Purpose: Cancer diagnostics and therapeutics are often based on clinically relevant markers that are expressed specifically in a malignant tissue at levels higher than in normal tissue. We examined potential markers for papillary thyroid carcinoma (PTC) by monitoring PTC-specific gene expression using cDNA microarray.

Experimental Design: Gene expression profiles for PTC tissue, normal thyroid tissue, and healthy peripheral blood cells were compared by use of a human 4000-gene cDNA microarray. Protein expressions of the up-regulated genes in PTC were examined in thyroid tissues by immunohistochemistry.

Results: Sixty-four genes were overexpressed in PTC tissue relative to normal thyroid tissue and healthy peripheral blood cells. The genes that were up-regulated in PTC were involved in cell cycle regulation, DNA damage response, angiogenesis, and oncogenesis. Among these genes, basic fibroblast growth factor and platelet-derived growth factor were identified by immunochemical methods as proteins that are specifically expressed at high levels in thyroid neoplasms. Basic fibroblast growth factor, which has been identified as a biomarker for PTC, was overexpressed in 54% of PTC cases, 67% of follicular thyroid carcinomas, and 36% of benign thyroid neoplasms. Platelet-derived growth factor was overexpressed in 81% of PTC cases and 100% of follicular carcinomas, but was immunonegative in normal thyroid tissues and benign thyroid neoplasms.

Conclusions: Platelet-derived growth factor may be a potential biomarker for PTC and follicular carcinoma. Expression profile analysis using a microarray followed by immunohistochemical study can be used to facilitate the development of molecular biomarkers for cancer.

Papillary thyroid carcinoma (PTC) is the most common malignant thyroid tumor, representing 80–90% of all thyroid malignancies. Papillary, follicular, and anaplastic thyroid carcinomas arise from follicular cells; medullary thyroid carcinomas arise from the parafollicular epithelium of the thyroid. PTC is usually well differentiated; however, the clinical behavior of PTC varies widely. For example, incidental microcarcinomas grow very slowly and are noninvasive or minimally invasive. On the other hand, invasive PTC with metastasis can be lethal. PTC often recurs many years after surgical removal. The prognosis for PTC is often favorable; however, ∼20% of PTC tumors recur (1), and some reach advanced stages. Postoperative follow-up for diagnosing recurrence is important for a favorable outcome. Although PTC can be diagnosed by fine-needle aspiration cytology, this method is not sufficiently reliable for predicting postoperative clinical outcome. Serum thyroglobulin (TG) has been monitored by immunoassay to detect the recurrence of differentiated thyroid carcinoma. However, measurement of serum TG is sometimes hindered by the presence of circulating factors, particularly anti-TG antibodies (2), and residual normal thyroid (NT) gland tissue that produces TG. Thus, a reliable diagnostic molecular marker for PTC would be extremely valuable for improving cancer detection and the prognosis of patients with PTC.

A common approach for identifying circulating cancer cells in peripheral blood is the reverse transcription-PCR (RT-PCR) method for amplifying tumor-specific mRNA (3, 4). Another possible approach is to measure abnormal or ectopic tissue-specific gene expression in peripheral blood. Previous studies have investigated the thyroid-specific markers TG, thyroid peroxidase (TPO), Ret/PTC1, and human sodium/iodide symporter (NIS) as diagnostic markers of thyroid carcinomas (5, 6, 7, 8). However, expression of TG is reportedly not thyroid specific and is not correlated with thyroid cancer (9, 10). Although RT-PCR detection of a marker for circulating cancer cells in peripheral blood would be an effective noninvasive method, a reliable marker for PTC has not yet been identified.

Use of a cDNA microarray is a powerful method for the quantitative analysis of cancer-specific gene expression (11, 12) that can detect altered gene expression associated with the pathology or the altered biology of cancer cells. A cDNA microarray can be used to identify potential diagnostic markers for cancer by measuring tumor-specific expression of thousands of genes in hundreds of tumors. Candidate genes for diagnostic markers can also be characterized by analyzing the gene expression profiles of a small number of cancer tissues in combination with the further large-scale immunohistochemical analysis of protein expression.

In this study we examined potential PTC-specific molecular markers that could be used to identify circulating cancer cells in peripheral blood or tumor-associated proteins in serum.

Tissue and RNA Extraction.

Thyroid tissue was obtained from patients undergoing surgery for thyroid disease at University Hospital, University of Tsukuba (Tsukuba, Japan). Tissues were immediately frozen in liquid nitrogen and stored at −80°C until use. The protocol was approved by the Institutional Review Board, and informed consent was obtained. Seven highly differentiated papillary carcinomas were chosen for microarray studies. Paired samples of normal and tumoral thyroid tissues from the same patients were studied. Histologically verified NT tissues were obtained from four patients undergoing thyroidectomy for a solitary benign thyroid neoplasm. Peripheral blood cells (PBCs) were obtained from whole blood of healthy volunteers. Total RNA was obtained with a commercial kit (ISOGEN; Nippon Gene, Tokyo, Japan) according to manufacturer’s protocol. The RNA quality was confirmed by subjecting the sample to electrophoresis on a formaldehyde-agarose gel and staining the 18S and 28S RNA bands for visualization.

Microarray Preparation.

