Lysine Propionylation To Boost Sequence Coverage and Enable a

Subscriber access provided by Grand Valley State | University

Article

Lysine Propionylation Boosts Proteome Sequence Coverage and Enables a “Silent SILAC” Labeling Strategy for Relative Protein Quantification Christoph Schräder, Shaun Moore, Aaron A. Goodarzi, and David C. Schriemer Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b01403 • Publication Date (Web): 05 Jul 2018 Downloaded from http://pubs.acs.org on July 8, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Lysine Propionylation Boosts Proteome Sequence Coverage and Enables a “Silent SILAC” Labeling Strategy for Relative Protein Quantification Christoph U. Schräder1, Shaun Moore1,2, Aaron A. Goodarzi1,2 and David C. Schriemer1,3* 1

Department of Biochemistry and Molecular Biology, University of Calgary, Calgary, Alberta, Canada T2N 4N1 Robson DNA Science Centre, Arnie Charbonneau Cancer Institute, University of Calgary, Calgary, Alberta, Canada T2N 4N1 3 Department of Chemistry, University of Calgary, Calgary, Alberta, Canada T2N 4N1 2

________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

ABSTRACT: Quantification in proteomics largely relies on the incorporation of stable isotopes, with protocols that either introduce the label through metabolic incorporation or chemical tagging. Most methods rely on the use of trypsin and/or LysC to generate labeled peptides. Although alternative proteases can enhance proteome coverage, generic quantitative methods that port over to such enzymes are lacking. Here we describe a quantification strategy amenable to most proteases, which involves propionylation of metabolically-labeled lysine, using a “silent SILAC” strategy that reveals isotopic labels on MS2 fragmentation in a TMT-like manner. We selectively propionylated lysine residues prior to digestion to generate pure ArgC-like digestion for trypsin and novel ArgN-like digestions for LysargiNase, by restricting digestion at lysine. The modification offers highly complementary sequence coverage, and even enhanced protein identification rates in certain situations (GluC digestion). Propionylated lysine residues were present in the majority of identified peptides generated from digests of cell lysates, and led to the consistent release of an intense cyclic imine reporter ion at m/z 140 using higher-energy collisional dissociation. We grew A549 cells in media containing either L-1-13C-lysine or L-6-13C-lysine, to generate proteins that share the same accurate mass when paired. Peptides were indistinguishable on the MS1 level, and upon fragmentation released reporter ions at m/z 140 and m/z 141, without otherwise affecting sequence ion mass. The quantification approach is independent of the number of peptide lysines and offers a new strategy for quantitative proteomics. _______________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________________

Stable isotope labeling by amino acids in cell culture (SILAC) is one of the most commonly used techniques for relative protein quantification.1 Through the incorporation of heavy and light versions of lysine (Lys) and/or arginine (Arg), proteins from cell cultures or even animals2 can be differentially tagged. Directly pooling lysates prior to workup can limit variability when compared to quantitative chemical labeling strategies such as TMT or iTRAQ (reviewed by Bantscheff et al.3). SILAC proteins are usually digested with trypsin (when labeled Lys and Arg are used) or LysC (when only labeled Lys is used). This leads to the release of peptides carrying exactly one isotopically tagged residue, and thus a defined mass spacing between differentially labeled peptides.

introduced a neutron-encoded mass signature (NeuCode) as an alternative. By incorporating isobaric Lys isotopologues that differ in mass in the mDa range, metabolic quantification approaching 39-plexing is theoretically feasible.6 It requires the use of state-of-theart FT-MS instruments, where the resolution requirements increase dramatically with the level of multiplexing due to issues with isotope beating.7 The technique is currently restricted to LysC digestion, since multiple Lys residues within the peptide would lead to an increased mass spacing among different states. Other proteases are used for digestion of NeuCode samples (e.g. GluC) but only for identification purposes, where the mass spacing can measure the number of labeled residues per peptide at the MS1 level.8

To avoid overlapping isotopic envelopes, the preferred minimal mass spacing of different isotopes is 4 Da. Caution has to be taken for larger peptides (approx. Mw > 2 kDa), as isotopic envelopes can now overlap and complicate data analysis. Either way, since isotopic differences are directly visible on the MS1 level, the signal complexity of already complex proteomics data sets is increased. It lowers peptide identification rates compared to unlabeled samples,4 and gets worse with increased multiplexing levels.5 Coon and co-workers

Relying on a single protease restricts quantification of the proteome.9 Proteomics is enabled primarily by trypsin, but it offers an incomplete picture of proteoform abundance. For example, several studies have revealed how the use of proteases targeting residues beyond Lys and Arg can increase (phospho)proteome coverage10 and the number of identified proteins.11-13 The interest in nontryptic methods as a means of improved proteoform representation will require new approaches to quantification. But, trypsin remains the protease of choice

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

because of its efficiency and selectivity. It consistently provides the highest number of peptide spectrum matches (PSMs) compared to other available proteases.14,15 Therefore, any new quantification method should retain applicability for trypsin as well.

