Abstract

Genomic copy number changes are frequently found in cancers and they have been demonstrated to contribute to carcinogenesis; and it is widely accepted that tumors with microsatellite instability (MSI) are genetically stable and mostly diploid. In the present study we compared the copy number alterations and the gene-expression profiles of microsatellite stable (MSS) and MSI colorectal tumors. A total number of 31 fresh-frozen primary tumors (16 MSS and 15 MSI) were used. Twenty-eight samples (15 MSS and 13 MSI) were analyzed with metaphase comparative genomic hybridization (CGH), nine of which plus one additional sample (4 MSS and 6 MSI) were further analyzed by cDNA-based array-CGH. Gene expression analysis was performed with six samples [3 MSS and 3 MSI, four of these used in metaphase CGH (mCGH) analysis] to identify differentially expressed genes possibly located in the lost or amplified regions found by CGH, stressing the biological significance of copy number changes. Metaphase and array-CGH analysis of two colon cancer cell lines (HTC116 and SW480, reported as MSI and MSS archetypes) gave comparable results. Alterations found by mCGH in MSS tumors were +20, +8q, −8p and −18q. Interestingly, 1p22, 4q26 and 15q21 were found deleted preferentially in MSS tumors, while 22q13 was found gained in MSI tumors. The regions of alterations identified by array-CGH were gains at 8q24, 16q24.3 and 20q13, and the loss of 5q21, appearing in the both types of tumors. Gene expression analysis revealed genes with specific associations with the copy number changes of the corresponding genomic regions. As a conclusion, colorectal cancer is a heterogeneous disease, demonstrated by the genomic profiles of individual samples. However, our data shows that copy number changes do not occur exclusively in the MSS phenotypes.

Introduction

The majority of colorectal cancers (CRC) are considered sporadic, having developed from an accumulation of mutations through the course of a lifespan. Genetic instability is a defining characteristic of most of cancer cells. Two forms of genetic instability have been identified in sporadic colorectal tumors: chromosomal instability (CIN), and microsatellite instability (MSI). Briefly, >80% of CRCs correspond to CIN tumors. They are caused by mutations in genes that are required to maintain chromosomal stability, leading to the accumulation of gains and losses of chromosomes (aneuploidy) as well as by structural chromosomal rearrangements ( 13 ). On the other hand, MSI tumors show mutations in the mismatch repair genes leading to mutations in microsatellite DNA sequences and in genes close to them ( 4 , 5 ). These tumors usually show a stable, diploid karyotype. Nevertheless, recent studies have demonstrated that MSI tumors can also present genetic imbalances ( 6 ), and both forms of genetic instability can coexist in some colorectal primary tumors ( 7 ) and cell lines ( 8 , 9 ).

The molecular cause of genomic instability is one of the most active research areas in cancer biology. MSI tumors are characterized by mutations in proofreading genes coding for mismatch repair enzymes, such as MLH1, MSH2, PMS1 and PMS2 ( 10 , 11 ), and by silencing the hMLH1 promoter by methylation. Somatic mutations in APC (5q), DCC , SMAD2 and SMAD4 (18q), TP53 (17p), KRAS (12p) and β -catenin (3p) genes are involved in CRC tumorigenesis ( 12 , 13 ). Other genes related to CIN tumors have also been reported. Thus, genes affecting the double-strand breakage repair mechanisms, e.g. MRE11, ATM, ATR, BRCA1 and BRCA2 , and genes involved in ploidy maintenance and the chromosomal segregation, e.g. BUB1B ( 14 ) and SMAD4 have been demonstrated to be involved in the genomic stability of cancer cells.

Classical cytogenetics from primary colorectal tumors reported that the most frequent numerical changes were, in order of decreasing frequency, +7, −18, −14, +13 and −22. The gains of chromosome arms 8q, 17q, 20q, and the losses of 8p and 17p are also unbalanced chromosomal alterations previously identified ( 15 , 16 ). To overcome G-banding limitations, conventional comparative genomic hybridization (CGH) has been used to detect aneuploidies and chromosomal alterations in colorectal tumors ( 1720 ). In these studies, CRCs have often shown extra copies of chromosomes 1, 13, and 20 and of chromosome arms 7p and 8q, whereas chromosome 4 and chromosome arms 8p and 10q were frequently underrepresented. Cytogenetic/CGH data are documented in the Mitelman Database of Chromosome Aberrations in Cancer ( http://cgap.nci.nih.gov/Chromosomes/Mitelman ), and at www.helsinki.fi/cmg ( 21 ).

Recently, microarray technologies have provided new ways to explore genomic imbalances and different gene expression ( 22 , 23 ). Array-based CGH represents a magnificent tool to study gene dosage imbalances in solid tumors ( 24 ). Recent studies using bacterial artificial chromosome (BAC)-arrays covering the human genome at about 1 Mb resolution have been reported in CRC ( 2527 ). Consistent frequencies of copy number changes were identified, including gains of chromosomes/chromosome arms 20q, 8q and 13. Losses of chromosomes/chromosome arms 18q, 17p, 18p, 8p and 14q were frequently found, whereas minor losses were detected at 1p, 4q, 5q and 15q. A recent study established a genetic signature which could be used in clustering CRCs according to Dukes' stages ( 28 ). Moreover, six differentially expressed genes were found in CIN and MSI phenotypes cell line analysis ( 29 ).

The purpose of the present study was to investigate genomic imbalance differences between microsatellite stability (MSS) and MSI CRCs, using metaphase CGH (mCGH), array-CGH (aCGH) and expression arrays to assess gene expression profiles. Moreover, two cell lines, HCT116 (MSI archetype) and SW480 (CIN archetype), whose molecular karyotype have been described in our laboratory ( 30 , 31 ) were also analyzed by both CGH methods. Our aim was to generate a comprehensive picture of the genomic alterations that occur differentially in both forms of colorectal tumors, and to identify chromosomal regions that may contain candidate genes important for the development of cancer cells.

Materials and methods

Tumor samples and cell lines

A total of 31 fresh-frozen colorectal tumors (16 MSS and 15 MSI) from Dukes' stages A–D were used for this study. Twenty-eight of these samples were used in the mCGH analysis, of which nine were further analyzed by aCGH and four by expression arrays. One sample was used for aCGH only, and two for expression analysis only. Samples were kindly provided by Dr Miguel A. Peinado from Institut de Recerca Oncològica, Hospital Duran i Reynals and Dr S. Schwartz Jr from Hospital Universitari Vall d'Hebron, Barcelona. Clinical characteristics are listed in Table I . MSI status was analyzed as previously described ( 32 ). None of the patients with MSI tumors showed hereditary non-polyposis colorectal (HNPCC)-related family history according to the Amsterdam Criteria ( 33 ). Additionally, two CRC cell lines were used in the present study: HCT116 and SW480 (ATCC; Manassas, VA). Both were cultured using DMEM medium supplemented with fetal bovine serum 10% and glutamine 1%. Genomic DNA was extracted by the phenol–chloroform method. Before extraction, evaluation by the pathologist determined that the proportion of tumor cells was higher than 60%. Reference DNA for CGH hybridizations was extracted from peripheral blood lymphocytes of healthy individuals.

