This Article Abstract Full Text (PDF) All Versions of this Article: 90/2/1163    most recent Author Manuscript (PDF) Submit a related Letter to the Editor Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Eszlinger, M. Articles by Paschke, R. Search for Related Content PubMed PubMed Citation Articles by Eszlinger, M. Articles by Paschke, R. Related Collections Endocrine Oncology Thyroid

Gene Expression Analysis Reveals Evidence for Increased Expression of Cell Cycle-Associated Genes and G_q-Protein-Protein Kinase C Signaling in Cold Thyroid Nodules

III. Medical Department (M.E., K.K., K.B., D.F., R.P.), and Interdisciplinary Center for Clinical Research Leipzig (K.K.), University of Leipzig, D-04103 Leipzig, Germany; Institute of Biometrics and Medical Informatics (J.L., S.K.), University of Magdeburg, D-39120 Magdeburg, Germany; and Interdisciplinary Center for Bioinformatics Leipzig (M.B.), University of Leipzig, D-04107 Leipzig, Germany

Address all correspondence and requests for reprints to: R. Paschke, M.D., III. Medical Department, University of Leipzig, Philipp-Rosenthal-Strasse 27, D-04103 Leipzig, Germany. E-mail: pasr{at}medizin.uni-leipzig.de.

	Abstract

Top Abstract Introduction Subjects and Methods Results Discussion References

 
In contrast to the molecular etiology of autonomously functioning thyroid nodules, the molecular cause of cold thyroid nodules (CTNs), their benign, functional inactive counterparts, are so far largely unknown. Because of the partially dedifferentiated phenotype of CTNs, alterations in signaling cascades that favor proliferation, but not differentiation, are likely candidates for tumor induction and progression. The importance of RAS mutations for the development of benign nodules with follicular histology is still in question. However, differentially expressed genes in the context of their signaling cascades could define aberrant signaling in CTNs. Therefore, we investigated gene expression in 22 CTNs and their normal surrounding tissue using Affymetrix GeneChips. Most prominently, data analysis revealed an increased expression of cell cycle-associated genes and a special relevance of protein kinase C signaling, whereas no evidence of RAS-MAPK signaling in CTNs was found. Moreover, we determined 31 differentially regulated genes in CTNs, including several histone mRNAs. Taken together, these results explain recent findings showing an increased proliferation in CTNs and draw attention to protein kinase C signaling, but away from RAS-MAPK signaling, as being involved in the etiology of CTNs.

	Introduction

Top Abstract Introduction Subjects and Methods Results Discussion References

 
IN CONTRAST TO autonomously functioning thyroid nodules (AFTN), the molecular etiology of cold thyroid nodules (CTNs) is, so far, poorly understood. They occur in about 5% of the population in iodine-deficient areas and need to be differentiated from carcinomas that occur in about 3% of CTNs (1). Mutations in genes that favor dedifferentiation (e.g. in the RAS/RAF/MEK/ERK/MAP kinase pathway) are likely candidates that could explain the molecular cause of CTNs. Moreover, data demonstrating a predominant monoclonal origin of benign thyroid adenomas support the assumption that somatic mutations could be the origin of CTNs (2, 3, 4, 5). A recent study of our group, investigating 41 well-defined CTNs, revealed a monoclonal origin in 61% of all informative cases but a RAS mutation in only one CTN (6). In addition to RAS mutations, BRAF mutations can also activate this pathway (7, 8). Although these mutations might play a major role in papillary thyroid carcinomas (PTCs) (9), they are absent in our set of CTNs (10). These results indicate that, in the case of CTNs, other targets initiating growth, but not hyperfunction, need to be defined. In this context, Vanvooren et al. (11) investigated the role of Epac and Rap1, two downstream targets of the cAMP cascade, in the development of CTNs. However, no mutations in these genes were found, indicating that the cAMP-Epac-Rap1 pathway is not a major target for mutations involved in the development of CTNs (11). Therefore, before further pursuing a broad sequencing approach, it can be helpful to define a molecular signature of CTNs and benign thyroid tumors in general. Expression profiling and subsequent identification of genes with differential expression between nodular and surrounding thyroid tissue could reduce the number of possible targets or unmask alternative candidates.

Recently, a limited investigation of 588 genes by cDNA expression arrays showed changes in the expression of several signal transducing components that seem to reflect a disturbed signaling system. However, the results of that study did not allow identification of specific signal transduction cascades that might be involved in nodular development (12). To gain a higher resolution, we here compare gene expression for approximately 10,000 full-length genes between CTNs and their corresponding normal surrounding tissue (ST) using the U95A Affymetrix GeneChip. In addition to analyzing single genes, we also studied the differential expression of gene sets. To overcome current limitations of simple visualization and to improve the statistical relevance of gene set analysis, we applied the Westfall-Young procedure to gene sets defined in the GenMAPP software (see Subjects and Methods).

Regulation of gene expression in CTNs was most consistent for a number of histone mRNAs. The increased expression of these histone mRNAs and of gene sets containing cell cycle-associated genes, like cyclin D1, cyclin H/cyclin-dependent kinase (CDK)7, and cyclin B, most likely define the molecular setup for a previously described increased proliferation in CTNs (13). Our data also suggest that altered gene expression of components pertaining to the RAS-MAPK cascade is of minor importance for the development of CTNs, because we find that the respective gene set does not show differential expression. However, the expression pattern of the G_q/11 protein and protein kinase (PK)C indicates a topic for further investigation.

	Subjects and Methods

Top Abstract Introduction Subjects and Methods Results Discussion References

 
All CTNs were identified by ultrasound and scintigraphy. All preoperatively identified nodules were also identified at surgery and postoperatively characterized by histology according to the World Health Organization criteria (14) (Table 1). Only patients without increased serum TSH levels and with solitary CTNs without lymphocytic thyroiditis were included in this study. Informed consent for the analysis was given by the patients.

Total RNA was isolated from 15 CTNs and their corresponding ST using TRIzol reagent (Life Technologies, Gaithersburg, MD) according to the manufacturer’s instructions. Afterward, the total RNA was purified with RNeasy kits (QIAGEN, Hilden, Germany) according to the RNA clean-up protocol.

