Pancreatic carcinomas display highly complex chromosomal abnormalities, including many structural and numerical aberrations. There is ample evidence indicating that some of these abnormalities, such as recurrent amplifications and homozygous deletions, contribute to tumorigenesis by altering expression levels of critical oncogenes and tumor suppressor genes. To increase the understanding of gene copy number changes in pancreatic carcinomas and to identify key amplification/deletion targets, we applied genome-wide array-based comparative genomic hybridization to 31 pancreatic carcinoma cell lines. Two different microarrays were used, one containing 3,565 fluorescence in situ hybridization-verified bacterial artificial chromosome clones and one containing 25,468 cDNA clones representing 17,494 UniGene clusters. Overall, the analyses revealed a high genomic complexity, with several copy number changes detected in each case. Specifically, 60 amplicons at 32 different locations were identified, most frequently located within 8q (8 cases), 12p (7 cases), 7q (5 cases), 18q (5 cases), 19q (5 cases), 6p (4 cases), and 8p (4 cases). Amplifications of 8q and 12p were mainly clustered at 8q23–24 and 12p11–12, respectively, whereas amplifications on other chromosome arms were more dispersed. Furthermore, our analyses identified several novel homozygously deleted segments located to 9p24, 9p21, 9q32, 10p12, 10q22, 12q24, and 18q23. The individual complexity and aberration patterns varied substantially among cases, i.e., some cell lines were characterized mainly by high-level amplifications, whereas others showed primarily whole-arm imbalances and homozygous deletions. The described amplification and deletion targets are likely to contain genes important in pancreatic tumorigenesis.

Pancreatic carcinoma is the fourth leading cause of cancer-related death in the United States and is characterized by a highly aggressive tumor phenotype leading to a median survival of 6 months and a 5-year survival rate of only 3% (1). At the cytogenetic level, pancreatic carcinomas show many structural and numerical abnormalities. Frequent chromosomal imbalances include trisomy 7 and 20; monosomy 18; partial or whole-arm gains of 3q, 5p, 8q, 11q, 12p, 17q, 19q, and 20q; and losses of 1p, 3p, 6q, 8p, 9p, 15q, 17p, 18q, and 19p (2, 3). These findings have been corroborated by a number of comparative genome hybridization (CGH) studies (4, 5, 6, 7, 8), which moreover have led to the identification of numerous regions showing genomic amplifications, most commonly observed within 5p, 7q21–31, 8q22-qter, 12p11–12, 19q, and 20q. Genomic amplifications and homozygous deletions are believed to contribute to tumorigenesis by altering expression levels of critical oncogenes and tumor suppressor genes. For these reasons CGH/fluorescence in situ hybridization (FISH; Refs. 5, 7, 8) and loss of heterozygosity analyses (9, 10, 11, 12) have been used to map frequently altered chromosomal regions in a wide variety of human neoplasms. The precise definition of chromosomal changes by these methods, however, has to date been hampered by the low resolution or inefficiency associated with these techniques. In cases in which more detailed analyses have been undertaken, regions harboring identical amplicons as determined by CGH have shown a considerable degree of heterogeneity (13).

The introduction of array-based CGH (14, 15) and the vast amounts of mapping data that have recently become available through the human genome effort have greatly facilitated the analyses of disease-related amplified or deleted chromosomal regions (16, 17, 18, 19). To increase understanding of the complex gene copy number changes occurring in pancreatic carcinomas and to identify key amplification/deletion target regions, we applied whole-genome array-based CGH on a panel of 31 pancreatic carcinoma cell lines. For this purpose we used two different microarray platforms, one containing 3,565 FISH-verified bacterial artificial chromosome (BAC) clones and one containing 25,648 different cDNA clones representing 17,494 UniGene clusters. Using this strategy, we successfully detected and confirmed the majority of previously known larger gene copy number alterations characteristic for pancreatic carcinomas. In addition, this approach resulted in the identification and precise molecular definition of several amplified as well as deleted regions likely to contain genes involved in pancreatic tumor formation and/or progression.

Tumor Cell Lines.

A total of 31 pancreatic carcinoma cell lines were included in the investigation. Fifteen of these, LPC1p–LPC8p and LPC10m–LPC15p, were low-passage cell lines (5–10 passages) established at the Department of Clinical Genetics in Lund as described by Gorunova et al.(3). The clinical and histopathological data for LPC1p–LPC12m have been reported previously (3, 9). LPC13p and LPC14p originated from moderately differentiated primary ductal adenocarcinomas from 60- and 78-year-old males, respectively, whereas LPC15p was derived from a more complex primary tumor in a 66-year-old female. The remainder of the investigated cases were obtained from American Type Culture Collection (Manassas, VA) and Deutsches Krebsforschungszentrum (Heidelberg, Germany) and consisted of AsPC-1, BxPC-3, Capan-2, CFPAC-1, DANG, HPAF-II, Hs700T, Hs766T, HupT3, HupT4, PANC-1, PaTu8902, PaTu8988S, PaTu8988T, SU.86.86, and SW1990. As a control for normal gene copy numbers, DNA from lymphocytes of healthy males was used.

Array-Based CGH.

Array-based CGH was performed with BAC and cDNA microarrays.3 The BAC arrays contained a total of 3,565 colony-purified FISH-verified clones, including ∼3,200 clones selected through an international collaboration to cover the genome with a 1-Mb resolution (20), and clones specifically selected for previous studies (21, 22, 23). Preparation, labeling, hybridization, and scanning procedures for BAC arrays were performed as described in detail by Vissers et al.(24).

The cDNA microarrays were obtained from the Swegene DNA microarray resource center at Lund University4 and consisted of 25,648 different cDNA clones representing 17,494 UniGene clusters. Labeling and hybridizations were performed essentially as described by Pollack et al.(25), and slide pre- and post-treatments were performed using the Universal Microarray Hybridization Kit (Corning, Acton, MA) according to instructions provided by the manufacturer. Fluorescence intensities were quantified on an Agilent G2565AA microarray scanner (Agilent Technologies, Palo Alto, CA).

Analyses of microarray images from both BAC and cDNA hybridizations were performed with GenePix Pro 4.0 (Axon Instruments, Inc., Foster City, CA). For each spot, the median pixel intensities minus the median local backgrounds for both dyes were used to obtain the ratio of test gene copy number to reference gene copy number. Data normalization for BAC arrays was performed for each array subgrid by applying linear regression, and data from cDNA arrays were normalized over the entire array by Lowess curve fitting (26) with a smoothing factor of 0.33. The normalizations were performed with the software package SAS, version 8.0 (SAS Institute, Cary, NC), and the web-based database BASE (27) for the BAC and cDNA arrays, respectively. Mapping information regarding clone locations, cytogenetic bands, and genomic contents were retrieved from the University of California Santa Cruz genome browser (July 2003 freeze)5 and UniGene build 164 at the National Center for Biotechnology Information.6

FISH.

The FISH analyses were performed essentially as described previously (28). Briefly, isolated BAC or P1-derived artificial chromosome DNA was labeled with either biotin- or digoxigenin-conjugated dUTP by use of the Megaprime DNA labeling kit (Amersham). After labeling, the DNA was purified on a Sepharose CL-6B column (Amersham). The hybridization signals were analyzed in a CytoVision Ultra system (Applied Imaging, Santa Clara, CA). The genomic clones used as probes were obtained from BACPAC Resources.7

Verifications of Homozygous Deletions.

Potential homozygous deletions were identified by use of the criterion that at least two adjacent BAC clones had to fulfill the deletion threshold ratio, determined as a value less than or equal to −3 SD of the particular hybridization. For sensitivity reasons only information from the BAC arrays was used in the identification of deletions. To independently verify homozygous deletions, primers for sequence tagged sites (STSs) within or between potentially lost BAC clones were obtained. All primer sequences were retrieved from UniSTS at the National Center for Biotechnology Information. Apart from the STSs contained near or within selected BAC clones, four STSs corresponding to known or putative tumor suppressor genes located close to regions suspected to be deleted were investigated. These genes, and their respective STSs, were CDKN2A (RH103023), SMAD4 (RH75869), RSU1 (RH103137), and DEC1 (SHGC-85730). All 31 cell lines were screened for all markers, and as a control for a functional PCR in cases with homozygous deletions, primers for ACTB were multiplexed in all reactions.