Table I.

Tumor collective

Tumor specimens 31 
Sex  
    Male 16 
    Female 15 
Age (years)  
    <50 
    50–70 18 
    >70 
Dukes' stage a  
    A 
    B 14 
    C 10 
    D 
MSI  
    − 16 
    + 15 
Tumor specimens 31 
Sex  
    Male 16 
    Female 15 
Age (years)  
    <50 
    50–70 18 
    >70 
Dukes' stage a  
    A 
    B 14 
    C 10 
    D 
MSI  
    − 16 
    + 15 
a

Dukes' stage was not determined in one case.

Table II.

Genomic alterations detected in human CRC cell lines by mCGH and aCGH

Cell Line  HCT116
 
    SW480
 
   
Alteration  Gains and amplifications
 
  Losses
 
  Gains and amplifications
 
  Losses
 
 
Chr
 
mCGH
 
aCGH
 
mCGH
 
aCGH
 
mCGH
 
aCGH
 
mCGH
 
aCGH
 
1p33pter – – 1p13.2p31   1p31.3qter 1p13.1p33 
 1q22 1q22 – 1q31.1q31.2   – 1q31.1q41 
       – 1q43 
  – 2q24.3q32.2 2q33pter – 2q37 2q37.1 
– 3p21.1 3p12 – 3p14.1p21 – 3p26 3p24.3pter 
     3q13.3qter 3q25.32 – 3q11.2q13.13 
     – 3q29   
  4q12q21.1 4q13.1   
   4q35 4q35.1     
– 5q35.1   5p15.2pter – 5p13.3q14 5p14.1q14.1 
     5q23.3qter 5q35.3   
– 6p23     6p12qter 
 – 6p21.31       
– 7q22.1     
 – 7q35   7pterq22 –   
8q21.1qter 8q24.2qter   8q24 8q24.2 8p22pter 8p21.3pter 
       8q11q23 8q13.3q23.1 
9q34 – – 9q21.2 9p21qter 9p13.3 9p21pter 9p21.1pter 
     – 9q34.2   
10 10q24qter 10q24.2q26.3     10 10 (10p12.1) 
11 11q13.3 11q13.1q13.3   11 11q12.3q13.3 – 11q14.3 
12   – 12p13.2 12pterq14 12pterq13.2 12q15qter – 
   – 12q22     
13   13q22 – 13 13q12.2   
14     14q23qter 14q24.3qter 14q12 14q12q22.3 
15 15q22.3 – – 15q13.3q21.3   15 15q21.3 
16 16p11p13.2 –       
 16q22qter 16q24.3       
17 17q 17q21.3qter   17 17 – 17q22 
18     18p – 18q 18q 
19 19 19q13.32q13.33   19q 19q13.12qter   
20 20q –   20 20p11.3qter   
21 21q22.3 –   21q –   
22 22q 22q12.3     – 22q11.1q11.2 
Cell Line  HCT116
 
    SW480
 
   
Alteration  Gains and amplifications
 
  Losses
 
  Gains and amplifications
 
  Losses
 
 
Chr
 
mCGH
 
aCGH
 
mCGH
 
aCGH
 
mCGH
 
aCGH
 
mCGH
 
aCGH
 
1p33pter – – 1p13.2p31   1p31.3qter 1p13.1p33 
 1q22 1q22 – 1q31.1q31.2   – 1q31.1q41 
       – 1q43 
  – 2q24.3q32.2 2q33pter – 2q37 2q37.1 
– 3p21.1 3p12 – 3p14.1p21 – 3p26 3p24.3pter 
     3q13.3qter 3q25.32 – 3q11.2q13.13 
     – 3q29   
  4q12q21.1 4q13.1   
   4q35 4q35.1     
– 5q35.1   5p15.2pter – 5p13.3q14 5p14.1q14.1 
     5q23.3qter 5q35.3   
– 6p23     6p12qter 
 – 6p21.31       
– 7q22.1     
 – 7q35   7pterq22 –   
8q21.1qter 8q24.2qter   8q24 8q24.2 8p22pter 8p21.3pter 
       8q11q23 8q13.3q23.1 
9q34 – – 9q21.2 9p21qter 9p13.3 9p21pter 9p21.1pter 
     – 9q34.2   
10 10q24qter 10q24.2q26.3     10 10 (10p12.1) 
11 11q13.3 11q13.1q13.3   11 11q12.3q13.3 – 11q14.3 
12   – 12p13.2 12pterq14 12pterq13.2 12q15qter – 
   – 12q22     
13   13q22 – 13 13q12.2   
14     14q23qter 14q24.3qter 14q12 14q12q22.3 
15 15q22.3 – – 15q13.3q21.3   15 15q21.3 
16 16p11p13.2 –       
 16q22qter 16q24.3       
17 17q 17q21.3qter   17 17 – 17q22 
18     18p – 18q 18q 
19 19 19q13.32q13.33   19q 19q13.12qter   
20 20q –   20 20p11.3qter   
21 21q22.3 –   21q –   
22 22q 22q12.3     – 22q11.1q11.2 

Amplifications are indicated in bold face.

Total RNA from six primary tumors (three MSS and three MSI) and from adjacent tissue with a normal histology, was isolated using the RNeasy Mini Kit according to the manufacturer's protocol (Qiagen; Hilden, Germany). Two samples included in the gene expression profiling were not analyzed by CGH. Prior to microarray hybridization, the integrity of total RNA was confirmed using a 2100 Bioanalyzer (Agilent Technologies; Palo Alto, CA).

Metaphase CGH

Metaphase CGH ( 34 ) was performed as previously described ( 35 ). Slides were analyzed using the Isis Imaging System, Metasystems software (Altlussheim, Germany). Ratio values >1.25 and <0.75 were considered as chromosomal gains and losses, respectively. Values were log 2 transformed for plotting purposes. Sex chromosomes were not taken into account in the analysis.

CGH on cDNA microarrays (aCGH)

Array-based CGH experiments were conducted as previously described on commercial cDNA microarrays (Human 1.0; Agilent Technologies) containing clones of approximately 13 000 genes ( 3638 ). Of these clones 982 had replicates on the array. Briefly, 20 μg of reference and sample DNA were digested with Alu1 and Rsa1 restriction enzymes (Life Technologies; Rockville, MD). Digested samples were purified using the basic phenol–chloroform method. Labeling was performed as previously described ( 23 ). Briefly, six μg of digested tumor and reference DNA were labeled with Cy3-dUTP and Cy5-dUTP (Amersham Biosciences; Piscataway, NJ), respectively, by RadPrime DNA labeling system (Gibco BRL; Gaithersburg, MD). Hybridization and post-hybridization washes were carried out as previously described ( 37 ). Slides were scanned using an Agilent laser confocal scanner (G2565AA). Microarray images were analyzed with the Feature Extraction software (version 6.1.1, Agilent Technologies) using the locally weighted linear-regression curve fit option. Only those probes with a known position in the genome (10 080 probes) according to genomic alignment information available at the University of California at Santa Cruz (UCSC) Genome Browser Database ( http://genome.ucsc.edu/ , April 2003 freeze) were selected. Filters, if any, were applied to the aCGH datasets. However, given the high density of spots per chromosome on the aCGH, visual interpretation of known genomic lesions in plots of raw data was obscured by even a small percentage of the outlier probes. Therefore, we computed LOWESS smoothing lines in the data with an ‘f’ span equal to 0.2 ( 39 ). These parameters were determined by comparisons between mCGH and aCGH results obtained in the laboratory. To identify statistically significant transitions in the copy number, we defined the variation of the copy number as log 2 ±2 SD of the middle 50% quantile of data. Also, we used a developed statistical segmentation algorithm termed circular binary segmentation (CBS), which parses the probe ratio data into segments of similar mean after taking the variance into account. The algorithm works by analyzing one chromosome at a time and, within that chromosome, recursively identifying the best possible segmentation. Each proposed split is accepted or rejected based on the probability that the difference in mean could have arisen by chance. This probability is determined using a randomization method ( 40 ).