Page 2 of 11

inhibitor cocktail tablet (Roche) and 200 µM PMSF. Cell lysate was left on ice for 20 minutes then centrifuged to collect cell debris. Protein concentration was determined by Bradford assay. All cell lysates were subsequently TCA-precipitated, washed with acetone and subsequently dissolved in H2O.

Chemical tagging of residues to alter peptide hydrophobicity or fragmentation is common in protein mass spectrometry.16 A recent study demonstrated that propionylation of E. coli lysates prior to tryptic digestion increases the number of identified proteins compared to using trypsin alone.17 Propionic anhydride selectively labels free N-termini and Lys ε-amino groups under controlled conditions.18 In histone research it is routinely used to trigger ArgC-like digestion, which then enables the quantification of post-translational modifications (PTMs) that frequently occur on the Lys-rich histone tails.19 Here, we evaluate proteome-wide propionylation of Lys for digests generated from trypsin, LysargiNase and GluC. It improves protein sequence coverage in all cases and at no cost to protein identification. Based on a very abundant reporter ion derived from propionylated Lys residues (Kpr), we establish a new metabolic labeling approach that combines the best features of metabolic and chemical labeling strategies, sharing elements of a strategy that makes use of immonium ion splitting.20

Propionylation of protein standards and whole cell lysates Propionylation of protein standard mix (BSA (Bos taurus), α-amylase (Bacillus amyloliquefaciens) and myoglobin (Equus caballus)) and whole cell lysates (HeLa or A549 cell lysates, grown and harvested as described above) was conducted as described elsewhere21 with the following modifications. Prior to derivatization of the three protein mix, the protein standards were dissolved in 50 mM NH4HCO3 and blended together in equimolar amounts yielding a final concentration of 10 µM each. The pH of TCA-precipitated human cell lysates was adjusted to 8.0 by the addition of 500 mM NH4HCO3, yielding a final concentration of 50 mM NH4HCO3. The protein concentration was adjusted to 2 µg µL-1. To benchmark SILAC-MS2 based quantification, TCAprecipitated A549 cells were pooled in ratios of 1:1 and 5:1 (1-13C-Lys / 6-13C-Lys). Next, all samples were incubated at 95°C for 15 min to promote denaturation. The derivatization reagent was prepared by mixing propionic anhydride (Sigma Aldrich, St Louis, MO, U.S.A.; product no. 240311) with acetonitrile (ACN) in 1:3 ratio (v/v). This solution was then added to the protein sample in a ratio of 1:3 (v/v) following incubation at 37°C for 20 min. NH4OH was added after the addition of propionic anhydride to a final pH of 8.5. The propionylation step was repeated to drive complete propionylation of Lys residues.22 After the final round of propionylation, samples were evaporated to dryness and subsequently resuspended in doubly-distilled water (H2Odd).

EXPERIMENTAL SECTION Cell Growth and Lysis Human HeLa S3 cells were grown and harvested as described elsewhere.12 The supernatant lysate was TCAprecipitated, frozen in liquid nitrogen and stored at -80 °C until further use. Protein concentration was determined by Bradford assay. L-1-13C-lysine (99%, product no. CLM-653), L-6-13Clysine (99 %, product no. CLM-632), L-arginine (L-Arg, unlabeled, product no. ULM-8347), DMEM media for SILAC (product no. DMEM-500) and dialyzed fetal bovine serum (product no. FBS-100) were purchased from Cambridge Isotope Laboratories, Inc. (Tewksbury, MA, USA). Human A549 adenocarcinoma cells (ATCC CCL-185), subsequently simply denoted as A549 cells, were grown in either: 500 mL DMEM (supplemented with 10 % fetal bovine serum and 1 % penicillin/streptomycin) and supplemented with 50 mg of L-Arg and 50 mg of L-1-13C-Lys; or 500 mL DMEM supplemented with 50 mg of L-Arg and 50 mg of L-613 C-Lys. A549 cells were seeded at 1x105 cells in 4 ml of respective labelling cell media for a total of seven doublings to ensure complete labelling. Media was exchanged every second day and cells were passaged once to ensure cell confluence remained below 90 %. 3x106 cells were used for cell lysis in 5X packed cell volume of RIPA buffer supplemented with a protease

Proteomic Sample Preparation All samples were digested using the FASP protocol.23 Digestions of HeLa whole cell lysates were carried out in two biological replicates and those of the protein standard mix were analyzed in two technical replicates. Merged cell lysates grown in SILAC media were analyzed in two biological replicates. HeLa cell lysates (untreated and propionylated) were digested using GluC, trypsin or LysargiNase. A549 cells (untreated and propionylated) were digested with trypsin alone, and the protein standard mix was digested with GluC only. Briefly, 30 µg of the protein standard mix or 100 µg of whole cell lysate – propionylated or untreated – was loaded onto a 10 kDa filter device and subsequently denatured, reduced and alkylated at pH 8.5 (50 mM