The quality and quantity of the total RNAs were examined on an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA) using the RNA 6.000 LabChip Kit (Agilent Technologies) according to the manufacturer’s instructions.

Microarray analysis

Ten micrograms of total RNA were used to prepare double-stranded cDNA (Superscript II, Life Technologies) primed with oligo-dT containing a T7 RNA polymerase promoter site (Genset SA, Paris, France). cDNA was purified by phenolchloroform extraction before in vitro transcription using the ENZO BioArray RNA transcript labeling kit (Affymetrix, Santa Clara, CA) to synthesize cRNA. After the in vitro transcription, unincorporated nucleotides were removed using the RNeasy kit (QIAGEN). The cRNA was fragmented and hybridized to Affymetrix GeneChip U95Av2.

The washing, staining, and scanning of the probe array was performed according to the manufacturer’s instructions.

Data analysis

Affymetrix GeneChip data, representing approximately 10,000 full-length genes, were processed using Microarray Suite 5.0 (MAS) from Affymetrix. After extraction of fluorescence intensities, data were scaled to normalize data for interarray comparison.

To detect differentially regulated genes, different methods of data analysis were used:

Empirical filtering. We manually selected genes, which were characterized by the Affymetrix software, by increased or decreased, and a signal log ratio greater than 0.585 or lower than –0.585 (i.e. these genes are at least 1.5-fold higher or lower expressed in the CTN than in the corresponding ST) in 18 of 22 patients, respectively.

Westfall-Young procedure. To avoid a high rate of false-positive results in this multitude of tests for the different genes, we added so-called multiple test procedures. After a logarithmic transformation of the data, we computed adjusted P values for each gene according to the Westfall-Young procedure, which imbeds univariate F tests into a permutation procedure (15). This procedure keeps the family-wise error rate in the strong sense, i.e. the lot of all selected genes may contain a false-positive gene with the probability , at most.

Cluster analysis. Hierarchical cluster analysis using the correlation metric and average linkage was performed by means of the Cluster and TreeView software (http://www.rana.lbl.gov) (16). Before clustering, the complete data set of all 12,625 ProbeSets was trimmed of genes that showed a differential expression for CTNs and STs in less than seven of 22 patients. This trimmed data set comprised 613 ProbeSets. Although loss of some information is possible, trimming the number of genes decreases the influence by genes with little or no differential expression onto the clustering procedure.

Application of the Westfall-Young strategy to gene sets. The procedure by Westfall and Young for multiple testing is applicable not only to single genes but also to sets of genes. It has been shown that combinations of genes more clearly reveal the effects of a disease than single genes (17). Therefore, we analyzed 51 gene sets with three to 150 genes based on NetAFFX annotations (GenMAPP collection) (18). We intended to recognize the relevant gene sets, that is, those sets in which some differences between CTNs and their ST are existent.

The P values of univariate and multivariate one-sample tests (19) were used for the assessment of the gene sets. All genes that did not attain significance in the univariate test of the level 0.05 were excluded from the multivariate consideration of a set. In the multivariate test, the number of principal components was applied that yielded the smallest P value. Thus, a characterization of a set has been obtained.

GenMAPP

GenMAPP (Gene Microarray Pathway Profiler, downloaded from www.genmapp.org) complements and extends available statistical and clustering algorithms (20). In contrast to statistical filters and pattern-finding algorithms (e.g. hierarchical clustering), GenMAPP analyzes the gene expression changes in the context of known biological pathways. Gene expression data, including P values, were imported into GenMAPP in a comma-separated-value (CSV-file) format. GenMAPP converts the expression data into a data set that can then be viewed on any MAPP with any number of color-coding criteria sets.

Real-time RT-PCR

The quantification of several differentially expressed genes by real-time RT-PCR was performed using a LightCycler (Roche, Mannheim, Germany) as previously described (12). The nucleotide sequences of the primers and PCR conditions are available on request. The differential expression of the investigated genes was calculated as the log(2)-ratio AFTN/ST. The determined ratios were also normalized to the ratio of the housekeeping gene ß-actin.

Western blot

Protein extracts were prepared according to Klose (21). Protein content was determined in all samples, using the commercially available Bradford reagent (Bio-Rad, Hercules, CA).

One hundred micrograms of protein were separated by 14% SDS-gel electrophoresis. Proteins were transferred onto a nitrocellulose membrane (Schleicher & Schuell, Dassel, Germany) and then blocked by incubation in blocking buffer composed of 3% nonfat-dry milk in Tris-buffered saline with Tween 20 (TBST) (150 mM NaCl; 25 mM Tris, pH 8.0; 0.2% Tween 20) for 45 min at room temperature. Blots were incubated with the anticellular retinoic acid binding protein (anti-CRABP)1 antibody (Abcam, Cambridgeshire, UK) at a dilution of 1:1000 in 2% BSA in TBST for 90 min at room temperature. Membranes were washed with TBST and treated with horseradish peroxidase-linked antirabbit IgG secondary antibody (1:3000) (New England Biolabs, Inc., Beverly, MA) in TBST containing 5% BSA for 45 min at room temperature. Proteins were detected using SuperSignal West Pico Chemiluminescent Substrate (Pierce, Rockford, IL). Blots were stripped and reprobed with anti-ß-actin antibody (1:500) (Sigma Chemical Co., St. Louis, MO) as described above.

	Results

Top Abstract Introduction Subjects and Methods Results Discussion References

 
GeneChips

On the basis of our recent examination of the gene expression pattern of AFTN (22), we planned the realization of this study. Statistical examination of data from five AFTN and their STs revealed the necessity to analyze at least 13 AFTNs and STs to detect 2-fold differences between AFTNs and STs with statistical reliability ( = 0.001, ß = 0.05, maximal logarithmic value of the SD = 1.60) (22). Because CTNs are most likely characterized by a higher variability, we generated gene expression profiles for 22 CTNs and their corresponding STs using U95Av2 GeneChips. Our recent study with AFTN showed a high correlation of the expression data for three identical samples (often termed technical replicates) (22). Therefore, no replicate hybridizations of the CTNs and their STs were performed in this study.