Gene expression microarrays

For gene expression profiling, GeneChip Human Genome U133 plus 2 oligonucleotide microarray (Affymetrix; Santa Clara, CA) was used, which contained 47 000 probe sets. Target cDNA preparations from total RNA, hybridization on the microarray, washing and staining with the antibody amplification procedure, and scanning were all carried out according to the manufacturer's instructions (Affymetrix). Arrays were scanned using an Affymetrix GeneChip Scanner 3000 and analyzed with Affymetrix Microarray Suite 5.0 (MAS 5.0). Detection values (present, marginal or absent) were determined by default parameters, and signal values were scaled by global methods to a target value of 100. Minimal quality control parameters for inclusion in the study included present calls >30% and a GAPDH 3′/5′ ratio <3. To process and normalize U133plus2 Affymetrix chips, and the RMA (Robust Multichip Averaging) algorithm were used ( 41 ). Differentially expressed genes between MSS and MSI were detected computing the B statistic ( 42 ). All these computations were performed using the Bioconductor package ( 43 ).

Results

mCGH

Metaphase CGH analysis was performed for 28 colorectal tumors (15 MSS and 13 MSI) and two cell lines (HCT116 and SW480). The chromosomal imbalances are represented in Figure 1 . The mean of chromosomal imbalances for MSS tumors (13 ± 8.46) was higher than the mean for MSI tumors (9 ± 4.58). Nevertheless, these differences were not statistically significant (0.05 < P < 0.1, Mann–Whitney U test). Only one primary tumor belonging to the MSI group did not show any chromosomal imbalance. Interestingly, three tumors included in the MSS group displayed <25% (≤4) chromosomal imbalances according to the mean, and they were not scored in the mean. The imbalances of separate chromosomal arms are represented in Figure 2 . The most frequent chromosome gains in MSS tumors were 20q (69%) (IC 95 = 93.7–43.7), 19 (63%) (IC 95 = 81.2–40.6), 16p (56%) (IC 95 = 81.2–31.2), 9q (50%), 16q (50%) (IC 95 = 75–25), 8q (44%), 17q (44%) (both IC 95 = 81.2–18.7), 12q (38%), 20p (38%) (both IC 95 = 65.6–15.6), 1 (31%), 7q (31%), 13q (31%) and 17p (31%) (all IC 95 = 53.1–6.25). The most frequent chromosome losses were 18q (63%), 4q (63%) (both IC 95 = 81.2–40.6), 15q (44%) (IC 95 = 81.2–18.7), 4p (38%), 5q (38%), 8p (38%) (all IC 95 = 65.6–15.6), 14q (31%) and Xq (31%) (both IC 95 = 53.1–6.25). Specific aneuploidies in MSS tumors were −4, −18q and +20q (all P < 0.05, χ 2 test), whereas the most frequent chromosome gains in MSI tumors were 12q (57%), 17q (57%), 9q (57%) (all IC 95 = 93.3–16.6), 1p (50%), 16q (50%), 19 (50%), 22q (50%) (all IC 95 = 86.6–13.3), 8q (36%), 16p (36%) and 20q (36%) (all IC 95 = 60–10). Chromosome losses in MSI tumors were very infrequent. Only chromosomes 5q, 2q and 4q (all 29%) (all IC 95 = 53.3–6.6) were lost. Minimal common altered regions were defined as overlapped regions for gains or losses detected in at least four primary tumors. In general, we found 10 minimal regions in MSI and 17 regions in MSS tumors. Gains at 1p35pter, 8q24, 9q34, 11q13, 12q24.3, 16q24.3, 17q21 and 17q25, and losses at 5q21 were common findings in both types of tumors. MSS tumors showed statistically significant, specific losses at 1p22 and 4q26 (all P < 0.05, χ 2 test), whereas 6q22q24, 8p23, 14q11q24 and 15q21 showed a tendency to be lost (all 0.05 < P < 0.1, χ 2 test). Minimal chromosome gains in MSS tumors appeared at 13q11q14, 13q34 (both 0.05 < P < 0.1, χ 2 test). Minimal chromosome changes specifically found in MSI tumors were the gain at 22q13.3. This difference was statistically significant ( P < 0.05, χ 2 test).

Fig. 1.

Idiogram showing the summary of copy number changes detected by mCGH in 28 colorectal carcinomas according to their microsatellite status, and two colorectal cancer cell lines, HCT116 and SW480. MSI and MSS profiles are indicated in dark green and dark blue, respectively. HCT116 and SW480 profiles are indicated in yellow and light blue, respectively.

Fig. 1.

Idiogram showing the summary of copy number changes detected by mCGH in 28 colorectal carcinomas according to their microsatellite status, and two colorectal cancer cell lines, HCT116 and SW480. MSI and MSS profiles are indicated in dark green and dark blue, respectively. HCT116 and SW480 profiles are indicated in yellow and light blue, respectively.

Fig. 2.

Overall frequencies of chromosomal gains and losses detected by mCGH according to the chromosome arm in MSS ( A ) and microsatellite unstable ( B ) tumors. Dotted bars indicate gains and black bars indicate losses.

Fig. 2.

Overall frequencies of chromosomal gains and losses detected by mCGH according to the chromosome arm in MSS ( A ) and microsatellite unstable ( B ) tumors. Dotted bars indicate gains and black bars indicate losses.

Array-CGH

Array-CGH was performed on 10 samples (4 MSS and 6 MSI), nine of which were analyzed by mCGH. The reliability of the analysis was evaluated by comparing the results obtained in the two cell lines, HCT116 and SW480. A total of 83% of losses and 71% of gains detected by mCGH were also detected by aCGH. In addition, multiple alterations not detected using mCGH were identified by the microarray-based method ( Table II ) ( Figure 3A–D ). In the SW480 cell line, genes residing within chromosomal regions apparently gained by mCGH, and showing a copy number ratio greater than 1.5-fold were located at 5q35.3 ( STK10 ), 7q33–q35 ( RhoGEF5) , 8q24.2 ( C-MYC ), 11q13.3 ( CCND1 and VEGFB ), 12p12.1 ( SIAT8A ) and 12q14 ( RAP1B ). HCT116 showed 8q24.2 ( C-MYC ), 11q13.3 ( CCND1 ) and 16q24.3 ( CDK10 ) as gains. Although the HTC-116 cell line is derived from a familial cancer, it is widely used in modeling sporadic MSI cancers (MSI archetype).

