Optimizing Thiophosphorylation in the Presence of Competing

Phosphorylation with MALDI-TOF-MS Detection

Laurie L. Parker,*,† Alexander B. Schilling,† Stephen J. Kron,‡ and Stephen B. H. Kent†

Department of Biochemistry and Molecular Biology, University of Chicago, CIS 201, 929 E. 57th Street,Chicago, Illinois 60637, and Center for Molecular Oncology, University of Chicago, Knapp R322,

924 E. 57th Street, Chicago, Illinois 60637

Received May 24, 2005

Abstract: Thiophosphorylation provides a metabolicallystable, chemically reactive phosphorylation analogue foranalyzing the phosphoproteome in vitro and in vivo. Wedeveloped a MALDI-TOF-MS based assay for optimizingthiophosphopeptide production by a kinase even in thepresence of Mg2+ and ATP. We found that Abl kinasethiophosphorylation rates can be ‘rescued’ using Mn2+ inthe presence of Mg2+. Under our ideal conditions, titrationof Mn2+ and ATPγS in the presence of Mg2+ allowedrelatively rapid, highly specific thiophosphorylation by Abltyrosine kinase, both as purified enzyme and in complexcell extracts.

Keywords: thiophosphorylation • MALDI-TOF-MS • kinase activ-ity • Bcr-Abl • phosphoproteomics

Introduction

Phosphate transfer is a major mechanism for cellular signal-ing, and is the subject of a wide body of research in the fieldof proteomics.1 Thiophosphorylation offers a metabolicallystable, chemically reactive phosphorylation analogue for in-vestigating kinase activity in vitro and in vivo.2-6 This techniqueprovides a number of potential advantages for detectingphosphorylation signals in vivo. For example, the phosphataseresistance of thiophosphoproteins7,8 may advantageously allowtransient cellular signaling species to accumulate to detectablelevels, facilitating analysis of low-level signaling events. Thio-phosphate also provides a chemoselectively reactive handle forfunctionalization of thiophosphorylated peptides and proteins,aiding in pull-down of these molecules from complex solu-tions.3,6 For example, recently Shokat and co-workers used acombination of a bulky ATPγS and a mutant kinase for specificthiophosphorylation and labeling of the substrates of thatkinase.5 Additionally, the ability to chemically distinguishbetween existing phosphorylation patterns and recently thio-phosphorylated molecules may lead to more informative time-resolution for studying the kinetics of in vivo signaling.

Thiophosphorylation is typically detected using S35-labeledATPγS or by reacting the phosphorothioate with e.g., fluores-

cent probes.2-4,6,9 The pKa of the phosphorothioate allowsselective labeling at low pH even in the presence of cysteinesand other nucleophilic residues in a peptide or protein.3 Oneof the challenges of using this tool is the poor competition ofATPγS against ATP, with Mg2+ as a kinase cofactor. Unfortu-nately, many kinases (especially tyrosine kinases) show sluggishkinetics with ATPγS: consequently, observing ‘real’ kinaseactivity and generating enough sample for practical use canbe difficult. Rate differences between phosphate and thiophos-phate transfer are typically described as the ‘thio-effect’ bycomparing the kcat values for phosphate vs thiophosphatetransfer.10,11 Catalysis of thiophosphate transfer with Mg2+ ascofactor is often inefficient, resulting in large thio-effects.Alternative divalent metals such as Mn2+ and Co2+ can improvethe efficiency of thiophosphorylation vs phosphorylationsthesemetals are thought to improve the activation of the terminalthiophosphate of ATPγS for transfer to the substrate viachelation with the thiol (which is poorly chelated by the ‘hard’ion Mg2+).10-12 However, kinetics of thiophosphorylation areoften still very slow under these conditions. Additionally, invitro optimization of the thio-effect does not address the issueof achieving thiophosphorylation in vivo, where competitionbetween ATP and ATPγS in the presence of Mg2+ remains amajor challenge.

