Unrestrictive identification of non-phosphorylation PTMs in yeast kinases by MS and PTMap

doi:10.1002/pmic.200900510

Proteomics. Author manuscript; available in PMC 2010 June 28.

Published in final edited form as:

Proteomics. 2010 March; 10(5): 896–903.

doi: 10.1002/pmic.200900510

PMCID: PMC2892973

NIHMSID: NIHMS209133

Unrestrictive identification of non-phosphorylation PTMs in yeast kinases by MS and PTMap

Kai Zhang,¹^* Yue Chen,² Zhihong Zhang,² Shengce Tao,^3,^4,⁵ Heng Zhu,^3,⁴ and Yingming Zhao²

¹ Department of Chemistry, Nankai University, Tianjin, P. R. China

² Ben May Department for Cancer Research, University of Chicago, Chicago, IL, USA

³ Departments of Pharmacology and Molecular Sciences, The Johns Hopkins University School of Medicine, Baltimore, Maryland, MD, USA

⁴ The High Throughput Biology Center, The Johns Hopkins University School of Medicine, Baltimore, Maryland, MD, USA

⁵ Shanghai Center for Systems Biomedicine, Shanghai Jiaotong University, Shanghai, P. R. China

Correspondence: Professor Yingming Zhao, The Ben May Department for Cancer Research, The University of Chicago, Chicago, IL, USA, Email: Yingming.Zhao/at/uchicago.edu, Fax: +1-713-218-7188

^*Additional corresponding author: Professor Kai Zhang, zhangkai730/at/yahoo.com.cn

The publisher's final edited version of this article is available at Proteomics

Abstract

A few protein PTMs, such as phosphorylation and ubiquitination, are known to be critical in regulation of protein kinase activities. However, the roles of other PTMs have not been extensively studied in kinases. Development of a comprehensive description of all types of PTMs and discovering novel in vivo PTMs in low abundance represent major analytical challenges. Toward this goal, we have developed a strategy for systematic and accurate identification of the full-spectrum of PTMs in yeast protein kinases. Our strategy involves isolation of GST-fused kinase proteins, MS analysis, and unrestrictive PTM identification by PTMap algorithm. Among the 30 purified yeast kinases, we identified 27 different types of PTMs, and 53 PTM sites, among which are 13 novel mass shifts that have not been previously reported, likely representing novel PTMs. These results represent a significant expansion of our current understanding of PTMs in kinases and suggest highly complex regulation of kinase function.

Keywords: Cell biology, Kinase, MS, Protein post-translational modifications, PTMap

1 Introduction

Protein phosphorylation is one of the most important and prevalent PTMs and is involved in regulation of diverse cellular functions, including cell division, growth, differentiation, and intercellular communication. [1, 2] Protein kinases typically catalyze phosphorylation of serine, threonine, or tyrosine residues by transferring the gamma phosphate group of ATP to substrate proteins. The family of protein kinases – the kinome – is encoded by approximately 518 ORFs in humans and by 122 ORFs in the budding yeast. [3–5] The proteins comprising the kinome have been considered to be highly attractive targets for drug development in a wide variety of diseases. [6]

Previous studies have indicated that in addition to phosphorylation, the activities and functions of protein kinases are regulated by diverse types of PTMs, such as ubiquitination, acetylation, glycosylation, and myristoylation. [7–11] Nevertheless, identities of PTMs other than phosphorylation have not been carefully examined in kinases. [12] A comprehensive description of the full-spectrum of PTMs is required to further characterize protein kinase functions, and to elucidate the mechanisms by which kinase pathways in biological processes are regulated.