Fig. 3.

Integration of mCGH (orange circles) and aCGH (blue line and red bar) profiles in the same plot. Blue line was obtained by computing LOWESS smoothing lines in the data with an ‘f’ span equal to 0.2. Red bar was obtained according to a developed statistical segmentation algorithm termed CBS. ( A and B ) Gain of 10q24q26.3 and amplification of chromosomal region 16q24.3 detected in the HCT116 human CRC cell line. ( C and D ) Gains of 7pterq22 and 14q24.3qter, respectively, in the SW480 human CRC cell line. ( E and F ) Coincidence of mCGH and aCGH profiles along the chromosome 8 and 20 in the primary tumors #53T and #167T, respectively.

Fig. 3.

Integration of mCGH (orange circles) and aCGH (blue line and red bar) profiles in the same plot. Blue line was obtained by computing LOWESS smoothing lines in the data with an ‘f’ span equal to 0.2. Red bar was obtained according to a developed statistical segmentation algorithm termed CBS. ( A and B ) Gain of 10q24q26.3 and amplification of chromosomal region 16q24.3 detected in the HCT116 human CRC cell line. ( C and D ) Gains of 7pterq22 and 14q24.3qter, respectively, in the SW480 human CRC cell line. ( E and F ) Coincidence of mCGH and aCGH profiles along the chromosome 8 and 20 in the primary tumors #53T and #167T, respectively.

The results of the tumor sample mCGH profiles corresponded well with the array results ( Table III ). In general, the coincidence of array and mCGH was ∼60% ( Figure 3E–F ). Two out of four MSS samples did not display any genetic alteration detected by aCGH. In only one MSS sample CMYC amplification was detected. The loss of chromosome 17p was found in one MSS sample. Target genes to be lost were XIAP associated factor-1 , PRPSAP2 , TP53 , CD68 and ALOXE3 . In MSI samples, aCGH an amplification peak at 1q31.3 was found, involving PLA2GA4 , TAF1A , ACTN2 and CR2 genes. A loss at 2p16p21 was detected in sample # 274T with the corresponding genes CYP1B1 , SLC8A1 and GPR75 . Deletions at 15q21q22 involved TRPM7 , RAB11A and MAP2K5 . A gained region at 17q25 involved PC2, MAFG, USP14, COLEC12 and VYES1 . Metaphase CGH was not performed on sample #123T because of lack of material, but aCGH showed a gain of chromosome 20, involving PDYN , GAL3ST1 , PROCR , SRC and PPP1R16B . Four cases, including one MSS and three MSI, showed the loss of 5q13.3q21, a chromosomal region found also by mCGH in 35% of samples. CCNH , FER and RASA1 genes were the most frequently involved in this region according to our aCGH results.

Table III.

Clinical data and copy number changes of 10 samples analyzed by metaphase and array-based CGH

Case  Patients
 
  DNA copy number changes by mCGH
 
  DNA copy number changes by aCGH
 
 

 
MSI status
 
Dukes' stage
 
Gains
 
Losses
 
Gains
 
Losses
 
50T – 1q31, 7p12, 17q25, 19, 20 3p12, 5q21q23.3 1q31, 7p21.3p21.2, 19p13.2p13.3, 19q13.42, 20 1p31, 2q13q14, 2q33.2, 3p12, 4q12q13, 5q15q23.1, 10q21.1, 15q14, 15q26.2 
53T – 8q, 12q21.3qter, 20q 1q21, 6q, 7q11, 8p, 17p, 18, 20p 8q, 12q12.2q21.2, 12q23.2qter, 16p13.3, 20q 6p12.1q23, 7cenq11.23, 8p, 17p11.2p13.1, 18p11.3q11.2, 18q21.1, 18q21.3qter, 20p 
62T – 11q13.3q13.4, 12q23qter 15q11q12 12 – 
135T – – 9p21 – – 
87T 8q24.3, 16q24.3, 17q25, 20q13.3 15q15q22 4q25q28, 6q16.3q21, 8q12q13, 16q23.3q24.1, 17q25.3, 18p11.2, 19q13.4, 20q13.33 15q21q22.2 
123T ND ND 13q12, 20 1p22p36, 1q, 4, 5q14q21, 9p23, 9q32q33, 10q12q21 
167T 3q29, 13q12.2qter, 14q21q24.1, 20 2p23pter, 2q37, 9p12p13 3q28q29, 13q11q34, 14q21.3q32.1, 20 1q41, 2q12qter, 2p23pter, 4, 15q21.3 
259T 1q42qter, 2q36, 3p21, 11q13, 12q24.3, 17, 19, 22q13 16p12.1q22 3p21.1p21.2, 12q24.21q24.31 16p12.2q12 
274T 1q, 16q24, 17q25, 18q22qter, 22q13.3 2p16p21 1q (peak at 1q31.3), 3q25.3q26.1, 18q22qter 2p22.1p16, 2q32.2, 5q21 
280T 1q32qter, 4p16, 8p22p23, 8q24.3, 12q23qter, 17, 19q, 20 2q23q33, 6q, 13q21q31, 14q13q21.1, 18q 1q42, 12q24.33 2p14, 3p14.2, 5q15q21.3, 6q22, 13q22.2q31.3, 14q13q21, 18q 
Case  Patients
 
  DNA copy number changes by mCGH
 
  DNA copy number changes by aCGH
 
 

 
MSI status
 
Dukes' stage
 
Gains
 
Losses
 
Gains
 
Losses
 
50T – 1q31, 7p12, 17q25, 19, 20 3p12, 5q21q23.3 1q31, 7p21.3p21.2, 19p13.2p13.3, 19q13.42, 20 1p31, 2q13q14, 2q33.2, 3p12, 4q12q13, 5q15q23.1, 10q21.1, 15q14, 15q26.2 
53T – 8q, 12q21.3qter, 20q 1q21, 6q, 7q11, 8p, 17p, 18, 20p 8q, 12q12.2q21.2, 12q23.2qter, 16p13.3, 20q 6p12.1q23, 7cenq11.23, 8p, 17p11.2p13.1, 18p11.3q11.2, 18q21.1, 18q21.3qter, 20p 
62T – 11q13.3q13.4, 12q23qter 15q11q12 12 – 
135T – – 9p21 – – 
87T 8q24.3, 16q24.3, 17q25, 20q13.3 15q15q22 4q25q28, 6q16.3q21, 8q12q13, 16q23.3q24.1, 17q25.3, 18p11.2, 19q13.4, 20q13.33 15q21q22.2 
123T ND ND 13q12, 20 1p22p36, 1q, 4, 5q14q21, 9p23, 9q32q33, 10q12q21 
167T 3q29, 13q12.2qter, 14q21q24.1, 20 2p23pter, 2q37, 9p12p13 3q28q29, 13q11q34, 14q21.3q32.1, 20 1q41, 2q12qter, 2p23pter, 4, 15q21.3 
259T 1q42qter, 2q36, 3p21, 11q13, 12q24.3, 17, 19, 22q13 16p12.1q22 3p21.1p21.2, 12q24.21q24.31 16p12.2q12 
274T 1q, 16q24, 17q25, 18q22qter, 22q13.3 2p16p21 1q (peak at 1q31.3), 3q25.3q26.1, 18q22qter 2p22.1p16, 2q32.2, 5q21 
280T 1q32qter, 4p16, 8p22p23, 8q24.3, 12q23qter, 17, 19q, 20 2q23q33, 6q, 13q21q31, 14q13q21.1, 18q 1q42, 12q24.33 2p14, 3p14.2, 5q15q21.3, 6q22, 13q22.2q31.3, 14q13q21, 18q 

