This workflow describes the performance of time- and cost-efficient enrichment of multiple protein post-translational modifications (PTMs) simultaneously for quantitative global proteomic analysis. The protocol utilizes peptide-level PTM enrichment with multiple conjugated antibodies, followed by data-independent acquisition mass spectrometry analysis to gain biological insights into PTM crosstalk.
Studying multiple post-translational modifications (PTMs) of proteins is a crucial step to understand PTM crosstalk and gain more holistic insights into protein function. Despite the importance of multi-PTM enrichment studies, few studies investigate more than one PTM at a time, due partially to the expenses, time, and large protein quantities required to perform multiple global proteomic analysis of PTMs. The "one-pot" affinity enrichment detailed in this protocol overcomes these barriers by permitting the simultaneous identification and quantification of peptides with lysine residues containing acetylation and succinylation PTMs with low amounts of sample input. The protocol involves preparation of protein lysate from mouse livers of SIRT5 knockout mice, performance of trypsin digestion, enrichment for PTMs, and performance of mass spectrometric analysis using a data-independent acquisition (DIA) workflow. Because this workflow allows for the enrichment of two PTMs from the same sample simultaneously, it provides a practical tool to study PTM crosstalk without requiring large amounts of samples, and it greatly reduces the time required for sample preparation, data acquisition, and analysis. The DIA component of the workflow provides comprehensive PTM-specific information. This is particularly important when studying PTM site localization, as DIA provides comprehensive sets of fragment ions that can be computationally deciphered to differentiate between different PTM localization isoforms.
A myriad of post-translational modifications dynamically regulate proteins and pathways through effects on the activity1, signaling2, and turnover3,4. For example, protein kinases are activated or deactivated by the addition of phosphate groups5, and histone acetylation and other modifications provide a mechanism to change the chromatin structure and serve as transcriptional regulatory mechanisms6,7. In recent years, evidence has mounted that multiple PTMs work in concert or compete to regulate protein function or activity8,9,10,11. Therefore, understanding PTM crosstalk is an emerging need in PTM research. However, most available proteomic workflows to identify and quantify PTM sites focus on single modifications, rather than the interplay of multiple modifications. The described workflow correlates specific protein modification "hot-spots" and lysine residues that are modified by multiple different PTMs.
There is a growing need in the scientific community for feasible methods to study multiple PTMs simultaneously12. Most methods to globally identify and quantify the sites of multiple types of PTMs are challenging due to the high costs and amount of tissue required12,13. Not only are multi-PTM enrichment experiments time-consuming in terms of sample preparation, data acquisition, and data analysis, but these studies typically require large and often prohibitive amounts of protein11. Described here is a protocol for simultaneous enrichment and analysis of multiple PTMs, which also addresses several of these barriers and enables large-scale PTM profiling and assessing crosstalk between various PTMs14. This one-pot workflow outlines a practical way for biomedical researchers to globally profile multiple PTMs, identify co-modified peptides, and study PTM crosstalk in an efficient and cost-effective way14,15.
Here, this method is showcased by examining mitochondrial protein acylation, which was first studied over 50 years ago16. It specifically concentrates on lysine acetylation17 and succinylation18, including the co-occurrence of these modifications on proteins and even co-modification at the peptide level. Since the study uses a sirtuin 5 (SIRT5) knockout mouse model, it was chosen to focus on enrichment of acetylation and succinylation sites. This decision was made because succinylation sites are targets of SIRT5 desuccinylase and are thus expected to show significant upregulation in KO mice, making them the most relevant PTMs in this case. Both PTMs are biologically relevant as recently summarized by Carrico et al.19. In general, acetylation shows important effects on gene expression and metabolism, and succinylation has been reported to regulate heart metabolism and function20.
