
==== Front
J Proteome Res
J Proteome Res
pr
jprobs
Journal of Proteome Research
1535-3893
1535-3907
American Chemical Society

37747260
10.1021/acs.jproteome.3c00505
Letter
EPURISp: Combining Enzymatic Digestion, Ultrafiltration, and Rapid In Situ Sample Purification for High-performance Proteomics
https://orcid.org/0000-0002-1830-506X
Lu Ping †¶
Shan Mengyuan †‡§¶
Ji Xuemeng ‡
Deng Fuqi †§∥
Wang Yan *†‡§∥⊥
† Tianjin Eye Hospital, Tianjin Eye Institute, Tianjin Key Laboratory of Ophthalmology and Visual Science, Tianjin 300020, China
‡ School of Medicine, Nankai University, Tianjin 300071, China
§ Nankai University Affiliated Eye Hospital, Nankai University, Tianjin 300020, China
∥ Clinical College of Ophthalmology, Tianjin Medical University, Tianjin 300070, China
⊥ Nankai University Eye Institute, Nankai University, Tianjin 300020, China
* Email: wangyan7143@vip.sina.com.
25 09 2023
06 10 2023
25 09 2024
22 10 33923400
11 08 2023
© 2023 The Authors. Published by American Chemical Society
2023
The Authors
https://creativecommons.org/licenses/by-nc-nd/4.0/ Permits non-commercial access and re-use, provided that author attribution and integrity are maintained; but does not permit creation of adaptations or other derivative works (https://creativecommons.org/licenses/by-nc-nd/4.0/).

High-performance liquid tandem mass spectrometry (HPLC-MS) is widely employed for protein analysis in biological systems. However, conventional proteomic sample pretreatment methods suffer from multiple steps and poor reproducibility. In this study, we introduce EPURISp (Enzymatic Digestion with Ultrafiltration and Rapid In-situ Sample Purification), a novel proteomic pretreatment technique that combines enzymatic digestion, ultrafiltration, and one-step temperature-controlled vacuum drying for efficient desalting. The EPURISp method exhibits excellent protein recovery rates across a wide range of molecular weights and hydrophilicity, surpassing traditional C18 desalting approaches. Practical proteomic analysis (PXD044209) utilizing EPURISp demonstrates the highest protein identification yield with remarkable reproducibility, which is particularly advantageous in membrane protein identification. Notably, EPURISp exhibits superior performance in minimizing oxidation and deamidation modifications compared with conventional FASP methods. This innovative EPURISp method represents a significant advancement in proteomics analysis, providing reliable and efficient results for mass spectrometry.

National Natural Science Foundation of China 10.13039/501100001809 82271118 Tianjin Eye Hospital NA YKPY2202 Tianjin Key Medical Discipline Construction Project NA TJYXZDXK-016A Tianjin Health Research Project NA TJWJ2023MS035 Tianjin Health and Technology Project NA TJWJ2022XK036 Nankai University Eye Institute NA NKYKK202214 Tianjin Diversified Investment Fund for Applied Basic Research NA 21JCZDJC01190 National Key Research and Development Program of China 10.13039/501100012166 2022YFC2404502 document-id-old-9pr3c00505
document-id-new-14pr3c00505
ccc-price
==== Body
pmcIntroduction

High-performance liquid tandem mass spectrometry (HPLC-MS) is widely used as an analytical tool for the qualitative and quantitative analysis of whole proteins in biological systems. For proteomic analysis, specific proteases are required to cleave proteins into peptides that are suitable for molecular weight detection. Currently, the common sample preparation method in proteomics involves using recombinant trypsin to cleave the proteins into peptides, and ammonium bicarbonate is one of the most frequently used salts in mass spectrometry preparation. A 50 mM aqueous ammonium bicarbonate solution can provide the ideal pH of 7.5–8.5 required for trypsin to work effectively.

However, the low tolerance of electrospray mass spectrometry for salts requires the removal of ammonium bicarbonate to avoid any negative impact on subsequent mass spectrometric analysis. Small molecule salts can cause significant ion suppression and reduced sensitivity of analytes due to their strong ionization effect in the electrospray system.1 Excessive salts can also generate a series of ion addition peaks, affecting the identification and grouping of peaks in MS spectra.2−4 Additionally, they can potentially corrode and contaminate the mass spectrometry system’s hardware, leading to damage and the need for cleaning.5 Therefore, it is recommended to include a desalting step in the pretreatment process of proteomics samples to avoid these potential issues and ensure reliable results.

