Accepted Manuscript Identification of peptidic substrates for the human kinase Myt1 using peptide microarrays Alexander Rohe, Charlott Platzer, Antonia Masch, Sandra Greiner, Claudia Henze, Christian Ihling, Frank Erdmann, Mike Schutkowski, Wolfgang Sippl, Matthias Schmidt PII: DOI: Reference:

S0968-0896(15)00421-6 http://dx.doi.org/10.1016/j.bmc.2015.05.021 BMC 12317

To appear in:

Bioorganic & Medicinal Chemistry

Received Date: Revised Date: Accepted Date:

7 April 2015 11 May 2015 12 May 2015

Please cite this article as: Rohe, A., Platzer, C., Masch, A., Greiner, S., Henze, C., Ihling, C., Erdmann, F., Schutkowski, M., Sippl, W., Schmidt, M., Identification of peptidic substrates for the human kinase Myt1 using peptide microarrays, Bioorganic & Medicinal Chemistry (2015), doi: http://dx.doi.org/10.1016/j.bmc.2015.05.021

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Title: Identification of peptidic substrates for the human kinase Myt1 using peptide microarrays Authors: Alexander Rohea,1, Charlott Platzera, Antonia Maschb, Sandra Greinera, Claudia Henzea, Christian Ihlingc, Frank Erdmannd, Mike Schutkowskib, Wolfgang Sippla, Matthias Schmidta,*

a

Institute of Pharmacy, Department of Medicinal Chemistry, Martin-Luther-University Halle-

Wittenberg, W.-Langenbeck-Str. 4, 06120 Halle, Germany b

Institute of Biochemistry and Biotechnology, Department of Enzymology, Martin-Luther-

University Halle-Wittenberg, Kurt-Mothes-Str. 3, 06120 Halle, Germany c.

Institute of Pharmacy, Department of Bioanalytics, Martin-Luther-University Halle-

Wittenberg, W.-Langenbeck-Str. 4, 06120 Halle, Germany d.

Institute of Pharmacy, Department of Pharmacology, Martin-Luther-University Halle-

Wittenberg, W.-Langenbeck-Str. 4, 06120 Halle, Germany

* to whom correspondence should be addressed: Dr. Matthias Schmidt, W.-Langenbeck-Str. 4, 06120 Halle, Germany Phone: +49 345 5525188 Fax: +49 345 55 27355 E-Mail: [email protected]

1

Current Address: Central Institute of the Bundeswehr Medical Service Munich, Ingolstädter Landstraße 102, 85748 Garching, Germany

Abstract: Myt1 kinase is a member of the Wee-kinase family involved in G2/M checkpoint regulation of the cell cycle. So far, no peptide substrate suitable for activity-based screening has been reported, hampering systematic development of Myt1 kinase inhibitors. Myt1 inhibitors had to be identified by using either binding assays or activity assays with expensive proteinous substrates. Here, a peptide microarray approach was used to identify peptidic Myt1 substrates. Wee1 kinase was profiled for comparison using the same technology. Myt1 hits from peptide microarray experiments were verified in solution by a fluorescence polarization assay and several peptide substrates derived from cellular proteins were identified. Subsequently, phosphorylation site determination was carried out by MS fragmentation studies and identified substrates were validated by kinase inhibitor profiling.

Keywords: Peptide microarray, Myt1 kinase, kinase substrate profiling, fluorescence polarization assay

Abbreviations: Cdk, cyclin-dependent kinase; FAM, carboxyfluorescein; FP, fluorescence polarization; FPIA, fluorescence polarization immunoassay; QQ-plot, quantile-quantile-plot; TBS, Tris-buffered saline; TFA, trifluoroacetic acid.

1. Introduction In the cell cycle, the decision to enter mitosis is mainly influenced by Cyclin-dependent kinase 1 (Cdk1). The activation of Cdk1 requires the binding of cyclin B and a phosphorylation at Thr161.1-2 The active protein kinase complex composed of Cdk1 and cyclin B (Cdk1/CycB) affects many cellular proteins and is, in turn, regulated by various feedback mechanisms,1 particularly inhibitory phosphorylations at Thr14 and Tyr15 of Cdk1.23

The kinases responsible for these modifications are members of the Wee kinase family,

which consists of Wee1, Myt1 and the less important Wee1B.4-5 While Wee1 exclusively mediates phosphorylation of Tyr15, Myt1 is dual-specific for Tyr15 as well as Thr14.4 For Wee1, a lot of biological data as well as published assays and inhibitors are available. In contrast, although it shows great potential as a drug target in anti-cancer therapy,6 Myt1 is less extensively studied. Because Myt1 is a membrane-associated enzyme that proved to be restrictive in terms of substrate acceptance, functional assays are difficult to perform,7 and binding assays were preferred for inhibitor screening purposes.8-9 Typically, functional Myt1 studies depend on whole proteins as substrates.10-11 We are interested in finding peptidic Myt1 substrates to expand the possibilities in functional approaches for kinase inhibitor studies. As known from the literature, Cdk1-derived peptides are not phosphorylated by Myt1,7,10 leading to the conclusion that simply deducing a sequence from the native substrate, omitting the 3D structure of the protein, is not necessarily successful. Other peptide sequences ‒though not associated with the native target protein ‒ may mimic the native substrate and can, therefore, be efficiently phosphorylated. Peptide microarrays were used to detect kinase substrates,12-14 to analyze subsite specificities,15-16 to identify phosphorylation sites within a substrate protein,14,17-18 to generate specific substrates,19 to profile kinase reactivities in cell lysates,20 and to uncover priming phosphorylations.14 Particularly, a collection of known phosphorylation-sites on peptide microarrays proved to be a useful tool for identification of peptidic kinase substrates.14,19,21 Such substrates are a prerequisite for high-throughput screenings in drug discovery projects.