The described protocol can be performed with a low amount of protein input material (e.g., 1 mg of protein lysate) and reduces the total duration of the experiment by reducing the time spent for sample processing, MS acquisition, and data analysis. A workflow schematic is provided in Figure 1. We have also used even lower amounts of starting material (down to 100 µg of protein lysate, scaling down the amounts of beads used accordingly), which as expected, reduces the overall yield of identified acylated peptides; however, it still provides highly valuable results and quantifiable acylated peptides.
While so-called top-down or middle-down workflows typically do not use proteolytic digestion approaches (and thus maintain the connectivity of multiple PTMs within one protein), this protocol focuses on a peptide-based affinity enrichment approach to gain extra depth and sensitivity for PTM identification and quantification (Figure 1). In addition, this peptide-centric workflow utilizes modern mass spectrometry methods, including 1) a combination of data-dependent acquisition (DDA) to generate spectral libraries, and 2) data-independent acquisition (DIA) for accurate PTM quantification in a label-free workflow.
DIA workflows overcome sampling stochasticity of typical DDA scan schemes by comprehensively fragmenting all peptide signals within the sampled m/z range21. This feature is also extremely beneficial in terms of site localization, because it is easier to gain information regarding which specific fragment ions are modified within the peptide. In addition, DIA workflows allow identification and quantification of minor PTM-peptide isoforms with identical precursor ions. DIA methods can also determine a specific PTM site localization within a peptide based on specific corresponding fragment ions that are comprehensively measured at all times. However, DDA approaches often utilize "dynamic exclusion" features that exclude multiple sampling of MS/MS for the same precursor ion, thus missing minor PTM isoforms.
The simultaneous enrichment strategy described here is ideally suited for studies that will benefit from the global profiling and quantification of multiple PTMs, examining PTM crosstalk, and understanding the dynamic interactions of post-translational modifications. Identification of multiple enriched PTMs in one combined workflow have been described by: global, serial or parallel enrichment of PTM containing proteolytic peptides, or alternatively by the analysis of intact proteins. In direct comparison with serial enrichment of acetylation and succinylation, the efficiency of the one-pot methodology was established as being very similar14. These alternative protocols require significant amounts of starting material and time and can be prohibitively expensive. In contrast, the one-pot protocol provides an inexpensive and efficient method for enrichment of more than one PTM with subsequent analysis and identification.
Mouse liver tissues are obtained from SIRT5 knockout mice and are used here as starting material. This protocol can also be performed for protein lysates from different tissues or cell culture experiments. This protocol can be applied to protein lysates obtained from tissues or cell culture pellets.
All experiments described in the protocol follow the guidelines of the Buck Institute Institutional Animal Research Committee.
1. Extraction of protein from homogenized tissue and digestion with protease
2. Desalting of non-enriched proteolytic peptides via large-scale solid-phase extraction
3. Simultaneous enrichment of K-acetylated and K-succinylated peptides with immunoaffinity beads
4. Elution of the peptides bound to antibody beads
5. Desalting of enriched peptides
NOTE: The pH of samples should be less than 4 for optimal binding to the tip containing C18 resin. Use a 10 µL pipette tip containing 0.6 µL of C18 resin and a 10 µL pipette to control the tip.
6. Data acquisition using DDA and DIA
NOTE: Analyze the samples with DDA and DIA LC-MS/MS methods, which can be adjusted depending on the available mass spectrometric instrument. Here, the samples were analyzed using a nano-LC 2D HPLC system coupled to a high-resolution mass spectrometer.
7. Data analysis
NOTE: Some data analysis settings should be changed and tailored to the specific experiment. For example, the protein database (FASTA file) selected depends on the species from which the sample was prepared (here, Mus musculus). Below, the data analysis for mouse samples enriched for acetylated and succinylated peptides is described.
8. Data visualization of modified peptides and assessment of PTM site localization
Figure 1 shows a general diagram of the workflow, including harvesting the tissue from mouse livers, using 1 mg of protein for digesting the protein lysate with trypsin, incubating peptides with antibody-conjugated beads, acquiring the samples on the MS, and finally performing DIA/SWATH analysis of the data using various quantitative proteomics software packages (academic and commercial).
