Clean and Complete Protein Digestion with an Autolysis Resistant Trypsin for Peptide Mapping

Abstract

Peptide mapping requires cleavage of proteins in a predictable fashion so that target protein-specific peptides can be reliably identified and quantified. Trypsin, a commonly used protease in this process, can also undergo self-cleavage or autolysis, thereby reducing the effectivity and even cleavage specificity at lysine and arginine residues. Here, we report highly efficient and reproducible peptide mapping of biotherapeutic monoclonal antibodies. We highlight the properties of a homogeneous chemically modified trypsin on thermal stability, a 54% increase in melting temperature with an 84% increase in energy required for unfolding, an indication of more thermally stable trypsin, >90% retained intact mass peak area after exposure to digestion conditions confirming autolysis resistance, 10× more intensity for intact enzyme compared to trypsin of similar source and narrower molecular weight distribution with LC-MS indicative of low degradation compared to 3 other types of trypsin. Finally, we show the utility of this autolysis-resistant trypsin in characterizing biotherapeutic monoclonal antibodies consistently and reliably showing a >30% reduction in missed cleavage for a short-duration protein digestion time of 30 min compared to heterogeneously modified trypsin of a similar source.

Introduction

Disease treatments are based on both small-molecule drugs and large-molecule-based biotherapeutics or biologics. The emerging trend indicates that biologics development is accelerating, with a projected growth of 9.25% in the next 5 years.¹ In-depth characterization of these biologics, including protein therapeutics, is a requirement for patient safety and maintaining regulatory standards. This inevitably involves identifying, analyzing, and quantifying critical quality attributes through peptide mapping approaches.² These attributes include sequence identification, protein purity levels, and post-translational modifications (PTMs). Peptide mapping often involves the cleavage of proteins into peptides of a suitable size for subsequent separation by liquid chromatography and sequencing by mass spectrometry on LC-MS instruments.³⁻⁵ One of the most common enzymes used in peptide mapping protocols is trypsin. This highly specific enzyme cleaves the peptide backbone on the C-terminal sides of lysine and arginine.

Qualitative and quantitative peptide mapping analysis of protein therapeutics requires a trypsin with predictable and reproducible protein digestion behavior with minimal autolysis (or self-hydrolysis). Various trypsin digestion protocols have been developed starting with sample preparation (denaturation, reduction, alkylation), desalting, and digestion conditions (pH, temperature, time, buffer types and concentrations of buffers, addition of divalent ions).^2−4,6−8 However, challenges such as protein sample complexity, stability, longer digestion times, variability of digestion-related post-translation modifications, and complex data analysis remain impediments to achieving reproducibility and routine-practice status. Longer digestion time was necessitated by trypsin self-hydrolysis (autolysis) during optimal digestion conditions. The enzyme-protein ratio is reduced to overcome autolysis, leading to slower digestion rates and a longer digestion time to achieve complete cleavage. When autolysis is unmitigated, a higher enzyme-protein ratio leads to a higher background of autolytic trypsin peptides and even nonspecific cleavage of target protein.^9,10 Furthermore, digestion protocols are also required to minimize nonspecific sample degradation and undesired PTMs.³

There have been many attempts to increase trypsin stability through the use of different species of trypsin (bovine vs porcine), the production of recombinant enzymes, the immobilization or chemical modifications of the enzyme (Figure 1).¹¹⁻¹⁹ Immobilization of trypsin on a solid support helps minimize autolysis through spatial separation of trypsin molecules. Unfortunately, most of the techniques for trypsin conjugation to a solid support are performed under conditions (such as a high pH, temperature, buffer type, and concentrations) that favor autolysis. Therefore, this process requires an enzyme inhibitor during attachment and removal after attachment for activity restoration. Immobilization can adversely change the configuration of the active enzyme, affecting the digestion performance. Moreover, the solid support surface may have high nonspecific binding properties, especially for hydrophobic peptides, resulting in a significant loss peptide response. Reduction in autolysis can also be achieved by chemical modifications, such as acetylation and methylation of lysine. However, such modifications, if heterogeneous, can lead to incomplete protection. Improvement of the autolysis resistance can ensure consistent digestion kinetics and cleavage performance.

Figure 1.

