NATH, a novel gene overexpressed in papillary thyroid ... - Nature

ª

Oncogene (2002) 21, 5056 – 5068 2002 Nature Publishing Group All rights reserved 0950 – 9232/02 $25.00 www.nature.com/onc

NATH, a novel gene overexpressed in papillary thyroid carcinomas Øystein Fluge*,1,4, Ove Bruland2, Lars A Akslen3, Jan E Varhaug4 and Johan R Lillehaug1 1

Department of Molecular Biology, University of Bergen, N-5020 Bergen, Norway; 2Department of Medical Genetics and Molecular Medicine, Haukeland Hospital, University of Bergen, N-5021 Bergen, Norway; 3Department of Pathology, The Gade Institute, Haukeland Hospital, University of Bergen, N-5021 Bergen, Norway; 4Department of Surgery, Haukeland Hospital, University of Bergen, N-5021 Bergen, Norway

In this study a replica cDNA screening (RCS) approach to identify genes differentially expressed in papillary thyroid carcinomas (PTC) was used, as compared to non-neoplastic thyroid tissues. RCS is based on hybridization of radioactively labeled cDNA probes made from the biopsies to replica membranes with 15 000 clones from a PTC cDNA library. Among the genes overexpressed in PTC, and especially in clinically aggressive tumors with histologic evidence of poorly differentiated or undifferentiated areas, a novel gene named NATH was found. NATH has two mRNA species, 4.6 and 5.8 kb, both harboring the same open reading frame encoding a putative protein of 866 amino acids. The NATH protein is homologous to yeast N-acetyltransferase (NAT)1 and to mouse NARG1 (mNAT1) and contains four tetratricopeptide repeat (TPR) domains, suggesting that NATH may be part of a multiprotein complex. Overlapping RT – PCR fragments from several PTC biopsies confirmed the NATH mRNA sequence. Northern blots, semiquantitative RT – PCR experiments, TaqMan real-time RT – PCR experiments, and in situ hybridization verified the overexpression of NATH mRNA localized to tumor cells in PTC biopsies. NATH was expressed at a low level in most human adult tissues, including the normal thyroid gland. Increased NATH expression was seen especially in a Burkitt lymphoma cell line and in adult human testis. Recombinant in vitro expression showed that NATH protein was located mainly in the cytoplasm, and was present as a single protein band of the expected 105 kDa molecular weight. Heterologous expression of NATH in the papillary carcinoma cell line (NPA) and 293 cells did not alter the cellular proliferation rate. The biological function of NATH remains to be elucidated, but the overexpression in classic PTC and especially in poorly differentiated or undifferentiated components may indicate a function in the progression of papillary thyroid carcinomas. Oncogene (2002) 21, 5056 – 5068. doi:10.1038/sj.onc. 1205687 Keywords: thyroid carcinoma; overexpression; NATH

*Correspondence: Ø Fluge, Department of Molecular Biology, University of Bergen, Thormøhlens gt. 55, N-5020 Bergen, Norway; E-mail: [email protected] Received 5 June 2001; revised 15 May 2002; accepted 20 May 2002

Introduction Papillary thyroid carcinoma (PTC), which is the most common histologic type of thyroid cancer, generally has a good prognosis. However, in subsets of patients, especially when tumors show evidence of poorly differentiated or undifferentiated components (i.e. solid structure, increased nuclear atypia, mitotic activity, tumor necrosis, vascular invasion) the risk of aggressive behavior is increased (Pilotti et al., 1995; Kurozumi et al., 1998; Nishida et al., 2002). Among characteristics at the molecular level, the ret gene is of particular importance in thyroid carcinomas (Santoro et al., 1999). ret rearrangements are found almost exclusively in papillary thyroid carcinomas (Santoro et al., 1992), and nude mice with targeted expression of ret/PTC-1, or ret/PTC-3, develop thyroid tumors with resemblance of PTC (Santoro et al., 1996; Powell et al., 1998). On the other hand, TP53 mutations are prevalent in poorly differentiated thyroid tumors, especially in the highly malignant undifferentiated (anaplastic) carcinomas (Fagin et al., 1993) but uncommon in differentiated PTC. Tyrosine kinase growth factor receptors and their cognate ligands such as Met (Di Renzo et al., 1992) and hepatocyte growth factor (HGF) (Trovato et al., 1998); ErbB family members (Haugen et al., 1996) and heregulins (Fluge et al., 2000); vascular endothelial growth factor (VEGF) receptors and VEGFs (Bunone et al., 1999), have all been reported as overexpressed in PTC. Increased expression of fibronectin (Ryu et al., 1997; Takano et al., 1998) and cathepsin B (Shuja et al., 1999) is also associated with the PTC phenotype. The cyclindependent kinase inhibitor p27/kip1 protein (Resnick et al., 1998) and the adhesion molecule E-cadherin (Scheumman et al., 1995; von Wasielewski et al., 1997) are examples of genes down-regulated in PTC versus non-neoplastic thyroid tissue. A recent microarray study, using oligonucleotides representing more than 12 000 genes, showed highly consistent gene expression profiles in PTC (Huang et al., 2001). Among the genes up-regulated were several previously known such as fibronectin and Met, and several others such as CITED1 (Cbp/p300-interacting transactivator1), EPS8 (epidermal growth factor receptor kinase substrate 8), and RIL (37 kDa LIM domain protein). Consistently down-regulated in PTC versus non-neoplastic thyroid

NATH overexpression in papillary thyroid carcinomas Ø Fluge et al

5057

tissue were several genes related to specialized thyroid functions such as TPO (thyroid peroxidase), DIO1 (type 1 5-iodothyronine deiodinase) and DIO2 (type 2 iodothyronine deiodinase). In spite of several known mutations and changes in gene expression, the molecular biology of PTC is still largely unresolved. The search for differentially expressed genes in malignant tumors versus non-neoplastic tissue is a strategy to find parameters of diagnostic and prognostic significance and to elucidate molecular mechanisms of value as possible targets for future therapy. Using a replica cDNA screening (RCS) protocol, we screened approximately 15 000 clones from a PTC cDNA library, and found several known and unknown differentially expressed genes. Here we describe a novel gene which is overexpressed in PTC as compared to non-neoplastic thyroid tissue. We have named this gene NATH due to its amino acid similarity to Saccharomyces cerevisiae N-acetyltransferase (NAT)1 (Mullen et al., 1989). Results Identification of differentially expressed genes The cDNA species in the amplified PTC cDNA library had a size range of 500 – 5000 bps, with a mean of approximately 1200 bps (Figure 1a). To reduce the number of false positive gene candidates, a secondary and a tertiary screening were performed using ordered macroarray membranes and hybridization with complex cDNA probes. The number of gene candidates was reduced from 15 000 in the primary screening to 180 in a secondary screening, to 70 after DNA sequencing and finally to 40 in the tertiary screening. The latter consisted of dot blot array hybridization experiments based on cDNA probes from 20 tumor and eight non-neoplastic thyroid biopsies. The fibronectin (Figure 1b) and the cathepsin B genes are examples of known genes identified by the RCS protocol to be consistently overexpressed in PTC, in accordance with the literature. NATH is a novel gene up-regulated in PTC The candidate gene B2-39 (NATH) was found to be overexpressed in PTC (Figure 1b,c). As assessed by hybridization, and normalized according to the sum of expression levels for several non-differentially expressed genes, mean NATH expression was significantly different from the non-neoplastic thyroid tissue group both in the ‘classic’ PTC (P=0.02) and in the aggressive PTC (P=0.008) groups. Also, there was a significant difference between mean NATH expression in the ‘classic’ and aggressive PTC groups (P=0.008). Using the 1.5 kb fragment identified from the PTC cDNA library, and combining overlapping human expressed sequence tags (EST), two putative NATH transcripts of 5.5 and 4.3 kb were constructed in silico. Five primer pairs (Table 1, Figure 2a) were designed and used for RT – PCR experiments with cDNA from