Empirical filtering

Empirical filtering of the data-sets of the 22 CTNs and their STs revealed three up- and 22 down-regulated genes (Tables 2 and 3). The family member O H2A histone is the gene with the strongest increase of expression in CTNs in comparison with the STs (signal log ratio of 1.5). Moreover, family member L H2A histone and family member A H2B histone are characterized by an increased expression in at least 18 of 22 CTNs. Apolipoprotein D and member 21 of the small inducible cytokine subfamily A show strongest regulation in the group of down-regulated genes (signal log ratios, –3.9 and –3.6, respectively, Table 2). Down-regulated genes are characterized by a higher consistency in their expression (22 genes met the filter criterion). In addition, they also show a stronger differential expression.

At a significance level of P < 0.01, the Westfall-Young procedure detected 50 differentially expressed genes. Here we find an overlap of 21 differentially expressed genes between empirical filtering and statistical analysis; whereas at a significance level of P < 0.001, the Westfall-Young procedure detected 22 differentially expressed genes with an overlap of 16 genes to the empirical filtering (Tables 2 and 3). Statistical analysis of the expression data revealed three additional, up-regulated genes (P < 0.001, CTN vs. ST): carbonic anhydrase XII, H2A histone family member O, and RNA binding motif protein 3 (Table 2). In the group of down-regulated genes, we detected also three additional genes with a P-value < 0.001: macrophage stimulating 1, reelin, and fibulin 5 (Table 3).

Hierarchical clustering

Hierarchical clustering was used to interpret the patterns of gene expression between CTNs and their normal STs. The software organizes gene expression data so that the genes (rows) and tissue samples (columns) are arranged according to the degree of similarity in their pattern of gene expression.

For the expression level of 613 genes (trimmed data set, CTNs vs. STs), clustering divided all tissue samples into two major clusters (data not shown): 21 CTNs and 22 STs grouped within cluster 1 and cluster 2, respectively. Only CTN 74 and its ST (ST 74) populate a separate subtree beyond cluster 2. The group of up-regulated genes shows a five-gene overlap with our empirical and statistical algorithms (four histones and carbonic anhydrase XII). Other additional histone family members, cyclin D1 and D2, adenylate cyclase 9, RGS 20, fibroblast growth factor-7, and others show higher mRNA expression in the CTNs than in their STs. In the cluster of down-regulated genes, we found 18 genes that overlap with empirical filtering and statistical analysis. Additionally, this gene cluster contains a further two members of the complement system, IGF binding protein (IGFBP) 5, CDK inhibitor 1C, and fibroblast growth factor receptor 1, among others.

Westfall-Young procedure applied to gene sets

The statistical analysis of 51 predefined gene sets (GenMAPP collection) revealed five significant groups. Named according to the respective genes, these groups are: cell cycle, G protein-coupled receptors class A rhodopsin-like, complement activation classical, G protein-signaling, and Wnt-signaling. In contrast, the MAPK cascade showed no significant differences between CTNs and their STs (Table 4).

GenMAPP software was used for visualization of significant gene sets detected by the Westfall-Young statistic.

The cell cycle is mapped with an increased expression of cyclin B, cyclin D, cyclin H, and CDK1 and 7. Furthermore, CDK inhibitor 1 (p21); retinoblastoma-associated protein (Rb)-1; DNA-dependent PK; mitotic spindle assembly checkpoint protein (MAD2L1); cell division cycle protein (Cdc)-20; Wee1-like PK (WEE1); transcription factors DP-1, E2F4, and E2F5; and histone deacetylase-1 are characterized by an increased expression in CTNs, whereas growth arrest and DNA-damage-inducible protein GADD45, transcription factor E2F1, histone deacetylase-4, and DNA replication licensing factor MCM5 are lower expressed in CTNs than in their STs (data not shown).

In the gene set of class A (rhodopsin-like)–G protein-coupled receptors, 13 genes show a differential expression between CTNs and their STs; the -1B- and the ß-3-adrenergic receptors, the thrombin receptor, CXC chemokine receptor 4, and chemokine (C-C motif) receptor 1 are higher expressed in CTNs than in the STs, whereas the endothelin type b receptor-like protein 2, the endothelin A receptor, and the dopamine receptor D5 show a decreased expression in CTNs.

The group of complement activation-associated genes is characterized by a decreased expression of complement component 1r and 1s, complement component 6 and 7, and an increased expression of complement component 1q and 2 in CTNs in comparison with the STs.

The fourth significant group comprising genes involved in G protein signaling demonstrates slight, but significant, increases in the expression of PKA subunit C-ß; PKC ß1; PKC ; phosphodiesterases 1B, 1C, and 4D; G_i; and G_q in CTNs. In contrast, G₁₁, R-RAS, and the PKC µ and show a weaker expression in CTNs than in their STs.

The significance of the Wnt-signaling group can be mainly attributed to the differential expression of cyclin D1 and D2, PKC , PKC µ, and transcripts, which also contribute to the differential expression of the sets cell cycle and G protein signaling.

Real-time RT-PCR

The GeneChip expression data of eight genes (RasGAP 1, apolipoprotein D, p85 PI-3-K, p90 RSK2, IGFBP 6, DIO 1, TPO, and metallothionein 1G) were verified by real-time RT-PCR. These experiments show a high concordance between the GeneChip data and the real-time RT-PCR (Fig. 1), even in cases with a low differential expression.

View larger version (11K): [in this window] [in a new window]   FIG. 1. Comparison of the differential gene expression of eight genes determined by GeneChip analysis and real-time RT-PCR. The graph shows a high correlation between the GeneChip data (white bars) and the results of the real-time RT-PCR (black bars).

The expression of the CRABP1 in CTNs and their ST determined by GeneChip hybridization was also investigated by Western blot analysis. These experiments show a strongly decreased expression of the CRABP1 protein in CTNs in comparison with their STs (Fig. 2).

View larger version (19K): [in this window] [in a new window]   FIG. 2. CRABP1 content of tissue extracts of three CTNs in comparison with their STs. The quantity of protein engaged in the Western blot was normalized to the housekeeping gene ß-actin.