Figure 2A shows how the timeline of the workflow and the amounts of sample and protein required, compared to alternative methods currently being used for multi-PTM enrichment studies. The one-pot method can be performed in half as much time and with half the number of samples as these alternative methods. Compared to the two single-PTM enrichment method, the one-pot protocol also requires half the amount of protein.
This protocol has been shown to be a feasible and cost-effective alternative. Figure 2B shows that the median coefficient of variation (CV) for modified peptide areas was lower in the one-pot method than in the single-PTM and serial-PTM enrichments. Figure 2C,D shows that, while comparing the one-pot PTM and single-PTM enrichment methods, no noteworthy differences were apparent between the correlations of site-level quantifications for the two modifications. This was also true for the peptide-level and fragment-level correlations. The same observation held for all three correlations when comparing the one-pot and serial-PTM enrichments. All underlying MS raw data and processed Excel results sheets associated with a recent report by Basisty et al.14 are available and can be downloaded from MassIVE (MSV00081906) and ProteomeXchange (PXD008640).
In general, while antibody enrichment strategies may show certain limitations, such as potential epitope occlusion or limited specificity, the antibodies used in this study are mixtures of independently generated clones and thus provide wider ranges of specificities.
Experimental results document the possibility of detecting and assessing PTM crosstalk. Figure 3 displays data from a successful enrichment and illustrates an example for a peptide containing multiple and different acyl modifications visualizing PTM crosstalk. Figure 3A shows a peptide that is acetylated on one lysine residue and succinylated on the other, and Figure 3B shows the same peptide succinylated at both lysines. This demonstrates that the same lysine residue can be modified with both acylation groups, and there is a possibility of crosstalk occurring at that site. Figure 4 displays the number of lysine residues that were identified as described in sections 7.1-7.3 to carry both modifications, also pointing towards possible PTM crosstalk.
As Figure 5 demonstrates, processing DIA PTM datasets with quantitative proteomics software allows us to pinpoint which specific lysine residues are modified. This is a concept known as site localization, which is an essential step for any analysis for determination of possible PTM crosstalk. Figure 5 displays two potential isoforms along with the confirming and refuting ions for each that could be visualized and assessed as described in sections 8.1-8.3 (specifically steps 8.3.2 and 8.3.3). Based on this information, we were able to confidently identify which of the two isoforms was present in the original sample. The MS/MS spectrum of the confirmed isoform KQYGEAFEKacR demonstrates clearly that the y ions (y2-y5) containing the acetylated lysine residue, which shifted by 42 m/z (the increment mass of an acetyl group), confirmed the specific lysine residue in the peptide that was modified.
Figure 1: Typical workflow for one-pot enrichment of PTMs. Tissue (here, livers) are harvested from SIRT5 KO and wild-type (WT) mice, and proteins are lysed, trypsin-digested into peptides, and desalted. Peptides are then enriched by immunoaffinity with combinations of succinyl- and acetyl-antibody beads. Parallel MS workflows measure both 1) small aliquots of whole lysate protein expression changes (for protein normalization) and 2) enriched acyl-containing peptides for acylation site identification (DDA-MS) and site localization, followed by quantification (DIA-MS). Please click here to view a larger version of this figure.
Figure 2: Comparison of one-pot workflow with alternative methods. (A) Comparison of time, costs, and materials required for the one-pot workflow, serial-PTM enrichment, and two single-PTM enrichments. (B) Comparison of CVs between the one-pot workflow, single acetyl-lysine PTM enrichment, and single succinyl-lysine PTM enrichment. Spearman correlation analysis comparing the acyl peptide peak areas obtained from the one-pot workflow, and the single-PTM enrichments: corresponding plots of the log2 peak area results for (C) acetylation sites and (D) succinylation sites. Regression slopes and correlation factors are indicated in the individual panels14. Two independent biological replicates were processed for each of the conditions. This figure has been modified from Basisty et al.14. Please click here to view a larger version of this figure.