several PTC biopsies as templates (Figure 2b). DNA sequencing verified the 5.5 kb NATH mRNA sequence. The complete nucleotide sequence of the 5.5 kb transcript is present in the DDBJ/EMBL/ GenBank Nucleotide Sequence Database under the accession number AJ314788. NATH mRNA contained a 2598 nts long open reading frame, encoding a polypeptide of 866 amino acids with predicted molecular weight 101 kDa (Figure 3, and GenBank accession no. AJ314788). Database search indicated that the NATH gene is located on chromosome 4, spanning at least 24 exons in a 90 kb genomic DNA region. Blast search showed the putative NATH protein to be homologous to Saccharomyces cerevisiae N-acetyltransferase (NAT)1 (Mullen et al., 1989) (854 amino acids, 25% identity). NATH is also homologous to the (partial) sequences of mouse NARG1 (Sugiura et al., 2001) and mouse Tubedown1 (Tbdn-1) (Gendron et al., 2000) proteins, in the Nterminal and C-terminal parts, respectively (Figure 4). No evidence of alternatively spliced NATH transcripts affecting the open reading frame was detected from either PTC biopsies, or from the papillary carcinoma cell line NPA, in the RT – PCR experiments (Figure 2c). One expressed sequence tag (EST) (BE504732) suggested the presence of an alternative splicing of 1202 nts located in the 3’-untranslated region (Figure 2a). The semiquantitative RT – PCR experiments using cDNA of PTC or non-tumor biopsies as template, and normalized to internally coamplified b-actin, demonstrated that NATH was clearly overexpressed in different PTC biopsies as compared to non-neoplastic thyroid tissue (representative experiments in Figure 2d). TaqMan quantitative real-time RT – PCR experiments, also normalized using b-actin as internal standard, verified the NATH overexpression in PTC (Figure 2e). As a morphological analysis of NATH expression, in situ hybridization experiments were performed using a mixture of two antisense biotinylated oligonucleotides and paraffin-embedded specimens of PTC and non-neoplastic thyroid tissues. These experiments showed that NATH was expressed markedly in the tumor cells of PTC (Figure 5a, seen in more detail in b), while only scattered and weak expression was seen in adjacent non-neoplastic thyroid follicular cells (Figure 5c,d). Parallel controls omitting the probe, or using a sense biotinylated NATH probe, were negative (Figure 5e). Northern blots also showed expression in the papillary carcinoma cell line NPA and in the anaplastic thyroid carcinoma cell line ARO. The major NATH mRNA was approximately 4.6 kb, while a minor 5.8 kb mRNA was also present (Figure 6a). Identical mRNA sizes were found in Northern blot experiments using total RNA from the PTC biopsies. Identical hybridization results were obtained using two different NATH probes. The multi-tissue mRNA dot blot experiment (Figure 6b) showed that the NATH gene was expressed at a low level in most adult tissues, including the normal thyroid, liver, pancreas, mammary and salivary glands, Oncogene


5058

Figure 1 Identification of NATH by Replica cDNA Screening (RCS). (a) Total RNA was extracted from papillary carcinoma biopsies and used for cDNA library construction as described. Lanes 2 – 4 show the amplified ds cDNA from three histologically representative papillary carcinomas. The lengths of the cDNA species were in the range 500 – 5000 bps. (b) Parts of dot blot arrays showing hybridization of radioactively labeled cDNA probes made from a papillary carcinoma (upper lane) and a non-neoplastic thyroid tissue specimen (lower lane) to DNA fragments specific for fibronectin cDNA, NATH cDNA, and a non-differentially expressed gene (denoted C-21). (c) Relative NATH mRNA expression in papillary carcinomas and non-neoplastic thyroid biopsies determined by dot blot hybridization, using the NATH-specific cDNA fragment on the membranes and hybridization to labeled cDNA probes from the indicated biopsies, as described. The white bars represent the normal thyroid tissue biopsies, the grey bars represent the group of differentiated papillary carcinomas, and the black bars represent the group of clinically aggressive papillary carcinomas. Normalization was made relative to the sum of the mRNA expression levels for several genes, as described. Abbreviations: M, molecular weight marker; PC, papillary carcinoma; NT, non-neoplastic thyroid biopsy; T2, T3, T4 refer to the thyroid carcinoma tumor staging (TNM) system; PC-ld, papillary carcinoma with evidence of poorly differentiated areas; PCr-M1, lymph node recurrence of papillary carcinoma in a patient with synchronous distant metastases

lung, ovary, urogenital system, and upper gastrointestinal tract (Figure 6b). A somewhat higher level of NATH expression was seen in parts of the human brain, the heart, colon, bone marrow, and in several leukemia and carcinoma cell lines. The highest level of NATH expression was found in adult testis, and in a Burkitt lymphoma cell line. Reprobing the membrane with two different probes from non-overlapping parts of NATH mRNA gave similar hybridization results. The putative NATH protein harbors four tetratricopeptide repeat (TPR) domains in two tandems, located between amino acid residues 46 – 113, and 374 – 441, respectively. A possible bipartite nuclear localization signal was predicted between residues 612 – 628 (KKNAEKEKQQRNQKKKK). The NATH open reading frame was cloned into pcDNA3.1-Myc-His vector, encoding a fusion protein of 897 amino acids with a C-terminal Myc-epitop and His-tag, and used for transient transfection of NPA and 293 cells. Western immunoblot experiments using a monoclonal anti-Myc antibody showed a specific Oncogene

protein band of the predicted molecular weight 105 kDa, present in cell lysates of transfected cells, which was absent in non-transfected cells (Figure 7b). Repeated experiments did not reveal any direct alteration of cellular proliferation rate in NPA or 293 cells, when comparing cells transfected with NATH and cells transfected with an empty plasmid. Representative results are shown in Figure 7b,c. Assessed by cotransfection with a LacZ-encoding plasmid, 40 – 45% of both 293 and NPA cells were successfully transfected. From the X-Gal staining of individual cells, and also from the Western immunoblot experiments, the level of heterologous protein expression was clearly higher in 293 than in NPA cells. In accordance with the immunoblot experiments, immunocytochemistry of Cos-7 cells transfected with the NATH/pcDNA3.1-Myc-His plasmid showed a predominant cytoplasmic staining (Figure 7a, panels 1, 2). Untransfected cells, and transfected cells omitting the primary antibody, were negative (Figure 7a, panel 3).


