Identification and Validation of Differential Phosphorylation Sites of the

Apr 14, 2015 - Granulosa cell tumor (GCT) is a rare form of ovarian cancer classified as a sex cord–stromal tumor. The c.402C→G missense mutation ...
0 downloads 14 Views 4MB Size
Download PDF
Article pubs.acs.org/jpr

Identification and Validation of Differential Phosphorylation Sites of the Nuclear FOXL2 Protein as Potential Novel Biomarkers for AdultType Granulosa Cell Tumors Dae-Shik Suh,†,# Hoon Kyu Oh,‡,# Jae-Hong Kim,§ Seeun Park,§ Eunkyoung Shin,§ Kangseok Lee,∥ Yong-Hak Kim,*,⊥ and Jeehyeon Bae*,§ †

Department of Obstetrics and Gynecology, Asan Medical Center, University of Ulsan College of Medicine, Seoul 138-736, Korea Department of Pathology, ⊥Department of Microbiology, Catholic University of Daegu School of Medicine, Daegu 705-718, Korea § School of Pharmacy, ∥Department of Life Science, Chung-Ang University, Seoul 156-756, Korea ‡

S Supporting Information *

ABSTRACT: Granulosa cell tumor (GCT) is a rare form of ovarian cancer classified as a sex cord−stromal tumor. The c.402C→G missense mutation in the FOXL2 gene that changes cysteine 134 to tryptophan (C134W) is found in more than 97% of adult-type GCTs, and the C134W FOXL2 mutant is hyperphosphorylated. We identified three differential phosphorylation sites, at serine 33 (S33), tyrosine 186 (Y186), and serine 238 (S238), of the C134W mutant by tandem mass spectrometry. Among these sites, antibodies were raised against the pS33 and pY186 epitopes using specific peptides, and they were tested by immunostaining tissue microarrays of archival adult-type GCT specimens, other tumors, and normal tissues. The pS33 antibody showed greater sensitivity and specificity for the detection of adulttype GCTs than that of the other phospho and nonphospho antibodies. The specificity of the pS33 antibody to the pS33 epitope was further confirmed by enriching the pS33 peptide by affinity chromatography using the immobilized antibody, followed by mass spectrometric and western blot analyses from whole cell lysates of the adult-type GCT cell line, KGN. pS33 FOXL2 immunostaining levels were significantly higher in adult-type GCTs than those in other tumors and tissues. The receiver operating characteristic curve analysis of pS33 FOXL2 showed high sensitivity (1.0) and specificity (0.76) to adult-type GCTs with a cutoff score of >30% positive cells, and the area under the curve was 0.96. This suggests the potential of pS33 FOXL2 to serve as a new biomarker for the diagnosis of adult-type GCT. KEYWORDS: Diagnosis, FOXL2, granulosa cell tumor, phosphorylation, serine 33

■

INTRODUCTION

and function, and FOXL2 knockout causes mature mouse ovaries to develop into testes.10 We previously reported that FOXL2 is a putative tumor suppressor of GCTs and that the C134W mutant exhibits defective apoptotic activities.11 Determination of the 402C→G mutational status of FOXL2 has recently garnered attention as a precise method to identify adult-type GCTs.7,8,12−14 Molecular techniques to identify the 402C→G mutation have been reported, including Sanger sequencing, pyrosequencing, TaqMan real-time polymerase chain reaction (PCR)-based allelic discrimination assay, and MALDI-TOF-mass spectrometry.7,8,13,15−18 A very recent study showed that the C134W mutant of human FOXL2 (UniProt KB no. P58102) presents posttranslational modifications, affecting its protein stability, transcriptional activity, and tumor suppressive activity.13 In

Ovarian sex cord−stromal tumors (SCSTs) are relatively rare cancers that constitute about 8% of ovarian malignancy, and granulosa cell tumors (GCTs) account for 90% of SCSTs.1,2 GCTs are classified as juvenile- or adult-type, and approximately 95% of tumors are adult-type GCTs, with a median age of disease presentation of 50−54 years.3,4 GCTs are characterized by slow tumor growth and a propensity for frequent relapse with long delays.3,4 Little information was available regarding the development of GCTs until the recent discovery of FOXL2 c.402C→G, a somatic missense mutation that leads to a p.C134W change.5 Remarkably, the 402C→G mutation is found in more than 97% of adult-type GCTs, but it is not observed in other tumors,5−8 indicating that this mutation is likely linked to the etiology of GCT development. FOXL2 is an evolutionarily conserved forkhead domaincontaining transcription factor that is highly expressed in ovarian granulosa cells.9 It is involved in ovarian development © XXXX American Chemical Society

Received: November 30, 2014

A

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

previously described.22 Each cellular fraction was incubated in 10 mM Tris-HCl (pH 7.0) containing 25 mM KCl, 1.1 mM MgCl, 0.1 mM EDTA, 1 mM DTT, 1 mM sodium azide, 2 mM ATP, and 35% glycerol for 12 h at 37 °C, and subsamples (10 μg of protein each) were harvested at 0, 3, 6, and 12 h and frozen at −80 °C, pending analysis of the relative levels of wildtype and C134W mutant FOXL2 proteins by SDS-PAGE and western blotting.

particular, serine 33 (S33) of the C134W mutant is hyperphosphorylated by the activity of glycogen synthase kinase-3β (GSK3β), as demonstrated in patients with adulttype GCT who are heterozygotes for the 402C→G genotype.13 However, to date, the contribution of post-translational modifications to the diagnosis is unknown, mainly due to the lack of a coherent pipeline connecting biomarker discovery with well-established methods for validation. Tumor tissues are necessary for the validation of specific GCT biomarkers, but the performance of biopsies in patients with GCT in the clinic is limited. To overcome this limitation, archival tissue collections are useful for accelerating the discovery and validation of identified biomarkers by a combined method of in situ histopathologic analysis and immunohistochemistry in a tissue microarray format.19−21 In the present study, we aimed to identify and validate phosphorylation sites of the C134W FOXL2 mutant as potential biomarkers for the diagnosis of adult-type GCT. Nanoflow liquid chromatography−tandem mass spectrometry (nLC−MS/MS) and tissue microarrays (TMAs) were employed to identify and verify FOXL2 phosphorylation sites, which were subsequently used to detect and diagnose adult-type GCT by immunostaining. A selected pS33-site biomarker was further tested with individual sample sets of various types of ovarian cancer, including adult-type GCT, juvenile-type GCT, Sertoli−Leydig cell tumor, fibrothecoma, serous, endometrioid, clear cell, and undifferentiated ovarian cancer. The receiver operating characteristic (ROC) curve was analyzed to evaluate the diagnostic accuracy for differentiating adult-type GCT from other types of cancers and noncancerous tissues.