MALDI-TOF mass spectrometry has been used as a detectionmethod for kinase assays in vitro.13-16 These nonradioactiveassays allow for the direct detection of phosphorylation ofpeptide substrates, and result in qualitative determination ofphosphorylation. However, quantification of product yieldsbased on the ratios of peptide to phospho-peptide peak areasmust be carefully calibrated due to the inherent differences inionization efficiency of phosphorylated, thiophosphorylatedand unmodified peptides.

In this work, we have developed a MALDI-TOF massspectrometry based assay for optimizing the level of thiophos-phopeptide production by a kinase even in the presence ofMg2+ and ATP. Mass spectrometry detection of thiophospho-vs phosphopeptides revealed that for Abl kinase, ATPγS utiliza-tion efficiency could be rescued by adding relatively low levelsof Mn2+ in the presence of Mg2+ and ATP. Titration of Mn2+

and ATPγS allowed relatively rapid, highly specific thiophos-phorylation by Abl tyrosine kinases, both as purified enzymesand in complex K562 human chronic myeloid leukemia cellextracts. The assay for optimal thiophosphorylation conditionswas also demonstrated for PKA, revealing Co2+ as the best

* To whom correspondence should be addressed. Tel: (773) 834-9989.Fax: (773) 702-0439. E-mail: lparker@uchicago.edu.

† Department of Biochemistry and Molecular Biology.‡ Center for Molecular Oncology.

10.1021/pr050150e CCC: $30.25 2005 American Chemical Society Journal of Proteome Research 2005, 4, 1863-1866 1863Published on Web 09/15/2005

divalent cation for ensuring high thiophosphorylation levels inthe presence of ATP for this serine/threonine kinase.

Experimental Section

PKA kinase (active domain) and biotinylated PKA substratepeptide were obtained from Calbiochem. Biotinylated PKAsubstrate peptide was obtained from Calbiochem. The StrepII-tagged Abl substrate peptide was synthesized using Boc SPPSand purified on a Waters PrepLC 4000 system (RP-C18column). The synthesized peptide was characterized using LC/MS (Agilent 1100 series LC/MSD Trap). Assay analyses wereperformed with the Applied Biosystems Voyager 4700 MALDI-TOF/TOF mass spectrometer.

Solution Phase Assay. Assay was performed in 96-well plateformat in a total volume of 30 µL. The kinase reaction bufferwas composed of 50 mM Tris-HCl (pH 7.5), 30 µg/µL BSA and1 mM DTT. Peptide concentration was 100 µM (3 nmol total).Metal cofactors and ATP and/or ATPγS were added as specified.0.34 µg (10 pmol) of Abl (recombinantly produced kinasedomain, 34 kDa) was used for each well. Reactions wereperformed at 30 °C for 30 min then quenched with 10 µL EDTA(0.5 M, pH 8.5). The resulting solutions were transferred to 5µL Streptactin-sepharose beads (IBA) pre-aliquoted into a 96-well filter bottom plate (Millipore) and allowed to equilibrateat RT for 15 min to bind peptide to beads. Beads were thenwashed on a vacuum manifold five times with PBS buffer and5 times with distilled water and allowed to dry partially. Boundpeptides were eluted with 10 µL acetonitrile/H2O (75/25%) andcollected into a 96-well plate via centrifugation (2 min, 1500rpm). The resulting mixture of unphosphorylated, phospho-rylated and thiophosphorylated peptide was mixed with 6 µLmatrix solution (75/25/0.1% acetonitrile/H2O/TFA, 10 mg/mLCHCA) and spotted in triplicate on a stainless steel MALDItarget (Applied Biosystems). Spots were analyzed in linearnegative mode. Data were processed using a macro providedby Melanie Lin of Applied Biosystems to extract mass andintensity (calculated as area under the isotope cluster curve)into tab-delimited text files.

Solid-Phase Assay. Assay was performed in 96-well filterbottom plates. Streptactin-sepharose beads (6 µL) were equili-brated with kinase buffer for 1-2 min and excess buffer wasremoved with a vacuum manifold. Peptide (20 nmol) was addedand bound to the beads for 5 min. Beads were washed andadditional reagents (kinase buffer as above, ATP, ATPγS, metalas indicated in plots) and recombinant Abl (2.1 µg, ∼60 pmol)were added. Reactions were performed at 30 °C for 30 min,and samples were quenched, processed and analyzed as above.