A 5-μl aliquot of sequence-verified human 3968 cDNA (Invitrogen, Carlsbad, CA) was added to 2.5 ml of Terrific broth (Difco, Detroit, MI) containing ampicillin and incubated at 37°C overnight. Plasmid DNA was extracted from each culture in a 96-well format by use of an alkaline-SDS method with a MultiScreen plate (Millipore, Bedford, MA). The microarray membrane was prepared with human 3968 cDNA, which contains genes named in the UniGene database (National Center for Biotechnology Information, Bethesda, MD). The cDNA insert was diluted and amplified by standard PCR protocols using the primer set GF-FW (5′-CTG CAA GGC GAT TAA GTT GGG TAA C-3′) and GF-RV (5′-GTG AGC GGA TAA CAA TTT CAC ACA GGA AAC AGC-3′). The amplification products, including thyroid-specific cDNAs, were purified by use of Multi Screen-PCR plates (Millipore) and spotted on Hybond-N Plus membranes (Amersham, Buckinghamshire, United Kingdom) with a GTMASS arrayer (NLE, Nagoya, Japan).

Hybridization and Data Analysis.

A 2-μg portion of total RNA was dissolved in 8 μl of RNase-free water; 2 μl of oligo(dT) primer (1 μg/μl; Invitrogen) were then added. The sample was heated for 10 min at 70°C and immediately chilled on ice. The following were then added to the sample: 5 μl of first-strand cDNA synthesis buffer, 1 μl of DTT (0.1 m final concentration), 1.5 μl of deoxynucleotide triphosphates (20 mm dATP, dGTP, and dTTP), 1.5 μl of SuperScript (200 units/μl; Invitrogen), and [α-33P]dCTP (3000 Ci/mmol; Amersham). The reaction was incubated at 37°C for 90 min. Labeled cDNA was then purified by use of Bio-Spin 6 chromatography columns (Bio-Rad Laboratories, Hercules, CA) and denatured for 3 min at 95°C. The microarray was prehybridized at 42°C for 2 h in MicroHyb solution (Invitrogen) with 0.5 μg/ml poly(dA) and 1 μg/ml COT1 DNA (Invitrogen). The probes were hybridized at 42°C overnight in the same solution. The membrane microarray was washed twice at 50°C with 2× SSC-1% SDS for 20 min and at room temperature with 0.5× SSC-1% SDS for 15 min. The membrane microarray was exposed to an imaging plate, which was scanned with a BAS 5000 imaging analyzer (Fuji Film, Tokyo, Japan). Spot intensity was quantified with ArrayVision 5.0 software (Imaging Research, Ontario, Canada). GeneSpring 5.0 software (Silicon Genetics, Redwood, CA) was used to normalize values for each gene and each microarray, and for data analysis.

The accuracy of microarray data was determined by six repeat analyses of the same RNA from NT tissue. Probes were stripped by boiling with 0.5% SDS before rehybridization.

Northern Blot Analysis.

Aliquots containing 10 μg of total RNA were electrophoresed on a 1% formaldehyde-agarose gel, transferred to a nylon membrane, and immobilized by baking the filters at 80°C for 2 h. The membrane was prehybridized with Church’s solution (0.5 m sodium phosphate-7% SDS-1 mm EDTA) at 68°C for 2 h. The DNA fragments for adrenergic receptor β kinase 1, myeloid leukemia factor, TG, TPO, thyroid-stimulating hormone receptor (TSHR), and glyceraldehyde-3-phosphate dehydrogenase were amplified from cDNA and sequenced by a standard protocol. After sequence confirmation, the DNA fragment was labeled with [α-32P]dCTP by a random primer cDNA labeling method with the commercial Megaprime DNA Labeling System (Amersham). The heat-denatured probe was added to the prehybridization solution and incubated at 68°C overnight.

Real-Time Quantitative PCR for Determination of mRNA Levels.

The gene expression levels of NIS, TPO, and the Ret proto-oncogene (RET) were examined by real-time quantitative PCR. A 1-μg aliquot of total RNA for each sample was reverse-transcribed in a 100-μl reaction volume with a commercial High-Capacity cDNA Archive Kit (Applied Biosystems PE) according to manufacturer’s protocol. The quantitative PCR reaction was carried out in 96-well microtiter plates on the ABI Prism Sequence Detector 7900 (Applied Biosystems) with a commercial Assays-on-Demand kit (Applied Biosystems). The probe sequences for PCR were as follows: 5′-GGC GCT CTG CAC CAG ATC ATC ACC C-3′ for TPO; 5′-GGC TCC TGG GAG GGT GAG GGT CTG CC-3′ for RET; and 5′-AGC TGC CTG ACA GGC CCC ACC AAG C-3′ for NIS. For each sample of each gene, PCR amplification was performed in triplicate with 18S rRNA used as an endogenous control. The mRNA content of each target was determined simultaneously in three paired normal/tumoral samples studied by DNA array analysis.

Immunohistochemistry.

Archival formaldehyde-fixed, paraffin-embedded blocks were obtained from Tsukuba University Hospital. Sections were cut 5-μm thick, deparaffinized, and rehydrated. Antigen retrieval for platelet-derived growth factor (PDGF) antibody staining was performed with 0.01 m citrate buffer. Endogenous peroxidase activity was inactivated with a blocking agent. After washing, the sections were exposed to a 1:100 solution of antihuman basic fibroblast growth factor (bFGF) monoclonal antibody (WAKO Pure Chemical Industries, Osaka, Japan) or PDGF polyclonal antibody (WAKO Pure Chemical Industries) at 4°C overnight and incubated with peroxidase-labeled dextran polymer (Envision Plus System; DAKO, Kyoto, Japan) for 1 h. Staining without primary antibodies was used as a negative control. For visualization, tissue sections were soaked in medium containing 3,3-diaminobenzidine tetrahydrochloride and H2O2 and were counterstained with hematoxylin. Cytoplasmic staining was assessed by identifying and scoring 100 cells within a randomly selected light microscopic field (×400 magnification). The cytoplasm of those cells was scored as 0 (no staining), 1 (mild staining), 2 (moderate staining), or 3 (heavy staining) by two independent observers. The results are presented as proportions of 1000 cells (fields from multiple slides were examined) from which the histochemical score for determining the level of expression was then calculated (13). Expression was considered to be positive when the histochemical score was >75.