Here, we report a systemic study of the non-phosphorylation PTMs in protein kinases (Fig. 1). Our strategy combines three main steps to achieve high-confidence and full-spectrum PTM identification. First, GST-tagged yeast kinases were affinity-purified and resolved on SDS-PAGE. Second, the kinase proteins were in-gel digested and analyzed by nano-HPLC/MS/MS. Finally, MS/MS were aligned with corresponding protein sequences using PTMap, we recently developed an algorithm for unrestrictive identification of all possible PTMs, whether known or unknown. Using this strategy, we identified 53 non-phosphorylation PTM sites in 30 yeast kinases. In addition to acetylation and methylation, some previously unknown mass shifts were also found, suggesting possible novel PTMs. Our results show that the complexity of PTMs in kinases is more diverse than previously thought. Our study also demonstrates that this strategy allows for the discovery of novel, unknown PTMs with low stoichiometry.

Figure 1

Schemes for PTM analysis of yeast kinases.

2 Materials and methods

2.1 Materials

Water and ACN were from Fisher Scientific (Pittsburgh, PA, USA). TFA was from Sigma-Aldrich (St. Louis, MO, USA). Colloidal Blue Staining Kit was from Invitrogen (Carlsbad, CA, USA). Sequencing-grade trypsin was from Promega (Madison, WI, USA). C18 ZipTips were from Millipore (Bedford, MA, USA).

2.2 Preparation of protein kinase

Expression and purification of protein kinases were carried out as described. [4] Briefly, protein kinase genes were cloned into a high-copy URA3 expression vector which produces GST fusion proteins. Thirty yeast expression strains containing plasmids of interest in which the correct yeast ORFs were fused in-frame into GST (Supporting Information Table S1) were isolated. The proteins were over-expressed in yeast, and were purified from yeast extracts using GST-agarose beads.

The recombinant fusion proteins were eluted from GST-agarose beads in a sample-loading buffer (20mM Tris-HCL (pH 6.8), 3% glycerol, 1% SDS, 0.03% bromophenol blue, and 2% mercaptoethanol). The supernatants were collected after centrifugation and dried down to 30 μL in a Speed-Vac prior to SDS-PAGE.

2.3 Protein in-gel digestion

The purified GST-kinases were first resolved by SDS-PAGE and visualized by colloidal Coomassie staining (Fixative solution is Proto Blue Safe colloidial coomassie G-250 staining solution, it includes: citric acid, ammonium sulfate, ethanol and coomassie blue G-250.). Gel bands of interest were excised and subjected to in-gel digestion as described previously. [12] Briefly, the gel band was sliced into small pieces (~1 mm) and destained in 25 mM ammonium bicarbonate in ethanol/water (50:50 v/v). The destained gel pieces were washed three times in 50% ethanol for 1 h, and twice in water for 20 min. The gel pieces were dehydrated in ACN and dried in a SpeedVac (Thermo Fisher, Waltham, MA). Dried gel pieces were immersed in 200 μL of 50 mM ammonium bicarbonate (pH 8.0) containing 0.2 μg porcine-modified trypsin (Promega) and incubated overnight at 37°C. Tryptic peptides were sequentially extracted from the gel pieces with 50% ACN (ACN/water/TFA, 50:45:5 v/v/v) and 75% ACN (ACN/water/TFA, 75:24:1, v/v/v). The peptide extracts were pooled, dried in a SpeedVac, and desalted using a μ-C18 Ziptip before HPLC/MS/MS analysis.

2.4 HPLC/MS/MS analysis

HPLC/MS/MS analysis was carried out by nano-HPLC/LTQ MS as described previously [13]. Briefly, each tryptic digest was dissolved in 10 μL of HPLC buffer A (0.1% v/v acetic acid in water), 2 μL of which were injected into an Agilent HPLC system (Agilent, Palo Alto, CA). Peptides were separated on a homemade capillary HPLC column (100 mm length × 75 mm inner diameter) containing Jupiter C12 resin (4 μm particle size, 90-Å pore diameter (Phenomenex, St. Torrance, CA, USA)) and electrosprayed directly into the mass spectrometer using a nanospray source. The LTQ mass spectrometer was operated in a data-dependant mode, cycling between acquisition of one MS spectrum followed by acquisition of MS/MS spectra of the ten strongest ions in that MS spectrum.