Illustrations of trypsin. (A) Sequence alignment of bovine versus porcine trypsin. (B) Crystal structure of porcine trypsin with the catalytic triad and aspartic acid binding residues marked with colored highlighting. (C) Crystal structure of porcine trypsin with residues containing primary amines highlighted in blue. (D) Cartoon representation of particle immobilized trypsin. Crystal structures images were created using Mol* and PDB 1S85 (Nucleic Acids Research, doi: 10.1093/nar/gkab314).

Thermal stability also influences the enzyme performance. In solution, differential scanning calorimetry (DSC) provides information about the biomolecular thermostability of a protein sample over a wide range of temperatures. Digestion conditions such as buffer types, pH, reagent additives, and concentrations can be included in the DSC experimental setup to understand the enzyme behavior. These studies can reveal unfolding events from tertiary, secondary, and primary structures at various temperatures when the enzymes undergo degradation. The experimental output in the form of unfolding and melting temperatures in the thermogram indicates the presence of multiple configurations or stabilization differences. When DSC data are combined and correlated with intact MS data, it can help to determine the stability of enzymes under different conditions, such as chemical modification or immobilization.

An in-depth comparison of analytical grade trypsin types was conducted via intact reversed-phase liquid chromatography-mass spectrometry (RPLC-MS), enzyme autolysis, and solution phase differential scanning calorimetry-based thermal stability studies. Intact mass analysis was used to evaluate the heterogeneity among the trypsin types according to their observed molecular weight distributions. Differential scanning calorimetry was used to show differences in melting temperatures, including calcium chloride-based stabilization. Finally, the data from the enzyme incubation by itself (without target protein) helped us understand the extent of autolysis occurring in various trypsin types under optimal digestion conditions. Subsequently, we demonstrate a rapid, high-efficiency, and reproducible peptide mapping workflow for infliximab using autolysis-resistant trypsin. In summary, a homogeneously methylated, recombinant porcine trypsin with enhanced autolysis resistance enables a high stoichiometric ratio of enzyme to target protein for faster and cleaner digestion, leading to minimal interference from autolytic peptides during peptide mapping experiments.

Experimental Section

Materials and Reagents

The Remicade (Infliximab) sample (S/N 100044793302, lot 21D042P1) was purchased from AmerisourceBergen (Besse Medical) (West Chester, OH) and reconstituted with water as directed (10 mg/mL). 8 M guanidine hydrochloride (catalog no. 24115) and trifluoroacetic acid (TFA) were purchased from Thermo Fisher Scientific. Pierce dithiothreitol (DTT) and iodoacetamide (IAM) were purchased from Thermo Fisher Scientific in no-weigh format (Cat #A39255 and A39271, respectively) and reconstituted with water to obtain 0.5 M stock solutions. The digestion buffer (100 mM Tris, 10 mM CaCl₂, pH 7.5; catalog no. 186010110) was from Waters Corp. (Milford, MA). Unique types of analytical grade lyophilized trypsins were acquired from various sources. Unmodified animal-derived bovine trypsin (Trypsin A; TPCK treated) was purchased from Worthington Biochemical Corporation (Cat LS003740; Lakewood, NJ). Animal-derived bovine trypsin with acetylated residues (Trypsin B) was obtained from New England Biolabs (Cat P8101; Ipswich, MA). Unmodified recombinant porcine trypsin (Trypsin C) was acquired from Roche Diagnostics (Cat RTRYP-RO; Mannheim, Germany). Animal-derived methylated porcine trypsin (Trypsin D) was purchased from Promega Corporation (Cat v5280; Madison, WI). Recombinant porcine trypsin with homogeneous and near-complete methylation (Trypsin E) was acquired from Waters Corporation (Cat 186010106; Milford, MA). Immobilized trypsin was obtained from Waters Corporation and was prepared using Trypsin A and 3.5 μm bridged ethylene hybrid particles (BEH).

Differential Scanning Calorimetry

The solution phase differential scanning calorimetry experiments were performed using 100 mM Tris-HCl buffer pH 7.5 (Sigma-Aldrich, St Louis, MO, USA)) with (10 or 100 mM) or without CaCl₂ (Sigma-Aldrich, St Louis, MO, USA). The buffers were prepared and used at Waters Corporation (Milford, MA) within 2 days. Benzamidine (Sigma-Aldrich) was used to stabilize trypsin which is prone to autolysis. Trypsin samples were prepared by dissolving in buffer (with or without CaCl₂) at 1 mg/mL concentration and sonicating at room temperature for 1 min. The samples were degassed with a degassing station for 5 min under vacuum, and benzamidine was added in a 1:4 enzyme-to-inhibitor ratio on a molar basis. Trypsin samples, when analyzed without benzamidine, were highly prone to autolysis. Experiments were performed with a DSC 60200 instrument (TA Instruments, New Castle, DE) using Dscrun software to collect data and Nanoanalyze software (TA) for data processing. All data sets were processed using polynomial background subtraction, and overlay graphs were prepared with MS Excel.

