- Email: [email protected]
Genomic Array Technology
CHAPTER 33
Genomic Array Technology Heike Fiegler and Nigel P. Carter The Wellcome Trust Sanger Institute Wellcome Trust Genome Campus Hinxton, Cambridge CB10 1SA, United Kingdom
I. II. III. IV. V. VI. VII. VIII. IX.
Introduction Construction of Genomic Clone Arrays Using DOP-PCR Amplification Validation of Array Performance and Data Analysis Array CGH Array CGH for Cytogenetic Analyses Array Painting Application of Array CGH and Array Painting for Complete Cytogenetic Analyses ChIP on Genomic Clone Arrays Conclusion References
I. Introduction The sequencing of the human genome in the public domain has generated an extensive resource of mapped and sequenced clones that are revolutionizing molecular cytogenetics. This clone resource finds particular use in the construction of large insert clone DNA microarrays spotted onto glass microscope slides. Originally, these arrays were produced to replace metaphase chromosomes in comparative genomic hybridization (CGH) experiments to detect copy number changes along the genome at an increased resolution (Albertson and Pinkel, 2003). In this way, the resolution of CGH becomes limited only by the insert size and the density of the mapped sequences used. However, because of the increased sensitivity and resolution compared to conventional methodologies, genomic arrays are also being applied for cytogenetic studies of microdeletion and microduplication, as well as for rapid mapping of breakpoints (Fiegler et al., 2003b). METHODS IN CELL BIOLOGY, VOL. 75 Copyright 2004, Elsevier Inc. All rights reserved. 0091-679X/04 $35.00
769
770
Heike Fiegler and Nigel P. Carter
II. Construction of Genomic Clone Arrays Using DOP-PCR Amplification Cosmid, P1, P1 artificial chromosome (PAC), and bacterial artificial chromosome (BAC) clones are being widely used for the construction of genomic DNA microarrays. DNA was originally extracted from large-scale cultures (Pinkel et al., 1998; Solinas-Toldo et al., 1997), which becomes a costly and time-consuming process when expanded to the number of clones required to construct an array with a resolution of 1 Mb (3500 clones). To overcome this problem, several approaches have been applied to amplify the clone DNA enzymatically for spotting, thereby removing the requirement for large-scale DNA preparations. These have included methods such as linker adapter polymerase chain reaction (PCR) (Snijders et al., 2001), rolling circle PCR using Phi29 (Buckley et al., 2002), or degenerate oligonucleotide primed PCR (DOP-PCR) (Hodgson et al., 2001) using an amine-modified version of the standard DOP-PCR primer 6MW (Telenius et al., 1992a,b). Escherichia coli genomic DNA, however, is a common contaminant of DNA preparations of large insert clones. The degree of contamination has been estimated by real-time PCR to be between 6% and 26%, depending on the method used for purification (Foreman and Davis, 2000). This contaminating E. coli DNA will reduce the capacity of each probe spotted on the array to hybridize with the DNA of interest and may contribute to an increased nonspecific background signal. To overcome this problem, we have designed three new DOP-PCR primers (DOP1, DOP2, and DOP3) that were chosen to be eYcient in amplifying human genomic DNA but ineYcient in amplifying the contaminating E. coli DNA (Fiegler et al., 2003a). The sequence and the frequency of the six bases at the 30 end of the new primers in human and E. coli DNA compared to the standard DOP-PCR primer 6MW is shown in Table I. The new primers retain the ability to amplify human sequence but demonstrate poor amplification of the E. coli sequence compared to the standard DOP-PCR primer 6MW. Furthermore, the 30 ends of the primers DOP1 and DOP3 are the reverse complement of each other, which allows the amplification from the complementary strand in the opposite direction, thereby
Table I Degenerate Oligonucleotide Primed Polymerase Chain Reaction Primer for the Amplification of Clones Containing Large Human Inserts Primer
30 sequence
Escherichia coli
Human
DOP1 DOP2 DOP3 6MW
CTAGAA TAGGAG TTCTAG ATGTGG
0.02 0.05 0.02 0.40
0.63 0.49 0.63 0.65
33. Genomic Array Technology
771
providing diVerent but complementary representations of the template sequence. The use of these three new DOP-PCR primers, particularly in combination, revealed a significant increase in sensitivity and reproducibility in genomic hybridizations compared to arrays constructed with the standard DOP-PCR primer 6MW (Fiegler et al., 2003a). Following this strategy, we have constructed a large insert clone DNA microarray composed of 3523 sequencing clones selected from the Golden Path of the human genome (Lander et al., 2001) located at approximately 1-Mb intervals across the human genome. The clones were picked from libraries held at the Wellcome Trust Sanger Institute to cover each chromosome omitting the short arms of acrocentric chromosomes. Essential for the correct interpretation of the array CGH experiments is the exact mapped position of each clone along the chromosomes; although this information is available through various genome browsers on a clone-by-clone basis, it is not easily assembled for large clone collections. We have, therefore, generated a database (Cytoview) within the Ensembl genome browser (Cytoview, www.ensembl.org/homo_sapiens/cytoview) that displays the 1-Mb clone set in relation to the Golden Path sequencing clones. Cytoview also oVers the possibility to download not only lists of clones but also their mapping information, including fluorescence in situ hybridization (FISH) and BAC end data when available. In addition, Ensembl provides automatic updating of all this information with every new assembly of the human genome (Fiegler et al., 2003a).