Page 3 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


NH4HCO3 buffer). Addition of buffer was followed by a centrifugation step for 30 min at 14,000 rpm. This was repeated three times. The enzyme was then added to an estimated enzyme-to-substrate ratio of 1:50 (w/w) in case of trypsin and LysargiNase, and 1:25 for GluC. Samples were then incubated overnight at 37 °C for trypsin and LysargiNase, and at room temperature for GluC.

injected and peptides were eluted using a 95 min gradient running linearly from 5 to 30% B followed by an increase to 40% B within 12 min. For column regeneration, the gradient was ramped up to 95 % B within 10 min and kept for 10 min. Charge states between 2+ and 6+ were selected for HCD fragmentation and the maximum injection time at the MS2 level was set to 50 ms with an AGC target value of 5x104. A549 lysate digests were analyzed with the same configuration. Data acquisition was controlled in all experiments with Xcalibur software (v. 4.1.31.9) and all raw data are publicly available via Chorus (https://chorusproject.org/pages/index.html; ID 1457).

Peptides were eluted from the filter device in three steps. First, the filter unit was centrifuged for 30 min at 14,000 rpm. Next, 50 µL of 50 mM NH4HCO3 was added to the filter unit, followed by a second centrifugation step. Final elution was performed by adding 50 µL 0.5 M NaCl to the filter device and subsequent centrifugation. All eluates were collected in the same vial. Prior to mass spectrometric data acquisition, all samples were desalted and concentrated using ZipTips and lyophilization, then reconstituted in 0.1 % FA.

Data Analysis All unprocessed data files were loaded into PEAKS studio24 (version 8.5; Bioinformatics Solutions) and precursor masses were corrected in the software. Peptides were identified by database search and de novo sequencing, using the SwissProt database and including entries for BSA, α-amylase and myoglobin. For the analysis of human cell lysates, the search was restricted to Homo sapiens. Digests of the protein standard mix were searched against the whole database. Mass error tolerance of precursor and fragment ions was set to 10 ppm and 0.02 Da, respectively. Carbamidomethylation of Cys was set as a fixed modification and oxidation of Met as a variable modification for all conditions. To test labeling efficiency, propionylation of Lys, Ser, Thr and Tyr was allowed as a variable modification in the propionylation of the protein standard mix. For propionylated HeLa samples, propionylation of Lys residues (+56.0262 Da) was subsequently included as a fixed modification. For isotopically labeled A549 cell line untreated samples, 13Clabeled Lys (+1.0034 Da) was included as a fixed modification. Propionylated and labeled A549 cell culture samples required the use of 13C-labeled Lys together with propionylation (+ 57.0296 Da) as a fixed modification. Cleavage sites were restricted in two ways. For digests of untreated samples digested with trypsin we restricted the search to P1:Lys/Arg as usual, and for LysargiNase to P1’:Lys/Arg. Corresponding digests of the propionylated samples were restricted to P1:Arg and P1’:Arg, respectively. Untreated and propionylated GluC-digests were both restricted to P1: Glu. We allowed a maximum of three missed cleavage sites for all conditions/proteases used. To evaluate the cleavage specificity, enzyme was set to ‘none’ and cleavage sites were determined as described elsewhere.12 The peptide score threshold was decreased until a false discovery rate (FDR) of 1% on the peptide level was reached. One unique peptide along with a maximum FDR of 1% was set as the lower threshold for identification of protein groups.

nanoLC-MS/MS GluC-digested (untreated or propionylated) protein standard mix was analyzed on an Orbitrap Fusion Lumos Tribrid mass spectrometer, coupled to an EASY-nLC 1200 system, and equipped with an Easy-Spray source (Thermo Fisher Scientific Bremen, Germany). Peptides were chromatographically separated using a PepMap C18 RSLC column (2 µm, 100 Å, 0.075 x 500 mm, flow rate of 300 nL min-1). Approximately 0.1 µg digested protein was injected. The gradient consisted of solvent A (97% H2Odd, 3% ACN, 0.1% FA) and B (3% H2Odd, 97% ACN, 0.1% FA). Peptides were eluted using a 38 min gradient running linearly from 5 to 30% B followed by an increase to 40% B within 7 min. For column regeneration, the gradient was ramped to 95% B within 10 min and kept at 95% B for 5 min. Spray voltage was set to 2200 V. The automatic gain control (AGC) target value for MS1 data acquisition was set to 4x105 and spectra were collected in the m/z range between 350 to 1500 at a resolution of 120,000 and a maximum injection time of 50 ms. Datadependent HCD MS/MS (32% normalized collision energy, NE) was performed on the most abundant precursor ions at Top Speed with a 3 s duty cycle time using a minimum intensity of 5x104 and selecting a charge state between 2+ and 8+. The isolation window was set to 1.2 Th. MS2 spectra were acquired at a resolution of 15,000 and a maximum injection time of 100 ms. The AGC target value was set to 1x105 and the lower mass limit was set to m/z 80 to include low mass reporter ions. Triggered ions were dynamically excluded for 45 s with a tolerance of 10 ppm. HeLa cell lysate digests (untreated and propionylated, digested with either GluC, LysargiNase or trypsin) were analyzed with the same instrumental setup using essentially the same method described above with the following modifications. A sample amount of 0.5 µg was

Peptide Quantification



Peptide ratios were quantified by extracting the relative intensities of the ions m/z 140.1070 and m/z 141.1104 with a mass tolerance of 0.003 Da from mzXML files, generated using RawConverter.25 Scan numbers were subsequently matched to identified peptides in PEAKS and the ratio of the two isotopes was determined.