In addition, the information about substrate subsite specificity offers valuable hints for integration of kinases into their biological environment and signal transduction network by enabling prediction of in vivo downstream targets.14,16,22 As a starting point for the discovery of Myt1 kinase substrates, we used high-content peptide microarrays displaying 2304 peptides derived from human phosphorylation sites.21 The respective 13meric peptides with the phosphorylation site in the middle position were extracted from the databases SwissProt and Phosphobase.23-24 For more comprehensive insights into the Wee kinase family, Myt1 and Wee1 were investigated in parallel. Subsequently, the Myt1 substrates detected on the peptide microarray were subjected to further experiments in solution phase to validate the respective peptide substrates in the nonimmobilized form.

2. Materials and Methods 2.1 Reagents All

reagents

were

purchased

from

Sigma

unless

stated

otherwise.

Peptides

(GTDEGIYDVPLLG, GMSRDIYSTDYYR, SSSIDEYFSEQPL, DGHEYIYVDPMQL) were purchased from Thermo as unmodified peptides with purities >95%. The fluorescent probe (6-FAM)KI(pY)VV was from IKFZ Leipzig (Germany). Myt1 in full-length was expressed and purified as described in Rohe et al.7 Full-length Wee1 was obtained as GST-fusion protein from Invitrogen. Complete EDTA-free protease inhibitor cocktail was from Roche (Basel, Switzerland).

2.2 Preparation of microarrays All peptides carried a linker at their N-terminus (N-(3-(2-(2-(3- amino-propoxy)-ethoxy)ethoxy)-propyl)-succinamic acid). Peptide derivatives were synthesized via glycine ester

linkage on cellulose membranes in a parallel manner using SPOT synthesis technology according to Wenschuh et al. acetylated

using

acetic

25-26

After each coupling, remaining amino functions were

anhydride

in

dimethyl

formamide

in

the

presence

of

diisopropylethylamine to prevent deletion sequences. After removal of the last Fmocprotecting group with 20% piperidine in dimethyl formamide, coupling with Boc-protected anthranilic acid-transformed full-length peptides with free N-terminus, but not N-terminally acetylated, truncated side products from incomplete coupling, into 2-aminobenzoylderivatives. After subsequent trifluoroacetic acid-mediated side chain deprotection, the cellulose bound peptide esters were transferred into 96 well microtiter filtration plates (Millipore, Bedford, Massachusetts, USA) and treated with 200 µL of aqueous triethylamine (2.5%) in order to cleave the peptides off the cellulose. Peptide-containing triethylamine solution was filtered off, quality controlled by LC-MS, and solvent was removed by evaporation under reduced pressure. Resulting peptide derivatives (50 nmol) were redissolved in 25 µL of printing solution (70% DMSO, 25% 0.2 M sodium acetate pH 4.5, 5% glycerol) and transferred into 384-well microtiterplates. Peptide derivatives were deposited onto epoxy-functionalized glass slides (PolyAn GmbH, Berlin, 3D-Epoxy, 25x75x1 mm) using a contact printer equipped with 16 SMP2 pins (Telechem/ArrayIt). Our buffer conditions resulted in selective covalent bond formation between epoxy-functions and amino groups of 2-amino-benzoyl-derivatives. Printed peptide microarrays were kept at room temperature for 5 hours. Each microarray with dried deposited spots was analysed using Axon 4000B microarray scanner. Microarrays that passed this quality control were quenched for 1 hour with sodium citrate buffered 1% BSA solution at 42°C, washed extensively with water followed by ethanol, resulting in ‘purified peptide’ spots, essentially free of deletion sequences (due to acetylation steps during synthesis) and truncated sequences (due to chemoselective immobilization). Printed microarrays were dried using a microarray centrifuge and stored at 4 °C.

2.3 Microarray assays

Two microarray slides were positioned face-to-face, separated by two spacers (0.3 mm) as described in Thiele et al.18 350 µl reaction mixture containing kinase (80 nM) in sterile filtered standard kinase buffer (50 mM Tris-HCl pH 7.5, 40 mM NaCl, 10 mM MgCl2, 1 mM DTT, 0.04% Triton X-100, Complete EDTA-free) together with 250 µM ATP was transferred bubble-free into the resulting reaction space between the two microarrays. Subsequently, the microarray sandwich was incubated at 30°C for 2 h in a humid chamber. After incubation, the microarrays were washed five times with 50 ml TBS (25 mM Tris-HCl pH 7.2, 137 mM NaCl) containing 0.1% Triton X-100 for 5 min followed by washing steps with deionized water. Finally, the microarray slides were spun dry at room temperature. For antibody detection, Tecan's Hyb Station HS 400 was used in a semi-automatic procedure. All steps were carried out at 25°C and the protocol was as follows: After 2 washing steps with TBS-T (TBS + 0.1% Tween-20; 2 min wash, 2 min soak), a wash and soak step with TBS was performed. Subsequently, the primary antibody (Anti-pTyr antibody P-100, Cell Signaling Technology) at 10 µg/ml in TBS-T containing 3% BSA was injected and hybridized for 1 h. After 5 washing steps with TBS-T and another step with TBS, the secondary antibody (Goat anti-mouse IgG #35515 as Dylight 649 conjugate (Pierce) at a concentration of 1 µg/ml in TBS-T containing 3% BSA) was injected and hybridized for 30 min. After five additional washing steps (TBS-T), the microarrays were rinsed with TBS and deionized water. Subsequent drying under a stream of gaseous nitrogen yielded microarray slides ready for detection. For negative controls, the microarrays were subjected to the same procedure without addition of kinase. The peptide microarrays were scanned at 635 nm (red channel) and 532 nm (green channel) using GenePix 4000B (Molecular Devices) at a PMT gain of 600 and 100% laser power. The response from the red channel was used to quantify the response, while the green channel was used to align the array grid in analysis. Images were analyzed utilizing the spot recognition software package GenePix Pro 7.1. Quality control settings were as described before.27 Both median foreground and median background intensities of individual peptide spots from the GenePix result files were used for subsequent analysis.