The current popular method for desalination is through the use of a reversed-phase C18 desalination resin column.6,7 The desalting process is divided into three stages: binding, washing, and elution. During the binding stage, hydrophobic interactions between the peptide and the reversed-phase column in the highly aqueous mobile phase cause the peptide to be bound to the column. Salt buffer is then removed through repeated washing. Finally, the hydrophobic interaction between the peptide and C18 is disrupted by an organic solvent to elute the peptide. However, this desalting column method has limitations. The method may result in sample loss due to the varying adsorption abilities of different peptides to C18. Peptides with strong hydrophilicity, such as phosphorylated peptides, may not bind well to C18 resin, leading to loss during the binding stage, while peptides with strong hydrophobicity may bind tightly to the resin but be difficult to elute, leading to loss in the elution stage.8−10 This selective preference of samples during the desalting process can compromise the objectivity of the mass spectrometry results. Additionally, the organic solvent used in the elution process can dissolve the polymer in the plastic product, leading to polymer contamination of the peptide sample and a shift in the mass spectrometry peak.1,11 Furthermore, peptides dissolved in the elution reagents cannot be directly used as mass spectrometry samples and require vacuum centrifugation to remove the organic solvent, followed by resolubilization with 0.1% formic acid. In conclusion, the commonly used desalting column preparation method in proteomics analysis is associated with poor reproducibility and low protein species detection, making it imperative to seek alternative methods for desalting.

We present a novel desalting method offering an alternative to the commonly used desalting column approach. The new method involves one-step temperature-controlled vacuum drying and deliquescing the ammonium bicarbonate after enzymatic digestion, resulting in desalinated peptide samples ready for direct analysis using mass spectrometry. Prior to analysis, the samples are simply dissolved in 0.1% formic acid, eliminating the need for multiple steps and reducing the risk of sample loss associated with the desalting column approach. This innovative desalting method overcomes the limitations of the conventional desalting column approach, such as poor reproducibility and low detection of protein species. It represents a promising advancement in proteomics analysis, offering improved accuracy and increased sensitivity.

Experimental Section

Workflow of EPURISp for Protein Digestion and Desalination

This study introduces a novel proteomic pretreatment method, EPURISp (Enzymatic Digestion with Ultrafiltration and Rapid In-situ Sample Purification), which integrates enzymatic digestion, ultrafiltration, and desalting through a one-step temperature-controlled vacuum drying process. The method’s procedure is outlined as shown (Figure 1). To begin, the protein sample (100 μg) is reduced and alkylated by diluting it to 400 μL using 6 M guanidine hydrochloride (Sigma-Aldrich, USA). Then, 10 μL of 500 mM TCEP (Tris(2-carboxyethyl)phosphine) and 8 μL of 1 M iodoacetamide (Thermo Fisher Scientific, USA) are added, and the mixture is vigorously vortexed. The reaction is incubated for 30 min at room temperature, shielded from light. Next, the sample is transferred to a 10 kDa ultrafiltration tube (Amicon Ultra, Millipore, USA) and centrifuged at 8000g for 20 min at room temperature to trap proteins. Subsequently, the protein sample is washed by adding 400 μL of 50 mM ammonium bicarbonate (Merck, Germany) and subjected to centrifugation five times to remove the buffer and ensure clean protein isolation. Afterward, the protein sample is dissolved in 200 μL of 50 mM ammonium bicarbonate in the ultrafiltration tube, and 2 μg of recombinant trypsin (Promega, USA) is added for digestion. The mixture is incubated overnight at 40 °C. Following digestion, the mixture is centrifuged at 8000g for 5 min, and the filtrate is collected. The filtrate is then dehydrated and deliquesced by drying it using a temperature-controlled vacuum centrifuge at 40 °C and 0.1 MPa for 30 min, effectively removing the ammonium bicarbonate. For analysis, 50 μL of a 0.1% formic acid solution (Sigma-Aldrich, USA) is added to the dried sample prior to loading it into a clean vial for LC-MS analysis.

Figure 1 Workflow overview of EPURISp. This workflow comprises four main steps: (1) Sample preparation: mix 100 μg of the protein sample with 10 μL of 500 mM TCEP and 8 μL of 1 M iodoacetamide in 400 μL of 6 M guanidine hydrochloride. Incubate for 30 min at room temperature. (2) Ultrafiltration and washing: transfer the sample to a 10 kDa ultrafiltration tube, centrifuge at 8000g for 20 min, and wash with 400 μL of 50 mM ammonium bicarbonate five times to isolate clean proteins. (3) Enzymatic digestion: dissolve the protein sample in 200 μL of 50 mM ammonium bicarbonate, add 2 μg of recombinant trypsin, and incubate overnight at 40 °C. After digestion, collect the filtrate following centrifugation. (4) Dehydration and analysis prep: dehydrate the filtrate using a vacuum centrifuge at 40 °C and 0.1 MPa for 30 min to remove ammonium bicarbonate. For analysis, 50 μL of 0.1% formic acid was added before loading into a clean vial for LC-MS analysis.

