Sensitive and Accurate Quantitation of Phosphopeptides Using TMT

Oct 12, 2017 - An EASY-nLC 1000 UPLC system (Thermo Fisher Scientific, San Jose, CA) with a 50 cm EASY-Spray Column was used to separate 1 μg of enri...
0 downloads 10 Views 930KB Size
Subscriber access provided by LAURENTIAN UNIV

Technical Note

Sensitive and accurate quantitation of phosphopeptides using TMT isobaric labeling technique Xiaoyue Jiang, Ryan Bomgarden, Joseph Brown, Devin L. Drew, Aaron M. Robitaille, Rosa Viner, and Andreas R. Huhmer J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.7b00610 • Publication Date (Web): 12 Oct 2017 Downloaded from http://pubs.acs.org on October 12, 2017

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 free 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 accessible to all readers and 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.

Journal of Proteome Research 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 24

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

Journal of Proteome Research

Sensitive and accurate quantitation of phosphopeptides using TMT isobaric labeling technique Xiaoyue Jiang1#, Ryan Bomgarden2, Joseph Brown1, Devin L Drew1, Aaron M. Robitaille1, Rosa Viner1*, Andreas R Huhmer1 1

Thermo Fisher Scientific, San Jose, CA; 2Thermo Fisher Scientific, Rockford, IL

#

Current affiliation: Rinat-Pfizer Inc., South San Francisco, CA

*Corresponding author: [email protected]

ABSTRACT Phosphorylation is an essential post-translational modification for regulating protein function and cellular signal transduction. Mass spectrometry (MS) combined with isobaric tandem mass tags (TMT) have become a powerful platform for simultaneous, large scale phospho-proteome site identification and quantitation. In order to improve the accuracy of isobaric tag-based quantitation in complex proteomic samples, MS3-based acquisition methods such as Synchronous Precursor Selection (SPS) have been used. However, the method suffers from lower peptide identification rates when applied to enriched phosphopeptide samples compared to unmodified samples due to differences in phosphopeptide fragmentation patterns during tandem MS. We developed and optimized two new acquisition methods for analysis of TMT labeled multiplexed phosphoproteome samples, which resulted in more phosphopeptide identifications with less ratio

1 ACS Paragon Plus Environment

Journal of Proteome Research

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

distortion when compared to previous methods. We also applied these improved methods to a largescale study of phosphorylation levels in A549 cell lines treated with insulin or insulin growth factor 1( IGF-1). Overall, 3,378 protein groups and 12,465 phosphopeptides were identified, of which 10,436 were quantified across 10 samples without pre-fractionation. The accurate measurement enabled us to map to numerous signaling pathways including mechanistic target of rapamycin (mTOR), epidermal growth factor receptor (EGFR, ErbB), and insulin signaling pathways.

Key words: phosphopeptides, phosphorylation, TMT, isobaric, multiplexed quantitation, multistage activation, neutral loss, ratio distortion

INTRODUCTION Reversible phosphorylation is essential post translational modification (PTM) in the regulation of cellular protein activity, subcellular localization, degradation and complex formation. Furthermore, it is the key chemistry in cellular signaling 1. Detection of phosphopeptides and phosphorylation site mapping by mass spectrometry (MS) have been widely adopted in the past decade 2-7. Compared to unmodified peptides, MS analysis of phosphopeptides is more challenging due to substoichiometric levels of phosphopeptides, the lability of the phosphate group in collision-induced dissociation (CID) of peptides during tandem MS 8-10 and poor assignment of phosphopeptide sites by peptide database search algorithms. However, these issues have been significantly improved by using phosphopeptide enrichment techniques such as titanium dioxide11 and immobilized metal affinity chromatography (IMAC) 12

, adoption of new fragmentation methods including HCD 13,14, ETD 15, EThcD 16, and multistage

activation (MSA) 17, and through the use of PTM search algorithms like Byonic 18 and site localization tools such as ptmRS 19 and Ascore 7.

2 ACS Paragon Plus Environment

Page 2 of 24

Page 3 of 24

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

Journal of Proteome Research

The MS-based quantitative analysis of phosphoproteomes is of increasing interest to biologists as it provides information on phosphorylation site dynamics. Label free or isotopic labeling (e.g. SILAC, dimethylation) are standard methods used for peptide quantitation 20-23. While the label free approach is straightforward, it requires close monitoring of liquid chromatography (LC) and MS over a long period of time to ensure reproducibility and accuracy. Precursor isotopic labeling has limited sample multiplexing and results in more complicated MS1 spectra, and therefore is not ideal for global phosphoproteome analysis.