In the case of radioisotopic readout, 400 µl of the reaction mixture containing kinase in assay 33

buffer together with 10 µM ATP and 18.5 MBq [γ- P]-ATP were transferred into the reaction space between the two microarray slides. The kinase reaction proceeded for 4 hours at ambient temperature. Subsequently, the microarrays were washed 3 times with TBS (50 mM Tris-HCl, pH 7.5, 150 mM NaCl), 2 times with phosphoric acid, pH 2.0, 2 times with deionized water and 2 times with methanol. All washing steps were performed for 5 min each and finalized by air drying of the microarrays. The incorporated radioactivity was detected by exposure of the microarrays for 2-8 d to imaging plates (Fuji BAS-MS) followed by readout with a phosphor imager (FujiFilm FLA-7000). A control experiment with radioisotopically labelled ATP in buffer but without kinase did not yield any signal after phosphor imaging, demonstrating that there is no unspecific binding of ATP/[γ-33P]-ATP to the immobilized peptides.

2.4 Peptide synthesis Peptides were synthesized on Fmoc-Rink MBHA basically as described in Coin et al. following a standard Fmoc/tBu based strategy with acetic anhydride capping steps, resulting in N-terminally acetylated peptides.28 After final cleavage of side-chain protecting groups and cleavage from the resin by trifluoroacetic acid (TFA) containing 2% water, the TFA was removed using a rotary evaporator. The crude product was precipitated by addition of cold diethyl ether, carefully filtered off and washed with cold diethyl ether. Solvent residues were removed in vacuo in an exsiccator. Purification was carried out by preparative RP-HPLC (Shimadzu, Kyoto, Japan; LC-10AD, SIL-HT auto sampler). A prepacked 7.8 x 300 mm XTerra RP-18 (7 µm) column from Waters (Milford, MA, USA) was used and the UV-Vis detector SPD-M10A VP PDA was set to 220 nm. A shallow gradient over 45 min from water containing 0.1% TFA to 95% acetonitrile was run and the eluate collected and fractionated. The determination of the purities was performed by analytical HPLC (Shimadzu, Kyoto,

Japan; LC-10AD, SIL-HT auto sampler) with an XTerra RP18 column (3.5 µm 3.9 x 100 mm) from Waters (Milford, MA, USA). Spectra and chromatograms are given in the supplement.

2.5 Determination of peptide concentrations Concentrations of all non-phospho peptide solutions used in fluorescence assays were determined

by

measuring absorbance at 280

nm

by

using

a

NanoVue

Plus

spectrophotometer (GE Healthcare). For each peptide, molar absorptivity (ε280) was calculated according to the Edelhoch method,29 taking into account contributions from tyrosine and tryptophane present in the primary structure. Prior to calculation, nonprotein absorbance at 280 nm was subtracted as described by Mach et al.30 Tyrosine and tryptophane absorptivities amount to 1280 and 5690 M-1*cm-1, respectively. This procedure is affected by an estimated error of 5% at most.31 Phospho peptides were totally hydrolyzed by hydrochloric acid and the free amino acids were determined by ninhydrin-based assay as reported before.32

2.6 Solution phase kinase reactions Kinase reactions were carried out in 96-well half area non binding surface plates (Corning) in a final assay volume of 40 µl. Up to 200 µM peptide substrate was dissolved in the same buffer as used in microarray assays. ATP was present at 500 µM. Kinase reactions were started by the addition of kinase (final concentration 40 nM), proceeded for 2 h at 30°C (gentle shaking) and were terminated by addition of 10 µl stop solution (100 mM EDTA, 50 mM Tris-HCl, pH 7.5). Specific controls containing only substrate plus ATP as well as controls containing only kinase plus ATP were carried along on each plate. For inhibitor studies, all dilutions were made from 10 mM stock solutions in DMSO and had a final assay concentration of 5 µM with 0.5% DMSO in the assay buffer. PD173952 was purchased from Sigma, sunitinib was from Pfizer, CEP-701 (Lestaurtinib) and Dasatinib were from LC laboratories.

Detection of phosphotyrosine-containing peptides was performed by adding 10 µl of a solution

containing anti-pY antibody (pTyr-P-100, Cell

signaling #9411)

and (6-

FAM)KI(pY)VV to a final concentration of 45 nM and 2.5 nM, respectively. After equilibration on a shaker for 1 h at room temperature, fluorescence polarization was measured by a PolarStar OMEGA plate reader (BMG) at 485 nm (Ex) and 520 nm (Em). Suitable controls were carried along to allow for specific background correction. For control and calibration purposes, wells containing fluorescein in buffer (pH 7.5) were also carried along. Fluorescence intensity of all wells was monitored for reasons of quality control.