Protein Recovery Assay

The proteins used in this study, cytochrome C (approximately 12.4 kDa) (Sigma-Aldrich, USA), green fluorescent protein (GFP) (approximately 27 kDa) (Invitrogen, USA), antithrombin III (approximately 58 kDa) (Merck Millipore, Germany), transferrin (approximately 80 kDa) (Roche, Switzerland), immunoglobulin G (IgG) (approximately 150 kDa) (Abcam, UK), and fibronectin (typically above 200 kDa) (BD Biosciences, USA), were all obtained from commercial sources. The reducing agent-compatible BCA protein assay kit was purchased from Thermo Fisher Scientific (USA). Three different methods were used for protein recovery: In-solution, FASP (Expedeon, UK), and EPURISp. After recovery, the proteins were quantified using a reducing agent-compatible BCA protein assay kit (Thermo Fisher Scientific, USA) according to the manufacturer’s protocol. The absorbance was read on a spectrophotometer (Agilent Technologies, USA) at a wavelength of 562 nm.

Peptide Quantification before and after the Desalination Process

Two different methods were employed for peptide desalination: C18 column desalting and EPURISp deliquescence desalination. For C18 column desalting, the peptide samples were loaded onto C18 columns (Thermo Fisher Scientific, USA). The columns were then washed to remove salts, and peptides were eluted by using a gradient of acetonitrile (Sigma-Aldrich, USA) in water (Milli-Q, Merck, Germany). For the EPURISp deliquescence desalination method, the peptide samples were processed through our proprietary EPURISp desalination protocol, which involves dissolving the peptide sample in a high salt concentration solution (Sigma-Aldrich, USA) followed by selective precipitation of salts. The peptides used in this study are leu-enkephalin (555.68 Da) (Sigma-Aldrich, USA), vasopressin (1084.23 Da) (Bachem, Switzerland), somatostatin (approximately 1637.89 Da) (Bachem, Switzerland), glucagon (approximately 3483.87 Da) (Sigma-Aldrich, USA), and LL-37 (Cat. No. LL37-05MG) (Novus Biologicals, USA), and TAT cell-penetrating peptide (CAS No: 697226-52-1) (AnaSpec, USA) were purchased from commercial sources. The C18 columns for desalting were acquired from the Waters Corporation (USA). Before and after the desalination process, the peptide concentrations were measured using high-performance liquid chromatography (HPLC) (Agilent Technologies, USA). A reversed-phase column (Waters Corporation, USA) was used, and the eluent was monitored for absorbance at 220 nm.

Protein Extraction and Digestion

Mouse intestinal tissues were washed twice with PBS buffer to remove any residual contents and then mixed with a protease inhibitor cocktail (Beyond, China) and PBS. The cells were lysed by using ultrasonic waves (200W) in an ice bath until the suspension became clear. The resulting lysate was purified through a microporous filter with a pore size of 0.22 μm (Millipore, USA). The protein concentration was determined using a BCA assay kit (Thermo Fisher Scientific, USA). 50 μg portion of protein was mixed with 1 μL of 200 mM tris(2-carboxyethyl)phosphine (Aladdin, China) and 1 μL of 500 mM iodoacetamide (Aladdin, China), and the total volume was brought to 100 μL with 6 M guanidine hydrochloride (Aladdin, China). The mixture was then incubated in the dark at room temperature for 40 min. The protein was subsequently digested overnight at 37 °C using 2 μg of recombinant trypsin (from Promega, USA) in a 15 kDa ultrafiltration tube (Millipore, USA). The resulting filtrate was collected by centrifugation and desalted by using a desalting column (Millipore, USA). The desalted sample was then analyzed by mass spectrometry after the addition of formic acid.

Measuring Ammonium Concentration

The Nessler reagent spectrophotometric method for measuring ammonium concentration involves the following steps: Sample preparation by diluting the sample with deionized water, addition of a certain amount of Nessler reagent (Sigma-Aldrich, USA) to the sample, closure of the bottle and shaking, allowing the mixture to stand for 5–10 min, measurement of the absorbance at a wavelength of 630 nm in the ultraviolet–visible spectral range using a spectrophotometer (Thermo Fisher Scientific, USA), preparation of a standard curve by treating a series of ammonium standard solutions with the same method, measuring the absorbance values, and creating a standard curve, and finally, calculation of the sample concentration by using the absorbance value of the sample to look up the corresponding ammonium ion concentration on the standard curve.

LC-MS/MS Analysis

The digested peptide samples were analyzed using a Q Exactive Plus mass spectrometer (e.g., Thermo Fisher Scientific, USA) coupled with an EASY nanoliquid chromatography system (EASY nLC 1200, Thermo Scientific, USA) equipped with an EASY nanoelectrospray interface. The chromatographic setup included a Pepmap nanotrap column (C18, 5 μm, 100 Å, 100 μm × 2 cm, Thermo Fisher Scientific, USA) and a EASY-Spray column (Pepmap RSLC, C18, 2 μm, 100 Å, 50 μm × 15 cm, Thermo Fisher Scientific, USA). Solvent A (0.1% formic acid) and solvent B (80% CH3CN/0.1% formic acid) were used in the chromatographic gradients as follows: 0–8% B for 3 min, 8–28% B for 42 min, 28–38% B for 5 min, and 38–100% B for 10 min.