■

Mass Spectrometry Analysis

Purified wild-type FOXL2 and C134W mutant proteins from the nucleus were isolated from SDS-PAGE gels. The isolated proteins were reduced with 10 mM DTT at 60 °C for 15 min and then alkylated with 100 mM iodoacetamide at room temperature in the dark. The proteins were digested for 24 h by sequencing-grade trypsin (Promega, Madison, WI, USA) at 37 °C, and the tryptic peptides were extracted as previously described.23 The dried peptides were dissolved in 10 μL of 0.4% acetic acid and analyzed on a Thermo Scientific Velos Pro mass analyzer connected to a reversed-phase Magic C18AQ column (i.d. 75 μm × o.d. 150 μm × L75 mm) and an EASYnLC 1000 system. The chromatographic conditions involved a 90 min linear gradient from 5 to 40% acetonitrile in 0.1% formic acid buffer solution at 0.3 μL min−1. Survey full-scan MS spectra (m/z 300−2000) were followed by data-dependent MS2 and MS3 scans of the most intense ions determined from the preview scans with the following parameters: isolation width, ± 1.5 m/z; collision energy, 35%; dynamic exclusion duration, 30 s; and neutral loss of 79.98/z. The acquired MS2 data were analyzed by the Proteome Discoverer program, version 1.3, with the database for Homo sapiens (UniProt proteome ID, UP000005640) and the common Repository of Adventitious Proteins (ftp://ftp.thegpm.org/fasta/cRAP) that are present either by accident or through unavoidable contamination of protein samples. The Sequest search parameters were set as follows: precursor mass tolerance, 0.8 Da; fragment mass error, 1 Da; fixed carbamidomethylation of cysteine; and variable phosphorylation of serine, threonine, and tyrosine. Peptides identified by tandem mass spectrometry were filtered for a probability greater than 0.99 and a target-decoy false-discovery rate (FDR) less than 0.01. In order to evaluate the pS33 peptide enrichment by the anti-pS33 antibody, the tryptic digest of the KGN cell lysate was subjected to affinity chromatography with immobilized antibodies, as described below, and selected ion monitoring (SIM) of the peptide ions of interest was performed with MS/MS confirmation during a 40 min linear gradient from 5 to 40% acetonitrile in 0.1% formic acid buffer solution at 0.3 μL min−1.

EXPERIMENTAL PROCEDURES

Cell Culture and Protein Purification

Human adult-type GCT-derived KGN cells (1 × 108) were transfected with 30 μg of plasmid encoding either wild-type FOXL2 or the C402G mutant gene, as previously described.11 After incubating the cells for 24 h in Dulbecco’s modified Eagle’s medium (DMEM)/F12 medium (Caisson, North Logan, UT, USA), cell lysates were prepared and subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and western blotting to analyze FOXL2 protein expression using an anti-FOXL2 antibody (Abcam, Cambridge, MA, USA). Protein concentrations were determined using a Pierce BCA protein assay reagent kit (Life Technologies, Rockford, IL, USA). Cells were washed with 1× PBS and then disrupted using cytosolic extraction buffer A [10 mM HEPES, pH 7.9, 10 mM KCl, and 0.1 mM ethylenediaminetetraacetic acid (EDTA)] containing 1 mM dithiothreitol (DTT) and 10% IGEPAL CA-630 for 5 min at room temperature. After centrifugation at 1000g for 5 min at 4 °C, the cytoplasmic fractions were transferred into new containers, and the pellets containing the nuclear fractions were solubilized in nuclear lysis buffer B (20 mM HEPES, pH 7.9, 0.4 M NaCl, 1 mM EDTA, and 10% glycerol) containing 1 mM DTT. The nuclear proteins were then obtained by centrifugation at 10 000g for 10 min at 4 °C. To enrich the nuclear wild-type and C134W mutant FOXL2 proteins, the nuclear proteins were treated with a protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO, USA) and a phosphatase inhibitor cocktail (Sigma-Aldrich) for 5 min at room temperature and then subjected to immunoprecipitation with an anti-FOXL2 antibody, as

Tissue Sampling and TMA Construction

The TMA was constructed with 1 mm cores of archival paraffin specimens from 18 patients diagnosed with adult-type GCT at Daegu Catholic University Medical Center (DCUMC), Daegu, Republic of Korea, between 2001 and 2011 (Supporting Information Table S1). The median age of the patients was 49.5 years, with a range of 17−82 years. A control set of single cores of 11 other cancers and normal tissues, including two normal ovarian tissues, one normal uterine endometrial tissue, one normal uterine myometrial tissue, one normal breast tissue, one normal prostate tissue, one normal tonsil tissue, one ovarian serous carcinoma, one ovarian mucinous carcinoma, one uterine endometrial adenocarcinoma, and one colon adenocarcinoma, were obtained from the Standard Tissue Collection of Human Cancers and Diseases at the Pathology B

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research Laboratory of DCUMC. Additional TMAs were produced that consisted of 1 mm paraffin tissue blocks of serous ovarian cancer specimens (SECA, n = 37), endometrioid ovarian cancer specimens (EMCA, n = 16), clear cell ovarian cancer specimens (CLCA, n = 9), undifferentiated ovarian cancer specimens (UNCA, n = 2), prostate cancer specimens (PC, n = 2), noncancerous tonsil specimens (n = 2), noncancerous endometria specimens (n = 2), noncancerous colon specimens (n = 2), and noncancerous ovary specimens (n = 2), which were collected between 2000 and 2008 in the Pathology Laboratory of DCUMC. All tissues used were reexamined by hematoxylin and eosin staining of 5 μm thick sections of paraffin-embedded cores before TMA construction. In addition, individually tested paraffin materials, including eight cases of adult-type GCTs and five cases each of juvenile-type GCT, Sertoli−Leydig cell tumor, fibrothecoma, and serous ovarian carcinoma, were obtained from the Bio-Resource Center at the Seoul Asan Medical Center and reviewed by a pathologist. The present study was reviewed and approved by the DCUMC and the Seoul Asan Medical Center Institutional Review Board. Informed consent was obtained from all subjects who participated in this study.