Page 4 of 11

myoglobin), using GluC for digestion. Propionylated peptides can be more hydrophobic than their unlabeled counterparts,27 and we too observed a shift in peptide elution times using C18 RP-HPLC, and minimal effect on sequence coverage (Figure S1A,B). From the identified peptides, we selected five with a varying degree of Lys residues in order to evaluate the completeness of propionylation and undesired over-propionylation that may occur on amino acids such as Ser, Thr or Tyr.27 We observed almost complete labeling of Lys residues (>95 %), even when multiple Lys residues were present (Figure S2A,B), confirming the high selectivity of the reaction.

RESULTS AND DISCUSSION Validation of propionylation selectivity Derivatizing the ε-amino group of Lys residues to trigger ArgC-like digestion is an emerging technique in proteomics to supplement current digestion strategies.17,26 While additional handling steps can diminish sample quality due to unwanted side reactions, propionylation of Lys residues appears remarkably efficient.18 Based on this, we aimed to evaluate how proteome-wide propionylation would influence identification rates for proteases beyond trypsin.

Effect of propionylation at the proteome level To evaluate the effect of propionylation on protein identification metrics for complex samples, we used trypsin, LysargiNase (which mirrors trypsin’s specificity,28), and GluC as a protease targeting residues beyond Lys.29

We first confirmed incorporation efficiency on a less complex sample of three proteins (BSA, α-amylase and

Figure 1: Identification metrics as a function of work-up protocol. Number of triggered MS2 scans is shown on the right y-axis, represented in grey bars. Number of PSMs, identified peptides and protein groups are shown on the left y-axis. Error bars represent standard deviation (n=2).

While LysargiNase can cleave N-terminal to methylated and dimethylated Lys residues, acetylation prevents cleavage28. Thus, the enzyme should generate an ArgNlike digestion after Lys propionylation, which we confirmed (Figure S3).17 The specificity of GluC was not affected by the derivatization step. Interestingly, detected peptides generated by GluC were enriched in propionylated Lys residues (Figure S3), probably because propionylation reduces the charge of Lys-containing peptides, which in turn facilitates HCD-based fragmentation.30 In keeping with this observation, propionylation led to a 12% increase in MS2-triggered

peptides with a charge state of 2+ for Propionylation had little effect on charge trypsin and LysargiNase (Figure S4A), modified Lys is replaced with an Arg carrier.

GluC digests. states for both because every as the charge

We next evaluated proteome coverage across all three enzymes (with and without propionylation). Tryptic digests of untreated cell lysate yielded the highest number of PSMs and identified peptides, which is in good agreement with previous studies.12-14,31 Propionylation slightly reduced the number of peptides identified for both


Page 5 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


trypsin and LysargiNase, but the number of protein groups is basically unaffected (Figure 1). Derivatization also did not appear to influence trends in fragment ion formation (Figure S4B).

propionylation generated a total coverage of 363,250 amino acids, higher than untreated lysate digested with GluC, LysargiNase and trypsin combined (359,112 amino acids, Figure 3A,B).

Figure 2: Venn diagram visualization of protein groups. Untreated and propionylated HeLa lysate digests with each of the indicated proteases are compared. Only protein groups, which were identified in both biological replicates, were taken into account.

And perhaps not surprisingly given our observations above, all metrics improved in the GluC digestion upon propionylation (Figure 1). Part of the boost may result from a better distributed elution of Lys-containing peptides under standard chromatographic conditions, as peptides with internal free Lys residues tended to elute at earlier time points (Figure S5). Considering only those protein groups identified in both biological replicates, we observed that propionylation provided on average more than 600 new protein groups regardless of the enzyme used (Figure 2). This gain in protein groups cannot be achieved by replicate analysis of the same sample (Figure S6). It offers an easier way to extend proteome coverage: tryptic digestion with and without propionylation provided 327 more protein groups than conventional trypsin, LysargiNase and GluC digests combined (Figure S7). The additional sequence coverage achieved by propionylation was remarkable. For each protease, more than 50% of the respective sequence coverage from the propionylated state was unique (Figure 3A). Here again, tryptic digestion with and without

Figure 3: Venn diagram visualization of sequence coverage, represented as total amino acid count. (A) Comparison of untreated vs propionylated HeLa lysate digested with each of the indicated proteases, along with the relative and absolute amount of amino acids. The combined total number of amino acids is shown in bold. (B) The observed sequence coverage for peptides identified from untreated HeLa lysate digested with GluC, LysargiNase or trypsin. All representation used peptides for which we obtained at least two PSMs among the two biological replicates combined.

