DNA gains targeting the 3q chromosome are common in head and neck squamous cell carcinomas, as well as in lung, ovarian, and cervical cancer. Several candidate oncogenes located on 3q were proposed, i.e., PIK3CA, p63, and eIF-5A2. However, none of these genes was found included in a narrow high-level amplification. Recently, microarray-based comparative genomic hybridization (array CGH) was developed for high-resolution screening of deletions and amplifications in tumor genomes. In this study, by microarray-based comparative genomic hybridization, we found a narrow 3q25.3 high-level amplification in a head and neck cancer cell line. We precisely delimited the 3-Mb length-amplified segment by semiquantitative PCR and measured the transcriptional level of every gene (RefSeq full-length mRNA) located inside this segment by cDNA microarray and quantitative reverse transcription-PCR. Four genes were overexpressed in three head and neck cancer cell lines with increased DNA copy number, compared with a control tongue cell line. We extended the transcriptional analysis of these four genes to 20 head and neck squamous cell carcinomas. Only one gene, cyclin L (ania-6a), is commonly overexpressed in primary tumors compared with corresponding normal tissues. This cyclin was previously pinpointed as a candidate for a role in promoting cell cycle entry. Thus, we propose cyclin L as a candidate oncogene in head and neck cancer.

Aneuploidy has been commonly observed in human cancer for several decades. However, solid tumors usually present very complex karyotypes, hampering the complete identification of chromosomal rearrangements by conventional cytogenetics. Since its introduction (1), CGH4 has facilitated the depiction of previously undefined recurrent and specific chromosomal imbalances in most of tumor types or stages. Nevertheless, by its low spatial resolution (5 to 15 Mb), CGH is not suitable to detect or delineate small deletions and amplifications, rendering very difficult the route toward oncogene or tumor suppressor gene identification.

FISH and loss of heterozygosity analysis can efficiently complement CGH, but these techniques require the prior selection of specific probes or markers at limited chromosomal loci. Recently, a new high-resolution screening method, microarray-based CGH, was developed to reliably detect genome-wide deletions and amplifications (2, 3, 4, 5). This new tool represents a promising way toward the identification of new candidate genes involved in tumor genesis and progression.

Previous CGH studies on head and neck cancer (6, 7, 8, 9) showed the prevalence of gains or amplifications targeting the long arm of chromosome 3 (in >50% of cases) and the 11q13 locus (>40%). 11q13 harbors one definitive critical oncogene, cyclin D1 (CCND1), and several additional coamplified genes like INT2 and HST(10, 11).

3q overrepresentations, which are present in a large majority of head and neck tumors, are as well frequent in tumors from lung, esophagus, ovary, and uterine cervix (for a review, see Ref. 12). On this chromosomal arm, several candidate oncogenes were proposed, i.e., (a) PIK3CA, at 3q26, in ovarian, cervical, and oral cancer (8, 13, 14), (b) p63, at 3q28, in oral and lung squamous cell carcinoma (15), and (c) eIF-5A2, at 3q26, in ovarian cancer (16). However, no restricted high level 3q amplification was fully characterized, and none of these or other candidate oncogenes was described yet as definitive critical targets of 3q amplifications.

In this study, we combined genome and transcriptional analysis to identify four genes at 3q25.3, specifically amplified and overexpressed in a head and neck cancer cell line. By quantitative RT-PCR, we showed that only one of these genes, cyclin L, is significantly overexpressed in a population of 20 primary tumors, compared with their normal counterparts. We propose, in accordance with structural and functional data, that cyclin L is a candidate oncogene in head and neck cancer. Our results illustrate how a combination of genome and transcriptome approaches using DNA microarrays can streamline the identification of new genes involved in cancer pathogenesis.

Biological Samples.

The head and neck cancer cell lines CAL 27 (CRL-2095), FaDu (HTB-43), and the tongue fibroblast cell line Hs 677.Tg (CRL-7408) are available at the American Type Culture Collection. The head and neck cancer cell line CAL 33 (ACC 447) can be obtained at the German Collection of Microorganisms and Cell Cultures (DSMZ). CAL 33 and FaDu are derived from moderately differentiated squamous cell carcinomas from tongue and hypopharynx, respectively (17, 18). CAL 27 was established from a poorly differentiated squamous cell carcinoma of the tongue and forms well-differentiated tumors when injected in nude mice (17). DNA and metaphase chromosomes were prepared from these cell lines using standard protocols. Twenty HNSCCs without distant metastasis (M0) from patients undergoing surgery as a primary treatment, without previous radiation or chemotherapy, were selected. They are composed of two distinct subgroups: (a) 9 well-differentiated tumors without regional lymph node involvement (N0); and (b) 11 tumors with variable N status and poorly differentiated (n = 9) or undifferentiated (n = 2). Normal mucosas of the upper aerodigestive tract, corresponding usually to uvula, were obtained at ≥5 cm from the tumor. Resected samples were immediately frozen and stored in liquid nitrogen. Specimens were obtained after patient consent and approval of the regional ethics committee.

Clones Selection.

Eighty-five clones (BACs or PACs) from the RPCI-11, RPCI-4, and RPCI-5 libraries were purchased from the Roswell Park Cancer Institute. We performed a first selection of clones mapped on chromosome 3 according to their STS content (provided by the San Antonio Genome Center; working draft data were not available at this time).5 We obtained a low coverage of the whole chromosome 3 and a higher coverage for the 3q26-qter region. These first selected clones were validated by PCR (for STS content) or by FISH mapping. Second, we increased the 3q25-qter coverage using the clone collection published by Cheung et al.(19). Third, we used the draft sequence to include some clones containing 3q candidate (onco)genes or mapped at 3q25-qter loci previously poorly represented. We finally synthesized mapping information about these clones using the April 2001 version of the Human Genome Working Draft from the UCSC (Table 1).6 In summary, 19 of 85 selected clones are issued from the collection described by Cheung et al.(19), and we chose the 66 remaining ones.

Preparation of the BAC/PAC Array.

Cloned DNA was extracted from bacterial cultures (500 ml) using Nucleobond AX500 cartridges, following the instructions of the manufacturer (Macherey-Nagel). After ethanol precipitation, DNA was resuspended in sterile water and quantified using Picogreen dsDNA quantification kit (Molecular Probes) on a Bio-Rad Fluoromark fluorometer. Every clone was then fragmented by sonification (we obtained fragments of 1–10 kb), ethanol precipitated, and resuspended at an estimated final concentration of 1 μg/μl in 75% formamide, 25% water. Cloned DNA was then spotted on CMT-GAPS-coated slides (Corning) in six replicates using a custom built arrayer. The slides were allowed to dry during at least 2 h and then heated 3 h at 85°C. The slides were then stored at room temperature, in the dark, in argon.

Array CGH.

Genomic DNA was labeled by nick-translation in a 50-μl reaction mix containing: 2 μg of DNA; 0.05 mm dATP and dGTP; 0.01 mm dCTP and dTTP; 0.04 mm Cy3- or Cy5-dUTP and dCTP; 1 μl of Kornberg polymerase (5 units/μl, Roche Diagnostics); 2–20 ng of DNase I; 10 mm β-mercaptoethanol; 5 mm MgCl2; 0.1% BSA; and 50 mm Tris-HCl (pH 8.0). The reaction product size was 0.1–1kb. The probe was then purified using a PCR purification kit (Qiagen). Test and control DNA (2 μg of each probe) were ethanol precipitated in the presence of 100 μg of Cot-1 DNA (Life Technologies, Inc.) and 20 μg of salmon sperm DNA. The pellet was dried and resuspended in 50 μl of 50% formamide, 2× SSC, 16% dextran sulfate, and 2% SDS (pH 7.0). The probes were denatured at 76°C during 5 min, and repetitive sequences were blocked by preannealing at 37°C during 3 h. An arrayed slide was blocked during 20 min at 42°C in 1% BSA, 0.2% SDS, 5× SSC, rinsed in 2× SSC, and dehydrated in ethanol. A 2.4-cm2 open hybridization chamber (Hybaid) was fixed on it, then the 50 μl of preannealed mix was applied and hybridized at 37°C on a slowly rocking table during 16–20 h. After hybridization, slides were rinsed and washed in 50% formamide, 2× SSC at 45°C during 15 min, then in SDS 0.1%, 0.1× SSC at 60°C during 15 min, dehydrated in ethanol, and air dried.