Genotyping

Generation of Primary Antibodies

Immunohistochemical Analysis

Rabbit polyclonal antibodies for the detection of pS33, pY186, and nonphosphorylated FOXL2 proteins were generated and purified using synthetic peptides corresponding to the following identified epitopes: GRTVKEPEGPPPpS33PGKGGGGGGGTAPEKPDPAQKPPYSYV (pS33), GRTVKEPEGPPPS33PGKC (non-pS33), GLFGAGGAAGGCGVAGAGADGpY 186 GYLAPPK (pY186), and GLFGAGGAAGGCGVAGAGADGY186GYLAPPK (non-pY186). In addition, a rabbit polyclonal antibody to pS263 FOXL2 (Abcam) was also tested for TMA immunostaining. The specificity of antibody binding to FOXL2 was tested by western blot analysis and compared with that of a mouse monoclonal anti-FOXL2 antibody (Abcam).

Immunohistochemical staining for phospho and nonphospho FOXL2 was performed on a TMA and on slides with individual tissues. The procedures were carried out using the Bond polymer intense detection system (Leica Microsystems, VIC, Australia) according to the manufacturer’s instructions with minor modifications. In brief, 5 μm thick tissue sections of the TMA were deparaffinized using Bond dewax solution, and an antigen retrieval procedure was performed using Bond ER solution for 30 min at 100 °C. The endogenous peroxidase activity was inhibited by treatment with hydrogen peroxide for 5 min. Slides were then treated for 15 min at room temperature with the rabbit phospho and nonphospho antibodies using the biotin-free polymeric horseradish peroxidase (HRP)-linker antibody conjugate system in a Bond-Max automatic slide stainer. Immunohistochemical analyses of individual tissue slides were performed as described previously. 13 For quantitative analysis of the immunostaining, the 400× images were acquired using an Olympus IX73 microscope (Olympus, Tokyo, Japan) equipped with a DP73 digital camera and cellSens software (Olympus). After image acquisition, the proportion of positive cells was calculated from the immunostaining intensity of the nucleus relative to the cytoplasm. The proportion and intensity scores were determined by a modified Allred scoring method. The proportion score of the bottom 1% was set to zero, and that for the top 99% of positive cells was scored from 1 to 5 in 20% intervals. The intensity score was scaled from 0 to 3 with the nucleus staining nonspecifically (0), weakly (1), moderately (2), or strongly (3). For ROC curve analysis, the proportion and intensity scores were then summed to produce total scores of 0 to 8.

For FOXL2 sequence analysis, paraffin-embedded ovary sections were collected in microtubes, and preheated xylene was added for 10 min at 65 °C. The tubes were then centrifuged at 13 000g for 5 min, and the supernatant was discarded. This procedure was repeated three times to ensure that no paraffin residue remained. The pellet was washed in 100% ethanol (Sigma-Aldrich). Genomic DNA was extracted from the deparaffinized tissues using an Intron G-DEXTM genomic DNA extraction kit (Intron, Seongnam, Korea) according to the manufacturer’s protocol. The concentration and quality of genomic DNA were determined with an ND1000 spectrophotometer (Thermo Scientific, Waltham, MA, USA). The genomic DNA obtained was used to amplify the 1138 bp fragment of the FOXL2 gene by PCR using the following primers: 5′-CTAGAATTCAAATGATGGCCAGCTACCCC-3′ and 5′-CTACTCGAGTCAGAGATCGAGGCGCGAATG-3′. The PCR products were purified using a MEGAspin agarose gel extraction kit (Intron) and sequenced using a BigDye Terminator v3.1 Cycle sequencing kit (Applied Biosystems, Foster City, CA, USA) and an ABI Prism 3730XL genetic analyzer (Applied Biosystems).

Antibody Immobilization and Affinity Chromatography

Purified antibodies (300 μg each) against synthetic peptides containing pS33 and non-pS33 residues were biotinylated using Thermo Scientific Sulfo-NHS-LC-Biotin and bound to 100 μL of avidin agarose according to the manufacturer’s protocols. The prepared antibody resins were applied for the enrichment of the FOXL2 protein and peptide in whole cell lysates and tryptic digest of KGN cells with various protein concentrations (0.01−10 μg) to 10 μL of each column’s resin equilibrated with 10 volumes of 1× PBS containing 0.02% Triton X-100. Sample loading was performed by a Dynal sample mixer with rotation for 1 h at room temperature, and the loaded resins were washed with 10 volumes of 1× PBS before the antigen was eluted by 3 volumes of 0.1 M glycine buffer at pH 2.8. The eluted protein was neutralized by mixing it with 4 μL of 100 mM Tris buffer prior to SDS-PAGE and western blot analysis. To determine the specificity of the antibodies’ binding to the epitopes, 1 μg of tryptic peptides from whole cell lysate was loaded onto 10 μL of resin from each column, and the columns were eluted by the same procedures as above. The eluted peptide was treated with 0.1% trifluoroacetic acid for cleanup using Millipore Ziptips C18 and vacuum-dried. Samples were dissolved in 0.4% acetic acid and analyzed as described above by mass spectrometry with SIM of m/z 710.8 and 750.8 to detect the peptides containing pS33 and non-pS33 residues with charge states of +2.

ROC Curve

Optimal cutoff values for phospho or nonphospho FOXL2 biomarkers were obtained by ROC curve analysis using the TMA assay results. A strong relationship between a selected biomarker and adult-type GCT was validated by the ROC curve analysis-generated cutoff values in the overall TMA validation set. C

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

Figure 1. Purification and stability of wild-type (WT) and C134W mutant FOXL2 proteins in cytoplasmic (WCE) and nuclear (NUC) fractions of KGN cells. (A) Purification steps for WT and C134W mutant FOXL2 proteins in the nuclear fractions by immunoprecipitation (IP) with the antiFOXL2 antibody. Asterisks indicate bands that probably arise from protein degradation. (B) Western blot assay for degradation of WT and C134W mutant FOXL2 proteins in the cytoplasmic and nuclear fractions of KGN cells during a 12 h incubation in in vitro proteasomal degradation buffer systems. Immunoprecipitated proteins obtained from the nuclear fractions of the WT and C134W mutant vector-transformed cells were included as experimental controls under the same in vitro degradation conditions. (C) Plots of relative intensities of WT and C134W mutant FOXL2 proteins in western blot assays of products from (B) that show the stability of cytoplasmic and nuclear FOXL2 and C134W proteins under the in vitro assay conditions. (D) Coomassie staining and western blot analysis of WT and C134W mutant FOXL2 proteins in the nuclear fraction of KGN cells. A covalent modification of WT FOXL2 was detected at about 85 kDa.