Differentially expressed genes

The gene expression experiments were performed on six samples (3 MSS and 3 MSI) of which four were used in mCGH analysis. More than 100 genes appeared to be differentially expressed when comparing three MSS to three MSI tumors. Clustering analysis allowed us to describe a total of 28 genes that were specifically over- or under-expressed in MSS or MSI samples ( Figure 4 ). The gene ontology database (AmiGO, http://www.godatabase.org/cgi-bin/amigo/go.cgi ) was used to postulate tentative gene functions. The genomic locations of the over- and under-expressed genes were sought to obtain possible correlation with CGH data, and thus find the possible target genes of the amplified and lost regions detected by CGH.

Fig. 4.

Gene expression clustering analysis. ‘Rows’ are individual genes and ‘Columns’ are colorectal samples according to their microsatellite status.

Fig. 4.

Gene expression clustering analysis. ‘Rows’ are individual genes and ‘Columns’ are colorectal samples according to their microsatellite status.

Discussion

The genomic instability of CRC is widely accepted. It is unlikely that the DNA copy number alterations observed in colorectal tumors occur non-randomly, but rather they involve particular regions of the genome. In the present study, genome-wide chromosomal imbalances and gene expression profiles in MSS tumors were compared with those in MSI tumors by using mCGH, aCGH and expression microarrays. The present study shows chromosomal differences, putative imbalanced genes and possible target gene identification by expression profiling in two phenotypes of colorectal tumors and cell lines categorized according to their MSS. The aCGH results were comparable to mCGH, as exemplified with the two cell lines. However, aCGH showed new chromosomal imbalances, especially losses. In addition, most of the regions undetected by aCGH (1p36, 9q34, 11q13, 12q24, 16p, 19 and 22) belonged to chromosomal areas that have been reported as problematic in mCGH, due to the GC-content ( 44 ). In general, it is suggested that MSI colorectal tumors show a stable karyotype ( 1 ). Recent aCGH studies performed on MSI tumors showed gains at 20q, 8p, and 8q24.3, and losses at 18q21.1q21.2 ( 27 ). Our aCGH data suggested particular copy number alterations affecting other chromosomes in MSI tumors ( Table III ). Most of the MSS colorectal tumors are suggested to show CIN mainly characterized by the prevalence of aneuploidy and structural alterations ( 2 , 3 , 45 ). Our data showed three MSS tumors with a very low number of chromosomal imbalances, similar to the MSI-CIN − group of colorectal tumors studied by Jones et al . ( 26 ). Nevertheless, we did not detect the loss of chromosome 20p, the only change that the authors found to be significantly over-represented in MSI-CIN − cancers compared with MSI-CIN + tumors.

In agreement with previous data of mCGH ( 17 , 46 , 47 ), we detected gains of 20q and 8q and loss of 18q as the most frequent alterations in MSS cancers. Gains of 17q, 20p, 7q and 13q, and losses of 4, 15q, 5q, 8p, and 14q, in order of decreasing frequency, were also detected. Minimal chromosomal imbalances found to be significantly specific to MSS tumors were losses at 1p22 and 4q26. The former has been postulated as a prominent deletion in CRC ( 48 ). The deletion of chromosome 4 has already been shown to be involved in the adenoma to invasive carcinoma progression ( 48 ) and the involvement of 4q26 has also been proven ( 49 ). The regions 6q22q24, 8p23, 14q11q24 and 15q21 showed a tendency to be lost in MSS tumors, while 13q11q14 and 13q34 tended to be gained. The presence of a putative tumor suppressor gene in the distal part of chromosome 8p has been proposed ( 50 ), and the region 15q21, although infrequently affected in CRC, is suggested to harbor a tumor suppressor gene ( 51 ). Interestingly, several down-regulated genes at 15q21 have been reported, including THBS1 and ITA3 ( 28 ), and this region has been linked to adenoma-carcinoma progression ( 52 ).

The results of the present study show that gains of 17q, 22q, 8q and 20q and the losses of 5q, 2q, and 4q occur frequently in MSI CRCs. Previous findings reported gains of 4q and 8q, and losses of 9q, 1p and 11q ( 53 ), or genomic stability ( 54 ) as the most common features in MSI tumors. However, in the present study only one of the 15 MSI tumors was genetically stable. A preferential gain of 22q13 was also observed. A previous mCGH report found MSI CRCs chromosomally stable, whereas the non-MSI tumors showed 7, 13 and 20q gained and 17, 18 and 9p lost ( 52 , 54 ).

Several minimal regions occurring both in MSS and MSI tumors were found, e.g. gains at 8q24, 16q24.3, 17q21, 17q25 and 20q13.2q13.3, and loss of 5q21. The latter region is known to contain the tumor suppressor gene APC , which mutations and allelic imbalances are considered the earliest and most prevalent genetic changes in CRC, coinciding with the CIN pathway ( 1 , 55 ). According to several authors ( 56 , 57 ), loss of heterozygosity at 5q is also present in patients with HNPCC. Considering that the Amsterdam Criteria are liable to miss many cases of HNPCC, the MSI samples may include some HNPCC patients. In the present study 29% (8/28) of the samples showed loss of 5q21 on mCGH, and 40% (4/10) by aCGH ( Figure 1 , Table III ). However, as it is well known that allelic imbalances affect more than one single gene, several candidate genes believed to be involved in the loss of 5q14q21 were identified by aCGH, e.g. CCNH , FER and RASA1 . FER is a member of the SRC tyrosine kinase gene family, and has been associated with the loss of 5q21 in e.g. myeloid leukemia ( 58 ) and lung cancer ( 59 ). Moreover, it has been suggested that Fer may be a potential target for modulators of malignant cell growth ( 60 ). We detected gains of 8q24 in both MSS and MSI tumors; however, the resolution of mCGH did not allow delineation of the target of amplification, which has been reported to vary according to MSI status ( 27 ). Recent findings have also demonstrated alterations in chromosome 8 in MSS and MSI tumors ( 25 ). The gain in chromosome 20q is thought to be an early event occurring in the transition from adenoma-carcinoma progression ( 17 ), and it has been associated with poor survival in several studies ( 20 , 47 ). The gain of 20q13.2q13.3 has already been associated with colorectal carcinogenesis and metastatic progression in CRC ( 6163 ). A gain of 16q24.3 was also frequently gained in our samples, including the HCT116 cell line ( Figure 3B ). Array-CGH data showed several putative target genes, including CDK10 and DPEP1 , which were also found to be over-expressed in MSS samples with a 16q24.3 gain. DPEP1 has already been suggested as a tumor-specific molecular marker for disseminated colon tumor cells ( 64 ). Gene expression analysis revealed differences between MSS and MSI tumors. MAGE and CTNNB1 were detected over-expressed in MSI tumors. CTNNB1 (encoding for the activating somatic β-catenin) mutations are common in HNPCC tumors ( 65 ). In recent studies, MAGE has been described as a biomarker to detect micrometastasis ( 66 ) and it has been significantly associated with the vessel emboli in CRC ( 67 ). However, none of the samples analyzed here, and which showed over-expression of MAGE , presented metastasis. Further studies will be required to demonstrate the role of those genes that appear to be differentially expressed between MSS and MSI tumors according to our results.