Image Capture and Analysis.

The array was scanned using a ScanArray 3000 machine (Packard Bioscience), and the signals obtained in the Cy3 and Cy5 channels were quantified using ImaGene 4.0 software. We excluded from analysis the clones that displayed signal intensities lower than twice the local background values for more than two of six replicates. After subtraction of local background, we calculated for each clone the median of the ratios between the two signals. The final experiment is composed of two independent hybridizations with Cy3 and Cy5 swapping. Each final ratio is the geometric mean of the two individual ones. To determine the genomically normal ratios, we calculated iteratively the SD of the Log2 ratios by including one by one the points with increasing distance to the modal value (beginning with the 20% modal values). We plotted the resulting SD against the successive iterations. SD first increases slowly with a regular slope for the genomically normal values, then suddenly increases and becomes irregular when reaching the unbalanced values. This slope abrupt change can be detected by the sudden increase of the SD second derivative curve. Then, to assign normal copy number to the array data, individual ratios were normalized to 1 for the mean value of the previously determined genomically normal population. We used the 3 SD upper and lower thresholds (shown in Fig. 1 B) to determine gains and losses.

Chromosome CGH and FISH.

Chromosome CGH was performed as described previously (8, 20). For FISH analysis, the clone 25 was labeled by nick-translation with digoxigenine-11-dUTP and ethanol-precipitated in the presence of Cot-1 fraction of normal DNA (Life Technologies, Inc.) and salmon sperm DNA (Sigma Chemical Co.). The probe was then hybridized at 37°C to CAL 27 metaphase chromosomes for 24 h and detected using mouse antidigoxin (Sigma Chemical Co.) and goat antimouse conjugated to Cy3 (Jackson ImmunoResearch Laboratory) antibodies. Chromosomes were counterstained with 4,6-diamidino-2-phenylindole.

Semiquantitative Genomic PCR.

PCR amplification for the relative evaluation of gene copy number was performed using decreasing amounts (∼30, 10, and 3 ng) of genomic DNA in 20-μl reactions containing 0.5 μm of each primer, 0.2 mm dNTP, 2 mm MgCl2, and 1 unit of Taq DNA polymerase (Perkin-Elmer). Oligonucleotides used as primers are listed on Table 2. For each primer pair, 25 cycles of amplification were performed with denaturation at 95°C for 30 s, annealing at 60°C for 30 s, and elongation at 72°C for 30 s. Five μl from each sample were loaded on 1% agarose gels, and DNA was detected with ethidium bromide. After image acquisition with a Typhoon scanner (Molecular Dynamics), specific amplified bands were quantified with the TotalLab v1.1 software.

Expression Microarray.

We obtained Human 1 cDNA Microarrays from Agilent Technologies. The array is composed of >12,000 cDNAs representing 8,363 different UniGene clusters. Each UniGene cluster was localized on the genome thanks to the April 2001 version of the Human Genome Working Draft from the UCSC.6 Total RNA was extracted from the CAL 27, CAL 33, and Hs 677.Tg cell lines using RNA-Solv Reagent (Omega Biotek), and RNA integrity was assessed on an agarose gel. Reverse transcription with direct incorporation of Cy3- or Cy5-dCTP was performed using 20 μg of RNA, as indicated by the manufacturer. Labeled cDNAs were cohybridized under a coverslip on the array for 16 h at 65°C. The slide was washed in 0.5× SSC, 0.01% SDS for 5 min, in 0.06× SSC for 2 min, then dehydrated in ethanol and air-dried. Scanning and quantification were performed as described for array CGH. For each spot, the intensities were corrected by subtracting the local background, and the ratios (CAL 27/Hs 677.Tg or CAL 27/CAL 33) were averaged for each UniGene cluster and normalized on the average value of ratios for each chromosome.

Quantitative RT-PCR.

Total RNAs were extracted using RNeasy kit (Qiagen, Les Ulis). First-strand cDNA synthesis was performed on 1 μg of RNA in 20 μl of a reaction mixture with 0.3 μg of hexanucleotide used as primer (Boerhinger-Mannheim), 200 units of SuperScript II RNase H reverse transcriptase (Life Technologies, Inc.), and 40 units of RNasin (Promega). The mixture was incubated at 42°C for 1 h and then heated to 70°C for 15 min to inactivate reverse transcriptase. A total of 0.2 μl of the reverse transcriptase products was used for PCR amplification on the Roche Lightcycler machine using the standard Sybr GreenI PCR protocol provided by the manufacturer. Oligonucleotides used as primers (annealing at 60°C) are listed on Table 2. Levels of expression were measured at least three times independently for each gene and normalized to the expression level of GAPDH for cell lines and to the averaged expression levels of GAPDH and UBB for primary tumors and normal tissues.

A Narrow 3q25.3 Amplification Detected by Array CGH.

We constructed a BAC/PAC array composed of 85 clones mapped on chromosome 3 (Table 1), with a particularly high coverage for the 3q25-qter region (one clone every 0.8 Mb). We compared chromosome and array CGH results for a human head and neck cancer cell line, CAL 27. The general evolution is similar for the two chromosome 3 profiles, showing a loss of the short arm (Fig. 1, A and B). In contrast, the results differ for the long arm. In the array CGH experiment, two adjacent clones at 3q25.31 (clones 25 and 26) display ratios higher than 2 and are surrounded by clones in the normal range (at steady level around 1) when the chromosome CGH profile shows variations within the normal range (Fig. 1, A–C). We interpret this narrow peak as DNA amplification between positions 166.8 and 173.0 Mb (largest candidate region of amplification between the two internal ends of the first excluded clones 24 and 27). We noted that clone 26 may be near the distal end of the amplicon because its ratio is lower than the clone 25 one. FISH of clone 25 on CAL 27 metaphase chromosomes confirmed the presence of numerous copies of this sequence (Fig. 1, D and E).

Delineation of the 3q25.3 Amplicon by Gene-specific Semiquantitative PCR.

The 6.2-Mb length largest candidate region of amplification defined by array CGH contains 15 RefSeq genes but is not fully sequenced in the UCSC April 2001 version of working draft assembly (21).6 Thus we downloaded the corresponding sequence published by Venter et al.(22).7 Annotation of this sequence using Genome Cryptographer led to the identification of one additional RefSeq gene, cyclin L (ania-6a), in a small gap of the public sequence draft assembly (169.59–169.64 Mb).8

We refined the mapping of the amplification by gene-specific semiquantitative PCR. A first screening delimited a smaller candidate interval of amplification between 167.3 and 170.9 Mb (Fig. 2,A and B). We then analyzed every gene located in this interval (every RefSeq mRNA, except SET that was mapped at multiple loci, and two additional full-length mRNAs, DKFZp434J214 and LOC115396; Fig. 2,C). We precisely mapped the borders of the core amplicon at 167.7–168 Mb and 170.2–170.9 Mb (Fig. 2 D). The 3.2-Mb length-amplified segment contains eight genes: ACATN, GMPS, KCNAB1, SSR3, DKFZp434J214, cyclin L, PTX-3, and LOC115396.

Transcriptional Consequences of the 3q25.3 Amplification.

To investigate the transcriptional consequences of this amplification, we analyzed gene expression levels in CAL 27 compared with the control tongue cell line Hs 677.Tg using a cDNA microarray. From 8363 Unigene clusters represented at least once on this array, 438 were mapped on chromosome 3. Interestingly, expression levels of most chromosome 3 genes seem to be similar in both cell lines with 6% of the genes overexpressed in CAL 27 (17 genes with ratios higher than 2; 6 with ratios higher than 4; Fig. 3,A). Of these, the three most overexpressed genes, ACATN (ratio: 4.9), GMPS (4.8), and cyclin L (10.7), are mapped in the core amplicon delineated by genomic PCR (Fig. 3, A and B). Two other genes mapped in the core amplicon and present on the chip were not expressed (KCNAB1) or not overexpressed (PTX-3; Fig. 3,B). These three genes seem to be overexpressed consequently to the 3q25.3 amplification. This was confirmed by a comparison between two tongue cancer cell lines, CAL 27 and CAL 33, carrying a 3q25 high-level amplification and a 3q25 low-level gain, respectively (Fig. 3 B; Ref. 8): ACATN, GMPS, and cyclin L are the three only overexpressed genes located in the core amplicon.