Surprisingly, the overall peptide length or mass was not significantly altered by propionylation (Figure S8). Of course, Lys propionylation shouldn’t influence GluC



patterns or average length,14,32 as the residue is not involved in the enzymatic process, but we anticipated an increase for both LysargiNase and trypsin. The lack of an increase does not appear to be due to altered missed cleavages. Digestion at Arg is more efficient than at Lys for both enzymes (Figure S9).28 An efficient ArgN-like digestion strategy (LysargiNase) and ArgC-like digestion strategy (trypsin) should move the average peptide length from ~18 aa to ~27 aa.15 We are not the first to observe a smaller than anticipated shift using alternative digestion strategies.32 The lack of an expected shift suggests additional biases in sample processing, ion selection/fragmentation and/or computational methods, suggesting that the drive for improved coverage might focus on these issues rather than alternative enzymes. Nevertheless, based on our results, the derivatization of Lys residues might be considered the first option for deeper proteome coverage rather than other commonly employed enzymes such as GluC or LysC.11

Page 6 of 11

upon propionylation – two distinct highly abundant reporter ions at m/z 140 and m/z 141 (Figure 5A). This “silent SILAC” method would avoid any signal splitting in the MS1 spectrum, in favor of reporter ion generation in the MS2 spectrum, in a process analogous to TMT labeling.

Fragmentation analysis of propionylated peptides Peptides release amino-acid specific immonium ions upon collisionally-induced dissociation due to multiple fragmentation reactions. These low-mass ions can act as a ‘fingerprint’ for the presence of an amino acid within a peptide.33 Aromatic amino acids yield the most abundant immonium ions, closely followed by histidine.34 While the actual immonium ion of Lys at m/z 101 is usually only observed to minor extent, there are other useful breakdown products (Figure S10A-C), particularly at slightly elevated collision energies.35 A protonated α-amino-εcaprolactam (m/z 129) is a cyclic residue formed through a combination of b/y cleavage, and a protonated cyclic imine (m/z 84) represents a further rearrangement involving just the side-chain. Because both carry the εamino group, modifications lead to a distinct mass shift.3638 The propionylated state generated a cyclic imine at m/z 140 of considerably higher abundance than its unmodified version at m/z 84, and higher also than the α-amino-εcaprolactam (Figure S11). A systematic investigation into the relative abundance of low-mass ions among the different sample conditions confirmed that m/z 140 shows, by far, the highest absolute intensity among all immonium and related ions (Figure 4).

Figure 4: Relative intensity of low-mass ions vs the onset of the given intensity, enumerated in Peptide Spectrum Matches (PSM). Data from GluC-digested propionylated HeLa cell lysate. Identities provided in the figure show the immonium ion along with its corresponding m/z value, together with the labeled cyclic imine reporter ion (Rep-Kpr-140).

To gauge the utility of such a strategy, we note that Lys constitutes approx. 5.7 % of the human proteome, which leads to a mean of 26 Lys residues per protein (Figure S12). We observed that approximately two-thirds of all identified peptides generated by either GluC, LysargiNase or trypsin from propionylated HeLa cells had at least one Lys residue (Figure S13). Thus, a large fraction of peptides is available for quantification. Peptides with multiple labeled Lys residues generate a single reporter ion (Figure S14). In a SILAC-inspired experimental setup, we incorporated the two 13C-Lys amino acids into the growth media for human A549 cell cultures. Except for the additional propionylation step, which is performed on the combined protein extracts at the lysate level, the workflow is the same as for the original SILAC approach (Figure 5B). Incorporation efficiency of the two isotopes after seven cell doublings was determined to be approximately 98 % based on inspection of MS1 spectra (Figure S15A). Differentially labeled peptides co-elute and show the same fragment ion spectra except for the reporter ion, which is either at m/z 84 (for 1-13C-Lys) or m/z 85 (for 613 C-Lys). As expected, the sequence ions for the matched samples are indistinguishable (Figure S15B). We pooled

A “Silent SILAC” approach for quantification Immonium ions have been used in the past for protein quantification, limited mostly by the low intensity of most candidates.20 An exception involves acetylated Lys, which generates an abundant reporter ion at m/z 126.39,40 The high abundance of m/z 140 suggests that a useful quantification method can be paired with the proteome identification attributes of propionylation. A metabolic labeling protocol using Lys isotopes of the same accurate mass, namely 1-13C-Lys and 6-13C-Lys, should yield –


Page 7 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the cell lysates in a defined ratio of either 1:1 or 5:1 and applied the propionylation workflow. We chose ArgClike digestion with trypsin to explore quantification attributes.

Thus, interferences do not appear to be a concern with the method. Duplicate analysis of each mixed sample generated approximately 15,000 peptides and 3,300 protein groups on average (Figure S17). We selected the most intense subset of reporter ion intensities for ratio determination, defined by those peptides containing two or more internal Lys residues, or one Lys located at the Nterminus.41 This allowed us to quantify almost 5000 peptides for both ratios (Figure 6).

