Feasibility Study on Measuring Selected Proteins ... - ACS Publications

Article pubs.acs.org/jpr

Feasibility Study on Measuring Selected Proteins in Malignant Melanoma Tissue by SRM Quantification Charlotte Welinder,*,†,‡ Göran Jönsson,† Christian Ingvar,§,∥ Lotta Lundgren,†,§ Bo Baldetorp,† Håkan Olsson,†,§,⊥ Thomas Breslin,† Melinda Rezeli,¶ Bo Jansson,† Thomas Laurell,‡,¶ Thomas E. Fehniger,‡,¶ Elisabet Wieslander,† Krzysztof Pawlowski,†,# and György Marko-Varga‡,¶,□ Departments of †Oncology, ∥Surgery, and ⊥Cancer Epidemiology, Clinical Sciences, and ‡Centre of Excellence in Biological and Medical Mass Spectrometry, Lund University, 221 85 Lund, Sweden § Skåne University Hospital, 221 85 Lund, Sweden ¶ Clinical Protein Science & Imaging, Biomedical Center, Department of Measurement Technology and Industrial Electrical Engineering, Lund University, 221 84 Lund, Sweden # Department of Experimental Design and Bioinformatics, Faculty of Agriculture and Biology, Warsaw University of Life Sciences, Nowoursynowska 159, 02-776 Warszawa, Poland □ First Department of Surgery, Tokyo Medical University, 6-7-1 Nishishinjiku Shinjiku-ku, Tokyo, 160-0023 Japan S Supporting Information *

ABSTRACT: Currently there are no clinically recognized molecular biomarkers for malignant melanoma (MM) for either diagnosing disease stage or measuring response to therapy. The aim of this feasibility study was to develop targeted selected reaction monitoring (SRM) assays for identifying candidate protein biomarkers in metastatic melanoma tissue lysate. In a pilot study applying the SRM assay, the tissue expression of nine selected proteins [complement 3 (C3), T-cell surface glycoprotein CD3 epsilon chain E (CD3E), dermatopontin, minichromosome maintenance complex component (MCM4), premelanosome protein (PMEL), S100 calcium binding protein A8 (S100A8), S100 calcium binding protein A13 (S100A13), transgelin-2 and S100B] was quantified in a small cohort of metastatic malignant melanoma patients. The SRM assay was developed using a TSQ Vantage triple quadrupole mass spectrometer that generated highly accurate peptide quantification. Repeated injection of internal standards spiked into matrix showed relative standard deviation (RSD) from 6% to 15%. All nine target proteins were identified in tumor lysate digests spiked with heavy peptide standards. The multiplex SRM peptide assay panel was then measured and quantified on a set of frozen MM tissue samples obtained from the Malignant Melanoma Biobank collected in Lund, Sweden. All nine proteins could be accurately quantified using the new SRM assay format. This study provides preliminary data on the heterogeneity of biomarker expression within MM patients. The S100B protein, which is clinically used as the pathology identifier of MM, was identified in 9 out of 10 MM tissue lysates. The use of the targeted SRM assay provides potential advancements in the diagnosis of MM that can aid in future assessments of disease in melanoma patients. KEYWORDS: malignant melanoma, cancer tissue, protein sequencing, proteomics, genes, mRNA, mass spectrometry

1. INTRODUCTION

early detection, identifying disease progression, or monitoring treatment of MM. Novel targeted drugs and immunotherapy with antibodies have improved the possibilities to treat patient with disseminated MM. Both Ipilimumab and BRAF inhibitors have shown prolonged survival in phase II controlled trials.3−7 In the search for new disease markers, the developments and deliveries within the genomics and proteomics research fields are essential.