2.5 Protein sequence database search and manual verification

All MS/MS spectra were searched against the NCBI-nr protein sequence database using the PTMap software. [14] A PTMap score of 1.0 and S_Unmatched scores of 4.0:10.0 (high:low mass ranges) were used to filter false positives. [14] High mass range is defined as those MS peaks in the mass range between doubly changed parent m/z and singly charged parent m/z, whereas low mass range as 0 to doubly changed parent m/z. Parameters for the PTMap database search included a mass error of ±4.0 Da for parent ions, a mass error of ±0.6 Da for fragment ions, and six allowed missed proteolytic cleavages. All modified peptides identified with a PTMap score >1 were further examined manually using the previously described rules to ensure accuracy of the peptide identifications. [14]

3 Results and discussion

3.1 Expression and purification of kinase proteins

Previous studies based on the yeast proteome chip platform have demonstrated the robustness and effectiveness of using GST-kinase fusion proteins to systematically characterize kinase functions. [4, 5] We took advantage of this approach in undertaking a comprehensive study of kinase PTMs. Cloning of GST-kinase fusion proteins in high-copy URA3 vectors enabled us to produce sufficient quantities of almost all of the yeast kinase proteins for subsequent MS analysis. Using this approach, we purified 30 yeast kinases (Supporting Information Table S1 and Fig. S1), which were confirmed by MS analysis.

3.2 Identification of non-phosphorylation PTMs in kinase

To determine the sites of diverse non-phosphorylation PTMs in yeast kinases, the GST-kinase proteins were analyzed using a procedure that we have described previously. [13, 14] Briefly, the purified kinase proteins were resolved in SDS-PAGE gels, excised, and subjected to in-gel digestion and nano-HPLC/MS/MS analysis using an LTQ mass spectrometer.

We carried out unrestricted sequence alignment using the PTMap algorithm to analyze MS/MS datasets of the yeast kinase proteins. PTMap, recently developed in our laboratory, is the first sequence alignment strategy that emphasizes unmatched peaks for identifying false positives. [14] It incorporates several unique features to improve sensitivity and accuracy of peptide identification, including a unique procedure for peak selection, automatic mass-shift adjustment, and precise PTM localization. This algorithm is able to identify any PTM, irrespective of prior knowledge. [14]

Sequence alignment and subsequent manual inspection using a procedure described previously [14] led to the identification of 53 sites in 30 yeast kinases, which corresponded to 27 types of non-phosphorylation PTMs after removal of artifacts arising from in vitro chemical modifications, which were frequently observed (e.g. oxidation and acrolein addition) (Tables 1 and and2,2, and Supporting Information Table S2). Of the 27 types of non-phosphorylation PTMs, 13 could not be attributed to any known PTMs (http://www.unimod.org) (Table 2), and therefore likely represent novel PTMs. All MS/MS spectra for the modified peptides are summarized and shown in Supporting Information Table S2 and Fig. S2.

Table 1

Methylated and acetylated peptide sequences in yeast kinases

Table 2

Mass shifts of the modified residues in yeast kinases

3.3 Identification of acetylation in kinases

A recent study showed that mitogen-activated protein kinase kinase (MAPKK6J) can be acetylated at Ser207 and Thr211 by YopJ, a bacterial acetyltransferase. [10] Serine acetylation was subsequently found at Ser222 and Ser226 of mitogen-activated protein kinase kinase 1/2 (MEK2). [11] The covalent acetylation modification prevents phosphorylation of these two serine residues and inhibits activation of the kinase [10, 11] To determine if additional O-acetylation sites are present in protein kinases in vivo, we focused on identifying new acetylation sites using the PTMap analysis. We found four acetylated serine residues and three acetylated threonine residues in seven kinases (Table 1).