Statistical Analysis

than that of wild-type FOXL2 (t1/2 = 8.9 h). In contrast, both protein levels in the nuclear fractions remained relatively constant up to 12 h, similar to the nuclear immunoprecipitated proteins in the in vitro assay. This finding suggests that the cytoplasmic C134W mutant protein undergoes rapid proteasomal degradation, which contributes to its decreased stability.13 In the nucleus, FOXL2 requires UBC9-mediated SUMOylation at lysine 25 (K25) for stabilization and transcriptional activation.24 However, the C134W mutant protein is resistant to K25 SUMOylation due to phosphorylation at S33.13 In this study, we observed that nuclear-localized wild-type FOXL2 formed a stable protein with a higher molecular weight, possibly indicating multiple binding with about four SUMO proteins (about 48 kDa), whereas we observed a negligible amount of covalent modification of the C134W mutant (Figure 1D).

All P values derived from t-tests are two-sided, and P < 0.05 was considered to be statistically significant. Statistical analyses were performed using SPSS, v. 17.0 (SPSS, Inc., Chicago, IL, USA), and SigmaPlot (Systat Software, San Jose, CA, USA).

■

RESULTS

Expression and Purification of Wild-Type FOXL2 and C134W Mutant Proteins

Because FOXL2 is a transcription factor, we specifically aimed to identify its phosphorylation profile in the nucleus. Wild-type FOXL2 and C134W mutant genes were expressed in KGN cells, and cytoplasmic and nuclear proteins were fractionated (Figure 1A). SDS-PAGE and western blotting showed that the expression level of the C134W mutant vector in the cytoplasm of KGN cells was higher than that of the wild-type FOXL2 vector, but their protein levels were similar in the nuclear fractions. When proteasomal degradation was assayed in vitro, the half-life (t1/2) for the cytoplasmic C134W mutant protein was 4.6 h (Figure 1B,C), which was approximately 2-fold lower

Identification of FOXL2 Phospho Sites by Tandem Mass Spectrometry

To identify differential phosphorylation sites in FOXL2, monomeric bands of wild-type FOXL2 and C134W mutant proteins were isolated from the SDS-PAGE gel of immunoD

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

Figure 2. Mass spectrometry analysis of FOXL2 phosphorylation. (A) Tandem mass spectra of phosphorylated and nonphosphorylated peptides with the same amino acid sequence, 23TVKEPEGPPPSPGK36, showing different ionization efficiencies and cross-correlation (Xcorr) values for the effectiveness and accuracy of electrospray ionization and collision-induced fragmentation. Log-intensity, log(I), of the precursor MS2 ion peak is shown as 100% in each spectrum. (B) Selected ion monitoring chromatograms of m/z 750.8 and 710.8 in 1 μg of trypsin-digested KGN cell lysate (upper), pS33 antibody-enriched sample (middle), or non-pS33 antibody-enriched sample (bottom) obtained from affinity chromatography. Elution times of pS33- and non-pS33-containing peptides were confirmed by examining the MS2 spectra and the tandem mass spectra in (A).

For a quantitative assessment, the corresponding nonphospho-peptides were considered to be counterparts of the post-translational modification. However, we identified only a single peptide spectrum that matches the non-pS33 peptide, TVKEPEGPPPS33PGK, in the two samples because it is unusually difficult to sequence the proline-rich peptide by collision-induced fragmentation due to the high efficiency of cleavage at the amide bond at the N-terminus of proline residues, which results in a relatively low abundance of fragments arising from cleavage at other amide bonds.25 Hence, it was difficult to compare the pS33 peptide levels with those of the corresponding non-pS33 peptides in this study. Against the high efficiency of ion trap collision-induced dissociation, phosphorylation of the proline-rich peptide at S33

precipitated fractions (Figure 1A), and trypsinized peptides were analyzed by MS2. After filtering the peptide spectrum matches for a peptide probability of 0.99 and target-decoy FDR < 0.01, the identified peptides shown in Supporting Information Table S2 cover 49.7 and 46.3% of the total amino acid sequences of wild-type FOXL2 and the C134W mutant, respectively. From the acquired MS2 and MS3 data, we identified three phospho-peptides: TVKEPEGPPPpS33PGK (pS33), LFGAGGAAGGC#GVAGAGADGYGpY186LAPPK (pY186), and ASCQMAAAAAAAAAAAAAAGPGpS238PGAAAVVK (pS238) (Supporting Information Figure S1), among which pS33 peptides were frequently detected in both samples, whereas single pY186 and pS238 peptides were found only in the wild-type FOXL2 sample. E

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

Figure 3. Evaluation of phospho and nonphospho FOXL2 by TMA. (A) Scatter histograms of the percentage of positive cells determined by immunohistochemical staining of nuclear FOXL2 with specific antibodies raised against pS33, non-pS33, pY183, non-pY186, and pS263. A statistical difference from two-tailed t-tests was observed between adult-type GCTs (n = 35−36 blocks) and other tissues (n = 11 blocks) in the TMA regarding nuclear pS33 FOXL2 staining (***P < 0.001). Representative images of the same tissue blocks (A1) captured from the 400× TMA slide images in Supporting Information Figure S3A−E are also shown. (B, C) ROC curve analyses of the proportion of positive cells (B) and of the total score (C) show similar AUC (area under the curve) values. Total immunostaining scores are calculated as the sum of the proportion score in the range of 0−5 and the intensity score within the range of 0−3. (D) Scatter histograms of the percentage of positive cells determined by immunohistochemical staining of nuclear FOXL2 with the pS33 antibody. A statistical difference from two-tailed t-tests was observed between the TMA data of adult-type GCTs (A) and the TMA data of other types of cancers and normal tissues (Supporting Information Figure S5) (***P < 0.001). SECA, serous ovarian cancer; EMCA, endometrioid ovarian cancer; CLCA, clear cell ovarian cancer; UNCA, undifferentiated ovarian cancer; and Other/Normal, other cancer and normal tissues of ovary, endometrium, colon, and tonsil.

accumulation of pS33 peptides in the C134W mutant protein when compared to that of the wild-type FOXL2 protein.

appeared to offer great promise for generating more accurate sequence information, although with a low ionization efficiency of the fragments (Figure 2A). Comparison of the relative pS33 peptide levels in the two samples after normalization to the numbers of the most abundant unmodified peptides, LTLSGIYQYIIAK (m/z = 742, z = +2) and GLAGPAASYGPYTR (m/ z = 691, z = +2), revealed an approximately 2.6-fold higher

Determination of the Specificity of the pS33 Antibody by Affinity Chromatography with Immobilized Antibodies

