Quenching Trypsin Is Unnecessary in Filter-Based Bottom-Up Proteomics

Abstract

Quenching digestions in proteomics prior to analysis is routine in order to eliminate residual protease activity. Residual activity leads to overdigestion, nonspecific star-activity, and back-exchange in isotopic ¹⁸O quantitation. Chemical and isobaric labeling (e.g., TMT/iTRAQ) of proteins or peptides for mass spectrometry-based proteomics is generally incompatible with ubiquitous postdigestion acidification. This necessitates buffer exchange and pH adjustments. We demonstrate that quenching is unnecessary with peptides generated from protein filter-traps, as trypsin activity and intact trypsin are negligible in the eluate from these preparations. Labeling can be directly performed on enzymatic digests from these methods, improving recovery, throughput, and ease of automation.

Introduction

Mass spectrometry (MS)-based proteomics is a powerful platform for measuring changes in biological systems at great depth.^1,2 Improvements in large-scale, bottom-up quantitative proteomic technologies has allowed for the characterization of expression dynamics for thousands of proteins from complex biological samples.³⁻⁵

Identification of proteins using bottom-up techniques utilizes protein digestion to generate peptides for analysis. Trypsin, a serine protease which cleaves predominantly after K and R residues, is the dominant enzyme for this approach. Quenching the reaction is crucial to eliminate deleterious enzymatic activity including overdigestion, back-exchange in ¹⁸O isotopic labeling, and catalytic formation of pseudotrypsin which nonspecifically cleaves proteins.⁶⁻⁹ Residual activity degrades the experimental results and impacts precision and accuracy in measurements.

Many quantitative proteomics studies utilization of isobaric/isotopic labeling to quantify proteins from tagged or labeled peptides.¹⁰⁻¹³ Isotopic/isobaric labeling chemistries are pH dependent and typically require pH > 7 for high efficiency.^12,14 This process is made cumbersome by postdigestion acidification necessary to terminate protease activity.^2,15−17 Acidification and subsequent desalting and drying after proteolysis are time-consuming and cause reduced recovery detrimental to mass-limited samples or low abundant proteins.¹⁸

Filter-based methods (i.e., S-Traps, FASP, MAP) for protein preparation of bottom-up proteomics samples have become widely popular because they successfully incorporate rapid protein isolation and waste removal.¹⁹⁻²⁵ Proteins trapped by the filter are digested with a protease, and the resulting peptides are eluted from the filter for analysis. There is strong development in improving filter-based methods for direct compatibility with diverse sample compositions.²⁶ We hypothesized that intact trypsin remains bound to the filter and not eluted with the peptides. This obviates the necessity of “quenching” the digest.

Here, we investigate the necessity of postdigestion acidification with samples generated from suspension traps (S-Traps) often used in bottom-up proteomics preparations. In this study, we subject several samples with analytically relevant amounts of trypsin to S-Trap filtration/bottom-up proteomics preparation: trypsin, trypsin with bovine serum albumin (BSA), positive control, and negative control. In order to test if enzymatic activity is detectable in the filtered digestion products, N_α-benzoyl-dl-arginine 4-nitroanilide hydrochloride (BAPNA) was added directly to the eluate. BAPNA is a chromogenic substrate; once hydrolyzed by trypsin, it will release the p-nitroaniline chromophore which can be detected by colorimetric analysis to measure enzymatic activity.²⁷

We also sought to confirm absence of trypsin using matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS). Trypsin was incubated on an S-Trap in the absence of BSA, simulating digestion conditions. Filters were eluted, intact BSA was added, and the sample was incubated (Figure 1A). If any trypsin is eluted, the BSA will form peptides detectable by MALDI-MS. A positive control of BSA incubated with trypsin was also performed, demonstrating typical BSA peptide spectra.

Figure 1.

Trypsin enzymatic activity in the eluent of filter-based preparations. (A) Spectrophotometric preparation of filtered and unfiltered samples prior to analysis in triplicate via kinetic plate reader assay. (B) Cleavage of BAPNA measured by at A_405nm was recorded for 90 min; error bars ±1 σ. (C) MALDI-MS samples were incubated with trypsin (BSA and Test) and eluted, and intact BSA was added to the eluate for an additional 90 min. Tryptic peptides (inset is K.YLYEIAR.R 927.493 m/z) of BSA in positive control; no residual digestion is observed of BSA in test. Peak at 1425.886 m/z is likely Tween-20.