Isobaric mass tagging (e.g., Tandem Mass Tag™ (TMT™)24 or isobaric Tag for Relative and Absolute Quantification (iTRAQ®)25 has become a common technique in MS for relative quantification of proteins26,27. Some advantages of TMT-based multiplexed relative quantification include higher sample multiplexing capability (up to 11), reduced overall experiment time and sample consumption, fewer missing quantitative values, and high reproducibility among samples. One limitation for isobaric labeling is the co-isolation of interference with precursor ions that can result in less accurate quantitation ratios 28

. The most accurate isobaric TMT quantitation on high dynamic range complex mixtures can be

accomplished using the Synchronous Precursor Selection (SPS) MS3 method 29,30. In this method, peptide identification is based on MS2 spectra generated by CID fragmentation in the ion trap, upon which multiple MS2 fragment ions isolated by SPS are subjected to HCD fragmentation at MS3 level for reporter ion quantitation. The SPS significantly reduces the co-isolated interference in MS1 spectra and improves the quantitation accuracy. This new technique has been quickly adopted in the community for many quantitative proteomics applications 31-34.

Recently, SPS MS3 technology has been applied to phosphoproteome quantitation in murine brain and liver tissues with promising results 35. In this study, over 17,800 unique phosphopeptides corresponding to near 14,000 phosphorylation forms were quantified from 24 fractions using the SPS MS3 method.

3 ACS Paragon Plus Environment

Journal of Proteome Research

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

However, the success rate of phosphopeptide identification and quantification per tandem MS spectra were much lower compared to unmodified peptides, especially when considering sample prefractionation and ~3 hour LC-MS analysis for each fraction. This was possibly due to limitations of the classic SPS MS3 acquisition method which commonly results in a high abundant neutral loss (NL) peak for phosphopeptides due to the lability of the phosphate group during CID MS2 peptide fragmentation. These spectra lack an adequate amount of fragment ions for phosphopeptide identification, PTM site localization and selection for the MS3 scan. Therefore, there is a need for improved methods which are optimized to balance phosphopeptide identification, site localization, and quantitation. In this study, we developed two novel instrument methods to extract maximal sequencing and site mapping information from phosphopeptide NL peaks with the goal to provide high phosphopeptide identification success rates as well as accurate TMT based quantitation.

Experimental Section Tandem Mass Tagging Labeling Each of the TMT10plex reagents (0.8mg, Thermo Fisher Scientific, Rockford, IL) was resuspended in 41 µL of anhydrous acetonitrile, and added into 100 µg of HeLa protein digest standard (Thermo Fisher Scientific, Rockford, IL) dissolved in 100 µL of 100mM TEAB buffer. After 1 hour, the reaction was quenched by adding 8 µL of 5% hydroxylamine. Labeled peptides were mixed at ratios of 16:8:4:2:1:1:2:4:8:16 and dried down. Yeast digest (Promega, Madison, WI) was labeled with the last 5 channels of TMT10plex reagents following above procedures, mixed equimolar (0:0:0:0:0:1:1:1:1:1), and then spiked into the aforementioned TMT-labeled HeLa digest sample as interference. A549 cells were serum starved overnight before stimulation with 100nM insulin or 100ng/ml IGF-1 for 15 min. Following treatment, cells were lysed, digested and labeled with TMT10plex reagents in triplicate and pooled for a

4 ACS Paragon Plus Environment

Page 4 of 24

Page 5 of 24

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

Journal of Proteome Research

reference channel (control: 126, 127N, 127C; insulin-treated: 128N, 128C, 129N; IGF-1 treated :129C, 130N, 130C; pooled sample:131). Phosphopeptide Enrichment All labeled samples were enriched for phosphopeptides using a Fe-NTA Phosphopeptide Enrichment Kit (Thermo Fisher Scientific, Rockford, IL) according to manufacturing instructions. Briefly, 1 mg of labeled digests dissolved in binding/wash buffer was loaded onto an equilibrated, phosphopeptide enrichment spin column. Phosphopeptides were bound to the resin for 30 minutes at room temperature without mixing, and the column was washed three times with the same buffer to remove un-bound, nonphosphorylated peptides. Phosphopeptides were eluted using a basic elution buffer, and total peptide concentration was measured using Quantitative Colorimetric Peptide Assay (Thermo Fisher Scientific, Rockford, IL) before LC-MS analysis.