To address the relationship between gene copy number and expression level for every gene in the core amplicon, we measured by real-time RT-PCR the expression levels of ACATN, GMPS, cyclin L, and the three amplified genes not included on the chip, DKFZp434J214, SSR3, LOC115396, in Hs 677.Tg, CAL 27, CAL 33, and FaDu, the last two head and neck cancer cell lines carrying a low-level 3q gain (8). The six genes show a higher expression in CAL 27, which carries the 3q25 high-level amplification, than in other cancer cell lines (Fig. 3,C): they remain candidate to be activated by DNA amplification. However, in the three cancer cell lines, expression levels of the SSR3 and LOC115396 genes are not significantly increased compared with Hs 677.Tg, whereas ACATN, GMPS, cyclin L, and DKFZp434J214 are systematically overexpressed (Fig. 3 C).

Cyclin L is Overexpressed in Well-Differentiated HNSCC.

We investigated the expression of these four genes in 20 primary tumors from head and neck compared with matched normal tissues (Table 3). Cyclin L, contrary to ACATN, GMPS, and DKFZp434J214, is frequently overexpressed (8 of 20) and rarely underexpressed (1 of 20). Consistently, gene expression levels are significantly different between tumor and normal tissue samples only for the cyclin L gene (Wilcoxon signed rank test; P < 0.05). The expression level of cyclin L in tumors stratified by differentiation degree is presented in Fig. 3 D. Most of the well-differentiated tumors show cyclin L RNA increase. Interestingly, the only well-differentiated tumor that does not display a cyclin L increase does not present 3q25 gain. In contrast, no clear tendency is visible in poorly differentiated tumors. Accordingly, the cyclin L expression is significantly increased only for the well-differentiated tumor population (compared with normal tissues; Wilcoxon signed rank test; P < 0.05).

Using array CGH, we found in one head and neck cancer cell line, CAL 27, a 3q25.3 amplification that was previously not detected using chromosome CGH. Availability of human draft sequence made it possible to recapitulate the genes contained in this 3q25.3 amplified chromosomal segment. Our results illustrate that array CGH, combined to genome sequencing data, represents a very efficient high-resolution genome-screening method to revisit cytogenetic information from chromosome CGH, toward the identification of novel genes dysregulated in cancer.

The ability of array CGH to refine the mapping of DNA copy number changes in tumors has been readily recognized since its proof of principle in 1998 (3). Snijders et al.(5) chose to construct a microarray composed of 2400 clones for whole-genome exploration. Among them, 24 clones are located at 3q25-qter, resulting in a resolution of one clone every 2.3 Mb for this common region of DNA gains or amplifications in human solid tumors. We followed an alternative strategy: the construction of a focused chip dedicated to the 3q25-qter region, with a resolution of one clone every 0.8 Mb.

To confirm and complement array CGH results, we applied gene-specific semiquantitative PCR: eight genes are amplified at 3q25.3 in CAL 27. Using cDNA microarrays and/or quantitative RT-PCR, we showed that six genes seem to be overexpressed consequently to increased DNA copy number in the head and neck cancer cell lines (CAL 27, CAL 33, and FaDu): DKFZp434J214, LOC115396, SSR3, ACATN, GMPS, and cyclin L. These six genes remain good candidates to be submitted to gene dosage effect in this particular cell type. We found that four genes were overexpressed in CAL 27 compared with Hs 677.Tg: DKFZp434J214, ACATN, GMPS, and cyclin L. We chose to focalize our efforts in these last candidates.

Functionally, very little is known about the potential roles of ACATN and DKFZp434J214 in carcinogenesis. The ACATN gene product is an acetyl-CoA transporter required for the formation of O-acetylated gangliosides (23). The DKFZp434J214 gene encodes a hypothetical protein, which contains a poly(ADP-ribose) polymerase domain found in proteins involved in telomere maintenance (data available on-line at LocusID: 25976).9

In contrast, GMPS, which was previously identified as the partner gene of MML in a leukemia-associated translocation (24), codes for a guanosine 5′-monophosphate synthetase that exhibits elevated level of activity in rapidly proliferating cells such as neoplastic and regenerating tissues (25). Furthermore, GMPS belongs to a group of 70 genes with expression variations significantly associated with poor prognosis in breast primary tumors (26).

Initially, cyclin L was described as coding for two alternatively spliced mRNAs induced by dopamine and glutamate in adult rat striatum (27). Both transcripts show widespread rat and human tissue distribution and encode proteins containing a NH2-terminal cyclin box characteristic of the cyclins involved in cell cycle regulation. The longer protein interacts with the p110 PITSLRE cyclin-dependent kinase and is associated with different components of the RNA elongation/splicing machinery in nuclear speckles. The shorter one is localized in both nucleus and cytoplasm and has no partner identified yet.

We extended the expression analysis of these four genes to a larger population of 20 head and neck primary tumors compared with their normal counterparts. Only one gene, cyclin L, was found significantly overexpressed in primary tumors. Our amplicon mapping and expression data pinpoint cyclin L as an oncogene in head and neck cancer.

Cyclin L was previously demonstrated as an immediate early gene induced by several growth factors such as epidermal growth factor (27). Its 3′ untranslated mRNA region contains several AU-rich elements promoting rapid mRNA degradation, which are also present in the proto-oncogenes c-fos and c-myc(27, 28). Immediate early genes are inducible in both mitotic and postmitotic cells (29). In G0 cells, they regulate gene cascades enabling G1 progression (29). Accordingly, by large-scale microarray experiments on quiescent human fibroblasts driven to proliferate by serum stimulation (30), cyclin L was found to be rapidly and transiently induced (with ∼3-fold RNA increase) in a time window similar to c-fos (13-fold increase for comparison; data available on-line).10 Structural data and expression timing led the authors to propose for this cyclin (named EST AA016305) “a role in promoting the exit from G0(30).”

We found cyclin L overexpression in well-differentiated but not in poorly differentiated primary tumors. According to functional data, cyclin L could be induced in quiescent cells to promote cell cycle reentry. Our working hypothesis is that cyclin L overexpression could increase the pool of proliferating cells in well-differentiated tumors. Inversely, in poorly differentiated tumors, most cells would undergo continuous cell division cycles, rendering cyclin L overexpression unnecessary.

In summary: (a) 3q25 gains were previously found in >50% of HNSCCs by chromosome CGH (6, 7, 8, 9); (b) cyclin L is included in a narrow high-level 3q amplification in a head and neck cancer cell line; (c) it is overexpressed consequently to increased gene copy number; (d) cyclin L is overexpressed in differentiated primary tumors; and (e) this gene encodes a putative key regulator of the G0 to G1 transition (30). All these data together lead us to propose cyclin L as a candidate oncogene in head and neck cancer.

Chromosome CGH is a cytogenetics method with limited spatial resolution (1, 20). This is exemplified here by the detection, using array CGH, of a 3q25.3 amplification not found with chromosome CGH. Along this line, we are currently screening by array CGH a large set of primary tumors for high-level amplifications centered on 3q25. Recurrence of such amplifications as well as functional characterization of cyclin L in cellular models will definitively prove its implication as an oncogene in HNSCC.

In this study, by using DNA microarrays for simultaneous genome and transcriptome analyses, we identified one new gene, located at 3q25.3, amplified and overexpressed in head and neck cancer. A similar approach for identification of amplification target genes has been previously performed by using cDNA microarrays (31). This strategy, applied for systematic screening of DNA amplifications and their consequences in tumor genomes, should lead very rapidly to the identification of novel genes critical for cancer development.