The mass spectra were processed using Maxquant software (ver. 2.0.3.1) and searched against the UniProt database with specific parameters and instructions. For the modifications, carbamidomethylation of cysteine was set as a fixed modification, while oxidation (M) was considered as a variable modification. The search allowed for trypsin cleavage specificity with up to two miscleavage events. The precursor mass tolerance was set to 10 ppm, and the fragment mass tolerance was set to 0.02 Da.

To ensure reliable identifications, both protein and peptide identifications were required to have a maximum false discovery rate (FDR) of 1.0%. Protein identification was based on at least one unique peptide identification, and protein quantification was performed by calculating the median number of unique peptides for each protein.

Statistical Analysis

The experiments were conducted independently and repeated three times to ensure accuracy and reliability. The gene ontology (GO) analysis of the proteomics data was obtained from the online database DAVID (https://david.ncifcrf.gov). Statistical analysis was performed by using GraphPad Prism software (version 8.3). The significance of differences was assessed using the two-tailed unpaired Student’s t test, with a p-value less than 0.05 considered statistically significant.

Results and Discussion

EPURISp for Protein Digestion and Rapid In Situ Desalination

Based on the characteristic of ammonium bicarbonate decomposing in moist environments, we developed a proteinomics sample pretreatment strategy for the gradual removal of contaminants. Prior to enzymatic digestion, we eliminated reducing agents, alkylating reagents, and denaturants using ultrafiltration tubes. Following digestion, we employed a temperature-controlled vacuum centrifuge to simultaneously remove ammonium bicarbonate during the drying process. For the removal of detergents, reducing agents, and alkylating reagents used in cell and tissue lysis, we adopted a sequential washing approach (Figure 1). High-performance liquid chromatography (HPLC) analysis demonstrated that both TEPC and IAA concentrations decreased by at least 100,000-fold and fell below the detection threshold after four washing steps, indicating that five sequential washes were sufficient to reduce contaminants to a noninterfering level (Figure 2A). To optimize the conditions for ammonium bicarbonate deliquescence in the vacuum centrifuge (Figure 2B), we observed that higher chamber temperatures resulted in shorter drying times. For instance, a 50 mM ammonium bicarbonate aqueous solution (200 μL) containing 0.79 mg of ammonium bicarbonate could be completely dried within 20 min at 50 °C, while it took 1 h at 30 °C. However, after drying at both 30 and 50 °C, white ammonium bicarbonate salt residue was observed on the tube walls. This might be attributed to the rapid drying at 50 °C, which did not allow sufficient reaction time for deliquescence, whereas ammonium bicarbonate did not decompose rapidly at 30 °C. We discovered that vacuum centrifugation at 40 °C effectively removed ammonium bicarbonate from a 200 μL 50 mM solution prior to drying. Furthermore, we validated the method using real peptide samples. After the enzymatic reaction, a 200 μL filtrate was collected from the bottom of the ultrafiltration tube through centrifugation. The pH of the filtrate was weakly alkaline (pH 7.5–8.0). Subsequently, the peptide sample was dried using vacuum centrifugation at 40 °C and 0.1 MPa for 40 min. The peptide was redissolved in 200 μL of ultrapure water, reaching a pH of 7.0, indicating the removal of alkaline components. To directly confirm the removal of ammonium bicarbonate, the spectrophotometric Nessler reagent method for ammonium ions was employed, revealing a reduction in ammonium ion concentration from 50 mM to less than 10 nM. This confirmed the effectiveness of ammonium bicarbonate removal through thermal deliquescence (Figure 2C).

Figure 2 Validation and optimization of the ammonium bicarbonate removal process. (A) High-performance liquid chromatography (HPLC) analysis: TEPC and IAA concentration after five sequential washes. (B) Evaluation of ammonium bicarbonate deliquescence conditions at 0.1 MPa. (C) Ammonium ion concentration after thermal deliquescence. A 200 μL filtrate, post enzymatic reaction, was vacuum-dried at 40 °C and 0.1 MPa. After redissolving in ultrapure water and pH adjustment to 7.0, the ammonium ion concentration decreased from 50 mM to less than 10 nM, confirming effective ammonium bicarbonate removal. Dotted lines denote the 10 nM detection threshold. Data are from three independent replicates (n = 3).

While EPURISp and FASP share similarities in employing filter-based protein truncation methods, the key differences between them become apparent. EPURISp conducts more thorough washing before enzymatic digestion, requiring 5 washes as opposed to FASP typical 2 washes using ammonium bicarbonate solution.12 Additionally, EPURISp seamlessly integrates in situ sample desalting and purification methods, achieved through lightly heated vacuum centrifugation. In contrast, FASP typically necessitates an additional desalting column purification step. Unlike EPURISp and FASP, the suspension trapping (S-Trap) employs a three-dimensional porous material to trap denatured protein solutions.13,14 Due to its large (submicron) pore size, each centrifugation cycle with the S-Trap filter only takes 1 min, achieving higher washing efficiency than EPURISp and FASP.13,14 However, it is worth noting that ultrafiltration tubes currently have a lower cost compared to S-Trap, and S-Trap requires multiple washes with organic solvents such as formic acid and acetonitrile after enzymatic digestion.15 S-Trap also involves an additional desalting column purification step, which adds to experimental complexity. In summary, EPURISp introduces a distinct approach and a significant enhancement for efficient sample preparation.