Array-Based CGH.

Microarrays containing either BAC or cDNA clones were used to identify local genomic amplifications and homozygous deletions in a panel of 31 pancreatic cancer cell lines. Generally the BAC and the cDNA analyses correlated very well (Fig. 1). To further validate the microarray results and to estimate the correlation between gene copy number ratios and gene copies as determined by chromosome analysis, FISH analyses were performed on two cases using probes on 8p, 8q, and 20q. The results from these analyses are shown in Fig. 2. The array analysis of LPC3p indicated equal underrepresentation of 8p and 8q, shown by log2 ratios close to −0.35 over the entire chromosome (Fig. 2,A). FISH analysis revealed the presence of two normal chromosomes 8 in cells with chromosome numbers close to 60, i.e., near-triploid cells (Fig. 2,B). The equivalent analysis of LPC13p showed one normal chromosome 8 and three additional 8q signals on two other chromosomes (Fig. 2,C), also in accordance with the array findings, which showed average log2 ratios of −0.47 for 8p and 0.67 for 8q. However, the number of 8p:8q copies as determined by microarray analysis (1.4:3.2 copies) underestimated the gene copies obtained by FISH (1:4 copies). The 20q copy number profile of LPC3p indicated high-level amplification (8-fold) of a small segment within 20q13.2 (Fig. 2,A). The comparable FISH analysis showed three chromosomes 20, one normal and two with high-level amplification of the P1-derived artificial chromosome probe RP4-724E16, located at the center of the amplicon (Fig. 2, A and D). The corresponding analysis of LPC13p displayed the presence of two normal chromosomes 20 (Fig. 2,E), in agreement with the array results (average log2 ratio of 0.06). Thus, an overall good correlation between the array-based CGH and FISH results was observed. The whole-genome array-based CGH profiles showed that all of the investigated cases contained larger copy number alterations but that the aberration complexity varied substantially among individual cases. This is illustrated by the profiles for LPC13p, which harbors only a few changes, and for DANG, which in addition to numerous gains and losses also indicates the presence of several high-level amplifications and potential homozygous deletions (Fig. 3).

Genomic Amplifications.

Amplifications were defined as regions containing two or more adjacent clones displaying log2 gene copy number ratios ≥1.0 on at least one of the microarrays. For size determination, the amplicons were assigned with the distance between their flanking unamplified clones, i.e., the closest clones showing consistent log2 ratios <1.0, taking both array platforms into consideration. In total 60 amplicons at 32 different locations were detected (Table 1), showing log2 ratios between 1.0 and 3.5. The amplicons ranged from 0.4 to 38.1 Mb in size, with average and median sizes of 8.4 and 4.5 Mb, respectively, indicating that the majority of amplicons were rather small. Many of the smaller amplicons were well defined, showing sharp increases in gene copy number ratios compared with neighboring clones, whereas the larger amplicons typically showed central peaks flanked by gradual gene copy number decreases. The regions most frequently involved in amplifications were located at 6p21–22, 7q21–31, 8p11–12, 8q23–24, 12p11–12, 18q11–12, and 19q13.2.

Four cases showed amplifications of 6p21–22. These included two commonly amplified regions, each involving two cases (Table 1). The distal region, equivalent to the HupT3 amplicon that was entirely included in the SW1990 amplicon, was 4.4 Mb in size and included almost 400 genes/UniGene clusters. The proximal commonly amplified segment, corresponding to short peripheral sequences of the LPC3p and SU.86.86 amplicons (Fig. 4), was significantly smaller (0.4 Mb) than the distal region. This recurrently amplified segment contains 37 genes/UniGene clusters, including CCND3. The amplification in LPC3p was also investigated by FISH (Fig. 5 A), which confirmed the high-level amplification indicated by array-based CGH.

Amplicons within 7q were observed in five cases. These were heterogeneous with regard to both size and location (Table 1; Fig. 4), resulting in complex interpretation of commonly amplified regions. The first region, determined by the proximal amplicon border in LPC12m and the distal amplicon border in LPC14p, was located in 7q21.1–7q22.1, was 19.3 Mb in size, and contained 665 genes/UniGene clusters, including HGF and CDK6. Together with LPC12m and LPC14p, AsPC-1 and Hs700T contributed to amplification of two to three cases over the entire stretch (Table 1; Fig. 4). The second commonly amplified region, observed in two cases, was more easily defined and extended over 16.4 Mb within 7q22.3–7q31.3. This region, which was defined by the proximal border in DANG and distal border in Hs700T (Table 1; Fig. 4), contained almost 450 genes/UniGene clusters and included HGFR (MET).

A total of four different amplicons were observed within 8p: one in LPC11p/LPC11m, derived from a primary tumor and a metastasis from the same patient (9), one in HupT3, and two in DANG. All of these were small, in the range of 1.7–2.5 Mb (Table 1), and nonoverlapping. The profiles for DANG and HupT3 are shown in Fig. 4.

The chromosome arm most frequently involved in amplifications was 8q, for which eight cases revealed amplifications. On the basis of amplicon size and location these appeared to fall into two categories. In three of the cases, small local 8q24.21 amplifications were seen, 1.2–3.1 Mb in size, and included MYC. The remaining cases contained larger overlapping amplicons (Table 1). Some of the cases showed constant levels of amplification over the whole region, whereas others, such as HupT3, displayed large segments with high-level amplifications interrupted by shorter stretches of low-level amplifications, indicating a modular amplicon structure (Fig. 4). The commonly amplified region, affected in all five cases with larger 8q23–24 amplifications, was 15.5 Mb in size (Table 1) and harbored almost 600 genes/UniGene clusters, including MYC.

Seven cases showed amplifications of 12p. The amplicons ranged from 5.4 to 17.4 Mb in size and were located mainly within 12p11–12, except in cell line PaTu8902, in which the amplification was located at the distal 12p13. The cases with amplifications in 12p11–12 showed a commonly amplified segment extending from BCAT1 to PTHLH, a region 3.2 Mb in size. This segment contains nearly 150 genes/UniGene clusters, including KRAS2 and PPFIBP1. Interestingly, both LPC4p and DANG showed marked discontinuities in their levels of amplification in which the most amplified segments were located distal to the commonly amplified region. In LPC4p the amplification maximum extended from SOX5 to CMAS, whereas the overlapping maximum in DANG was located ever further distally, extending to the gene PIK3C2G at 12p12.3 (Fig. 4).

Five cases showed amplifications of 18q, ranging from 1.9 to 8.6 Mb in size (Table 1). LPC11p and LPC11m had an amplicon within the distal 18q12, a region of ∼7 Mb. Because the array-based CGH data suggested that this amplification was located at the very end or at close proximity to a deletion breakpoint (Fig. 1), BAC clones for the amplification and for proximal and distal segments were used for FISH. As seen in the inset in Fig. 5,B, the BAC clone RP11-91K12, representing the amplified region, is located at the very end of the deleted derivative chromosome 18. The interphase nucleus shows the presence of two normal and three derivative chromosomes 18, corroborating the copy number profiles. PaTu8988S and PaTu8988T showed a proximal amplicon within 18q11, ∼2 Mb in size, that overlapped with the distal 0.8 Mb of the amplicon in LPC6p (Table 1). Because PaTu8988T and -S, and LPC11p and -m represent primary and metastatic tumors derived from the same respective patients, this was the only region defined as recurrently amplified on 18q. The segment contains 17 genes/UniGene clusters.

