- Email: [email protected]
An antibody-like peptide that recognizes malignancy among thyroid nodules
Cancer Letters xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Cancer Letters journal homepage: www.elsevier.com/locate/canlet
An antibody-like peptide that recognizes malignancy among thyroid nodules Carolina Fernandes Reis a, Ana Paula Carneiro a, Carlos Ueira Vieira b, Patrícia Tiemi Fujimura b, Elaine Cristina Morari a, Sindeval José da Silva c, Luiz Ricardo Goulart b,⇑, Laura Sterian Ward a,⇑ a b c
Laboratory of Cancer Molecular Genetics, Faculty of Medical Sciences (FCM), University of Campinas (UNICAMP), Campinas, SP, Brazil Laboratory of Nanobiotechnology, Federal University of Uberlandia, Institute of Genetics and Biochemistry (INGEB), Uberlandia, MG, Brazil Department of Head and Neck Surgery, Clinical Hospital of Federal University of Uberlândia (HC-UFU), Uberlândia, MG, Brazil
a r t i c l e
i n f o
Article history: Received 28 September 2012 Received in revised form 1 February 2013 Accepted 18 February 2013 Available online xxxx Keywords: Thyroid cancer Phage display Biomarker Papillary tumor
a b s t r a c t There is an urgent need for biomarkers to identify malignant thyroid nodules from indeterminate follicular lesions. We have used a subtractive proteomic strategy to identify novel biomarkers by selecting ligands to goiter tissue from a 12-mer random peptide phage-displayed library using the BRASIL method (Biopanning and Rapid Analysis of Selective Interactive Ligands). After three rounds of selection, two highly reactive clones to the papillary thyroid tumor cell line NPA were further evaluated, and their specific binding to tumor proteins was confirmed using phage-ELISA. The antibody-like peptide CaT12 was tumor-specific, which was further tested by immunohistochemistry against TMAs (tissue microarrays) comprised of 775 human benign and malignant tissues, including 232 thyroid nodular lesions: 15 normal thyroid tissues, 53 nodular goiters (NG), 54 follicular adenomas (FA); 69 papillary thyroid carcinomas (PTC); and 41 follicular carcinomas (FC). CaT12 was able to identify PTC among thyroid nodular lesions with 91.2% sensitivity and 85.1% specificity, despite its non-specificity for thyroid tissues. Additionally, the CaT12 peptide helped characterize follicular lesions distinguishing the follicular variant of PTC (FVPTC) from FA with 91.9% accuracy; FVPTC from NG with 83.1% accuracy; FVPTC from the classic PTC with 57.7% accuracy; and FVPTC from FC with 88.7% accuracy. In conclusion, our strategy to select differentially expressed ligands to thyroid tissue was highly effective and resulted in a useful antibody-like biomarker that recognizes malignancy among thyroid nodules and may help distinguish follicular patterned lesions. Ó 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Thyroid cancer incidence rates have steadily increased over the past few decades all over the world [1] and 56,460 new patients are estimated to be diagnosed during 2012 in the USA [2]. The accessible use of sensitive imaging detection methods is mainly responsible for this increased incidence [3]. However, there are evidences that other factors, including the exposure to environmental carcinogenic factors, may have also contributed to this phenomenon [3,4]. We have actually experienced real epidemics of thyroid nodules referred to specialized diagnosis, and the cytological exam of fine-needle cell aspiration is the recommended meth⇑ Corresponding authors. Addressess: Laboratory of Nanobiotechnology, Institute of Genetics and Biochemistry, Federal University of Uberlandia (UFU), Campus Umuarama, Bl 2E, R. 248, 38400-902 Uberlandia, MG, Brazil (L.R. Goulart), Laboratory of Cancer Molecular Genetics, School of Medical Sciences, University of Campinas (FCM-UNICAMP), Rua Tessalia Vieira de Camargo 126, Barão Geraldo, 13083-887 Campinas, SP, Brazil. Tel.: +55 19 35218954; fax: +55 19 35218925 (L.S. Ward). E-mail address: [email protected] (L.S. Ward).
od by current guidelines to distinguish benign from malignant lesions [5]. In most cases, the clinical features of the patients based on ultrasonography and cytological findings enable appropriate case management [5]. However, up to one third of the nodules submitted to fine needle aspiration biopsies may present an ‘‘indeterminate’’ cytology, a pattern that remains burdened with interobserver variability and uncertainty regarding management [6]. Even the most experienced pathologists may have difficulties distinguishing cases of follicular variants of papillary thyroid carcinomas from simple benign hyperplasia, and many cases of follicular adenomas are submitted to unnecessary surgeries because it is virtually impossible to eliminate the risk of a follicular carcinoma in cytological analyses [7]. A series of molecular markers for malignancy have been proposed and many of them appear to be effective in reducing uncertainties [8,9]. However, clinical problems concerning diagnosis of thyroid malignancies are not solved, and certainly are not appropriate to large populations. In the present investigation, we have used a different approach to distinguish thyroid malignancies. We employed phage display
0304-3835/$ - see front matter Ó 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.canlet.2013.02.039
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
2
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
technology that has been successfully used to select targets for various diseases [10–12], including many cancer types [13–24], and, by using a direct selection (BRASIL) against goiter tissue, we were able to obtain a highly reactive and discrimination marker for thyroid cancer diagnosis.
anti-M13 monoclonal antibody (Amersham Pharmacia Biotech Benelux, Roosendaal, The Netherlands) using a 1:5000 dilution in blocking buffer. The peroxidase staining reaction was developed in the presence of SIGMAFAST™ OPD (o-phenylenediamine dihydrochloride) tablets (Sigma–Aldrich, São Paulo, SP, Brazil), and then incubated at room temperature for 5 min, stopped with diluted acid solutions, and the optical density (OD) values were measured on a microplate reader (TPREADER – Thermo plate) at 492 nm.
2. Materials and methods 2.5. Recombinant phage and synthetic peptide immunohistochemistry (IHC) Research Ethics Committees of three institutions approved this study: Federal University of Uberlândia (UFU), State University of Campinas (UNICAMP) and Fundação Antônio Prudente (Hospital AC Camargo, SP). All participants signed informed consent forms. 2.1. Cell culture and patients’ samples We have used the TPC-1 (Human thyroid papillary cancer) and NPA (Human papillary cancer) cells, gently provided by Prof. J.M. Cerutti, from the Federal University of São Paulo. In addition, we have developed a primary human cell culture using tissue obtained from a patient submitted to surgical removal of a multinodular goiter at the Surgery Center of UFU Hospital. Thyroid cancer cell lines and goiterderived cells were grown in RPMI 1640 medium supplemented with penicillin, streptomycin, and 10% fetal bovine serum (Nutricell, Campinas-SP) and incubated at 37 °C in a humidified atmosphere containing 5% CO2. Fresh tissues from patients with histologically-proved classic PTC and hyperplasia were obtained from the Federal University of Uberlândia (UFU) Clinics’ Hospital, and immediately frozen until use for protein binding testing. The selected clone was tested by immunohistochemistry in formalin fixed, paraffin embedded tissues from 775 patients reviewed for diagnostic confirmation in order to select the most representative areas designed to build tissue microarrays (TMA, Beecher InstrumentsÒ, Silver Springs, MD, USA). A first TMA was composed by 232 patient samples: 15 normal thyroid tissues, 53 nodular goiters, 54 follicular adenomas, 69 papillary thyroid carcinomas (36 cases of the classical form and 33 of the PTC follicular variant) and 41 follicular carcinomas. A second TMA was designed to validate our results, and was constituted by 281 other cancer tissues (189 breast, 57 prostate and 35 kidney), 41 benign lesions (12 nevus atypical and 29 nevus) and 221 non-neoplastic tissues (including normal stomach, ovary, breast, prostate, kidney, heart, muscle, brain, liver, lung and others). Patients’ clinical information was obtained from their clinical records. Aggressiveness at diagnosis was ascertained using the Tumor Node Metastasis and stage classification systems for DTC as recommended by the American Thyroid Association guidelines [5]. For statistical purpose we grouped patients classified as stage II, III and IV in one class named ‘‘higher stages’’. Patients were followed with periodic total body scans, serum TSH and thyroglobulin (Tg) measurements, according to a standard protocol that included X-ray, ultrasonography, computed tomography scan and other eventual procedures to detect distant metastasis for a period of 12– 298 months (43.50 ± 33.29 months). Patients presenting high non-stimulated serum Tg levels (>2 mg/dl) were submitted to image evaluation. We defined tumors as persistent/recurrent and/or presenting long distance metastasis, according to the aforementioned parameters. 2.2. Biopanning in cell surface Biopanning was performed in three rounds using BRASIL method (Biopanning and Rapid Analysis of Selective Interactive Interactive Ligands) [25] to verify phage binding to intact cells. 2.3. DNA sequencing of selected phage clones After three rounds, 96 blue colonies were randomly selected and their phage single-strand DNA were isolated by the iodide buffer extraction procedure (Instruction Manual Ph.D.-12 Phage Display Peptide Library Kit) and analyzed with MegaBace 1000 Genetic Analyzer (Amersham Biosciences) automatic capillary sequencer. The nucleotide sequence of the gene III insert was sequenced with 96 gIII sequencing primer 50 -CCC TCA TAG TTA GCG TAA CG-30 by automated dye terminator cycle sequencing. The amino acid sequence of the insert was deduced from the nucleotide sequence using DNA2PRO software from RELIC program [26,27]. 2.4. ELISA (enzyme-linked immunosorbent assay) After extraction of papillary cancer and goiter proteins with extraction buffer (Tris–HCl 20 mM pH7.2, EDTA 10 mM, EGTA 2 mM, sucrose 250 mM, DTT 1 mM, Benzamidine 1 mM) proteins were diluted in sodium carbonate buffer (10 lg/ml in 0.1 M NaHCO3) and immobilized into a 96-wells plate (Nunc Immuno Maxisorp, Roche Diagnostics) by incubation it overnight at 4 °C. That plate was subsequently washed three times with PBST 0.1%, and blocked with BSA 3% (blocking buffer) for three hours at 37 °C. Plate wells were washed again three times, and each well was incubated with 100 ll of phages supernatants for one hour at 37 °C. After five washes, recombinant phage particles were detected with peroxidase-conjugated