Intact Reversed-Phase LC-MS

Samples of intact trypsin were separated by reversed-phase liquid chromatography (ACQUITY Premier BSM, Waters Corporation) and eluting peaks were detected by UV absorbance and ESI-MS in positive ion mode using a quadrupole time-of-flight mass spectrometer operating with a mass resolution of 20k (Xevo G2-XS QToF, Waters Corporation). Lyophilized trypsin samples reconstituted in 18.2 MΩ water at 0.5 mg/mL were directly analyzed by injecting 1 μL injection onto a 2.1 × 100 mm ACQUITY Premier HSS T3 column kept at 60 °C (Waters Corporation). A gradient-based elution was performed using mobile phase A (0.045% (v/v) TFA in 18.2 MΩ water) and mobile phase B (0.045% (v/v) TFA in acetonitrile) at a flow rate of 0.6 mL/min. For our analysis, dilute TFA in the mobile phase provided the necessary sensitivity; however, DFA could also be applied with additional method development and an added benefit of increased sensitivity.²² A gradient of 0% to 40% B was applied over 11.56 min for elution. The QT of the ionization source was programmed to have a capillary voltage of 3 kV, a sample cone setting of 80 V, a source temperature of 100 °C, a desolvation temperature of 600 °C, and a desolvation gas flow of 800 L/h. Mass spectra were acquired from 100 to 4000 m/z at a 2 Hz scan rate. Chromatography and mass spectrometry data analysis was performed with MassLynx v4.1 (Waters Corporation).

Enzyme Autolysis Measurements

A solution containing 10 mM CaCl₂ and 0.1 M Tris, pH 7.5, was used to resuspend trypsin from different manufacturers to obtain 0.1 mg/mL concentration. About 200 μL of each trypsin sample was transferred to microcentrifuge tubes, incubated for 1 h at 37 °C, and inactivated with 20 μL of 1% formic acid. Each enzyme preparation (200 μL of 0.1 mg/mL) was immediately treated with 20 μL of 1% formic acid for unincubated samples. A large 50 μL injection was made into a ACQUITY Premier Peptide CSH C18 column, 130 Å, 1.7 μm, 2.1 × 150 mm (Waters Corporation) column for analysis of autolytic peptides at both high and low abundance. A higher injection volume was used to allow capture of low abundance peptides. LC mobile phase A consisted of 0.1% DFA in LC-MS grade water, and mobile phase B contained 0.07% DFA in LC-MS grade acetonitrile. A gradient elution at 65 °C contained a hold at 1% B for 5 min for sample loading, 40% B at 65 min, and 70% B at 68 min with a hold for 2 min before switching to 1% B for equilibration. The wavelength of the TUV detector was set at 219 nm, and the ACQUITY RDa detector was set to full scan with a mass range of 50–2000 m/z and positive polarity at a scan rate of 2 Hz. The cone voltage alternated between low energy (20 V) and high energy (60–120 V) to enable the fragmentation of peptide precursors. Data acquisition and processing were performed using UNIFI (version 3.0.0.15) operating on the waters_connect Informatics Platform.