2.7 Mass spectrometry For determination of phosphorylation sites, after solution phase kinase reactions according to section 2.6, peptides were separated on an Ultimate 3000 RSLCnano-HPLC system (Thermo Fisher Scientific) using reversed phase C18 columns (trap column: Acclaim PepMap C18, 100 µm x 10 mm, 3 µm, 100 Å, separation column: EASY-Spray column, Acclaim PepMap C18, 75 µm x 250 mm, 2 µm, 100 Å, Thermo Fisher Scientific). After washing the peptides on the trap column for 15 min with 0.1% TFA at a flow rate of 20 µL/min, peptides were eluted and separated using 120 min gradients from 1 to 40% solvent B (solvent A: 0.1% formic acid in water, solvent B: 0.08% formic acid in acetonitrile) at a flow rate of 300 nL/min. The nano-HPLC system was directly coupled to the nano-ESI source (EASY-Spray source, Thermo Fisher Scientific) of an Orbitrap Fusion Tribrid mass spectrometer (Thermo Fisher Scientific). MS and MS/MS data were acquired in datadependent MS/MS mode using targeted inclusion: Each high-resolution full scan (m/z 350 to 1850; R = 120000 at m/z 200) in the orbitrap analyzer was followed by CID, HCD and ETD product ion scans (quadrupole isolation window 2.0 Th) on the most intense signals of a targeted inclusion mass list containing the masses of the singly to quadruply protonated ions of the singly phosphorylated peptides. ETD fragmentation was triggered for charge states 24. MS/MS data were acquired in the orbitrap analyzer (R = 15000 at m/z 200) as well as in the linear ion trap. Data acquisition was controlled via XCalibur 3.0 (Thermo Fisher Scientific)

in combination with DCMS link 2.0 (Dionex). Identification of phosphorylation sites was done by Mascot search against the peptide sequences using the Proteome Discoverer 1.4 and phosphoRS 3.0. Phosphorylated peptides show a mass difference of 80 u compared to the respective unmodified peptides, for more details see EYRICH et al.33 Results were validated by manual spectra inspection. Mass spectra are given in the supplementary material.

3. Results and Discussion 3.1 Peptide Microarray Studies A high content peptide microarray approach as described before was used for Myt1 and Wee1 profiling.21 This peptide microarray displays 2304 peptides derived from human phosphorylation sites in three identical subarrays on two chemically modified microscope glass slides. Each peptide microarray was treated with full-length Myt1 and full-length Wee1 in the presence of ATP and the resulting phosphorylation reactions were visualized by means of anti phospho-tyrosine antibodies and fluorescent dye-labeled secondary antibodies. Also, negative controls without kinase were performed. Fluorescence readout is favorable over a radioisotopic readout, since the resolution of phosphor imagers (25-50 µm pixel size) is dramatically lower than for microarray fluorescence scanners (down to 2 µm pixel size). Notwithstanding, a radioisotopic readout procedure was used to verify the results from the pY-antibody based experiments. After fluorescence readout, the index I was calculated as a measure of response according to NAHTMAN et al.34 The index I is defined as follows, with ‫ݔ‬෤ being the median of foreground intensity (actual spot) or background intensity (area around the actual spot having 3fold the diameter of the respective spot), as indicated: ‫݃݋݈ = ܫ‬ଶ ൬

௫෤೑೚ೝ೐೒ೝ೚ೠ೙೏

௫෤್ೌ೎ೖ೒ೝ೚ೠ೙೏

൰ (Eq. 1)

As a measure of response, the ratiometric index is well defined even if the background signal is higher than the foreground signal (e.g. for empty spots as negative controls). Importantly, it provides the advantage of introducing homoscedasticity, facilitating statistical analysis. However, the importance of the visual inspection of every single spot cannot be overemphasized for such a ratiometric measure and has to be carried out carefully.27 For the identification of false positives, control experiments without kinase were run. A graphical display index I vs. ݈‫݃݋‬ଶ ൫‫ݔ‬෤௕௔௖௞௚௥௢௨௡ௗ ൯ was used for data evaluation. Using this methodology nine reproducible false-positive peptides with high response indices could be identified. With respect to NGO et al.,27 these peptides served as positive controls in further experiments and, moreover, were used to normalize readouts to allow for a better comparability between peptide microarrays. The two different microarrays were assessed individually and only the final results are presented together. After the kinase experiments, the responses were normalized using a two-step methodology. Firstly, the global median normalization was conducted.35 Secondly, a positive control-guided normalization as described by NGO et al. was carried out.27 Aligned dot plots of the normalized data for Myt1 and Wee1 are shown in Fig. 1. Both kinases were used in concentrations of 80 nM under identical conditions. The differences in range between Myt1 and Wee1 may arise from differences between the specific kinase activities of both preparations. To identify promising hits, quantile-quantileplots (QQ-plots, supplementary Fig. S1 and S2) were used as a simple means of cluster analysis for the two individual microarrays. Gaps at higher responses were identified visually and any index beyond the gap was considered to be a hit. The resulting hits were individually processed. Any peptides that appeared at least once in one of the antibody control experiments were removed, as well as peptides that contained a tyrosine residue only in terminal position or not at all in their sequences. Finally, a visual inspection of each individual spot in the scanned images was carried out to exclude false positives gained by artifacts. Altogether, 21 hits could be

identified for Wee1, whereas Myt1 yielded 11 substrate peptides. The overlap of both sets is represented by four peptides. A respective Venn-plot is shown in Fig. 2. 33

This result was confirmed by a radioisotopic experiment using [γ- P]-ATP and subsequent detection of incorporated

33

P. All of the peptides identified in the fluorescence-based readout

were positive in this radioisotopic approach (data not shown). Interestingly, the native phosphorylation site of both kinases, an N-terminal region of Cdk1, was identified as a substrate for Wee1 only. This matches former reports wherein Wee1 but not Myt1 accepted peptidic substrates derived from this site (Cdk18-20).7,10 Complete information on the identified substrates can be found in Table 1 (Wee1) and Table 2 (Myt1). Peptide names consist of an abbreviation of the protein the sequence was derived from, together with the position of the respective sequence in the corresponding protein. The top 50 peptide substrates for both kinases are provided in the supplementary material. The identified Myt1 substrate peptides were synthesized for solution phase verification of the peptide microarray results. However, to be able to screen these substrates for acceptance by Myt1, a suitable assay was required.