After deparaffinizing the tissues, the endogenous peroxidase was inactivated by incubating the sections with 3% hydrogen peroxide for three times, 5 min each. The tissue sections were then rinsed thoroughly in distilled water. For the biomarker exposure, sodium citrate buffer (0.1 M) was heated to 95 °C steamer, the slides were submerged into the buffer for 35 min, washed for 5 min in running water, and blocked with 3% skim milk (Molico Milk, Nestle, Brazil; 3% w/v in distilled water) for 30 min at room temperature. Incubation with the primary marker (phage or synthetic peptide) was then performed for 30 min at 37 °C and overnight at 4 °C. The recombinant phage CaT12, the synthetic peptide (Invitrogen, Carlsbad, CA, USA), and the wild phage (no peptide) were diluted into 100 ll of the Novocastra Universal IHC Blocking/Diluent 1000 (Leica Microsystems Inc, Buffalo Grove, IL, USA) containing 6 1011 pfu (viral particles/slide) or 1 lg synthetic peptide. As negative controls, duplicate sections were incubated with 3% skim milk proteins instead of specific primary antibodies. The sections were washed four times in PBS for 5 min each followed by one hour incubation with peroxidase-conjugated mouse anti-M13 secondary antibody or streptavidin. After a washing step in PBS, peroxidase activity was visualized by incubation in 3,30 diaminobenzidine tetrahydrochloride (Liquid DAB, Sigma, St. Louis, USA) for 5 min at room temperature. Sections were counterstained with hematoxylin. Positive and negative controls were run in the same reaction batch. Slides were independently scored by two experienced researchers (CFR and ECM), both blinded to tumor features, and confirmed by an experienced pathologist (JV). 2.6. Immunohistochemistry staining index Signals were considered positive when reaction products were localized in the expected cellular component. The staining index value was obtained by the sum of intensity and distribution scores, and values of a staining index equal or greater than 5 were considered positive. The criteria employed for the intensity score were: 0, no staining; 1, weak; 2, moderate; 3, strong. The criteria employed for the distribution score were: 0, no staining; 1, staining of <25% of the cells; 2, between 25% and 50% of the cells; 3, between 50% and 75% of the cells; and 4, staining of >75% of the cells. In addition, we employed the Automated Cellular Imaging System ACIS-III (Chroma Vision Medical Systems, Inc, DAKO, San Juan Capistrano, CA, USA) for IHC quantification. Briefly, each tissue spot was digitalized to the systems’ software, and a computer algorithm, considering the intensity and extension of the brown staining, attributed numerical values. 2.7. Streptavidin-bead precipitation assay TPC-1 cells were lysed according to the instructions of the RIPA Buffer (Thermo scientific, Rockford, IL, USA) kit; a protease inhibitor cocktail (1:100 dilution) (Sigma–Aldrich, São Paulo, SP, Brazil) was added and the protein concentration was determined using the BCA Protein Assay Kit (Thermo scientific). The synthetic peptide (1 mg/ml) was N-terminal labeled with biotin, and mixed with 200 ll packed Streptavidin-beads (Thermo scientific) for 1 h at 4 °C, and incubated with 500 lg of protein extracts for 4 h at room temperature. After washing four times with TBS (Tris Buffered Saline), the proteins were eluted from the beads by addition 0.1 M glycine HCl buffer, pH 3.0, twice for 10 min each. The eluate was quantified by the BCA Protein Assay Kit prior to proteins sequencing. 2.8. Protein sequencing by LC-MS/MS and target identification Eluted proteins were precipitated out of solution using the ProteoExtract kit (Calbiochem) and the protein pellet was left to dry overnight in a sterile fumehood. The lyophilized pellet was then resuspended in 50 mM ammonium bicarbonate (pH 8.0) and subjected to an in-solution tryptic digestion (Mike Myers, Cold Spring Harbor modified by Brett S. Phinney, UC Davis Proteomics Core, Davis, CA, USA). Digested peptides were then de-salted using aspire tips (Thermo-Fisher Scientific, RP30 tips) before being resuspended in loading buffer. Digested peptides were analyzed using a LTQ-FT (Thermo Fisher Scientific) coupled with a MG4 paradigm HPLC (Michrom, Auburn, CA, USA). The samples were loaded onto a Michrom cap trap (0.5 2 mm) to be de-salted online. The peptides were then separated using a Michrom Magic C18AQ (200 lm 150 mm) reversedphase column and eluted using a gradient for 60 min. Collision induced dissociation was applied to the peptide samples and data was acquired with an isolation width of 1, normalized collision energy of 35, and a resolution of 50,000. The spray voltage on the Michrom captive spray was set to 1.8 kV with a heated transfer capillary temperature of 200 °C.
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
3
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx Raw data was analyzed using X!Tandem and visualized using Scaffold (Proteome Software, version 3.01). Samples were searched in Uniprot human sequences databases, appended with the cRAP (commonly found laboratory contaminants) and reverse decoy databases.
ELISA was performed on cancer and goiter cases in order to analyze the specific protein binding properties and to select the best phage clones for immunohistochemistry testing. Clones CaT12 and CaT18 presented the highest absorbances for cancer proteins (Fig. 1A) when compared to goiter proteins (P < 0.0001). A wild phage used as control (M13) showed low reactivity. CaT12 and CaT18 presented a consensus sequence showed on Fig. 1B.
phage was used as a negative control (Fig. 2 and Table 1). The CaT12 clone showed higher positivity than M13 wild phage in thyroid cancer cases, as demonstrated in Fig. 2. Furthermore, the CaT12 synthetic peptide showed an increased signal (Fig. 2, A1), which was not followed by CaT18 peptide that presented a marginal positivity (data not shown). The cut-off values for CaT12 phage and CaT12 peptide were chosen based on the ROC curve analyses (Fig. 2B2; Fig. 2B4, respectively), which generated diagnostic parameters for the comparisons between cancer groups (Table 2). A significant difference (P < 0.0001) between cancer and benign diseases was found (Fig. 2B1 and B3), which was superior for the CaT12 phage as a biomarker. We observed an overall higher positivity using CaT12 peptide in goiters, FA and FC, whereas CaT12 phage was more frequently positive in CPTC and FCPTC, as demonstrated in Fig. 2. Visual and ACIS analysis generated very comparable results as demonstrated in Table 2 where we demonstrate the utility of the employed CaT12 phage and CaT12 peptide in the characterization of follicular patterned lesions. Both visual and ACIS were valid analytical methods, and presented similar values. In fact, both CaT12 phage and synthetic peptide were able to identify malignancy among follicular patterned lesions (goiter, FA, FVPTC and PTC) and discriminate these lesions with considerable accuracy, as demonstrated in Table 3. CaT12 was inversely correlated with patient’s age (Spearman r = 0.2469, p = 0.0197) and was directly correlated with the occurrence of metastasis (Spearman r = 0.3585, p = 0.0039) at follow-up. We were not able to establish any other association between CaT12 and any clinical or tumor feature or with the patients outcome. We further employed CaT12 peptide in other tissues, as shown in Fig. 3, demonstrating positivity in malignant tumors, and a weak positivity in benign and normal tissues as presented in Fig. 3C and Table 4.
3.3. Biomarker validation: phage and synthetic peptide immunohistochemistry
3.4. Immunohistochemistry validation
To confirm the ability of the selected clones to target cancer cells, phage clone CaT12 and CaT18 were amplified, purified and directly tested in IHC assays of the TMA built with cases of thyroid cancer, adenoma, goiter and normal thyroid tissues. The M13 wild
To validate CaT12 peptide we used different benign and malign tissues as demonstrated in Fig. 3. We found a weak or negative staining in benign lesions and in normal tissues, whereas tumors were positive, as demonstrated in Table 4.
2.9. Statistical analysis We employed GraphPrism 5.0 software to analyze the reactivity of the selected phage and the synthetic peptide. ANOVA test with Bonferroni Post-hoc test, was used to verify differences between means of OD value obtained by ELISA with phage clones. Differences between benign and papillary cancer tissues for both biomarker CaT12 phage and CaT12 peptide were tested with the Mann–Whitney U test (continuous variables and non-parametric analyses). Receiver operating characteristics (ROC) curves were constructed to assess sensitivity, specificity, accuracy, and respective areas under the curves (AUCs) with 95% confidence interval (CI). For visual analysis and ACIS imaging, Fisher’s test was employed to determine the relationship between CaT12 phage or synthetic peptide for the different disease groups in the TMA. We have used sensitivity and specificity parameters in the Fisher’s test to choose the cut-off value in the ROC curve analysis. Probability below 0.05 was considered significant.
3. Results 3.1. Biopanning and specific enrichment of NPA-bound phages The amount of phages obtained by in vitro biopanning was always higher than the output titers, indicating a successful enrichment and specific binding to cancer cells. 3.2. Phage ELISA for clone detection
Absorbance at 492 nm
(A) 0.20
*
0.15
Papillary carcinoma Goiter
*
0.10 0.05
C
C
aT
1 aT 2 C aT 3 C aT C aT 5 C aT 6 C aT C 7 aT 1 C 0 aT 11 C aT 1 C 2 aT 1 C 4 aT 17 C aT 1 C 8 aT 19 C aT 2 C 0 aT 2 C 1 aT 22 C aT 2 C 3 aT 24 C aT 2 C 5 aT 28 M 13
0.00
Tested phages
(B)
Fig. 1. Phage-ELISA and bioinformatics analyses of selected phage clones. (A) Phage-ELISA assays for cancer and goiter proteins. The Y-axis represents the average absorbance reading at 492 nm and the X-axis shows the groups that were incubated with the different peptides clones tested and the control of the reaction (M13). Clones that presented significantly higher absorbance when compared to the control are indicated (=p < 0.001; ANOVA, Bonferroni correction). (B) Consensus sequences formed by CaT12 and CaT18. Green letters = hydroxyl + amine + basic amino acids; red = small + hydrophobic amino acids; pink letter = basic amino acids. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
4
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
(A1)
(A2)
(A3)
(A4)
(A5)
(A6)
(B1)
(B2)
(B4) (B3)
Fig. 2. Immunohistochemistry with CaT12 phage and peptide in thyroid tissues. (A) Tissues of papillary thyroid cancer: (A1) CaT12 peptide; (A2) CaT12 phage; (A3) M13 wild phage. Tissues of nodular goiter: (A4) CaT12 peptide; (A5) CaT12 phage; (A6) M13 wild phage. (400). (B) Comparison of ACIS values in benign and papillary cancer tissues. (B1 and B3) Box-plot of CaT12 phage and peptide. (B2 and B4) represent ROC Curves of CaT12 phage and peptide, respectively.