Figure 6: Quantitative analysis of A549 cells labeled with Lys 13 13 isotopes 1- C or 6- C. (A) Representative fragment spectra of peptide KprFSYR (Uniprot no. O00571) for mixing ratios 1:1 (top) and 5:1 (bottom). (B) Zoomed view into the low mass region of peptide KprFSYR revealing the different abundance of reporter ions. (C) Box plots showing quantitative accuracy and precision for mixing ratios 1:1 and 5:1 across two biological replicates (theoretical ratios are shown as red, dashed lines). Number of peptides quantified is provided in the figure.

Although the quantification can be readily applied to all peptides containing internal Lys residues, we note that our approach suffers from the same precursor interference problem as other MS2-based quantitative methods, such as TMT or iTRAQ42. Overlapping peptides in the MS1 isolation window can generate chimeric spectra and thus contaminate the reporter ion signal. Since m/z 140 is abundant, we did observe it in spectra of peptides that did not contain Kpr. However, multiple solutions to this problem have been developed.43,44 For example,

Figure 5: MS2-based quantification strategy for propionylated samples. (A) Chemical structures of the two isobaric Lys isotopes, with reporter ions. (B) Workflow illustrating the MS2based quantification strategy for propionylated lysates.

We first examined the likelihood of interferences on the m/z 140 and m/z 141 channels. Other isobaric ions in this mass range can be readily distinguished in the orbitrap analyzer, as the smallest mass difference is 60 ppm and their overall abundance is rather low (Figure S16A,B).



synchronous selection of multiple precursors (SPS-MS3) can improve selectivity. The 32% NE used in our study is unlikely to diminish identification numbers40,45, but SPSMS3 would support independent selection of fragmentation energy, allowing for separate optimization of peptide identification and reporter ion generation. The approach has been used very successfully on complex proteomes, somewhat at the cost of peptide coverage.43,46

Page 8 of 11

Analysis of protein mix composed of BSA, α-amylase and myoglobin; illustration of propionylation efficiency; cleavage specificity analysis of trypsin, LysargiNase and GluC; characteristics of precursor and fragment ions for different HeLa digests; lysine abundance in GluC digested HeLa lysates; additional protein groups per replicate; venn diagram visualization of identified protein groups for untreated HeLa lysate; identified and in silico calculated peptide lengths and masses; missed cleavage site analysis; chemical structures of lysine and related ions formed in the gas phase upon dissociation; product ion spectra of peptide KKLFYSTFATDDRKE from HeLa lysate digests; in silico analysis of lysine residues per protein in the human proteome; number of Lys residues per peptide identified in HeLa lysate digests; product ion spectra of three peptides containing 10 propionylated lysine residues each; incorporation efficiency of heavy lysine into A549 cells; isobaric ions at m/z 140 and 141; identification metrics for labeled A549 cells after propionylation and ArgC-like digestion; chemical structures of isotopes used for a 4-plex approach

CONCLUSIONS We have shown how propionylation prior to digestion enables very clean ArgC-like digestion (trypsin) and a similarly clean ArgN-like digestion (LysargiNase). In the case of trypsin, it provides a simple alternative to the use of other enzymes in the pursuit of deep and complementary sequence coverage, and it should find its way into routine proteomics analysis. GluC peptides are better retained in reversed-phase chromatography, which results in an improved GluC digestion method as well. As to the use of anhydride chemistry, we demonstrate that the reaction products are remarkably clean, and the overall workflow is simple to implement on the blended sample. A highly intense and specific reporter ion at m/z 140 is readily observed for Kpr-containing peptides, which can be used for the accurate determination of relative peptide and protein abundances. Admittedly, there is a selectivity in the quantification method as only Lys-containing peptides report on abundance, but these are abundant across the proteome. Intriguing extensions to this “silent SILAC” method could include other middle-down strategies, especially those involving PTM analysis (e.g. histones). Even top-down proteomics should be possible, where the Lys content will be even higher on a percentage basis.

AUTHOR INFORMATION Corresponding author * David C. Schriemer: [email protected] Acknowledgements This work was supported by an NSERC Discovery Grant 298351-2010 (DCS). DCS acknowledges the additional support of the Canada Research Chair program, Alberta Ingenuity - Health Solutions and the Canada Foundation for Innovation.

It is possible to extend the method to a 4-plex experiment without sacrificing acquisition rates and peptide identification numbers. Through the use of two other Lys isotopes that have 18O incorporated in their backbone and 18 O labeled propionic anhydride, four propionylated Lys isotopes of the same accurate mass can be generated that yield four distinct reporter ions (Figure S18). Any application of the current method requires some caution, as Kpr is also a naturally occurring PTM in vivo, but usually only detectable after exhaustive immunoprecipitation enrichment.47 However, one can imagine replacing propionic anhydride with an alternative blocking agent, possibly containing other rare isotopes, to shift the reporter mass away from an unwanted m/z value, and increase the scope of this Silent SILAC method .