As a conclusion, CRC is a heterogeneous disease with many molecular phenotypes. In the present study we studied both microsatellite stable and unstable phenotypes by their copy number changes. Both the phenotypes were shown to have both stable and unstable chromosomal structures, suggesting involvement of multiple mechanisms in CRC carcinogenesis. Over-expression of CDK10 and DPEP1 was found in samples with amplification at the corresponding genomic region (16q24.3), and therefore they may be considered as putative target genes. However, more samples are needed for validation studies.

Correspondence may also be addressed to Jordi Camps Email: Jordi.Camps@uab.es

We thank laboratory technicians Àngels Niubó and Tiina Wirtanen for their excellent work. JC is a fellow of the Universitat Autònoma de Barcelona, and belongs to the research program EPICUR-Red. GA has a contract of re-incorporation of doctors from the Generalitat de Catalunya. This work has been supported by EPICUR-Red (FIS G03/174) from the Spanish Government, and the Generalitat de Catalunya (CIRIT 2001SGR-00201).

Conflict of Interest Statement : None declared.

References

1.
Lengauer,C., Kinzler,K.W. and Vogelstein,B. (
1998
) Genetic instabilities in human cancers.
Nature
  ,
396
,
643
–649.
2.
Ribas,M., Masramon,L., Aiza,G., Capellà,G., Miró,R. and Peinado,M.A. (
2003
) The structural nature of chromosomal instability in colon cancer cells.
FASEB J.
  ,
17
,
289
–291.
3.
Camps,J., Ponsa,I., Ribas,M., Prat,E., Egozcue,J., Peinado,M.A. and Miró,R. (
2005
) Comprehensive measurement of chromosomal instability in cancer cells: combination of fluorescence in situ hybridization and cytokinesis-block micronucleus assay.
FASEB J
  .,
19
,
828
–830.
4.
Ionov,Y., Peinado,M.A., Malkhosyan,S., Shibata,D. and Perucho,M. (
1993
) Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis.
Nature
  ,
363
,
558
–561.
5.
Aaltonen,L.A., Peltomaki,P., Leach,F.S. et al. (
1993
) Clues to the pathogenesis of familial colorectal cancer.
Science
  ,
260
,
812
–816.
6.
Tang,R., Changchien,C.R., Wu,M.C., Fan,C.W., Liu,K.W., Chen,J.S., Chien,H.T. and Hsieh,L.L. (
2004
) Colorectal cancer without high microsatellite instability and chromosomal instability—an alternative genetic pathway to human colorectal cancer.
Carcinogenesis
  ,
25
,
841
–846.
7.
Goel,A., Arnold,C.N., Niedzwiecki,D. et al. (
2003
) Characterization of sporadic colon cancer by patterns of genomic instability.
Cancer Res.
  ,
63
,
1608
–1614.
8.
Abdel-Rahman,W.M., Katsura,K., Rens,W., Gorman,P.A., Sheer,D., Bicknell,D., Bodmer,W.F., Arends,M.J., Wyllie,A.H. and Edwards,P.A. (
2001
) Spectral karyotyping suggests additional subsets of colorectal cancers characterized by pattern of chromosome rearrangement.
Proc. Natl Acad. Sci. USA
  ,
98
,
2538
–2543.
9.
Camps,J., Morales,C., Prat,E., Ribas,M., Capellà,G., Egozcue,J., Peinado,M.A. and Miró,R. (
2004
) Genetic evolution in colon cancer KM12 cells and metastatic derivates.
Int. J. Cancer
  ,
110
,
869
–874.
10.
Fishel,R., Lescoe,M.K., Rao,M.R., Copeland,N.G., Jenkins,N.A., Garber,J., Kane,M. and Kolodner,R. (
1993
) The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer [Erratum (1994) Cell , 77, 167.].
Cell
  ,
75
,
1027
–1038.
11.
Bronner,C.E., Baker,S.M., Morrison,P.T. et al. (
1994
) Mutation in the DNA mismatch repair gene homologue hMLH1 is associated with hereditary non-polyposis colon cancer.
Nature
  ,
368
,
258
–261.
12.
Kinzler,K.W. and Vogelstein,B. (
1996
) Lessons from hereditary colorectal cancer.
Cell
  ,
87
,
159
–170.
13.
Cahill,D.P., Lengauer,C., Yu,J., Riggins,G.J., Willson,J.K., Markowitz,S.D., Kinzler,K.W. and Vogelstein,B. (
1998
) Mutations of mitotic checkpoint genes in human cancers.
Nature
  ,
392
,
300
–303.
14.
Hanks,S., Coleman,K., Reid,S. et al. (
2004
) Constitutional aneuploidy and cancer predisposition caused by biallelic mutations in BUB1B.
Nat. Genet.
  ,
36
,
1159
–1161.
15.
Bardi,G., Sukhikh,T., Pandis,N., Fenger,C., Kronborg,O. and Heim,S. (
1995
) Karyotypic characterization of colorectal adenocarcinomas.
Genes Chromosomes Cancer
  ,
12
,
97
–109.
16.
Muleris,M., Dutrillaux,A.M., Olschwang,S., Salmon,R.J. and Dutrillaux,B. (
1995
) Predominance of normal karyotype in colorectal tumors from hereditary non-polyposis colorectal cancer patients.
Genes Chromosomes Cancer
  ,
14
,
223
–226.
17.
Ried,T., Knutzen,R., Steinbeck,R., Blegen,H., Schrock,E., Heselmeyer,K., du Manoir,S. and Auer,G. (
1996
) Comparative genomic hybridization reveals a specific pattern of chromosomal gains and losses during the genesis of colorectal tumors.
Genes Chromosomes Cancer
  ,
15
,
234
–245.
18.
Meijer,G.A., Hermsen,M.A., Baak,J.P., van Diest,P.J., Meuwissen,S.G., Belien,J.A., Hoovers,J.M., Joenje,H., Snijders,P.J. and Walboomers,J.M. (
1998
) Progression from colorectal adenoma to carcinoma is associated with non-random chromosomal gains as detected by comparative genomic hybridisation.
J. Clin. Pathol.
  ,
51
,
901
–909.
19.
Al-Mulla,F., Keith,W.N., Pickford,I.R., Going,J.J. and Birnie,G.D. (
1999
) Comparative genomic hybridization analysis of primary colorectal carcinomas and their synchronous metastases.
Genes Chromosomes Cancer
  ,
24
,
306
–314.
20.
Aragane,H., Sakakura,C., Nakanishi,M. et al. (
2001
) Chromosomal aberrations in colorectal cancers and liver metastases analyzed by comparative genomic hybridization.
Int. J. Cancer
  ,
94
,
623
–629.
21.
Knuutila,S., Autio,K. and Aalto,Y. (
2000