Differential expression of cell cycle-associated genes (GenMAPP set, e.g. different cyclins, CDKs and p53) was compared with the labeling index (labeled vs. nonlabeled cells) for PCNA and KI-67 recently determined on slices of 14 CTNs used in this study (13). Interestingly, we found a high correlation between the labeling index of PCNA and the signal log ratios of cyclin H (r = –0.624, P = 0.011), Cip1 (r = –0.57, P = 0.021), and p53 (r = 0.56, P = 0.023).

	Discussion

Top Abstract Introduction Subjects and Methods Results Discussion References

 
Data analysis

Our analysis of differential gene expression reveals 31 genes that are characterized by a differential mRNA expression between CTNs and their normal STs. Comparing results of the three analysis algorithms (i.e. empirical filtering, statistical analysis according to Westfall-Young, and hierarchical clustering), we confirm our recent findings (22) showing a high overlap between these three algorithms. We found an overlap of 42% (21 of 50 genes) between the empirical filtering and the statistical analysis with P < 0.01, and 73% (16 of 22 genes) between the empirical filtering and the statistical analysis with P < 0.001, respectively. These data show a high concordance of our empirical filter algorithm and the statistical analysis according to Westfall-Young. Most prominently, we found a consistently increased expression of several histone mRNAs in CTNs in all three analysis algorithms. These transcripts very likely present new marker molecules for benign hypofunctional thyroid tumors. Moreover, using our advanced application of the Westfall-Young procedure to gene sets, we found significant differences between CTNs and their STs in five different gene sets. Among these, the two gene sets cell cycle and G protein signaling seem especially important in the context of CTNs.

Comparison of the expression pattern of the differentially expressed genes with molecular, histologic, and clinical data

We compared the expression pattern of the differentially expressed genes between men and women, adenomas and adenomatous nodules. However, we could not find any correlation between the gene expression pattern and the mentioned properties of the CTNs with the Westfall-Young procedure.

In addition, we compared the chromosomal localization of the differentially expressed genes with previously reported loci for multinodular goiter (MNG) (23, 24, 25). Interestingly, we found an overlap at the MNG-1 locus at 14q31 with the down-regulated genes fibronectin leucine-rich transmembrane protein (FLRT) 2 and fibulin (FBLN) 5 and at the recently identified susceptibility locus at chromosome 2q with IGFBP 5. Though, so far, it is unclear whether the MNG-1 locus predisposes to euthyroid goiter, because a recent study failed to confirm this locus in additional families (24). Moreover, the functions of FLRT 2 and FBLN 5 in the thyroid are unknown. The precise functions and patterns of IGFBP 5 expression have not been determined so far. IGFBP 5 inhibits growth via G2/M cell cycle arrest and induces apoptosis in human breast cancer cell lines (26) but has also been implicated in growth stimulation of osteoblast proliferation (27) and of prostate cancer cells (28). In the thyroid, a decreased IGFBP 5 mRNA expression has been shown in goiter, whereas IGFBP 5 is increased in papillary thyroid cancer in comparison with normal thyroid tissue (29). Despite the fact that functions of IGFBP 5 are multiple, these interesting findings suggest a further investigation of IGFBP 5 in the context of thyroid pathophysiology.

Cell cycle and complement-associated genes

It has recently been suggested that combinations of genes that individually do not show powerful differential regulation could reveal additional aspects of a disease (17). Subtle, but coordinated, changes in expression might be detected more readily by combining measurements across multiple members of each gene set (17). However, differential expression of gene sets is often established only from graphical visualization. However, before discussing the biological relevance of gene sets, statistical significance needs to be proven. We therefore applied the permutation procedure according to Westfall-Young not only to the analysis of single genes but also to gene sets. Such an approach implies that alterations in gene expression might more clearly manifest at the level of biological pathways or coregulated gene sets, rather than individual genes (17).

We found a significantly increased expression pattern of several cell cycle-associated genes in CTNs (Fig. 3). More specifically, cyclin D1 supports a progression into S phase and is characterized by a significantly increased expression in CTNs. Although the S phase cyclins E and A do not show a changed mRNA expression in CTNs, other S phase markers also suggest an increased cell cycle progression: all applied univariate algorithms revealed an increased expression of several histone mRNAs, which are modulated by changing their half-life in response to the rate of DNA replication during S phase (30). Furthermore, the increased expression of cyclin H and CDK7 illustrates the increased cell cycle activity in CTNs: cyclin H regulates CDK7, the catalytic subunit of the CDK-activating kinase complex, which subsequently activates the cyclin-associated kinases CDC2/CDK1, CDK2, CDK4, and CDK6 by threonine phosphorylation. However, not only stimulating, but also cell cycle-inhibiting, genes are characterized by an increased expression in CTNs [e.g. the CDK inhibitor 1 (Cip1, p21)], which very likely reflects a feedback from increased proliferation. Moreover, our gene expression data fit nicely with recent findings of our group, demonstrating an increased proliferation in CTNs: Krohn et al. (13) could show a 10- to 20-fold increase of the labeling index of proliferating cell nuclear antigen (PCNA) and Ki-67 epitopes in CTNs, compared with a 2- to 3-fold increase in AFTNs (31). Interestingly, we found a high correlation between the labeling index of PCNA and the signal log ratios of cyclin H (r = –0.624, P = 0.011), Cip1 (r = –0.57, P = 0.021), and p53 (r = 0.56, P = 0.023). Negative correlation of the signal log ratio for Cip1 and PCNA labeling clearly reflects the typical inhibitory function of Cip1 on cyclin/CDK complexes (32). However, the discrepant correlation for Cip1 (negative) and p53 (positive) indicates that either the well-established regulation of Cip1 by p53 (33) is defective or that activation of Cip1 in CTNs is, at least in part, p53 independent (34). A further explanation for this discrepancy might be a selective inhibition of bound p53 from activating Cip1 transcription, which has been shown for myc, that functions as a switch between induction of cell cycle arrest and apoptotic cell death (35).