Using synthetic peptides with or without phospho-serine at position 33 of the FOXL2 sequence, rabbit polyclonal antibodies were generated and purified. These antibodies were biotinylated and bound to avidin agarose for affinity F

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

Figure 4. Individual validation of pS33 FOXL2 (C134W) as a biomarker for differentiating adult-type GCTs from different sex cord−stromal tumors (SCSTs) and other ovarian cancer types. (A) Genotyping results of eight adult-type GCTs, five juvenile-type GCTs, five fibrothecomas, and five serous ovarian carcinomas for the FOXL2 402C→G mutation. The heterozygous 402C→G mutation found in all cases of adult-type GCTs is shaded at the nucleotide 402C/402G position in FOXL2. (B) Immunohistochemical staining of pS33 (left panels) and non-pS33 (right panels) FOXL2 in ovarian SCSTs, including eight cases of adult-type GCTs, five cases of juvenile-type GCTs, five cases of Sertoli−Leydig cell tumors, and five cases of fibrothecomas. Five cases of serous ovarian carcinoma are included as the most abundant form of ovarian cancer. (C) Evaluation of pS33 FOXL2 for its ability to differentiate adult-type GCTs from other types of ovarian cancers. Scatter histograms represent the percentage of positive cells determined by pS33 and non-pS33 FOXL2 immunostaining in nuclei. A statistically significant difference in the levels of positive cells between adulttype GCTs and other types of ovarian cancers is shown at ***P < 0.001.

chromatography. The prepared antibody resins specifically enriched pS33 FOXL2 and non-pS33 FoxL2 from up to a 10 μg loading of whole cell lysate of KGN cells with no apparent cross-reactivity to each other’s antigen, as shown in Supporting Information Figure S2. The relative intensity of the pS33

antibody-enriched FOXL2 band normalized to total FOXL2 levels, detected using a mouse monoclonal anti-FOXL2 antibody, was about 7-fold higher than that of the non-pS33 antibody-enriched band. This result indicates that the pS33 antibody is very sensitive in the detection of pS33 FOXL2. G

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research In order to further determine the specificity of this antibody to the pS33 epitope, the antibody-bound resins were used for peptide enrichment from 1 μg of a trypsin-digested KGN whole cell lysate. The SIM chromatograms of trypsinized peptides containing pS33 and non-pS33 proved that the pS33 antibody was specifically bound to a pS33-containing peptide (m/z 750.8), showing no signal for the non-pS33-containing peptide (m/z 710.8) as the counter partner (Figure 2B). The elution time for a peptide of interest was confirmed by examining the MS2 spectra with a SEQUEST preliminary score (Sp) of more than 400. The efficiency of pS33 recovery by affinity chromatography using the pS33 antibody was nearly 100%, as estimated by the SIM analysis of 1 μg of a trypsinized sample of whole cell lysate. In contrast, non-pS33 antibody resins produced no mass signal corresponding to pS33- or nonpS33-containing peptide by the same procedure.

the increase in the ratio of pS33 FOXL2 rather than the level of FOXL2 (intensity) in the nucleus relative to the cytoplasm of ovarian granulosa cells. To further evaluate the specificity of the pS33 antibody to detect adult-type GCTs, another TMA that consists of multiple other types of ovarian cancers, including serous, endometrioid, clear cell, and undifferentiated types, prostate cancer, and normal tissues in ovary, endometrium, colon, and tonsil, was produced (Supporting Information Figure S5). By an automatic immunostaining process, the pS33 antibody resulted in significantly low proportions of positive cells in the TMA of other cancers and normal tissues compared to the results obtained for the TMA of adult-type GCTs (Figure 3D). Of interest, these data indicate that immunohistochemical analysis using pS33 FOXL2 is useful to distinguish adult-type GCTs from other types of ovarian cancer, including serous, endometrioid, clear cell, and undifferentiated types (Supporting Information Figure S6). In the TMA validation set, pS33 FOXL2 immunostaining had a specificity of 75% and a sensitivity of 100% for the detection of adult-type GCTs with a cutoff score of 30% positive cells.

Differentiation of Adult-Type GCT and Other Tissues by Immunostaining TMAs with Phospho- and Nonphospho-Peptide Antibodies

The association of FOXL2 pS33 with GCT development has been previously demonstrated.13 However, it was unclear whether and which phosphorylation sites were differently distributed in the nuclei of human cells derived from adult-type GCTs and other types of cancer and normal tissues. To specifically detect the accumulation of FOXL2 phosphorylated at a specific site in human cell nuclei, specific phospho-peptide antibodies were generated and purified using synthetic peptides corresponding to the pS33 and pY186 epitopes, and their performances in terms of specificity and sensitivity were tested by comparison of the TMA immunostaining results, including the results of the counter nonphospho-peptide antibodies and an unidentified pS263 FOXL2 antibody (Supporting Information Figure S3). The proportion of cells staining positive in each block was microscopically (400× magnitude) determined to have stronger staining in the nucleus relative to that in the cytoplasm. The pS33 antibody revealed a significantly higher proportion of positive cells in adult-type GCT relative to that observed in the normal and other tissues (median 57% vs 14%; P < 0.001), but all other antibodies used in this study showed no significant difference in this comparison (Figure 3A). From the preliminary TMA immunostaining, the pS33 antibody showed good specificity (0.909) and sensitivity (0.800) for the detection of adult-type GCTs with a cutoff of 36.5% positive cells by ROC curve analysis and an area under the curve (AUC) of 0.91, significantly higher than other AUC values (Figure 3B). All adult-type GCTs examined showed a heterozygous 402C→G mutation of FOXL2 by DNA sequencing (Supporting Information Figure S4). These results indicated that pS33 FOXL2 (C134W) is a potential biomarker for differentiating adult-type GCT from normal and other tissues by nuclear immunostaining. ROC curve analysis of the total scores of the TMA assay results, calculated as the sum of the proportion score (0−5) of positive cells and the intensity score (0−3 scales) of nuclear immunostaining, yielded AUCs similar to those determined from the percentage of positive cells (Figure 3C). However, the calculated total score had a lower sensitivity to adult-type GCT tissues (n = 36) in the TMA training set (n = 47) than the proportion of positive cells. The FOXL2 intensity in images of the entire TMA appeared to vary widely between different cell types even in the same tissues. This implies that adult-type GCT is strongly related to