Results/Discussion

After 90 min of analysis, there is no noticeable increase in activity at 405 nm outside of the positive control (Figure 1B). The two trypsin-containing samples increase absorbance readings over the 2 h period, but activity never returns to that of the positive control. From these data, acidification is not necessary with peptides generated from protein filter traps, and isobaric/chemical labeling can be carried out immediately after protein digestion.^11,13,25 From our MALDI-MS data, we see no evidence of protein digestion in our test sample, in which trypsin was eluted from filter and BSA was added (Figure 1C). This contrasts with our positive control sample, in which BSA was incubated with trypsin before elution. The positive control generated numerous tryptic BSA digestion fragments including K.YLYEIAR.R at 927.493 [M + H]⁺m/z (inset). Our test sample includes one significant peak at 1425.886 m/zM_r = 1425.8539 m/z. This is likely polysorbate-20 and does not match albumin, keratins, or known trypsin autolysis fragments.

Conclusion

Mass spectrometry-based proteomics has diverse sample preparation, chemical-labeling, and processing procedures which require widely divergent conditions. Unnecessary volumetric steps add complexity and reduce sensitivity, in particular, quenching serial reactions. We demonstrate that for trypsin digestion using filter-based proteomics preparation quenching steps (typically acidification) are unnecessary.

Methods

For spectrophotometry, BSA, trypsin, and BAPNA were prepared to a 2 mg/mL concentration. BSA and trypsin were dissolved in 50 mM ABC; BAPNA was dissolved in dimethyl sulfoxide (DMSO).

Four samples were prepared (Figure 1A): 50 mM ABC (negative control), 1 μg trypsin in 50 mM ABC (trypsin), 1 μg of trypsin and 50 μg of BSA in 50 mM ABC (trypsin + BSA), and 1 μg of trypsin in 50 mM ABC (positive control). All samples were diluted to 150 μL using 50 mM ABC, and samples 1–3 were individually loaded on S-Trap Mini spin columns, with collection microcentrifuge tubes below. Positive control did not undergo filtration. Columns and collection tubes were spun at 2000 rcf for 60 s. 100 μL of S-Trap Binding Buffer was spun through filter samples. 100 μL of 50 mM ABC was used to wash (negative control). ABC was added to balance volume as needed.

25 μg of 2 mg/mL BAPNA was added to all four samples, which were then transferred to a 96-well plate and incubated at 37 °C for 15 min. Three replicates of each treatment were performed. Absorbance at 405 nm was recorded over a 2-h time course, reading every 2.5 min. For MS, 1.4 μg of trypsin in 100 μL of 100 mM TEAB was incubated in an S-Trap Mini for 2 h at 37 °C with 100 μg reduced and alkylated BSA in the positive control^12,21 eluted in 2 × 60 μL of TEAB. 100 μg of BSA in 20 μL of TEAB was added to the test eluate to detect residual trypsin activity, incubated for 90 min, and quenched with 10 μL of 10% TFA. 1 μL of sample was spotted on a steel target for MALDI-MS overlaid with 1 μL of CHCA, and 8000 summed spectra were acquired in reflectron mode on an Ultraflextreme (Bruker) at 500 Hz from 100 to 3000 m/z.²²

Reagents

LC-MS water, methanol, and acetonitrile were from J.T. Baker (Radnor, PA). Ammonium bicarbonate (ABC) was purchased from Acros Organics (Geel, Belgium). Trypsin Promega (Madison, WI), BSA, BAPNA, α-cyano-4-hydroxycinnamic acid (CHCA), and triethylammonium bicarbonate (TEAB) were from Sigma-Aldrich (St. Louis, MO). Trifluoroacetic acid, DMSO, and 96-well plates were from Thermo Scientific (Rockford, IL). S-Trap Minis were from ProtiFi (Farmingdale, NY). Plate reader was a Biotek Epoch Agilent Technologies (Santa Clara, CA). S-Trap Binding Buffer was 100 mM TEAB in 90% methanol. (CHCA) was saturated in 50/50 acetonitrile/water.

Author Contributions

^# C.B.M. and D.D.H. contributed equally to this work. C.B.M.: Conceptualization, formal analysis, investigation, methodology, writing–original draft, writing–review and editing, visualization; D.D.H.: Conceptualization, writing, review, editing, and visualization. M.M.C.: Conceptualization, funding acquisition, writing–review and editing. Authors declare no conflicts of interest.

The authors declare no competing financial interest.