5059 Table 1 Name cDNA library AnkT16V Ank – F M13Fmod RT – PCR NATH – F1 NATH – F2 NATH – F3 NATH – F4 NATH – F5 Actin – F Cloning T7 NATH – Bam – F In-situ hybridization NATH – ISH1 NATH – ISH2 TaqMan RT – PCR NATH – Taq – F NATH – probe

Sequence (5’ – 3’)

Primers and probes used in this study

Positiona Name

gaccacgcgtatcgatgtcgact16v gaccacgcgtatcgatgtcgac aacgacggccagtgaattgt cggcggcagcgttaagtgag agcttcgacctgcgcagagag aggccttggacacagcagacaga tgccacacagattatcagctgcc tgcactgtgatgtggcaccaact ggcaccacaccttctaca

1 – 19 633 – 653 1512 – 1530 2493 – 2515 3596 – 3618 330 – 347

taatacgactcactataggg gagctcggatccgaaacgatgccggccgtg agcctcccg

Sequence (5’ – 3’)

SMA – G3c SMA – Fc M13R

aagcagtggtaacaacgcagagtacgcggg aagcagtggtaacaacgcagagt caggaaacagctatgac

NATH – R1 NATH – R2 NATH – R3 NATH – R4 NATH – R5 Actin – R

ggcccagttttcaggatttctctc ccgctgatgttccttcccttgta ggctaccatcatacaaggcttcca tcgtgaacaaggcctgctgag acgggtggcatctgcatcttaag aggaaggctggaagagtg

Positiona

PCR-prod. Ann (bp) tpb

500 – 5000 65 56 985 – 962 1634 – 1612 2648 – 2625 3762 – 3742 4943 – 4921 869 – 852

BGH – R tagaaggcacagtcgagg NATH – Not – R gactcgagcggccgcccaatttca ttggccagttcttc

985 1002 1143 1270 1348 540

56 58 56 59 59 56

2800 2663

56 58

75

60

gcgcaggtcgaagctgaagtaactgataccd 647 – 618 ggcctcatccatccaccttgcagcttctttd 1510 – 1481 cttactacagattcaaatgcgagatc actgatacctcgtttccctgtaaccctcae

571 – 596 NATH – Taq – R cgcaggtcgaagctgaag 598 – 626

646 – 629

a

Position relative to NATH mRNA (GenBank accession AJ314788), except for b-actin (GenBank NM_001101). bAnnealing temperature in PCR (8C). cSequences from the ‘SMART’ procedure (Clontech). dBiotinylated 5’. eFAM (6-carboxyfluorescein)-labeled 5’, and TAMRA (carboxytetramethylrhodamine)-labeled 3’

Discussion We studied genes differentially expressed in papillary thyroid carcinomas when compared to non-neoplastic thyroid tissue specimens. Histologically representative tumor and non-tumor biopsies were used, and the patients were selected to ensure subgroups with different clinical behavior. On the basis of a Replica cDNA Screening (RCS) protocol, a novel gene named NATH was found to be upregulated in the PTC group as compared to nonneoplastic thyroid tissue. The 5.5 kb NATH transcript harbored an open reading frame encoding a putative 866 amino acid protein, which was homologous to S. cerevisiae NAT1 (Mullen et al., 1989), mouse NARG1 (Sugiura et al., 2001), and to mouse Tbdn-1 (Gendron et al., 2000). The NATH protein contains four tetratricopeptide repeat (TPR) domains in two tandems. TPR are degenerate 34 amino acid repeats often associated with protein-protein interactions (Blatch and Lassle, 1999), and are present in diverse proteins important e.g. for cell cycle regulation (Lamb et al., 1994), transcription (Gounalaki et al., 2000; Nakatsu et al., 2000), and also in chaperone complexes (Scheufler et al., 2000). The presence of TPR domains in NATH suggests that this protein may be part of a multiprotein complex. The RNA expression data using macroarray membranes and labeled cDNA probes from the biopsies, supported by the semiquantitative RT – PCR assays, quantitative real-time TaqMan RT – PCR experiments, and in situ hybridization, provide evidence that NATH is overexpressed in the PTC tumor group as compared to non-neoplastic thyroid tissue. The hybridization data indicate NATH overexpression especially in the group of clinically aggressive tumors,

but also to a lesser extent in the ‘classic’ PTC group. In the hybridization series, expression data were normalized according to the sum of expression levels for several non-differentially expressed genes, including two encoding ribosomal proteins. Northern blot experiments also showed NATH expression in the PTC cell line NPA and in the anaplastic thyroid carcinoma cell line ARO. NATH was expressed at very low levels in most adult tissues including the normal thyroid, although expression at a somewhat higher level was seen in parts of the human brain, the heart, bone marrow and in several cell lines. The highest level of NATH expression was, however, recorded in human testis and in a Burkitt lymphoma cell line, both characterized by a high cellular proliferation rate. These data, and the relative overexpression primarily in aggressive and dedifferentiated thyroid carcinomas suggest that NATH may be involved in cellular proliferation at some level. Interestingly, mouse tbdn1 was expressed in M1 myeloid leukemia cells and was down-regulated during LIF-induced differentiation (Gendron et al., 2000), while mouse NARG-1 (mNAT1) was up-regulated in neonatal brain regions with neuronal proliferation, and further down-regulated by NMDA receptor activation (Sugiura et al., 2001). However, our NATH transfection experiments of papillary thyroid carcinoma (NPA) cells and 293 cells did not seem to alter the cellular proliferation rate. The NATH transfection efficacy assessed by cotransfection with a LacZ plasmid was adequate both in 293 and NPA cells. If NATH is related to cellular proliferation, these results suggest that overexpression of NATH alone may not be sufficient, as suggested from the bioinformatics results indicating that NATH may be part of a protein complex. Oncogene