Cytological specimens were obtained intraoperatively by fine-needle aspiration. Cells were smeared on the slide and fixed in formaldehyde. The PDGF protein expression was examined as described above.

Microarray Quality Control.

We estimated the reproducibility and accuracy of the microarray data by carrying out six repeat experiments using the 3968 cDNA microarray and a NT cDNA probe. Data were collected for the same probe hybridized to four membranes, two of which were then stripped and rehybridized, generating six data sets (data not shown). The same membranes were used in experiments 1 and 5 and in experiments 2 and 6. The average normalized spot intensity for each data set was 1.0, and the SD was 0.7.

Hierarchical Clustering Analysis.

Expression profiles were analyzed for seven PTC samples and seven NT samples, including three pairs of normal/tumoral thyroid tissues. The cutoff for differential expression associated with PTC was set at 2.0-fold. Clustering analysis was performed with the model-based approach with GeneSpring software and showed a similar expression pattern among PTC samples (Fig. 1). Most of the NT tissue samples were clustered in one group, and the tumors were clustered in another group. The NT and PTC groups were separated by the differential expression of genes; these differences were determined to be statistically significant by the t test (<0.05). Several genes showed up-regulated expression patterns in PTC (Table 1). The proteins encoded by the altered genes are associated with cell cycle regulators, growth regulators, oncogenes, and DNA damage regulators.

Corroboration of Gene Expression.

Northern blot analysis was used to confirm the results of the microarray studies. Northern blots were prepared with the samples used in the microarray study plus one follicular carcinoma sample and two follicular adenoma samples (Fig. 2,A). The blots were hybridized with probes for β-adrenergic receptor kinase 1, myeloid leukemia factor, and TG. The results confirmed microarray data showing that these genes are up-regulated in PTC (Fig. 2 B). Expression of TPO was down-regulated in PTC and follicular carcinoma samples. The gene TSHR was expressed at a lower level in two of the PTC samples than in the NT samples. The mRNA for TPO, TSHR, and TG was not detected in PBCs.

We analyzed the differential expression of three genes related with thyroid disease in five paired normal/tumoral thyroid tissue samples by quantitative PCR. The levels of expression determined by RT-PCR correlated with the results obtained by microarray analysis (data not shown). Wild-type RET showed a minute amount of mRNA in both normal and well-differentiated PTC tissues.

PTC-Specific Gene Expression.

Genes that were differentially expressed in PTC compared with paired normal tissue obtained from the same patients were also detected. We focused on those genes that showed up-regulated expression in at least one NT/PTC pair and statistically different expression from other NT samples. This screening identified 64 genes showing a >2-fold increase in expression. Table 2 summarizes the tumor-related genes that showed PTC-specific expression (PTC/NT ratio).

TG, PDGF, and FGF expression levels are shown in Fig. 3,A. The TG expression pattern was predominant in PTC samples; however, its up-regulated signal was also seen in healthy PBCs. On the other hand, PDGF and FGF expression was low in PBCs. We compared expression of the thyroid-specific genes TG, TPO, TSHR, and NIS and of RET between paired PTC and NT samples (Fig. 3 B). TG was up-regulated in PTC samples, whereas the levels of TPO, RET, and NIS were reduced. The expression of TSHR was down-regulated in two PTC samples.

PTC-Specific Protein Expression.

Because bFGF and PDGF were identified as potential diagnostic tools for PTC from their gene expression, their protein expression was also examined in thyroid tissues. Protein expression was examined by immunohistochemical analysis of 91 thyroid tissue sections [55 differentiated PTCs, 4 follicular variant PTCs, 6 follicular thyroid carcinomas, 5 hyperfunctioning thyroid tissues, 11 benign thyroid neoplasms, and 10 NT samples]; these samples included the tissues analyzed by microarray. The results of the immunohistochemical analysis are summarized in Table 3. Thirty-four of 59 (54%) PTCs were positive for bFGF, but none of the 10 NT samples were positive. Positive staining for bFGF was also observed in hyperfunctioning thyroid tissues and benign neoplasms (Fig. 4). Thyroid PTC and follicular thyroid carcinoma neoplasms showed increased proportions of positive immunostaining for PDGF: 48 of 59 (81%) cases of PTC and 6 of 6 (100%) cases of follicular thyroid carcinoma. On the other hand, follicular cells in NT tissue from a multinodular goiter and an adenoma were negative for PDGF immunostaining. As shown in Fig. 5, immunochemical analysis showed that neoplastic cells retrieved by fine-needle aspiration biopsy were immunoreactive for PDGF.