3.5. Identification of the putative binding protein to the CaT12 peptide In order to further identify the target protein that could be binding to the CaT12 peptide, we have bound the biotinylated CaT12
peptide to streptavidin-conjugated beads and thyroid cancer proteins were magnetic captured. After magnetic separation, proteins were eluted and sequenced by mass spectrometry (LC-MS/MS). A list of proteins with the highest hits and number of matches is
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
5
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
Table 1 CaT12 recombinant phage and synthetic peptide positivity analyses considering visual and ACIS quantitative immunohistochemical scores for benign nodules (Nodular Goiter and Follicular Adenomas) and differentiated thyroid cancer histological subtypes. CaT12
Normal
Goiter
FA
CPTC
FVPTC
FC
P/N
Positivity (%)
P/N
Positivity (%)
P/N
Positivity (%)
P/N
Positivity (%)
P/N
Positivity (%)
P/N
Positivity (%)
Visual Phage Peptide
0/15 0/15
0 0
9/53 12/42
16.98 28.57
9/54 19/48
16.67 39.58
36/36 31/33
100 93.94
26/32 18/25
81.25 72.00
7/41 27/39
20.00 69.23
Acis Phage Peptide
0/2 0/15
0 0
13/47 12/44
27.66 27.27
8/54 21/47
14.81 44.68
34/36 32/35
94.44 91.43
27/33 22/30
81.82 73.33
10/41 28/40
24.39 70.00
Abbreviations: FA = follicular adenoma; CPTC = classic type PTC; FVPTC = follicular variant PTC; FC = follicular carcinoma; P = positive number of patients; N = total number of patients. ACIS’s cut-off used for the peptide was >57.87, and for the phage was >45.68.
Table 2 Diagnostics’ parameters for the CaT12 phage and CaT12 peptide immunohistochemistry analyses taking into account the benign nodules (Nodular Goiter and Follicular Adenomas) and differentiated thyroid cancer histological subtypes. CaT12
Cut-off
CPTC
CPTC and FVPTC
CPTC, FVPTC and FC
SEN (%)
SPE (%)
PPVa/NPVb or ACCc(%)
SEN (%)
SPE (%)
PPVa/NPVb or ACCc (%)
SEN (%)
SPE (%)
PPVa/NPVb or ACCc(%)
66.67a 100b 50.00a 97.37b
91.18
85.12
64.22
85.12
84.48
70.48
77.50a 94.50b 61.25a 89.16b
78.35
70.48
79.55a 72.54b 71.03a 77.89b
94.17c 83.53c
88.41 83.08
79.61 68.87
91.15c 79.65c
55.00 78.10
79.61 68.87
Visual Phage
–
100
85.12
Peptide
–
93.94
70.48
Acis Phage Peptide
>45.68 >57.87
94.44 91.43
79.61 68.87
72.50c 76.23c
Abbreviations: NG = nodular goiter, FA = follicular adenoma, CPTC = classic type PTC; FVPTC = follicular variant PTC; FC = follicular carcinoma, SEN = sensibility, SPE = specificity. a PPV = positive predictive value. b NPV = negative predictive value. c ACC = accuracy. The cut-off employed in the Visual score was determined by Fisher’s exact test. The cut-off employed in the quantitative immunohistochemistry analysis by ACIS was determined by the ROC curve analysis (see Methods description).
Table 3 Diagnostic value of CaT12 phage and CaT12 peptide for discriminating different histopathological subtypes, according to ACIS immunohistochemical values. Variable Phage CPTC FVPTC FVPTC FA FVPTC FC FC FA FA Goiter FVPTC Goiter Peptide CPTC FVPTC FVPTC FA FVPTC FC FC FA FA Goiter FVPTC Goiter
P*
Sensibility (%)
Specificity (%)
Accuracy (%)
0.2691 <0.0001* <0.0001* 0.6253 <0.0001* <0.0001*
91.67 81.82 78.79 24.39 85.19 81.82
18.18 85.19 75.61 85.19 27.66 72.34
57.74 91.92 88.77 52.94 76.36 83.11
0.0199* 0.0254* 0.8680 0.0512* 0.0583 <0.0001*
91.43 76.67 73.33 70.00 46.68 73.33
26.67 55.32 30.00 55.32 72.73 70.45
66.86 65.18 51.17 62.18 61.53 78.48
Abbreviations: FA = follicular adenoma; CPTC = classic type PTC; FVPTC = follicular variant PTC; FC = follicular carcinoma. * P < 0.05.
presented in Table 5, and apparently the peptide is a cytokeratinspecific ligand. 4. Discussion Identification of malignant thyroid nodules from indeterminate follicular lesions is difficult and biomarkers can barely distinguish follicular variants of papillary thyroid carcinomas from simple benign hyperplasia. We have used a subtractive phage display selection against goiter tissue aiming the improvement of thyroid cancer diagnosis. We have chosen this technology due to its
successful applications in biomarkers’ discovery, especially when short random peptides are selected against specific target molecules, such as those with antitumor activity [20] or for drug delivery [28]. A highly reactive peptide, CaT12, although not specific for thyroid tissues, was able to identify malignancy among thyroid nodular lesions and characterize follicular patterned lesions. In fact, its accuracy to identify follicular patterned lesions is comparable to most of the currently used immunohistochemical markers employed in differential diagnosis of indeterminate nodules [29,30]. Moreover, this peptide is associated to features of DTC aggressiveness, such as age and the occurrence of metastasis, suggesting a potential clinical use. We observed different results concerning the CaT12 recombinant clone and its synthetic peptide, with increased staining and positivity for the phage clone in cancer samples. This fact may be due to the specific epitope conformation generated by the fusion of the peptide with the N-terminal region of the bacteriophage pIII protein. On the other hand, higher reactivity of the peptide was observed in controls (goiter and adenomas), probably because of the lower affinity of the peptide alone, without the conformational rearrangement of the peptide-pIII fusion. Therefore, there is a potential contribution of phage-coat proteins for the peptide affinity as discussed elsewhere [31,32]. CaT12 strongly stained both the membrane and the cytoplasm of papillary thyroid cancer cells. Since normal thyroid cells do not stain, its protein target may be produced during physiological cellular growth. Benign thyroid lesions, such as follicular adenomas and hyperplasia, presented a very weak staining, reinforcing the hypothesis that we are dealing with a protein involved in cellular growth. We have identified several protein targets of the
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
6
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
Fig. 3. Immunohistochemistry with CaT12 peptide in different tissues. Cancer tissues: Positive: (A1) kidney; (A2) prostate; (A3) breast. Negative: (B1) kidney; (B2) prostate; (B3) breast. Non-neoplastic tissues: Positive: (C1) bladder; (C2) respiratory epithelium; (C3) small intestine. Negative: (D1) parathyroid; (D2) nevus; (D3) muscle.
CaT12 ligand, mostly cytokeratins, which are currently being tested. In fact, serum cytokeratin fragments have been found in thyroid nodule patients [33] and have even been associated with the risk of metastases [34]. Interestingly, cytokeratin-19 immunohistochemistry is largely employed in the differential diagnosis of follicular patterned lesions [29], although it was not recognized in our mass spectrometry analysis. However, due to the smaller specificity and
binding affinity of the synthetic peptide in comparison to the recombinant phage, it is possible that the peptide may partially bind to a conserved cytokeratin domain, which may explain the alignment (hits) of digested protein fragments with many cytokeratin sequences. On the other hand, the higher specificity found for the CaT12 recombinant phage suggests that its fused peptide sequence may be partially complemented by the amino terminal sequence of the pIII phage capsid protein improving its binding affinity.
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx Table 4 Immunohistochemistry positivity of the CaT12 peptide in different tissues. Histopathological diagnosis
Visual analysis N
Positive cases (%)
Non-malignant Nevus atypical Nevus communes Nonneoplastic tissues* Normal breast Normal prostate Normal kidney Total
12 29 206 5 5 5 262
0 3.45 15.53 0 0 60 13.74
Malignant Breast cancer Prostate cancer Kidney cancer Total
189 57 35 281
53.96 70.18 74.29 59.79
* Different human nonneoplastic tissues including pancreas, liver, lung, breast, prostate and brain.
Table 5 Identification of putative protein targets of the Cat12 peptide ligand by LC-MS/MS. Protein name Keratin, type Keratin, type Keratin, type Keratin, type Keratin, type Hornerin Keratin, type Keratin, type
II cytoskeletal 1 I cytoskeletal 9 I cytoskeletal 10 I cytoskeletal 14 II cytoskeletal 2 II cytoskeletal 6A II cytoskeletal 5
Acession number
MW (KDa)
P04264 P35527 P13645 P02533 P35908 Q86YZ3 P02538 P13647
66 62 59 52 65 282 60 62
In conclusion, we have designed a novel approach to select specific peptide ligands to thyroid cancer cells, which allowed us to identify a highly reactive antibody-like peptide that maybe readily used in the cellular diagnosis of thyroid nodules, discriminating indeterminate benign follicular lesions from carcinomas with high accuracy. This specific ligand may also be a potential therapeutic agent or drug carrier for future treatment strategies of thyroid cancer. Conflict of Interests The authors declare no conflict of interests. Acknowledgements The present work was supported by the Brazilian funding agencies: Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). The authors would like to thank Dr. Janete Maria Cerutti (Unifesp) for donating the NPA and TPC1 cell lines, Dr. José Vasallo (A.C. Camargo) for performing final IHC analysis, and A.C. Camargo Hospital for providing tissue microarrays. References [1] L. Davies, H.G. Welch, Increasing incidence of thyroid cancer in the United States, 1973–2002, JAMA: J. Am. Med. Assoc. 295 (2006) 2164–2167. [2] R. Siegel, C. DeSantis, K. Virgo, K. Stein, A. Mariotto, T. Smith, D. Cooper, T. Gansler, C. Lerro, S. Fedewa, C. Lin, C. Leach, R.S. Cannady, H. Cho, S. Scoppa, M. Hachey, R. Kirch, A. Jemal, E. Ward, Cancer treatment and survivorship statistics, CA: Cancer J. Clin. 62 (2012) 220–241. [3] L.S. Ward, H. Graf, Thyroid cancer: increased occurrence of the disease or simply in its detection?, Arq Bras. endocrinol. Metabol. 52 (2008) 1515–1516.