3.2 A Homogenous FP Based pTyr Assay With respect to our expertise gained in former studies,7-8 it was decided to use a homogenous fluorescence polarization based immunoassay (FPIA). Because of its satisfying performance in the peptide microarray studies, the same antibody was used in the FPIA to detect phosphorylated tyrosine residues. As demonstrated with a phosphopeptide microarray approach phosphotyrosine is the key motif for recognition by this antibody with minimal subsite specificity caused by surrounding sequences.36 Therefore, the FPIA should be suitable for detecting any phosphorylated tyrosine residues.

No reliable peptidic Myt1 substrates have been reported so far which led to the expectation that sensitivity could be another important issue in this case. Therefore, a competitive FPIA was used instead of a direct one. A suitable fluorescent probe was needed for the FPIA having a molecular weight as low as possible to improve the dynamic range of the final assay. The free fluorescent probe is determined to show a low fluorescence polarization (FP) due to fast rotation (Brownian motion). Binding of the fluorescent probe to the anti-pY-antibody dramatically decreases the mobility of the probe, leading to an increase in FP. If a mixture of antibody-probe complex (high FP) is added to a kinase in the presence of ATP, any phospho-tyrosine residues formed will displace the fluorescent probe from the anti-pY-antibody, resulting in a lowered FP. Due to the high molecular mass of the antibody, the mass of the probe does not significantly affect the top plateau of a displacement curve. The bottom plateau, however, depends totally on the intrinsic anisotropy of the probe. A heavy probe (large substrate peptide or protein consisting of many amino acids) will yield a narrow dynamic range, while a low molecular weight probe (containing only few amino acids) will yield an increased dynamic range, thereby improving the overall assay performance. As a compromise between molecular weight and providing a suitable antibody epitope, a labeled pentapeptide (6-FAM)KI(pY)VV was used as a probe. The dissociation constant Kd towards the antibody was determined to 41 nM by nonlinear regression. The maximum dynamic range (Δrmax) was 0.15 (anisotropy). For more information on the actual assay development, the reader be referred to the supplement. Briefly, the final assay conditions were as follows: Kinase reactions were carried out in 40 µl assay volume and the reactions were terminated by addition of 10 µl stop solution containing EDTA. After addition of 10 µl detection solution containing antibody and fluorescent probe, measurements were taken when equilibrium was reached. Final concentrations of antibody and fluorescent probe were 45 nM and 2.5 nM, respectively. The dynamic range of the assay was 0.08 (anisotropy) which is equal to 123 mP (polarization).

Several different peptides containing a phosphotyrosine were used to check the assay for its functionality. All test peptides were able to displace the fluorescent probe and could be detected in a submicromolar range with good certainty, allowing application of this assay to the evaluation of peptide substrates.

3.3 Myt1 Solution Phase Verification of Peptidic Substrates The 11 Myt1 substrates derived from peptide microarray studies were synthesized and initially tested in concentrations up to 200 µM in kinase experiments containing 500 µM ATP in standard kinase buffer. An important aspect to be considered is the limited specificity of the FPIA detection principle. Any phosphotyrosine residue generated will competitively displace

the

fluorescent

probe.

Many

posttranslational

modifications

including

phosphorylations are reported for Myt1.37-40 One of these putative modifications is directed at a tyrosine residue (Tyr97) which might interfere with the assay by narrowing the assay window.37 To circumvent a large impact of such tyrosine phosphorylations within the Myt1 kinase, it was decided to run the experiments at a kinase concentration of 40 nM. Results of the FPIA with the 11 different Myt1 peptide substrates derived from the peptide microarray approach are summarized in Table 3. All Myt1 peptide substrates led to a detectable displacement of the fluorescent probe. These results could be verified by qualitative mass spectrometry (MS spectra are available as supplementary data) for 10 out of the 11 identified substrates. Substrate SC4A381-393 showed incoherent results and was therefore excluded from further experiments. Six out of the 10 remaining substrates led to significant displacement (>50%) of the fluorescent probe. Substrates with displacement >50% that did not encounter any solubility issues were tested for substrate dose-dependency. These experiments were performed at shorter incubation time to make sure that the product formation was still in the linear phase. The resulting fits enabled the deviation of apparent Km values (Table 3). Negative control experiments without kinase showed no effect of the substrate peptides on the binding of the fluorescent probe to

the anti-pY-antibody. The deduced Km,app values have to be handled cautiously, because the affinity of the anti-pY-antibody is assumed to be equal for all phosphotyrosine-containing peptides. As expected, substrates with high indexes in microarray experiments were generally more likely to perform well in solution phase. The peptide EFS247-259 was the highest response substrate in microarray as well as solution phase experiments. Qualitatively, all tested peptides appear to be accepted as Myt1 substrates in a dosedependent manner.

3.4 Substrate validation We used mass spectrometry to determine the phosphoacceptor site for the best substrate peptides. After in vitro kinase reaction, the respective solutions were subjected to MS studies. All peptides showed phosphorylations sites (marked with an asterisk) on the expected

tyrosine

positions:

PDGFRB575-587

(DGHEYIY*VDPMQL)

and

(DGHEY*YIVDPMQL), RET1090-1102 (YPNDSVY*ANWMLS), EFS247-259 (GTDEGIY*DVPLLG), INSR1179-1191

(GMTRDIY*ETDYYR)

and

(GMTRDIYETDY*YR)

and

TRKA670-682

(GMSRDIY*STDYYR), (supplementary data). Some of these peptides showed modifications not only on tyrosine residues but also additionally on threonine and serine residues (data not shown). Myt1 is a dual-specific kinase that is generally characterized as a threonine and tyrosine modifying kinase.4 However, Myt1 is also known to modify serine residues, although this

kind

of

posttranslational

modification

has

only

been

described

for

Myt1

autophosphorylation so far.41 These serine phosphorylations could not be prevented in the presence of promiscuous inhibitors of many Ser/Thr-kinases, such as sunitinib or lestaurtinib. These inhibitors do not affect Myt1,8,42 demonstrating that a Ser/Thr-kinase contamination in the Myt1 preparation is highly unlikely to be responsible for the detected Serphosphorylation. Two substrate peptides (EFS247-259 and PDGFRB575-587) were investigated with respect to the inhibition profile. These peptides were selected with respect to their solubility and Km,app value