Liquid Chromatography and Mass Spectrometry An EASY-nLC 1000 UPLC system (Thermo Fisher Scientific, San Jose, CA) with a 50cm EASY-Spray Column was used to separate 1µg of enriched phosphopeptides using a 210min gradient at a flow rate of 300 nL/min. Peptides were analyzed on Orbitrap Fusion or Orbitrap Fusion Lumos (Thermo Fisher Scientific, San Jose, CA) mass spectrometers. Spray voltage was set to 1.8-2 kV, RF lens level was set at 60% on Fusion and 30% on Lumos mass spectrometer, ion transfer tube temperature at 275 °C. Full scan resolution was set to 120K with maximum ion injection time of 50ms and automated gain control (AGC) target of 4e5. Four methods were used for comparison including traditional HCD MS2, SPS MS3, and two novel methods named MSA SPS MS3 and NL triggered MS3. The detailed instrument methods are shown in Table 1.

5 ACS Paragon Plus Environment

Journal of Proteome Research

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

Page 6 of 24

Table 1. Different instrument method settings. OT: Orbitrap; IT: Ion trap; MSA: Multistage activation; NL: Neutral loss

MS2

SPS MS3

MSA SPS MS3

NL triggered MS3

dd MS2

HCD OT

CID IT

CID OT

CID IT

Isolation width (Th)

0.7

0.7

0.7

0.7

Collision Energy

38

35

35

35

MSA

-

-

Yes/NL mass 97.9763

-

Resolution

60K

Turbo

30K

Turbo

Ion Injection Time

105

50

60

50

1e5

1e4

5e4

5e4

-

400-1200

400-1200

-

-

-18 to +5

-18 to +5

-

-

TMT

TMT

-

-

-

-

97.9763/79.9658

HCD OT

HCD OT

HCD OT

(msec) Automated gain control (AGC) MS3 filters Precursor selection Range Precursor Ion Exclusion Isobaric Tag Loss Exclusion Targeted Loss dd MS3

6 ACS Paragon Plus Environment

Page 7 of 24

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

Journal of Proteome Research

Isolation Width

-

2

2

2

# of Notches

-

10

10

1

Collision Energy

-

65

65

38

Resolution

-

60K

60K

60K

Scan Range

-

100-500

100-500

100-2000

Ion Injection Time

-

105

105

120

-

1e5

1e5

1e5

(msec) Automated gain control (AGC)

Data Analysis LC-MS data were analyzed using Proteome Discoverer software v.2.1 (Thermo Fisher Scientific) with the SEQUEST HT search engine. Precursor mass tolerance was set to 10ppm. Fragment ion tolerance was 0.02 Da when using the Orbitrap analyzer and 0.6Da when using the Ion trap analyzer. For the phosphopeptide identification using the NL method, both MS2 and MS3 spectra were used for database searching and then combined for identification. Carbamidomethylation on cysteine (+57.021 Da) and TMT6 tags on N termini as well as lysine (+229.163 Da) were set as static modifications. Dynamic modifications included oxidation on methionine (+15.995 Da) and phosphorylation on serine, threonine and tyrosine (+79.966 Da). Data were searched against a UniProt human or a combined database of human and yeast with a 1% FDR criteria using Percolator. Two 10-plex technical replicates were acquired for two proteome experiments with the average and standard deviation reported. For interference samples containing human and yeast proteins, all the reported numbers were filtered for human.