5060

Figure 2 NATH expression in papillary thyroid carcinomas. (a) Schematic presentation (not to scale) of primer positions (see also Table 1) used for RT – PCR and cloning, and the positions of antisense oligonucleotides used for in situ hybridization (ISH), indicated relative to NATH cDNA. The open reading frame is indicated by the start codon ‘atg’ and the stop codon ‘tga’. The positions of the initially cloned cDNA fragment (1.5 kb) and of the putative alternative splicing event in the 3’-untranslated region are indicated. Abbreviations as in Figure 1. (b) NATH specific RT – PCR products obtained using cDNA templates from papillary carcinomas. The five different primer pairs used (Table 1) are indicated below each lane. The five overlapping PCR products, from several papillary carcinoma templates, were sequenced to verify the NATH mRNA sequence. (c) NATH specific DNA fragments obtained by nested PCR using primer pairs covering the putative open reading frame. The RT – PCR products were generated using cDNA templates from papillary carcinoma biopsies or from a papillary carcinoma cell line (NPA). The primer pairs used in the nested reactions are indicated below each lane. (d) Semiquantitative assessment of NATH mRNA in papillary carcinomas and non-neoplastic thyroid tissues using a duplex RT – PCR with the primers NATH-F1/R1 amplifying a 985 bp NATH cDNA fragment, and a primer pair amplifying a 540 bp b-actin fragment included as an internal control of mRNA amount and quality. The numbers of PCR cycles are shown below each panel. Representative results from two papillary carcinomas and one non-neoplastic thyroid specimen are shown. The NATH specific cDNA fragment (denoted N) and the b-actin cDNA fragment (denoted A) are indicated by arrows. (e) Results from Taqman real-time quantitative RT – PCR experiments on selected papillary carcinomas and non-neoplastic thyroid biopsies, normalized to b-actin mRNA as internal standard, as described

The transfection experiments showed that the recombinant NATH protein was present mainly in the cytoplasm, and occasionally in the nucleus. A putative Oncogene

bipartite nuclear localization signal was found between amino acid residues 612 – 628. The putative complex partner TE2 (human homolog of yeast ARD1) was also


5061

Figure 3 NATH cDNA (open reading frame) and amino acid sequence of the putative NATH protein. The single underlined regions denote the tetratricopeptide repeat (TPR) domains. The double underlining refer to the putative nuclear localization signal. The complete nucleotide sequence of the 5.5 kb NATH cDNA is given in GenBank accession no. AJ314788

predicted in silico to be nuclearly located (not shown). Thus, we speculate that NATH may be shuttling between the nucleus and the cytoplasm. The presence of the heterologous 105 kDa protein band in extracts of transfected cells showed that the suggested open reading frame may indeed translate into the predicted protein, with no evidence of proteolytic cleavage. Northern blot experiments showed that the major NATH mRNA was approximately 4.6 kb with a minor transcript of 5.8 kb. Using a nested RT – PCR approach covering the complete open reading frame,

no evidence of alternative splicing affecting the open reading frame could be detected. A putative alternative splicing event removing 1202 nts from the 3’untranslated region may account for the presence of two bands in the Northern data. This splicing event is supported by the presence of an EST (BE504732), but was not detected in our RT – PCR experiments because the primer NATH-R5 was located within the actual spliced region permitting amplification only of the longer transcript. The 5.5 kb NATH cDNA sequenced therefore represents the longer 5.8 kb mRNA incomOncogene


5062

Figure 4 Alignment (ClustalW) of NATH, mouse NARG1, mouse Tubedown-1, and yeast NAT1

pletely sequenced in the 5’-untranslated region. In contrast, mouse Tbdn-1 mRNA was found to be 3.4 kb (Gendron et al., 2000), while Northern blot of mouse NARG1 mRNA (Sugiura et al., 2001) are consistent with our data. At the protein level, mouse Tbdn-1 was reported to be 69 kDa, suggested to result from an alternative translation initiation predicted to encode a protein of 594 amino acids. The protein sequence of mouse NARG-1 was only partial, but with identity to NATH in the N-terminal 137 amino acids. Whether mouse Tbdn-1 and human NATH are derived from the same gene, remains to be clarified. Studies describing function of the putative homologous genes suggest a relevance of NATH also in human tumor biology. In yeast, the NAT1 and ARD1 genes encode proteins acting in concert to achieve protein NAT activity (Mullen et al., 1989; Park and Szostak, 1992). Both Saccharomyces NAT1 and mouse Tbdn-1 were found to harbor NAT-activity (Lee et al., 1990; Gendron et al., 2000). However, in yeast, the main catalytic unit seems to be ARD1 (Polevoda et al., 1999), a member of the GCN5-related N-acetyltransferase family (Neuwald and Landsman, 1997). The ARD1/NAT1 complex acetylates several proteins important for mating phenotype and for cell cycle control, regulating entry into the G0 phase (Park and Szostak, 1992). Although more than one-half of eucaryotic proteins are N-acetylated (Driessen et al., 1985), the limited but similar phenotype divergence of either NAT1- or ARD1 null mutants (Mullen et al., Oncogene

1989) suggests that only a subset of proteins is critically dependent on these proteins for adequate function in yeast. It is tempting to suggest that the human counterpart of the yeast NAT1/ARD1 complex is the NATH described herein and a human homolog of ARD1 such as TE2 (Tribioli et al., 1994). The role of specific protein acetylation as an important mechanism regulating cellular differentiation, growth and apoptosis, is increasingly recognized. For example, acetylation may enhance the DNA binding and transcriptional properties of p53 (Liu et al., 1999), MyoD (Sartorelli et al., 1999), and of GATA-1 (Boyes et al., 1998). It is well known that histone acetyltransferases are important chromatin modifiers and transcriptional cofactors. Histone acetylation modifiers have also been implicated in the pathogenesis of malignant disease (Mahlknecht and Hoelzer, 2000). NATH was selected among several unknown genes found to be overexpressed in PTC. The replica cDNA screening protocol proved to be useful for the identification of differentially expressed genes, by an initial screening of 15 000 colonies and through subsequent screening of candidate genes using macroarray membranes and complex cDNA probes. The hybridization data obtained from nylon array membranes and complex cDNA probes have been shown to be in good consistency with Northern experiments (Bernard et al., 1996). In support for the validity of the RCS procedure, we identified several


5063

Figure 5 NATH expression demonstrated by in situ hybridization. In situ hybridization experiments were performed on selected representative samples of papillary carcinomas and normal thyroid tissue, using a mixture of two NATH-specific antisense biotinylated oligonucleotides and paraffin-embedded tissue sections, as described. Shown are NATH expression in a classic papillary carcinoma in (a) and (b), and in adjacent non-neoplastic thyroid tissue (on the same tissue section) in (c) and (d). In (e) is shown a negative control, processed in parallel, using a sense NATH biotinylated oligonucleotide

known genes previously reported as overexpressed in PTC, among these were the fibronectin and cathepsin B genes (Takano et al., 1998; Shuja et al., 1999). When screening for genes differentially expressed in tumor and non-tumor tissues, the histopathological evaluation of biopsies for representativity is important. Even in biopsy material from representative tumor areas, the gene candidates isolated may stem from other cell types such as stromal cells, endothelial cells, or lymphocytes (Perou et al., 2000). Therefore, when tissue samples are used to generate cDNA probes, a hybridization series with probes made from many individual tumor and non-tumor biopsies should be performed, in order to minimize variations due to cellular heterogeneity in individual tissue samples, and to identify natural variation in mRNA expression level of genes with high inherent expression variability.