Performances of the EPURISp in Proteins’ Recovery of Various Molecular Weight before Protein Digestion

The molecular weight distribution of proteins varies greatly, ranging from a few thousand to several hundred thousand Daltons. To investigate any preference of our EPURISp method for protein recovery prior to enzymatic digestion, we conducted a comparative analysis of protein recovery rates using the In-solution, FASP, and EPURISp methods during the predigestion steps. Protein quantification was performed by using a reducing agent-compatible BCA protein assay kit. The proteins tested in this study included cytochrome C (approximately 12.4 kDa), green fluorescent protein (GFP) (approximately 27 kDa), antithrombin III (approximately 58 kDa), transferrin (approximately 80 kDa), immunoglobulin G (IgG) (approximately 150 kDa), and fibronectin (with a wide range of molecular weights, typically above 200 kDa).

The results clearly demonstrated that all three methods, In-solution, FASP, and EPURISp, achieved protein recovery rates of over 80% for proteins with different molecular weights (Figure 3). Overall, there was no significant difference in protein recovery rates between the FASP and EPURISp methods across the ranges of molecular weights tested. However, it is worth noting that the In-solution method exhibited slightly better recovery rates for low molecular weight proteins, which may be attributed to the use of a 10 kDa molecular weight cutoff filter.

Figure 3 Protein recovery rates of various molecular weights using In-solution, FASP, and EPURISp methods. The protein recovery rates were evaluated for six proteins with diverse molecular weights: cytochrome C (approximately 12.4 kDa), green fluorescent protein (GFP) (approximately 27 kDa), antithrombin III (approximately 58 kDa), transferrin (approximately 80 kDa), immunoglobulin G (IgG) (approximately 150 kDa), and fibronectin (with a broad molecular weight range, typically above 200 kDa). The experiments were performed in triplicate, and the results are presented using violin plots.

Specifically, for the low molecular weight protein Cytochrome C, the average recovery rates for FASP and EPURISp were 85% and 83%, respectively, while the In-solution method achieved a recovery rate of 96%. Nevertheless, as the molecular weight of the proteins increased, the recovery rates of both FASP and EPURISp improved significantly. For instance, for Green Fluorescent Protein (GFP) with a molecular weight of approximately 27 kDa, both FASP and EPURISp achieved recovery rates of around 93%, while the In-solution method achieved a recovery rate of 97% for the GFP protein. Moreover, for larger proteins with molecular weights exceeding 50 kDa, such as Antithrombin III, the recovery rates using FASP and EPURISp were not significantly different from those using the In-solution method, reaching approximately 98%.

Performance of EPURISp in Peptide Recovery of Various Sizes and Hydrophilicity during Desalination

In order to investigate the potential preferences of our EPURISp method for the in-solution desalination of peptides with varying molecular weights following enzymatic digestion, we conducted a comparative analysis of the recovery rates between deliquescence desalination and C18 desalting. Peptide concentrations were measured using HPLC both before and after the desalination process. The peptides subjected to testing included leu-enkephalin (molecular weight: 555.68 Da), consisting of five amino acids, vasopressin (molecular weight: 1084.23 Da), consisting of nine amino acids, somatostatin (molecular weight: approximately 1637.89 Da), consisting of 14 amino acids, and glucagon (molecular weight: approximately 3483.87 Da), consisting of 29 amino acid residues (Figure 4A). The results showed that the recovery rates achieved through C18 column desalting ranged from 75% to 90%, with no discernible correlation to the molecular weight of the peptides. It is plausible that the loss of samples during C18 column desalting could be attributed to factors such as column binding properties and elution efficiency. In contrast, the EPURISp deliquescence desalination method consistently achieved recovery rates above 96% for all peptides tested, irrespective of their molecular weights. This observation indicates that the EPURISp deliquescence desalination method is highly effective in recovering peptides of varying molecular weights.

Figure 4 Peptide recovery comparison. (A) Peptide recovery rates with different molecular weights using two desalination methods: C18 column desalting and EPURISp deliquescence desalination. Peptide concentrations were assessed via HPLC pre- and post desalination. Tested peptides: leu-enkephalin (555.68 Da), vasopressin (1084.23 Da), somatostatin (approximately 1637.89 Da), and glucagon (approximately 3483.87 Da). Triplicate experiments; results in violin plots. (B) Comparative recovery analysis of highly hydrophilic and hydrophobic peptides with C18 column desalting and EPURISp deliquescence desalination. Representative peptides: LL-37, a hydrophobic antimicrobial peptide, and TAT cell-penetrating peptides are rich in hydrophilic amino acids. The experiment was performed in triplicate, and the results are presented as violin plots.