7
[4] B. Aschebrook-Kilfoy, M.H. Ward, M.M. Sabra, S.S. Devesa, Thyroid cancer incidence patterns in the United States by histologic type, 1992–2006, Thyroid: Official J. Am. Thyroid Assoc. 21 (2011) 125–134. [5] D.S. Cooper, G.M. Doherty, B.R. Haugen, R.T. Kloos, S.L. Lee, S.J. Mandel, E.L. Mazzaferri, B. McIver, F. Pacini, M. Schlumberger, S.I. Sherman, D.L. Steward, R.M. Tuttle, Revised American thyroid association management guidelines for patients with thyroid nodules and differentiated thyroid cancer, Thyroid: Official J. Am. Thyroid Assoc. 19 (2009) 1167–1214. [6] S. Bose, A.E. Walts, Thyroid fine needle aspirate: a post-Bethesda update, Adv. Anat. Pathol. 19 (2012) 160–169. [7] C.C. Wang, L. Friedman, G.C. Kennedy, H. Wang, E. Kebebew, D.L. Steward, M.A. Zeiger, W.H. Westra, Y. Wang, E. Khanafshar, G. Fellegara, J. Rosai, V. Livolsi, R.B. Lanman, A large multicenter correlation study of thyroid nodule cytopathology and histopathology, Thyroid: Official J. Am. Thyroid Assoc. 21 (2011) 243–251. [8] E.K. Alexander, G.C. Kennedy, Z.W. Baloch, E.S. Cibas, D. Chudova, J. Diggans, L. Friedman, R.T. Kloos, V.A. LiVolsi, S.J. Mandel, S.S. Raab, J. Rosai, D.L. Steward, P.S. Walsh, J.I. Wilde, M.A. Zeiger, R.B. Lanman, B.R. Haugen, Preoperative diagnosis of benign thyroid nodules with indeterminate cytology, New England J. Med. 367 (2012) 705–715. [9] L. Yip, C. Farris, A.S. Kabaker, S.P. Hodak, M.N. Nikiforova, K.L. McCoy, M.T. Stang, K.J. Smith, Y.E. Nikiforov, S.E. Carty, Cost impact of molecular testing for indeterminate thyroid nodule fine-needle aspiration biopsies, J. Clin. Endocrinol. Metab. 97 (2012) 1905–1912. [10] S. Carnazza, C. Foti, G. Gioffre, F. Felici, S. Guglielmino, Specific and selective probes for Pseudomonas aeruginosa from phage-displayed random peptide libraries, Biosensors Bioelectron. 23 (2008) 1137–1144. [11] C.G. Nicol, L. Denby, O. Lopez-Franco, R. Masson, C.A. Halliday, S.A. Nicklin, A. Kritz, L.M. Work, A.H. Baker, Use of in vivo phage display to engineer novel adenoviruses for targeted delivery to the cardiac vasculature, FEBS Lett. 583 (2009) 2100–2107. [12] X. Yu, G.P. Owens, D.H. Gilden, Rapid and efficient identification of epitopes/ mimotopes from random peptide libraries, J. Immunol. Methods 316 (2006) 67–74. [13] F. Bai, J. Liang, J. Wang, Y. Shi, K. Zhang, S. Liang, L. Hong, H. Zhai, Y. Lu, Y. Han, F. Yin, K. Wu, D. Fan, Inhibitory effects of a specific phage-displayed peptide on high peritoneal metastasis of gastric cancer, J. Mol. Med. (Berl) 85 (2007) 169– 180. [14] L. Chen, Y. Wang, X. Liu, S. Dou, G. Liu, D.J. Hnatowich, M. Rusckowski, A new TAG-72 cancer marker peptide identified by phage display, Cancer Lett. 272 (2008) 122–132. [15] B. Du, M. Qian, Z. Zhou, P. Wang, L. Wang, X. Zhang, M. Wu, P. Zhang, B. Mei, In vitro panning of a targeting peptide to hepatocarcinoma from a phage display peptide library, Biochem. Biophys. Res. Commun. 342 (2006) 956–962. [16] A. Dybwad, P. Lambin, M. Sioud, M. Zouali, Probing the specificity of human myeloma proteins with a random peptide phage library, Scand. J. Immunol. 57 (2003) 583–590. [17] A.L. Goenaga, Y. Zhou, C. Legay, H. Bougherara, L. Huang, B. Liu, D.C. Drummond, D.B. Kirpotin, C. Auclair, J.D. Marks, M.A. Poul, Identification and characterization of tumor antigens by using antibody phage display and intrabody strategies, Mol. Immunol. 44 (2007) 3777–3788. [18] S. Hu, X. Guo, H. Xie, Y. Du, Y. Pan, Y. Shi, J. Wang, L. Hong, S. Han, D. Zhang, D. Huang, K. Zhang, F. Bai, H. Jiang, H. Zhai, Y. Nie, K. Wu, D. Fan, Phage display selection of peptides that inhibit metastasis ability of gastric cancer cells with high liver-metastatic potential, Biochem. Biophys. Res. Commun. 341 (2006) 964–972. [19] S. Liang, T. Lin, J. Ding, Y. Pan, D. Dang, C. Guo, M. Zhi, P. Zhao, L. Sun, L. Hong, Y. Shi, L. Yao, J. Liu, K. Wu, D. Fan, Screening and identification of vascularendothelial-cell-specific binding peptide in gastric cancer, J. Mol. Med. (Berl) 84 (2006) 764–773. [20] A.L. Matsuo, A.S. Tanaka, M.A. Juliano, E.G. Rodrigues, L.R. Travassos, A novel melanoma-targeting peptide screened by phage display exhibits antitumor activity, J. Mol. Med. (Berl) 88 (2010) 1255–1264. [21] C. Mazuet, D. Lerouge, M.A. Poul, N. Blin, Breast carcinoma specific antibody selection combining phage display and immunomagnetic cell sorting, Biochem. Biophys. Res. Commun. 348 (2006) 550–559. [22] W.D. Jia, H.C. Sun, J.B. Zhang, Y. Xu, Y.B. Qian, J.Z. Pang, L. Wang, L.X. Qin, Y.K. Liu, Z.Y. Tang, A novel peptide that selectively binds highly metastatic hepatocellular carcinoma cell surface is related to invasion and metastasis, Cancer Lett. 247 (2007) 234–242. [23] T. Oyama, K.F. Sykes, K.N. Samli, J.D. Minna, S.A. Johnston, K.C. Brown, Isolation of lung tumor specific peptides from a random peptide library: generation of diagnostic and cell-targeting reagents, Cancer Lett. 202 (2003) 219–230. [24] L. Zang, L. Shi, J. Guo, Q. Pan, W. Wu, X. Pan, J. Wang, Screening and identification of a peptide specifically targeted to NCI-H1299 from a phage display peptide library, Cancer Lett. 281 (2009) 64–70. [25] R.J. Giordano, M. Cardo-Vila, J. Lahdenranta, R. Pasqualini, W. Arap, Biopanning and rapid analysis of selective interactive ligands, Nat. Med. 7 (2001) 1249–1253. [26] J. Huang, B. Ru, P. Dai, Bioinformatics resources and tools for phage display, Molecules 16 (2011) 694–709. [27] S. Mandava, L. Makowski, S. Devarapalli, J. Uzubell, D.J. Rodi, RELIC – a bioinformatics server for combinatorial peptide analysis and identification of protein-ligand interaction sites, Proteomics 4 (2004) 1439–1460. [28] B. Du, H. Han, Z. Wang, L. Kuang, L. Wang, L. Yu, M. Wu, Z. Zhou, M. Qian, Targeted drug delivery to hepatocarcinoma in vivo by phage-displayed specific binding peptide, Mol. Cancer Res.: MCR 8 (2010) 135–144.
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
8
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
[29] P.S. de Matos, A.P. Ferreira, F. de Oliveira Facuri, L.V. Assumpcao, K. Metze, L.S. Ward, Usefulness of HBME-1, cytokeratin 19 and galectin-3 immunostaining in the diagnosis of thyroid malignancy, Histopathology 47 (2005) 391–401. [30] K. Sethi, S. Sarkar, S. Das, B. Mohanty, M. Mandal, Biomarkers for the diagnosis of thyroid cancer, J. Exp. Therap. Oncol. 8 (2010) 341–352. [31] C.F. Barbas, Phage Display: a Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 2001. [32] M.B. Zwick, L.L. Bonnycastle, K.A. Noren, S. Venturini, E. Leong, C.F. Barbas 3rd, C.J. Noren, J.K. Scott, The maltose-binding protein as a scaffold for monovalent
display of peptides derived from phage libraries, Anal. Biochem. 264 (1998) 87–97. [33] L. Giovanella, L. Ceriani, A. Ghelfo, M. Maffioli, Circulating cytokeratin 19 fragments in patients with benign nodules and carcinomas of the thyroid gland, Int. J. Biol. Markers 23 (2008) 54–57. [34] M. Appetecchia, A. Mecule, M. Ducci, L. Palma, M. Castelli, Serum cytokeratins determination in differentiated thyroid carcinoma, J. Exp. Clin. Cancer Res.: CR 20 (2001) 253–256.