In conclusion, we have used a replica cDNA screening system for the identification of differentially expressed genes in PTC as compared to non-neoplastic thyroid tissue. Among several candidate genes found to be upregulated in PTC we report the identification of the novel gene NATH, with similarity to the yeast NAT1 and mouse NARG1 or Tbdn-1 genes. NATH mRNA harbors an open reading frame encoding a protein of 866 amino acids, including four TPR domains. When exogenously expressed, NATH was located in the cytoplasm and also occasionally in the nucleus of transfected cells. NATH showed a very low level of expression in most adult tissues, with the highest level of expression recorded in the testis and a Burkitt lymphoma cell line, both characterized by a high cellular proliferation rate. NATH overexpression in the NPA cell line did not alter the cellular proliferation rate per se. The Oncogene


5064

Figure 6 NATH RNA expression assessed by Northern blot and multi-mRNA blot. (a) Northern blot of total RNA from the papillary thyroid carcinoma cell line NPA and the anaplastic thyroid carcinoma cell line ARO. The probe was generated from a 2457 bp NATH specific template covering most of the open reading frame, as described. (b) A human multiple tissue expression array membrane was used (Clontech, cat. #7775-1). The hybridization experiments were performed twice with two different probes, as described. The numbers indicate: 1, adult testis; 2, Burkitt’s lymphoma (Daudi) cell line; 3, colorectal carcinoma cell line (SW 480); 4, adult corpus callosum; 5, adult thyroid gland; 6, human genomic DNA; 7, controls including yeast total RNA, E. coli rRNA, E. coli DNA, Poly r(A), and human Cot-1 DNA were all negative

biological function of NATH remains to be elucidated, but its sequence similarity to NAT1 and NARG1, and the overexpression especially in some aggressive tumors, may implicate a function in the progression of PTC. Materials and methods cDNA library from PTC Total RNA was purified from the fresh frozen thyroid tissue biopsies using a modified Chomczynski and Sacchi procedure (Chomczynski and Sacchi, 1987). High quality RNA, isolated from three histologically representative PTC, one differentiated ‘classic’ and two PTC including areas of poorly differentiation, were used for cDNA library construction. The cDNA was made using a modified ‘SMART’-system (Clontech, Palo Alto, CA, USA) as follows: 5 mg total RNA was ethanolprecipitated, washed, and dissolved in 15 ml diethylpyrocarbonate-(DEPC) treated water with 20 U RNasin (Promega, Madison, WI, USA). First-strand cDNA was synthesized in a total volume of 30 ml, using Superscript II reverse transcriptase, followed by RNase H treatment (Gibco-BRL, Gaithersberg, MD, USA). The cDNA synthesis was primed using AnkT16V and the Sma-G3 ‘switch’-primer (Clontech) in combination (Table 1). Two ml of the cDNA were used as a template in a hot-start PCR using the Advantage2 polymerase (Clontech), with primers Ank-F and Sma-F (Table 1). The PCR was performed on a GeneAmp 2400 thermocycler (Perkin-Elmer) programmed for 15 cycles of 958C for 20 s, 658C for 1 min, and 708C for 4 min. The amplified doublestranded cDNA was then cloned in 50 separate ligation and transformation reactions using the TOPO-TA Cloning kit (pCR2.1-TOPO) with Top10 ultracompetent E. coli (Invitrogen, San Diego, CA, USA). Approximately 20 000 individual colonies were obtained and pooled in LB/ampicillin-medium and titrated for number of bacteria per millilitre. Replicas 3 – 5000 bacteria from the cDNA library were seeded onto prewetted 13 cm nylon membranes (HybondNX) (AmerOncogene

sham, Chicago, IL, USA). These ‘master’ membranes were incubated at 378C overnight, on LB-agar containing 50 mg/ml ampicillin allowing the colonies to grow directly onto the membrane. Four master membranes totaling at least 15 000 individual bacterial colonies were made. Replicas were then made from each master membrane using direct contact transmission to a new prewetted HybondNX membrane. Four replica sets were made, each set consisting of a replica from each of the four master membranes. The master membranes were stored on LB-agar at 48C for backup and later isolation of colonies of interest. The 16 replica membranes were subjected to lysis of the bacterial colonies for 3 min on a 3M Whatman paper soaked in 10% SDS (bacterial side up), followed by denaturation on 3M paper soaked in 0.5 M NaOH and 1.5 M NaCl, and neutralization on 3M paper soaked in 1.5 NaCl and 0.5 M Tris-HCl, pH 7.2. Finally, the membranes were air-dried at room temperature before UV-crosslinking. Primary, secondary and tertiary screening Each replica set, comprising a total of 15 000 colonies on four membranes, was hybridized with radioactively labeled firststrand cDNA constructed from total RNA isolated from either a PTC or a non-neoplastic thyroid biopsy. The cDNA probes were labeled as follows: 5 mg total RNA was ethanolprecipitated, washed, and dissolved in 12 ml DEPC-treated H2O with addition of 20 U RNasin and 5 mM oligo-dT16. First-strand cDNA synthesis was performed at 428C for 60 min using Superscript II reverse transcriptase. The nucleotide concentration was 0.33 mM each of dATP, dGTP and dTTP. Five ml [a-32P]-dCTP (370 mBq/ml) (Amersham) was added in a total reaction volume of 30 ml. The reverse transcription was followed by RNase H treatment. Each probe contained approximately 76107 c.p.m. Prehybridization for 5 h, followed by hybridization for 30 h, was carried out at 678C, in 56SSPE (0.75 M NaCl, 0.05 M NaH2PO4×H2O, 5 mM Na2×EDTA, pH 7.4), 106Denhardt’s, and 2% sodium dodecyl sulphate (SDS), with 0.5 mg/ml sheared and sonicated herring sperm DNA, and the 32P-labeled and denatured cDNA probe. The membranes were washed four times in 26SSC (0.3 M NaCl,


5065

Figure 7 NATH in vitro expression, immunostaining and cell proliferation assay. (a) Subcellular localization of heterologous NATH. Cos-7 cells were grown on coverslips and transfected with the NATH/pcDNA3.1-Myc-His plasmid. A mouse anti-Myc primary antibody and FITZ-conjugated anti-mouse IgG1 secondary antibody were used, followed by immunofluorescence observation with a confocal laser scanning microscope, as described. Panels 1 and 2 show transfected cells. Untransfected cells are shown in panel 3. (b) and (c) Heterologous NATH expression in human embryonal kidney 293 cells (b) and in papillary thyroid carcinoma (NPA) cells (c). Cell proliferation assays were performed in 293 and NPA cells as described, each cell line transfected with either NATH/pcDNA3.1-Myc-His plasmid (denoted 293-NATH and NPA-NATH, respectively) or an empty plasmid (denoted 293pcDNA and NPA-pcDNA, respectively). At each time point, the cells from each dish were counted 30 times, symbols and error bars indicate means and standard deviations. To the right in (b) and (c) are shown Western immunoblot experiments with protein extracts of untransfected and transfected cells separated on 10% SDS-polyacrylamide gels and electroblotted. The monoclonal antiMyc-HRP antibody was used as described. The arrows indicate the 105 kDa NATH specific protein band in transfected cells