In this study microarray analysis showed that the genes bFGF and PDGF were overexpressed in PTC. This result confirms previous reports that the bFGF protein is detected in PTCs (14, 15); however, studies of bFGF or PDGF at the mRNA level or of PDGF at the protein level have not been done. The bFGF isolated from PTC samples was reported to be similar to bFGF from other sources and was reported to play an important role in the angiogenesis and proliferation of parenchymal cells (16). In this study, we found that PDGF was a potential diagnostic marker of PTC. The expression of PDGF was detected in PTCs but not in normal tissues by immunohistochemical staining. The bFGF antibody displayed increased protein levels in malignant neoplasms as well as in benign thyroid diseases. On the other hand, PDGF showed a high frequency of positive staining in cases of thyroid malignancy (81% of PTCs), and its expression was specifically detected in malignant but not benign thyroid diseases. These results showed that PDGF is a potential diagnostic marker. Fujimoto et al.(17) previously suggested that serum bFGF could be used as a highly specific diagnostic marker for renal cell carcinoma. Similarly, serum PDGF might be used as a diagnostic marker for PTC, and this possibility is under investigation.

Cytological specimens from fine-needle aspirates were also immunoreactive for PDGF. Physicians make their diagnosis of PTC on the basis of morphology from fine-needle aspirate biopsy samples, but not immunochemistry. Immunostaining these samples for PDGF may be helpful for identifying PTC. On the basis of our results for fine-needle aspirate cytology, thyroid tumors could be more accurately diagnosed by a combination of morphological diagnosis and detection of PDGF expression.

Microarray analysis was performed on 4000 human cDNA arrays and RNA from 7 PTC samples and 7 normal samples, including 3 PTC/NT pairs. The reproducibility of the microarray data was confirmed. Clustering analysis classified samples as PTC or NT and identified 64 genes whose expression is up-regulated in PTC but low or undetectable in NT and PBCs. Although more clinical cases need to be studied, the expression profiles of selected genes in these PTC samples may be representative of this type of cancer cell.

Of the 64 genes with up-regulated expression, several genes may be candidates as diagnostic markers for thyroid carcinoma. Such markers could be particularly useful for monitoring patients postoperatively. An ideal biomarker for cancer in its early diagnosis should have both tumor and tissue specificity. Our microarray screening did not give any information about tissue specificity. Although a protein that shows both tumor and tissue specificities might be rare, a detailed study of the candidate genes detected by the microarray in this study might successfully find such biomarkers.

Previous studies have similarly identified genes up-regulated in PTC [i.e., fibronectin 1 (18), galectin-3 (19), S100 calcium-binding proteins (20, 21), and insulin receptor (22)], indicating that these genes may be involved in the development of thyroid carcinogenesis. From previous microarray studies (23, 24), it is difficult to identify common up-regulated genes in PTC. Overlapping gene expression among microarray studies may depend on the microarray probe design, which may explain some of the overlap between studies. However, several of the genes related to cell cycle, carcinogenesis, oncogene, and growth factors showed similar expression patterns in our and other studies.

The thyroid follicular cell is a highly differentiated cell that responds to thyroid-stimulating hormone, traps iodine, and synthesizes TG and TPO. Thus, TG, TPO, and TSHR can be used as markers for the differentiation of follicular cells. In this study, TG was expressed at a high level in PTC. However variable levels of TG expression have been observed in thyroid malignancies (25, 26, 27, 28, 29, 30). Some studies reported that the expression was not thyroid specific (9, 10); in fact, we found TG expression in healthy PBCs. We selected three thyroid-related genes for additional analysis to validate the microarray expression data. In this study, we found that the gene expression levels of TPO and NIS were decreased; this supports previous findings (28, 30, 31, 32). The gene RET, a tyrosine kinase receptor, may be a major proto-oncogene in thyroid carcinogenesis: RET rearrangement is the most common genetic alternation found in PTC (33, 34, 35). In this study we did not detect RET gene expression in NT tissue and found different levels of expression in PTC samples. These data coincide with previous reports using RT-PCR (36).

In summary, this study demonstrates that microarray technology can be used to develop molecular markers for cancer diagnosis. Some of the genes shown to have strong expression might be used as serum tumor markers of thyroid cancers. In particular, gene expression profiling was used to identify genes differentially regulated in PTC. From this set of genes, PDGF was identified as a strong candidate diagnostic marker for PTC because it was highly expressed in 81% of PTC cases at the protein level. Antibodies to PDGF could be extremely useful in future studies to evaluate the clinical potential of PDGF as a diagnostic marker for PTC. The expression profiling data presented here may also improve our understanding of thyroid carcinogenesis.

Grant support: Supported in part by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports, and Culture of Japan.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Kazuhiko Uchida, Department of Biochemistry and Molecular Oncology, Institute of Basic Medical Sciences, University of Tsukuba, 1-1-1 Tennoudai, Tsukuba, Ibaraki 305-8575, Japan. Phone: 81-29-853-3115/3070; Fax: 81-29-853-3271; E-mail: yanoyuki@rg7.so-net.ne.jp

Fig. 1.

Clustering diagram for human 3968 genes. Genes with similar expression patterns are clustered together. The lengths of the tree branches reflect the degree of similarity of gene expression. Most papillary thyroid carcinoma (PTC) samples are clustered in one group, and the normal thyroid (NT) samples are clustered in another group. Each horizontal block represents a single gene. The color scale correlates color with relative expression. The basal expression of each gene is represented in yellow. A shift toward red indicates an increase in expression, whereas a shift toward blue indicates a decrease in expression. PBC, peripheral blood cells.

Fig. 1.

Clustering diagram for human 3968 genes. Genes with similar expression patterns are clustered together. The lengths of the tree branches reflect the degree of similarity of gene expression. Most papillary thyroid carcinoma (PTC) samples are clustered in one group, and the normal thyroid (NT) samples are clustered in another group. Each horizontal block represents a single gene. The color scale correlates color with relative expression. The basal expression of each gene is represented in yellow. A shift toward red indicates an increase in expression, whereas a shift toward blue indicates a decrease in expression. PBC, peripheral blood cells.