Metastatic malignant melanoma is a disease with unfavorable prognosis and has one of the highest incidence rates globally. In Sweden approximately 2800 new patients are diagnosed every year and almost 500 patients die of disseminated melanoma disease annually. The 5-year survival rate in metastatic melanoma is around 5% and the median survival is only 6− 10 months.1,2 At the time of diagnosis 10−15% of the patients are diagnosed with disseminated disease and hence a poor prognosis. In contrast to most other malignancies, MM is also common in young people and not only in the elderly group. There are no blood or tissue biomarkers currently available for © 2014 American Chemical Society

Received: August 27, 2013 Published: February 3, 2014 1315

dx.doi.org/10.1021/pr400876p | J. Proteome Res. 2014, 13, 1315−1326

Journal of Proteome Research

Article

Table 1. Chromosome Localization and Nucleic Acid Sequence Positions for the Selected Target Peptides gene name

gene ID (NCBI)

chromosome localization

nucleic acid sequence position

peptide sequence

C3

718

19p13.3-p13.2

CD3E

916

11q23.3

DPT

1805

1q12-q23

MCM4 PMEL S100A8

4173 6490 6279

8q11.21 12q13-q14 1q21.3

S100A13

6284

1q21.3

S100B TAGLN2

6285 8407

21q22.3 1q21-q25

23224−23251 9573−9620 23186−23224 10928−10934 9880−9924 14908−14931 28157−28189 11168−11219 4989−5015 952−971 723−746 678−707 7711−7736 7738−7773 15084−15117 2810−2855 5133−5156 6181−6222

TGLQEVEVK TELRPGETLNVNFLLR SSLSVPYVIVPLK DLYSGLNQR ERPPPVPNPDYEPIR YFESVLDR GATTTFSAVER TSVLAAANPIESQWNPK NQDWLGVSR GADVWFK GNFHAVYR ALNSIIDVYHK DSLSVNEFK ELVTQQLPHLLK SLDVNQDSELK ELINNELSHFLEEIK ENFQNWLK DDGLFSGDPNWFPK

markers of autoinflammation, where both tissue and serum concentrations correlates to disease activity, during both local and systemic inflammation.21 The CD3E is part of the T cell receptor-CD3 complex. CD3E is particularly important in T cell development and crucial for normal immune function.22−24 Dermatopontin is an extracellular matrix protein, possibly involved in cellular adhesion, and has been implicated in cancer, such as having a role in metastasis25 and therapy resistance.26 The MCM4 belongs to the Mini chromosome maintenance complex protein family, unwinds DNA at the initiation of replication during cell division, and is a critical mediator of cellular proliferation and tumor growth. This increased tumor cell proliferation is a hallmark of malignancy and also a prognostic factor in several tumor types.27 Within recent MM studies, high expression of MCM4 has been associated with poor prognosis for melanoma patients.16 PMEL is involved in pigment formation, and levels of PMEL have been proposed as complementary marker to, e.g., S100B, in order to strengthen the MM diagnosis.28 Gene expression profiling has shown S100A8 to be more expressed in MM cells than in normal epithelial melanocytes.29 S100A13 has been suggested to be a new angiogenic and prognostic marker in melanoma. S100A13 is expressed in melanocytic lesions when angiogenesis is switched on and may cooperate with VEGF-A in supporting the formation of new blood vessels, favoring the shift from radial to vertical growth.30 Transgelin-2 is a homologue of transgelin and a member of family of actin-binding proteins. Transgelin-2 has shown to be overexpressed in colorectal cancer and has been suggested as new biomarker for prediction, progression, and prognosis for the disease.31 The S100B protein target was included into the SRM assay, since it is the tissue biomarker that is currently used in some clinics for pathological MM diagnosis.32,33 In Table 1, the nine prospective protein biomarkers used in this study are listed along with corresponding target peptide sequences and chromosome localization. The assay panel was employed on tissue lysates in a small MM cohort, and the targeted proteins were quantified.