View larger version (20K): [in this window] [in a new window]   FIG. 3. Diagram of the cell cycle, depicting the genes with the most significant differential expression between CTNs and their normal STs according to the Westfall-Young procedure applied to gene sets. Lines, Increased gene expression in CTNs; dotted line, decreased gene expression in CTNs; arrows, genes involved in cell cycle progression; circles, genes involved in cell cycle retardation. GADD, Growth arrest and DNA-damage-inducible; MAD, mitotic arrest deficient; Cdc, cell division control; Cip, CDK interacting protein.

G_q-PKC cascade

The significantly different G protein signaling gene set is particularly interesting: four of 17 significant genes of this group are PKC isoforms. Whereas we found an up-regulated PKC ß and expression in CTNs, PKC µ and are down-regulated. Furthermore, the gene set analysis revealed an increased expression of the G_q protein. Although the cause of this increased G_q and PKC expression is unknown up to now, these findings are of special interest because it has been shown that a long-term PKC stimulation (days) of ovine, porcine, and dog thyroid cells in culture with 12-O-tetradecanoyl-phorbol-13-acetate causes a general loss of thyroid-specific functions [e.g. loss of iodide transport and thyroglobulin iodination (37, 38, 39, 40, 41)]. Furthermore, Kraiem et al. (42) could show that the PKC inhibits TSH-mediated human thyroid cell differentiation, whereas TSH attenuates PKC-mediated mitogenesis. All patients with CTNs that we investigated were euthyroid. Therefore, an increased TSH signaling as a cause of the increased expression of G_q and PKC mRNA can be excluded. Because no activating mutations in either G_q or G₁₁-subunits were found in benign and malignant thyroid tumors and are unlikely to be responsible for elevated PLC and PKC activities (43), PKC activation may be due to stimulation by other growth factor receptors (44).

In contrast to AFTN, in which mutations of the TSH receptor or the G_s protein are the main cause of increased proliferation and hyperfunction, mutations in genes that favor dedifferentiation are most likely to be the origin of CTNs. However, a general relevance of RAS mutations for the development of CTNs is unlikely: Esapa et al. (45) reviewed several studies and found the frequency of mutations in the RAS oncogene of follicular adenomas (not scintigraphically characterized) in the range of 18–52%. In our set of 41 CTNs, only a single RAS mutation was detectable (6). In this context, our findings showing no significant differences in MAPK signaling components in CTNs in comparison with their STs (P = 0.87) are of particular interest. These results are in line with a recent in vitro study that indicates that the dedifferentiated phenotype of CTNs is unlikely to be the result of an activated RAS signaling (46).

CRABP1

Experimental data of several studies provide evidence that differentiated functions of thyrocytes and of iodide metabolism can be reinduced by retinoic acid (47, 48, 49). Because CRABP1 encodes a high-affinity CRABP, which regulates the availability of retinoic acid to its nuclear receptors and is also involved in retinoic acid catabolism (50), the 4-fold decreased mRNA expression and the decreased protein expression of CRABP1 in CTNs might impact on the partly dedifferentiated and hypofunctioning phenotype of CTNs. This assumption is supported by our recent finding showing an approximate 2-fold increased expression of CRABP1 in AFTN in comparison with their normal ST (unpublished observation) using GeneChips. Interestingly, Huang et al. (51) revealed a decreased expression of CRABP1 in PTC in comparison with their normal counterparts, and recently they found that hypermethylation of CRABP1-promoter CpG islands plays a role in silencing of the human CRABP1 gene expression in tumor cells and possibly in PTC tumors (52). Because retinoic acid (the CRABP1-ligand) treatment of thyroid carcinoma cell lines affects thyroid-specific functions (e.g. type I 5'-deiodinase and sodium/iodide-symporter), cell-cell and cell-matrix interaction (e.g. E-cadherin), differentiation markers (e.g. alkaline phosphatase, CD 97), growth, and tumorigenicity (for review, see Ref. 53), Huang et al. (52) conclude that the hypermethylated and inactivated CRABP1 is likely involved in the pathogenesis of sporadic PTC. However, CRABP1 transcripts are also decreased in follicular thyroid carcinogenesis (54); and instead of supporting a link to malignant thyroid tumors, our data also show a similar reduction of CRABP1 in benign thyroid lesions. Because our results make reduced expression of CRABP1 a common feature of CTN, PTC, and follicular thyroid carcinoma, we conclude a more general role of CRABP1 in thyroid epithelial cell differentiation.

In conclusion, our data support findings showing a stronger increase in proliferation in CTNs than in AFTNs. Furthermore, our gene expression data suggest a major relevance of G_q–PKC-signaling in CTNs; whereas, by means of investigating the differential mRNA expression between CTNs and their corresponding STs, no evidence of aberrant RAS-MAPK signaling was found in our set of RAS mutation negative CTNs on the level of gene expression. This does not rule out that this cascade is regulated by posttranslational modifications in CTNs. Further experiments have to evaluate the initiating event for the increased PKC signaling in CTNs and the relevance of increased G_q- and PKC mRNA expression for the development of CTNs. Moreover, the influence of CRABP1 down-regulation onto thyroid differentiation has to be evaluated in further experiments, too.

	Acknowledgments

 
We are grateful to Gudrun Bethge, Eileen Bösenberg, and Petra Süptitz for their excellent technical assistance.

	Footnotes

 
This work was supported by a grant from the Deutsche Forschungsgemeinschaft (DFG/Pa423/3–2), the Krebshilfe, and the BMBF Interdisciplinary Center for Clinical Research at the University of Leipzig (01KS9504, Projects B20 and B14, Z16-CHIP2, Z16-CHIP10). Microarray analysis was done at the IZKF Leipzig core facility (01KS9504, Project Z03). K.K. is supported by IZKF Leipzig, BMB-F, Interdisciplinary Center for Clinical Research at the University of Leipzig 9504 (01KS9504, Project Z3).

First Published Online November 2, 2004

Abbreviations: AFTN, Autonomously functioning thyroid nodules; CDK, cyclin-dependent kinase; CRABP, cellular retinoic acid binding protein; CTN, cold thyroid nodule; IGFBP, IGF binding protein; MNG, multinodular goiter; PK, protein kinase; PTC, papillary thyroid carcinoma; ST, surrounding tissue; TBST, Tris-buffered saline with Tween 20.