Close modal
Fig. 2.

Northern blot analysis. A, β-adrenergic receptor kinase 1 (ADRBK1) and myeloid leukemia factor 2 (MLF2) expression in papillary thyroid carcinoma (PTC) and normal thyroid (NT) samples. Thyroid peroxidase (TPO), thyroid-stimulating hormone receptor (TSHR), and thyroglobulin (TG) expression is apparent in PTC samples. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is shown as the control. B, comparison of MLF2 expressions in the microarray and Northern blot assays. FA, follicular adenoma; FC, follicular carcinoma.

Fig. 2.

Northern blot analysis. A, β-adrenergic receptor kinase 1 (ADRBK1) and myeloid leukemia factor 2 (MLF2) expression in papillary thyroid carcinoma (PTC) and normal thyroid (NT) samples. Thyroid peroxidase (TPO), thyroid-stimulating hormone receptor (TSHR), and thyroglobulin (TG) expression is apparent in PTC samples. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is shown as the control. B, comparison of MLF2 expressions in the microarray and Northern blot assays. FA, follicular adenoma; FC, follicular carcinoma.

Close modal
Fig. 3.

Gene expression levels in samples. Normalized intensities of the microarray data are shown. A, representation of up-regulated gene expression in papillary thyroid carcinoma (PTC). Increased expression in PTC is shown by thyroglobulin (TG), basic fibroblast growth factor (bFGF), and platelet-derived growth factor (PDGF). TG shows increased expression in peripheral blood cells (PBC). Samples 5, 6, and 7 are obtained from paired normal thyroid (NT)/PTC tissues. B, thyroid-related gene expression in paired NT/PTC tissues. TG expression is increased in PTC compared with paired NT tissue. Thyroid peroxidase (TPO), thyroid-stimulating hormone receptor (TSHR), wild-type Ret proto-oncogene (RET), and sodium/iodide symporter (NIS) gene expression is down-regulated in most cases of PTC.

Fig. 3.

Gene expression levels in samples. Normalized intensities of the microarray data are shown. A, representation of up-regulated gene expression in papillary thyroid carcinoma (PTC). Increased expression in PTC is shown by thyroglobulin (TG), basic fibroblast growth factor (bFGF), and platelet-derived growth factor (PDGF). TG shows increased expression in peripheral blood cells (PBC). Samples 5, 6, and 7 are obtained from paired normal thyroid (NT)/PTC tissues. B, thyroid-related gene expression in paired NT/PTC tissues. TG expression is increased in PTC compared with paired NT tissue. Thyroid peroxidase (TPO), thyroid-stimulating hormone receptor (TSHR), wild-type Ret proto-oncogene (RET), and sodium/iodide symporter (NIS) gene expression is down-regulated in most cases of PTC.

Close modal
Fig. 4.

Overexpression of basic fibroblast growth factor (bFGF) in papillary thyroid carcinoma. Thyroid tissues are stained with anti-bFGF antibody. Immunopositivity is visualized as a brown stain, whereas nuclear staining is blue. bFGF staining is localized primarily in tumor cells with weak staining for stroma (A). Positive bFGF staining is localized in the cytoplasm in cases of papillary thyroid carcinoma (B and C). Hyperfunctioning thyroid tissue (D) and thyroid follicular adenoma (E) showed weak immunopositive staining. No staining was observed in the normal follicular cell sample (F). Original magnifications are, as follows: A, ×100; B–F, ×400.

Fig. 4.

Overexpression of basic fibroblast growth factor (bFGF) in papillary thyroid carcinoma. Thyroid tissues are stained with anti-bFGF antibody. Immunopositivity is visualized as a brown stain, whereas nuclear staining is blue. bFGF staining is localized primarily in tumor cells with weak staining for stroma (A). Positive bFGF staining is localized in the cytoplasm in cases of papillary thyroid carcinoma (B and C). Hyperfunctioning thyroid tissue (D) and thyroid follicular adenoma (E) showed weak immunopositive staining. No staining was observed in the normal follicular cell sample (F). Original magnifications are, as follows: A, ×100; B–F, ×400.

Close modal
Fig. 5.

Overexpression of platelet-derived growth factor (PDGF) in papillary thyroid carcinoma (PTC). Immunopositivity is visualized as a brown stain, whereas nuclear staining is blue. Strong staining for anti-PDGF antibody is observed only in PTC tumor cells (A and B). The parafollicular cells of hyperfunctioning thyroid tissue show a weakly positive immunoreaction (C). Follicular adenoma (D) and normal thyroid (E) are negative for PDGF immunostaining. Positive staining for anti-PDGF antibody is observed in cytological specimens from PTC obtained by fine-needle aspiration biopsy (F). Original magnification for all panels is ×400.

Fig. 5.

Overexpression of platelet-derived growth factor (PDGF) in papillary thyroid carcinoma (PTC). Immunopositivity is visualized as a brown stain, whereas nuclear staining is blue. Strong staining for anti-PDGF antibody is observed only in PTC tumor cells (A and B). The parafollicular cells of hyperfunctioning thyroid tissue show a weakly positive immunoreaction (C). Follicular adenoma (D) and normal thyroid (E) are negative for PDGF immunostaining. Positive staining for anti-PDGF antibody is observed in cytological specimens from PTC obtained by fine-needle aspiration biopsy (F). Original magnification for all panels is ×400.