Fig. 1.

Detection of a narrow high-level 3q25.3 amplification in the head and neck cancer cell line CAL 27 by array CGH. A, CAL 27 chromosome 3 profile obtained by chromosome CGH. The profile shows a loss on the short arm of chromosome 3 (3p) and a normal representation of the long arm (3q). B, CAL 27 chromosome 3 profile obtained by array CGH. The profile displays, in accordance with chromosome CGH profile, the 3p underrepresentation, but pinpoints in addition a narrow amplification between 166.8 and 173.0 Mb (clones 25 and 26), and a low-level gain at the centromeric part of 3q (clones 12 and 13). For thresholds used, see “Materials and Methods.” C, examples of arrayed spots cohybridized with CAL 27 (first row in green) and control DNA (second row in red). The composite image results from the merging of the two previous ones (third row). Examples from clones 1–11 are red in the composite image showing an underrepresentation of the short arm. Examples from clones 14–24 and 27–85 are yellow showing a balanced state. Clones 25 and 26 are bright green and are contained in the amplification. D and E, FISH on CAL 27 chromosomes. D, hybridization signals of the clone 25 (in red) on a metaphase spread were observed in four separated chromosomal regions in this nearly diploid cell line (average number of chromosomes/cell: 43; Ref. 17). Two signals (marker I) are located on an isochromosome 3q known to be present in this cell line (8). In addition, two different marker chromosomes (II and III) depict two large and/or split signals, suggesting the presence of local duplications of the 3q25.3 sequence. E, examples of markers I, II, and III found in other spreads.

Fig. 1.

Detection of a narrow high-level 3q25.3 amplification in the head and neck cancer cell line CAL 27 by array CGH. A, CAL 27 chromosome 3 profile obtained by chromosome CGH. The profile shows a loss on the short arm of chromosome 3 (3p) and a normal representation of the long arm (3q). B, CAL 27 chromosome 3 profile obtained by array CGH. The profile displays, in accordance with chromosome CGH profile, the 3p underrepresentation, but pinpoints in addition a narrow amplification between 166.8 and 173.0 Mb (clones 25 and 26), and a low-level gain at the centromeric part of 3q (clones 12 and 13). For thresholds used, see “Materials and Methods.” C, examples of arrayed spots cohybridized with CAL 27 (first row in green) and control DNA (second row in red). The composite image results from the merging of the two previous ones (third row). Examples from clones 1–11 are red in the composite image showing an underrepresentation of the short arm. Examples from clones 14–24 and 27–85 are yellow showing a balanced state. Clones 25 and 26 are bright green and are contained in the amplification. D and E, FISH on CAL 27 chromosomes. D, hybridization signals of the clone 25 (in red) on a metaphase spread were observed in four separated chromosomal regions in this nearly diploid cell line (average number of chromosomes/cell: 43; Ref. 17). Two signals (marker I) are located on an isochromosome 3q known to be present in this cell line (8). In addition, two different marker chromosomes (II and III) depict two large and/or split signals, suggesting the presence of local duplications of the 3q25.3 sequence. E, examples of markers I, II, and III found in other spreads.

Close modal
Fig. 2.

Fine mapping of the CAL 27 3q25.3 amplicon using gene-specific semiquantitative PCR. A and C, scanned gel image of PCR products using three dilutions of DNA input from CAL 27 (left) and a control (normal female, right). Primers specific for genes located within the 3q-amplified region detected by array CGH were used. Genes are named according to their position rank on the chromosome 3, refer to Table 2. For all of the lanes, input increase results in higher amount of PCR product, showing that the reaction does not reach the plateau phase at 10 ng. For control DNA, the equivalent amounts of PCR products obtained for each gene demonstrate that all of the primer pairs present similar efficiencies. B and D, quantification of PCR products from three independent experiments with 10 ng of DNA input for CAL 27 (———; mean ± SE) and control DNA (— — —). A, for CAL 27, with 3 ng of input, only bands 3, 6, and 9 are clearly visible, showing increased copy number for the three corresponding genes. The quantification of PCR products for the intermediate dilution (B) define a candidate interval from 167.0–170.9 Mb. C, similarly, exhaustive analysis of the genes mapped in the candidate interval shows that all genes from 3–10 present increased copy number. The quantification of PCR products for the intermediate dilution (D) refines the core amplicon mapping from 167.7–168 Mb to 170.2–170.9 Mb. ▪ represent positions of clones from the BAC/PAC array included in the amplification; □ positions of clones excluded from amplification; numbers under squares refer to clone numbers in Table 1. Note that the gene 19 is mapped in the clone 27 and excluded from the amplification by array CGH (Fig. 1 B).

Fig. 2.

Fine mapping of the CAL 27 3q25.3 amplicon using gene-specific semiquantitative PCR. A and C, scanned gel image of PCR products using three dilutions of DNA input from CAL 27 (left) and a control (normal female, right). Primers specific for genes located within the 3q-amplified region detected by array CGH were used. Genes are named according to their position rank on the chromosome 3, refer to Table 2. For all of the lanes, input increase results in higher amount of PCR product, showing that the reaction does not reach the plateau phase at 10 ng. For control DNA, the equivalent amounts of PCR products obtained for each gene demonstrate that all of the primer pairs present similar efficiencies. B and D, quantification of PCR products from three independent experiments with 10 ng of DNA input for CAL 27 (———; mean ± SE) and control DNA (— — —). A, for CAL 27, with 3 ng of input, only bands 3, 6, and 9 are clearly visible, showing increased copy number for the three corresponding genes. The quantification of PCR products for the intermediate dilution (B) define a candidate interval from 167.0–170.9 Mb. C, similarly, exhaustive analysis of the genes mapped in the candidate interval shows that all genes from 3–10 present increased copy number. The quantification of PCR products for the intermediate dilution (D) refines the core amplicon mapping from 167.7–168 Mb to 170.2–170.9 Mb. ▪ represent positions of clones from the BAC/PAC array included in the amplification; □ positions of clones excluded from amplification; numbers under squares refer to clone numbers in Table 1. Note that the gene 19 is mapped in the clone 27 and excluded from the amplification by array CGH (Fig. 1 B).

Close modal
Fig. 3.

Transcriptional analysis of chromosome 3 genes in head and neck cancer cell lines and primary tumors. A, differential expression of chromosome 3 genes between CAL 27 and a normal cell line by cDNA microarray analysis. Expression ratio data are represented as a function of the position of UniGene clusters on chromosome 3 according to the UCSC April 2001 working draft assembly. Individual data points have been arbitrarily connected with a line. Seventeen genes localized on chromosome 3 are overexpressed in CAL 27 with a ratio higher than 2. The three genes with the highest ratios are located in the array CGH largest candidate region of amplification delimited by a gray box. B, differential expression of genes located in the array CGH largest candidate region of amplification between CAL 27 and the control cell line Hs 377.Tg (on the left) or between CAL 27 and CAL 33, a head and neck cancer cell line with a low-level 3q gain (on the right), by cDNA microarray analysis. The core amplicon refined by the genomic PCR and delimited by a gray box contains ACATN (number 3), GMPS (number 4), and cyclin L (number 8). These three genes are overexpressed in CAL 27 compared with the normal cell line and CAL 33 (ratios > 2). Other genes are not overexpressed in both cases. KCNAB1 and IL12A are not expressed in every cell line. For gene number assignment, refer to Table 2. C, relative expression levels of ACATN, GMPS, cyclin L, DKFZp434J214, SSR3, and LOC115396 measured by quantitative RT-PCR in CAL 33, FaDu, and CAL 27 cancer cell lines compared with Hs 677.Tg normal cell line. Results from at least three independent experiments were normalized on GAPDH. For each gene, relative expression (mean ± SE) was obtained by normalization to Hs 677.Tg. SSR3 and LOC115396 are not overexpressed in any cancer cell line. In contrast, ACATN, GMPS, DKFZp434J214, and cyclin L are overexpressed in CAL 33, FaDu, and CAL 27 (unpaired t test: ∗, P < 0.05; ∗∗, P < 0.01). Expression of these four genes in CAL 27 is higher than in CAL 33 and FaDu (unpaired t test: P < 0.01, data not shown). D, fold change in cyclin L expression measured by real-time quantitative RT-PCR for head and neck primary tumors relative to their normal corresponding tissues. Well- and poorly differentiated tumors show distinct patterns of expression ratios. Cyclin L is overexpressed in the well-differentiated tumor population compared with the normal counterpart (Wilcoxon signed rank test; P < 0.05) but not in the poorly differentiated one. Results from three independent experiments were normalized on averaged expression of GAPDH and UBB. Each bar represents the ratio of cyclin L expression level in one tumor to its normal counterpart. 3q25 representation was evaluated by chromosome CGH in four tumors (8). +, tumors with 3q25 gain, −, tumor with normal 3q representation.