Results and Discussion

We set out to find conditions for using MALDI-TOF toquantify the efficency of thiophosphorylation in the presenceof competing phosphorylation. As a model system we chosethe tyrosine kinase c-Abl,17 for which thiophosphorylation isreported to be undetectable.12,18,19 Addition of ‘thiophilic’divalent cations to improve Abl thiophosphorylation had notbeen reported. Mimicking cellular conditions (necessitating thepresence of Mg2+ and ATP) was desirable for our overall methoddevelopment. Typical thiophosphorylation assays do not ad-dress detection of competing phosphorylation. MALDI-TOF MSoffered direct detection of the ratio of thiophosphorylated tophosphorylated peptide products, allowing us to determineconditions for competitive thiophosphorylation in the presence

of ATP. The Abl kinase domain was used in conjunction withan ‘optimal’ substrate peptide20 (in italics) carrying an N-terminal affinity tag (Strep-tag)21 (WSHPQFEK)EAIYAAPFAKKK(1) to optimize the conditions for thiophosphorylation, andthese conditions were applied to selective thiophosphorylationof the peptide substrate 1 by cell extract. To address theinherent spot-to-spot variability in sample crystallization, allexperiments were performed in duplicate and each was spottedin triplicate (i.e., six measurements per ratio). Linear mode wasused for quantification of peak ratios, because reflectron moderevealed fragmentation of the thiophosphopeptide with lossof HPO2S (see Supporting Information). Negative mode waschosen over positive mode because we find it improves thesensitivity for detection of low-level thiophosphorylation com-pared to positive mode.

It was important to establish calibration curves to quantifythe relationship between phosphorylated and thiophosphor-ylated peptide. This was accomplished by combining knownratios of the two purified materials and observing the resultingsignal intensity ratio in the mass spectrometer (Figure 1). Thio-to phospho-peptide signal ratios were consistently 85% of theactual ratio in the sample, thus this correction factor wasapplied to the ratios observed in the rest of the analyses.

We assessed the question of competition between ATPγS andATP (1:1) in the conversion of 1 using the Abl kinase domain.Mn2+, Co2+, and Ni2+ were chosen as examples of ‘thiophilic’metals previously reported in the literature. In preliminaryexperiments, Cu2+ and Hg2+ were also investigated but resultedin precipitation problems in the assay. Divalent metal ionsMg2+, Mn2+, Co2+, and Ni2+ (5 mM) were used alone and incombination with Mg2+ (10 mM Mg2+:1 mM other M2+). Thereactions were performed in solution as described above.Results are summarized in Table 1. As expected, Mg2+ alonegreatly favored phosphorylation. Mn2+ alone resulted in verylow overall conversion (<5%), but did improve thiophospho-

Figure 1. Calibration curve representing signal intensity ratiosof thiophosphopeptide to phosphopeptide in both linear positiveand negative modes (calculated as area under the curve) com-pared to the ideal ratio.

Table 1. Addition of Mn2+ with Mg2+ Resulted in a 2-3-FoldImprovement in Thiophosphorylation by Abl Kinase DomainCompared to Mg2+ Alone

Optimizing Thiophosphorylation technical notes

1864 Journal of Proteome Research • Vol. 4, No. 5, 2005

rylation (to approximately 1:1). Co2+ and Ni2+ alone alsoresulted in minimal conversion to either product. With Mg2+

