Byos program, and validating the peptide masses using MS/MS.52

We have confirmed Asn deamidation in H:N55, H:N84, H:N318, H:N387, H:N392, H:N393, H:N437, L:N30, and L:N137 of trastuzumab. To quantitatively analyze the deamidation of the ASQDVN30TAVAWYQQK peptide from L:N30, we first identified its unmodified (top, Figure 4) and deamidated (bottom, Figure 4) forms using MS/MS results. A shift of 0.984 Da in the y10 and b6 ions, as depicted in Figure 4, confirms deamidation in L:N30.

E1MPSV_2024_v15n2_107_6_f0001.png 이미지

Figure 4. Annotated MS2 spectra of a deamidated N30 (top) and its corresponding unmodified form (bottom).

Figure 5 shows the deconvoluted XIC of the ASQDVN30TAVAWYQQK peptide. The deamidated form of N30 exhibits a retention time centered at approximately 35.1 min, as indicated by the peak. The unmodified ASQDVN30TAVAWYQQK peptide is denoted by the solid line peak around 33.4 min.

E1MPSV_2024_v15n2_107_6_f0002.png 이미지

Figure 5. A series of XICs quantifying L:N30 deamidation of L:T3 peptide for each stress condition.

The area under the extracted ion chromatogram (XIC) corresponding to the molecular weight of each confirmed peptide served as the primary data (Figure 6). The ratio of deamidation based on XIC area was calculated using the XIC area values of deamidated and wildtype peptides. Specifically, the XIC area values for each deamidated and wildtype peptide charge state were summed. Furthermore, the XIC area values for missed cleavage peptides and non-specific peptides containing the site of Asn deamidation were aggregated for each respective charge state. The modification percentage was then calculated by dividing the XIC area of modified peptides by the sum of the XIC areas of modified and unmodified peptides (Table 3).

Table 3. Calculation of the deamidation ratio for L:N30

E1MPSV_2024_v15n2_107_7_t0001.png 이미지

E1MPSV_2024_v15n2_107_7_f0001.png 이미지

Figure 6. ETD mechanism for generating reporter ions​​​​​​​

The quantification of Asn deamidation in H:N55, H:N84, H:N318, H:N387, H:N392, H:N393, H:N437, and L:N137 of Trastuzumab followed the same procedure employed for L:N30 (Table 4).

Table 4. Deamidation ratio of stressed trastuzumab.​​​​​​​

E1MPSV_2024_v15n2_107_7_t0002.png 이미지

Mass spectrometry alone cannot differentiate Asp isomerization because the molecular weight remains unchanged even if Asp is isomerized. However, employing Tandem Mass ETD allows the identification of reporter ions specific to isomerized Asp. Using this technique, we confirmed the presence of Asp reporter ions in the XIC peaks of each validated peptide.

The ETD method involves electron absorption by amino acids from radical cations. This leads to the cleavage of the bond between the amino group and the alpha carbon, generating c and z• ions. Therefore, isoAsp generates c +57 and z• −57 fragmentation ions owing to the breakdown of the bond between the alpha and beta carbons, a phenomenon absent in Asp. Thus, detecting c +57 and z• −57 fragmentation ions confirms isoAsp formation (Figure 6).57-59

Thus, ETD analysis revealed isoAsp sites in trastuzumab at H:N55, H:D102, H:N137, H:N318, and H:N437, with the proportion of each isoAsp detailed in Table 5. The deamidated H:T6 peptide containing the H:N55 site exhibited three XIC peaks. Confirmation of isoAsp presence was achieved by detecting z• −57 reporter ions in the MS/MS spectra of peaks (a) and (c) (Figure 7), enabling differentiation between isomerization and deamidation peaks in the XIC. The isomerization ratio was determined using the same method as the deamidation ratio (Table 5). Moreover, it was observed that the size of the isomerization peak in the XIC increased with the duration of stress (Figure 8). The quantification of isoAsp content in H:D102, H:N137, H:N318, and H:N437 of trastuzumab followed the identical procedure used for H:N55.