Fig. 3.

Transcriptional analysis of chromosome 3 genes in head and neck cancer cell lines and primary tumors. A, differential expression of chromosome 3 genes between CAL 27 and a normal cell line by cDNA microarray analysis. Expression ratio data are represented as a function of the position of UniGene clusters on chromosome 3 according to the UCSC April 2001 working draft assembly. Individual data points have been arbitrarily connected with a line. Seventeen genes localized on chromosome 3 are overexpressed in CAL 27 with a ratio higher than 2. The three genes with the highest ratios are located in the array CGH largest candidate region of amplification delimited by a gray box. B, differential expression of genes located in the array CGH largest candidate region of amplification between CAL 27 and the control cell line Hs 377.Tg (on the left) or between CAL 27 and CAL 33, a head and neck cancer cell line with a low-level 3q gain (on the right), by cDNA microarray analysis. The core amplicon refined by the genomic PCR and delimited by a gray box contains ACATN (number 3), GMPS (number 4), and cyclin L (number 8). These three genes are overexpressed in CAL 27 compared with the normal cell line and CAL 33 (ratios > 2). Other genes are not overexpressed in both cases. KCNAB1 and IL12A are not expressed in every cell line. For gene number assignment, refer to Table 2. C, relative expression levels of ACATN, GMPS, cyclin L, DKFZp434J214, SSR3, and LOC115396 measured by quantitative RT-PCR in CAL 33, FaDu, and CAL 27 cancer cell lines compared with Hs 677.Tg normal cell line. Results from at least three independent experiments were normalized on GAPDH. For each gene, relative expression (mean ± SE) was obtained by normalization to Hs 677.Tg. SSR3 and LOC115396 are not overexpressed in any cancer cell line. In contrast, ACATN, GMPS, DKFZp434J214, and cyclin L are overexpressed in CAL 33, FaDu, and CAL 27 (unpaired t test: ∗, P < 0.05; ∗∗, P < 0.01). Expression of these four genes in CAL 27 is higher than in CAL 33 and FaDu (unpaired t test: P < 0.01, data not shown). D, fold change in cyclin L expression measured by real-time quantitative RT-PCR for head and neck primary tumors relative to their normal corresponding tissues. Well- and poorly differentiated tumors show distinct patterns of expression ratios. Cyclin L is overexpressed in the well-differentiated tumor population compared with the normal counterpart (Wilcoxon signed rank test; P < 0.05) but not in the poorly differentiated one. Results from three independent experiments were normalized on averaged expression of GAPDH and UBB. Each bar represents the ratio of cyclin L expression level in one tumor to its normal counterpart. 3q25 representation was evaluated by chromosome CGH in four tumors (8). +, tumors with 3q25 gain, −, tumor with normal 3q representation.

Close modal

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

1

Supported by grants from the ‘Comité Régional du Haut-Rhin de la Ligue contre le Cancer,’ by the program ‘carte d’identité des tumeurs’ from the ‘Ligue Nationale contre le Cancer,’ and by funds from Institut National de la Santé et de la Recherche Médicale, Centre National de la Recherche Scientifique, and the ‘Hôpital Universitaire de Strasbourg.’ R. R. was supported by a fellowship from the ‘Association pour la Recherche sur le Cancer’ and T. H. by a fellowship from the ‘Ministère de l’Education Nationale, de la Recherche et des Technologies.’

4

The abbreviations used are: CGH, comparative genomic hybridization; FISH, fluorescence in situ hybridization; RT-PCR, reverse transcription-PCR; UCSC, University of California, Santa Cruz; STS, sequence-tagged site; BAC, bacterial artificial chromosome; PAC, P1 artificial chromosome; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; HNSCC, head and neck squamous cell carcinoma.

5

Internet address: apollo.uthscsa.edu.

6

Internet address: genome.ucsc.edu.

7

Internet address: publication.celera.com.

8

Internet address: shark.ucsf.edu/gc.

9

Internet address: www.ncbi.nlm.nih.gov/LocusLink.

10

Internet address: genome-www.stanford.edu/serum.

Table 1

Chromosome 3 location of the 85 genomic clones used for array CGH according to the April 2001 version of the UCSC draft assembly