7 ACS Paragon Plus Environment

Journal of Proteome Research

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 ptmRS node 19 was used to localize phosphorylation sites with a probability of 90% or higher considered as a confident phosphorylation site. Neutral loss fragments were identified with a mass tolerance of 0.2 Da for the ion trap analyzer and 0.02 Da for the Orbitrap analyzer, compared to the expected mass calculated based on precursor m/z and charge state. To assess intensity of the neutral loss peak, mass range of 400-2000 m/z was used for HCD MS2 spectra to exclude TMT reporter ions. For quantitation, both unique and razor peptides were used. Reporter ion abundances were corrected for isotopic impurities based on the manufacturer’s data sheets. Signal-to-noise (S/N) values were used to represent the reporter ion abundance with an co-isolation threshold of 75% and an average reporter S/N threshold of 10 and above required for quantitation spectra to be used. The S/N values of peptides, which were summed from the S/N values of the PSMs, were summed to represent the abundance of the proteins. Phosphopeptide identifications and quantitation were imported into Perseus 1.5.6.0 for t-test statistical analysis (FDR90%) phosphopeptide samples, over 63% of CID spectra exhibited a neutral loss (NL) peak as the most intense peak in the spectra. In contrast, only 24% of HCD spectra have NL peaks as the base peak (Figure 1c vs 1a). The strong presence of the neutral loss peak in CID results in the loss of sequence information, potentially compromising the ability to identify the phosphopeptide sequence or localize phosphorylation to a specific residue 36. Method optimization to improve phosphopeptide identification and quantitation accuracy To evaluate the identification and quantitation aspects of phosphoproteomes, a two species phosphoproteome sample was constructed as described in Methods section. HeLa cell digest labeled with TMT10plex reagents at ratios of 16:8:4:2:1:1:2:4:8:16 were mixed with a yeast digest sample labeled with the last 5 channels (1:1:1:1:1) as interference. The mixture was processed using Fe-NTA affinity chromatography to generate a mixture highly enriched for phosphopeptides. For the mixture, the first five channels of HeLa digests should be free of interference and the last five channels should have interfered by yeast proteome. To assess the quantitation accuracy of different acquisition methods, reporter ion ratios of the last five channels (131/129N, 131/129C, 131/130N, 131/130C) were compared to the ratios of the first five channels (126/127N, 126/127C, 126/128N, 126/128C). The number of phosphopeptides and the number of confident phosphorylated sites were also monitored. In addition, the quantitation rate defined as the ratio of phosphopeptides with all 10 channels present for quantitation against all identified phosphopeptides was compared. 10 ACS Paragon Plus Environment

Page 11 of 24

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

Journal of Proteome Research

Multistage activation (MSA) SPS MS3 Recently, an extra multistage activation (MSA) step for ion trap-based instruments has been developed to improve CID phosphopeptide fragmentation and phosphorylation site localization 17. MSA involves simultaneous activation of the precursor ion and the resultant neutral loss product ion during a single CID-MS2 event. Therefore, the observed MS2 product ion spectrum contains a ‘‘composite’’ of the product ions generated by fragmentation of both the precursor ion and the initial neutral loss product ion (Figure 1d). Notably, as the spectrum gets more complex, we found that utilizing the Orbitrap analyzer for detecting MSA CID MS2 spectra was better than the ion trap analyzer as seen by an increase in phosphopeptide identifications and site localization (Figure S2). The high resolution of the Orbitrap enables better assignment of peptide fragments, their charge states and identifying co-eluting peptides that results in an increased ptmRS probability score used for phospho-site localization. By combining MSA with the SPS MS3 method (top 10 notches) for phosphoproteomics quantitation, we observed a 16% increase in phosphopeptide identifications and 30% more phosphopeptides with confidently localized sites compared to the standard SPS MS3 method (Figure 2A). This increase was due in part to improved spectra quality with MSA compared to regular CID fragmentation (Figure S3). In addition to improved peptide identifications, this method also provided a very high peptide quantitation success rate (96%) (Figure 2B, red bar vs. blue bar), similar to what is typically observed with the HCD MS2 method (Figure 2B), while maintaining high quantitative accuracy (Figure 2C). Figure 2

11 ACS Paragon Plus Environment

Journal of Proteome Research

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

Neutral loss (NL) triggered method As stated previously, due to the dominant NL peak observed for many phosphopeptide spectra, CID fragment ions are typically not sufficient for peptide identification nor confident phosphorylation site mapping. Using a NL triggered MS3 event with moderate collision energy (HCD CE38) allowed us to obtain both the sequence information and the quantitation from MS3 spectrum (Figure 1e). In this method, regular MS2 spectra are first generated from the precursor using CID. Then, if the neutral loss peak is detected, a MS3 scan is performed on the most abundant ion in the spectrum. This method could be described as triggering the top 1 peak in the classic SPS MS3 method, but with a wider mass range used for the MS3 spectrum to extract peptide sequence information in addition to quantitation. The quantitation resulting from this method is accurate since the MS3 reporter ions are generated from a unique NL MS2 fragment, which already has potentially co-eluting interference removed (Figure 2c). Our results from this method were consistent with a recent report showing that selecting single notch 12 ACS Paragon Plus Environment