E1MPSV_2024_v15n2_107_8_f0001.png 이미지

Figure 7. Extracted ion chromatogram and ETD spectra of the H:T6 peptide from the 2-week sample. The reporter ions of isoAsp (z5−57) are observed only in peaks (a) and (c), not in peak (b)

E1MPSV_2024_v15n2_107_9_f0001.png 이미지

Figure 8. A series of XICs quantifying H:N55 isomerization of the H:T6 peptide for each stress condition

Table 5. The isoAsp ratio of stressed trastuzumab.​​​​​​​

E1MPSV_2024_v15n2_107_9_t0001.png 이미지

Table 4 summarizes the deamidation sites of each Asn in trastuzumab, organized by enzyme using peptide mapping methods, along with their respective proportions. Similarly, the isomerization ratios of Asn were also compiled using the same method (Table 5).

In the peptide mapping results processed with trypsin, the deamidation of N30 at CDR1 increased from 7.76% to 87.88% over 2 weeks. Similarly, the deamidation of H:N55 at CDR2 increased from 4.75% to 15.00%, a slightly lower increase than that of L:N30. Isomerization of H:D102 in CDR3 was not detected after trypsin treatment. In the Fc region of trastuzumab, H:N392 or N393 deamidation increased from 0% to 8.25% after 2 weeks of stress. Furthermore, deamidation was confirmed in five additional amino acids: H:N84 increased from 1.04% to 1.71%, H:N318 from 2.94% to 4.17%, H:N387 from 1.08% to 8.57%, H:N437 from 0.66% to 0.81%, and L:N137 from 0.19% to 0.34%. Additionally, isomerization was observed in three amino acids: H:N55 increased from 4.73% to 14.71%, H:N318 from 1.15% to 1.31%, and H:N437 from 0.43% to 0.86% after 2 weeks of stress.

In the peptide mapping results using chymotrypsin, the deamidation of L:N30 in CDR1 increased from 0.19% to 81.72% after 2 weeks. Similarly, the deamidation of H:N55 in CDR2 increased from 1.87% to 6.53%, slightly lower than L:N30. Isomerization of H:D102 in CDR3 increased from 20.86% to 47.23% after 2 weeks. Deamidation of H:N392 in the Fc region of trastuzumab also increased from 0% to 14.39% after 2 weeks of stress. Additionally, deamidation was observed in four additional amino acids: H:N84 increased from 1.25% to 1.65%, H:N318 from 3.90% to 5.69%, H:N437 from 6.58% to 33.60%, and L:N137 from 1.24% to 2.03%. Furthermore, isomerization was detected in H:N55, increasing from 0.1% to 0.26% after 2 weeks of stress.

The peptide mapping results obtained with Glu-C showed no evidence of deamidation of L:N30 in CDR1 following Glu-C treatment. However, the deamidation of H:N55 in CDR2 increased from 0% to 14.10% after 2 weeks. Isomerization of H:D102 in CDR3 was not detected after Glu-C treatment. In the Fc region of trastuzumab, the deamidation of H:N392 increased from 1.28% to 14.47% after 2 weeks of stress. Additionally, deamidation was observed in two additional amino acids: H:N437 increased from 0.88% to 1.08%, and L:N137 increased from 0.03% to 0.06%. Furthermore, isomerization occurred in three amino acids: H:N55 increased from 0 to 16.00%, H:N137 increased from 0.06% to 0.08%, and H:N437 increased from 0.96% to 1.04% after 2 weeks of stress.

Discussion

Studies are underway to qualitatively and quantitatively assess the frequency and extent of Asn deamidation and Asp isomerization.41,60 While relative quantification is performed using ion chromatography and CE, qualitative analysis has limitations. Even with ion chromatography or CE application, confirming qualitatively Asn deamidation and Asp isomerization at specific protein sites and measuring their extent remains challenging.43,61,62 In this study, the qualitative identification of modified peptides within each peak in the CEX analysis relied on previous research for estimation, and the lack of reference data in charge variant analysis hindered estimation.