No.CloneSTSaAccession no.Location (Mb)b
RP11-204C23  AC018814 2.9 
RP11-121D3  AC016963 3.4 
RP11-245E5  AC018542 28.6 
RP11-11L6  AC021751 34.7 
RP11-103N21  AC018351 35.9 
RP11-286G5  AC020624 41.7 
RP11-111P21  AC063916 50.8 
RP11-122D19  AC018354 59.5 
RP11-169G24  AC018920 61.6 
10 RP11-94D19  AC016965 67.6 
11 RP11-124O2  AC018352 107.5 
12 RP4-710G11 D3S3695  117.1 
13 RP4-671C12 D3S1610  121.5 
14 RP4-730E1 D3S1316  150.8 
15 RP11-4H14  AC068429 152.0 
16 RP11-438D8  AC010184 152.5 
17 RP11-190C21  AC072028 153.8 
18 RP5-1164J18 D3S3546  154.1 
19 RP5-924I19 D3S1593  157.0 
20 RP11-649A16  AC027060 158.9 
21 RP11-464E15  AC021059 160.9 
22 RP11-385G14  AC011103 163.1 
23 RP11-529G21  AC015991 165.3 
24 RP11-316N24  AC041012 166.6 
25 RP5-854L10 D3S1275  168.5 
26 RP11-197N19  AC020630 170.1 
27 RP11-227J5  AC024221 173.1 
28 RP11-16N11  AC069224 173.6 
29 RP11-203L15  AC016920 173.9 
30 RP11-816J6  AC078802 183.0 
31 RP4-701G18 D3S1564  184.0 
32 RP11-517A22 D3S3094  185.4 
33 RP11-198D7 D3S1548  187.0 
34 RP11-477P16  AC008120 188.0 
35 RP11-89J17  AC016936 188.6 
36 RP11-414N11 D3S3676  190.6 
37 RP11-120G16 SHGC-15953  191.1 
38 RP11-238P22 D3S3097  191.1 
39 RP5-1020C14 D3S3511  191.4 
40 RP11-2L8  AC007953 191.9 
41 RP11-569L16  AC021975 192.5 
42 RP11-245C23  AC076966 192.8 
43 RP11-229I14 stSG4294  193.4 
44 RP11-118F4  AC048335 193.6 
45 RP11-657G5  AC073277 194.4 
46 RP11-544D12  AC068638 195.6 
47 RP11-259I19  AC024104 196.9 
48 RP11-139K4  AC021055 197.1 
49 RP11-235O18 stSG8306  197.5 
50 RP11-12L14  AC013631 198.4 
51 RP11-14I2  AC068769 198.4 
52 RP11-80E16 KIAA0208  198.9 
53 RP11-488M12  AC061705 198.9 
54 RP11-125E8  AC078797 199.0 
55 RP11-329B9  AC016975 199.3 
56 RP11-683J13  AC069420 200.0 
57 RP11-410F19  AC009247 200.5 
58 RP11-379C23  AC007917 200.8 
59 RP5-875F16 D3S1602  200.8 
60 RP11-208N14  AC007920 201.9 
61 RP11-119E13  AC012149 202.1 
62 RP11-211G3  AC072022 202.8 
63 RP11-30O15  AC021414 203.0 
64 RP11-426J22  AC069226 203.7 
65 RP11-297K7  AC009319 204.0 
66 RP11-67E18  AC055711 204.1 
67 RP11-115I8  AC018516 204.5 
68 RP11-352A18  AC024103 204.8 
69 RP4-756C5 D3S1294  205.2 
70 RP11-53D15  AC016966 205.2 
71 RP11-789N15  AC080006 205.6 
72 RP11-303B22 D3S1314  205.7 
73 RP11-398D6 D3S1314  205.7 
74 RP11-525C11  AC068645 207.1 
75 RP11-608P9  AC025512 207.5 
76 RP4-714P24 AFM308yfl  207.7 
77 RP11-590F24  AC079847 208.3 
78 RP11-91M9  AC026671 208.3 
79 RP11-532L22  AC016947 208.9 
80 RP11-143P4  AC021846 209.2 
81 RP11-528A4  AC069421 209.9 
82 RP11-338O10  AC069213 212.1 
83 RP11-436M6  AC009099 212.3 
84 RP4-654H18 D3S1305  212.5 
85 RP11-496H1  AC024560 214.5 
No.CloneSTSaAccession no.Location (Mb)b
RP11-204C23  AC018814 2.9 
RP11-121D3  AC016963 3.4 
RP11-245E5  AC018542 28.6 
RP11-11L6  AC021751 34.7 
RP11-103N21  AC018351 35.9 
RP11-286G5  AC020624 41.7 
RP11-111P21  AC063916 50.8 
RP11-122D19  AC018354 59.5 
RP11-169G24  AC018920 61.6 
10 RP11-94D19  AC016965 67.6 
11 RP11-124O2  AC018352 107.5 
12 RP4-710G11 D3S3695  117.1 
13 RP4-671C12 D3S1610  121.5 
14 RP4-730E1 D3S1316  150.8 
15 RP11-4H14  AC068429 152.0 
16 RP11-438D8  AC010184 152.5 
17 RP11-190C21  AC072028 153.8 
18 RP5-1164J18 D3S3546  154.1 
19 RP5-924I19 D3S1593  157.0 
20 RP11-649A16  AC027060 158.9 
21 RP11-464E15  AC021059 160.9 
22 RP11-385G14  AC011103 163.1 
23 RP11-529G21  AC015991 165.3 
24 RP11-316N24  AC041012 166.6 
25 RP5-854L10 D3S1275  168.5 
26 RP11-197N19  AC020630 170.1 
27 RP11-227J5  AC024221 173.1 
28 RP11-16N11  AC069224 173.6 
29 RP11-203L15  AC016920 173.9 
30 RP11-816J6  AC078802 183.0 
31 RP4-701G18 D3S1564  184.0 
32 RP11-517A22 D3S3094  185.4 
33 RP11-198D7 D3S1548  187.0 
34 RP11-477P16  AC008120 188.0 
35 RP11-89J17  AC016936 188.6 
36 RP11-414N11 D3S3676  190.6 
37 RP11-120G16 SHGC-15953  191.1 
38 RP11-238P22 D3S3097  191.1 
39 RP5-1020C14 D3S3511  191.4 
40 RP11-2L8  AC007953 191.9 
41 RP11-569L16  AC021975 192.5 
42 RP11-245C23  AC076966 192.8 
43 RP11-229I14 stSG4294  193.4 
44 RP11-118F4  AC048335 193.6 
45 RP11-657G5  AC073277 194.4 
46 RP11-544D12  AC068638 195.6 
47 RP11-259I19  AC024104 196.9 
48 RP11-139K4  AC021055 197.1 
49 RP11-235O18 stSG8306  197.5 
50 RP11-12L14  AC013631 198.4 
51 RP11-14I2  AC068769 198.4 
52 RP11-80E16 KIAA0208  198.9 
53 RP11-488M12  AC061705 198.9 
54 RP11-125E8  AC078797 199.0 
55 RP11-329B9  AC016975 199.3 
56 RP11-683J13  AC069420 200.0 
57 RP11-410F19  AC009247 200.5 
58 RP11-379C23  AC007917 200.8 
59 RP5-875F16 D3S1602  200.8 
60 RP11-208N14  AC007920 201.9 
61 RP11-119E13  AC012149 202.1 
62 RP11-211G3  AC072022 202.8 
63 RP11-30O15  AC021414 203.0 
64 RP11-426J22  AC069226 203.7 
65 RP11-297K7  AC009319 204.0 
66 RP11-67E18  AC055711 204.1 
67 RP11-115I8  AC018516 204.5 
68 RP11-352A18  AC024103 204.8 
69 RP4-756C5 D3S1294  205.2 
70 RP11-53D15  AC016966 205.2 
71 RP11-789N15  AC080006 205.6 
72 RP11-303B22 D3S1314  205.7 
73 RP11-398D6 D3S1314  205.7 
74 RP11-525C11  AC068645 207.1 
75 RP11-608P9  AC025512 207.5 
76 RP4-714P24 AFM308yfl  207.7 
77 RP11-590F24  AC079847 208.3 
78 RP11-91M9  AC026671 208.3 
79 RP11-532L22  AC016947 208.9 
80 RP11-143P4  AC021846 209.2 
81 RP11-528A4  AC069421 209.9 
82 RP11-338O10  AC069213 212.1 
83 RP11-436M6  AC009099 212.3 
84 RP4-654H18 D3S1305  212.5 
85 RP11-496H1  AC024560 214.5 
a

Nonsequenced clones are mapped according to their STS content.

b

Central position of the clone (average of clone end mapping values).

Table 2

Primer sequences for genomic and transcriptional analysis of genes located within the 3q25.3 amplicon detected in CAL 27