Individual Validation of pS33 FOXL2 as a Biomarker for Differentiating Adult-Type GCT from Juvenile-Type GCT and Sertoli−Leydig Cell Tumors

pS33 FOXL2 immunostaining of the TMA indicates that pS33 FOXL2 may be a potential new biomarker for the diagnosis of adult-type GCT by nuclear immunohistochemical staining. Because accurate diagnosis of adult-type GCTs among other SCSTs in the clinical setting is a difficult task, we further investigated the diagnostic performance of pS33 antisera in differentiating adult-type GCTs from other types of SCSTs as well as serous ovarian cancer, which is the most abundant type of ovarian cancer. We tested the pS33 antibody on individual tissue slides from eight cases of adult-type GCTs and five cases each of juvenile-type GCTs, Sertoli−Leydig cell tumors, fibrothecoma, and serous ovarian cancer by immunohistochemical staining. We confirmed that all eight of the adult-type GCTs were heterozygotes for 402C and 402G by DNA sequencing, whereas the samples of other types of cancers did not present this mutation (Figure 4A). Individual adult-type GCTs showed an intense immunostaining for the pS33 FOXL2 (C134W) biomarker in the nucleus relative to the other samples, whereas all five tumor types showed no significant differences in immunostaining of non-pS33 FOXL2 (Figure 4B). These data are similar to the TMA immunostaining results, which suggest that nuclear pS33 FOXL2 immunohistochemical staining is useful to differentiate adult-type GCTs from other types of ovarian cancer such as juvenile-type GCTs, Sertoli−Leydig cell tumors, fibrothecoma, and serous ovarian carcinomas.

■

DISCUSSION Morphological diagnosis of ovarian SCSTs can be challenging due to their various histological patterns.3,7,8,12,13,26−28 Therefore, multiple immunohistochemical markers such as inhibin, calretinin, CD56, melan-A, p53, and epithelial membrane antigen have been employed to accurately diagnose GCTs, but their usefulness is limited.17,26,29−32 More recently, the identification of a common FOXL2 mutation, 402C→G, in adult-type GCTs has allowed a molecular diagnosis of GCT by determining the presence of this somatic mutation.7,8,12,13,15−17,32,33 H

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Journal of Proteome Research

■

In a previous study, we observed the hyperphosphorylation of residue S33 in the C134W FOXL2 mutant in adult-type GCTs, which leads to MDM2-mediated ubiquitination and subsequent degradation.13 In this study, we identified three FOXL2 phosphorylation sites on the S33, Y186, and S238 residues. FOXL2 and its C134W mutant protein phosphorylated on these sites accumulated in the nuclear fraction. Among these sites, pS33 presents biological relevance, as it relates to the transactivation ability and stability of FOXL2 proteins in ovarian granulosa cells.13,34 Despite great advances in mass spectrometry-based proteomics, there are still many unanswered questions about the extent, localization, and site-specific stoichiometry of post-translational modifications and their corresponding pathophysiologies.35 In this study, we evaluated the effectiveness of pS33 as a potential new biomarker for the detection of adult-type GCTs by nuclear immunohistochemical staining of human cells in a TMA platform and in individual tissue slides of different SCSTs, including adult-type GCTs, juvenile-type GCTs, Sertoli−Leydig cell tumors, and fibrothecoma, as well as serous ovarian cancer because it is the most abundant form of ovarian cancer. Our results indicate that pS33 levels could be employed to diagnose adult-type GCTs containing the C402G mutation that produces the C134W FOXL2 mutant protein. As demonstrated herein, an immunostaining assay using the anti-pS33-FOXL2 antibody showed both high sensitivity and high specificity for the discrimination of adult-type GCT compared that using other antibodies. In addition, ROC curve analysis of the percentage of positive cells showed that this method has a substantial diagnostic value (AUC = 0.91; Figure 3B), whereas immunohistochemical analysis using anti-FOXL2 and other peptide antibodies did not effectively discriminate adult-type GCTs from others. In particular, our results suggest that pS33 FOXL2 immunostaining can also be a useful approach for the diagnosis of adult-type GCT among other ovarian SCCTs, including juvenile-type GCTs, Sertoli−Leydig cell tumors, fibrothecoma, and various ovarian cancers of epithelial origin. Further in situ validation of this new diagnostic tool employing additional paraffin-embedded tissue blocks and slides collected from multiple medical facilities is warranted for the application of the pS33 antibody in the detection and diagnosis of adulttype GCTs. Our findings suggest that pS33 FOXL2 (C134W) could serve as an accurate new biomarker for the diagnosis of adulttype GCT, especially when accurate discrimination between GCTs and other types of ovarian cancer is difficult. In a clinical setting, this immunostaining method can be easily and routinely adopted to discriminate adult-type GCTs from other tumor types with high specificity and selectivity. Genotyping of the FOXL2 gene could accurately discriminate adult-type GCTs, virtually all of which present the 402C→G mutation. However, the pS33 FOXL2 (C134W) immunostaining assay has advantages over the use of a DNA extraction procedure followed by PCR-mediated DNA amplification and subsequent sequence determination steps and does not require additional laboratory settings required for genotyping. Therefore, we propose that the pS33 site provides a new, convenient, and useful diagnostic tool to discriminate adult-type GCTs, and its practical utility should be evaluated in the near future.

Article

ASSOCIATED CONTENT

S Supporting Information *

Table S1: Histopathological data of tissue microarray (TMA) composed of adult-type granulosa cell tumors (GCTs) and other types of cancer and normal tissues. Table S2: Tandem mass spectrometry results obtained from monomeric bands of wild-type FOXL2 and C134W mutant proteins in the nuclear fraction of KGN cells. Table S3: Sensitivity and specificity of phospho- and nonphospho-peptide antibodies used for the discrimination of adult-type GCTs and other tissues, with cutoff points analyzed by ROC curve analysis. Figure S1: MS2 and MS3 spectra of phospho-peptides identified from monomeric bands of wild-type and/or C134W mutant FOXL2 proteins in the nuclear fraction of KGN cells. Figure S2: Coomassie bluestained membrane and western blot images of whole cell lysate of KGN cells, pS33 antibody-enriched samples, and non-pS33 antibody-enriched samples obtained from affinity chromatography with immobilized pS33 and non-pS33 antibodies. Figure S3: Images of adult-type GCTs and other tissues in the TMA analyzed by FOXL2 immunostaining with a 1:200 dilution of antibody generated using phospho- or nonphospho-peptide. Figure S4: Genotyping results of adult-type GCTs in the TMA list. Figure S5: Images of the TMA containing different types of ovarian cancer, prostate cancer, and normal tissues analyzed by FOXL2 immunostaining with a 1:200 dilution of antibody generated using phospho-S33 peptide. Figure S6: ROC curve analysis of the percentage of positive cells determined by pS33 FOXL2 immunohistochemical staining in the TMA validation set of adult-type GCTs and other types of ovarian cancer, including serous, endometrioid, clear cell, and undifferentiated ovarian cancers. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/ pr501230b.