Figure 3: Crosstalk between acetylation and succinylation modifications of lysine residues. MS/MS spectra of tryptic peptides from mitochondrial 3-ketoacyl-CoA thiolase that show the same amino acid sequence but have been modified at two lysine residues with different PTMs. (A) MS/MS of peptide AANEAGYFNEEMAPIEVKsuccTKacK and (B) MS/MS of peptide AANEAGYFNEEMAPIEVKsuccTKsuccK. This figure has been modified from Basisty et al.14. Please click here to view a larger version of this figure.
Figure 4: Overlap and crosstalk between the acetylated and succinylated lysine residues–specific examples in protein complexes. (A) Venn diagram displaying overlap between 2,235 acetylation and 2,173 succinylation sites. Of these, 943 sites were both acetylated and succinylated. Liver from a SIRT5 (de-succinylase) knockout mouse was analyzed, and many succinylation sites were identified. In fact, they were more abundant than normally observed in mouse liver (modified peptides were filtered at a Q value of <0.05). (B) Protein complexes showing the percentage of their subunits containing both acetylated and succinylated sites (bold red line represents the significance as determined by Fisher's exact test). (C) Diagram of ATP synthase complex: protein subunits in red depict the subunits that contain both acetylated and succinylated sites. This figure has been modified from Basisty et al.14. Please click here to view a larger version of this figure.
Figure 5: Quantitative proteomics software deciphers the peptide site localization of PTMs. Based on MS/MS fragmentation of the peptide, it is possible to provide information about the specific lysine residue the acetyl group is modifying. This showcases the ability of the software to offer valuable information on site localization of PTMs. (A) Two possibilities of lysine residue modification and PTM site localization: KQYGEAFEKacR (left) and KacQYGEAFEKR (right). "Confirming" and "refuting" fragment ions are shown for each of the potential site localization isoforms of the peptide. Based on this information, confirming scores and refuting scores are assigned, confirming the presence of isoform KQYGEAFEKacR in the sample. (B) MS/MS spectrum corresponding to the confirmed isoform KQYGEAFEKacR indicating that all y ions including the acetylated lysine residue (y2 and higher) carry an increment mass of +42 m/z, which corresponds to the acetyl modification. Observed b ions do not contain the modification. (C) Extracted ion chromatogram (XIC) with abundant peak areas resulting from y2 and y3 ions, both of which make up the acetylation site in the confirmed isoform KQYGEAFEKacR. Please click here to view a larger version of this figure.
This protocol describes a novel technique for simultaneous multiple PTM enrichment to more effectively understand PTM crosstalk. Alternative methods of reaching this objective tend to be prohibitively time-consuming and expensive, and they require large amounts of protein to be successful11,13. This protocol presents a multi-PTM enrichment workflow that involves incubation in antibody-conjugated beads for two PTMs at once to improve the overall efficiency of the experiment. This method also involves the use of DDA for spectral library generation and DIA MS acquisitions to detect and quantify the peptides present with reduced interference from fragment ions22,23. Software programs, such as MS database search engines are used to analyze and quantify data from DDA acquisitions, whereas Quantitative Proteomics Software24 and specific DIA Quantitative Analysis Software25 are necessary to interpret the complex spectra produced by DIA acquisitions.
There are several critical steps within this protocol that should be followed carefully. As the main goal of the protocol is to enrich for multiple PTMs simultaneously, the antibody-affinity enrichment step (section 2) is critical to success of the experiment. When performing washes on the beads, it is necessary to ensure none of the beads are aspirated accidentally. Ensuring the urea concentration has been diluted to 1 M prior to digestion with trypsin (step 1.8) is also necessary. Although 8 M urea is required earlier in the protocol for protein solubilization, urea concentrations above 1 M will inhibit trypsin's enzymatic activity. In addition, it is important to consistently check the pH of the sample throughout the protocol. This is especially important prior to digestion. If the pH of the sample and the trypsin solution are not neutralized appropriately prior to the incubation, it can result in an inefficient digestion wherein many cleavage sites may be missed, resulting in fewer peptide identifications.