Therefore, LC-MS based peptide mapping is a prominent experimental approach to address these challenges.40 This method enables the identification of amino acids where Asn deamidation and Asp isomerization occur in protein sequences and quantification through the measurement of ionized peptide fragments. Consequently, qualitative and quantitative analyses of Asn deamidation and Asp isomerization can be conducted simultaneously. Furthermore, it is possible to estimate the extent of these modifications by examining the peptides where Asn deamidation and Asp isomerization occur and measuring their approximate ratios.63

Because each peptide varies in molecular size and degree of ionization, mass spectrometry may not always be suitable for quantitative peptide analysis. Therefore, for relative quantification using mass spectrometry, the peptides under examination must have appropriate sizes and charges for mass spectrometric analysis. Fortunately, when measuring the ratio of Asn deamidation and Asp isomerization, relative quantification is possible because the sizes of native and modified peptides are nearly identical. Therefore, peptides containing these modifications must be cleaved to sizes and charges suitable for mass spectrometric analysis to accurately measure the occurrence ratio of Asn deamidation and Asp isomerization within proteins. However, when Asn deamidation occurs, there is a change of 1 in the peptide's charge, resulting in differences in ionization levels between peptides with and without Asn deamidation. Additionally, peptides with various charges are generated during ionization, necessitating the differentiation of all peptides by charge, and their quantified values confirmed in XIC should be summed accordingly (Table 3).

This study underscores the critical importance of the type and efficiency of the enzymes employed during protein cleavage in quantitative peptide analysis through peptide mapping. Miss-cleaved or non-specific peptides may be generated depending on the type and efficiency of the enzyme used in the protein cleavage process. Indeed, in this experiment, a significant presence of peptides with misscleavage or non-specific sites was observed. Therefore, to obtain accurate quantitative values, it is imperative to sum the XICs of miss-cleaved peptides or non-specific peptides according to each charge state (Table 3).

The importance of enzyme selection in peptide mapping analysis was evident from examining the experimental results obtained with trypsin, Glu-C, and chymotrypsin (Table 4). The trend in Asn deamidation ratios generally exhibited similarity across the various enzyme treatments, as depicted in Table 4. However, significant discrepancies arose with the Glu-C treatment (Table 4). This disparity likely occurs because Glu-C cleaves the Glu-C-terminal region of protein sequences. However, in some cases, it can also cleave the Asp C-terminal region. Particularly in the analysis of Asn deamidation, if Asn undergoes deamidation, it converts to Asp, thereby exposing the Glu-C cleavage site Asp and resulting in cleavage at that site. In such instances, if deamidation occurs in the L:G4 peptide, N29 is substituted with D29, forming the “VD” peptide, which remains unidentified in peptide mapping. Consequently, while the deamidation ratio of L:N30 was 90% when analyzed with trypsin and chymotrypsin, it remained undetected when analyzed with Glu-C.

When comparing trypsin and chymotrypsin treatments, similar Asn deamidation ratios were observed across 0, 1, and 2-week periods when the sizes of peptides containing the same Asn deamidation site were comparable. Chymotrypsin, with a higher number of cleavage sites by default, can break down proteins into smaller fragments. While this may offer an advantage for relative quantification owing to the smaller peptide sizes, the abundance of non-specific sites poses challenges in adequately extracting these peptides using software, inevitably leading to a degree of error. Therefore, it is anticipated that trypsin-treated peptide maps generally provide a sufficient number of peptides of appropriate sizes. Still, supplementation with chymotrypsin may be considered in cases where this is not achieved.

The experimental findings regarding Asp isomerization with trastuzumab are summarized in Table 5, delineating observations based on the enzyme utilized. As depicted in the results, Asp isomerization was detected at H:N55, H:N318, and H:N437 with trypsin treatment, at H:N55, H:N137, and H:N437 with Glu-C treatment, and solely at H:N55 with chymotrypsin treatment. However, even at H:N55, the confirmed ratio was low when chymotrypsin was employed, indicating its unsuitability for analyzing Asp isomerization. Peptides treated with chymotrypsin generally exhibited very low intensity in ETD. Attempts were made to improve this, but optimal conditions were not found. In contrast, for trypsin and Glu-C analyses at H:N55, trypsin exhibited 10.03% at week 1 and 14.08% at week 2. Glu-C analysis demonstrated 8.12% at week 1 and 16.00% at week 2, yielding relatively similar results between the two enzymes. However, despite distinct peaks with matching molecular weights on the RP column of UPLC, uncertainty persists regarding whether these peaks represent 100% isoAsp, even when reporter ions are integrated. Additionally, the intensity of reporter ions was insufficient, and irregularly shaped reporter ions were not detected.