RankGeneGenome analysisExpression analysis
Chr. 3 mapping (Mb)Forward primerReverse primerMicroarrayaForward primerReverse primer
MME 167.0 5′-AGTGCCTGGATCCCCTCTGT-3′ 5′-CATCAGCTCCCAACCCTTTG-3′ Yes   
KIAA1069 167.7 5′-GACCACCAAACATGCAACGA-3′ 5′-TTTCCACAAGGCTGCTCTCC-3′ No   
ACATN 168.0 5′-CTTCGTGGAAATGCAAAAGGA-3′ 5′-GGTTACCACCCGTGACTACTGC-3′ Yes 5′-AAAAGCCCCACAGAGCTTCC-3′ 5′-TCGGTATTCCCAAGCAAACG-3′ 
GMPS 168.1 5′-AGCCAGATGCCGGTGATTT-3′ 5′-CTCATTGCCAGGTGTTGCAG-3′ Yes 5′-ACGCCAAGCTGATTTTGAGG-3′ 5′-TTCCTGGGGGCTTTGATGTT-3′ 
KCNAB1 168.4 5′-CCGCAACTGCTCAACCTCTC-3′ 5′-TTCATGCCAGTCTGCTTTGC-3′ Yes   
SSR3 169.1 5′-CTTGACCCTGCCATGTTGGT-3′ 5′-AGCTGGATCCCAAGGGCTAA-3′ No 5′-GCACAGAAGAGGGAGGATGC-3′ 5′-GAGGGCGATGAGTCCTGATG-3′ 
DKFZp434J214 169.2 5′-ACTGGGTGGGGAACAGCATT-3′ 5′-GCCTCCATTCCATTTGGTCA-3′ No 5′-TACGAAGGCTGTCCACACCA-3′ 5′-GGGAACCCCACCAAGTGTCT-3′ 
cyclin L 169.6 5′-TAGGCGGAGTCGATCTGGAA-3′ 5′-CCATGGTGCTTGCTTTTATGG-3′ Yes 5′-ACTCCAAGCCCCCTGATCCT-3′ 5′-TGGCAACGGAATCTGAAGTG-3′ 
PTX-3 170.1 5′-CGTGCATCCAGTGAGACCAA-3′ 5′-ACAGTGGTAGCCGCCAGTTC-3′ Yes   
10 LOC115396 170.2 5′-TGAAGACCTCAAAGTCTGGAGAAA-3′ 5′-GTCCCCAGCTCGTGAAAGAT-3′ No 5′-CCGGGCTTTCGAAATCTTCA-3′ 5′-CCTGCCATTTTAGGCCCTTC-3′ 
11 LOC51319 170.9 5′-CCCACATGCAGGAACAGTGA-3′ 5′-CATCTCCGTCCCCACCATTA-3′ Yes   
12 MLF1 171.3 5′-CCTGGCATGCAGTTCTCTCA-3′ 5′-AGGTTTGCCAGCCTGCTTCT-3′ No   
13 EFG 171.4   Yes   
14 RARRES1 171.4   Yes   
15 FLJ14153 171.5 5′-AGCTGGAGCCCGTGTAATCA-3′ 5′-TCTGCTTGAGTGGGGACCAT-3′ No   
16 SCHIP-1 172.5 5′-TCCAGTGATTGGCCTTTGCT-3′ 5′-TTCACTACGAGGCAGGCATGT-3′ No   
17 IL12A 172.8   Yes   
18 BRD7 172.9   Yes   
19 CAP-C 173.2 5′-AGTCCCACTGTTGAGGA-3′ 5′-AATGGCCCACTGAAGAA-3′ n.d.b   
Control UBB    n.d. 5′-GGTGGACGTGGTTGGTGATT-3′ 5′-TCTTGCCGGTCAGGGTCTT-3′ 
Control GAPDH    n.d. 5′-TTGCCCTCAACGACCACTTT-3′ 5′-TGGTGGTCCAGGGGTCTTAC-3′ 
RankGeneGenome analysisExpression analysis
Chr. 3 mapping (Mb)Forward primerReverse primerMicroarrayaForward primerReverse primer
MME 167.0 5′-AGTGCCTGGATCCCCTCTGT-3′ 5′-CATCAGCTCCCAACCCTTTG-3′ Yes   
KIAA1069 167.7 5′-GACCACCAAACATGCAACGA-3′ 5′-TTTCCACAAGGCTGCTCTCC-3′ No   
ACATN 168.0 5′-CTTCGTGGAAATGCAAAAGGA-3′ 5′-GGTTACCACCCGTGACTACTGC-3′ Yes 5′-AAAAGCCCCACAGAGCTTCC-3′ 5′-TCGGTATTCCCAAGCAAACG-3′ 
GMPS 168.1 5′-AGCCAGATGCCGGTGATTT-3′ 5′-CTCATTGCCAGGTGTTGCAG-3′ Yes 5′-ACGCCAAGCTGATTTTGAGG-3′ 5′-TTCCTGGGGGCTTTGATGTT-3′ 
KCNAB1 168.4 5′-CCGCAACTGCTCAACCTCTC-3′ 5′-TTCATGCCAGTCTGCTTTGC-3′ Yes   
SSR3 169.1 5′-CTTGACCCTGCCATGTTGGT-3′ 5′-AGCTGGATCCCAAGGGCTAA-3′ No 5′-GCACAGAAGAGGGAGGATGC-3′ 5′-GAGGGCGATGAGTCCTGATG-3′ 
DKFZp434J214 169.2 5′-ACTGGGTGGGGAACAGCATT-3′ 5′-GCCTCCATTCCATTTGGTCA-3′ No 5′-TACGAAGGCTGTCCACACCA-3′ 5′-GGGAACCCCACCAAGTGTCT-3′ 
cyclin L 169.6 5′-TAGGCGGAGTCGATCTGGAA-3′ 5′-CCATGGTGCTTGCTTTTATGG-3′ Yes 5′-ACTCCAAGCCCCCTGATCCT-3′ 5′-TGGCAACGGAATCTGAAGTG-3′ 
PTX-3 170.1 5′-CGTGCATCCAGTGAGACCAA-3′ 5′-ACAGTGGTAGCCGCCAGTTC-3′ Yes   
10 LOC115396 170.2 5′-TGAAGACCTCAAAGTCTGGAGAAA-3′ 5′-GTCCCCAGCTCGTGAAAGAT-3′ No 5′-CCGGGCTTTCGAAATCTTCA-3′ 5′-CCTGCCATTTTAGGCCCTTC-3′ 
11 LOC51319 170.9 5′-CCCACATGCAGGAACAGTGA-3′ 5′-CATCTCCGTCCCCACCATTA-3′ Yes   
12 MLF1 171.3 5′-CCTGGCATGCAGTTCTCTCA-3′ 5′-AGGTTTGCCAGCCTGCTTCT-3′ No   
13 EFG 171.4   Yes   
14 RARRES1 171.4   Yes   
15 FLJ14153 171.5 5′-AGCTGGAGCCCGTGTAATCA-3′ 5′-TCTGCTTGAGTGGGGACCAT-3′ No   
16 SCHIP-1 172.5 5′-TCCAGTGATTGGCCTTTGCT-3′ 5′-TTCACTACGAGGCAGGCATGT-3′ No   
17 IL12A 172.8   Yes   
18 BRD7 172.9   Yes   
19 CAP-C 173.2 5′-AGTCCCACTGTTGAGGA-3′ 5′-AATGGCCCACTGAAGAA-3′ n.d.b   
Control UBB    n.d. 5′-GGTGGACGTGGTTGGTGATT-3′ 5′-TCTTGCCGGTCAGGGTCTT-3′ 
Control GAPDH    n.d. 5′-TTGCCCTCAACGACCACTTT-3′ 5′-TGGTGGTCCAGGGGTCTTAC-3′ 
a

Yes/no, present/not present on the cDNA microarray.

b

n.d., not determined.

Table 3

Changes in ACATN, cyclin L, GMPS, and DKFZp434J214 gene expression for 20 head and neck primary tumors compared with corresponding normal tissues

GeneExpression level changea (no. of cases)Wilcoxon signed rank test
DecreaseNo changeIncrease
ACATN 14 No difference 
GMPS 13 No difference 
DKFZp434J214 12 No difference 
cyclin L 11 Overexpression, P < 0.05 
GeneExpression level changea (no. of cases)Wilcoxon signed rank test
DecreaseNo changeIncrease
ACATN 14 No difference 
GMPS 13 No difference 
DKFZp434J214 12 No difference 
cyclin L 11 Overexpression, P < 0.05 
a

For each gene, expression levels in tumors and normal tissues were measured three times independently and normalized to the averaged expression levels of GAPDH and UBB. Unpaired bilateral t test was used to compare gene expression levels in tumors and corresponding normal tissues. Numbers of tumors displaying significant changes (P < 0.01) are given in decrease or increase categories.

We thank the International Human Genome Sequencing Consortium and especially Jim Kent and coworkers for data availability through the “golden path.” We also thank Celera Genomics. We thank Doulaye Dembele for statistical expertise, Christine Macabre for technical assistance, and the IGBMC core facilities. We also thank Bohdan Wasylyk and Julia Young for a kind gift and Jean-Louis Mandel for helpful discussion and comments.