Next, we conducted a comparative analysis of the recovery rates for highly hydrophilic and hydrophobic peptides using C18 column desalting and EPURISp deliquescence desalination. To represent the hydrophobic and hydrophilic peptides, we selected LL-37 (Cat. No. LL37-05MG), a naturally occurring antimicrobial peptide with broad-spectrum antibacterial activity, and TAT cell-penetrating peptide (CAS No: 697226-52-1), which contains a high proportion of hydrophilic amino acids and exhibits strong hydrophilicity. Our results demonstrated that C18 column desalting exhibited relatively poor recovery rates for both LL-37 and TAT, with average recovery rates of 45% and 39% respectively. In contrast, the EPURISp in-solution desalination method achieved recovery rates above 95% for both peptides, regardless of their hydrophilicity (Figure 4B). These findings suggest that the recovery rate in C18 column desalting is influenced by the hydrophilicity of the peptides. This may be attributed to the reliance of C18 column desalting on hydrophobic interactions to bind peptides to the C18 column resin and subsequent elution using hydrophilic solvents, which can result in preferential retention of peptides with different hydrophilic properties. On the other hand, in-solution desalination offers the advantage of in situ operation, which inherently prevents the loss of peptides during the desalting process. Our findings demonstrate that the EPURISp deliquescence desalination method does not lead to peptide loss and consistently achieves higher recovery rates compared to C18 column desalting.

Performance of EPURISp in Practical Proteomic Analysis

To begin with, mouse intestinal tissues were lysed using a 6 M urea solution. Subsequently, enzymatic digestion was performed by using three methods: In-solution, FASP, and EPURISp. Following digestion, the proteins were separated using high-performance liquid chromatography and analyzed using a QE mass spectrometer. Database searching and analysis were conducted by using Maxquant. In each method, three technical runs were performed and proteins detected with at least two or more peptides were used for subsequent analysis. Each method yielded different protein identification results, as shown in Figure 5A. Overall, the EPURISp method identified the highest number of proteins (2975 ± 52), followed by FASP (2964 ± 102), and finally, the In-solution method (2803 ± 57). The number of identified peptide segments for each method is illustrated in Figure 5B. Using the In-solution method, 10931 ± 16 unique peptide segments were identified. FASP and EPURISp identified 10981 ± 48 and 10959 ± 23 unique peptide segments, respectively. On average, each protein was identified with 3.90 peptides in the In-solution method, 3.70 peptides in the FASP method, and 3.68 peptides in the EPURISp method. This indicates that the EPURISp method exhibited the highest protein identification efficiency. Subsequently, the reproducibility of label-free quantitative analysis of the detected proteins was evaluated using ICC (Figure 5C). It was found that EPURISp showed the best reproducibility with an average ICC value of 0.622, and 1375 proteins had an ICC value greater than 0.4. FASP demonstrated slightly weaker reproducibility, with an average ICC value of 0.533, and 1180 proteins had an ICC value greater than 0.4. In contrast, the In-solution method exhibited the poorest reproducibility, with an average ICC value of only 0.477, and 1017 proteins had ICC values greater than 0.4.

Figure 5 EPURISp in practical proteomic analysis. (A) Comparison of protein identification with the three digestion methods: In-solution, FASP, and EPURISp. Mouse intestinal tissue lysates in 6 M urea underwent HPLC separation and QE mass spectrometry. Maxquant analysis considered proteins with at least two peptides. Each method had triplicate biological replicates. (B) Unique peptide segments identified for each digestion method. Triplicate experiments were conducted for unique peptide segments. (C) Reproducibility evaluation of label-free quantitative analysis with the intraclass correlation coefficient (ICC).

The EPURISp exhibits exceptional advantages in the context of label-free proteomics processing. For TMT-labeled proteomics, which shares the same experimental objectives in the steps preceding TMT labeling, EPURISp could also be employed for enzymatic digestion and desalting. However, TMT labeling necessitates the use of tetramethylammonium bicarbonate (TEAB) as the dissolution buffer for the labeling experiment, with the termination of the labeling reaction using 5% hydroxylamine. The thermal deliquescence of TEAB and hydroxylamine remains requiring further investigation. Nevertheless, the salt introduced by TMT labeling can be conservatively removed by using the traditional desalting column method.