Close modal
Table 1

Genes up-regulated in papillary thyroid carcinoma compared with normal thyroid (not exhaustive)

Cell cycle control proteins Tetracycline transporter-like protein 
 Cyclin H, cyclin D1, SPHAR gene for cyclin-related protein 
 Heme oxygenase (decycling) 1 
Oncogene v-raf, putative oncogene, vav 2 
 V-rel avian reticuloendotheliosis viral oncogene 
Growth factor FGF,a VEGF-B, PDGF, placental growth factor 
 BMP5, TGF, β-induced 
 FHF-1, hepatocyte growth factor-like protein 
Growth factor receptor bFGF receptor, FGF 4 receptor 
 Insulin-like growth factor 1 receptor, growth factor-bound protein 
Phosphatase Inositol polyphosphate phosphatase-like protein 1 
 Nuclear dual-specificity phosphatase 
 MAP kinase phosphatase, M-phase inducer phosphatase 2 
 Inositol polyphosphate 4-phosphatase 
 Protein phosphatase-1 inhibitor, protein-tyrosine phosphatase 
 l-3-Phosphoserine-phosphatase 
Cell cycle control proteins Tetracycline transporter-like protein 
 Cyclin H, cyclin D1, SPHAR gene for cyclin-related protein 
 Heme oxygenase (decycling) 1 
Oncogene v-raf, putative oncogene, vav 2 
 V-rel avian reticuloendotheliosis viral oncogene 
Growth factor FGF,a VEGF-B, PDGF, placental growth factor 
 BMP5, TGF, β-induced 
 FHF-1, hepatocyte growth factor-like protein 
Growth factor receptor bFGF receptor, FGF 4 receptor 
 Insulin-like growth factor 1 receptor, growth factor-bound protein 
Phosphatase Inositol polyphosphate phosphatase-like protein 1 
 Nuclear dual-specificity phosphatase 
 MAP kinase phosphatase, M-phase inducer phosphatase 2 
 Inositol polyphosphate 4-phosphatase 
 Protein phosphatase-1 inhibitor, protein-tyrosine phosphatase 
 l-3-Phosphoserine-phosphatase 
a

FGF, fibroblast growth factor; VEGF, vascular endothelial growth factor; PDGF, platelet-derived growth factor; TGF, transforming growth factor; bFGF, basic fibroblast growth factor; MAP, mitogen-activated protein.

Table 2

Tumor-related genes up-regulated in papillary thyroid carcinoma

Acc. no.aGene name/ProductionValueb (fold)
R83837 V-yes-1 Yamaguchi sarcoma viral related oncogene 62 
AA486289 DNA polymerase α 37 
T81103 Sp2 transcription factor 34 
AA045822 Guanine nucleotide regulatory protein (tim) 7.6 
N33955 Protein phosphatase Wip1 4.7 
AA456291 GTP-binding protein 4.3 
AA598611 Immediate-early response protein 3.6 
AA400476 Mitotic centromere-associated kinesin 3.2 
R53541 Melanoma-associated chondroitin sulfate proteoglycan 3.1 
AA682337 Vav 2 oncogene 3.0 
AA283693 Osteoclast-stimulating factor 2.9 
AA045699 Tumor-associated antigen CO-029 2.8 
AA082943 Cyclin G1 2.8 
AA703660 H2A histone family 2.8 
H97778 E-cadherin 2.8 
H24326 Regulator of G-protein signaling (RGS7) 2.7 
H54020 Splicing factor (9G8) 2.5 
AA126911 Heterogeneous nuclear ribonucleoprotein A1 2.4 
H20856 DNA repair helicase ERCC3 2.4 
AA488332 Cell cycle protein p38-2G4 2.4 
AA608514 Histone H3.3 2.3 
AA704255 Spinocerebellar ataxia 7 2.2 
H95392 Histone H2A.X 2.2 
R53787 Protein phosphatase 2A B56-epsilon (PP2A) 2.1 
AA778077 Chromodomain helicase DNA binding protein 3 2.0 
AA701502 Platelet-derived growth factor PDGF-A 2.0 
AA088861 LI-cadherin 2.0 
Acc. no.aGene name/ProductionValueb (fold)
R83837 V-yes-1 Yamaguchi sarcoma viral related oncogene 62 
AA486289 DNA polymerase α 37 
T81103 Sp2 transcription factor 34 
AA045822 Guanine nucleotide regulatory protein (tim) 7.6 
N33955 Protein phosphatase Wip1 4.7 
AA456291 GTP-binding protein 4.3 
AA598611 Immediate-early response protein 3.6 
AA400476 Mitotic centromere-associated kinesin 3.2 
R53541 Melanoma-associated chondroitin sulfate proteoglycan 3.1 
AA682337 Vav 2 oncogene 3.0 
AA283693 Osteoclast-stimulating factor 2.9 
AA045699 Tumor-associated antigen CO-029 2.8 
AA082943 Cyclin G1 2.8 
AA703660 H2A histone family 2.8 
H97778 E-cadherin 2.8 
H24326 Regulator of G-protein signaling (RGS7) 2.7 
H54020 Splicing factor (9G8) 2.5 
AA126911 Heterogeneous nuclear ribonucleoprotein A1 2.4 
H20856 DNA repair helicase ERCC3 2.4 
AA488332 Cell cycle protein p38-2G4 2.4 
AA608514 Histone H3.3 2.3 
AA704255 Spinocerebellar ataxia 7 2.2 
H95392 Histone H2A.X 2.2 
R53787 Protein phosphatase 2A B56-epsilon (PP2A) 2.1 
AA778077 Chromodomain helicase DNA binding protein 3 2.0 
AA701502 Platelet-derived growth factor PDGF-A 2.0 
AA088861 LI-cadherin 2.0 
a

Acc. no. indicates Unigene accession number.

b

Value shows the difference between the means of average difference of papillary thyroid carcinoma and normal thyroid tissues.