) Online access to CGH data of DNA sequence copy number changes [Erratum (2000) Am. J. Pathol. , 157, 1413.].
Am. J. Pathol.
  .,
157
,
689
.
22.
Pinkel,D., Segraves,R., Sudar,D. et al. (
1998
) High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays.
Nat. Genet.
  ,
20
,
207
–211.
23.
Pollack,J.R., Perou,C.M., Alizadeh,A.A., Eisen,M.B., Pergamenschikov,A., Williams,C.F., Jeffrey,S.S., Botstein,D. and Brown,P.O. (
1999
) Genome-wide analysis of DNA copy-number changes using cDNA microarrays.
Nat. Genet.
  ,
23
,
41
–46.
24.
Albertson,D.G. and Pinkel,D. (
2003
) Genomic microarrays in human genetic disease and cancer.
Hum. Mol. Genet.
  ,
12
,
R145
–R152.
25.
Nakao,K., Mehta,K.R., Fridlyand,J., Moore,D.H., Jain,A.N., Lafuente,A., Wiencke,J.W., Terdiman,J.P. and Waldman,F.M. (
2004
) High-resolution analysis of DNA copy number alterations in colorectal cancer by array-based comparative genomic hybridization.
Carcinogenesis
  ,
25
,
1345
–1357.
26.
Jones,A.M., Douglas,E.J., Halford,S.E., Fiegler,H., Gorman,P.A., Roylance,R.R., Carter,N.P. and Tomlinson,I.P. (
2005
) Array-CGH analysis of microsatellite-stable, near-diploid bowel cancers and comparison with other types of colorectal carcinoma.
Oncogene.
  ,
24
,
118
–129.
27.
Douglas,E.J., Fiegler,H., Rowan,A., Halford,S., Bicknell,D.C., Bodmer,W., Tomlinson,I.P. and Carter,N.P. (
2004
) Array comparative genomic hybridization analysis of colorectal cancer cell lines and primary carcinomas.
Cancer Res.
  ,
64
,
4817
–4825.
28.
Birkenkamp-Demtroder,K., Christensen,L.L., Olesen,S.H., Frederiksen,C.M., Laiho,P., Aaltonen,L.A., Laurberg,S., Sorensen,F.B., Hagemann,R. and ORntoft,T.F. (
2002
) Gene expression in colorectal cancer.
Cancer Res.
  ,
62
,
4352
–4363.
29.
Dunican,D.S., McWilliam,P., Tighe,O., Parle-McDermott,A. and Croke,D.T. (
2002
) Gene expression differences between the microsatellite instability (MIN) and chromosomal instability (CIN) phenotypes in colorectal cancer revealed by high-density cDNA array hybridization [Erratum (2002) Oncogene , 21, 7912.].
Oncogene
  ,
21
,
3253
–3257.
30.
Masramon,L., Ribas,M., Cifuentes,P., Arribas,R., Garcia,F., Egozcue,J., Peinado,M.A. and Miro,R. (
2000
) Cytogenetic characterization of two colon cell lines by using conventional G-banding, comparative genomic hybridization, and whole chromosome painting.
Cancer Genet. Cytogenet.
  ,
121
,
17
–21.
31.
Camps,J., Mrasek,K., Prat,E., Weise,A., Starke,H., Egozcue,J., Miro,R. and Liehr,T. (
2004
) Molecular cytogenetic characterisation of the colorectal cancer cell line SW480.
Oncol. Rep.
  ,
11
,
1215
–1218.
32.
Gonzalez-Garcia,I., Moreno,V., Navarro,M., Marti-Rague,J., Marcuello,E., Benasco,C., Campos,O., Capella,G. and Peinado,M.A. (
2000
) Standardized approach for microsatellite instability detection in colorectal carcinomas.
J. Natl. Cancer Inst.
  ,
92
,
544
–549.
33.
Vasen,H.F., Watson,P., Mecklin,J.-P. and Lynch,H.T. (
1999
) New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the International Collaborative group on HNPCC.
Gastroenterology
  ,
116
,
1453
–1456.
34.
Kallioniemi,A., Kallioniemi,O.P., Sudar,D., Rutovitz,D., Gray,J.W., Waldman,F. and Pinkel,D. (
1992
) Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors.
Science
  ,
258
,
818
–821.
35.
Prat,E., Bernues,M., Caballin,M.R., Egozcue,J., Gelabert,A. and Miro,R. (
2001
) Detection of chromosomal imbalances in papillary bladder tumors by comparative genomic hybridization.
Urology
  ,
57
,
986
–992.
36.
Hyman,E., Kauraniemi,P., Hautaniemi,S. et al. (
2002
) Impact of DNA amplification on gene expression patterns in breast cancer.
Cancer Res.
  ,
62
,
6240
–6245.
37.
Monni,O., Barlund,M., Mousses,S. et al. (
2001
) Comprehensive copy number and gene expression profiling of the 17q23 amplicon in human breast cancer.
Proc. Natl Acad. Sci. USA
  ,
98
,
5711
–5716.
38.
Atiye,J., Wolf,M., Kaur,S., Monni,O., Bohling,T., Kivioja,A., Tas,E., Serra,M., Tarkkanen,M. and Knuutila,S. (
2005
) Gene amplifications in osteosarcoma-CGH microarray analysis.
Genes Chromosomes Cancer
  ,
42
,
158
–163.
39.
Cleveland,W.S. (
1979
) Robust locally weighted regression and smoothing scatterplots.
J. Am. Stat. Assoc.
  ,
74
,
829
–836.
40.
Olshen,A.B. and Venkatraman,E.S. (
2002
) Change-point analysis of array-based comparative genomic hybridization data. American Statistical Association Proceedings of the Joint Statistical Meetings , 2530–2535.
41.
Irizarry,R.A., Hobbs,B., Collin,F., Beazer-Barclay,Y.D., Antonellis,K.J., Scherf,U. and Speed,T.P. (
2003
) Exploration, normalization, and summaries of high density oligonucleotide array probe level data.
Biostatistics
  ,
4
,
249
–264.
42.
Lonnstedt,I. and Speed,T.P. (
2002
) Replicated microarray data.
Stat. Sinica
  ,
12
,
31
–46.
43.
Gentleman,R.C., Carey,V.J., Bates,D.M. et al. (
2004
) Bioconductor: open software development for computational biology and bioinformatics.
Genome Biol.
  ,
5
,
R80
.
44.
el-Rifai,W., Larramendy,M.L., Bjorkqvist,A.M., Hemmer,S. and Knuutila,S. (
1997
) Optimization of comparative genomic hybridization using fluorochrome conjugated to dCTP and dUTP nucleotides.
Lab. Invest.
  ,
77
,
699
–700.
45.
Roschke,A.V., Stover,K., Tonon,G., Schaffer,A.A. and Kirsch,I.R. (
2002
) Stable karyotypes in epithelial cancer cell lines despite high rates of ongoing structural and numerical chromosomal instability.
Neoplasia
  ,
4
,
19
–31.
46.
Korn,W.M., Yasutake,T., Kuo,W.L., Warren,R.S., Collins,C., Tomita,M., Gray,J. and Waldman,F.M. (
1999