Amplifications of 19q were seen in five cases. SU.86.86 contained two separate amplicons, one proximal and 6.5 Mb in size, and one very small distal, ∼0.4 Mb in size (Table 1). The proximal amplifications included the amplifications seen in LPC6p, PANC-1, and PaTu8988T. The overlapping segment, amplified in at least three cases, was 1.1 Mb in size and included 78 genes/UniGene clusters. AKT2, known to be amplified in pancreatic carcinomas, was included in the PANC-1 and SU.86.86 but not in the LPC6p and PaTu8988T amplifications. The amplicon in LPC3p was distal to the commonly amplified segments in SU.86.86, LPC6p, PANC-1, and PaTu8988T (Fig. 4).

Homozygous Deletions.

Thirteen segments were identified as candidate regions for homozygous deletions. Examples of such segments are shown in Fig. 6,A. PCR analysis of STSs within the identified regions confirmed that seven segments, located within 9p24, 9p21, 9q32, 10p12, 10q22, 12q24, and 18q23, were homozygously lost in at least one case. The deletions were searched for previously described potential tumor suppressor genes in an interval extending 2 Mb upstream and downstream from the marker used to verify the homozygous deletion, which revealed that DEC1 was included in the 9q32 deletion in BxPC-3 (Fig. 6,B). The tumor cell lines were also screened for homozygous deletions of SMAD4 and CDKN2A, both known to be homozygously deleted in pancreatic cancer. CDKN2A was deleted in 11 of the 31 cases and SMAD4 in 3 cases (Fig. 6 B). As for the number of homozygous deletions present in the individual cases, BxPC-3 was the most affected case, showing loss of both alleles at six different genomic locations.

We performed array-based CGH to map genome-wide DNA copy number alterations in a panel of 31 pancreatic carcinoma cell lines. Because the most common chromosomal imbalances have previously been identified in pancreatic carcinomas (2, 3), we focused on the more extreme forms of chromosome imbalances, i.e., local high-level amplifications and homozygous deletions. As the analysis of gene copy numbers directly on pancreatic tumor biopsies is hampered by their strong desmoplastic reaction, resulting in a large fraction of nontumorous cells in the biopsies, the present analysis was performed on pancreatic cancer cell lines. Two different microarrays were used, one containing 3,565 well-characterized BAC clones and one containing 25,468 different cDNA clones. Generally the data from these analyses correlated very well and provided convincing support for one another. Because of uneven clone distribution for certain genomic regions, the arrays furthermore complemented each other and the combined sets provided good coverage over the entire genome. To validate the results we performed single-copy FISH analyses using nine different BAC clones. The FISH analyses corroborated the array results in every instance, which showed that results obtained with array analysis were reliable in identifying both low-level gains and losses of larger regions and high-level local amplifications. However, comparison of gene copies determined by FISH with copy numbers obtained with array-based CGH indicated that the latter sometimes underestimated the real gene copy numbers, similar to what has been noted by others using the same techniques (15, 25).

In total 60 amplifications at 32 different genomic locations were detected. These ranged from 0.4 to 38.1 Mb in size and showed log2 test:reference ratios between 1.0 and 3.5. The nature of the amplicons varied from normally distributed shapes with imprecise borders to amplicons with well-defined borders and large log2 ratio differences for amplified clones compared with their nonamplified neighbors. Some profiles showed clear stepwise increments in the copy number ratios within amplicons, such as the 12p amplicons in LPC4p and DANG, or the division of larger amplicons into smaller segments, as observed in the 8q amplification in HupT3 (Fig. 5). The presence of modular structures could imply that the mechanisms causing these amplifications are related to breakage-fusion-bridge cycles, which are known to generate block-by-block amplified sequences (29). In fact, in a previous FISH analysis of the 12p amplicon in LPC5m we demonstrated the presence of such repeated segments resulting in a high-level amplification (13).

A total of 21 of the 31 investigated cell lines had acquired detectable local amplifications. These were found in 10 of the LPC cell lines and 11 of the established cell lines. The number of amplicons per case ranged from 1 to 6, with the exception of HupT3, in which 10 local amplifications were observed. No differences were seen between the low-passage and the established cell lines with respect to the number of amplifications or in the distribution of amplicon sizes. Thus, the low-passage and established cell lines were comparable with respect to amplicon contents.

Segments that showed a reduction in gene copy number ratios corresponding to at least −3 SD for the respective hybridizations for a minimum of two consecutive BAC clones were selected as potential homozygous deletion targets. In total, 13 candidate segments were found, of which 7 were verified by STS analysis, all previously unreported in pancreatic cancer. Only one of these contained a previously described putative tumor suppressor gene, DEC1(30). Although the applied criteria identified several deletions, this analysis was by no means exhaustive. This was apparent by the inconsistency between PCR and the array-based CGH in detecting homozygous deletions of CDKN2A and SMAD4. However, in several cases where CDKN2A was homozygously deleted, the single BAC clone CTD-2097K16, which covers the tumor suppressor gene, showed log2 ratios below −3 SD. Moreover, because the sizes of homozygous deletions are expected to be quite small, a higher density array may be needed to fully appreciate the extent of homozygous deletions.

Thirteen cases were found to have at least one homozygous deletion. Of these, the majority had one to two deletions, the exception being BxPC-3, in which at least six were present. BxPC-3 was also exceptional in that no local amplifications were detected in this cell line, only whole chromosome-arm changes. This points to the possibility that genomic/karyotypic evolution may be directed toward many local amplifications and few homozygous deletions, as in the case of HupT3, or toward many homozygous deletions and few or no local amplifications, as in the case of BxPC-3. On the basis of cytogenetic investigations, a similar division of karyotypic pathways into those dominated by gains and those dominated by losses is seen in several solid tumor types (31, 32, 33). Although BxPC-3 and HupT3 may represent extremes, these could, at the molecular level and in progressed tumor types such as pancreatic carcinomas, correspond to tumor subgroups that may show a pronounced amplification or deletion phenotype.

Several regions of the genome were involved in recurrent amplifications. Among these, the amplicons within 8q23–24 and 12p11–12 were significantly more clustered than amplicons on other chromosome arms. The occurrence in pancreatic cancer of 8q amplifications and, more specifically of MYC, has been reported previously (34, 35). In the present study, amplifications of distal 8q were identified in eight cases that formed two apparent groups. The first group included five cases, which showed larger amplified regions, 17–33 Mb in size. The commonly amplified region in these cases was 15.5 Mb in size, a chromosomal section including almost 600 genes/UniGene clusters. The second category included three cases with small amplicons, ≤3.1 Mb in size. Their recurrently amplified segment was determined to a size of 1.2 Mb, a sequence containing MYC as the only fully annotated gene.

The 12p amplifications were homogeneous with regard to location, i.e., the majority of the amplicons were located within 12p11–12. A commonly amplified 3.2-Mb region was identified that contained 147 genes/UniGene clusters, including KRAS2 and PPFIBP1. We had previously characterized this region by use of FISH and STS mapping techniques, and the results from the present investigation confirm our previous findings (13). However, because the arrays used in the present study provide a higher resolution and better coverage than the previously used mapping techniques, the amplicons in LPC4p, LPC5m, DANG, and SU.86.86 were found to extend more distally, and to overlap with the commonly amplified 12p segment described in testicular germ cell tumors (23, 36), and to include the DAD-R, SOX5, and EKI1 genes (37).

Amplifications in other locations of the genome were more complex with regard to amplicon organization. The analysis of 6p revealed two commonly amplified regions, a larger region in 6p21.3–22.1 and a smaller region in 6p21.1, including CCND3. Within 7q, two larger recurrently amplified regions were mapped. These included HGF and HGFR (MET), respectively, which have been shown to be simultaneously overexpressed in pancreatic cancer (38), suggesting autocrine cell signaling. The two genes were, however, not concomitantly amplified in any of the cases in this investigation. The AKT2 oncogene, located at 19q13.2, is amplified in 10% of pancreatic carcinomas (39). In the present study five cases showed amplifications of 19q, two of which included AKT2.