Monoclonal Antibody Digestion and LC-MS Peptide Mapping

Peptide mapping analyses were performed by liquid chromatography and mass spectrometry (LC-MS) on a BioAccord System (Waters Corporation). The instrument configuration includes an ACQUITY UPLC I-Class PLUS coupled to an ACQUITY UPLC TUV optical detector and an ACQUITY RDa time-of-flight (TOF) mass detector. The data were acquired and processed with UNIFI (version 3.0.0.6) operating on the waters_connect Informatics Platform. Briefly, the sample was diluted to 1 mg/mL in denaturation/reduction buffer, with final concentrations of 5 M guanidine hydrochloride, 250 mM Tris, pH 7.5, and 3 mM DTT. It was incubated at room temperature for 30 min to reduce disulfide bonds. Then IAM was added to the sample with a final concentration of 7 mM for alkylation in the dark at room temperature for 20 min. The sample was split, and immediately, the buffer was exchanged in a pH 7.5 digestion buffer using desalting columns, according to the manufacturer’s instructions (Zeba spin, Thermo Fisher Scientific, Waltham, MA). Sample concentrations were checked with a Nanodrop after desalting. A typical sample recovery of 75–95% was observed for these devices. Volumes equivalent to 50 μg of protein were transferred to new microcentrifuge tubes for trypsin digestion. Trypsin was added in the appropriate ratio of 1:5 (weight/weight). (For 50 μg of protein, digestions in the 1:5 ratio receive 10 μg of trypsin.) The 1:5 digestions were incubated at 37 °C for 30 min, and then the trypsin was inactivated with 10% acetic acid (to a final ∼0.16%) and diluted to 0.2 mg/mL with mobile phase A for analysis.

The BioAccord system was fitted with a C18 column (ACQUITY Premier CSH 130 Å, 1.7 μm, 2.1 × 100 mm, Waters Corporation) set at 60 °C, with mobile phases A (0.1% formic acid in water) and B (0.1% formic acid in acetonitrile), flowing at 0.2 mL/min. 1 μg of each digested sample was loaded onto the column, and the peptides were chromatographically separated with a linear gradient of 1–35% B for 50 min. (The method includes an initial hold at 1% B for 1 min before the gradient starts. The gradient is followed by a ramp up to 85% B over 6 min, hold at 85% B for 4 min, return to 1% B over 6 min, and column re-equilibration at 1% B for 13 min.) The TUV detector acquired data at 214 nm. The mass spectrometry data was collected on the RDa detector set to ESI positive mode, full MS scan with fragmentation, in the m/z range 50–2000. The capillary voltage was set at 1.2 kV, the cone voltage at 30 V, with a collisional energy ramp from 60 to 120 V to generate fragment ions. The desolvation temperature was set at 350 °C. Data was collected with the intelligent data capture (IDC) on.

Results and Discussion

Heterogeneity Analysis by Intact RPLC-MS

Much like a biotherapeutic protein is subjected to extensive characterization to ensure patient safety, we thought it could be worthwhile to thoroughly characterize the physicochemical properties of a common enzyme reagent used for its analysis, namely, analytical grade trypsins. One means to that end is to study their heterogeneity by reverse phase chromatography and intact mass analysis. Five unique types of lyophilized trypsins were reconstituted and immediately separated under acidic trifluoroacetic acid ion pairing conditions. Figure 2 (top left) shows the chromatogram obtained for unmodified animal-derived bovine trypsin (Trypsin A).

Figure 2.

Trypsin purity and homogeneity of methylation as observed by intact protein reversed phase LC-MS. Acidified samples of trypsin were injected as 0.5 μg quantities onto a 2.1 × 100 mm 1.8 μm particle C₁₈ RPLC column and eluted across an 11.6 min gradient of 0 to 40% ACN with 0.045% TFA acidified mobile phase, a column temperature of 60 °C, and flow rate of 0.6 mL/min. Eluting peaks were electrospray into a QToF mass spectrometer operating with a 2 Hz scan rate and acquisition window set to 100 to 4000 m/z.

The retention time window corresponding to intact trypsin exhibited two predominant species, and a summed, deconvoluted mass spectrum resulting from this region showed a pronounced signal for both unmodified bovine trypsin and a 16 Da heavier oxidized isoform. The most abundant MS signal could, in fact, be attributed to the oxidized isoform where a molecular weight (MW) of 23,311.9 Da was observed, which is in reasonable agreement with the predicted MW. Table S1 shows different trypsin species from different sources, their molecular formula, and corresponding theoretical molecular weights, while Table S2 shows a list of different trypsin species from bovine and porcine sources, unmodified and chemically modified to reduce autolysis used in this study. Next, a sample of acetylated animal-derived bovine trypsin (Trypsin B) was characterized. Both its chromatogram and the deconvoluted mass spectrum displayed a series of increasingly more retained and increasingly heavier peaks. The assignment of the deconvoluted masses proved that there were 42 Da mass differences between these peaks, which corresponds to the added mass of acetyl groups. The most abundant signal was found to correspond to an oxidized, monoacetylated molecule, although significant amounts of ion signal were also observed for unoxidized and dual-acetylated isoforms. Since bovine trypsin contains 15 primary amines, the average state of modification for Trypsin B is thus at most 20%. An unmodified, recombinant porcine trypsin (Trypsin C) was also analyzed, and slightly different results, given that homogeneous chromatographic and MS data were observed. The predominant molecular weight observed for this trypsin was 23,464.1 Da, a mass that is only 0.9 Da heavier than the average mass predicted for porcine trypsin containing six disulfide bonds.