A few modifications to the protocol may be helpful when preparing samples. For a protein lysate procured from 1 mg of starting material, a quarter of the antibody beads provided in each PTM scan tube can be used as a cost-effective alternative. A greater amount of starting material can be used for better results, as long as the amount of antibody beads used are increased proportionately. Another modification that can enhance the workflow is to digest samples with another protease in addition to trypsin. This modification would result in more variability in the peptides cleaved, providing increased coverage of protein residues. Although not necessary, it is recommended that trypsin be one of the enzymes used for PTM analysis due to its high cleavage specificity26.
A limitation of this protocol is that the PTMs being studied need to have similar chemistries in order to be enriched simultaneously14. The procedures for the different antibody-conjugated beads must be similar, utilizing similar solvents and solutions, including similar elution conditions, and preferably from the same vendor. For this reason, the method outlined here consistently uses acetylation and succinylation as an example, both of which use antibody-conjugated beads (Cell Signaling Technology, Inc). Although this method can theoretically be applied to any number of PTMs, additional studies would be needed to assess the exact limitation of the protocol in this respect. Furthermore, as this is an antibody-based enrichment method, the method can only provide a relative quantification of PTM sites.
In comparison with existing multi-PTM enrichment methods, this workflow is a more feasible and cost-effective alternative. From this experiment, it was observed that the efficacy of this method compares extremely well with alternative methods, such as individual or serial enrichments. Figure 2B shows that the median CV for modified peptide peak areas was actually decreased in the one-pot method in comparison to single-PTM enrichments and serial-PTM enrichments13. We further analyzed the experimental results assessing site-level quantifications for acetylation or succinylation. Additionally, Spearman correlation analysis (Figure 2C,D) demonstrated that the one-pot PTM enrichment performed similarly to the single-PTM enrichment workflows. This was also true for the peptide- and fragment-level correlations. The same observation held for all three correlations when comparing one-pot with serial-PTM enrichment.
This protocol allows researchers to make fascinating biological insights into PTM crosstalk in a quick and cost-effective manner. The DIA component of the workflow allows researchers to understand more about PTMs, as it provides information about site localization and overcomes challenges such as low site occupancy of PTMs. Precursor ions tend to be excluded with DDA, which is particularly important when studying PTMs, as site occupancy is often low to the point where these peptides do not get selected for MS/MS but still contain crucial information. Follow-up experiments could be performed to assess the upper limit of how many PTMs can be enriched simultaneously using this method. A future improvement of this workflow may include development of more advanced software platforms to further automate the analysis of site localization and PTM site occupancy.
The authors have nothing to disclose.
We acknowledge the support from the NIH shared instrumentation grant for the TripleTOF system at the Buck Institute (1S10 OD016281). This work was also supported by the National Institute of Allergy and Infectious Disease (R01 AI108255 to B.S.) and the National Institute of Diabetes and Digestive and Kidney Diseases (R24 DK085610 to Eric Verdin; R01 DK090242 to Eric Goetzman). X.X. was supported by a grant from the National Institutes of Health (NIH grant T32GM8806, to Judith Campisi and Lisa Ellerby), N.B. was supported by a postdoctoral fellowship from the Glenn Foundation for Medical Research.