(Table 3). EFS247-259 was the substrate having the lowest Km,app and PDGFRB575-587 was the substrate having the highest Km,app as determined by FPIA. Both substrates did not show any solubility issues. Since there is no selective Myt1 inhibitor available so far, specificity has to be gained from a combination of known kinase inhibitors. The Myt1 inhibitor analysis is based on the results of Davis et al. who set up inhibition profiles for > 80% of the human protein kinome.42 It was found that, using dasatinib and CEP-701 (lestaurtinib) as inhibitors, only three tyrosine kinases showed the same inhibition profile as Myt1 (EPHA8, EPHB3 and ErbB2). Dasatinib inhibits Myt1, while the generally very promiscuous inhibitor CEP-701 should not affect the catalytic activity of Myt1. Additionally, PD-173952, a pyrido[2,3d]pyrimidine-based inhibitor closely related to PD-173955, was used. This compound is appromixately 10fold more potent toward Myt1 than dasatinib (Ki values 8 nM and 73 nM, respectively).8,43 Taking PD173952 into account, only two kinases exhibit the same inhibition profile (EPHA8 and EPHB3). Exclusively for Myt1, the inhibitory potency can be ranked as follows: PD173952 > dasatinib >> CEP-701. In view of this background, kinase reactions in the presence or absence of 5 µM CEP-701, dasatinib or PD173952 were carried out with both substrates, EFS247-259 and PDGFRB575-587. Anisotropy values were converted into relative conversion using controls without substrate and conversion controls without inhibitor. The results are displayed in Fig. 3. For both substrates, the relative conversion is dramatically decreased in the presence of dasatinib. In presence of the even more potent Myt1 inhibitor PD173952, the relative conversion drops down to baseline values. The presence of CEP-701 does not considerably affect the relative conversion. These results exactly match the inhibition profile which could be expected from the literature for Myt1.

5. Conclusion We report several peptides that can be used as Myt1 kinase substrates in vitro. The availability of such substrates compatible with high throuphput screenings enables studies

not possible so far and may boost inhibitor development. Additionally, the proteins corresponding to the identified peptides may be potential downstream targets of Myt1 kinase.14 Recently, a new cellular downstream target for Wee1 was discovered, vastly expanding its biological role.43 Similarly, besides the assay development, the peptide studies presented may help to clarify cellular processes and further define the biological role of Myt1 and other proteins. Future work has to resolve whether the results reported herein can be transferred to an actual cellular environment.

6. Acknowledgements We thank Dr. Schierhorn and Christina Gersching from the Martin-Luther-University HalleWittenberg for MALDI-MS measurements. This work was supported by the European Regional Development Fund of the European Commission.

7. References [1] Lindqvist, A.; Rodríguez-Bravo, V.; Medema, R. H. J. Cell Biol. 2009, 185, 193. [2] Booher, R. N.; Holman, P. S.; Fattaey, A. J Biol. Chem. 1997, 272, 22300. [3] Parker, L. L.; Piwnica-Worms, H. Science 1992, 257, 1955. [4] Mueller, P. R.; Coleman, T. R.; Kumagai, A.; Dunphy, W. G. Science 1995, 270, 86. [5] Nakanishi, M.; Ando, H.; Watanabe, N.; Kitamura, K.; Ito, K.; Okayama, H.; Miyamoto, T.; Agui, T.; Sasaki, M. Genes Cells 2000, 5, 839. [6] Chow, J. P. H.; Poon, R. Y. C. Oncogene 2013, 32, 4778. [7] Rohe, A.; Erdmann, F.; Bäßler, C.; Wichapong, K.; Sippl, W.; Schmidt, M. Bioorg. Med. Chem. Lett. 2012, 22, 1219. [8] Rohe, A.; Henze, C.; Erdmann, F.; Sippl, W.; Schmidt, M. Assay Drug Dev. Technol. 2014, 12, 136.

[9] Rohe, A.; Göllner, C.; Wichapong, K.; Erdmann, F.; Al-Mazaideh, G. M.; Sippl, W.; Schmidt, M. Eur. J. Med. Chem. 2013, 61, 41. [10] Kristjánsdóttir, K.; Rudolph, J. Anal. Biochem. 2003, 316, 41. [11] Rohe, A.; Göllner, C.; Erdmann, F.; Sippl, W.; Schmidt, M. J. Enzyme Inhib. Med. Chem. 2014, DOI: 10.3109/14756366.2014.926343. [12] Mah, A. S.; Elia, A. E.; Devgan, G.; Ptacek, J.; Schutkowski, M.; Snyder, M.; Yaffe, M. B.; Deshaies, R. J. BMC Biochem. 2005, 6, 22. [13] Schonberg, A.; Bergner, E.; Helm, S.; Agne, B.; Dunschede, B.; Schunemann, D.; Schutkowski, M.; Baginsky, S. PLoS ONE 2014, 9, e108344. [14] Schutkowski, M.; Reimer, U.; Panse, S.; D., L.; Lizcano, J. M.; Alessi, D. R.; SchneiderMergener, J. Angew. Chem. Int. Ed. 2004, 43, 2671. [15] Lizcano, J. M.; Deak, M.; Morrice, N.; Kieloch, A.; Hastie, C. J.; Dong, L.; Schutkowski, M.; Reimer, U.; Alessi, D. R. J. Biol. Chem. 2002, 277, 27839. [16] Rychlewski, L.; Kschischo, M.; Dong, L.; Schutkowski, M.; Reimer, U. J. Mol. Biol. 2004, 336, 307. [17] Wolf, A.; Rietscher, K.; Glass, M.; Huttelmaier, S.; Schutkowski, M.; Ihling, C.; Sinz, A.; Wingenfeld, A.; Mun, A.; Hatzfeld, M. J. Cell. Sci. 2013, 126, 1832. [18] Thiele, A.; Zerweck, J.; Weiwad, M.; Fischer, G.; Schutkowski, M. Methods Mol. Biol. 2009, 570, 203. [19] Papadopoulos, C.; Arato, K.; Lilienthal, E.; Zerweck, J.; Schutkowski, M.; Chatain, N.; Muller-Newen, G.; Becker, W.; de la Luna, S. J. Biol. Chem. 2011, 286, 5494. [20] Thiele, A.; Weiwad, M.; Zerweck, J.; Fischer, G.; Schutkowski, M. Methods Mol. Biol. 2010, 669, 173. [21] Panse, S.; Dong, L.; Burian, A.; Carus, R.; Schutkowski, M.; Reimer, U.; SchneiderMergener, J. Mol. Divers. 2004, 8, 291. [22] Schutkowski, M.; Reineke, U.; Reimer, U. ChemBioChem 2005, 6, 513.