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
Contents lists available at SciVerse ScienceDirect
Cancer Letters journal homepage: www.elsevier.com/locate/canlet
An antibody-like peptide that recognizes malignancy among thyroid nodules Carolina Fernandes Reis a, Ana Paula Carneiro a, Carlos Ueira Vieira b, Patrícia Tiemi Fujimura b, Elaine Cristina Morari a, Sindeval José da Silva c, Luiz Ricardo Goulart b,⇑, Laura Sterian Ward a,⇑ a b c
Laboratory of Cancer Molecular Genetics, Faculty of Medical Sciences (FCM), University of Campinas (UNICAMP), Campinas, SP, Brazil Laboratory of Nanobiotechnology, Federal University of Uberlandia, Institute of Genetics and Biochemistry (INGEB), Uberlandia, MG, Brazil Department of Head and Neck Surgery, Clinical Hospital of Federal University of Uberlândia (HC-UFU), Uberlândia, MG, Brazil
a r t i c l e
i n f o
Article history: Received 28 September 2012 Received in revised form 1 February 2013 Accepted 18 February 2013 Available online xxxx Keywords: Thyroid cancer Phage display Biomarker Papillary tumor
a b s t r a c t There is an urgent need for biomarkers to identify malignant thyroid nodules from indeterminate follicular lesions. We have used a subtractive proteomic strategy to identify novel biomarkers by selecting ligands to goiter tissue from a 12-mer random peptide phage-displayed library using the BRASIL method (Biopanning and Rapid Analysis of Selective Interactive Ligands). After three rounds of selection, two highly reactive clones to the papillary thyroid tumor cell line NPA were further evaluated, and their specific binding to tumor proteins was confirmed using phage-ELISA. The antibody-like peptide CaT12 was tumor-specific, which was further tested by immunohistochemistry against TMAs (tissue microarrays) comprised of 775 human benign and malignant tissues, including 232 thyroid nodular lesions: 15 normal thyroid tissues, 53 nodular goiters (NG), 54 follicular adenomas (FA); 69 papillary thyroid carcinomas (PTC); and 41 follicular carcinomas (FC). CaT12 was able to identify PTC among thyroid nodular lesions with 91.2% sensitivity and 85.1% specificity, despite its non-specificity for thyroid tissues. Additionally, the CaT12 peptide helped characterize follicular lesions distinguishing the follicular variant of PTC (FVPTC) from FA with 91.9% accuracy; FVPTC from NG with 83.1% accuracy; FVPTC from the classic PTC with 57.7% accuracy; and FVPTC from FC with 88.7% accuracy. In conclusion, our strategy to select differentially expressed ligands to thyroid tissue was highly effective and resulted in a useful antibody-like biomarker that recognizes malignancy among thyroid nodules and may help distinguish follicular patterned lesions. Ó 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Thyroid cancer incidence rates have steadily increased over the past few decades all over the world [1] and 56,460 new patients are estimated to be diagnosed during 2012 in the USA [2]. The accessible use of sensitive imaging detection methods is mainly responsible for this increased incidence [3]. However, there are evidences that other factors, including the exposure to environmental carcinogenic factors, may have also contributed to this phenomenon [3,4]. We have actually experienced real epidemics of thyroid nodules referred to specialized diagnosis, and the cytological exam of fine-needle cell aspiration is the recommended meth⇑ Corresponding authors. Addressess: Laboratory of Nanobiotechnology, Institute of Genetics and Biochemistry, Federal University of Uberlandia (UFU), Campus Umuarama, Bl 2E, R. 248, 38400-902 Uberlandia, MG, Brazil (L.R. Goulart), Laboratory of Cancer Molecular Genetics, School of Medical Sciences, University of Campinas (FCM-UNICAMP), Rua Tessalia Vieira de Camargo 126, Barão Geraldo, 13083-887 Campinas, SP, Brazil. Tel.: +55 19 35218954; fax: +55 19 35218925 (L.S. Ward). E-mail address: [email protected] (L.S. Ward).
od by current guidelines to distinguish benign from malignant lesions [5]. In most cases, the clinical features of the patients based on ultrasonography and cytological findings enable appropriate case management [5]. However, up to one third of the nodules submitted to fine needle aspiration biopsies may present an ‘‘indeterminate’’ cytology, a pattern that remains burdened with interobserver variability and uncertainty regarding management [6]. Even the most experienced pathologists may have difficulties distinguishing cases of follicular variants of papillary thyroid carcinomas from simple benign hyperplasia, and many cases of follicular adenomas are submitted to unnecessary surgeries because it is virtually impossible to eliminate the risk of a follicular carcinoma in cytological analyses [7]. A series of molecular markers for malignancy have been proposed and many of them appear to be effective in reducing uncertainties [8,9]. However, clinical problems concerning diagnosis of thyroid malignancies are not solved, and certainly are not appropriate to large populations. In the present investigation, we have used a different approach to distinguish thyroid malignancies. We employed phage display
0304-3835/$ - see front matter Ó 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.canlet.2013.02.039
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
2
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
technology that has been successfully used to select targets for various diseases [10–12], including many cancer types [13–24], and, by using a direct selection (BRASIL) against goiter tissue, we were able to obtain a highly reactive and discrimination marker for thyroid cancer diagnosis.
anti-M13 monoclonal antibody (Amersham Pharmacia Biotech Benelux, Roosendaal, The Netherlands) using a 1:5000 dilution in blocking buffer. The peroxidase staining reaction was developed in the presence of SIGMAFAST™ OPD (o-phenylenediamine dihydrochloride) tablets (Sigma–Aldrich, São Paulo, SP, Brazil), and then incubated at room temperature for 5 min, stopped with diluted acid solutions, and the optical density (OD) values were measured on a microplate reader (TPREADER – Thermo plate) at 492 nm.
2. Materials and methods 2.5. Recombinant phage and synthetic peptide immunohistochemistry (IHC) Research Ethics Committees of three institutions approved this study: Federal University of Uberlândia (UFU), State University of Campinas (UNICAMP) and Fundação Antônio Prudente (Hospital AC Camargo, SP). All participants signed informed consent forms. 2.1. Cell culture and patients’ samples We have used the TPC-1 (Human thyroid papillary cancer) and NPA (Human papillary cancer) cells, gently provided by Prof. J.M. Cerutti, from the Federal University of São Paulo. In addition, we have developed a primary human cell culture using tissue obtained from a patient submitted to surgical removal of a multinodular goiter at the Surgery Center of UFU Hospital. Thyroid cancer cell lines and goiterderived cells were grown in RPMI 1640 medium supplemented with penicillin, streptomycin, and 10% fetal bovine serum (Nutricell, Campinas-SP) and incubated at 37 °C in a humidified atmosphere containing 5% CO2. Fresh tissues from patients with histologically-proved classic PTC and hyperplasia were obtained from the Federal University of Uberlândia (UFU) Clinics’ Hospital, and immediately frozen until use for protein binding testing. The selected clone was tested by immunohistochemistry in formalin fixed, paraffin embedded tissues from 775 patients reviewed for diagnostic confirmation in order to select the most representative areas designed to build tissue microarrays (TMA, Beecher InstrumentsÒ, Silver Springs, MD, USA). A first TMA was composed by 232 patient samples: 15 normal thyroid tissues, 53 nodular goiters, 54 follicular adenomas, 69 papillary thyroid carcinomas (36 cases of the classical form and 33 of the PTC follicular variant) and 41 follicular carcinomas. A second TMA was designed to validate our results, and was constituted by 281 other cancer tissues (189 breast, 57 prostate and 35 kidney), 41 benign lesions (12 nevus atypical and 29 nevus) and 221 non-neoplastic tissues (including normal stomach, ovary, breast, prostate, kidney, heart, muscle, brain, liver, lung and others). Patients’ clinical information was obtained from their clinical records. Aggressiveness at diagnosis was ascertained using the Tumor Node Metastasis and stage classification systems for DTC as recommended by the American Thyroid Association guidelines [5]. For statistical purpose we grouped patients classified as stage II, III and IV in one class named ‘‘higher stages’’. Patients were followed with periodic total body scans, serum TSH and thyroglobulin (Tg) measurements, according to a standard protocol that included X-ray, ultrasonography, computed tomography scan and other eventual procedures to detect distant metastasis for a period of 12– 298 months (43.50 ± 33.29 months). Patients presenting high non-stimulated serum Tg levels (>2 mg/dl) were submitted to image evaluation. We defined tumors as persistent/recurrent and/or presenting long distance metastasis, according to the aforementioned parameters. 2.2. Biopanning in cell surface Biopanning was performed in three rounds using BRASIL method (Biopanning and Rapid Analysis of Selective Interactive Interactive Ligands) [25] to verify phage binding to intact cells. 2.3. DNA sequencing of selected phage clones After three rounds, 96 blue colonies were randomly selected and their phage single-strand DNA were isolated by the iodide buffer extraction procedure (Instruction Manual Ph.D.-12 Phage Display Peptide Library Kit) and analyzed with MegaBace 1000 Genetic Analyzer (Amersham Biosciences) automatic capillary sequencer. The nucleotide sequence of the gene III insert was sequenced with 96 gIII sequencing primer 50 -CCC TCA TAG TTA GCG TAA CG-30 by automated dye terminator cycle sequencing. The amino acid sequence of the insert was deduced from the nucleotide sequence using DNA2PRO software from RELIC program [26,27]. 2.4. ELISA (enzyme-linked immunosorbent assay) After extraction of papillary cancer and goiter proteins with extraction buffer (Tris–HCl 20 mM pH7.2, EDTA 10 mM, EGTA 2 mM, sucrose 250 mM, DTT 1 mM, Benzamidine 1 mM) proteins were diluted in sodium carbonate buffer (10 lg/ml in 0.1 M NaHCO3) and immobilized into a 96-wells plate (Nunc Immuno Maxisorp, Roche Diagnostics) by incubation it overnight at 4 °C. That plate was subsequently washed three times with PBST 0.1%, and blocked with BSA 3% (blocking buffer) for three hours at 37 °C. Plate wells were washed again three times, and each well was incubated with 100 ll of phages supernatants for one hour at 37 °C. After five washes, recombinant phage particles were detected with peroxidase-conjugated