Page 12 of 24

Page 13 of 24

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

Journal of Proteome Research

method for phosphopeptide provides the most accurate quantitation 35. An additional benefit of the NL triggered method is that MS3 spectra (HCD) can be used to complement MS2 spectra (CID) for peptide identification, greatly improving the identification rate over the standard SPS MS3 method (Figure S4). Overall, the phosphopeptide identification and site localization improved 33% and 40% respectively compared to the standard SPS MS3 method (Figure 2a). As only a single notch is used for generating reporter ions for the quantitation, this method is less sensitive compared to MSA SPS method but is only 17% lower than HCD MS2 method (Figure 2b). Comparison of two new methods In agreement with previous results, both MSA and NL triggered methods showed significant phosphopeptide identification improvement over the classic SPS MS3 method. Comparing these two new methods, we observed 6,230 (55%) peptide sequences overlapping, with 1,798 unique peptide sequences from MSA and 3,316 from NL triggered strategy (Figure S5). The detected phosphopeptides from two methods showed similar representation of phosphorylated tyrosine, serine, and threonine with no considerable compositional motif bias (Figure S6). Our results were also consistent with the relative abundance distribution for phosphorylation sites (pS, pT and pY) identified in other large scale phosphoproteome studies 22. These two novel methods characterize more phosphopeptides by activating the predominant NL fragment in phosphopeptide MS2 spectra to produce more sequencing information of the precursor ion. The MSA method produces a MS2 spectrum that is the combination of MS2 and neutral loss peak fragmentation using CID for MS2 level identification. The NL triggered strategy, on the other hand, fragments the top 1 MS2 ion with HCD for MS3 level identification. In our method, a lower HCD collision energy (CE38) is used compared to classic SPS MS3 method (CE65) to help balance reporter ion and fragment ion relative abundances. Overall, these two complementary dissociation methods (MSA SPS 13 ACS Paragon Plus Environment

Journal of Proteome Research

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

MS3 and NL triggered MS3 spectra) result in more phosphopeptide identifications and better site mapping. In theory, the quantitation of isobaric tag reporter ions should be obtainable as long as there is phosphopeptide identification. However, filters such as precursor isolation purity and average reporter ion S/N threshold (see Methods for details) are routinely applied to assure accurate quantitation. This filtering can reduce the success rate of quantitation which corresponds to the sensitivity of the acquisition method. The MSA method using 10 notches combined for the MS3 spectra has 96% success rate of quantitation compared to only 75% using a NL triggered MS3 spectrum produced from a single notch (Figure 2b, red bar Vs blue bar). Therefore, the number of notches selected directly determines the sensitivity for the quantitation at MS3 level. In general, we found that in the NL triggered strategy that the peak triggered for fragmentation was most typically the NL peak, even for phosphotyrosine containing peptides. This contrasts with previous reports where phosphotyrosine peptide spectra did not have a dominant NL peak 37. However, other reports have indicated that NL is a very common occurrence in phosphotyrosine-containing peptides that have been labeled with isobaric tags 38. This is consistent with our results which have dominant NL ions for all TMT-labeled phosphopeptide (STY) species. This may explain why the NL triggered method using single notch provided a more accurate quantitation result compared to the MSA method using additional notches, as the single NL peak represents the bulk of the precursor ion population 35. Notably, in the event of precursor phosphopeptide co-isolating with another interfering phosphopeptide, the NL triggered method may not effectively remove the interfering NL peak from precursor NL peak, when both phosphopeptides share same charge and phosphoric acid loss. However, based on our data and previous studies 35,39, reporter ion interference caused only by co-isolated species of the same charge

14 ACS Paragon Plus Environment

Page 14 of 24

Page 15 of 24

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

Journal of Proteome Research

and mass as a selected precursor, is unlikely. Most importantly, both of these new methods showed significant quantitation accuracy improvement over MS2-based quantitation (Figure 2c). Overall, our two new methods provide significant improvements for phosphopeptide identification, while maintaining highly accurate quantitation. To maximize the benefits of these methods, it is recommended to use both methods and then combine the results for protein quantitation. This approach offers more values for better statistics and the extra identifications from the NL triggered method can become candidates for future targeted quantitation-based methods. If using both methods is not possible, MSA is preferred over NL triggered method when sample injection amount is limited (