[23] Boeckmann, B.; Bairoch, A.; Apweiler, R.; Blatter, M. C.; Estreicher, A.; Gasteiger, E.; Martin, M. J.; Michoud, K.; O'Donovan, C.; Phan, I.; Pilbout, S.; Schneider, M. Nucleic Acids Res. 2003, 31, 365. [24] Kreegipuu, A.; Blom, N.; Brunak, S. Nucleic Acids Res. 1999, 27, 237. [25] Frank, R. Tetrahedron 1992, 48, 9217. [26] Wenschuh, H.; Volkmer-Engert, R.; Schmidt, M.; Schulz, M.; Schneider-Mergener, J.; Reineke, U. Biopolymers 2000, 55, 188. [27] Ngo, Y.; Advani, R.; Valentini, D.; Gaseitsiwe, S.; Mahdavifar, S.; Maeurer, M.; Reilly, M. J. Immunol. Methods 2009, 343, 68. [28] Coin, I.; Beyermann, M.; Bienert, M. Nat. Protoc. 2007, 2, 3247. [29] Edelhoch, H. Biochemistry 1967, 6, 1948. [30] Mach, H.; Middaugh, C. R.; Lewis, R. V. Anal. Biochem. 1992, 200, 74. [31] Gill, S. C.; von Hippel, P. H. Anal. Biochem. 1989, 182, 319. [32] Starcher, B. Anal. Biochem. 2001, 292, 125. [33] Eyrich, B.; Sickmann, A.; Zahedy, R.P. Proteomics 2011, 11, 554. [34] Nahtman, T.; Jernberg, A.; Mahdavifar, S.; Zerweck, J.; Schutkowski, M.; Maeurer, M.; Reilly, M. J. Immunol. Methods 2007, 328, 1. [35] Quintana, F. J.; Hagedorn, P. H.; Elizur, G.; Merbl, Y.; Domany, E.; Cohen, I. R. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 14615. [36] Zerweck, J.; Masch, A.; Schutkowski, M. Methods Mol. Biol. 2009, 524, 169. [37] Nousiainen, M.; Silljé, H. H. W.; Sauer, G.; Nigg, E. A.; Körner, R. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 5391. [38] Dephoure, N.; Zhou, C.; Villén, J.; Beausoleil, S. A.; Bakalarski, C. E.; Elledge, S. J.; Gygi, S. P. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 10762. [39] Ruiz, E. J.; Vilar, M.; Nebreda, A. R. Curr. Biol. 2010, 20, 717. [40] Daub, H.; Olsen, J. V.; Bairlein, M.; Gnad, F.; Oppermann, F. S.; Körner, R.; Greff, Z.; Kéri, G.; Stemmann, O.; Mann, M. Mol. Cell 2008, 31, 438. [41] Liu, F.; Stanton, J. J.; Wu, Z.; Piwnica-Worms, H. Mol Cell. Biol. 1997, 17, 571.

[42] Davis, M. I.; Hunt, J. P.; Herrgard, S.; Ciceri, P.; Wodicka, L. M.; Pallares, G.; Hocker, M.; Treiber, D. K.; Zarrinkar, P. P. Nat. Biotech. 2011, 29, 1046. [43] Wichapong, K.; Rohe, A.; Platzer, C.; Slynko, I.; Erdmann, F.; Schmidt, M.; Sippl, W. J. Chem. Inf. Model. 2014, 54, 881. [44] Mahajan, K.; Fang, B.; Koomen, J. M.; Mahajan, N. P. Nat. Struct. Biol. 2012, 19, 930. [45] The Uniprot Consortium. Nucleic Acids Res. 2012, 40, D71.

Figure captions:

Fig. 1: Aligned dot plots of the mean normalized response index for the individually assessed peptide microarrays I and II after incubation with Wee1 or Myt1 (displayed are Median ± interquartile range). Each column contains 1152 data points representing mean values of triplicate measurements.

Fig. 2: Venn plot of peptide substrates identified in microarray experiments for Wee1 and Myt1. For more information on the peptides refer to the upcoming tables.

Fig. 3: Substrate validation by inhibition profiles for EFS247-259 (A) and PDGFRB575-587 (B). Kinase reactions were carried out in presence or absence of various kinase inhibitors. Values are normalized against controls and displayed as means ± SEM (n=3).