The amplifications located in 8p and 18q, respectively, showed short or no overlaps and were significantly smaller than the average amplicons. The small sizes of these amplicons are interesting in light of the fact that 8p and 18q are chromosome arms frequently lost in pancreatic cancer (3, 40). This could indicate that 8p and 18q contain large stretches of sequences that are preferentially lost in the tumor environment and that this restricts the sizes of gained/amplified 8p and 18q segments, a hypothesis also supported by the two amplicons at 9p (1.1 Mb) and 15q (2.1 Mb), which also show frequent loss in pancreatic carcinoma. Nevertheless, the sizes of these amplicons have most probably made them undetectable, or uncertain, in previous CGH analyses. Notably, the 18q amplifications in LPC11p/LPC11m were located close to or at the deletion breakpoint. Similar colocalization of genomic amplifications and deletion breakpoints has been described in head and neck cancer (41).

More than half of the detected amplicons were smaller than 5 Mb, and one-third were smaller than 2.5 Mb, indicating that the majority of amplifications are less than a cytogenetic band in size. Furthermore, because the detection level of amplified segments by the current arrays is estimated to be 0.5–1.0 Mb, several even smaller amplicons may have remained undetected in the present investigation. Pancreatic carcinomas therefore contain many more amplifications than previously appreciated. This and the finding of homozygous deletions in at least 13 of the cases indicate a high level of genomic plasticity in pancreatic carcinomas.

Grant support: Swedish Cancer Society, the American Cancer Society, European Commission COST Action B19, the Crafoord Foundation, the John and Augusta Persson Foundation, the Royal Physiographic Society, and the Erik Philip-Sörensen Foundation.

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

Requests for reprints: Markus Heidenblad, Department of Clinical Genetics, Lund University Hospital, SE-221 85 Lund, Sweden. Phone: 46(46) 17 33 98; Fax: 46(46) 13 10 61; E-mail: [email protected]

3

The data sets are available at http://www.klingen.lu.se/E/research.

4

http://swegene.onk.lu.se.

5

http://genome.ucsc.edu.

6

http://www.ncbi.nih.gov.

7

http://bacpac.chori.org/home.htm.

Fig. 1.

Array-based comparative genome hybridization copy number profiles for LPC11p generated from the bacterial artificial chromosome (left) and the cDNA (right) microarray hybridizations. Profiles derived from the latter are displayed as a moving average (symmetric four nearest neighbors). A, genome-wide profiles showing the normalized log2 ratios for test versus reference gene copy numbers for all chromosomes. The clones have been ordered from chromosome (Chr.) 1 to Y (indicated by vertical bars), and within each chromosome according to their position on the University of California Santa Cruz genome browser (July 2003 freeze). B, enlargements of the 18q copy number profiles from the two hybridizations. The profiles show log2 ratios for test versus reference copy numbers for clones displayed at their position (in Mb) according to the University of California Santa Cruz genome browser (July 2003 freeze). The profiles from both analyses clearly show a proximal gene copy number increase, followed by a distinct amplification peak, and a distal copy number decrease.

Fig. 1.

Array-based comparative genome hybridization copy number profiles for LPC11p generated from the bacterial artificial chromosome (left) and the cDNA (right) microarray hybridizations. Profiles derived from the latter are displayed as a moving average (symmetric four nearest neighbors). A, genome-wide profiles showing the normalized log2 ratios for test versus reference gene copy numbers for all chromosomes. The clones have been ordered from chromosome (Chr.) 1 to Y (indicated by vertical bars), and within each chromosome according to their position on the University of California Santa Cruz genome browser (July 2003 freeze). B, enlargements of the 18q copy number profiles from the two hybridizations. The profiles show log2 ratios for test versus reference copy numbers for clones displayed at their position (in Mb) according to the University of California Santa Cruz genome browser (July 2003 freeze). The profiles from both analyses clearly show a proximal gene copy number increase, followed by a distinct amplification peak, and a distal copy number decrease.

Close modal
Fig. 2.

Fluorescence in situ hybridization (FISH) validation experiments. A, array-based comparative genome hybridization copy number profiles showing 8p, 8q, and 20q in cases LPC3p (purple) and LPC13p (blue). The profiles of the log2 ratios for test versus reference gene copy numbers were generated from the hybridizations to cDNA clones, which within each chromosome arm are presented as a moving average (symmetric four nearest neighbors) and according to their positions (in Mb) on the University of California Santa Cruz genome browser (July 2003 freeze). Positions and colors of the bacterial (BAC) and P1-derived artificial chromosome (PAC) probes used for FISH analyses are shown below the respective profiles. B, FISH hybridization to LPC3p using the BAC probes RP11-79H13 located on 8p (red) and RP11-10G10 on 8q (green). C, FISH hybridization to LPC13p using the BAC probes RP11-79H13 located on 8p (red) and RP11-10G10 on 8q (green). D, FISH hybridization to LPC3p using the PAC probes RP4-724E16 located on 20q13.2 (red) and RP4-511B24 on 20q12 (green). E, FISH hybridization to LPC13p using the PAC probes RP4-724E16 located on 20q13.2 (red) and RP4-511B24 on 20q12 (green).

Fig. 2.

Fluorescence in situ hybridization (FISH) validation experiments. A, array-based comparative genome hybridization copy number profiles showing 8p, 8q, and 20q in cases LPC3p (purple) and LPC13p (blue). The profiles of the log2 ratios for test versus reference gene copy numbers were generated from the hybridizations to cDNA clones, which within each chromosome arm are presented as a moving average (symmetric four nearest neighbors) and according to their positions (in Mb) on the University of California Santa Cruz genome browser (July 2003 freeze). Positions and colors of the bacterial (BAC) and P1-derived artificial chromosome (PAC) probes used for FISH analyses are shown below the respective profiles. B, FISH hybridization to LPC3p using the BAC probes RP11-79H13 located on 8p (red) and RP11-10G10 on 8q (green). C, FISH hybridization to LPC13p using the BAC probes RP11-79H13 located on 8p (red) and RP11-10G10 on 8q (green). D, FISH hybridization to LPC3p using the PAC probes RP4-724E16 located on 20q13.2 (red) and RP4-511B24 on 20q12 (green). E, FISH hybridization to LPC13p using the PAC probes RP4-724E16 located on 20q13.2 (red) and RP4-511B24 on 20q12 (green).

Close modal
Fig. 3.

Genome-wide array-based comparative genome hybridization copy number profiles for LPC13p and DANG. The log2 ratio profiles for test versus reference gene copy numbers are based on the hybridizations to cDNA clones, which have been ordered from chromosome (Chr.) 1 to Y, and within each chromosome according to their position in the University of California Santa Cruz genome browser (July 2003 freeze). The profiles are displayed as a moving average (symmetric four nearest neighbors).

Fig. 3.

Genome-wide array-based comparative genome hybridization copy number profiles for LPC13p and DANG. The log2 ratio profiles for test versus reference gene copy numbers are based on the hybridizations to cDNA clones, which have been ordered from chromosome (Chr.) 1 to Y, and within each chromosome according to their position in the University of California Santa Cruz genome browser (July 2003 freeze). The profiles are displayed as a moving average (symmetric four nearest neighbors).

Close modal
Fig. 4.

Composite gene copy number profiles showing amplifications of 6p, 7q, 8p, 8q, 12p, and 19q. The log2 ratio profiles for test versus reference gene copy numbers are derived from the cDNA array hybridizations, and each panel demonstrates the copy number profiles of two different cases, displayed as a moving average (symmetric four nearest neighbors). Clone positions, ideograms, and amplicon bars, which provide an overview of all detected amplicons on the respective chromosome arms, are illustrated according to their positions/sizes (in Mb) in the University of California Santa Cruz genome browser (July 2003 freeze).

Fig. 4.

Composite gene copy number profiles showing amplifications of 6p, 7q, 8p, 8q, 12p, and 19q. The log2 ratio profiles for test versus reference gene copy numbers are derived from the cDNA array hybridizations, and each panel demonstrates the copy number profiles of two different cases, displayed as a moving average (symmetric four nearest neighbors). Clone positions, ideograms, and amplicon bars, which provide an overview of all detected amplicons on the respective chromosome arms, are illustrated according to their positions/sizes (in Mb) in the University of California Santa Cruz genome browser (July 2003 freeze).