EPURISp Demonstrates Advantages in Identifying Membrane Proteins

By utilizing Venn diagrams (Figure 6A), we compared the influence of each sample preparation method on the identification of the tissue proteome components. For each method, proteins detected in at least two out of three technical runs were selected for subsequent analysis. Remarkably, we discovered that 2428 proteins were consistently identified across all three methods, with additional specific proteins found only in individual methods. Notably, the In-solution method yielded 107 unique proteins, whereas both FASP and EPURISp exhibited a comparable number of specific proteins, with 198 and 188, respectively. Analyzing membrane proteins presents inherent challenges due to their hydrophobic nature, lipid bilayer integrity, and presence of multiple transmembrane domains (TMDs). To assess the strengths and weaknesses of the three methods in identifying membrane proteins, we conducted GO analysis to compare the proportion of specific proteins identified by each method in the GOTERM_CC_DIRECT category (Figure 6B). The GO enrichment analysis of uniquely identified proteins revealed that the In-solution method did not exhibit any enrichment of specific proteins in membrane-related categories. Conversely, FASP identified 12 significantly enriched proteins in GO:0000139 (Golgi membrane), 18 in GO:0005743 (mitochondrial inner membrane), and 27 in GO:0005789 (endoplasmic reticulum membrane). EPURISp identified 12 significantly enriched proteins in GO:0005743 (mitochondrial inner membrane), 7 in GO:0005765 (lysosomal membrane), and 5 in GO:0031902 (late endosome membrane). These findings indicate that both EPURISp and FASP offer advantages over the In-solution method in identifying membrane proteins. EPURISp and FASP demonstrate proficiency in analyzing the mitochondrial inner membrane, while FASP excels in analyzing Golgi and endoplasmic reticulum membranes. Moreover, EPURISp exhibits higher analytical capabilities specifically for lysosomal and late endosome membranes.

Figure 6 Comparison of sample preparation methods for identifying membrane proteins. (A) Venn diagrams illustrate the influence of each sample preparation method on the identification of tissue proteome components. Proteins detected in at least two out of three technical runs were selected for subsequent analysis. The experiment was independently performed in triplicate as biological replicates. (B) GO analysis to evaluate the efficacy of the three methods in identifying membrane proteins. A gene ontology (GO) analysis was conducted on the specific proteins identified by each approach in the GOTERM_CC_DIRECT category.

EPURISp Aids in Reducing Unfavorable Modifications of Amino Acid Residues

Oxidation and deamidation modifications are commonly observed in natural protein and peptide samples, and studying these modifications is essential to understanding the inherent protein stresses. However, modifications that occur during sample preprocessing can pose a challenge, as they mask the intrinsic protein modifications in the sample. To evaluate the impact of these unfavorable modifications introduced during sample processing, we compared three methods: In-solution, FASP, and EPURISp, with respect to their effects on oxidation and deamidation modifications. We identified and quantified varying numbers of monoxidation and deamidation modifications, while dioxidation and trioxidation modifications were not detected. Among the three methods, EPURISp and In-solution exhibited relatively lower levels of monooxidation modifications (Figure 7A). Additionally, the EPURISp method also demonstrated a reduction in nonessential deamidation modifications (Figure 7B), identifying the fewest proteins with deamidation modifications (473 ± 8), followed by In-solution (485 ± 23), and finally the FASP method (544 ± 23). Considering that hydroxyl groups in organic solvents can form hydrogen bonds with amide bonds, the higher levels of deamidation modifications observed in the In-solution and FASP methods may be attributed to the use of organic solvents during the desalting process. In conclusion, the EPURISp method effectively reduces the unfavorable deamidation and oxidation modifications of amino acid residues.

Figure 7 Impact of Sample Preparation Methods on Amino Acid Residue Modifications. (A) Comparison of monooxidation modifications introduced by three sample preparation methods: In-solution, FASP, and EPURISp. (B) Impact of the three methods on deamidation modifications of amino acid residues. The experiment was independently performed in triplicate (N = 3) as biological replicates.

Conclusions

In this study, we introduced a novel proteomic pretreatment technique, EPURISp (Enzymatic Digestion with Ultrafiltration and Rapid In-situ Sample Purification), which combines enzymatic digestion, ultrafiltration, and one-step temperature-controlled vacuum drying for efficient desalting. The EPURISp method demonstrated excellent protein recovery rates across a wide range of molecular weights and hydrophilicity, outperforming traditional C18 desalting approaches. Practical proteomic analysis utilizing EPURISp yielded the highest protein identification yield with remarkable reproducibility, which is particularly advantageous in identifying membrane proteins.

EPURISp showed superior performance in minimizing oxidation and deamidation modifications of amino acid residues compared to conventional FASP methods. This highlights the method’s effectiveness in reducing unfavorable modifications introduced during sample processing and ensuring the accurate identification of intrinsic protein modifications.

The EPURISp method represents a significant advancement in proteomics analysis, providing reliable and efficient results for mass spectrometry. By eliminating the need for multiple desalting steps and reducing the risk of sample loss, EPURISp offers improved accuracy and increased sensitivity for protein and peptide identification.

In conclusion, EPURISp is a promising and innovative approach for high-performance proteomics, enhancing the quality and reproducibility of mass spectrometry analysis. Its versatility and efficiency make it a valuable tool for researchers in the field of proteomics, enabling deeper insights into complex biological systems and protein modifications. The EPURISp method opens new possibilities for advancing proteomics research and contributes to a better understanding of the intricate world of proteins in biological systems.