The molecular profiling efforts of MM have improved the understanding of disease biology at the molecular level.8−13 A number of older markers associated with MM (e.g., S-100B and 5-S-cysteinyldopa) are under investigation, but their relevance to melanoma progression, clinical outcome, and the selection of optimal treatment strategies still needs to be established. Gene signatures for individual patients have been proposed to classify MM into distinct subtypes with different clinical outcome.14−18 For example. Winnepenninckx et al. analyzed 58 primary cutaneous melanoma and identified a 254 gene expression signature associated with a 4-year distant metastasis-free survival.16 Recently, Jönsson et al. and Harbst et al. identified four tumor subgroups based on unsupervised clustering of gene expression data. These groups could be significantly linked to patient outcome.14,15 These gene signatures might be useful for identifying new potential candidates for follow-up work at the protein level. Quantification and understanding of protein markers is obviously relevant since proteins are in most cases the functional molecule in the cell. Further, proteins are subjected to a wide variety of chemical modifications after translation, and these post-translational modifications are critical to the protein function. Finally most drug target interactions take place with proteins. From the published gene expression data eight proteins were selected:14,15 complement 3 (C3), T-cell surface glycoprotein CD3 epsilon chain E (CD3E), dermatopontin, minichromosome maintenance complex component (MCM4), premelanosome protein (PMEL), S100 calcium binding protein A8 (S100A8), S100 calcium binding protein A13 (S100A13), and transgelin-2. Complement 3 is a component of the human immune system, and its degradation leads to inhibition of the complement pathway. Recent findings in the literature show that complement proteins C3, C4, and C5 may aid tumor growth through immunosuppression, and Markiewski et al. suggest an insidious relationship between the complement system and cancer.19 Complement 3 may also be considered as an acute phase reactant protein and has been reported as a biomarker in many diseases including cancer.20 The three S100 protein forms (S100A8, S100A13, and S100B) have been targeted. The S100 proteins have been suggested as new 1316



Article

2. MATERIALS AND METHODS

and 14 peptides in each mixture, respectively. The theoretical concentration of each peptide was 50 fmol/μL. The transition lists were created in Skyline v1.2 software (MacCoss Lab). Primarily, high numbers of transitions, all possible y-ion series that matches the criteria (from m/z > precursor-2 to last ion-2, precursor m/z exclusion window: 20 Th), were selected for each peptide at both 2+ and 3+ charge states. The peptide mixtures were analyzed by nano LC−MS/MS using a TSQ Vantage triple quadrupole mass spectrometer equipped with an Easy n-LC II pump (Thermo Scientific, Waltham, MA). The samples were injected onto an Easy C18-A1 precolumn (Thermo Scientific, Waltham, MA), and following online desalting and concentration the tryptic peptides were separated on a 75 μm × 150 mm fused silica column packed with ReproSil C18 (3 μm, 120 Å from Dr. Maisch GmbH, Germany). Separations were performed in a 45-min linear gradient from 10% to 35% acetonitrile containing 0.1% formic acid, at the flow rate of 300 nL/min. The MS analysis was conducted in positive ion mode with the spray voltage and declustering potential set to 1750 V and 0, respectively. The transfer capillary temperature was set to 270 °C, and tuned Slens value was used. SRM transitions were acquired in Q1 and Q3 operated at unit resolution (0.7 fwhm), and the collision gas pressure in Q2 was set to 1.2 mTorr. The cycle time was 2.5 s in the nonscheduled methods and 1.5 s in the scheduled methods. The best transitions (2−3 per precursor) were selected by manual inspection of the data in Skyline, and scheduled transition lists were created for the final assays. Collision energies were optimized for each transition. The collision energy was ramped round the predicted value in 3 steps on both sides, in 2 V increments. The selected transitions were tested in real matrix also by spiking the heavy peptide mixtures into human malignant melanoma tissue digests.

2.1. Clinical Samples

Ethical approval for this study was granted by Institutional Review Board at Lund University Hospital, approval number DNR 191/2007, 101/2013. Lymph nodes containing metastasis of malignant melanoma (stage III) were obtained surgically from 10 MM patients undergoing treatment at the Skåne University Hospital, Sweden. The fresh specimens were carefully divided into two portions by a pathologist. One portion of the tumor containing nodes was fixed in formalin, paraffin embedded, and sectioned for pathological evaluation. Another portion of the node judged tumor-rich was snap frozen and stored at −80 °C in the local malignant melanoma Biobank until use with the informed consent of each patient donor. The frozen specimens were used as described below for both protein expression analysis and histological comparisons of the same specimens. The clinical information describing the respective patients is summarized in Table 2. Table 2. Clinical Information of Patient Characteristicsa

a

tumor

sex

age at metastases

MM35 MM98 MM504 MM687 MM787 MM812 MM813 MM825 MM829 MM835

male male male male male male female female male female

55 75 54 74 81 51 54 66 55 36

age at primary melanoma

Breslow class

Clark

stage

54 73 NA 72 78 NA 54 64 49 32

3 4 NA 1 2 NA 2 2 1 3

4 4 NA 2 4 NA 3 4 2 3

3 3 NA 3 3 3 3 3 3 3

2.4. Standard Curves of Synthetic Peptides

NA: not available.