After deparaffinizing the tissues, the endogenous peroxidase was inactivated by incubating the sections with 3% hydrogen peroxide for three times, 5 min each. The tissue sections were then rinsed thoroughly in distilled water. For the biomarker exposure, sodium citrate buffer (0.1 M) was heated to 95 °C steamer, the slides were submerged into the buffer for 35 min, washed for 5 min in running water, and blocked with 3% skim milk (Molico Milk, Nestle, Brazil; 3% w/v in distilled water) for 30 min at room temperature. Incubation with the primary marker (phage or synthetic peptide) was then performed for 30 min at 37 °C and overnight at 4 °C. The recombinant phage CaT12, the synthetic peptide (Invitrogen, Carlsbad, CA, USA), and the wild phage (no peptide) were diluted into 100 ll of the Novocastra Universal IHC Blocking/Diluent 1000 (Leica Microsystems Inc, Buffalo Grove, IL, USA) containing 6 1011 pfu (viral particles/slide) or 1 lg synthetic peptide. As negative controls, duplicate sections were incubated with 3% skim milk proteins instead of specific primary antibodies. The sections were washed four times in PBS for 5 min each followed by one hour incubation with peroxidase-conjugated mouse anti-M13 secondary antibody or streptavidin. After a washing step in PBS, peroxidase activity was visualized by incubation in 3,30 diaminobenzidine tetrahydrochloride (Liquid DAB, Sigma, St. Louis, USA) for 5 min at room temperature. Sections were counterstained with hematoxylin. Positive and negative controls were run in the same reaction batch. Slides were independently scored by two experienced researchers (CFR and ECM), both blinded to tumor features, and confirmed by an experienced pathologist (JV). 2.6. Immunohistochemistry staining index Signals were considered positive when reaction products were localized in the expected cellular component. The staining index value was obtained by the sum of intensity and distribution scores, and values of a staining index equal or greater than 5 were considered positive. The criteria employed for the intensity score were: 0, no staining; 1, weak; 2, moderate; 3, strong. The criteria employed for the distribution score were: 0, no staining; 1, staining of <25% of the cells; 2, between 25% and 50% of the cells; 3, between 50% and 75% of the cells; and 4, staining of >75% of the cells. In addition, we employed the Automated Cellular Imaging System ACIS-III (Chroma Vision Medical Systems, Inc, DAKO, San Juan Capistrano, CA, USA) for IHC quantification. Briefly, each tissue spot was digitalized to the systems’ software, and a computer algorithm, considering the intensity and extension of the brown staining, attributed numerical values. 2.7. Streptavidin-bead precipitation assay TPC-1 cells were lysed according to the instructions of the RIPA Buffer (Thermo scientific, Rockford, IL, USA) kit; a protease inhibitor cocktail (1:100 dilution) (Sigma–Aldrich, São Paulo, SP, Brazil) was added and the protein concentration was determined using the BCA Protein Assay Kit (Thermo scientific). The synthetic peptide (1 mg/ml) was N-terminal labeled with biotin, and mixed with 200 ll packed Streptavidin-beads (Thermo scientific) for 1 h at 4 °C, and incubated with 500 lg of protein extracts for 4 h at room temperature. After washing four times with TBS (Tris Buffered Saline), the proteins were eluted from the beads by addition 0.1 M glycine HCl buffer, pH 3.0, twice for 10 min each. The eluate was quantified by the BCA Protein Assay Kit prior to proteins sequencing. 2.8. Protein sequencing by LC-MS/MS and target identification Eluted proteins were precipitated out of solution using the ProteoExtract kit (Calbiochem) and the protein pellet was left to dry overnight in a sterile fumehood. The lyophilized pellet was then resuspended in 50 mM ammonium bicarbonate (pH 8.0) and subjected to an in-solution tryptic digestion (Mike Myers, Cold Spring Harbor modified by Brett S. Phinney, UC Davis Proteomics Core, Davis, CA, USA). Digested peptides were then de-salted using aspire tips (Thermo-Fisher Scientific, RP30 tips) before being resuspended in loading buffer. Digested peptides were analyzed using a LTQ-FT (Thermo Fisher Scientific) coupled with a MG4 paradigm HPLC (Michrom, Auburn, CA, USA). The samples were loaded onto a Michrom cap trap (0.5 2 mm) to be de-salted online. The peptides were then separated using a Michrom Magic C18AQ (200 lm 150 mm) reversedphase column and eluted using a gradient for 60 min. Collision induced dissociation was applied to the peptide samples and data was acquired with an isolation width of 1, normalized collision energy of 35, and a resolution of 50,000. The spray voltage on the Michrom captive spray was set to 1.8 kV with a heated transfer capillary temperature of 200 °C.
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
3
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx Raw data was analyzed using X!Tandem and visualized using Scaffold (Proteome Software, version 3.01). Samples were searched in Uniprot human sequences databases, appended with the cRAP (commonly found laboratory contaminants) and reverse decoy databases.
ELISA was performed on cancer and goiter cases in order to analyze the specific protein binding properties and to select the best phage clones for immunohistochemistry testing. Clones CaT12 and CaT18 presented the highest absorbances for cancer proteins (Fig. 1A) when compared to goiter proteins (P < 0.0001). A wild phage used as control (M13) showed low reactivity. CaT12 and CaT18 presented a consensus sequence showed on Fig. 1B.
phage was used as a negative control (Fig. 2 and Table 1). The CaT12 clone showed higher positivity than M13 wild phage in thyroid cancer cases, as demonstrated in Fig. 2. Furthermore, the CaT12 synthetic peptide showed an increased signal (Fig. 2, A1), which was not followed by CaT18 peptide that presented a marginal positivity (data not shown). The cut-off values for CaT12 phage and CaT12 peptide were chosen based on the ROC curve analyses (Fig. 2B2; Fig. 2B4, respectively), which generated diagnostic parameters for the comparisons between cancer groups (Table 2). A significant difference (P < 0.0001) between cancer and benign diseases was found (Fig. 2B1 and B3), which was superior for the CaT12 phage as a biomarker. We observed an overall higher positivity using CaT12 peptide in goiters, FA and FC, whereas CaT12 phage was more frequently positive in CPTC and FCPTC, as demonstrated in Fig. 2. Visual and ACIS analysis generated very comparable results as demonstrated in Table 2 where we demonstrate the utility of the employed CaT12 phage and CaT12 peptide in the characterization of follicular patterned lesions. Both visual and ACIS were valid analytical methods, and presented similar values. In fact, both CaT12 phage and synthetic peptide were able to identify malignancy among follicular patterned lesions (goiter, FA, FVPTC and PTC) and discriminate these lesions with considerable accuracy, as demonstrated in Table 3. CaT12 was inversely correlated with patient’s age (Spearman r = 0.2469, p = 0.0197) and was directly correlated with the occurrence of metastasis (Spearman r = 0.3585, p = 0.0039) at follow-up. We were not able to establish any other association between CaT12 and any clinical or tumor feature or with the patients outcome. We further employed CaT12 peptide in other tissues, as shown in Fig. 3, demonstrating positivity in malignant tumors, and a weak positivity in benign and normal tissues as presented in Fig. 3C and Table 4.
3.3. Biomarker validation: phage and synthetic peptide immunohistochemistry
3.4. Immunohistochemistry validation
To confirm the ability of the selected clones to target cancer cells, phage clone CaT12 and CaT18 were amplified, purified and directly tested in IHC assays of the TMA built with cases of thyroid cancer, adenoma, goiter and normal thyroid tissues. The M13 wild
To validate CaT12 peptide we used different benign and malign tissues as demonstrated in Fig. 3. We found a weak or negative staining in benign lesions and in normal tissues, whereas tumors were positive, as demonstrated in Table 4.
2.9. Statistical analysis We employed GraphPrism 5.0 software to analyze the reactivity of the selected phage and the synthetic peptide. ANOVA test with Bonferroni Post-hoc test, was used to verify differences between means of OD value obtained by ELISA with phage clones. Differences between benign and papillary cancer tissues for both biomarker CaT12 phage and CaT12 peptide were tested with the Mann–Whitney U test (continuous variables and non-parametric analyses). Receiver operating characteristics (ROC) curves were constructed to assess sensitivity, specificity, accuracy, and respective areas under the curves (AUCs) with 95% confidence interval (CI). For visual analysis and ACIS imaging, Fisher’s test was employed to determine the relationship between CaT12 phage or synthetic peptide for the different disease groups in the TMA. We have used sensitivity and specificity parameters in the Fisher’s test to choose the cut-off value in the ROC curve analysis. Probability below 0.05 was considered significant.
3. Results 3.1. Biopanning and specific enrichment of NPA-bound phages The amount of phages obtained by in vitro biopanning was always higher than the output titers, indicating a successful enrichment and specific binding to cancer cells. 3.2. Phage ELISA for clone detection
Absorbance at 492 nm
(A) 0.20
*
0.15
Papillary carcinoma Goiter
*
0.10 0.05
C
C
aT
1 aT 2 C aT 3 C aT C aT 5 C aT 6 C aT C 7 aT 1 C 0 aT 11 C aT 1 C 2 aT 1 C 4 aT 17 C aT 1 C 8 aT 19 C aT 2 C 0 aT 2 C 1 aT 22 C aT 2 C 3 aT 24 C aT 2 C 5 aT 28 M 13
0.00
Tested phages
(B)
Fig. 1. Phage-ELISA and bioinformatics analyses of selected phage clones. (A) Phage-ELISA assays for cancer and goiter proteins. The Y-axis represents the average absorbance reading at 492 nm and the X-axis shows the groups that were incubated with the different peptides clones tested and the control of the reaction (M13). Clones that presented significantly higher absorbance when compared to the control are indicated (=p < 0.001; ANOVA, Bonferroni correction). (B) Consensus sequences formed by CaT12 and CaT18. Green letters = hydroxyl + amine + basic amino acids; red = small + hydrophobic amino acids; pink letter = basic amino acids. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
4
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
(A1)
(A2)
(A3)
(A4)
(A5)
(A6)
(B1)
(B2)
(B4) (B3)
Fig. 2. Immunohistochemistry with CaT12 phage and peptide in thyroid tissues. (A) Tissues of papillary thyroid cancer: (A1) CaT12 peptide; (A2) CaT12 phage; (A3) M13 wild phage. Tissues of nodular goiter: (A4) CaT12 peptide; (A5) CaT12 phage; (A6) M13 wild phage. (400). (B) Comparison of ACIS values in benign and papillary cancer tissues. (B1 and B3) Box-plot of CaT12 phage and peptide. (B2 and B4) represent ROC Curves of CaT12 phage and peptide, respectively.