■

AUTHOR INFORMATION

Corresponding Authors

*(Y.-H.K.) Phone: 82-53-650-4338; Fax: 82-53-621-4106; Email: [email protected]. *(J.B.) Phone: 82-2-820-5604; Fax: 82-2-816-7338; E-mail: [email protected]. Author Contributions #

D.-S.S. and H.K.O. contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank all subjects who participated by donating tissue for this study. This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT, and Future Planning (2014R1A2A2A01006839, 2013R1A1A2060761), the Ministry of Education (2013R1A1A2061369). It was also supported by the NextGeneration BioGreen 21 Program (PJ01117703), Rural Development Administration, Republic of Korea, and by the National R&D Program for Cancer Control, Ministry for Health and Welfare, Republic of Korea (1220090).

■

REFERENCES

(1) Scully, R. E. Classification of human ovarian tumors. Environ. Health Perspect. 1987, 73, 15−25.

I

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research

granulosa cell tumor patients possessing the FOXL2 mutation. Int. J. Gynecol. Cancer 2010, 20, 1341−3. (19) Dube, V.; Grigull, J.; DeSouza, L. V.; Ghanny, S.; Colgan, T. J.; Romaschin, A. D.; Siu, K. W. Verification of endometrial tissue biomarkers previously discovered using mass spectrometry-based proteomics by means of immunohistochemistry in a tissue microarray format. J. Proteome Res. 2007, 6, 2648−55. (20) Liu, Y.; Luo, X.; Hu, H.; Wang, R.; Sun, Y.; Zeng, R.; Chen, H. Integrative proteomics and tissue microarray profiling indicate the association between overexpressed serum proteins and non-small cell lung cancer. PLoS One 2012, 7, e51748. (21) Marzinke, M. A.; Choi, C. H.; Chen, L.; Shih, I.-M.; Chan, D. W.; Zhang, H. Proteomic analysis of temporally stimulated ovarian cancer cells for biomarker discovery. Mol. Cell. Proteomics 2013, 12, 356−68. (22) Park, M.; Shin, E.; Won, M.; Kim, J. H.; Go, H.; Kim, H. L.; Ko, J. J.; Lee, K.; Bae, J. FOXL2 interacts with steroidogenic factor-1 (SF1) and represses SF-1-induced CYP17 transcription in granulosa cells. Mol. Endocrinol. 2010, 24, 1024−36. (23) Kim, Y. H.; Yu, M. H. Overexpression of reactive cysteinecontaining 2-nitrobenzoate nitroreductase (NbaA) and its mutants alters the sensitivity of Escherichia coli to reactive oxygen species by reprogramming a regulatory network of disulfide-bonded proteins. J. Proteome Res. 2012, 11, 3219−30. (24) Kuo, F. T.; Bentsi-Barnes, I. K.; Barlow, G. M.; Bae, J.; Pisarska, M. D. Sumoylation of forkhead L2 by Ubc9 is required for its activity as a transcriptional repressor of the steroidogenic acute regulatory gene. Cell Signalling 2009, 21, 1935−44. (25) Leymarie, N.; Berg, E. A.; McComb, M. E.; O’Connor, P. B.; Grogan, J.; Oppenheim, F. G.; Costello, C. E. Tandem mass spectrometry for structural characterization of proline-rich proteins: application to salivary PRP-3. Anal. Chem. 2002, 74, 4124−32. (26) Zhao, C.; Vinh, T. N.; McManus, K.; Dabbs, D.; Barner, R.; Vang, R. Identification of the most sensitive and robust immunohistochemical markers in different categories of ovarian sex cord-stromal tumors. Am. J. Surg. Pathol. 2009, 33, 354−66. (27) Stewart, C. J.; Alexiadis, M.; Crook, M. L.; Fuller, P. J. An immunohistochemical and molecular analysis of problematic and unclassified ovarian sex cord-stromal tumors. Hum. Pathol. 2013, 44, 2774−81. (28) Deavers, M. T.; Malpica, A.; Liu, J.; Broaddus, R.; Silva, E. G. Ovarian sex cord-stromal tumors: an immunohistochemical study including a comparison of calretinin and inhibin. Mod. Pathol. 2003, 16, 584−90. (29) Kommoss, F.; Schmidt, D. [Immunohistochemical sex cord markers. Description and use in the differential diagnosis of ovarian tumors]. Pathologe 2007, 28, 187−94. (30) Nofech-Mozes, S.; Ismiil, N.; Dube, V.; Saad, R. S.; Khalifa, M. A.; Moshkin, O.; Ghorab, Z. Immunohistochemical characterization of primary and recurrent adult granulosa cell tumors. Int. J. Gynecol. Pathol. 2013, 31, 80−90. (31) Rabban, J. T.; Zaloudek, C. J. A practical approach to immunohistochemical diagnosis of ovarian germ cell tumours and sex cord-stromal tumours. Histopathology 2013, 62, 71−88. (32) Kommoss, S.; Gilks, C. B.; Penzel, R.; Herpel, E.; Mackenzie, R.; Huntsman, D.; Schirmacher, P.; Anglesio, M.; Schmidt, D.; Kommoss, F. A current perspective on the pathological assessment of FOXL2 in adult-type granulosa cell tumours of the ovary. Histopathology 2013, 64, 380−8. (33) Benayoun, B. A.; Caburet, S.; Dipietromaria, A.; Georges, A.; D’Haene, B.; Pandaranayaka, P. J.; L’Hote, D.; Todeschini, A. L.; Krishnaswamy, S.; Fellous, M.; De Baere, E.; Veitia, R. A. Functional exploration of the adult ovarian granulosa cell tumor-associated somatic FOXL2 mutation p.Cys134Trp (c.402C>G). PLoS One 2010, 5, e8789. (34) Georges, A.; Benayoun, B. A.; Marongiu, M.; Dipietromaria, A.; L’Hote, D.; Todeschini, A. L.; Auer, J.; Crisponi, L.; Veitia, R. A. SUMOylation of the forkhead transcription factor FOXL2 promotes