Received June 29, 2004.

Accepted October 25, 2004.

	References

Top Abstract Introduction Subjects and Methods Results Discussion References

 

1. Belfiore A, La Rosa GL, Padova G, Sava L, Ippolito O, Vigneri R 1987 The frequency of cold thyroid nodules and thyroid malignancies in patients from an iodine-deficient area. Cancer 60:3096–3102[CrossRef][Medline]
2. Apel RL, Ezzat S, Bapat BV, Pan N, LiVolsi VA, Asa SL 1995 Clonality of thyroid nodules in sporadic goiter. Diagn Mol Pathol 4:113–121[Medline]
3. Fey MF, Peter HJ, Hinds HL, Zimmermann A, Liechti-Gallati S, Gerber H, Studer H, Tobler A 1992 Clonal analysis of human tumors with M27 ß, a highly informative polymorphic X chromosomal probe. J Clin Invest 89:1438–1444
4. Namba H, Matsuo K, Fagin JA 1990 Clonal composition of benign and malignant human thyroid tumors. J Clin Invest 86:120–125
5. Hicks DG, LiVolsi VA, Neidich JA, Puck JM, Kant JA 1990 Clonal analysis of solitary follicular nodules in the thyroid. Am J Pathol 137:553–562[Abstract]
6. Krohn K, Reske A, Ackermann F, Muller A, Paschke R 2001 Ras mutations are rare in solitary cold and toxic thyroid nodules. Clin Endocrinol (Oxf) 55:241–248[CrossRef][Medline]
7. Davies H, Bignell GR, Cox C, Stephens P, Edkins S, Clegg S, Teague J, Woffendin H, Garnett MJ, Bottomley W, Davis N, Dicks E, Ewing R, Floyd Y, Gray K, Hall S, Hawes R, Hughes J, Kosmidou V, Menzies A, Mould C, Parker A, Stevens C, Watt S, Hooper S, Wilson R, Jayatilake H, Gusterson BA, Cooper C, Shipley J, Hargrave D, Pritchard-Jones K, Maitland N, Chenevix-Trench G, Riggins GJ, Bigner DD, Palmieri G, Cossu A, Flanagan A, Nicholson A, Ho JW, Leung SY, Yuen ST, Weber BL, Seigler HF, Darrow TL, Paterson H, Marais R, Marshall CJ, Wooster R, Stratton MR, Futreal PA 2002 Mutations of the BRAF gene in human cancer. Nature 417:949–954[CrossRef][Medline]
8. Yuen ST, Davies H, Chan TL, Ho JW, Bignell GR, Cox C, Stephens P, Edkins S, Tsui WW, Chan AS, Futreal PA, Stratton MR, Wooster R, Leung SY 2002 Similarity of the phenotypic patterns associated with BRAF and KRAS mutations in colorectal neoplasia. Cancer Res 62:6451–6455[Abstract/Free Full Text]
9. Kimura ET, Nikiforova MN, Zhu Z, Knauf JA, Nikiforov YE, Fagin JA 2003 High prevalence of BRAF mutations in thyroid cancer: genetic evidence for constitutive activation of the RET/PTC-RAS-BRAF signaling pathway in papillary thyroid carcinoma. Cancer Res 63:1454–1457[Abstract/Free Full Text]
10. Krohn K, Paschke R 2004 BRAF mutations are not an alternative explanation for the molecular etiology of ras mutation negative cold thyroid nodules. Thyroid 14:359–361[CrossRef][Medline]
11. Vanvooren V, Allgeier A, Nguyen M, Massart C, Parma J, Dumont JE, Van Sande J 2001 Mutation analysis of the Epac–Rap1 signaling pathway in cold thyroid follicular adenomas. Eur J Endocrinol 144:605–610[Abstract]
12. Eszlinger M, Krohn K, Paschke R 2001 Complementary DNA expression array analysis suggests a lower expression of signal transduction proteins and receptors in cold and hot thyroid nodules. J Clin Endocrinol Metab 86:4834–4842[Abstract/Free Full Text]
13. Krohn K, Stricker I, Emmrich P, Paschke R 2003 Cold thyroid nodules show a marked increase in proliferation markers. Thyroid 13:569–575[CrossRef][Medline]
14. Hedinger C 1988 Histological typing of thyroid tumors. Berlin, Heidelberg, New York: Springer Verlag
15. Westfall PH, Young SS 1993 Resampling-based multiple testing: examples and methods for multiple P-value adjustment. New York: John Wiley, Sons
16. Eisen MB, Spellman PT, Brown PO, Botstein D 1998 Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95:14863–14868[Abstract/Free Full Text]
17. Mootha VK, Lindgren CM, Eriksson KF, Subramanian A, Sihag S, Lehar J, Puigserver P, Carlsson E, Ridderstrale M, Laurila E, Houstis N, Daly MJ, Patterson N, Mesirov JP, Golub TR, Tamayo P, Spiegelman B, Lander ES, Hirschhorn JN, Altshuler D, Groop LC 2003 PGC-1-responsive genes involved in oxidative phosphorylation are coordinately down-regulated in human diabetes. Nat Genet 34:267–273[CrossRef][Medline]
18. Liu G, Loraine AE, Shigeta R, Cline M, Cheng J, Valmeekam V, Sun S, Kulp D, Siani-Rose MA 2003 NetAffx: affymetrix probesets and annotations. Nucleic Acids Res 31:82–86[Abstract/Free Full Text]
19. Läuter J, Glimm E, Kropf S 1998 Multivariate tests based on left-spherically distributed linear scores. Ann Stat 26:1972–1988[CrossRef]
20. Dahlquist KD, Salomonis N, Vranizan K, Lawlor SC, Conklin BR 2002 GenMAPP, a new tool for viewing and analyzing microarray data on biological pathways. Nat Genet 31:19–20[CrossRef][Medline]
21. Klose J 1999 Fractionated extraction of total tissue proteins from mouse and human for 2-D electrophoresis. Methods Mol Biol 112:67–85[Medline]
22. Eszlinger M, Krohn K, Frenzel R, Kropf S, Tonjes A, Paschke R 2004 Gene expression analysis reveals evidence for inactivation of the TGF-ß signaling cascade in autonomously functioning thyroid nodules. Oncogene 23:795–804[CrossRef][Medline]
23. Bignell GR, Canzian F, Shayeghi M, Stark M, Shugart YY, Biggs P, Mangion J, Hamoudi R, Rosenblatt J, Buu P, Sun S, Stoffer SS, Goldgar DE, Romeo G, Houlston RS, Narod SA, Stratton MR, Foulkes WD 1997 Familial nontoxic multinodular thyroid goiter locus maps to chromosome 14q but does not account for familial nonmedullary thyroid cancer. Am J Hum Genet 61:1123–1130[CrossRef][Medline]
24. Bayer Y, Neumann S, Meyer B, Ruschendorf F, Reske A, Brix T, Hegedus L, Langer P, Nurnberg P, Paschke R 2004 Genome-wide linkage analysis reveals evidence for four new susceptibility loci for familial euthyroid goiter. J Clin Endocrinol Metab 89:4044–4052[Abstract/Free Full Text]
25. Neumann S, Willgerodt H, Ackermann F, Reske A, Jung M, Reis A, Paschke R 1999 Linkage of familial euthyroid goiter to the multinodular goiter-1 locus and exclusion of the candidate genes thyroglobulin, thyroperoxidase, and Na+/I– symporter. J Clin Endocrinol Metab 84:3750–3756[Abstract/Free Full Text]
26. Butt AJ, Dickson KA, McDougall F, Baxter RC 2003 Insulin-like growth factor-binding protein-5 inhibits the growth of human breast cancer cells in vitro and in vivo. J Biol Chem 278:29676–29685[Abstract/Free Full Text]