in the reaction buffer (alone or in combination with anothermetal), substrate conversion to thio- and/or phospho-peptidewas essentially quantitative in 30 min (with the exception ofMg2+/Ni2+ which gave only ∼30% conversion). The mostpositive results came from the combination of Mg2+/Mn2+:thiophosphorylation was improved 2-fold over Mg2+ alone, andimportantly, conversion was quantitative. This ‘rescue effect’using low levels of ‘thiophilic’ divalent metals in combinationwith Mg2+ is known for ribozymes containing thiophosphatelinkages,22-24 but has not previously been observed for kinases.The seemingly cooperative effect of Mg2+ and Mn2+ in rescuingthiophosphorylation may result from binary metal cofactorinteractions in the kinase active site.25 The optimized conditionsusing 10:1 Mg2+/Mn2+ allowed us to achieve very high levelsof thiophosphorylation in the presence of micromolar levelsof ATP (Figure 2). This should be compared to the reports inthe literature mentioning the failure of previous attempts todetect thiophosphorylation by Abl kinase.10,16,17

Additional characterization of key reactants (Mn2+ andATPγS) and in depth analysis of ATP and ATPγS competitionwas carried out in an on-bead assay. 1 was pre-bound to thebeads and subjected to the kinase reactions varying the levelsof Mn2+ or ATPγS (up to 48 conditions in duplicate for a totalof 96 reactions). The products were analyzed directly by fixinga small amount of beads to the MALDI target with matrix. Thematrix solution containing acetonitrile/H2O/TFA served to elutethe peptide from the beads. These studies showed that as littleas 0.5 mM Mn2+ in the presence of 10 mM Mg2+ promotedselectivity for thiophosphorylation, and that using the opti-mized Mg/Mn combination buffer (10:1), a ratio of 10:1 ATPγSto ATP was sufficient to achieve 1:1 ratios of thiophosphorylatedto phosphorylated product. Importantly, these reagent levelsare within a practical range for use in cell extracts. Wedemonstrated this using the optimized conditions (describedin Figure 2) with K562 chronic myeloid leukemia cell extracts(which overexpress the oncogenic fusion protein Bcr-Abl) tothiophosphorylate 1. Specificity was shown using the Bcr-Ablinhibitor Gleevec (Figure 3). Notably, 10 mM MgCl2 or 1 mMMnCl2 alone did not promote thiophosphorylation in theseconditions (Figure 4).

The high-throughput capabilities of this assay will in thefuture allow rapid, convenient determination of optimal thio-phosphorylation conditions for other kinases. The key advan-tage is the ability to observe both phosphorylation andthiophosphorylation in the same assay. For example, wedemonstrated the use of the ATP:ATPγS competition assayusing the catalytic subunit of the serine/threonine kinase PKA.

Thiophosphorylation of the substrate peptide LCGRTGRRNSI-NH2 (biotinylated on the N-terminus) was carried out accordingto the solution phase protocol described above. Although PKAwas reasonably efficient at thiophosphorylating its substrateeven with only Mg2+ (∼10-15% thiophosphopeptide producedat 1:1 ATP/ATPγS), a mild improvement in thiophosphorylationwas gained (15-20% thiophosphopeptide produced) by usingCo2+ in the reaction buffer (Table 2)showever, the reactionproceeded efficiently with Co2+ alone in the buffer and additionof Mg2+ only modestly improved the rate. We are currentlyscreening other kinases in this assay to determine if thethiophosphorylation ‘rescue’ observed for Abl kinase withMg2+/Mn2+ is a general effect for tyrosine kinases or if it isspecific to some kinases and not others. Our primary goal isto use these optimized conditions for Abl kinase to investigatethiophosphorylation of endogenous substrates by Bcr-Abl inK562 cell extracts, in response to changes in cellular states, withhigh throughput detection by MALDI-TOF mass spectrometry.The assay described here should facilitate access to thiophos-phorylation as a tool for phosphoproteomics in general, butespecially for systems such as ours that require optimizationof not only the ‘thiophilic’ cation required for the kinase ofinterest, but also ATPγS competition in the presence of ATPand Mg2+, a feature which will prove crucial for makingthiophosphorylation practical for in vivo use.

Acknowledgment. We thank Greg Flugum and DonWolfegher of the proteomics core facility for MALDI-MS

Figure 2. Optimized thiophosphorylation conditions: ATP (2.5µM ATPγS (1 mM), MgCl2 (10 mM), MnCl2 (1 mM). A: unphos-phorylated; B: phosphorylated; C: thiophosphorylated.