1 M Triethylammonium biocarbonate buffer (TEAB) | Sigma Aldrich, St. Louis, MO, USA | T7408 | |
Acetonitrile, Burdick and Jackson LC-MS grade | Burdick and Jackson, Muskegon, MI, USA | 36XL66 | |
Bioruptor sonicator | Diagenode, Denville, NJ, USA | B01020001 | |
C18 pre-column chip (200 µm x 6 mm ChromXP C18-CL chip, 3 um, 120 A) | SCIEX, Framingham, MA, USA | 5015841 | |
C18-CL chip (75 µm x 15 cm ChromXP, 3 µm, 300 Å) | SCIEX, Framingham, MA, USA | 804-00001 | |
Dithiothreitol (DTT) | Sigma Aldrich, St. Louis, MO, USA | D9779-5G | |
Eppendorf Thermomixer Compact | Eppendorf AG, Hamburg, Germany | T1317-1EA | |
Eppendorf Tube (2.0 mL Safelock) | Eppendorf AG, Hamburg, Germany | 22363352 | |
Formic acid | Sigma Aldrich, St. Louis, MO, USA | F0507-500ML | |
Indexed retention time (iRT) normalization peptide standard | Biognosys AG, Schlieren, Zurich, Switzerland | Ki-3002-2 | |
Iodoacetamide (IAA) | Sigma Aldrich, St. Louis, MO, USA | I1149-25G | |
mapDIA | web link | software for interference removal of DIA datasets | |
Methanol, Burdick and Jackson LC-MS grade | Burdick and Jackson, Muskegon, MI, USA | BJLC230-4 | |
PURELAB flex 1 ultrapure water dispenser | VWR International, Radnor, PA, USA | 89204-088 | |
mProphet in Skyline | incorporated in Skyline | integrated statistical algorithms for FDR assessments | |
Oasis HLB SPE cartridges | Waters Corp., Milford, MA, USA | WAT094225 | cartridges for desalting protein lysates, up to 50 mg material |
Phosphate buffered saline solution | Life Technologies | 10010023 | |
Pierce BCA Assay | Thermo Fisher Scientific, Waltham, MA, USA | 23225 | |
ProteinPilot 5.0 – 'MS database search engine' | SCIEX, Framingham, MA, USA | software download SCIEX | MS database search engine |
PTMScan Succinyl-Lysine Motif [Succ-K] Kit #13764 | Cell Signaling Technology | 13764 | antibody beads for affinity enrichment |
PTMScan Acetyl-Lysine Motif [Ac-K] Kit #13416 | Cell Signaling Technology | 13416 | antibody beads for affinity enrichment |
Sequencing-grade lyophilized trypsin | Life Technologies | 23225 | |
Skyline – 'Quantitative Proteomics Software' | MacCoss lab (academic) | open source software | Quantitative Proteomics Software (academic) |
Spectronaut – 'DIA Quantitative Analysis Software' | Biognosys AG, Schlieren, Zurich, Switzerland | Sw-3001 | DIA Quantitative Analysis Software / PTM site localization |
Thermo Scientific Savant SPD131DDA Speedvac Concentrator | Thermo Fisher Scientific, Waltham, MA, USA | SPD131DDA-115 | instrument to concentrate liquid volume of samples |
TissueLyser II | Qiagen, Hilden, Germany | 85300 | instrument for efficient lysis of tissue |
Trifluoroacetic acid (TFA) | Sigma Aldrich, St. Louis, MO, USA | T6508-1L | |
TripleTOF 6600: orthoganol quadrupole time-of-flight (QqTOF)mass spectrometer | SCIEX, Framingham, MA, USA | Per quote | high resolution mass spectrometer |
Ultra Plus nano-LC 2D HPLC system | SCIEX, Eksigent Division, Framingham, MA, USA | Model #845 | chromatographic separation system |
Urea | Thermo Fisher Scientific, Waltham, MA, USA | PI29700 | |
Water, Burdick and Jackson LC-MS | Burdick and Jackson, Muskegon, MI, USA | 600-30-76 | |
ZipTip C18 Pipette Tips, P10 | Merck Millipore Ltd, Tullagreen, Carrigtwohill, Co. Cork, IRL | ZTC18S096 | C-18 resin loaded tips for desalting of peptide mixtures |