Conclusion

We identified a consistent trend by juxtaposing the results of ion exchange chromatography and protein ion variant analysis performed on the same sample with the peptide mapping experiment results using trastuzumab. Specifically, we confirmed a decrease in the intact form of trastuzumab under harsh heat conditions, while acidic variants increased (Tables 1,2,4, and 5). This trend aligns with the increasing tendencies of Asn deamidation and Asp isomerization observed in the peptide mapping experiment (Tables 1,2,4, and 5). This finding underscores the reliability of assessing deamidation rates through peptide mapping, thereby enhancing the clarity of experimental interpretations.

In this study, we employed LC-MS based peptide mapping to identify the extent of deamidation and isoAsp formation in trastuzumab. However, our method is based on relative quantification, comparing intact peptides with deamidated peptides or peptides containing isoAsp, and thus does not provide absolute quantification. Therefore, additional research is needed to validate the absolute quantification performance of this method. Furthermore, research should verify the general applicability of this method on various therapeutics beyond trastuzumab. Consequently, we are currently planning further experiments on diffierent protein therapeutics to solidify the utility of this method.

Conflict of interest

The authors declare no conflict of interest.

Acknowledgments

This work is supported by Binex Co., Ltd.

References

  1. Haberger, M.; Bomans, K.; Diepold, K.; Hook, M.; Gassner, J.; Schlothauer, T.; Zwick, A.; Spick, C.; Kepert, J.F.; Hienz, B.; Wiedmann, M.; Beck, H.; Metzger, P.; Molhoj, M.; Knoblich, C.; Grauschopf, U.; Reusch, D.; Bulau, P., mAbs 2014, 6, 327-339. https://doi.org/10.4161/mabs.27876.
  2. Gupta, S.; Jiskoot, W.; Schoneich, C.; Rathore, A.S., Journal of Pharmaceutical Sciences 2022, 111, 903-918. https://doi.org/10.1016/j.xphs.2021.11.024. 
  3. Furman, J.L.; Chiu, M.; Hunter, M.J., American Association of Pharmaceutical Scientists 2015, 17, 111-120. https://doi.org/10.1208/s12248-014-9681-9. 
  4. Schmid, I.; Bonnington, L.; Gerl, M.; Bomans, K.; Thaller, A.L.; Wagner, K.; Schlothauer, T.; Falkenstein, R.; Zimmermann, B.; Kopitz, J.; Hasmann, M.; Bauss, F.; Haberger, M.; Reusch, D.; Bulau, P., Communications Biology 2018, 1, 28. https://doi.org/10.1038/s42003-018-0032-8. 
  5. Haberger, M.; Heidenreich, A.-K.; Hook, M.; Fichtl, J.; Lang, R.; Cymer, F.; Adibzadeh, M.; Kuhne, F.; Wegele, H.; Reusch, D.; Bonnington, L.; Bulau, P., Journal of the American Society for Mass Spectrometry 2021, 32, 2062-2071. https://doi.org/10.1021/jasms.0c00446. 
  6. Nugrahadi, P.P.; Hinrichs, W.L.J.; Frijlink, H.W.; Schoneich, C.; Avanti, C., Pharmaceutics 2023, 15, 935. https://doi.org/10.3390/pharmaceutics15030935. 
  7. Harding, J.J.; Beswick, H.T.; Ajiboye, R.; Huby, R.; Blakytny, R.; Rixon, K.C., Mechanisms of Ageing and Development 1989, 50, 7-16. https://doi.org/10.1016/0047-6374(89)90054-7. 