Figure 3. Thiophosphorylation of 1 by Bcr-Abl in K562 cellextracts (1) with and (2) without inhibitor. A: unphosphorylated;B: phosphorylated; C: thiophosphorylated.

Figure 4. Thiophosphorylation of 1 by Bcr-Abl in K562 cellextracts with Mg2+ and Mn2+ alone. No thiophosphorylation isobserved with Mg2+ alone, and no conversion is observed forMn2+ alone. A: unphosphorylated; B: phosphorylated; C: thio-phosphorylated.

Table 2. Co2+ Provided a Slight Increase inThiophosphorylation by PKA, However Conversion WasQuantitative with and without Mg2+

technical notes Parker et al.

Journal of Proteome Research • Vol. 4, No. 5, 2005 1865

training and technical support, and Evan Nair-Gill and VivianTien for help with initial assay development. Abl kinase domainwas a kind gift from Markus Seeliger and John Kuriyan (UC-Berkeley). Financial support provided by NIH CardiovascularPathophysiology training grant T32 HL0723, the University ofChicago NSF MRSEC and R33 CA103235. S.J.K. is a Scholar ofthe Leukemia and Lymphoma Society.

Supporting Information Available: Additional graphsfrom sample analysis, peptide synthesis/characterization de-tails. This material is available free of charge via the Internetat http://www.pubs.acs.org.

References

(1) Hunter, T. Signaling-2000 and beyond. Cell 2000, 100 (1), 113-127.

(2) Jeong, S.; Nikiforov, T. T. Kinase assay based on thiophospho-rylation and biotinylation. Biotechniques 1999, 27 (6), 1232-1238.

(3) Kwon, S. W.; Kim, S. C.; Jaunbergs, J.; Falck, J. R.; Zhao, Y. Selectiveenrichment of thiophosphorylated polypeptides as a tool for theanalysis of protein phosphorylation. Mol. Cell Proteomics 2003,2 (4), 242-247.

(4) Sun, I. Y.; Allfrey, V. G. In vivo thiophosphorylation of chromo-somal proteins. Recovery and analysis of HeLa histones andderivative phosphopeptides. J. Biol. Chem. 1982, 257 (3), 1347-1353.

(5) Allen, J. J.; Lazerwith, S. E.; Shokat, K. M. Bio-orthogonal affinitypurification of direct kinase substrates. J. Am. Chem. Soc. 2005,127 (15), 5288-5289.

(6) Nikiforov, T. T.; Simeonov, A. M. Application of fluorescencepolarization to enzyme assays and single nucleotide polymor-phism genotyping: Some recent developments. Comb. Chem.High T. Scr. 2003, 6 (3), 201-212.

(7) Hiriyanna, K. T.; Baedke, D.; Baek, K. H.; Forney, B. A.; Kordiyak,G.; Ingebritsen, T. S. Thiophosphorylated substrate analogues arepotent active site-directed inhibitors of protein-tyrosine phos-phatases. Anal. Biochem. 1994, 223 (1), 51-58.

(8) Cassel, D.; Glaser, L. Resistance to phosphatase of thiophospho-rylated epidermal growth factor receptor in A431 membranes.Proc. Natl. Acad. Sci. U.S.A. 1982, 79 (7), 2231-2235.

(9) Cassidy, P.; Hoar, P. E.; Kerrick, W. G. Irreversible thiophospho-rylation and activation of tension in functionally skinned rabbitileum strips by [35S]ATP gamma S. J. Biol. Chem. 1979, 254 (21),11148-11153.

(10) Grace, M. R.; Walsh, C. T.; Cole, P. A. Divalent ion effects andinsights into the catalytic mechanism of protein tyrosine kinaseCsk. Biochemistry 1997, 36 (7), 1874-1881.