Close modal
Fig. 5.

Fluorescence in situ hybridization validation experiments. A, hybridization of the bacterial artificial chromosome (BAC) probes RP11-136C6 (red) and RP11-174E21 (green) to the 6p21 amplification in LPC3p. The image shows the metaphase chromosomes of a near-triploid genome containing two normal chromosomes 6 and two marker chromosomes demonstrating high-level amplification of the BAC clone RP11-136C6, located at the center of the amplicon. B, hybridization to LPC11p using the BAC probes RP11-79G13 (yellow), RP11-91K12 (red), and RP11-89M10 (green), located in proximal to distal order within 18q. The inset shows metaphase chromosomes of a normal chromosome 18, for which all three signals are present, and a marker chromosome showing presence of the proximal probe (RP11-79G13; yellow), amplification of the central probe (RP11-91K12; red), and deletion of the distal probe (RP11-89M10; green). The interphase nucleus below the inset illustrates the presence of two normal chromosomes 18, seen as three-color clusters to the upper left and right, and three marker chromosomes seen as red and yellow clusters located between and below the normal chromosomes 18.

Fig. 5.

Fluorescence in situ hybridization validation experiments. A, hybridization of the bacterial artificial chromosome (BAC) probes RP11-136C6 (red) and RP11-174E21 (green) to the 6p21 amplification in LPC3p. The image shows the metaphase chromosomes of a near-triploid genome containing two normal chromosomes 6 and two marker chromosomes demonstrating high-level amplification of the BAC clone RP11-136C6, located at the center of the amplicon. B, hybridization to LPC11p using the BAC probes RP11-79G13 (yellow), RP11-91K12 (red), and RP11-89M10 (green), located in proximal to distal order within 18q. The inset shows metaphase chromosomes of a normal chromosome 18, for which all three signals are present, and a marker chromosome showing presence of the proximal probe (RP11-79G13; yellow), amplification of the central probe (RP11-91K12; red), and deletion of the distal probe (RP11-89M10; green). The interphase nucleus below the inset illustrates the presence of two normal chromosomes 18, seen as three-color clusters to the upper left and right, and three marker chromosomes seen as red and yellow clusters located between and below the normal chromosomes 18.

Close modal
Fig. 6.

Homozygous deletion results. A, copy number profiles from hybridizations to bacterial artificial chromosome (BAC) clones, showing log2 ratios of test to reference gene copy numbers along chromosome 9 for cases BxPC-3 and HupT4. Clones in profiles and the ideograms below are displayed according to their positions/sizes (in Mb) in the University of California Santa Cruz genome browser (July 2003 freeze). The horizontal bars indicate the −3 SD thresholds in the respective hybridizations and the BAC clones at the large copy number decreases, the sequences from which sequence tagged sites (STSs) were selected for verifications. B, PCR verifications of homozygous deletions. The image shows agarose gel photos of ACTB, the control for functional PCR, and of STSs deleted in at least one cell line for all cases displaying at least one homozygous deletion. To the right, the names of the STS markers are given, as are the BAC clones or genes they were derived from, and the cytogenetic bands in which they are located according to the University of California Santa Cruz genome browser (July 2003 freeze).

Fig. 6.

Homozygous deletion results. A, copy number profiles from hybridizations to bacterial artificial chromosome (BAC) clones, showing log2 ratios of test to reference gene copy numbers along chromosome 9 for cases BxPC-3 and HupT4. Clones in profiles and the ideograms below are displayed according to their positions/sizes (in Mb) in the University of California Santa Cruz genome browser (July 2003 freeze). The horizontal bars indicate the −3 SD thresholds in the respective hybridizations and the BAC clones at the large copy number decreases, the sequences from which sequence tagged sites (STSs) were selected for verifications. B, PCR verifications of homozygous deletions. The image shows agarose gel photos of ACTB, the control for functional PCR, and of STSs deleted in at least one cell line for all cases displaying at least one homozygous deletion. To the right, the names of the STS markers are given, as are the BAC clones or genes they were derived from, and the cytogenetic bands in which they are located according to the University of California Santa Cruz genome browser (July 2003 freeze).