27. Richman C, Baylink DJ, Lang K, Dony C, Mohan S 1999 Recombinant human insulin-like growth factor-binding protein-5 stimulates bone formation parameters in vitro and in vivo. Endocrinology 140:4699–4705[Abstract/Free Full Text]
28. Miyake H, Nelson C, Rennie PS, Gleave ME 2000 Overexpression of insulin-like growth factor binding protein-5 helps accelerate progression to androgen-independence in the human prostate LNCaP tumor model through activation of phosphatidylinositol 3'-kinase pathway. Endocrinology 141:2257–2265[Abstract/Free Full Text]
29. Stolf BS, Carvalho AF, Martins WK, Runza FB, Brun M, Hirata Jr R, Jordao NE, Soares FA, Postigo-Dias J, Kowalski LP, Reis LF 2003 Differential expression of IGFBP-5 and two human ESTs in thyroid glands with goiter, adenoma and papillary or follicular carcinomas. Cancer Lett 191:193–202[CrossRef][Medline]
30. Marzluff WF, Duronio RJ 2002 Histone mRNA expression: multiple levels of cell cycle regulation and important developmental consequences. Curr Opin Cell Biol 14:692–699[CrossRef][Medline]
31. Krohn K, Emmrich P, Ott N, Paschke R 1999 Increased thyroid epithelial cell proliferation in toxic thyroid nodules. Thyroid 9:241–246[Medline]
32. Roninson IB 2002 Oncogenic functions of tumour suppressor p21(Waf1/Cip1/Sdi1): association with cell senescence and tumour-promoting activities of stromal fibroblasts. Cancer Lett 179:1–14[CrossRef][Medline]
33. el Deiry WS, Tokino T, Velculescu VE, Levy DB, Parsons R, Trent JM, Lin D, Mercer WE, Kinzler KW, Vogelstein B 1993 WAF1, a potential mediator of p53 tumor suppression. Cell 75:817–825[CrossRef][Medline]
34. Liu S, Bishop WR, Liu M 2003 Differential effects of cell cycle regulatory protein p21(WAF1/Cip1) on apoptosis and sensitivity to cancer chemotherapy. Drug Resist Updat 6:183–195[CrossRef][Medline]
35. Seoane J, Le HV, Massague J 2002 Myc suppression of the p21(Cip1) Cdk inhibitor influences the outcome of the p53 response to DNA damage. Nature 419:729–734[CrossRef][Medline]
36. Blanchin S, Estienne V, Durand-Gorde JM, Carayon P, Ruf J 2003 Complement activation by direct C4 binding to thyroperoxidase in Hashimoto’s thyroiditis. Endocrinology 144:5422–5429[Abstract/Free Full Text]
37. Vassart G, Dumont JE 1992 The thyrotropin receptor and the regulation of thyrocyte function and growth. Endocr Rev 13:596–611[CrossRef][Medline]
38. Haye B, Aublin JL, Champion S, Lambert B, Jacquemin C 1985 Tetradecanoyl phorbol-13-acetate counteracts the responsiveness of cultured thyroid cells to thyrotropin. Biochem Pharmacol 34:3795–3802[CrossRef][Medline]
39. Takasu N, Shimizu Y, Yamada T 1987 Tumour promoter 12-O-tetradecanoylphorbol 13-acetate and epidermal growth factor stimulate proliferation and inhibit differentiation of porcine thyroid cells in primary culture. J Endocrinol 113:485–487[Abstract]
40. Roger PP, Reuse S, Servais P, Van Heuverswyn B, Dumont JE 1986 Stimulation of cell proliferation and inhibition of differentiation expression by tumor-promoting phorbol esters in dog thyroid cells in primary culture. Cancer Res 46:898–906[Abstract/Free Full Text]
41. Bachrach LK, Eggo MC, Mak WW, Burrow GN 1985 Phorbol esters stimulate growth and inhibit differentiation in cultured thyroid cells. Endocrinology 116:1603–1609[Abstract]
42. Kraiem Z, Sadeh O, Yosef M, Aharon A 1995 Mutual antagonistic interactions between the thyrotropin (adenosine 3',5'-monophosphate) and protein kinase C/epidermal growth factor (tyrosine kinase) pathways in cell proliferation and differentiation of cultured human thyroid follicles. Endocrinology 136:585–590[Abstract]
43. Ringel MD, Saji M, Schwindinger WF, Segev D, Zeiger MA, Levine MA 1998 Absence of activating mutations of the genes encoding the -subunits of G₁1 and Gq in thyroid neoplasia. J Clin Endocrinol Metab 83:554–559[Abstract/Free Full Text]
44. Berridge MJ 1993 Inositol trisphosphate and calcium signalling. Nature 361:315–325[CrossRef][Medline]
45. Esapa CT, Johnson SJ, Kendall-Taylor P, Lennard TW, Harris PE 1999 Prevalence of Ras mutations in thyroid neoplasia. Clin Endocrinol (Oxf) 50:529–535[CrossRef][Medline]
46. Gire V, Wynford-Thomas D 2000 RAS oncogene activation induces proliferation in normal human thyroid epithelial cells without loss of differentiation. Oncogene 19:737–744[CrossRef][Medline]
47. Van Herle AJ, Agatep ML, Padua III DN, Totanes TL, Canlapan DV, Van Herle HM, Juillard GJ 1990 Effects of 13 cis-retinoic acid on growth and differentiation of human follicular carcinoma cells (UCLA R0 82 W-1) in vitro. J Clin Endocrinol Metab 71:755–763[Abstract]
48. Schmutzler C, Brtko J, Bienert K, Kohrle J 1996 Effects of retinoids and role of retinoic acid receptors in human thyroid carcinomas and cell lines derived therefrom. Exp Clin Endocrinol Diabetes 104(Suppl 4):16–19
49. Schreck R, Schnieders F, Schmutzler C, Kohrle J 1994 Retinoids stimulate type I iodothyronine 5'-deiodinase activity in human follicular thyroid carcinoma cell lines. J Clin Endocrinol Metab 79:791–798[Abstract]
50. Donovan M, Olofsson B, Gustafson AL, Dencker L, Eriksson U 1995 The cellular retinoic acid-binding proteins. J Steroid Biochem Mol Biol 53:459–465[CrossRef][Medline]
51. Huang Y, Prasad M, Lemon WJ, Hampel H, Wright FA, Kornacker K, LiVolsi V, Frankel W, Kloos RT, Eng C, Pellegata NS, de la Chapelle A 2001 Gene expression in papillary thyroid carcinoma reveals highly consistent profiles. Proc Natl Acad Sci USA 98:15044–15049[Abstract/Free Full Text]
52. Huang Y, de la Chapelle A, Pellegata NS 2003 Hypermethylation, but not LOH, is associated with the low expression of MT1G and CRABP1 in papillary thyroid carcinoma. Int J Cancer 104:735–744[CrossRef][Medline]
53. Schmutzler C, Kohrle J 2000 Retinoic acid redifferentiation therapy for thyroid cancer. Thyroid 10:393–406[Medline]
54. Aldred MA, Ginn-Pease ME, Morrison CD, Popkie AP, Gimm O, Hoang-Vu C, Krause U, Dralle H, Jhiang SM, Plass C, Eng C 2003 Caveolin-1 and caveolin-2, together with three bone morphogenetic protein-related genes, may encode novel tumor suppressors down-regulated in sporadic follicular thyroid carcinogenesis. Cancer Res 63:2864–2871[Abstract/Free Full Text]