3.5. Identification of the putative binding protein to the CaT12 peptide In order to further identify the target protein that could be binding to the CaT12 peptide, we have bound the biotinylated CaT12
peptide to streptavidin-conjugated beads and thyroid cancer proteins were magnetic captured. After magnetic separation, proteins were eluted and sequenced by mass spectrometry (LC-MS/MS). A list of proteins with the highest hits and number of matches is
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
5
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
Table 1 CaT12 recombinant phage and synthetic peptide positivity analyses considering visual and ACIS quantitative immunohistochemical scores for benign nodules (Nodular Goiter and Follicular Adenomas) and differentiated thyroid cancer histological subtypes. CaT12
Normal
Goiter
FA
CPTC
FVPTC
FC
P/N
Positivity (%)
P/N
Positivity (%)
P/N
Positivity (%)
P/N
Positivity (%)
P/N
Positivity (%)
P/N
Positivity (%)
Visual Phage Peptide
0/15 0/15
0 0
9/53 12/42
16.98 28.57
9/54 19/48
16.67 39.58
36/36 31/33
100 93.94
26/32 18/25
81.25 72.00
7/41 27/39
20.00 69.23
Acis Phage Peptide
0/2 0/15
0 0
13/47 12/44
27.66 27.27
8/54 21/47
14.81 44.68
34/36 32/35
94.44 91.43
27/33 22/30
81.82 73.33
10/41 28/40
24.39 70.00
Abbreviations: FA = follicular adenoma; CPTC = classic type PTC; FVPTC = follicular variant PTC; FC = follicular carcinoma; P = positive number of patients; N = total number of patients. ACIS’s cut-off used for the peptide was >57.87, and for the phage was >45.68.
Table 2 Diagnostics’ parameters for the CaT12 phage and CaT12 peptide immunohistochemistry analyses taking into account the benign nodules (Nodular Goiter and Follicular Adenomas) and differentiated thyroid cancer histological subtypes. CaT12
Cut-off
CPTC
CPTC and FVPTC
CPTC, FVPTC and FC
SEN (%)
SPE (%)
PPVa/NPVb or ACCc(%)
SEN (%)
SPE (%)
PPVa/NPVb or ACCc (%)
SEN (%)
SPE (%)
PPVa/NPVb or ACCc(%)
66.67a 100b 50.00a 97.37b
91.18
85.12
64.22
85.12
84.48
70.48
77.50a 94.50b 61.25a 89.16b
78.35
70.48
79.55a 72.54b 71.03a 77.89b
94.17c 83.53c
88.41 83.08
79.61 68.87
91.15c 79.65c
55.00 78.10
79.61 68.87
Visual Phage
–
100
85.12
Peptide
–
93.94
70.48
Acis Phage Peptide
>45.68 >57.87
94.44 91.43
79.61 68.87
72.50c 76.23c
Abbreviations: NG = nodular goiter, FA = follicular adenoma, CPTC = classic type PTC; FVPTC = follicular variant PTC; FC = follicular carcinoma, SEN = sensibility, SPE = specificity. a PPV = positive predictive value. b NPV = negative predictive value. c ACC = accuracy. The cut-off employed in the Visual score was determined by Fisher’s exact test. The cut-off employed in the quantitative immunohistochemistry analysis by ACIS was determined by the ROC curve analysis (see Methods description).
Table 3 Diagnostic value of CaT12 phage and CaT12 peptide for discriminating different histopathological subtypes, according to ACIS immunohistochemical values. Variable Phage CPTC FVPTC FVPTC FA FVPTC FC FC FA FA Goiter FVPTC Goiter Peptide CPTC FVPTC FVPTC FA FVPTC FC FC FA FA Goiter FVPTC Goiter
P*
Sensibility (%)
Specificity (%)
Accuracy (%)
0.2691 <0.0001* <0.0001* 0.6253 <0.0001* <0.0001*
91.67 81.82 78.79 24.39 85.19 81.82
18.18 85.19 75.61 85.19 27.66 72.34
57.74 91.92 88.77 52.94 76.36 83.11
0.0199* 0.0254* 0.8680 0.0512* 0.0583 <0.0001*
91.43 76.67 73.33 70.00 46.68 73.33
26.67 55.32 30.00 55.32 72.73 70.45
66.86 65.18 51.17 62.18 61.53 78.48
Abbreviations: FA = follicular adenoma; CPTC = classic type PTC; FVPTC = follicular variant PTC; FC = follicular carcinoma. * P < 0.05.
presented in Table 5, and apparently the peptide is a cytokeratinspecific ligand. 4. Discussion Identification of malignant thyroid nodules from indeterminate follicular lesions is difficult and biomarkers can barely distinguish follicular variants of papillary thyroid carcinomas from simple benign hyperplasia. We have used a subtractive phage display selection against goiter tissue aiming the improvement of thyroid cancer diagnosis. We have chosen this technology due to its
successful applications in biomarkers’ discovery, especially when short random peptides are selected against specific target molecules, such as those with antitumor activity [20] or for drug delivery [28]. A highly reactive peptide, CaT12, although not specific for thyroid tissues, was able to identify malignancy among thyroid nodular lesions and characterize follicular patterned lesions. In fact, its accuracy to identify follicular patterned lesions is comparable to most of the currently used immunohistochemical markers employed in differential diagnosis of indeterminate nodules [29,30]. Moreover, this peptide is associated to features of DTC aggressiveness, such as age and the occurrence of metastasis, suggesting a potential clinical use. We observed different results concerning the CaT12 recombinant clone and its synthetic peptide, with increased staining and positivity for the phage clone in cancer samples. This fact may be due to the specific epitope conformation generated by the fusion of the peptide with the N-terminal region of the bacteriophage pIII protein. On the other hand, higher reactivity of the peptide was observed in controls (goiter and adenomas), probably because of the lower affinity of the peptide alone, without the conformational rearrangement of the peptide-pIII fusion. Therefore, there is a potential contribution of phage-coat proteins for the peptide affinity as discussed elsewhere [31,32]. CaT12 strongly stained both the membrane and the cytoplasm of papillary thyroid cancer cells. Since normal thyroid cells do not stain, its protein target may be produced during physiological cellular growth. Benign thyroid lesions, such as follicular adenomas and hyperplasia, presented a very weak staining, reinforcing the hypothesis that we are dealing with a protein involved in cellular growth. We have identified several protein targets of the
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
6
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
Fig. 3. Immunohistochemistry with CaT12 peptide in different tissues. Cancer tissues: Positive: (A1) kidney; (A2) prostate; (A3) breast. Negative: (B1) kidney; (B2) prostate; (B3) breast. Non-neoplastic tissues: Positive: (C1) bladder; (C2) respiratory epithelium; (C3) small intestine. Negative: (D1) parathyroid; (D2) nevus; (D3) muscle.
CaT12 ligand, mostly cytokeratins, which are currently being tested. In fact, serum cytokeratin fragments have been found in thyroid nodule patients [33] and have even been associated with the risk of metastases [34]. Interestingly, cytokeratin-19 immunohistochemistry is largely employed in the differential diagnosis of follicular patterned lesions [29], although it was not recognized in our mass spectrometry analysis. However, due to the smaller specificity and
binding affinity of the synthetic peptide in comparison to the recombinant phage, it is possible that the peptide may partially bind to a conserved cytokeratin domain, which may explain the alignment (hits) of digested protein fragments with many cytokeratin sequences. On the other hand, the higher specificity found for the CaT12 recombinant phage suggests that its fused peptide sequence may be partially complemented by the amino terminal sequence of the pIII phage capsid protein improving its binding affinity.
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx Table 4 Immunohistochemistry positivity of the CaT12 peptide in different tissues. Histopathological diagnosis
Visual analysis N
Positive cases (%)
Non-malignant Nevus atypical Nevus communes Nonneoplastic tissues* Normal breast Normal prostate Normal kidney Total
12 29 206 5 5 5 262
0 3.45 15.53 0 0 60 13.74
Malignant Breast cancer Prostate cancer Kidney cancer Total
189 57 35 281
53.96 70.18 74.29 59.79
* Different human nonneoplastic tissues including pancreas, liver, lung, breast, prostate and brain.
Table 5 Identification of putative protein targets of the Cat12 peptide ligand by LC-MS/MS. Protein name Keratin, type Keratin, type Keratin, type Keratin, type Keratin, type Hornerin Keratin, type Keratin, type
II cytoskeletal 1 I cytoskeletal 9 I cytoskeletal 10 I cytoskeletal 14 II cytoskeletal 2 II cytoskeletal 6A II cytoskeletal 5
Acession number
MW (KDa)
P04264 P35527 P13645 P02533 P35908 Q86YZ3 P02538 P13647
66 62 59 52 65 282 60 62
In conclusion, we have designed a novel approach to select specific peptide ligands to thyroid cancer cells, which allowed us to identify a highly reactive antibody-like peptide that maybe readily used in the cellular diagnosis of thyroid nodules, discriminating indeterminate benign follicular lesions from carcinomas with high accuracy. This specific ligand may also be a potential therapeutic agent or drug carrier for future treatment strategies of thyroid cancer. Conflict of Interests The authors declare no conflict of interests. Acknowledgements The present work was supported by the Brazilian funding agencies: Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). The authors would like to thank Dr. Janete Maria Cerutti (Unifesp) for donating the NPA and TPC1 cell lines, Dr. José Vasallo (A.C. Camargo) for performing final IHC analysis, and A.C. Camargo Hospital for providing tissue microarrays. References [1] L. Davies, H.G. Welch, Increasing incidence of thyroid cancer in the United States, 1973–2002, JAMA: J. Am. Med. Assoc. 295 (2006) 2164–2167. [2] R. Siegel, C. DeSantis, K. Virgo, K. Stein, A. Mariotto, T. Smith, D. Cooper, T. Gansler, C. Lerro, S. Fedewa, C. Lin, C. Leach, R.S. Cannady, H. Cho, S. Scoppa, M. Hachey, R. Kirch, A. Jemal, E. Ward, Cancer treatment and survivorship statistics, CA: Cancer J. Clin. 62 (2012) 220–241. [3] L.S. Ward, H. Graf, Thyroid cancer: increased occurrence of the disease or simply in its detection?, Arq Bras. endocrinol. Metabol. 52 (2008) 1515–1516.