To confirm these new acetylation sites, we first compared the fragmentation patterns of the in vivo-derived peptides bearing modifications and their unmodified counterparts. Similar MS/MS peak patterns were observed for each pair of acetylated and unmodified peptides as shown in Fig. 2 and Supporting Information Fig. S2. These covalent modifications were further validated by the stringent manual inspection procedure as described previously. [13] Furthermore, the identification of the new acetylation sites is supported by the evidence that these modifications lead to increased HPLC-retention time and side-chain hydrophobicity (Supporting Information Table S2). For example, the HPLC-retention time of NL_ACSFEETPDYEGYR was found to be 57.51 min as compared with 56.06 min for the unmodified peptide in a 100 min HPLC MS/MS run.

Figure 2

Identification of O-acetylated peptides in yeast kinases. (A) MS/MS of an in vivo-acetylated peptide NL_ACSFEETPDYEGYR″, and (B) its corresponding unmodified isoform. (C) MS/MS of an in vivo O-acetylated peptide “SEIDTANFDQEF_ACTK”, (more ...)

Taken together, the data confirmed acetylation sites at Ser328 (Yck2p), Ser178 (Hog1p), Ser537 (Dbf2p), Ser299 (Ybr028cp), Thr634 (Ypk2p), Thr861 (Cdc15p), and Thr303 (Cka1p). These kinases are known to be involved in various functions, including DNA repair, cell morphogenesis, and nuclear division (Table 1). Among them, Hog1p is also an MAPK protein. To the best of our knowledge, these O-acetylated sites have not been reported previously and their effects on phosphorylation are still unclear.

We believe that the serine and threonine acetylation identified in this study are not due to in vitro artifacts because of the following three reasons: (i) acetic acid was not used in staining or destaining buffer for SDS-PAGE gels; (ii) a very low concentration of acetic acid was used in the HPLC buffer during HPLC/MS/MS analysis; and (iii) acetic acid was avoided when the peptides were cleaned in ZipTip.

3.4 Identification of methylation in kinases

Protein methylation has been shown to play important roles in epigenetics and diseases. [15, 16] To our knowledge, methylation of protein kinases has not been previously reported. In this study, we identified three different types of methylation on three sites in yeast kinases: mono-methylation of Arg182 (Sch9p), Lys423 (Yak1p), and Lys635 (Ypk2p). We confirmed these modifications through careful manual inspection of MS/MS spectra and by comparison of HPLC-retention times to those of the corresponding unmodified peptides as described above (Fig. 3, Table 1, and Supporting Information Table S2). An interesting observation is that lysine methylation (Lys635) and threonine acetylation (Thr634) were identified in adjacent residues in Ypk2p. To confirm these PTMs in Ypk2p, we compared the MS/MS spectra and HPLC-retention time of the methylated peptides with the corresponding unmodified and acetylated peptides (Figs. 2 and and33 and Supporting information Table S2). The fragmentation patterns of lysine-methylated peptides are similar to those of corresponding unmodified and acetylated peptides; however, the identifications can be verified by stringent manual inspection. The HPLC-retention times for unmodified peptide, and Lysine-methylated and Thr-acetylated peptides were 44.31, 44.75, and 45.43 min, respectively, which suggests increasing hydrophobicity moving from unmodified to methylated to acetylated peptide. To obtain semi-quantification information, we evaluated the MS signal intensities of three peptide forms. The signal intensities of Lys-methylated and Thr-acetylated peptides are 1000 and 500 times lower, respectively, than the unmodified counterpart, suggesting that the two PTMs have very low stoichiometry. While it is known that Ypk2p participates in a signaling pathway required for optimal cell wall integrity, [17] it remains to be determined if the two PTMs have functional consequences.

Figure 3

Identification of methylated peptides in yeast kinases. (A) MS/MS of an in vivo methylated peptide “EAAAAAYGPDTDIP_MeR”, and (B) its corresponding unmodified isoform. (C) MS/MS of an in vivo methylated peptide “SEIDTANFDQEFT_MeK”, (more ...)