0.03 M Na3Citrate×2H2O, pH 7.0) and 1% SDS, and then twice in 0.16SSC and 0.5% SDS at 668C. The relative signal intensity of each individual spot was determined by phospho-

imaging using FLA-2000 (Fujifilm, Tokyo, Japan). Spots representing differentially expressed genes were identified by comparing corresponding replica membranes hybridized Oncogene


5066

against either a PTC or a non-neoplastic complex cDNA probe. The corresponding colonies were then isolated from the backup-membrane and cultured in LB-medium for plasmid purification and isolation of cDNA inserts by PCR. The PCR products of 180 colonies, each with putative differential expression in the primary screening, were obtained using M13 polylinker primers (Table 1) and purified (Qiaquick PCR Purification Kit; Qiagen, Santa Clara, CA, USA). Approximately 50 ng of each amplified DNA fragment was dot blotted on HybondN+ nylon membranes prewetted in 106SSC. The DNA was denatured, neutralized, and UV crosslinked. Hybridization experiments using complex cDNA probes made from another two independent PTC and two non-neoplastic thyroid tissue specimens were carried out in this secondary screening. Phospho-imaging assessed the mRNA expression levels for each candidate gene. After this secondary screening, 70 candidate gene fragments were sequenced using the ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit and the ABI 377 sequencer. The final (tertiary) screening was performed using the purified PCR-products of 40 gene candidates selected after sequencing. Hybridization was performed using labeled complex cDNA probes constructed from 20 tumor and 8 non-tumor thyroid biopsies. All patients in this study had a main diagnosis of PTC, i.e. predominantly classic PTC. However, some cases were selected due to the presence of poorly differentiated or undifferentiated components. In every case, precise histologic classification was performed on tissue adjacent to the part used for RNA extraction. Thus, the tumor biopsies were selected to ensure subgroups with different clinical behavior. One tumor group comprised 11 differentiated (classic) PTC of tumor stage pT2-pT4, according to the TNM system. A second group comprised nine clinically aggressive tumors, either primary PTC tumors with components of predominantly poorly differentiated or undifferentiated areas in the sample adjacent to the part used for RNA extraction (n=6), or local recurrences from PTC patients with synchronous distant metastases (n=3). New identical membranes were made for each of the 28 hybridization experiments. Each gene fragment was applied in duplicate on the membranes and the mean difference in hybridization signal intensity between parallel dots was approximately 7 – 8% with a standard deviation of 8 – 9%. The signal intensity of the TOPO-TA plasmid polylinker was used as background and subtracted. Normalization was made relative to the sum of the mRNA expression levels for several genes, including ribosomal protein S4 mRNA and ribosomal protein S25 mRNA. RT – PCR experiments The RT – PCR experiments were performed as follows: total RNA was extracted and subjected to first-strand cDNA synthesis primed by oligo-dT16 and random nucleotide hexamer, using Superscript II reverse transcriptase, and then RNase H treated. One ml of the first-strand reaction mixture was used as template in a 50 ml PCR (30 cycles), performed on a GeneAmp 2400 (Perkin Elmer) with Advantage2 polymerase (Clontech). The primers and annealing temperatures used for PCR are given in Table 1, and the primer positions relative to NATH mRNA are indicated in Figure 2a. Samples of interest from each RT – PCR experiment were purified and sequenced in both directions. To verify the differential gene expression detected in the hybridization experiments, semiquantitative duplex RT – PCR experiments with cDNA of tumor or non-tumor biopsies as template, were performed. The gene specific primers NATHOncogene

F1/NATH-R1 amplifying a 985 bp NATH fragment and the primer pair Actin-F/Actin-R (Table 1) amplifying a 540 bp bactin fragment were used, the latter included as an internal control of mRNA amount and quality. The first cycles detecting the co-amplified NATH and b-actin specific bands in ethidium-bromide stained agarose gels were compared. The PCR was performed in 100 ml initial reaction volume including 1 ml cDNA, annealing temperature was 578C, and 10 ml aliquots were removed every second PCR cycle, from cycle 13 until cycle 27. Negative controls omitting reverse transcriptase from cDNA synthesis, and controls omitting template from PCR, were run in parallel in the RT – PCR experiments. To investigate for the presence of alternative NATH splicing variants affecting the putative open reading frame, a nested PCR was performed. Using cDNA of several PTC biopsies and the NPA cell line as template, an initial PCR was performed using the forward primer NATH-F1 and the reverse primer NATH-R4 or -R5 (Table 1), for 10 cycles with annealing at 608C and extension time 4 min. A nested PCR was then performed using the forward primer NATH-Bam-F and one of the following reverse primers; NATH-R2, -R3, -R3B (Table 1, Figure 2a), annealing at 588C and extension for 4 min. TaqMan real-time quantitative RT – PCR experiments Total RNA from the fresh frozen biopsies was subjected to reverse transcription using TaqMan Reverse Transcription reagents (Applied Biosystems). The biopsies were mainly selected from those used in the hybridization series (three non-neoplastic and eight PTC) and including in addition four representative ‘classic’ papillary thyroid carcinomas. PCR primers and the TaqMan probe (Table 1) were designed using the Primer Express software version 1.5 (Applied Biosystems) and the amplicon spanned one exon boundary. Each reverse transcription reaction contained approximately 100 ng total RNA, 5 ml 106RT buffer, 2.5 ml 50 mM Random Hexamers, 1.25 ml Multiscribe Reverse Transcriptase (50 U/ml), 1 ml RNase Inhibitor (20 U/ml), 11 ml 25 mM MgCl2, 10 ml dNTP mixture and nuclease-free water made up to 50 ml. The reaction was incubated at 258C for 10 min, followed by 488C for 30 min, and then 958C for 5 min. Each real-time PCR contained 2.5 ml cDNA, 12.5 ml TaqMan Universal Master Mix (Applied Biosystems), 300 nM sense and anti-sense primers and 200 nM TaqMan probe (Table 1). Each sample was run in triplicate. Cycling parameters were 958C for 10 min, followed by 40 cycles of 958C for 15 s and 608C for 1 min. Standard curves were run on the same plate and the relative standard curve method was used to calculate the relative gene expression, as described (Heid et al., 1996). TaqMan b-actin Detection reagent (Applied Biosystems, prod # 4310855) was included as an endogenous normalization control to adjust for unequal amounts of RNA. Northern blot and multiple mRNA blot Northern blot of approximately 20 mg total RNA purified from PTC tissue specimens, from the PTC cell line NPA, and from the anaplastic thyroid carcinoma cell line ARO (de Nigris et al., 1998) were ethanol-precipitated, washed, and run under standard conditions on a 1% agarose formaldehyde denaturing gel, blotted onto Hybond-N+ nylon membranes (Amersham) and UV-fixed. Prehybridization and hybridization at 668C, and washing at 648C, were performed as described above. The radioactive probe was made using a purified PCR-product generated by the primers