Table 3

Overexpression of basic fibroblast growth factor (bFGF) and platelet-derived growth factor (PDGF) in thyroid tissues

TissuebFGFan (%)PDGF n (%)
Normal thyroid 0/10 (0) 0/10 (0) 
Hyperfunctioning tissue 3/5 (60) 1/5 (20) 
Multinodular goiter 2/6 (33) 0/6 (0) 
Adenoma 2/5 (40) 0/5 (0) 
Papillary carcinoma   
 Well-differentiated 30/55 (55) 45/55 (82) 
 Follicular variant 2/4 (50) 3/4 (75) 
Follicular carcinoma 4/6 (67) 6/6 (100) 
TissuebFGFan (%)PDGF n (%)
Normal thyroid 0/10 (0) 0/10 (0) 
Hyperfunctioning tissue 3/5 (60) 1/5 (20) 
Multinodular goiter 2/6 (33) 0/6 (0) 
Adenoma 2/5 (40) 0/5 (0) 
Papillary carcinoma   
 Well-differentiated 30/55 (55) 45/55 (82) 
 Follicular variant 2/4 (50) 3/4 (75) 
Follicular carcinoma 4/6 (67) 6/6 (100) 
a

Number of samples with positive staining/total number of samples, including the seven cases studied by DNA array analysis.