This article has been cited by other articles:

				K. Krause, M. Eszlinger, O. Gimm, S. Karger, C. Engelhardt, H. Dralle, and D. Fuhrer TFF3-Based Candidate Gene Discrimination of Benign and Malignant Thyroid Tumors in a Region with Borderline Iodine Deficiency J. Clin. Endocrinol. Metab., April 1, 2008; 93(4): 1390 - 1393. [Abstract] [Full Text] [PDF]

				M. Eszlinger, K. Krohn, A. Kukulska, B. Jarzab, and R. Paschke Perspectives and Limitations of Microarray-Based Gene Expression Profiling of Thyroid Tumors Endocr. Rev., May 1, 2007; 28(3): 322 - 338. [Abstract] [Full Text] [PDF]

				K. Krause, S. Karger, A. Schierhorn, S. Poncin, M.-C. Many, and D. Fuhrer Proteomic Profiling of Cold Thyroid Nodules Endocrinology, April 1, 2007; 148(4): 1754 - 1763. [Abstract] [Full Text] [PDF]

				K. Krause, S. Karger, O. Gimm, S.-Y. Sheu, H. Dralle, A. Tannapfel, K. W. Schmid, C. Dupuy, and D. Fuhrer Characterisation of DEHAL1 expression in thyroid pathologies Eur. J. Endocrinol., March 1, 2007; 156(3): 295 - 301. [Abstract] [Full Text] [PDF]

				J. Galle, D. Sittig, I. Hanisch, M. Wobus, E. Wandel, M. Loeffler, and G. Aust Individual Cell-Based Models of Tumor-Environment Interactions: Multiple Effects of CD97 on Tumor Invasion Am. J. Pathol., November 1, 2006; 169(5): 1802 - 1811. [Abstract] [Full Text] [PDF]

				M. Eszlinger, M. Wiench, B. Jarzab, K. Krohn, M. Beck, J. Lauter, E. Gubala, K. Fujarewicz, A. Swierniak, and R. Paschke Meta- and Reanalysis of Gene Expression Profiles of Hot and Cold Thyroid Nodules and Papillary Thyroid Carcinoma for Gene Groups J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1934 - 1942. [Abstract] [Full Text] [PDF]

				D Fuhrer, M Eszlinger, S Karger, K Krause, C Engelhardt, D Hasenclever, H Dralle, and R Paschke Evaluation of insulin-like growth factor II, cyclooxygenase-2, ets-1 and thyroid-specific thyroglobulin mRNA expression in benign and malignant thyroid tumours Eur. J. Endocrinol., May 1, 2005; 152(5): 785 - 790. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 90/2/1163    most recent Author Manuscript (PDF) Submit a related Letter to the Editor Purchase Article View Shopping Cart Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Eszlinger, M. Articles by Paschke, R. Search for Related Content PubMed PubMed Citation