5067

NATH-BamF/NATH-R3 (2457 bp) as template (Table 1), random hexamer primers, and the Klenow fragment of E. coli DNA polymerase I (Promega) (Feinberg and Vogelstein, 1983) with incorporation of [a-32P]dCTP (Amersham). Northern blot experiments were also performed using a radioactive probe derived from the 353 bp PCR-product generated using the primers NATH-F2/NATH-R1 (Table 1). A multi-tissue mRNA blot was purchased (Clontech, cat. # 7775-1) with polyadenylated RNA from 61 different adult human tissues, eight cell lines, and seven human fetal organs. The membrane was prehybridized and hybridized at 668C followed by washing at 648C, as described. This hybridization experiment was performed twice, using the two probes described above.

at 7208C overnight. Cells were permeabilized for 10 min with 0.1% Triton X-100, washed in PBS, then blocked in 0.5% BSA in PBS for 30 min. The samples were incubated for 1 h with the monoclonal anti-Myc primary antibody diluted 1 : 1000 in PBS/BSA. One coverslip was run in parallel omitting the primary antibody. After washing three times in PBS, the coverslips were incubated with a FITZ-conjugated anti-mouse IgG1 secondary antibody diluted 1 : 100 for 30 min, washed three times in PBS, and mounted with SlowFade (Molecular Probes, Eugene, OR, USA). Immunofluorescence was observed under a confocal laser scanning microscope (Leica TCSSP). Images were processed using Leica TCS NT software. Cell proliferation assay and Western immunoblotting

In situ hybridization In situ hybridization experiments were performed using the Innogenex ISH kit for biotin-labeled probes (InnoGenex, San Ramon, CA, USA), using paraffin-embedded tissue sections from representative PTC and non-neoplastic thyroid tissue. Briefly, sections were deparaffinized in xylene for 10 min, followed by rehydration through graded ethanol and deionized water (containing 0.2% RNase Block). Sections were then treated with Proteinase K for 10 min, followed by target retrieval using microwave and boiling for 15 min in citrate-buffer pH 6.0 (with RNase Block), then cooled for 20 min and rinsed in deionized water. Sections were postfixed in 1% formaldehyde/PBS (with RNase Block) for 10 min and rinsed in deionized water. Hybridization was performed in a Hybaid oven, with initial denaturation at 808C for 5 min, then at 378C for 1.5 h, using a mixture of two NATH-specific antisense biotinylated oligonucleotides (NATH-ISH1 and NATH-ISH2, Table 1) both at a concentration of 250 ng/ml hybridization buffer. Negative controls using a sense NATH probe, or no probe, were included in parallel. After hybridization, sections were washed in 26PBS/0.1% Tween-20, then in 16PBS/0.1% Tween-20 at 378C. Sections were then incubated with Power Block reagent for 5 min, followed by mouse monoclonal antiBiotin antibody for 20 min at RT. After a brief wash in 16PBS/0.1% Tween-20, sections were incubated with biotinylated goat anti-mouse secondary antibody for 20 min, washed in PBS/Tween, then incubated with streptavidin-conjugated horseradish peroxidase for 20 min, washed again, incubated with aminoethyl carbazole (AEC) substrate for 10 min, then washed in water, counterstained with Mayer’s hematoxylin, and mounted. Cloning, in vitro expression, and immunostaining RT – PCR was performed using the primer pair NATH-F1/ NATH-R4, followed by a nested PCR with primers NATHBam-F and NATH-Not-R, generating a 2633 bp PCRproduct (Table 1, Figure 2a,c). This fragment was cloned into the BamHI and NotI sites of pcDNA3.1-Myc-His (Invitrogen). Top10 competent E. coli (Invitrogen) were transformed with subsequent screening of bacterial colonies by PCR using the polylinker primers T7 and BGH reverse (Table 1). Plasmids harboring inserts of correct size (2800 bp) were sequenceverified using relevant primers (Table 1, Figure 2a). For immunocytochemistry, Cos-7 cells were grown on coverslips and transfected with the NATH/pcDNA3.1-MycHis plasmid, using Fugene-6 according to the descriptions (Roche Diagnostics, Basel, Switzerland). Forty-eight hours post-transfection, the cells were washed with PBS, fixed in 4% formaldehyde in PBS for 20 min, then stored in methanol

293 cells (6 million), and NPA cells (3 million), were plated on 9-cm dishes in high glucose DMEM containing 10% FCS the day before transfection. The transfection mixture contained 75 ml Lipofectamine2000, 30 mg pcDNA3.1NATH-Myc-His, or empty vector as a negative control, and was prepared according to the manufacturers recommendations (Invitrogen). Three mg MSCV LacZ was added as an internal control of transfection efficacy. Twenty-four hours post transfection the cells were trypsinized, and replated on 25 dishes at a density of 50 000 cells per 3-cm dish. From each individual transfection, trypsinized cells from two dishes were counted daily by an Improved Neubacher hematocytometer (Sigma) starting 48 h post transfection. Trypan blue was added at a final concentration of 0.2%. At each time point, the cells from each dish were counted 30 times. The proliferation assay was performed four times for both NPA and 293 cells. For Western immunoblot experiments, 48 h post-transfection cells were washed with PBS and harvested in extraction buffer (1% NP-40, 20 mM Tris-HCl pH 8.0, 137 mM NaCl, 1% glycerol, 2 mM EDTA, 50 mM NaF, 1 mM phenylmethyl-sulphonyl-fluoride (PMSF), 1 mM Na3VO4), sonicated, and centrifuged for 15 min at 13 000 r.p.m. Western immunoblot experiments were performed using protein extracts from cell lysates of transfected and untransfected cells (293 and NPA). Approximately 40 mg (293 cells) or 100 mg (NPA cells) protein were separated on 10% SDSpolyacrylamide gels and electroblotted (Fluge et al., 2000). The blocked membranes were incubated overnight at 48C with the monoclonal anti-Myc-HRP antibody (Invitrogen) diluted 1 : 5000 in 1% dry milk. Negative controls omitting the primary antibody were run in parallel. After several washes in PBS-Tween the membranes were subjected to enhanced chemiluminescence (ECL) (Amersham).

Acknowledgments The authors thank Ms Anne Marstad and Ms Benedicte Rasmussen for technical assistance. Drs Oddbjørn Straume and Hans Geir Eiken are thanked for fruitful discussions and advice. This work was supported by grants from The Norwegian Cancer Society (Ø Fluge, JE Varhaug), The Norwegian Research Council (JR Lillehaug), and the Cancer Locus, Faculty of Medicine, University of Bergen (JR Lillehaug, JE Varhaug).