A dilution series of heavy labeled synthetic peptide mixtures (in the theoretical concentration range of 6.25−200 fmol/μL) in tumor digest containing a constant amount of nonlabeled synthetic peptides (theoretical concentration of 100 fmol/μL) for characterization of the assay linearity were analyzed in triplicate. Calibration curves were generated by linear regression analysis of the peak area ratios (heavy/light) versus concentration ratios for all targeted peptides.

2.2. Sample Preparation

Proteins were extracted from frozen specimens obtained as above with AllPrep DNA/RNA/Protein Mini Kit (Qiagen) according to the manufacturer’s instructions. The extracted proteins were precipitated with ice-cold acetone to a final concentration of 80% acetone. Samples were incubated for 30 min at −20 °C followed by centrifugation at 16000g for 2 min. The supernatant was removed, and the protein pellets were allowed to air-dry. The dried protein pellets were resolved in 8 M urea in 50 mM ammonium bicarbonate (pH 7.6). Protein concentration was determined by the BCA method (Pierce). Of total proteins, 150 μg was reduced with 10 mM dithiolthreitol (1 h at 37 °C) and alkylated using 40 mM iodoacetamide (30 min, kept dark at room temperature) followed by buffer exchange with 50 mM ammonium bicarbonate buffer (pH 7.6) by using a 10 kDa cutoff spin filter (YM10 filter, AMICON). The samples were digested with sequencing grade trypsin (Promega, Madison, WI) in a ratio of 1:120 w/w (trypsin:protein) overnight at 37 °C.

2.5. Data Analysis

Data sets were imported into Skyline (version 1.2, http:// proteome.gs.washington.edu/software/skyline),34 and peaks were automatically integrated. After automatic integration of the data sets, the data were also manually inspected. Integration of the peaks was adjusted when signals were not intense and the software could not reliably determine the peaks. Interferences with the matrix, detector saturations, and variable peak area ratios in replicate samples were also investigated. Data from the individual tumor lysates are presented as mean of triplicates measurements ± standard deviation. 2.6. Histology of Tumors

2.3. Mass Spectrometric Analysis of the Synthetic Peptide Standards

For each patient, a portion from the sample analyzed for protein expression was evaluated histologically to determine the level of tumor development. Frozen tissue samples were sectioned on a cryostat into 6 μm thick sections, placed upon glass slides, dried at 37 °C for 30 min, and fixed with 100% methanol for 5 min. The sections were stained with Mayer’s hematoxylin-eosin,35,36 where protein-rich cytoplasma stains

SRM Assay Development. Crude peptides, both light and heavy, were supplied by Thermo Scientific (Ulm, Germany). The heavy petides were isotopically labeled on the C-terminal arginine or lysine residue (13C, 15N). Two mixtures were created from the crude peptides (Supplementary Table 1), 12 1317



Article

Figure 1. Histochemical images of the 10 lymph node metastasis. Frozen tumor samples were cryosectioned and stained with Mayer’s haematoxylineosin. The nuclei stain blue and protein-rich cytoplasma stains dark pink, while cytoplasma that is actively synthesizing proteins stains rich purple. The brown pigment seen scattered within clusters corresponds to focal hyperexpression of melanin by groups of melanoma cells.

required for building the disease understanding that is necessary for developing the next generation of diagnostics and therapy for MM patients. In this study, tumor tissue from the South Swedish MM biobank in Lund were utilized.37,38 Typical tissue morphology of biopsies derived from the MM patients’ tumors used in this

dark pink while cytoplasma that is actively synthesizing protein stains rich purple and the nucleus stains blue.