Close modal
Table 1

Summary of amplicons, ordered according to genomic localization

CaseCytogenetic regionap-Terminal borderabMb positionaq-Terminal borderabMb positionaSizea (Mb)
LPC5m 2p25.1 342211 (LOC129642, Hs. 90797) 9.02 624627 (RRM2, Hs. 226390) 10.29 1.27 
LPC15p 4p15.1-4q12 RP11-81H11 31.90 814443 (MGC3232, Hs. 8715) 57.75 25.85 
SW1990 6p21.32-6p22.2 665373 (TRIM38, Hs. 511746) 26.09 773246 (RING1, Hs. 202430) 33.23 7.14 
HupT3 6p21.32-6p22.1 1687138 (HIST1H2AM, Hs. 134999) 27.97 153411 (HLA-DRA, Hs. 409805) 32.48 4.51 
LPC3p 6p21.1-6p21.2 180082 (MDGA1, Hs. 403921) 37.65 795875 (Hs. 92858) 42.18 4.53 
SU.86.86 6p21.1 2321341 (PGC, Hs. 1867) 41.75 399532 (POLR1C, Hs. 5409) 43.54 1.79 
LPC15p 6q23.2-6q24.2 1493085 (CRSP3, Hs. 29679) 131.89 376764 (UTRN, Hs. 250607) 145.15 13.26 
HupT3 6q23.2-6q23.3 796531 (Hs. 453381) 132.20 1880757 (MAP3K5, Hs. 151988) 136.86 4.66 
DANG 7p12.1-7p13 1620818 (LOC285961, Hs. 31818) 45.50 RP11-91C9 52.17 6.67 
LPC14p 7q11.21-7q22.1 RP11-91C6 64.99 2321510 (TAF9, Hs. 289950) 99.32 34.33 
LPC12m 7q21.11-7q21.12 2445159 (SEMA3C, Hs. 171921) 79.98 287843 (GRM3, Hs. 512145) 86.11 6.13 
Hs700T 7q21.11-7q31.32 RP11-88D24 84.77 2028912 (WASL, Hs. 182469) 122.89 38.12 
AsPC-1 7q21.3-7q22.1 726086 (TFPI2, Hs. 438231) 93.13 743041 (G10, Hs. 380233) 98.62 5.49 
DANG 7q22.3-7q32.2 RP11-61C18 106.46 1412245 (CPA2, Hs. 89717) 129.47 23.01 
DANG 8p23.1 RP11-262B15 9.99 1573039 (CTSB, Hs. 135226) 11.74 1.75 
LPC11p 8p12 RP11-91P13 34.02 RP11-210F15 36.45 2.43 
LPC11m 8p12 RP11-91P13 34.02 RP11-210F15 36.45 2.43 
DANG 8p11.23-8p12 1626197 (Hs. 13861) 36.85 813188 (TACC1, Hs. 279245) 38.72 1.87 
HupT3 8p11.21 RP11-262I23 39.94 1663512 (POLB, Hs. 180107) 41.82 1.88 
LPC12m 8q23.3-8q24.3 1641243 (Hs. 130970) 112.73 243405 (MGC9718, Hs. 78547) 146.12 33.39 
LPC15p 8q23.3-8q24.3 30207 (Hs. 91381) 114.12 243405 (MGC9718, Hs. 78547) 146.12 32.00 
LPC6p 8q24.11-8q24.23 1557678 (EIF3S3, Hs. 127149) 117.73 462762 (Hs. 269586) 136.50 18.77 
HupT3 8q24.11-8q24.23 855728 (Hs. 187820) 118.89 RP11-122H7 138.27 19.38 
Hs700T 8q24.12-8q24.23 1055146 (Hs. 112981) 121.02 RP11-122H7 138.27 17.25 
LPC11p 8q24.21 RP11-128G18 127.64 70384 (Hs. 100425) 128.79 1.15 
LPC11m 8q24.21 RP11-128G18 127.64 RP11-79E8 130.72 3.08 
CFPAC 8q24.21 RP11-128G18 127.64 RP11-79E8 130.72 3.08 
HupT3 9p13.2-9p13.3 1376853 (CD72, Hs. 116481) 35.60 1517595 (MELK, Hs. 184339) 36.66 1.08 
DANG 9q21.12-9q21.13 213548 (Hs. 494167) 70.60 396006 (Hs. 291510) 72.82 2.22 
LPC12m 10q21.1-10q21.3 50982 (Hs. 407535) 53.40 RP11-280H23 68.62 15.22 
HupT3 10q22.3 RP11-21B16 78.83 450924 (Hs. 433072) 80.90 2.07 
LPC4p 11p11.2 RP11-206I1 45.23 1563147 (Hs. 192035) 47.08 1.85 
HupT3 11q14.1 RP11-91A3 80.18 RP11-13I14 95.17 14.99 
PaTu8902c 12p13.31-12pter RP11-110K11 0.17 RP11-166G2 5.61 5.44 
SU.86.86 12p11.21-12p13.1 120098 (Hs. 131933) 14.55 RP11-56J24 31.96 17.41 
DANG 12p11.22-12p12.3 897262 (DAT1, Hs. 301914) 16.59 503841 (PTHLH, Hs. 89626) 28.01 11.62 
LPC5m 12p11.22-12p12.1 2563391 (LDHB, Hs. 234489) 21.68 RP11-100P18 30.12 8.44 
LPC4pc 12cen-12p12.1 RP11-59N23 21.82 RP11-88P4 33.53 11.71 
LPC11p 12p11.1-12p12.1 1470169 (BCAT1, Hs. 438993) 24.86 RP11-88P4 33.36 8.50 
LPC11m 12p11.1-12p12.1 RP11-100C20 24.94 RP11-88P4 33.36 8.42 
HupT3 12q23.3-12q24.12 RP11-81H23 106.65 294196 (LNK, Hs. 13131) 110.30 3.65 
HupT3 12q24.31-12q24.32 1569801 (RNF10, Hs. 487883) 121.15 RP11-284P8 125.57 4.42 
LPC10m 15q26.1 795181 (LOC283761, Hs. 25314) 87.78 1055427 (Hs. 119280) 89.90 2.12 
LPC10m 17q23.2-17q23.3 1930780 (BZRAP1, Hs. 112499) 56.85 1628961 (Hs. 500449) 61.74 4.89 
HupT3 17q23.2-17q24.3 810948 (Hs. 439144) 60.50 321488 (CDC42EP4, Hs. 3903) 71.88 11.28 
LPC6p 18q11.1-18q11.2 RP11-96C18 16.98 854891 (Hs. 116770) 18.86 1.88 
PaTu8988T 18q11.2 RP11-18K7 18.07 51920 (OSBPL1A, Hs. 415753) 20.00 1.93 
PaTu8988S 18q11.2 RP11-18K7 18.07 RP11-94H23 20.25 2.18 
LPC11p 18q12.2-18q12.3 RP11-104N11 33.61 1031900 (Hs. 464986) 40.16 6.55 
LPC11m 18q12.2-18q12.3 RP11-104N11 33.61 RP11-19L3 42.18 8.57 
SU.86.86 19q13.12-19q13.2 47518 (UBA2, Hs. 511739) 39.65 1631209 (CYP2B7, Hs. 415794) 46.14 6.49 
PANC-1 19q13.2 2063982 (KCNK6, Hs. 240395) 43.51 257008 (PLD3, Hs. 257008) 45.56 2.05 
PaTu8988T 19q13.2 450912 (SIRT2, Hs. 375214) 44.06 853151 (RPS16, Hs. 397609) 44.62 0.56 
LPC6p 19q13.2 756450 (PAK4, Hs. 20447) 44.36 810558 (PSMC4, Hs. 211594) 45.17 0.81 
LPC3p 19q13.2-19q13.32 RP11-208I3 47.75 290536 (LOC229344, Hs. 374285) 51.07 3.32 
SU.86.86 19q13.33 795439 (NUP62, Hs. 437023) 55.11 2017860 (NAP1, Hs. 512843) 55.55 0.44 
LPC3p 20q13.2 1556872 (ZFP64, Hs. 504892) 51.39 RP11-172C21 54.56 3.17 
PaTu8988Tc 20q13.2-20qter RP11-10D18 51.33 2494000 (MYT1, Hs. 279562) 63.63 12.30 
LPC4p 22q13.1-22q13.2 283375 (CACNA1I, Hs. 125116) 38.33 RP11-258N5 41.31 2.98 
LPC11p Xq28 1631546 (MAGEA6, Hs. 441113) 150.49 134476 (SYBL, Hs. 24167) 153.60 3.11 