1
Kallioniemi A., Kallioniemi O-P., Sudar D., Rutovitz D., Gray J. W., Waldman F., Pinkel D. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors.
Science (Wash. DC)
,
258
:
818
-821,  
1992
.
2
Solinas-Toldo S., Lampel S., Stilgenbauer S., Nickolenko J., Benner A., Dohner H., Cremer T., Lichter P. Matrix-based comparative genomic hybridization: biochips to screen for genomic imbalances.
Genes Chromosomes Cancer
,
20
:
399
-407,  
1997
.
3
Pinkel D., Segraves R., Suda R. D., Clark S., Poole I., Kowbel D., Collins C., Kuo W. L., Chen C., Zhai Y., Dairkee S. H., Ljung B. M., Gray J. W., Albertson D. G. High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays.
Nat. Genet.
,
20
:
207
-211,  
1998
.
4
Albertson D. G., Ylstra B., Segraves R., Collins C., Dairkee S. H., Kowbel D., Kuo W. L., Gray J. W., Pinkel D. Quantitative mapping of amplicon structure by array CGH identifies CYP24 as a candidate oncogene.
Nat. Genet.
,
25
:
144
-146,  
2000
.
5
Snijders A. M., Nowak N., Segraves R., Blackwood S., Brown N., Conroy J., Hamilton G., Hindle A. K., Huey B., Kimura K., Law S., Myambo K., Palmer J., Ylstra B., Yue J. P., Gray J. W., Jain A. N., Pinkel D., Albertson D. G. Assembly of microarrays for genome-wide measurement of DNA copy number.
Nat. Genet.
,
29
:
263
-264,  
2001
.
6
Speicher M. R., Howe C., Crotty P., du Manoir S., Costa J., Ward D. C. Comparative genomic hybridization detects novel deletions and amplifications in head and neck squamous cell carcinomas.
Cancer Res.
,
55
:
1010
-1013,  
1995
.
7
Bockmuhl U., Schluns K., Kuchler I., Petersen S., Petersen I. Genetic imbalances with impact on survival in head and neck cancer patients.
Am. J. Pathol.
,
157
:
369
-375,  
2000
.
8
Redon R., Muller D., Caulee K., Wanherdrick K., Abecassis J., du Manoir S. A simple specific pattern of chromosomal aberrations at early stages of head and neck squamous cell carcinomas: PIK3CA but not p63 gene as a likely target of 3q26-qter gains.
Cancer Res.
,
61
:
4122
-4129,  
2001
.
9
Bockmuhl U., Schluns K., Schmidt S., Matthias S., Petersen I. Chromosomal alterations during metastasis formation of head and neck squamous cell carcinoma.
Genes Chromosomes Cancer
,
33
:
29
-35,  
2002
.
10
Jiang W., Kahn S. M., Tomita N., Zhang Y. J., Lu S. H., Weinstein I. B. Amplification and expression of the human cyclin D gene in esophageal cancer.
Cancer Res.
,
52
:
2980
-2983,  
1992
.
11
Muller D., Millon R., Velten M., Bronner G., Jung G., Engelmann A., Flesch H., Eber M., Methlin G., Abecassis J. Amplification of 11q13 DNA markers in head and neck squamous cell carcinomas: correlation with clinical outcome.
Eur. J. Cancer
,
33
:
2203
-2210,  
1997
.
12
Gebhart E., Liehr T. Patterns of genomic imbalances in human solid tumors (Review).
Int. J. Oncol.
,
16
:
383
-399,  
2000
.
13
Shayesteh L., Lu Y., Kuo W. L., Baldocchi R., Godfrey T., Collins C., Pinkel D., Powell B., Mills G. B., Gray J. W. PIK3CA is implicated as an oncogene in ovarian cancer.
Nat. Genet.
,
21
:
99
-102,  
1999
.
14
Ma Y. Y., Wei S. J., Lin Y. C., Lung J. C., Chang T. C., Whang-Peng J., Liu J. M., Yang D. M., Yang W. K., Shen C. Y. PIK3CA as an oncogene in cervical cancer.
Oncogene
,
19
:
2739
-2744,  
2000
.
15
Hibi K., Trink B., Patturajan M., Westra W. H., Caballero O. L., Hill D. E., Ratovitski E. A., Jen J., Sidransky D. AIS is an oncogene amplified in squamous cell carcinoma.
Proc. Natl. Acad. Sci. USA
,
97
:
5462
-5467,  
2000
.
16
Guan X. Y., Sham J. S., Tang T. C., Fang Y., Huo K. K., Yang J. M. Isolation of a novel candidate oncogene within a frequently amplified region at 3q26 in ovarian cancer.
Cancer Res.
,
61
:
3806
-3809,  
2001
.
17
Gioanni J., Fischel J. L., Lambert J. C., Demard F., Mazeau C., Zanghellini E., Ettore F., Formento P., Chauvel P., Lalanne C. M., Courdi A. Two new human tumor cell lines derived from squamous cell carcinomas of the tongue: establishment, characterization and response to cytotoxic treatment.
Eur. J. Cancer Clin. Oncol.
,
24
:
1445
-1455,  
1988
.
18
Rangan S. R. A new human cell line (FaDu) from a hypopharyngeal carcinoma.
Cancer (Phila.)
,
29
:
117
-121,  
1972
.
19
Cheung V. G., Nowak N., Jang W., Kirsch I. R., Zhao S., Chen X. N., Furey T. S., Kim U. J., Kuo W. L., Olivier M., et al Integration of cytogenetic landmarks into the draft sequence of the human genome.
Nature (Lond.)
,
409
:
953
-958,  
2001
.
20
du Manoir S., Schrock E., Bentz M., Speicher M. R., Joos S., Ried T., Lichter P., Cremer T. Quantitative analysis of comparative genomic hybridization.
Cytometry
,
19
:
27
-41,  
1995
.
21
International Human Genome Sequencing Consortium. Initial sequencing and analysis of the human genome.
Nature (Lond.)
,
409
:
860
-921,  
2001
.
22
Venter J. C., Adams M. D., Myers E. W., Li P. W., Mural R. J., Sutton G. G., Smith H. O., Yandell M., Evans C. A., Holt R. A., et al The sequence of the human genome.
Science (Wash. DC)
,
291
:
1304
-1351,  
2001
.
23
Kanamori A., Nakayama J., Fukuda M. N., Stallcup W. B., Sasaki K., Fukuda M., Hirabayashi Y. Expression cloning and characterization of a cDNA encoding a novel membrane protein required for the formation of O-acetylated ganglioside: a putative acetyl-CoA transporter.
Proc. Natl. Acad. Sci. USA
,
94
:
2897
-2902,  
1997
.
24
Pegram L. D., Megonigal M. D., Lange B. J., Nowell P. C., Rowley J. D., Rappaport E. F., Felix C. A. t(3;11) translocation in treatment-related acute myeloid leukemia fuses MLL with the GMPS (guanosine 5-prime monophosphate synthetase) gene.
Blood
,
96
:
4360
-4362,  
2000
.
25
Hirst M., Haliday E., Nakamura J., Lou L. Human GMP synthetase: protein purification, cloning, and functional expression of cDNA.
J. Biol. Chem.
,
269
:
23830
-23837,  
1994
.
26
van ’t Veer L. J., Dai H., van de Vijver M. J., He Y. D., Hart A. A., Mao M., Peterse H. L., van der Kooy K., Marton M. J., Witteveen A. T., Schreiber G. J., Kerkhoven R. M., Roberts C., Linsley P. S., Bernards R., Friend S. H. Gene expression profiling predicts clinical outcome of breast cancer.
Nature (Lond.)
,
415
:
530
-536,  
2002
.
27
Berke J. D., Sgambato V., Zhu P. P., Lavoie B., Vincent M., Krause M., Hyman S. E. Dopamine and glutamate induce distinct striatal splice forms of Ania-6, an RNA polymerase II-associated cyclin.
Neuron
,
32
:
277
-287,  
2001
.
28
Guhaniyogi J., Brewer G. Regulation of mRNA stability in mammalian cells.
Gene (Amst.)
,
265
:
11
-23,  
2001
.
29
Herschman H. R. Primary response genes induced by growth factors and tumor promoters.
Annu. Rev. Biochem.
,
60
:
281
-319,  
1991
.
30
Iyer V. R., Eisen M. B., Ross D. T., Schuler G., Moore T., Lee J. C., Trent J. M., Staudt L. M., Hudson J. J., Boguski M. S., Lashkari D., Shalon D., Botstein D., Brown P. O. The transcriptional program in the response of human fibroblasts to serum.
Science (Wash. DC)
,
283
:
83
-87,  
1999
.
31
Monni O., Barlund M., Mousses S., Kononen J., Sauter G., Heiskanen M., Paavola P., Avela K., Chen Y., Bittner M. L., Kallioniemi A. Comprehensive copy number and gene expression profiling of the 17q23 amplicon in human breast cancer.
Proc. Natl. Acad. Sci. USA
,
98
:
5711
-5716,  
2001
.