III. Validation of Array Performance and Data Analysis It is of great importance to prove the reproducibility of the genomic hybridization and to establish characteristics of the clones used to construct the array before applying this method in large sample studies. We, therefore, performed a series of validation experiments including self–self and male–female hybridizations. For a self–self hybridization, the linear ratio of all clones on the array (autosomal clones and chromosome X/Y clones) is expected to be 1:1, which corresponds to a ratio of 0 on a log2 scale (test vs reference, see Fig. 1). In contrast, for a female–male hybridization, the expected linear ratio for chromosome X clones is 2:1, although there should be no copy number changes detectable on autosomal clones. However, although most of the autosomal clones reported the expected 1:1 ratio, we have found several clones showing copy number changes between the two individuals (Fig. 2). Further investigation of these clones by FISH analysis revealed that some clones had been mismapped and could be subsequently placed correctly onto the genome. Others, however, mapped to their correct position, thereby suggesting copy number polymorphism between the two individuals used in the study. Although we were able to detect an increase of ratios for the clones representing chromosome X, most of those clones did not report the expected ratio of 2 (ratio
772
Heike Fiegler and Nigel P. Carter
Fig. 1 Genome-wide plot of ratios for a self–self hybridization.
Fig. 2 Genome-wide plot of ratios for a female–male hybridization.
value of 1 on a log2 scale, see Fig. 3). It is, however, interesting to note that all published methods for human array CGH underestimate female/male ratios for chromosome X clones. The mean linear ratios for chromosome X clones in female–male hybridizations reported by diVerent groups were found to be between 1.65 and 1.73 (Fiegler et al., 2003a; Snijders et al., 2001). It has been suggested that this underestimation of the true ratio on chromosome X clones might be due to incomplete suppression of shared repeat sequences and homology between chromosomes X and Y. It has also been suggested that the inactivation of one chromosome X in female DNA might impair the labeling reaction (Fiegler et al., 2003a). DNA derived from tumor samples often consists of a mixed population of cells (tumor cells and infiltrating normal cells), which can interfere with the correct detection of copy number changes. To test whether our arrays would be sensitive enough to detect single copy number changes even in a mixed cell population, we performed experiments hybridizing female DNA against a mix of 50% male and 50% female DNA onto a small custom array consisting of fifteen chromosome 1 and nine chromosome X clones, which previously reported ratios close to 2:1 in female–male hybridizations (Fig. 4). Although the expected linear ratio for chromosome 1 clones should again be 1:1, the ratios for the chromosome X clones
773
33. Genomic Array Technology
Fig. 3 Chromosome X plot of ratios for a female–male hybridization.
Fig. 4 Hybridization of female DNA versus a mix of 50% male and 50% female DNA. Chromosome 1 clones (open diamonds) and chromosome X clones ( filled diamonds) are shown. The error bars are the standard deviation of triplicates.
should now only be 1.33:1 (instead of 2:1 female vs male). As shown in Fig. 4, all chromosome X clones report an approximate ratio of 1.3, demonstrating that the arrays are indeed sensitive enough to detect single copy number changes even when DNA isolated from mixed populations consists of 50% tumor cells and 50% normal infiltrating cells. For clones to accurately report copy number changes in genomic hybridizations, the nonrepetitive sequence of each clone has to be unique to its location.