3. RESULTS AND DISCUSSION Clinical sample collections with detailed pathological characterization and clinical evaluation are key strategic resources 1318



Article

Table 3. Proteotypic Peptide Sequence and Selected SRM Transistions for the 10 Analysed Proteins accession no. P01024

P07766

Q07507

protein

peptide sequence

Complement 3

T-cell surface glycoprotein CD3 epsilon chain E

Dermatopontin

Q1

Q3

TGLQEVEVK

501.7769++

TELRPGETLNVNFLLR

624.6828+++

SSLSVPYVIVPLK

701.4212++

DLYSGLNQR

533.2698++

ERPPPVPNPDYEPIR

592.6407+++

YFESVLDR

514.7560++

GATTTFSAVER

570.2882++

731.3934+ [y6] 603.3348+ [y5] 474.2922+ [y4] 875.5098+ [y7] 761.4668+ [y6] 662.3984+ [y5] 1114.6871+ [y10] 1027.6550+ [y9] 928.5866+ [y8] 837.4213+ [y7] 674.3580+ [y6] 587.3260+ [y5] 889.4414+ [y7] 550.7722++ [y9] 676.3777+ [b6] 718.3730+ E [y6] 589.3304+ [y5]c 809.4152+ [y7] 708.3675+ [y6] 1283.6379+ [y11] 1212.6008+ [y10] 1098.5578+ [y9] 832.4312+ [y7] 717.4042+ [y6] 694.3559+ [y5] 579.3289+ [y4] 480.2605+ [y3] 645.3467+ [y5] 508.2878+ [y4] 437.2507+ [y3] 974.5306+ [y8] 887.4985+ [y7] 774.4145+ [y6] 836.4512+ [y7] 723.3672+ [y6] 636.3352+ [y5] 1077.6415+ [y9] 848.5352+ [y7] 607.3926+ [y5] 1047.4953+ [y9] 932.4684+ [y8] 833.3999+ [y7] 793.4147++ [y13] 736.8726++ [y12] 679.8512++ [y11] 835.4461+ [y6] 688.3777+ [y5] 560.3191+ [y4] 1194.5578+ [y10] 1047.4894+ [y9] 960.4574+ [y8]

P33991

Minichromosome maintenance complex component 4

TSVLAAANPIESQWNPK

913.4758++

P40967

Premelanosome protein

NQDWLGVSR

537.7700++

P05109

S100 calcium binding protein A8

GADVWFK

411.7109++

GNFHAVYR

482.2434++

ALNSIIDVYHK

636.8510++

DSLSVNEFK

519.7587++

ELVTQQLPHLLK

709.9219++

SLDVNQDSELK

624.3093++

Q99584

S100 calcium binding protein A13

P04271

S100 calcium binding protein B

ELINNELSHFLEEIK

609.9878+++

P37802

Transgelin-2

ENFQNWLK

539.7694++

DDGLFSGDPNWFPK

797.8623++

study are shown in Figure 1. The lymph node tumors showed almost complete replacement of normal follicular architecture by metastatic melanoma cells. The brown pigment seen scattered within clusters corresponds to focal hyperexpression of melanin by groups of melanoma cells. The histopathological characterization of metastatic melanomas has been extensively presented in recent studies, where similar morphologies were observed.39

However, from this initial phase of development, data from a larger clinical cohort with detailed pathological analysis would be required to provide an extended relevance to MM. 3.1. Selection of Transitions for SRM

The peptide sequences of the nine targeted proteins used in this study to build the SRM assay (C3, CD3E, dermatopontin, MCM4, PMEL, S100A8, S100A13, S100B and transgelin-2) were identified using Skyline software. Peptide sequences for all 1319



Article

Figure 2. continued

1320



Article

Figure 2. continued

1321



Article

Figure 2. (A) Resulting SRM mass spectra (left, endogenous specific transitions for the different peptides corresponding to the protein used for quantification and right, corresponding heavy peptides) and (B) histogram of protein quantification from 10 MM patients. Data from the individual tumor lysates are presented as mean of triplicate measurements ± standard deviation.

plotted against their theoretical concentrations. Corresponding calibration curves were generated by linear regression analysis of the peak areas (heavy/light) for all targeted peptides. A linear regression was found for all peptide sequences to be higher than 0.99 (R 2 values) within the investigated concentration range (data not shown). By repeated series of analysis using this SRM assay variations were found to range between 6% and 15%. Generation of SRM data by repeated injections typically gave chromatographic retention time index RSD values

Feasibility Study on Measuring Selected Proteins ... - ACS Publications

Recommend Documents