3.5 Novel PTMs in kinase

In addition to acetylation, methylation and other known PTMs (such as histidine oxidation, [18] and deamidation, [19]) (Figs. 4A and B), PTMap identified a series of novel mass shifts in yeast kinases that have not been described previously (http://www.unimod.org). These novel mass shifts include +74 Da at arginine, +54, and +42 Da at aspartic acids, +136 Da at glutamic acid, +96 Da and +34 Da at histidines, and +163, +108, +150 and +34 Da at tryptophans (Table 2). The MS/MS spectra of all the modified peptides are shown in Supporting Information Fig. S2. To learn more about these novel mass shifts in protein kinases, we compared the MS signal intensities and retention times for each pair of modified and unmodified peptides (Supporting Information Table S2). The modified peptides were usually much less abundant than the corresponding unmodified counterparts. For example, the MS signal intensity of LNEEW(+163)SSYLQR was 120-fold lower than the signal of LNEEWSSYLQR (Figs. 4C and D). In addition, an apparent shift in HPLC-retention time was caused by each PTM, which excludes the possibility that the precursor ions of the modified peptide came from in-source fragmentation of the corresponding unmodified peptide. For example, the retention of LNEEW(+163)SSYLQR increase to 66 from 56 min of its unmodified counterpart. Identification of the modified and unmodified forms of the peptides suggests that the mass shifts are probably caused by novel PTMs rather than by mutations in the gene. To discover potential PTMs as many as possible, especially for the low abundance of PTMs, the potein kinase was over-expressed in this study. Over-expression of certain protein may lead to addition of PTM to itself, especially for transferases such as HATs, which are capable of self-catalyzed modification. [20] Here, our study focused on the non-phosphorylation event of yeast kinases, and therefore, these PTMs are unlikely to be correlated to their autophosphorylation activity. However, it would be difficult to know if some of the PTMs are indirectly caused by the over-expression of the enzymes. Further studies may involve extensive quantification analysis to study the relationship between PTMs and protein over-expression.

Figure 4

Identification of PTM peptides in yeast kinases. (A) MS/MS of an in vivo mass-shift peptide “QMVIQE V₍₊₁₎QDFR”, as well as (B) its corresponding unmodified isoform. (C) MS/MS of an in vivo mass-shift peptide “LNEE₍₊₁₆₃₎WSSYLQR”, (more ...)

4 Concluding remarks

We have used a strategy that combines affinity-based protein purification with MS and unrestrictive PTM analysis to systematically study non-phosphorylation PTMs in yeast kinases. Among 30 kinases, we identified 27 different types of non-phosphorylation PTMs on 53 sites, including new O-acetylation and protein methylation sites. In addition, we identified 13 mass shifts that have not been documented previously, and are likely to represent novel protein modifications.

Protein PTMs serve to orchestrate the activity of diverse cellular processes. Our study demonstrates that the complexity of PTMs on single protein is more diverse han previously expected and it is possible that some of these modifications may participate in the functional regulation and physiological interaction of the kinases in vivo. Crosstalk between different PTMs may comprise another layer of cellular signaling and regulatory networks that fine-tune biological processes under various physiological conditions. It is expected that such phenomena may be more significant in higher eukaryotes. Future studies will focus on determining the nature of the novel PTMs, identification of their upstream modifying enzymes, and characterization of their physiological significance in cells.

Supplementary Material

Supplementary data

Click here to view.^{(4.8M, zip)}

Acknowledgments

This work was supported in part by the National Institutes of Health (to H. Z. and Y. Z.), the National Nature Science Foundation of China (Nos. 90919008, 20975053, U0932001) and National Basic Research Program of China (973 program, No. 2007CB914100).

Footnotes

The authors have declared no conflict of interest.