Two more samples of analytical grade trypsin were analyzed, yet a fourth type of trypsin was particularly interesting to evaluate. This type of trypsin (Trypsin D) has become one of the most widely used forms of trypsin. It is derived from the porcine pancreas and modified by reductive methylation that adds up to 2 methyl groups per residue containing a primary amine. LC analysis of Trypsin D resulted in a chromatogram with the widest range of peak retention times. In addition, the corresponding deconvoluted mass spectrum contained the most complex ion signal. Mass differences corresponding to methylation (+14 Da) were readily identified in a pattern of peaks with masses of 23,674.9 Da and up to and beyond 23,743.6 Da. These molecular weights can be attributed to porcine trypsin modified with 15 and up to 20 methyl groups. An intensity-weighted average extent of modification can be estimated to be about 17 methylations. Porcine trypsin contains 11 primary amines, so these results confirm that Trypsin D is heterogeneously methylated and contains isoforms with a percent modification as low as 68%. The last trypsin (Trypsin E) was manufactured by recombinant expression, purification, and subsequent methylation. The chromatogram of this reagent was the simplest of all trypsins tested, with a singular peak that represented more than 90% of the detected UV signal. Deconvoluted MS data also showed that this trypsin is one of the most homogeneous types. The most abundant MS signal for Trypsin E was deconvoluted to a molecular weight of 23,726.0 Da, which is close to the mass predicted for porcine trypsin modified with 19 methyl groups. The next most abundant ion signal can be attributed to an isoform containing 20 methyl groups. On average, this trypsin is estimated to be close to 90% modified. In this case, it is feasible that each of its individual amine-containing residues is present in a monomethylated, if not dimethylated, form. Although detailed studies on the heterogeneous forms have not been performed in this work, apart from the correlation confirmed between heterogeneous alkylation and susceptibility to autolysis, we reason that the most reliable, predictable, and repeatable performance of an enzyme reagent will come from a highly purified form of a known chemical composition.

Thermal Stability by Differential Scanning Calorimetry

Divalent ions such as Ca²⁺ ions have been reported to offer trypsin stabilizations.^6,20,21 The effect of different concentrations of CaCl₂ on the thermal stability of trypsin was evaluated. The maximum peak temperature, Tm, is the temperature at which 50% of trypsin molecules are unfolded and denatured. The right-hand side region of the peak represents the temperature range beyond which the trypsin is fully unfolding and denatured, the left-hand side of the peak represents the temperature range where trypsin remains active, and the thermogram y-axis is the measure of the energy required for unfolding. Irreversible damage/unfolding of trypsin is observed at temperatures <100 °C regardless of the source of trypsin or the chemical modification (Figures 3 and 4). Trypsin A, an animal-derived bovine trypsin, was tested with different CaCl₂ concentrations corresponding to typical digestion conditions. A maximum temperature change from 40 to 52 °C was observed, and an increase of 84% in energy requirement for 50% unfolding was observed when 100 mM CaCl₂ was added, indicating the stabilization of trypsin by CaCl₂. Although Tm remains unchanged between 10 and 100 mM CaCl₂, there was a 13% increase in energy requirement with the latter. Trypsin becomes less efficient once unfolded, and therefore the optimal performance temperature for trypsin A would be at a temperature <37 °C, before any unfolding begins. A broad peak observed in the thermogram indicates multiple forms of Trypsin A, consistent with the intact mass analysis shown in Figure 2(a), where multiple peaks are manifested with different molecular weights.

Figure 3.

Thermogram for different Typsin showing calcium chloride stabilization (a) Trypsin A with different concentrations of calcium chloride and (b) Trypsin E, with and without 100 mM calcium chloride. A small change in Tm with more energy is required for unfolding in the presence of calcium chloride for both Trypsin A and Trypsin E.