7
[4] B. Aschebrook-Kilfoy, M.H. Ward, M.M. Sabra, S.S. Devesa, Thyroid cancer incidence patterns in the United States by histologic type, 1992–2006, Thyroid: Official J. Am. Thyroid Assoc. 21 (2011) 125–134. [5] D.S. Cooper, G.M. Doherty, B.R. Haugen, R.T. Kloos, S.L. Lee, S.J. Mandel, E.L. Mazzaferri, B. McIver, F. Pacini, M. Schlumberger, S.I. Sherman, D.L. Steward, R.M. Tuttle, Revised American thyroid association management guidelines for patients with thyroid nodules and differentiated thyroid cancer, Thyroid: Official J. Am. Thyroid Assoc. 19 (2009) 1167–1214. [6] S. Bose, A.E. Walts, Thyroid fine needle aspirate: a post-Bethesda update, Adv. Anat. Pathol. 19 (2012) 160–169. [7] C.C. Wang, L. Friedman, G.C. Kennedy, H. Wang, E. Kebebew, D.L. Steward, M.A. Zeiger, W.H. Westra, Y. Wang, E. Khanafshar, G. Fellegara, J. Rosai, V. Livolsi, R.B. Lanman, A large multicenter correlation study of thyroid nodule cytopathology and histopathology, Thyroid: Official J. Am. Thyroid Assoc. 21 (2011) 243–251. [8] E.K. Alexander, G.C. Kennedy, Z.W. Baloch, E.S. Cibas, D. Chudova, J. Diggans, L. Friedman, R.T. Kloos, V.A. LiVolsi, S.J. Mandel, S.S. Raab, J. Rosai, D.L. Steward, P.S. Walsh, J.I. Wilde, M.A. Zeiger, R.B. Lanman, B.R. Haugen, Preoperative diagnosis of benign thyroid nodules with indeterminate cytology, New England J. Med. 367 (2012) 705–715. [9] L. Yip, C. Farris, A.S. Kabaker, S.P. Hodak, M.N. Nikiforova, K.L. McCoy, M.T. Stang, K.J. Smith, Y.E. Nikiforov, S.E. Carty, Cost impact of molecular testing for indeterminate thyroid nodule fine-needle aspiration biopsies, J. Clin. Endocrinol. Metab. 97 (2012) 1905–1912. [10] S. Carnazza, C. Foti, G. Gioffre, F. Felici, S. Guglielmino, Specific and selective probes for Pseudomonas aeruginosa from phage-displayed random peptide libraries, Biosensors Bioelectron. 23 (2008) 1137–1144. [11] C.G. Nicol, L. Denby, O. Lopez-Franco, R. Masson, C.A. Halliday, S.A. Nicklin, A. Kritz, L.M. Work, A.H. Baker, Use of in vivo phage display to engineer novel adenoviruses for targeted delivery to the cardiac vasculature, FEBS Lett. 583 (2009) 2100–2107. [12] X. Yu, G.P. Owens, D.H. Gilden, Rapid and efficient identification of epitopes/ mimotopes from random peptide libraries, J. Immunol. Methods 316 (2006) 67–74. [13] F. Bai, J. Liang, J. Wang, Y. Shi, K. Zhang, S. Liang, L. Hong, H. Zhai, Y. Lu, Y. Han, F. Yin, K. Wu, D. Fan, Inhibitory effects of a specific phage-displayed peptide on high peritoneal metastasis of gastric cancer, J. Mol. Med. (Berl) 85 (2007) 169– 180. [14] L. Chen, Y. Wang, X. Liu, S. Dou, G. Liu, D.J. Hnatowich, M. Rusckowski, A new TAG-72 cancer marker peptide identified by phage display, Cancer Lett. 272 (2008) 122–132. [15] B. Du, M. Qian, Z. Zhou, P. Wang, L. Wang, X. Zhang, M. Wu, P. Zhang, B. Mei, In vitro panning of a targeting peptide to hepatocarcinoma from a phage display peptide library, Biochem. Biophys. Res. Commun. 342 (2006) 956–962. [16] A. Dybwad, P. Lambin, M. Sioud, M. Zouali, Probing the specificity of human myeloma proteins with a random peptide phage library, Scand. J. Immunol. 57 (2003) 583–590. [17] A.L. Goenaga, Y. Zhou, C. Legay, H. Bougherara, L. Huang, B. Liu, D.C. Drummond, D.B. Kirpotin, C. Auclair, J.D. Marks, M.A. Poul, Identification and characterization of tumor antigens by using antibody phage display and intrabody strategies, Mol. Immunol. 44 (2007) 3777–3788. [18] S. Hu, X. Guo, H. Xie, Y. Du, Y. Pan, Y. Shi, J. Wang, L. Hong, S. Han, D. Zhang, D. Huang, K. Zhang, F. Bai, H. Jiang, H. Zhai, Y. Nie, K. Wu, D. Fan, Phage display selection of peptides that inhibit metastasis ability of gastric cancer cells with high liver-metastatic potential, Biochem. Biophys. Res. Commun. 341 (2006) 964–972. [19] S. Liang, T. Lin, J. Ding, Y. Pan, D. Dang, C. Guo, M. Zhi, P. Zhao, L. Sun, L. Hong, Y. Shi, L. Yao, J. Liu, K. Wu, D. Fan, Screening and identification of vascularendothelial-cell-specific binding peptide in gastric cancer, J. Mol. Med. (Berl) 84 (2006) 764–773. [20] A.L. Matsuo, A.S. Tanaka, M.A. Juliano, E.G. Rodrigues, L.R. Travassos, A novel melanoma-targeting peptide screened by phage display exhibits antitumor activity, J. Mol. Med. (Berl) 88 (2010) 1255–1264. [21] C. Mazuet, D. Lerouge, M.A. Poul, N. Blin, Breast carcinoma specific antibody selection combining phage display and immunomagnetic cell sorting, Biochem. Biophys. Res. Commun. 348 (2006) 550–559. [22] W.D. Jia, H.C. Sun, J.B. Zhang, Y. Xu, Y.B. Qian, J.Z. Pang, L. Wang, L.X. Qin, Y.K. Liu, Z.Y. Tang, A novel peptide that selectively binds highly metastatic hepatocellular carcinoma cell surface is related to invasion and metastasis, Cancer Lett. 247 (2007) 234–242. [23] T. Oyama, K.F. Sykes, K.N. Samli, J.D. Minna, S.A. Johnston, K.C. Brown, Isolation of lung tumor specific peptides from a random peptide library: generation of diagnostic and cell-targeting reagents, Cancer Lett. 202 (2003) 219–230. [24] L. Zang, L. Shi, J. Guo, Q. Pan, W. Wu, X. Pan, J. Wang, Screening and identification of a peptide specifically targeted to NCI-H1299 from a phage display peptide library, Cancer Lett. 281 (2009) 64–70. [25] R.J. Giordano, M. Cardo-Vila, J. Lahdenranta, R. Pasqualini, W. Arap, Biopanning and rapid analysis of selective interactive ligands, Nat. Med. 7 (2001) 1249–1253. [26] J. Huang, B. Ru, P. Dai, Bioinformatics resources and tools for phage display, Molecules 16 (2011) 694–709. [27] S. Mandava, L. Makowski, S. Devarapalli, J. Uzubell, D.J. Rodi, RELIC – a bioinformatics server for combinatorial peptide analysis and identification of protein-ligand interaction sites, Proteomics 4 (2004) 1439–1460. [28] B. Du, H. Han, Z. Wang, L. Kuang, L. Wang, L. Yu, M. Wu, Z. Zhou, M. Qian, Targeted drug delivery to hepatocarcinoma in vivo by phage-displayed specific binding peptide, Mol. Cancer Res.: MCR 8 (2010) 135–144.
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039
8
C.F. Reis et al. / Cancer Letters xxx (2013) xxx–xxx
[29] P.S. de Matos, A.P. Ferreira, F. de Oliveira Facuri, L.V. Assumpcao, K. Metze, L.S. Ward, Usefulness of HBME-1, cytokeratin 19 and galectin-3 immunostaining in the diagnosis of thyroid malignancy, Histopathology 47 (2005) 391–401. [30] K. Sethi, S. Sarkar, S. Das, B. Mohanty, M. Mandal, Biomarkers for the diagnosis of thyroid cancer, J. Exp. Therap. Oncol. 8 (2010) 341–352. [31] C.F. Barbas, Phage Display: a Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 2001. [32] M.B. Zwick, L.L. Bonnycastle, K.A. Noren, S. Venturini, E. Leong, C.F. Barbas 3rd, C.J. Noren, J.K. Scott, The maltose-binding protein as a scaffold for monovalent
display of peptides derived from phage libraries, Anal. Biochem. 264 (1998) 87–97. [33] L. Giovanella, L. Ceriani, A. Ghelfo, M. Maffioli, Circulating cytokeratin 19 fragments in patients with benign nodules and carcinomas of the thyroid gland, Int. J. Biol. Markers 23 (2008) 54–57. [34] M. Appetecchia, A. Mecule, M. Ducci, L. Palma, M. Castelli, Serum cytokeratins determination in differentiated thyroid carcinoma, J. Exp. Clin. Cancer Res.: CR 20 (2001) 253–256.
Please cite this article in press as: C.F. Reis et al., An antibody-like peptide that recognizes malignancy among thyroid nodules, Cancer Lett. (2013), http:// dx.doi.org/10.1016/j.canlet.2013.02.039