References

1. Cohen P. The origins of protein phosphorylation. Nature Cell Biol. 2002;4:E127–E130. [PubMed]

2. Manning G, Plowman GD, Hunter T, Sudarsanam S. Evolution of protein kinase signalling from yeast to man. Trends Biochem Sci. 2002;27:514–520. [PubMed]

3. Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S. The protein kinase complement of the human genome. Science. 2002;298:1912–1934. [PubMed]

4. Zhu H, Klemic JF, Chang S, Bertone P, et al. Analysis of Yeast protein kinases using protein chips. Nature Genet. 2000;26:283–289. [PubMed]

5. Ptacek J, Devgan G, Michaud G, Zhu H, Zhu H, et al. Global analysis of protein phosphorylation in yeast. Nature. 2005;438:679–684. [PubMed]

6. Cohen P. Protein kinases – the major drug targets of the twenty-first century. Nat Rev Drug Discov. 2002;1:309–315. [PubMed]

7. Contessa JN, Bhojani MS, Freeze HH, Rehemtulla A, Lawrence TS. Inhibition of N-linked glycosylation disrupts receptor tyrosine kinase signaling in tumor cells. Cancer Res. 2008;68:3803–3809. [PubMed]

8. Wang ML, Panasyuk G, Gwalter J, Nemazanyy I, et al. Regulation of ribosomal protein S6 kinases by ubiquitination. Biochem Biophys Res Communicat. 2008;369:382–387. [PubMed]

9. Nagar B, Hantschel O, Young MA, Scheffzek K, et al. Structural basis for the autoinhibition of c-Abl tyrosine kinase. Cell. 2003;112:859–871. [PubMed]

10. Mukherjee S, Keitany G, Li Y, Wang Y, et al. Yersinia YopJ acetylates and inhibits kinase activation by blocking phosphorylation. Science. 2006;312:1211–1214. [PubMed]

11. Mittal R, Peak-Chew SY, McMahon HT. Acetylation of MEK2 and I kappa B kinase (IKK) activation loop residues by YopJ inhibits signaling. Proc Natl Acad Sci USA. 2006;103:18574–18579. [PubMed]

12. Witze ES, Old WM, Resing KA, Ahn NG. Mapping protein post-translational modifications with mass spectrometry. Nat Methods. 2007;4:798–806. [PubMed]

13. Chen Y, Kwon SW, Kim SC, Zhao YM. Integrated approach for manual evaluation of peptides identified by searching protein sequence databases with tandem mass spectra. J Proteome Res. 2005;4:998–1005. [PubMed]

14. Chen Y, Chen W, Cobb M, Zhao Y. PTMap, a novel algorithm for systematic identification of all the possible post-translational protein modifications in a protein. Proc Natl Acad Sci USA. 2009;106:761–766. [PubMed]

15. Borrelli E, Nestler EJ, Allis CD, Sassone-Corsi P. Decoding the epigenetic language of neuronal plasticity. Neuron. 2008;60:961–974. [PMC free article] [PubMed]

16. Martin C, Zhang Y. Mechanisms of epigenetic inheritance. Curr Opin Cell Biol. 2007;19:266–272. [PubMed]

17. Roelants FM, Torrance PD, Bezman N, Thorner J. Pkh1 and pkh2 differentially phosphorylate and activate ypk1 and ykr2 and define protein kinase modules required for maintenance of cell wall integrity. Mol Biol Cell. 2002;13:3005–3028. [PMC free article] [PubMed]

18. Lee JW, Helmann JD. The PerR transcription factor senses H₂O₂ by metal-catalysed histidine oxidation. Nature. 2006;440:363–367. [PubMed]

19. Wright HT. Nonenzymatic deamidation of asparaginyl and glutaminyl residues in proteins. Crit Rev Biochem Mol Biol. 1991;26:1–52. [PubMed]

20. Karanam B, Jiang LH, Wang L, Kelleher NL, Cole PA. Kinetic and mass spectrometric analysis of p300 histone acetyltransferase domain autoacetylation. J Biol Chem. 2006;281:40292–40301. [PubMed]