REFERENCES

(1) Ong, S. E.; Blagoev, B.; Kratchmarova, I.; Kristensen, D. B.; Steen, H.; Pandey, A.; Mann, M. Mol Cell Proteomics 2002, 1, 376-386. (2) Krüger, M.; Moser, M.; Ussar, S.; Thievessen, I.; Luber, C. A.; Forner, F.; Schmidt, S.; Zanivan, S.; Fassler, R.; Mann, M. Cell 2008, 134, 353-364. (3) Bantscheff, M.; Lemeer, S.; Savitski, M. M.; Kuster, B. Anal Bioanal Chem 2012, 404, 939-965. (4) Tebbe, A.; Klammer, M.; Sighart, S.; Schaab, C.; Daub, H. Rapid Commun Mass Sp 2015, 29, 795801. (5) Merrill, A. E.; Hebert, A. S.; MacGilvray, M. E.; Rose, C. M.; Bailey, D. J.; Bradley, J. C.; Wood, W. W.; El Masri, M.; Westphall, M. S.; Gasch, A. P.; Coon, J. J. Mol Cell Proteomics 2014, 13, 2503-2512.

ASSOCIATED CONTENT Supporting Information


Page 9 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Wein, S.; Afjehi-Sadat, L.; Garcia, B. A. Proteomics 2015, 15, 1459-1469. (23) Wisniewski, J. R.; Zougman, A.; Nagaraj, N.; Mann, M. Nat. Methods 2009, 6, 359-362. (24) Zhang, J.; Xin, L.; Shan, B.; Chen, W.; Xie, M.; Yuen, D.; Zhang, W.; Zhang, Z.; Lajoie, G. A.; Ma, B. Mol. Cell. Proteomics 2012, 11, M111 010587. (25) He, L.; Diedrich, J.; Chu, Y. Y.; Yates, J. R., 3rd. Anal Chem 2015, 87, 11361-11367. (26) Sanders, J. D.; Greer, S. M.; Brodbelt, J. S. Anal Chem 2017, 89, 11772-11778. (27) Meert, P.; Govaert, E.; Scheerlinck, E.; Dhaenens, M.; Deforce, D. Proteomics 2015, 15, 2966-2971. (28) Huesgen, P. F.; Lange, P. F.; Rogers, L. D.; Solis, N.; Eckhard, U.; Kleifeld, O.; Goulas, T.; Gomis-Ruth, F. X.; Overall, C. M. Nat. Methods 2015, 12, 55-58. (29) Breddam, K.; Meldal, M. Eur J Biochem 1992, 206, 103-107. (30) Frese, C. K.; Altelaar, A. F. M.; Hennrich, M. L.; Nolting, D.; Zeller, M.; Griep-Raming, J.; Heck, A. J. R.; Mohammed, S. J. Proteome Res. 2011, 10, 23772388. (31) Tsiatsiani, L.; Giansanti, P.; Scheltema, R. A.; van den Toorn, H.; Overall, C. M.; Altelaar, A. F.; Heck, A. J. J Proteome Res 2017, 16, 852-861. (32) Cristobal, A.; Marino, F.; Post, H.; van den Toorn, H. W.; Mohammed, S.; Heck, A. J. Anal Chem 2017, 89, 3318-3325. (33) Falick, A. M.; Hines, W. M.; Medzihradszky, K. F.; Baldwin, M. A.; Gibson, B. W. J Am Soc Mass Spectr 1993, 4, 882-893. (34) Michalski, A.; Neuhauser, N.; Cox, J.; Mann, M. J Proteome Res 2012, 11, 5479-5491. (35) Yalcin, T.; Harrison, A. G. J Mass Spectrom 1996, 31, 1237-1243. (36) Couttas, T. A.; Raftery, M. J.; Bernardini, G.; Wilkins, M. R. J. Proteome Res. 2008, 7, 2632-2641. (37) Schräder, C. U.; Heinz, A.; Majovsky, P.; Schmelzer, C. E. H. J Am Soc Mass Spectr 2015, 26, 762-773. (38) Trelle, M. B.; Jensen, O. N. Anal Chem 2008, 80, 3422-3430. (39) Nakayasu, E. S.; Wu, S.; Sydor, M. A.; Shukla, A. K.; Weitz, K. K.; Moore, R. J.; Hixson, K. K.; Kim, J. S.; Petyuk, V. A.; Monroe, M. E.; Pasa-Tolic, L.; Qian, W. J.; Smith, R. D.; Adkins, J. N.; Ansong, C.