CaseCytogenetic regionap-Terminal borderabMb positionaq-Terminal borderabMb positionaSizea (Mb)
LPC5m 2p25.1 342211 (LOC129642, Hs. 90797) 9.02 624627 (RRM2, Hs. 226390) 10.29 1.27 
LPC15p 4p15.1-4q12 RP11-81H11 31.90 814443 (MGC3232, Hs. 8715) 57.75 25.85 
SW1990 6p21.32-6p22.2 665373 (TRIM38, Hs. 511746) 26.09 773246 (RING1, Hs. 202430) 33.23 7.14 
HupT3 6p21.32-6p22.1 1687138 (HIST1H2AM, Hs. 134999) 27.97 153411 (HLA-DRA, Hs. 409805) 32.48 4.51 
LPC3p 6p21.1-6p21.2 180082 (MDGA1, Hs. 403921) 37.65 795875 (Hs. 92858) 42.18 4.53 
SU.86.86 6p21.1 2321341 (PGC, Hs. 1867) 41.75 399532 (POLR1C, Hs. 5409) 43.54 1.79 
LPC15p 6q23.2-6q24.2 1493085 (CRSP3, Hs. 29679) 131.89 376764 (UTRN, Hs. 250607) 145.15 13.26 
HupT3 6q23.2-6q23.3 796531 (Hs. 453381) 132.20 1880757 (MAP3K5, Hs. 151988) 136.86 4.66 
DANG 7p12.1-7p13 1620818 (LOC285961, Hs. 31818) 45.50 RP11-91C9 52.17 6.67 
LPC14p 7q11.21-7q22.1 RP11-91C6 64.99 2321510 (TAF9, Hs. 289950) 99.32 34.33 
LPC12m 7q21.11-7q21.12 2445159 (SEMA3C, Hs. 171921) 79.98 287843 (GRM3, Hs. 512145) 86.11 6.13 
Hs700T 7q21.11-7q31.32 RP11-88D24 84.77 2028912 (WASL, Hs. 182469) 122.89 38.12 
AsPC-1 7q21.3-7q22.1 726086 (TFPI2, Hs. 438231) 93.13 743041 (G10, Hs. 380233) 98.62 5.49 
DANG 7q22.3-7q32.2 RP11-61C18 106.46 1412245 (CPA2, Hs. 89717) 129.47 23.01 
DANG 8p23.1 RP11-262B15 9.99 1573039 (CTSB, Hs. 135226) 11.74 1.75 
LPC11p 8p12 RP11-91P13 34.02 RP11-210F15 36.45 2.43 
LPC11m 8p12 RP11-91P13 34.02 RP11-210F15 36.45 2.43 
DANG 8p11.23-8p12 1626197 (Hs. 13861) 36.85 813188 (TACC1, Hs. 279245) 38.72 1.87 
HupT3 8p11.21 RP11-262I23 39.94 1663512 (POLB, Hs. 180107) 41.82 1.88 
LPC12m 8q23.3-8q24.3 1641243 (Hs. 130970) 112.73 243405 (MGC9718, Hs. 78547) 146.12 33.39 
LPC15p 8q23.3-8q24.3 30207 (Hs. 91381) 114.12 243405 (MGC9718, Hs. 78547) 146.12 32.00 
LPC6p 8q24.11-8q24.23 1557678 (EIF3S3, Hs. 127149) 117.73 462762 (Hs. 269586) 136.50 18.77 
HupT3 8q24.11-8q24.23 855728 (Hs. 187820) 118.89 RP11-122H7 138.27 19.38 
Hs700T 8q24.12-8q24.23 1055146 (Hs. 112981) 121.02 RP11-122H7 138.27 17.25 
LPC11p 8q24.21 RP11-128G18 127.64 70384 (Hs. 100425) 128.79 1.15 
LPC11m 8q24.21 RP11-128G18 127.64 RP11-79E8 130.72 3.08 
CFPAC 8q24.21 RP11-128G18 127.64 RP11-79E8 130.72 3.08 
HupT3 9p13.2-9p13.3 1376853 (CD72, Hs. 116481) 35.60 1517595 (MELK, Hs. 184339) 36.66 1.08 
DANG 9q21.12-9q21.13 213548 (Hs. 494167) 70.60 396006 (Hs. 291510) 72.82 2.22 
LPC12m 10q21.1-10q21.3 50982 (Hs. 407535) 53.40 RP11-280H23 68.62 15.22 
HupT3 10q22.3 RP11-21B16 78.83 450924 (Hs. 433072) 80.90 2.07 
LPC4p 11p11.2 RP11-206I1 45.23 1563147 (Hs. 192035) 47.08 1.85 
HupT3 11q14.1 RP11-91A3 80.18 RP11-13I14 95.17 14.99 
PaTu8902c 12p13.31-12pter RP11-110K11 0.17 RP11-166G2 5.61 5.44 
SU.86.86 12p11.21-12p13.1 120098 (Hs. 131933) 14.55 RP11-56J24 31.96 17.41 
DANG 12p11.22-12p12.3 897262 (DAT1, Hs. 301914) 16.59 503841 (PTHLH, Hs. 89626) 28.01 11.62 
LPC5m 12p11.22-12p12.1 2563391 (LDHB, Hs. 234489) 21.68 RP11-100P18 30.12 8.44 
LPC4pc 12cen-12p12.1 RP11-59N23 21.82 RP11-88P4 33.53 11.71 
LPC11p 12p11.1-12p12.1 1470169 (BCAT1, Hs. 438993) 24.86 RP11-88P4 33.36 8.50 
LPC11m 12p11.1-12p12.1 RP11-100C20 24.94 RP11-88P4 33.36 8.42 
HupT3 12q23.3-12q24.12 RP11-81H23 106.65 294196 (LNK, Hs. 13131) 110.30 3.65 
HupT3 12q24.31-12q24.32 1569801 (RNF10, Hs. 487883) 121.15 RP11-284P8 125.57 4.42 
LPC10m 15q26.1 795181 (LOC283761, Hs. 25314) 87.78 1055427 (Hs. 119280) 89.90 2.12 
LPC10m 17q23.2-17q23.3 1930780 (BZRAP1, Hs. 112499) 56.85 1628961 (Hs. 500449) 61.74 4.89 
HupT3 17q23.2-17q24.3 810948 (Hs. 439144) 60.50 321488 (CDC42EP4, Hs. 3903) 71.88 11.28 
LPC6p 18q11.1-18q11.2 RP11-96C18 16.98 854891 (Hs. 116770) 18.86 1.88 
PaTu8988T 18q11.2 RP11-18K7 18.07 51920 (OSBPL1A, Hs. 415753) 20.00 1.93 
PaTu8988S 18q11.2 RP11-18K7 18.07 RP11-94H23 20.25 2.18 
LPC11p 18q12.2-18q12.3 RP11-104N11 33.61 1031900 (Hs. 464986) 40.16 6.55 
LPC11m 18q12.2-18q12.3 RP11-104N11 33.61 RP11-19L3 42.18 8.57 
SU.86.86 19q13.12-19q13.2 47518 (UBA2, Hs. 511739) 39.65 1631209 (CYP2B7, Hs. 415794) 46.14 6.49 
PANC-1 19q13.2 2063982 (KCNK6, Hs. 240395) 43.51 257008 (PLD3, Hs. 257008) 45.56 2.05 
PaTu8988T 19q13.2 450912 (SIRT2, Hs. 375214) 44.06 853151 (RPS16, Hs. 397609) 44.62 0.56 
LPC6p 19q13.2 756450 (PAK4, Hs. 20447) 44.36 810558 (PSMC4, Hs. 211594) 45.17 0.81 
LPC3p 19q13.2-19q13.32 RP11-208I3 47.75 290536 (LOC229344, Hs. 374285) 51.07 3.32 
SU.86.86 19q13.33 795439 (NUP62, Hs. 437023) 55.11 2017860 (NAP1, Hs. 512843) 55.55 0.44 
LPC3p 20q13.2 1556872 (ZFP64, Hs. 504892) 51.39 RP11-172C21 54.56 3.17 
PaTu8988Tc 20q13.2-20qter RP11-10D18 51.33 2494000 (MYT1, Hs. 279562) 63.63 12.30 
LPC4p 22q13.1-22q13.2 283375 (CACNA1I, Hs. 125116) 38.33 RP11-258N5 41.31 2.98 
LPC11p Xq28 1631546 (MAGEA6, Hs. 441113) 150.49 134476 (SYBL, Hs. 24167) 153.60 3.11 
a