Figure 4.

Thermogram for different trypsins tested without inhibitor (dotted line) and with inhibitor (solid line) in 100 mM Tris buffer pH 7.5 with 100 mM calcium chloride. Comparison of different trypsin sources (a) bovine vs porcine shows stabilization by the inhibitor and different chemical modifications.

Furthermore, trypsin A is not chemically modified; therefore, it is prone to autolysis. The resulting peptides will presumably have lower unfolding temperatures.

Immobilization has been reported to result in the thermal stabilization of trypsin. The Supporting Information provides details on immobilization of Trypsin A on hybrid silica particles BEH as described in Materials and Method. Figure S1 shows the thermal stability of Trypsin A immobilized on solid particles, where unfolding temperature is increased from 60 to 80 °C in the presence of CaCl_2. With immobilized Trypsin A, Tm increased by 60% reaching 83 °C, increasing its usability temperature before unfolding begins with the peak extending to 100 °C, indicating some very stable configurations. Figure 3(b) shows the impact of adding 100 mM CaCl₂ to Trypsin E, a chemically modified trypsin, to reduce autolysis. The Tm for Trypsin E is similar at 73 °C irrespective of the absence or presence of CaCl₂, with a 90% increase in the maximum energy required for 50% unfolding of trypsin in the presence of Ca²⁺, indicative of stabilization. A narrow peak observed with Trypsin E indicates homogeneity in the trypsin population, and presumably, due to the absence of autolysis. As mentioned above, trypsin activity decreases once it starts to unfold, which is indicated by Tm, Trypsin E has a higher Tm than Trypsin A, allowing usability without reduction in activity at temperatures up to 57 °C. Trypsin E may benefit from the adding 100 mM CaCl₂, compared to Trypsin A, as indicated by the 84% higher energy requirement for 50% unfolding after the adding CaCl₂.

Figure 4 shows a thermal stability comparison of several trypsins and the impact of adding an inhibitor in the presence of 100 mM CaCl₂. The peak temperature shifted to higher values for all trypsins analyzed. These shifts were more significance for Trypsins A–C than Trypsins D, E, indicative of additional thermal stabilization. The data obtained in the presence of the inhibitor indicated the absence of autolysis and are in good agreement with the intact mass data shown in Figure 2. Both animal-derived bovine Trypsin A and Trypsin B exhibited two peaks with different temperatures, indicating the presence of two distinct trypsin configurations in equal amounts. This is also supported by the molecular weight data obtained by RPLC-MS.

In contrast, animal-derived porcine Trypsins C–E have one prominent configuration and a small but significant percentage of the less stable configuration, as indicated by the peak profiles in the thermograms. Figure 4(e) shows a narrow Tm peak for Trypsin E, indicating a homogeneous population also supported by the narrow molecular weight distribution shown in Figure 2(e).

Autolysis Resistant Trypsin E

The autolysis tendencies of Trypsins A, B, C, D, and E were evaluated by comparing the UV (219 nm) peak areas of intact trypsin peaks and the increase in the number of tryptic peptides with or without the incubation of trypsin preparation at 37 °C for 1 h in the absence of any target protein. In this condition, where no other protein is available, trypsin can only act upon itself if the enzyme is susceptible to self-hydrolysis. By monitoring the changes in the intact trypsin signal and the appearance of new tryptic peptides, we could understand the autolysis tendency of each trypsin type. Trypsins A, B, and E showed clear intact trypsin peaks with or without incubation at 37 °C for 1 h, while the intact trypsin UV signal was weaker for Trypsins C and D (Figure 5).

Figure 5.

Comparison of the peak areas of intact trypsin from various sources. Resuspended trypsin from each source was incubated at 37 °C for 1 h before inactivation (Orange). Control samples (blue) were inactivated and analyzed by LC-UV to measure the absorbance at 219 nm.

Furthermore, Trypsin E exhibited a signal at least 2.5× higher compared to trypsin per unit mass of enzyme preparation. Its enhanced autolysis resistance is the most likely reason for Trypsin E’s efficient digestion and higher activity. Trypsin A and B are 10× higher than Trypsin C and D from similar sources. Moreover, >90% retention of the intact trypsin UV signal after incubation at 37 °C for 1 h, indicating superior autolysis resistance of Trypsin E. This observation is further supported by the insignificant increase in the number of enzyme origin tryptic peptides (17 vs 19) in the case of Trypsin E after incubation at 37 °C for 1 h (Figure 6) compared to Trypsin A, B, and D.