(11) Sondhi, D.; Xu, W.; Songyang, Z.; Eck, M. J.; Cole, P. A. Peptideand protein phosphorylation by protein tyrosine kinase Csk:

insights into specificity and mechanism. Biochemistry 1998, 37(1), 165-172.

(12) Zou, K.; Miller, W. T.; Givens, R. S.; Bayley, H., Caged Thiophos-photyrosine Peptides. Angew. Chem. Int. Ed. Engl. 2001, 40 (16),3049-3051.

(13) Min, D. H.; Su, J.; Mrksich, M. Profiling kinase activities by usinga peptide chip and mass spectrometry. Angew Chem. Int. Ed. Engl.2004, 43 (44), 5973-5977.

(14) Zeller, M.; Essmann, F.; Janicke, R. U.; Schulze-Osthoff, K.; Konig,S. A rapid nonradioactive peptide phosphorylation assay. J. Exp.Ther. Oncol. 2003, 3 (2), 59-61.

(15) Houseman, B. T.; Huh, J. H.; Kron, S. J.; Mrksich, M. Peptide chipsfor the quantitative evaluation of protein kinase activity. Nat.Biotechnol. 2002, 20 (3), 270-274.

(16) Matsumoto, H.; Kahn, E. S.; Komori, N. Nonradioactive phos-phopeptide assay by matrix-assisted laser desorption ionizationtime-of-flight mass spectrometry: Application to calcium/cal-modulin-dependent protein kinase II. Anal. Biochem. 1998, 260(2), 188-194.

(17) Pendergast, A. M. The Abl family kinases: Mechanisms ofregulation and signaling. Adv. Cancer Res. 2002, 85, 51-100.

(18) Martin, K.; Steinberg, T. H.; Cooley, L. A.; Gee, K. R.; Beechem, J.M.; Patton, W. F. Quantitative analysis of protein phosphorylationstatus and protein kinase activity on microarrays using a novelfluorescent phosphorylation sensor dye. Proteomics 2003, 3 (7),1244-1255.

(19) Martin, K.; Steinberg, T. H.; Goodman, T.; Schulenberg, B.; Kilgore,J. A.; Gee, K. R.; Beechem, J. M.; Patton, W. F. Strategies and solid-phase formats for the analysis of protein and peptide phospho-rylation employing a novel fluorescent phosphorylation sensordye. Comb. Chem. High T. Scr. 2003, 6 (4), 331-339.

(20) Songyang, Z.; Carraway, K. L., 3rd; Eck, M. J.; Harrison, S. C.;Feldman, R. A.; Mohammadi, M.; Schlessinger, J.; Hubbard, S.R.; Smith, D. P.; Eng, C. Catalytic specificity of protein-tyrosinekinases is critical for selective signaling. Nature 1995, 373 (6514),536-539.

(21) Skerra, A.; Schmidt, T. G. Use of the Strep-Tag and streptavidinfor detection and purification of recombinant proteins. MethodsEnzymol. 2000, 326, 271-304.

(22) Shan, S.; Kravchuk, A. V.; Piccirilli, J. A.; Herschlag, D. Definingthe catalytic metal ion interactions in the Tetrahymena ribozymereaction. Biochemistry 2001, 40 (17), 5161-5171.

(23) Basu, S.; Strobel, S. A. Thiophilic metal ion rescue of phospho-rothioate interference within the Tetrahymena ribozyme P4-P6domain. RNA 1999, 5 (11), 1399-1407.

(24) Cunningham, L. A.; Li, J.; Lu, Y. Spectroscopic evidence for inner-sphere coordination of metal ions to the active site of a ham-merhead ribozyme. J. Am. Chem. Soc. 1998, 120 (18), 4518-4519.

(25) Cheng, Y.; Zhang, Y.; McCammon, J. A. How Does the cAMP-Dependent Protein Kinase Catalyze the Phosphorylation Reac-tion: An ab Initio QM/MM Study. J. Am. Chem. Soc. 2005, 127,1553-1562.

PR050150E

Optimizing Thiophosphorylation technical notes

1866 Journal of Proteome Research • Vol. 4, No. 5, 2005