) Chromosome arm 20q gains and other genomic alterations in colorectal cancer metastatic to liver, as analyzed by comparative genomic hybridization and fluorescence in situ hybridization.
Genes Chromosomes Cancer
  ,
25
,
82
–90.
47.
De Angelis,P.M., Stokke,T., Beigi,M., Mjaland,O. and Clausen,O.P. (
2001
) Prognostic significance of recurrent chromosomal aberrations detected by comparative genomic hybridization in sporadic colorectal cancer.
Int. J. Colorectal Dis.
  ,
16
,
38
–45.
48.
Knosel,T., Schluns,K., Stein,U., Schwabe,H., Schlag,P.M., Dietel,M. and Petersen,I. (
2004
) Chromosomal alterations during lymphatic and liver metastasis formation of colorectal cancer [Erratum (2004) Neoplasia , 6, 686.].
Neoplasia
  ,
6
,
23
–28.
49.
Arribas,R., Risques,R.A., Gonzalez-Garcia,I., Masramon,L., Aiza,G., Ribas,M., Capella,G. and Peinado,M.A. (
1999
) Tracking recurrent quantitative genomic alterations in colorectal cancer: allelic losses in chromosome 4 correlate with tumor aggressiveness.
Lab. Invest.
  ,
79
,
111
–122.
50.
Flanagan,J.M., Healey,S., Young,J., Whitehall,V., Trott,D.A., Newbold,R.F. and Chenevix-Trench,G. (
2004
) Mapping of a candidate colorectal cancer tumor-suppressor gene to a 900-kilobase region on the short arm of chromosome 8.
Genes Chromosomes Cancer
  ,
40
,
247
–260.
51.
Park,W.S., Park,J.Y., Oh,R.R. et al (
2000
) A distinct tumor suppressor gene locus on chromosome 15q21.1 in sporadic form of colorectal cancer.
Cancer Res.
  ,
60
,
70
–73.
52.
Hermsen,M., Postma,C., Baak,J. et al. (
2002
) Colorectal adenoma to carcinoma progression follows multiple pathways of chromosomal instability.
Gastroenterology
  ,
123
,
1109
–1119.
53.
Li,L.S., Kim,N.G., Kim,S.H. et al. (
2003
) Chromosomal imbalances in the colorectal carcinomas with microsatellite instability.
Am. J. Pathol.
  ,
163
,
1429
–1436.
54.
Schlegel,J., Stumm,G., Scherthan,H., Bocker,T., Zirngibl,H., Ruschoff,J. and Hofstadter,F. (
1995
) Comparative genomic in situ hybridization of colon carcinomas with replication error.
Cancer Res.
  ,
55
,
6002
–6005.
55.
Shih,I.M., Zhou,W., Goodman,S.N., Lengauer,C., Kinzler,K.W. and Vogelstein,B. (
2001
) Evidence that genetic instability occurs at an early stage of colorectal tumorigenesis.
Cancer Res.
  ,
61
,
818
–822.
56.
Young,J., Simms,L.A., Biden,K.G. et al. (
2001
) Features of colorectal cancers with high-level microsatellite instability occurring in familial and sporadic settings: parallel pathways of tumorigenesis.
Am. J. Pathol.
  ,
159
,
2107
–2116.
57.
Wu,C., Akiyama,Y., Imai,K. et al. (
1994
) DNA alterations in cells from hereditary non-polyposis colorectal cancer patients.
Oncogene
  ,
9
,
991
–994.
58.
Morris,C., Heisterkamp,N., Hao,Q.L., Testa,J.R. and Groffen,J. (
1990
) The human tyrosine kinase gene (FER) maps to chromosome 5 and is deleted in myeloid leukemias with a del(5q).
Cytogenet. Cell Genet.
  ,
53
,
196
–200.
59.
Ueno,K., Kumagai,T., Kijima,T., Kishimoto,T. and Hosoe,S. (
1998
) Cloning and tissue expression of cDNAs from chromosome 5q21–22 which is frequently deleted in advanced lung cancer.
Hum. Genet.
  ,
102
,
63
–68.
60.
Orlovsky,K., Theodor,L., Malovani,H., Chowers,Y. and Nir,U. (
2002
) Gamma interferon down-regulates Fer and induces its association with inactive Stat3 in colon carcinoma cells.
Oncogene
  ,
21
,
4997
–5001.
61.
Hidaka,S., Yasutake,T., Takeshita,H. et al. (
2000
) Differences in 20q13.2 copy number between colorectal cancers with and without liver metastasis.
Clin. Cancer Res.
  ,
6
,
2712
–2717.
62.
Laiho,P., Hienonen,T., Karhu,A. et al. (
2003
) Genome-wide allelotyping of 104 Finnish colorectal cancers reveals an excess of allelic imbalance in chromosome 20q in familial cases.
Oncogene
  ,
22
,
2206
–2214.
63.
Rooney,P.H., Boonsong,A., McFadyen,M.C., McLeod,H.L., Cassidy,J., Curran,S. and Murray,G.I. (
2004
) The candidate oncogene ZNF217 is frequently amplified in colon cancer.
J. Pathol.
  ,
204
,
282
–288.
64.
McIver,C.M., Lloyd,J.M., Hewett,P.J. and Hardingham,J.E. (
2004
) Dipeptidase 1: a candidate tumor-specific molecular marker in colorectal carcinoma.
Cancer Lett.
  ,
209
,
67
–74.
65.
Miyaki,M., Iijima,T., Kimura,J., Yasuno,M., Mori,T., Hayashi,Y., Koike,M., Shitara,N., Iwama,T. and Kuroki,T. (
1999
) Frequent mutation of beta-catenin and APC genes in primary colorectal tumors from patients with hereditary nonpolyposis colorectal cancer.
Cancer Res.
  ,
59
,
4506
–4509.
66.
Kufer,P., Zippelius,A., Lutterbuse,R. et al. (
2002
) Heterogeneous expression of MAGE-A genes in occult disseminated tumor cells: a novel multimarker reverse transcription-polymerase chain reaction for diagnosis of micrometastatic disease.
Cancer Res.
  ,
62
,
251
–261.
67.
Li,M., Yuan,Y.H., Han,Y., Liu,Y.X., Yan,L., Wang,Y. and Gu,J. (
2005
) Expression profile of cancer-testis genes in 121 human colorectal cancer tissue and adjacent normal tissue.
Clin. Cancer Res.
  ,
11
,
1809
–1814.

Author notes

Laboratory of Cytogenetics, Departament de Biologia Cel·lular, Fisiologia i Immunologia and Institut de Biotecnologia i de Biomedicina, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain, 1Unitat d'Antropologia, Departament de Biologia Animal, de Biologia Vegetal i d'Ecologia, Universitat Autònoma de Barcelona, Bellaterra, Spain, 2Programa de Bioinformàtica i Genòmica, Centre de Regulació Genòmica (CRG), Barcelona, Spain and 3Department of Pathology, Haartman Institute and HUSLAB, POB 21 (Haartmaninkatu 3), FI-00014, University of Helsinki and Helsinki University Central Hospital, Helsinki, Finland