  8. Hovorka, S.W.; Schoneich, C., Journal of Pharmaceutical Sciences 2001, 90, 253-269. https://doi.org/10.1002/1520-6017(200103)90:3<253::AID-JPS1>3.0.CO;2-W. 
  9. Frokjaer, S.; Otzen, D.E., Nature Reviews Drug Discovery 2005, 4, 298-306. https://doi.org/10.1038/nrd1695. 
  10. Russell, R.G.G.; Watts, N.B.; Ebetino, F.H., Osteoporosis International 2008, 19, 733-759. https://doi.org/10.1007/s00198-007-0540-8. 
  11. Gervais, D., Journals of Chemical technology and Biotechnology 2016, 91, 569-575. https://doi.org/10.1002/jctb.4850. 
  12. Ying, Y.; Li, H., Methods 2022, 200, 42-57. https://doi.org/10.1016/j.ymeth.2020.06.009. 
  13. Reissner, K.J.; Aswad, D.W., Cellular and Molecular Life Sciences 2003, 60, 1281-1295. https://doi.org/10.1007/s00018-003-2287-5. 
  14. Hao, P.; Adav, S.S.; Gallart-Palau, X.; Sze, S.K., Mass spectrometry review 2017, 36, 677-692. https://doi.org/10.1002/mas.21491. 
  15. Giorgianni, F.; Desiderio, D.M.; Beranova-Giorgianni, S., Electrophoresis 2003, 24, 253-259. https://doi.org/10.1002/elps.200390021. 
  16. Gervais, D., Journals of Chemical technology and Biotechnology 2016, 91, 569-575. https://doi.org/10.1002/jctb.4850. 
  17. Yang, H.; Zubarev, R.A., Electrophoresis 2010, 31, 1764-1772. https://doi.org/10.1002/elps.201000027. 
  18. Lu, X.; Nobreg, R.P.; Lynaugh, H.; Jain, T.; Barlow, K.; Boland, T; Sivasubramanian, A.; Vasquez, M.; Xu, Y., mAbs 2019, 11, 45-57. https://doi.org/10.1080/19420862.2018.1548233. 
  19. Pace, A.L.; Wong, R.L.; Zhang, Y.T.; Kao, Y.-H.; Wang, Y. J., Journal of Pharmaceutical Sciences 2013, 102, 1712-1723. https://doi.org/10.1002/jps.23529. 
  20. Goulet, D.R.; Atkins, W.M., Journal of Pharmaceutical Sciences 2020, 109, 74-103. https://doi.org/10.1016/j.xphs.2019.05.031. 
  21. Liu, Z.; Gunasekaran, K.; Wang, W.; Razinkov, V.; Sekirov, L.; Leng, E.; Sweet, H.; Foltz, I.; Howard, M.; Rousseau, A.-M.; Kozlosky, C.; Fanslow, W.; Yan, W., Journal of Biological Chemistry 2014, 289, 3571-3590. https://doi.org/10.1074/jbc.M113.513366. 
  22. Huang, L; Lu, J.; Wroblewski, V.J.; Beals, J.M.; Riggin, R.M., Anal. Chem. 2005, 77, 1432-1439. https://doi.org/10.1021/ac0494174. 
  23. Reissner, K.J.; Aswad, D.W., Cellular and Molecular Life Sciences 2003, 60, 1281-1295. https://doi.org/10.1007/s00018-003-2287-5. 
  24. Bickel, F.; Griaud, F.; Kern, W.; Kroener, F.; Gritsch, M.; Dayer, J.; Barteau, S.; Denefeld, B.; Kao-Scharf, C.-Y.; Lang, M.; Slupska-Muanza, I.; Schmidt, C.; Berg, M.; Sigg, J.; Boado, L.; Chelius, D., mAbs 2023, 15, e2151075. https://doi.org/10.1080/19420862.2022.2151075. 
  25. Wang, J.; Guo, C.; Meng, Z.; Zwan, M.D.; Chen, X.; Seelow, S.; Lundstrom, S.L.; Rodin, S.; Teunissen, C.E.; Zubarev, R.A., Alzheimer's & Dementia 2022, 19, 1491-1502. https://doi.org/10.1002/alz.12735. 
  26. Rouse, J.C.; Stephens, E.; Philip, S.; English, A.M.; Marzilli, L.A., LCGC 2017, 16, 6-14. https://www.chromatographyonline.com/view/minimizing-methodinduced-deamidation-and-isomerization-during-antibody-characterization-ensure-op-0. 