Data Availability Statement

All data sets generated for this study are included in the manuscript. All of the MS proteomics data have been deposited to ProteomeXchange and can be accessed with the accession PXD044209 (URL: https://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD044209).16,17

Author Contributions

¶ P.L. and M.Y.S. contributed equally to this work. Y.W. conceived the research. P.L., F.D. and M.Y.S. designed and executed the study. X.M.J. was responsible for data analysis. All authors contributed to manuscript editing.

Tianjin Key Medical Discipline (Specialty) Construction Project, Nankai University Eye Institute (NKYKK202214), the Science and Technology Fund of Tianjin Eye Hospital (YKPY2202), Tianjin Health Research Project (TJWJ2023MS035), The National Program on Key Research Project of China (2022YFC2404502), the National Natural Science Foundation of China (82271118), the Tianjin Diversified Investment Fund for Applied Basic Research (21JCZDJC01190), the Tianjin Health and Technology Project (TJWJ2022XK036), and the Tianjin Key Medical Discipline (Specialty) Construction Project (TJYXZDXK-016A).

The authors declare no competing financial interest.

Notes

The study was approved by the Ethics Committee of our department, and written informed consent was obtained from all participants before the study.
==== Refs
References

LaCava J. ; et al. Affinity proteomics to study endogenous protein complexes: pointers, pitfalls, preferences and perspectives. BioTechniques 2015, 58 (3 ), 103–119. 10.2144/000114262.25757543
Wheeler A. R. ; et al. Digital microfluidics with in-line sample purification for proteomics analyses with MALDI-MS. Analytical chemistry 2005, 77 (2 ), 534–540. 10.1021/ac048754+.15649050
Kim K. H. ; et al. Online matrix removal platform for coupling gel-based separations to whole protein electrospray ionization mass spectrometry. J. Proteome Res. 2015, 14 (5 ), 2199–2206. 10.1021/pr501331q.25836738
Wang C. ; et al. Simplified quantitative glycomics using the stable isotope label Girard’s reagent p by electrospray ionization mass spectrometry. J. Proteome Res. 2014, 13 (2 ), 372–384. 10.1021/pr4010647.24274809
Righetti P. G. ; et al. Protein Equalizer Technology: The quest for a “democratic proteome. Proteomics 2006, 6 (14 ), 3980–3992. 10.1002/pmic.200500904.16800034
Zhang G. ; et al. Quantitative proteomics analysis reveals proteins and pathways associated with anthocyanin accumulation in barley. Food chemistry 2019, 298 , 124973 10.1016/j.foodchem.2019.124973.31261005
Zhu Z. ; et al. Integrated proteomics and metabolomics link acne to the action mechanisms of cryptotanshinone intervention. Frontiers in Pharmacology 2021, 12 , 700696 10.3389/fphar.2021.700696.34539397
Liu Q. ; et al. Graphene and graphene oxide sheets supported on silica as versatile and high-performance adsorbents for solid-phase extraction. Angew. Chem., Int. Ed. 2011, 50 (26 ), 5913–5917. 10.1002/anie.201007138.
Kecskemeti A. ; et al. Development of an enzymatic reactor applying spontaneously adsorbed trypsin on the surface of a PDMS microfluidic device. Anal. Bioanal. Chem. 2017, 409 , 3573–3585. 10.1007/s00216-017-0295-9.28299417
Li L. ; et al. A novel method to isolate protein N-terminal peptides from proteome samples using sulfydryl tagging and gold-nanoparticle-based depletion. Anal. Bioanal. Chem. 2016, 408 , 441–448. 10.1007/s00216-015-9136-x.26577085
Shieh I. F. ; Lee C.-Y. ; Shiea J. Eliminating the interferences from TRIS buffer and SDS in protein analysis by fused-droplet electrospray ionization mass spectrometry. J. Proteome Res. 2005, 4 (2 ), 606–612. 10.1021/pr049765m.15822941
Wiśniewski J. R. ; et al. Universal sample preparation method for proteome analysis. Nat. Methods 2009, 6 (5 ), 359–362. 10.1038/nmeth.1322.19377485
Zougman A. ; Selby P. J. ; Banks R. E. Suspension trapping (STrap) sample preparation method for bottom-up proteomics analysis. Proteomics 2014, 14 (9 ), 1006–1000. 10.1002/pmic.201300553.24678027
HaileMariam M. ; et al. S-Trap, an ultrafast sample-preparation approach for shotgun proteomics. J. Proteome Res. 2018, 17 (9 ), 2917–2924. 10.1021/acs.jproteome.8b00505.30114372
Varnavides G. ; et al. In search of a universal method: a comparative survey of bottom-up proteomics sample preparation methods. J. Proteome Res. 2022, 21 (10 ), 2397–2411. 10.1021/acs.jproteome.2c00265.36006919
Ma J. ; et al. iProX: an integrated proteome resource. Nucleic acids research 2019, 47 (D1 ), D1211–D1217. 10.1093/nar/gky869.30252093
Chen T. ; et al. iProX in 2021: connecting proteomics data sharing with big data. Nucleic Acids Res. 2022, 50 (D1 ), D1522–D1527. 10.1093/nar/gkab1081.34871441