Figure 6.

Comparison of the autolytic tryptic peptides in the control (blue) and after incubation at 37 °C (orange). For the control, trypsin was rapidly inactivated with formic acid (see experimental methods) soon after thawing.

The tryptic peptides present in the control samples of Trypsin E are presumed to have been generated during the methylation process. Such tryptic peptides are also seen with methylated versions of porcine Trypsin D, where >30 tryptic peptides were observed, and acetylated Trypsin B, where >20 peptides were observed. Although no significant change in the number of tryptic peptides is observed for Trypsins C and D after incubation, the amount of intact trypsin is also much smaller for this source (Figure 5). The lower number of tryptic peptides observed for Trypsin A is possibly due to the difference in autolysis mechanism between the bovine sourced (Trypsins A and B) and porcine source (Trypsins C, D, and E) trypsin. The availability of higher amounts of intact trypsin per unit mass of enzyme preparation and its enhanced autolysis resistance are the most likely reason for efficient digestion and higher activity of Trypsin E.

Peptide Mapping of a Monoclonal Antibody

Peptide mapping data must be as clean and straightforward as possible to ease the burden of biopharmaceutical characterization. Data analysis time and effort are minimized when only the expected tryptic cleavage peptides are generated. In addition, if the peptide mapping data are to be used in any regulatory filings, each peak must be confidently assigned. When extra autolysis or unexpected cleavage peptide peaks are present, the user must manually investigate when a standard software search does not assign them. To test the impact of autolysis-resistant trypsin in use for typical biopharmaceutical characterization, a denatured, reduced, alkylated, and desalted sample of infliximab was digested with a high 1:5 enzyme-to-protein ratio for 30 min at 37 °C with heterogeneously methylated animal-derived porcine trypsin (Trypsin D) or homogeneously methylated porcine trypsin (Trypsin E). The resulting TIC chromatograms (Figure 7) exhibited clear differences in the presence of trypsin autolysis peaks (red arrows) and baseline differences. Of the total integrated TIC, Trypsin D produced a signal that was only 88.5% related to infliximab peptides resulting from tryptic cleavages. The remaining 11.5% of the area is made up of trypsin autolysis, missed cleavages, and other unmatched species.

Figure 7.

Comparison of 1:5 enzyme-to-protein ratio digestion of infliximab using Trypsin D (heterogeneously methylated, animal-derived porcine trypsin) versus Trypsin E (homogeneously methylated, recombinant porcine trypsin). Red arrows mark the most significant trypsin autolysis peaks.

Meanwhile, Trypsin E produced 96.4% of the integrated TIC as infliximab-matched tryptic cleavage peptides. Moreover, it yielded a much cleaner chromatographic baseline with intense intact trypsin observed after digestion, as indicated in Figure 7, in agreement with low autolysis data in Figures 5 and 6.

Conclusions

This study has elucidated several insights about trypsins’ physicochemical properties with and without chemical modification. Additionally, the addition of CaCl₂ was also evaluated for thermal stabilization. The physical data correlates with observations of autolysis resistance and monoclonal antibody peptide mapping analysis. Chemical modification can stabilize Trypsin against thermal denaturation, as indicated by the change in the temperature of Tm < 60 °C for modified Trypsins (B, D, and E). However, as observed with Trypsin E and confirmed with RPLC-MS, homogeneous modification is necessary to eliminate autolysis and to provide cleaner and high sequence coverage digestion needed for robust peptide mapping studies. Such trypsin’s enhanced thermal stability and autolysis resistance also make them a versatile candidate for rapid high-temperature digestion or simple one-pot digestion protocols, thereby reducing digestion time and nonspecific cleavages.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jproteome.4c00598.

• Table S1: Shows a table of different trypsin species from different sources, formula and corresponding theoretical molecular weights; Table S2: List of different trypsin species from bovine and porcine sources, unmodified and chemically modified to reduce autolysis; Figure S1: shows the thermal stability of immobilized bovine sourced trypsin’s, the impact of inhibitor and calcium chloride stabilization; Materials/Methods S1: Description of materials and method used to prepare the immobilized trypsin on Waters BEH particles (PDF)

The authors declare no competing financial interest.