(6) Hebert, A. S.; Merrill, A. E.; Bailey, D. J.; Still, A. J.; Westphall, M. S.; Strieter, E. R.; Pagliarini, D. J.; Coon, J. J. Nat. Methods 2013, 10, 332-334. (7) Overmyer, K. A.; Tyanova, S.; Hebert, A. S.; Westphall, M. S.; Cox, J.; Coon, J. J. Nature protocols 2018, 13, 293-306. (8) Rose, C. M.; Merrill, A. E.; Bailey, D. J.; Hebert, A. S.; Westphall, M. S.; Coon, J. J. Anal Chem 2013, 85, 5129-5137. (9) Peng, M.; Taouatas, N.; Cappadona, S.; van Breukelen, B.; Mohammed, S.; Scholten, A.; Heck, A. J. Nat. Methods 2012, 9, 524-525. (10) Gonczarowska-Jorge, H.; Loroch, S.; Dell'Aica, M.; Sickmann, A.; Roos, A.; Zahedi, R. P. Anal Chem 2017, 89, 13137-13145. (11) Bekker-Jensen, D. B.; Kelstrup, C. D.; Batth, T. S.; Larsen, S. C.; Haldrup, C.; Bramsen, J. B.; Sorensen, K. D.; Hoyer, S.; Orntoft, T. F.; Andersen, C. L.; Nielsen, M. L.; Olsen, J. V. Cell systems 2017, 4, 587-599 e584. (12) Schräder, C. U.; Lee, L.; Rey, M.; Sarpe, V.; Man, P.; Sharma, S.; Zabrouskov, V.; Larsen, B.; Schriemer, D. C. Mol Cell Proteomics 2017, 16, 1162-1171. (13) Swaney, D. L.; Wenger, C. D.; Coon, J. J. J. Proteome Res. 2010, 9, 1323-1329. (14) Giansanti, P.; Tsiatsiani, L.; Low, T. Y.; Heck, A. J. Nat. Protoc. 2016, 11, 993-1006. (15) Tsiatsiani, L.; Heck, A. J. FEBS J 2015, 282, 26122626. (16) Leitner, A. Analytica chimica acta 2018, 1000, 2-19. (17) Golghalyani, V.; Neupartl, M.; Wittig, I.; Bahr, U.; Karas, M. J Proteome Res 2017, 16, 978-987. (18) Meert, P.; Dierickx, S.; Govaert, E.; De Clerck, L.; Willems, S.; Dhaenens, M.; Deforce, D. Proteomics 2016, 16, 1970-1974. (19) Garcia, B. A.; Mollah, S.; Ueberheide, B. M.; Busby, S. A.; Muratore, T. L.; Shabanowitz, J.; Hunt, D. F. Nature protocols 2007, 2, 933-938. (20) Colzani, M.; Schutz, F.; Potts, A.; Waridel, P.; Quadroni, M. Mol Cell Proteomics 2008, 7, 927-937. (21) Lin, S.; Garcia, B. A. Nuclesomes, Histones & Chromatin, Pt A 2012, 512, 3-28. (22) Sidoli, S.; Yuan, Z. F.; Lin, S.; Karch, K.; Wang, X.; Bhanu, N.; Arnaudo, A. M.; Britton, L. M.; Cao, X. J.; Gonzales-Cope, M.; Han, Y.; Liu, S.; Molden, R. C.;



International journal of proteomics 2014, 2014, 730725. (40) Zolg, D. P.; Wilhelm, M.; Schmidt, T.; Medard, G.; Zerweck, J.; Knaute, T.; Wenschuh, H.; Reimer, U.; Schnatbaum, K.; Kuster, B. Mol Cell Proteomics 2018. (41) Hohmann, L. J.; Eng, J. K.; Gemmill, A.; Klimek, J.; Vitek, O.; Reid, G. E.; Martin, D. B. Anal Chem 2008, 80, 5596-5606. (42) Ow, S. Y.; Salim, M.; Noirel, J.; Evans, C.; Rehman, I.; Wright, P. C. J Proteome Res 2009, 8, 5347-5355. (43) McAlister, G. C.; Nusinow, D. P.; Jedrychowski, M. P.; Wuhr, M.; Huttlin, E. L.; Erickson, B. K.; Rad, R.; Haas, W.; Gygi, S. P. Anal. Chem. 2014, 86, 71507158. (44) Wenger, C. D.; Lee, M. V.; Hebert, A. S.; McAlister, G. C.; Phanstiel, D. H.; Westphall, M. S.; Coon, J. J. Nat. Methods 2011, 8, 933-935. (45) Zolg, D. P.; Wilhelm, M.; Schnatbaum, K.; Zerweck, J.; Knaute, T.; Delanghe, B.; Bailey, D. J.; Gessulat, S.; Ehrlich, H. C.; Weininger, M.; Yu, P.; Schlegl, J.; Kramer, K.; Schmidt, T.; Kusebauch, U.; Deutsch, E. W.; Aebersold, R.; Moritz, R. L.; Wenschuh, H.; Moehring, T., et al. Nat. Methods 2017, 14, 259-262. (46) Wühr, M.; Guttler, T.; Peshkin, L.; McAlister, G. C.; Sonnett, M.; Ishihara, K.; Groen, A. C.; Presler, M.; Erickson, B. K.; Mitchison, T. J.; Kirschner, M. W.; Gygi, S. P. Curr Biol 2015, 25, 2663-2671. (47) Sun, M.; Xu, J.; Wu, Z.; Zhai, L.; Liu, C.; Cheng, Z.; Xu, G.; Tao, S.; Ye, B. C.; Zhao, Y.; Tan, M. J Proteome Res 2016, 15, 4696-4708.


Page 10 of 11

Page 11 of 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC


Lysine Propionylation To Boost Sequence Coverage and Enable a

Recommend Documents