  27. Harris, R.J.; Kabakoff, B.; Macchi, F.D.; Shen, F.J.; Kwong, M.; Andya, J.D.; Shire, S.J.; Bjork, N.; Totpal, K.; Chen, A.B., J. Chromatogr. B Biomed. Appl. 2021, 752, 233-245. https://doi.org/10.1016/S0378-4347(00)00548-X. 
  28. Farjami, A.; Siahi-Shadbad, M.; Akbarzadehlaleh, P.; Molavi, O., Chromatographia 2018, 81, 1649-1660. https://doi.org/10.1007/s10337-018-3627-9. 
  29. Weitzhandler, M.; Farnan, D.; Horvath, J.; Rohrer, J.S.; Slingsby, R.W.; Avdalovic, N.; Pohl, C., J. Chromatogr. A 1998, 828, 365-72. https://doi.org/10.1016/s0021-9673(98)00521-4. 
  30. Feketea, S.; Beck, A.; Fekete, J.; Guillarme, D., Journal of Pharmaceutical and Biomedical Analysis 2015, 102, 282-289. https://doi.org/10.1016/j.jpba.2014.09.032. 
  31. Kim, D.G.; Kim, H.J.; Kim, H.J., Archives of Pharmacal Reserch 2016, 39, 1472-1481. https://doi.org/10.1007/s12272-016-0818-5. 
  32. Moritz, B.; Schnaible, V.; Kiessig, S.; Heyne, A.; Wild, M.; Finkler, C.; Christians, S.; Mueller, K.; Zhang, L.; Furuya, K.; Hassel, M.; Hamm. M.; Rustandi, R.; He, Y.; Solano, O.S.; Whitmore, C.; Park, S.A.; Hansen, D.; Santos, M.; Lies, M., J. Chromatogr. B 2015, 983-984, 101-110. https://doi.org/10.1016/j.jchromb.2014.12.024. 
  33. Kaur, H.; Beckman, J.; Zhang, Y.; Li, Z.J.; Szigeti, M.; Guttman, A., TrAC 2021, 144, 116407. https://doi.org/10.1016/j.trac.2021.116407. 
  34. Kumar, R.; Guttman, A.; Rathore, A.S., Electrophoresis 2022, 43, 143-166. https://doi.org/10.1002/elps.202100182. 
  35. Zhao, S.S.; Chen, D.D.Y., Electrophoresis 2014, 35, 96-108. https://doi.org/10.1002/elps.201300372. 
  36. Tamizi, E.; Jouyban, A., Electrophoresis 2015, 36, 831-858. https://doi.org/10.1002/elps.201400343. 
  37. Chen, T.; Chen, Y.; Stella, C.; Medley, C.D.; Gruenhagen, J.A.; Zhang, K., J. Chromatogr. B 2016, 1032, 39-50. https://doi.org/10.1016/j.jchromb.2016.07.023. 
  38. Ascione, A.; Belfiore, M.; Vesterinen, J.; Buda, M.; Holtkampe, W.; Luciani, F., mAbs 2024, 16, 2313737. https://doi.org/10.1080/19420862.2024.2313737. 
  39. Annan, R.S.; Carr, S.A., Journal of Protein Chemistry 1997, 16, 391-402. https://doi.org/10.1023/A:1026384605285. 
  40. Wang, W.; Meeler, A.R.; Bergerud, L.T.; Hesselberg, M.; Byrne, M.; Wu, Z., International Journal of Mass Spectrometry 2012, 312, 107-113. https://doi.org/10.1016/j.ijms.2011.06.006. 
  41. Spanov, B.; Olaleye, O.; Mesurado, T.; Govorukhina, N.; Jungbauer, A.; Merbel, N.C.V.D.; Lingg, N.; Bischoff, R., J. Chromatogr. A 2023, 95, 3951-3958. https://doi.org/10.1021/acs.analchem.2c03275. 
  42. Huang, L.; Lu, J.; Wroblewski, V.J.; Beals, J.M.; Riggin, R.M., Anal. Chem. 2005, 77, 1432-1439. https://doi.org/10.1021/ac0494174. 
  43. Shi, R.L.; Xiao, G.; Dillon, T.M.; Ricci, M.S; Bondarenko, P.V., mAbs 2020, 12, e1739825. https://doi.org/10.1080/19420862.2020.1739825. 
  44. Diepold, K.; Bomans, K.; Wiedmann, M.; Zimmermann, B; Petzold, A.; Schlothauer, T.; Mueller, R.; Moritz, B.; Stracke, J.O.; Molhoj, M.; Reusch, D.; Bulau, P., PLoS One 2012, 7, e30295. https://doi.org/10.1371/journal.pone.0030295. 