Map positions and cytogenetic locations are based on data available through the University of California Santa Cruz genome browser (July 2003 freeze).

b

For amplicon borders, the closest positions of neighboring unamplified clones have been used. The boundaries are based on a composite analysis of the bacterial artificial chromosome and cDNA arrays and are denoted as the positions of either a bacterial artificial chromosome or an IMAGE clone. IMAGE clone numbers are accompanied with the UniGene clusters they are associated to (UniGene build 164) and gene symbols if accessible.

c

Because no unamplified clones next to chromosomal ends/centromeres could be detected in these amplifications, the most distal/proximal positions of the outermost amplified clones have been listed.

We thank Irene M. Janssen and Bodil Strömbeck for expert technical assistance, Huub Straatman for statistical support, and Srinivas Veerla for bioinformatics support.

1
Bardeesy N, DePinho RA Pancreatic cancer biology and genetics.
Nat Rev Cancer
,
2
:
897
-909,  
2002
.
2
Griffin CA, Hruban RH, Morsberger LA, et al Consistent chromosome abnormalities in adenocarcinoma of the pancreas.
Cancer Res
,
55
:
2394
-9,  
1995
.
3
Gorunova L, Höglund M, Andrén-Sandberg Å, et al Cytogenetic analysis of pancreatic carcinomas: intratumor heterogeneity and nonrandom pattern of chromosome aberrations.
Genes Chromosomes Cancer
,
23
:
81
-99,  
1998
.
4
Solinas-Toldo S, Wallrapp C, Müller-Pillasch F, Bentz M, Gress T, Lichter P Mapping of chromosomal imbalances in pancreatic carcinoma by comparative genomic hybridization.
Cancer Res
,
56
:
3803
-7,  
1996
.
5
Fukushige S, Waldman FM, Kimura M, et al Frequent gain of copy number on the long arm of chromosome 20 in human pancreatic adenocarcinoma.
Genes Chromosomes Cancer
,
19
:
161
-9,  
1997
.
6
Mahlamäki EH, Höglund M, Gorunova L, et al Comparative genomic hybridization reveals frequent gains of 20q, 8q, 11q, 12p, and 17q, and losses of 18q, 9p, and 15q in pancreatic cancer.
Genes Chromosomes Cancer
,
20
:
383
-91,  
1997
.
7
Curtis LJ, Li Y, Gerbault-Seureau M, et al Amplification of DNA sequences from chromosome 19q13.1 in human pancreatic cell lines.
Genomics
,
53
:
42
-55,  
1998
.
8
Ghadimi BM, Schröck E, Walker RL, et al Specific chromosomal aberrations and amplification of the AIBI nuclear receptor coactivator gene in pancreatic carcinomas.
Am J Pathol
,
154
:
525
-36,  
1999
.
9
Jonson T, Gorunova L, Dawiskiba S, et al Molecular analyses of the 15q and 18q SMAD genes in pancreatic cancer.
Genes Chromosomes Cancer
,
24
:
62
-71,  
1999
.
10
Abe T, Makino N, Furukawa T, et al Identification of three commonly deleted regions on chromosome arm 6q in human pancreatic cancer.
Genes Chromosomes Cancer
,
25
:
60
-4,  
1999
.
11
Kimura M, Abe T, Sunamura M, Matsuno S, Horii A Detailed deletion mapping on chromosome arm 12q in human pancreatic adenocarcinoma: identification of a 1-cM region of common allelic loss.
Genes Chromosomes Cancer
,
17
:
88
-93,  
1996
.
12
Hilgers W, Tang DJ, Sugar AY, Shekher MC, Hruban RH, Kern SE High-resolution deletion mapping of chromosome arm 1p in pancreatic cancer identifies a major consensus region at 1p35.
Genes Chromosomes Cancer
,
24
:
351
-5,  
1999
.
13
Heidenblad M, Jonson T, Mahlamäki EH, et al Detailed genomic mapping and expression analyses of 12p amplifications in pancreatic carcinomas reveal a 3.5-Mb target region for amplification.
Genes Chromosomes Cancer
,
34
:
211
-23,  
2002
.
14
Solinas-Toldo S, Lampel S, Stilgenbauer S, et al Matrix-based comparative genomic hybridization: biochips to screen for genomic imbalances.
Genes Chromosomes Cancer
,
20
:
399
-407,  
1997
.
15
Pinkel D, Segraves R, Sudar D, et al High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays.
Nat Genet
,
20
:
207
-11,  
1998
.
16
Albertson DG, Ylstra B, Segraves R, et al Quantitative mapping of amplicon structure by array CGH identifies CYP24 as a candidate oncogene.
Nat Genet
,
25
:
144
-6,  
2000
.
17
Bruder CE, Hirvela C, Tapia-Paez I, et al High resolution deletion analysis of constitutional DNA from neurofibromatosis type 2 (NF2) patients using microarray-CGH.
Hum Mol Genet
,
10
:
271
-82,  
2001
.
18
Wilhelm M, Veltman JA, Olshen AB, et al Array-based comparative genomic hybridization for the differential diagnosis of renal cell cancer.
Cancer Res
,
62
:
957
-60,  
2002
.
19
Veltman JA, Fridlyand J, Pejavar S, et al Array-based comparative genomic hybridization for genome-wide screening of DNA copy number in bladder tumors.
Cancer Res
,
63
:
2872
-80,  
2003
.
20
Cheung VG, Nowak N, Jang W, et al Integration of cytogenetic landmarks into the draft sequence of the human genome.
Nature (Lond)
,
409
:
953
-8,  
2001
.
21
Veltman JA, Schoenmakers EF, Eussen BH, et al High-throughput analysis of subtelomeric chromosome rearrangements by use of array-based comparative genomic hybridization.
Am J Hum Genet
,
70
:
1269
-76,  
2002
.
22
Veltman JA, Jonkers Y, Nuijten I, et al Definition of a critical region on chromosome 18 for congenital aural atresia by array CGH.
Am J Hum Genet
,
72
:
1578
-84,  
2003
.
23
Zafarana G, Grygalewicz B, Gillis AJ, et al 12p-amplicon structure analysis in testicular germ cell tumors of adolescents and adults by array CGH.
Oncogene
,
22
:
7695
-701,  
2003
.
24
Vissers LELM, de Vries BBA, Osoegawa K, et al Array-based comparative genomic hybridization for genome wide detection of submicroscopic chromosome abnormalities.
Am J Hum Genet
,
73
:
1261
-71,  
2003
.
25
Pollack JR, Perou CM, Alizadeh AA, et al Genome-wide analysis of DNA copy-number changes using cDNA microarrays.
Nat Genet
,
23
:
41
-6,  
1999
.
26
Saal LH, Troein C, Vallon-Christersson J, Gruvberger S, Borg Å, Peterson C BioArray Software Environment (BASE): a platform for comprehensive management and analysis of microarray data.
Genome Biol
,
3
: SOFTWARE0003 
2002
.
27
Yang YH, Dudoit S, Luu P, et al Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation.
Nucleic Acids Res
,
30
:
e15
2002
.
28
Höglund M, Johansson B, Pedersen-Bjergaard J, Marynen P, Mitelman F Molecular characterization of 12p abnormalities in hematologic malignancies deletion of KIP1, rearrangement of TEL, and amplification of. CCND2.
Blood
,
87
:
324
-30,  
1996
.
29
Gisselsson D, Pettersson L, Höglund M, et al Chromosomal breakage-fusion-bridge events cause genetic intratumor heterogeneity.
Proc Natl Acad Sci USA
,
97
:
5357
-62,  
2000
.
30
Nishiwaki T, Daigo Y, Kawasoe T, Nakamura Y Isolation and mutational analysis of a novel human cDNA, DEC1 (deleted in esophageal cancer 1), derived from the tumor suppressor locus in 9q32.
Genes Chromosomes Cancer
,
27
:
169
-76,  
2000
.
31
Höglund M, Säll T, Heim S, Mitelman F, Mandahl N, Fadl-Elmula I Identification of cytogenetic subgroups and karyotypic pathways in transitional cell carcinoma.
Cancer Res
,
61
:
8241
-6,  
2001
.
32
Höglund M, Gisselsson D, Hansen GB, Säll T, Mitelman F, Nilbert M Dissecting karyotypic patterns in colorectal tumors: two distinct but overlapping pathways in the adenoma-carcinoma transition.
Cancer Res
,
62
:
5939
-46,  
2002
.
33
Höglund M, Gisselsson D, Hansen GB, Säll T, Mitelman F Ovarian carcinoma develops through multiple modes of chromosomal evolution.
Cancer Res
,
63
:
3378
-85,  
2003
.
34
Armengol G, Knuutila S, Lluis F, Capella G, Miro R, Caballin MR DNA copy number changes and evaluation of MYC, IGF1R, and FES amplification in xenografts of pancreatic adenocarcinoma.
Cancer Genet Cytogenet
,
116
:
133
-41,  
2000
.
35
Mahlamäki EH, Bärlund M, Tanner M, et al Frequent amplification of 8q24, 11q, 17q, and 20q-specific genes in pancreatic cancer.
Genes Chromosomes Cancer
,
35
:
353
-8,  
2002
.
36
Mostert MC, Verkerk AJMH, van den Pol M, et al Identification of the critical region of 12p over-representation in testicular germ cell tumors of adolescents and adults.
Oncogene
,
16
:
2617
-27,  
1998
.
37
Zafarana G, Gillis AJ, van Gurp RJ, et al Coamplification of DAD-R, SOX5, and EKI1 in human testicular seminomas, with specific overexpression of DAD-R, correlates with reduced levels of apoptosis and earlier clinical manifestation.
Cancer Res
,
62
:
1822
-31,  
2002
.
38
Ebert M, Yokoyama M, Friess H, Buchler MW, Korc M Coexpression of the c-met proto-oncogene and hepatocyte growth factor in human pancreatic cancer.
Cancer Res
,
54
:
5775
-8,  
1994
.
39
Cheng JQ, Ruggeri B, Klein WM, et al Amplification of AKT2 in human pancreatic cancer cells and inhibition of AKT2 expression and tumorigenecity by antisense RNA.
Proc Natl Acad Sci USA
,
93
:
3636
-41,  
1996
.
40
Hahn SA, Seymour AB, Hoque ATMS, et al Allelotype of pancreatic adenocarcinoma using xenograft enrichment.
Cancer Res
,
55
:
4670
-5,  
1995
.
41
Jin Y, Höglund M, Jin C, et al FISH characterization of head and neck carcinomas reveals that amplification of band 11q13 is associated with deletion of distal 11q.
Genes Chromosomes Cancer
,
22
:
312
-20,  
1998
.

Supplementary data