1
Loh KC, Greenspan FS, Gee L, Miller TR, Yeo PP Pathological tumor-node-metastasis (pTNM) staging for papillary and follicular thyroid carcinomas: a retrospective analysis of 700 patients.
J Clin Endocrinol Metab
,
82
:
3553
-62,  
1997
.
2
Spencer CA, Wang CC Thyroglobulin measurement. Techniques, clinical benefits, and pitfalls.
Endocrinol Metab Clin North Am
,
24
:
841
-63,  
1995
.
3
Mattano LA, Jr., Moss TJ, Emerson SG Sensitive detection of rare circulating neuroblastoma cells by the reverse transcriptase-polymerase chain reaction.
Cancer Res
,
52
:
4701
-5,  
1992
.
4
Burchill SA, Bradbury MF, Pittman K, Southgate J, Smith B, Selby P Detection of epithelial cancer cells in peripheral blood by reverse transcriptase-polymerase chain reaction.
Br J Cancer
,
71
:
278
-81,  
1995
.
5
Tanaka K, Otsuki T, Sonoo H, et al Semi-quantitative comparison of the differentiation markers and sodium iodide symporter messenger ribonucleic acids in papillary thyroid carcinomas using RT-PCR.
Eur J Endocrinol
,
142
:
340
-6,  
2000
.
6
Tallini G, Ghossein RA, Emanuel J, et al Detection of thyroglobulin, thyroid peroxidase, and RET/PTC1 mRNA transcripts in the peripheral blood of patients with thyroid disease.
J Clin Oncol
,
16
:
1158
-66,  
1998
.
7
Ringel MD, Ladenson PW, Levine MA Molecular diagnosis of residual and recurrent thyroid cancer by amplification of thyroglobulin messenger ribonucleic acid in peripheral blood.
J Clin Endocrinol Metab
,
83
:
4435
-42,  
1998
.
8
Ditkoff BA, Marvin MR, Yemul S, et al Detection of circulating thyroid cells in peripheral blood.
Surgery
,
120
:
959
-64, discussion 964–5 
1996
.
9
Bojunga J, Roddiger S, Stanisch M, et al Molecular detection of thyroglobulin mRNA transcripts in peripheral blood of patients with thyroid disease by RT-PCR.
Br J Cancer
,
82
:
1650
-5,  
2000
.
10
Takano T, Miyauchi A, Yoshida H, Hasegawa Y, Kuma K, Amino N Quantitative measurement of thyroglobulin mRNA in peripheral blood of patients after total thyroidectomy.
Br J Cancer
,
85
:
102
-6,  
2001
.
11
Schena M, Shalon D, Davis RW, Brown PO Quantitative monitoring of gene expression patterns with a complementary DNA microarray.
Science (Wash DC)
,
270
:
467
-70,  
1995
.
12
DeRisi J, Penland L, Brown PO, et al Use of a cDNA microarray to analyse gene expression patterns in human cancer.
Nat Genet
,
14
:
457
-60,  
1996
.
13
Kenneth SM, Larry SM, Edwin BC, John K, Kenneth SM Estrogen receptor analyses: correlation of biochemical and immunohistochemical methods using monoclonal antireceptor antibodies.
Arch Pathol Lab Med
,
109
:
716
-21,  
1985
.
14
Shingu K, Fujimori M, Ito K, et al Expression of fibroblast growth factor-2 and fibroblast growth factor receptor-1 in thyroid diseases: difference between neoplasms and hyperplastic lesions.
Endocr J
,
45
:
35
-43,  
1998
.
15
Kodama M, Daa T, Kashima K, Yokoyama S, Nakayama I, Noguchi S Immunohistochemical localization of acidic and basic fibroblast growth factors in human benign and malignant thyroid lesions.
Jpn J Clin Oncol
,
24
:
66
-73,  
1994
.
16
Daa T, Kodama M, Kashima K, Yokoyama S, Nakayama I, Noguchi S Identification of basic fibroblast growth factor in papillary carcinoma of the thyroid.
Acta Pathol Jpn
,
43
:
582
-9,  
1993
.
17
Fujimoto K, Ichimori Y, Yamaguchi H, et al Basic fibroblast growth factor as a candidate tumor marker for renal cell carcinoma.
Jpn J Cancer Res
,
86
:
182
-6,  
1995
.
18
Takano T, Matsuzuka F, Sumizaki H, Kuma K, Amino N Rapid detection of specific messenger RNAs in thyroid carcinomas by reverse transcription-PCR with degenerate primers: specific expression of oncofetal fibronectin messenger RNA in papillary carcinoma.
Cancer Res
,
57
:
3792
-7,  
1997
.
19
Fernandez PL, Merino MJ, Gomez M, et al Galectin-3 and laminin expression in neoplastic and non-neoplastic thyroid tissue.
J Pathol
,
181
:
80
-6,  
1997
.
20
Fabien N, Fusco A, Santoro M, Barbier Y, Dubois PM, Paulin C Description of a human papillary thyroid carcinoma cell line. Morphologic study and expression of tumoral markers.
Cancer
,
73
:
2206
-12,  
1994
.
21
Mitselou A, Vougiouklakis TG, Peschos D, Dallas P, Boumba VA, Agnantis NJ Immunohistochemical study of the expression of S-100 protein, epithelial membrane antigen, cytokeratin and carcinoembryonic antigen in thyroid lesions.
Anticancer Res
,
22
:
1777
-80,  
2002
.
22
Frittitta L, Sciacca L, Catalfamo R, et al Functional insulin receptors are overexpressed in thyroid tumors: is this an early event in thyroid tumorigenesis?.
Cancer (Phila)
,
85
:
492
-8,  
1999
.
23
Huang Y, Prasad M, Lemon WJ, et al Gene expression in papillary thyroid carcinoma reveals highly consistent profiles.
Proc Natl Acad Sci USA
,
98
:
15044
-9,  
2001
.
24
Wasenius VM, Hemmer S, Kettunen E, Knuutila S, Franssila K, Joensuu H Hepatocyte growth factor receptor, matrix metalloproteinase-11, tissue inhibitor of metalloproteinase-1, and fibronectin are up-regulated in papillary thyroid carcinoma: a cDNA and tissue microarray study.
Clin Cancer Res
,
9
:
68
-75,  
2003
.
25
Hoang-Vu C, Dralle H, Scheumann G, et al Gene expression of differentiation- and dedifferentiation markers in normal and malignant human thyroid tissues.
Exp Clin Endocrinol
,
100
:
51
-6,  
1992
.
26
Takano T, Hasegawa Y, Matsuzuka F, et al Gene expression profiles in thyroid carcinomas.
Br J Cancer
,
83
:
1495
-502,  
2000
.
27
Pauws E, Moreno JC, Tijssen M, Baas F, de Vijlder JJ, Ris-Stalpers C Serial analysis of gene expression as a tool to assess the human thyroid expression profile and to identify novel thyroidal genes.
J Clin Endocrinol Metab
,
85
:
1923
-7,  
2000
.
28
Lazar V, Bidart JM, Caillou B, et al Expression of the Na+/I− symporter gene in human thyroid tumors: a comparison study with other thyroid-specific genes.
J Clin Endocrinol Metab
,
84
:
3228
-34,  
1999
.
29
Ringel MD, Anderson J, Souza SL, et al Expression of the sodium iodide symporter and thyroglobulin genes are reduced in papillary thyroid cancer.
Mod Pathol
,
14
:
289
-96,  
2001
.
30
Tanaka K, Sonoo H, Yamamoto Y, et al Changes of expression level of the differentiation markers in papillary thyroid carcinoma under thyrotropin suppression therapy in vivo immunohistochemical detection of thyroglobulin, thyroid peroxidase, and thyrotropin receptor.
J Surg Oncol
,
75
:
108
-16,  
2000
.
31
Ohta K, Endo T, Onaya T The mRNA levels of thyrotropin receptor, thyroglobulin and thyroid peroxidase in neoplastic human thyroid tissues.
Biochem Biophys Res Commun
,
174
:
1148
-53,  
1991
.
32
Fabbro D, Di Loreto C, Beltrami CA, Belfiore A, Di Lauro R, Damante G Expression of thyroid-specific transcription factors TTF-1 and PAX-8 in human thyroid neoplasms.
Cancer Res
,
54
:
4744
-9,  
1994
.
33
Mai KT, Vaccani JP, Thomas J, Odell PF Immunostaining for ret oncogene proteins in papillary thyroid carcinoma: a correlation of ret immunoreactivity and potential of lymph node metastasis.
Thyroid
,
11
:
859
-63,  
2001
.
34
Kitamura Y, Minobe K, Nakata T, et al Ret/PTC3 is the most frequent form of gene rearrangement in papillary thyroid carcinomas in Japan.
J Hum Genet
,
44
:
96
-102,  
1999
.
35
Bongarzone I, Pierotti MA, Monzini N, et al High frequency of activation of tyrosine kinase oncogenes in human papillary thyroid carcinoma.
Oncogene
,
4
:
1457
-62,  
1989
.
36
Bunone G, Uggeri M, Mondellini P, Pierotti MA, Bongarzone I RET receptor expression in thyroid follicular epithelial cell-derived tumors.
Cancer Res
,
60
:
2845
-9,  
2000
.