  45. Bults, P.; Bischoff, R.; Bakker, H.; Gietema, J.A.; Merbel, N.C.V.D., Anal. Chem. 2016, 88, 1871-1877. https://doi.org/10.1021/acs.analchem.5b04276. 
  46. Wang, Y.; Li, X.; Liu, Y.-H.; Richardson, D.; Li, H.; Shameem, M.; Yang, X., mAbs 2016, 8, 1477-1486. https://doi.org/10.1080/19420862.2016.1226715. 
  47. Ponniah, G.; Kita, A.; Nowak, C.; Neill, A.; Kori, Y.; Rajendran, S.; Liu, H., Anal. Chem. 2015, 87, 9084-9092. https://doi.org/10.1021/acs.analchem.5b02385. 
  48. Nahta, R.; Esteva, F.J., Cancer Letters 2006, 232, 123-138. https://doi.org/10.1016/j.canlet.2005.01.041. 
  49. Harries, M.; Smith, I., Endocrine-Related Cancer 2022, 9, 75-85. https://doi.org/10.1677/erc.0.0090075. 
  50. Spanov, B.; Olaleye, O.; Lingg, N.; Bentlage, A.E.H.; Govorukhin, N.; Hermans, J.; Merbel, N.V.D.; Vidarsson, G.; Jungbauer, A.; Bischoff, R., J. Chromatogr. A 2021, 1655, 462506. https://doi.org/10.1016/j.chroma.2021.462506. 
  51. PerkinElmer, Inc, Protein Charge Variant Assay User Guide, 2022, PN CLS140162 Rev. E. 
  52. PROTEIN METRICS, Quantification of Modifications in the Peptide Workflows, 2024, https://support.proteinmetrics.com/hc/en-us/articles/20700549870612-Quantification-of-Modifications-in-the-Peptide-Workflows. 
  53. Johnson, K.A.; Paisley-Flango, K.; Tangarone, B.S.; Porter, T.J.; Rouse, J.C., Anal. Biochem. 2007, 360, 75-83. https://doi.org/10.1016/j.ab.2006.10.012. 
  54. Zhang, T.; Bourret, J.; Cano, T., J. Chromatogr. A 2011, 1218, 5079-5086. https://doi.org/10.1016/j.chroma.2011.05.061. 
  55. Khawli, L.A.; Goswami, S.; Hutchinson, R.; Kwong, Z.W.; Yang, J.; Wang, X.; Yao, Z.; Sreedhara, A.; Cano, T.; Tesar, D.; Nijem, I.; Allison, D.E.; Wong, P.Y.; Kao, Y.-H.; Quan, C.; Joshi, A.; Harris, R.J.; Motchnik, P., MAbs. 2010, 2, 613-624. https://doi.org/10.4161/mabs.2.6.13333. 
  56. Wheeler, T.D.; Sun, J.L.; Pleiner, S.; Geier, H.; Dobberthien, P.; Studts, J.; Singh, R.; Fathollahi, B., Anal. Chem. 2014, 86, 5416-5424. https://doi.org/10.1021/ac500497n. 
  57. Sargaeva, N.P.; Lin, C.; O'Connor, P.B., Anal. Chem. 2009, 81, 9778-9786. https://doi.org/10.1021/ac901677t. 
  58. Chan, W.Y.K.; Chan, T.W.D.; O'Connor, P.B., Journal of the American Society for Mass Spectrometry 2010, 21, 1012-1015. https://doi.org/10.1016/j.jasms.2010.02.002. 
  59. Ni, W.; Dai S.; Karger, B.L.; Zhou, Z.S., Anal. Chem. 2010, 82, 7485-7491. https://doi.org/10.1021/ac101806e. 
  60. Sargaeva, N.P.; Lin, C.; O'Connor, P.B., Anal. Chem. 2011, 83, 6675-6682. https://doi.org/10.1021/ac201223d. 
  61. Sadakane, Y.; Yamazaki, T.; Nakagomi, K.; Akizawa, T.; Fujii, N.; Tanimura, T.; Kaneda, M.; Hatanaka, Y., Journal of Pharmaceutical and Biomedical Analysis 2003, 30, 1825-1833. https://doi.org/10.1016/S0731-7085(02)00525-3. 
  62. Zhang, W.; Czupryn, M.J., Journal of Pharmaceutical and Biomedical Analysis 2003, 30, 1479-1490. https://doi.org/10.1016/S0731-7085(02)00479-X. 
  63. PROTEIN METRICS, Byos PTM Percent Modification Calculation, 2024, https://support.proteinmetrics.com/hc/en-us/articles/20611739819796-Byos-PTM-Percent-Modification-Calculation.