(2) Colombo, N.; Parma, G.; Zanagnolo, V.; Insinga, A. Management of ovarian stromal cell tumors. J. Clin. Oncol. 2007, 25, 2944−51. (3) Schumer, S. T.; Cannistra, S. A. Granulosa cell tumor of the ovary. J. Clin. Oncol. 2003, 21, 1180−9. (4) Jamieson, S.; Fuller, P. J. Molecular pathogenesis of granulosa cell tumors of the ovary. Endocr. Rev. 2012, 33, 109−44. (5) Shah, S. P.; Kobel, M.; Senz, J.; Morin, R. D.; Clarke, B. A.; Wiegand, K. C.; Leung, G.; Zayed, A.; Mehl, E.; Kalloger, S. E.; Sun, M.; Giuliany, R.; Yorida, E.; Jones, S.; Varhol, R.; Swenerton, K. D.; Miller, D.; Clement, P. B.; Crane, C.; Madore, J.; Provencher, D.; Leung, P.; DeFazio, A.; Khattra, J.; Turashvili, G.; Zhao, Y.; Zeng, T.; Glover, J. N.; Vanderhyden, B.; Zhao, C.; Parkinson, C. A.; JimenezLinan, M.; Bowtell, D. D.; Mes-Masson, A. M.; Brenton, J. D.; Aparicio, S. A.; Boyd, N.; Hirst, M.; Gilks, C. B.; Marra, M.; Huntsman, D. G. Mutation of FOXL2 in granulosa-cell tumors of the ovary. N. Engl. J. Med. 2009, 360, 2719−29. (6) Jamieson, S.; Butzow, R.; Andersson, N.; Alexiadis, M.; UnkilaKallio, L.; Heikinheimo, M.; Fuller, P. J.; Anttonen, M. The FOXL2 C134W mutation is characteristic of adult granulosa cell tumors of the ovary. Mod. Pathol. 2010, 23, 1477−85. (7) Al-Agha, O. M.; Huwait, H. F.; Chow, C.; Yang, W.; Senz, J.; Kalloger, S. E.; Huntsman, D. G.; Young, R. H.; Gilks, C. B. FOXL2 is a sensitive and specific marker for sex cord-stromal tumors of the ovary. Am. J. Surg. Pathol. 2011, 35, 484−94. (8) McCluggage, W. G.; Singh, N.; Kommoss, S.; Huntsman, D. G.; Gilks, C. B. Ovarian cellular fibromas lack FOXL2 mutations: a useful diagnostic adjunct in the distinction from diffuse adult granulosa cell tumor. Am. J. Surg. Pathol. 2013, 37, 1450−5. (9) Pisarska, M. D.; Bae, J.; Klein, C.; Hsueh, A. J. Forkhead L2 is expressed in the ovary and represses the promoter activity of the steroidogenic acute regulatory gene. Endocrinology 2004, 145, 3424− 33. (10) Uhlenhaut, N. H.; Jakob, S.; Anlag, K.; Eisenberger, T.; Sekido, R.; Kress, J.; Treier, A. C.; Klugmann, C.; Klasen, C.; Holter, N. I.; Riethmacher, D.; Schutz, G.; Cooney, A. J.; Lovell-Badge, R.; Treier, M. Somatic sex reprogramming of adult ovaries to testes by FOXL2 ablation. Cell 2009, 139, 1130−42. (11) Kim, J. H.; Yoon, S.; Park, M.; Park, H. O.; Ko, J. J.; Lee, K.; Bae, J. Differential apoptotic activities of wild-type FOXL2 and the adult-type granulosa cell tumor-associated mutant FOXL2 (C134W). Oncogene 2011, 30, 1653−63. (12) Kim, M. S.; Hur, S. Y.; Yoo, N. J.; Lee, S. H. Mutational analysis of FOXL2 codon 134 in granulosa cell tumour of ovary and other human cancers. J. Pathol. 2010, 221, 147−52. (13) Kim, J. H.; Kim, Y. H.; Kim, H. M.; Park, H. O.; Ha, N. C.; Kim, T. H.; Park, M.; Lee, K.; Bae, J. FOXL2 posttranslational modifications mediated by GSK3beta determine the growth of granulosa cell tumours. Nat. Commun. 2014, 5, 2936. (14) Maillet, D.; Goulvent, T.; Rimokh, R.; Vacher-Lavenu, M. C.; Pautier, P.; Alexandre, J.; Pujade-Laurraine, E.; DevouassouxShisheboran, M.; Treilleux, I.; Ray-Coquard, I.; Savina, A. Impact of a second opinion using expression and molecular analysis of FOXL2 for sex cord-stromal tumors. A study of the GINECO group & the TMRO network. Gynecol. Oncol. 2014, 132, 181−7. (15) Geiersbach, K. B.; Jarboe, E. A.; Jahromi, M. S.; Baker, C. L.; Paxton, C. N.; Tripp, S. R.; Schiffman, J. D. FOXL2 mutation and large-scale genomic imbalances in adult granulosa cell tumors of the ovary. Cancer Genet. 2011, 204, 596−602. (16) Gershon, R.; Aviel-Ronen, S.; Korach, J.; Daniel-Carmi, V.; Avivi, C.; Bar-Ilan, D.; Barshack, I.; Meirow, D.; Ben-Baruch, G.; Cohen, Y. FOXL2 C402G mutation detection using MALDI-TOF-MS in DNA extracted from Israeli granulosa cell tumors. Gynecol. Oncol. 2011, 122, 580−4. (17) Kommoss, S.; Anglesio, M. S.; Mackenzie, R.; Yang, W.; Senz, J.; Ho, J.; Bell, L.; Lee, S.; Lorette, J.; Huntsman, D. G.; Blake Gilks, C. FOXL2 molecular testing in ovarian neoplasms: diagnostic approach and procedural guidelines. Mod. Pathol. 2013, 26, 860−7. (18) Lee, S.; Kim, T. H.; Won, M.; Ko, J. J.; Rho, J.; Lee, K.; Bae, J. Absence of a FOXL2 mutation (402C→G) in the blood of adult-type J

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX

Article

Journal of Proteome Research its stabilization/activation through transient recruitment to PML bodies. PLoS One 2011, 6, e25463. (35) Sharma, K.; D’Souza, R. C.; Tyanova, S.; Schaab, C.; Wisniewski, J. R.; Cox, J.; Mann, M. Ultradeep human phosphoproteome reveals a distinct regulatory nature of Tyr and Ser/Thr-based signaling. Cell Rep. 2014, 8, 1583−94.

K

DOI: 10.1021/pr501230b J. Proteome Res. XXXX, XXX, XXX−XXX