Table 1: Peptide substrates from microarray experiments identified for Wee1 in alphabetical order. Peptides also positive for Myt1 are highlighted (bold). Peptide

Sequence

Uniprot ID 45

Corresponding protein full name

ADAM15

VMLGAGYWYRARL

Q13444

ADAM 15 (A disintegrin and metalloproteinase domain 15)

CD79a182-194

YEDENLYEGLNLD

P11912

Membrane-bound immunoglobulin associated protein (CD79a)

9-21

KIGEGTYGVVYKG

P06493

Cell division control protein 2, Cyclin-dependent kinase 1 (Cdk1)

DDR1

786-798

GMSRNLYAGDYYR

Q08345

Epithelial discoidin domain receptor 1 (Tyrosine kinase DDR1)

DDR1

791-803

LYAGDYYRVQGRA

Q08345

Epithelial discoidin domain receptor 1 (Tyrosine kinase DDR1)

602-614

APGMKVYIDPFTY

P54753

Ephrin type-B receptor 3 (Tyrosine-protein kinase hEK-2)

VAENPEYLSEFSL

Q15303

ERBB-4 receptor protein-tyrosine kinase (p180erbB4) (cell surface receptor HER4).

573-585

YIEDEDYYKASVT

Q14289

Focal adhesion kinase 2 (FADK 2) (Proline-rich tyrosine kinase 2)

1159-1171

DIYETDYYRKGGK

P08069

Insulin-like growth factor I receptor

IKKE

166-178

DDEKFVSVYGTEE

Q14164

Inhibitor of nuclear factor kappa-B kinase epsilon subunit (IkBKE) (IKK-epsilon)

INR1

475-487

SSSIDEYFSEQPL

P17181

Interferon-alpha/beta receptor alpha chain (IFNalpha-REC)

1179-1191

GMTRDIYETDYYR

P06213

Insulin receptor (IR)

DSEMTGYVVTRWY

P53778

MAP kinase p38 gamma, ,MAP kinase 12

709-721

CDK1

EPB3

ERBB4 FAK2

1278-1290

IG1R

INSR

179-191

MAPK12 MMD

224-236

WKYLYRSPTDFMR

Q15546

Monocyte to macraphage differentiation factor

747-759

RDINSLYDVSRMY

P16885

Phospholipase C-gamma-2

111-123

NNEEESSYSYEEI

P12579

RNA polymerase alpha subunit (Phosphoprotein P, Protein P)

670-682

IYSTDYYRVGGRT

P04629

High affinity nerve growth factor receptor (TRK1, TrkA)

670-682

GMSRDIYSTDYYR

P04629

High affinity nerve growth factor receptor (TRK1, Trk-A)

675-687

DIYSTDYYRVGGR

P04629

High affinity nerve growth factor receptor (TRK1, TrkA)

TRKB

696-708

GMSRDVYSTDYYR

Q16620

BDNF/NT-3 growth factor receptor (Trk-B)

TRKB

700-712

DVYSTDYYRVGGH

Q16620

BDNF/NT-3 growth factor receptor (Trk-B)

PLCG2

PROTP TRKA

TRKA TRKA

Table 2: Peptide substrates from microarray experiments identified for Myt1 in alphabetical order. Peptides also positive for Wee1 are highlighted (bold). Peptide EAA1

719-731

EFS 247-259 INR1

Sequence

Uniprot ID 44

Corresponding protein full name

AFLLESTMNEYYR

Q16099

Glutamate receptor ionotropic kainate 4 (GluK4, EAA1)

GTDEGIYDVPLLG

O43281

Embryonal FYN-associated substrate (hEFS).

475-487

SSSIDEYFSEQPL

P17181

Interferon-alpha/beta receptor alpha chain (IFNalpha-REC)

1179-1191

INSR

GMTRDIYETDYYR

P06213

Insulin receptor (IR)

176-188

DEEMTGYVATRWY

Q15759

MAP kinase p38 beta, MAP kinase 11

176-188

DDEMTGYVATRWY

Q16539

MAP kinase p38, MAP kinase 14

WKYLYRSPTDFMR

Q15546

Monocyte to macrophage differentiation factor

DGHEYIYVDPMQL

P09619

Platelet-derived growth factor receptor beta

MAPK11 MAPK14 MMD

224-236

PDGFRB

575-587

1090-1102

RET

YPNDSVYANWMLS

P07949

Proto-oncogene tyrosine-protein kinase receptor Ret

381-393

NPNYGYTSYDTFS

P35499

Sodium channel protein type 4 subunit alpha

670-682

GMSRDIYSTDYYR

P04629

High affinity nerve growth factor receptor (TRK1, Trk-A)

SC4A

TRKA

Table3: Solution phase verification of the identified Myt1 peptide substrates via fluorescence polarization immunoassay. The microarray response (as normalized index) is given for comparison.

Peptide Microarray response Displacement of the (alphabetical (normalized index) fluorescent probe at order) 200 µM peptide 2.96 EAA1719-731 5%1 12.91 EFS247-259 >95% 475-487 3.71 INR1 21% 7.04 INSR1179-1191 >95% 3.25 MAPK11176-188 5%2 176-188 3.83 MAPK14 14%3 4.93 MMD224-236 40% 575-587 5.19 PDGFRB 84% 1090-1102 3.27 RET 93% 3.45 SC4A381-393 >95% 5.06 TRKA670-682 >95% 1 peptide concentration reduced to 100 µM (solubility issues) 2 peptide concentration reduced to 150 µM (solubility issues) 3 peptide concentration reduced to 50 µM (solubility issues)

Km,app [µM] not determined 6.8 ± 1.4 not determined 11.9 ± 3.6 not determined not determined not determined 46.0 ± 5.7 8.7 ± 2.0 44.3 ± 8.4 18.3 ± 3.4

Fig. 1

Fig. 2

Fig. 3

Graphical abstract