Accession number GenBank accession for NATH mRNA: AJ314788. Oncogene


5068

References Bernard K, Auphan N, Granjeaud S, Victorero G, SchmittVerhulst AM, Jordan BR and Nguyen C. (1996). Nucleic Acids Res., 24, 1435 – 1442. Blatch GL and Lassle M. (1999). Bioessays, 21, 932 – 939. Boyes J, Byfield P, Nakatani Y and Ogryzko V. (1998). Nature, 396, 594 – 598. Bunone G, Vigneri P, Mariani L, Buto S, Collini P, Pilotti S, Pierotti MA and Bongarzone I. (1999). Am J. Pathol., 155, 1967 – 1976. Chomczynski P and Sacchi N. (1987). Anal. Biochem., 162, 156 – 159. de Nigris F, Visconti R, Cerutti J, Califano D, Mineo A, Santoro M, Santelli G and Fusco A. (1998). Cancer Res., 58, 4745 – 4751. Di Renzo MF, Olivero M, Ferro S, Prat M, Bongarzone I, Pilotti S, Belfiore A, Costantino A, Vigneri R and Pierotti MA. (1992). Oncogene, 7, 2549 – 2553. Driessen HP, de Jong WW, Tesser GI and Bloemendal H. (1985). CRC Crit. Rev. Biochem., 18, 281 – 325. Fagin JA, Matsuo K, Karmakar A, Chen DL, Tang SH and Koeffler HP. (1993). J. Clin. Invest., 91, 179 – 184. Feinberg AP and Vogelstein B. (1983). Anal. Biochem., 132, 6 – 13. Fluge Ø, Akslen LA, Haugen DR, Varhaug JE and Lillehaug JR. (2000). Int. J. Cancer, 87, 763 – 770. Gendron RL, Adams LC and Paradis H. (2000). Dev. Dyn., 218, 300 – 315. Gounalaki N, Tzamarias D and Vlassi M. (2000). FEBS Lett., 473, 37 – 41. Haugen DRF, Akslen LA, Varhaug JE and Lillehaug JR. (1996). Cancer Res., 56, 1184 – 1188. Heid CA, Stevens J, Livak KJ and Williams PM. (1996). Genome Res., 6, 986 – 994. Huang Y, Prasad M, Lemon WJ, Hampel H, Wright FA, Kornacker K, LiVolsi V, Frankel W, Kloos RT, Eng C, Pellegata NS and de la Chapelle A. (2001). Proc. Natl. Acad. Sci. USA, 98, 15044 – 15049. Kurozumi K, Nakao K, Nishida T, Nakahara M, Ogino N and Tsujimoto M. (1998). World J. Surg., 22, 1237 – 1242. Lamb JR, Michaud WA, Sikorski RS and Hieter PA. (1994). EMBO J., 13, 4321 – 4328. Lee FJ, Lin LW and Smith JA. (1990). J. Biol. Chem., 265, 11576 – 11580. Liu L, Scolnick DM, Trievel RC, Zhang HB, Marmorstein R, Halazonetis TD and Berger SL. (1999). Mol. Cell. Biol., 19, 1202 – 1209. Mahlknecht U and Hoelzer D. (2000). Mol. Med., 6, 623 – 644. Mullen JR, Kayne PS, Moerschell RP, Tsunasawa S, Gribskov M, Colavito-Shepanski M, Grunstein M, Sherman F and Sternglanz R. (1989). EMBO J., 8, 2067 – 2075. Nakatsu Y, Asahina H, Citterio E, Rademakers S, Vermeulen W, Kamiuchi S, Yeo JP, Khaw MC, Saijo M, Kodo N, Matsuda T, Hoeijmakers JH and Tanaka K. (2000). J. Biol. Chem., 275, 34931 – 34937. Neuwald AF and Landsman D. (1997). Trends Biochem. Sci., 22, 154 – 155.

Oncogene

Nishida T, Katayama S and Tsujimoto M. (2002). Am. J. Surg., 183, 80 – 86. Park EC and Szostak JW. (1992). EMBO J., 11, 2087 – 2093. Perou CM, Sorlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees CA, Pollack JR, Ross DT, Johnsen H, Akslen LA, Fluge Ø, Pergamenschikov A, Williams C, Zhu SX, Lonning PE, Borresen-Dale AL, Brown PO and Botstein D. (2000). Nature, 406, 747 – 752. Pilotti S, Collini P, Manzari A, Marubini E and Rilke F. (1995). Semin. Diagn. Pathol., 12, 249 – 255. Polevoda B, Norbeck J, Takakura H, Blomberg A and Sherman F. (1999). EMBO J., 18, 6155 – 6168. Powell DJ, Russell J, Nibu K, Li G, Rhee E, Liao M, Goldstein M, Keane WM, Santoro M, Fusco A and Rothstein JL. (1998). Cancer Res., 58, 5523 – 5528. Resnick MB, Schacter P, Finkelstein Y, Kellner Y and Cohen O. (1998). Mod. Pathol., 11, 735 – 739. Ryu S, Jimi S and Takebayashi S. (1997). Cancer Lett., 121, 189 – 193. Santoro M, Carlomagno F, Hay ID, Herrmann MA, Grieco M, Melillo R, Pierotti MA, Bongarzone I, Della Porta G and Berger N. (1992). J. Clin. Invest., 89, 1517 – 1522. Santoro M, Chiappetta G, Cerrato A, Salvatore D, Zhang L, Manzo G, Picone A, Portella G, Santelli G, Vecchio G and Fusco A. (1996). Oncogene, 12, 1821 – 1826. Santoro M, Melillo RM, Carlomagno F, Visconti R, De Vita G, Salvatore G, Fusco A and Vecchio G. (1999). Biochimie, 81, 397 – 402. Sartorelli V, Puri PL, Hamamori Y, Ogryzko V, Chung G, Nakatani Y, Wang JY and Kedes L. (1999). Mol. Cell, 4, 725 – 734. Scheufler C, Brinker A, Bourenkov G, Pegoraro S, Moroder L, Bartunik H, Hartl FU and Moarefi I. (2000). Cell, 101, 199 – 210. Scheumman GF, Hoang-Vu C, Cetin Y, Gimm O, Behrends J, von Wasielewski R, Georgii A, Birchmeier W, von Zur M and Dralle H. (1995). J. Clin. Endocrinol. Metab., 80, 2168 – 2172. Shuja S, Cai J, Iacobuzio-Donahue C, Zacks J, Beazley RM, Kasznica JM, O’Hara CJ, Heimann R and Murnane MJ. (1999). Thyroid, 9, 569 – 577. Sugiura N, Patel RG and Corriveau RA. (2001). J. Biol. Chem., 276, 14257 – 14263. Takano T, Matsuzuka F, Miyauchi A, Yokozawa T, Liu G, Morita S, Kuma K and Amino N. (1998). Br. J. Cancer, 78, 221 – 224. Tribioli C, Mancini M, Plassart E, Bione S, Rivella S, Sala C, Torri G and Toniolo D. (1994). Hum. Mol. Genet., 3, 1061 – 1067. Trovato M, Villari D, Bartolone L, Spinella S, Simone A, Violi MA, Trimarchi F, Batolo D and Benvenga S. (1998). Thyroid, 8, 125 – 131. von Wasielewski R, Rhein A, Werner M, Scheumann GF, Dralle H, Potter E, Brabant G and Georgii A. (1997). Cancer Res., 57, 2501 – 2507.