774
Heike Fiegler and Nigel P. Carter
Regions of homology or duplication (gene families, paralogs, pseudogenes, segmental duplications, etc.) will contribute to the hybridization signal and could potentially mask copy number changes at the specific locus. Although the presence of duplications can be detected by, for example, FISH analysis onto metaphase chromosomes, only a small proportion of the Golden Path clones have been FISH mapped so far, and because this method is time consuming, it is not suited for the analysis of the whole 1-Mb clone set (3500 clones). As an alternative, we have developed a method that allows the analysis of the response characteristics for all clones directly by hybridization on the array. The basis of this method is to use PCR to amplify flow-sorted chromosomes and add copies of chromosomal material to a series of normal versus normal genomic hybridizations, thereby increasing the copy number of a specific chromosome in the test sample. The hybridization profile for each clone can then be generated by plotting a standard curve of ratio versus additional chromosomal material. By repeating these hybridizations for each chromosome in turn, clones that do not respond appropriately to the addition of the chromosome can be identified easily (Fig. 5). In addition, reduction in the slope of the response curve identifies clones that share homologous sequences with other sites of the genome. Unfortunately, intrachromosomal duplications will not be identified using this strategy (Fiegler et al., 2003a). Another important factor for the reliable detection of single copy number changes in genome-wide array CGH experiments is the successful suppression of repeat sequences present in both target and probe DNA. This is achieved by use of high-quality human Cot1 DNA. The Cot1 fraction of human genomic DNA
Fig. 5 Characteristics for chromosome 1 clones in response to additional copies of chromosome 1 in a normal versus normal hybridization.
33. Genomic Array Technology
775
consists largely of rapidly annealing sequences such as small interspersed nuclear elements (SINEs) and large interspersed nuclear elements (LINEs) repeats. These highly repetitive elements contain either repeat sequences consisting of less than 500-bp elements (SINEs, e.g., Alu repeats) or sequences consisting of more than 5-kb elements (LINEs). When diVerentially labeled male and female DNA was cohybridized in two independent experiments with Cot1 DNA derived from two batches, we observed diVerent hybridization profiles after analyzing the arrays (Fig. 6A). Although co-hybridization with Cot1 DNA derived from batch 1 showed a clear copy number gain on chromosome X clones (expected linear ratio, 2:1; observed mean ratio of all chromosome X clones, 1.72:1), only a slight change in copy number could be seen when batch 2 was used to suppress repeat sequences (observed mean ratio of all chromosome X clones, 1.43:1). Although the expected ratio of 2:1 could not be obtained with any of the Cot1 samples, the observed ratio of 1.72:1, however, compares well with previously reported values of 1.73:1 and 1.65:1, as mentioned previously. Agarose gel electrophoresis of the two batches revealed that the average size of the DNA fragments varied between 100 bp (batch 2) and 400 bp (batch 1) (Fig. 6B). This suggests that Cot1 DNA containing larger DNA fragments (up to 400 bp) might be more eYcient at annealing to the repeat sequences present in both probe and target sequences and in generating specific
Fig. 6 Performance of diVerent batches of human Cot1 DNA in a female versus male hybridization. (A) Ratio plot for Cot1 DNA batch 1; (B) ratio plot for Cot1 DNA batch 2; (C) agarose gel electrophoresis of diVerent samples of Cot1 DNA batches 1 and 2.
776
Heike Fiegler and Nigel P. Carter
hybridization signals. However, the fragment size alone is not the only indicator of Cot1 DNA quality because we have also seen similar variation in the ability of Cot1 DNA batches containing DNA fragments of the same size (data not shown) to suppress repeat sequences.
IV. Array CGH Tumor development and progression is often associated with dramatic copy number changes. Although regions of DNA amplification commonly harbor oncogenes, regions of deletion can potentially harbor tumor suppressor genes. Although considerable information about genome-wide copy number changes in tumors has been provided by conventional CGH, its resolution is limited to typically 10–20 Mb and at best to approximately 3–5 Mb (Lichter et al., 2000). However, as mentioned earlier, by replacing the metaphase chromosomes as the hybridization targets with spatially mapped sequences arrayed onto glass slides, the resolution becomes limited only by the size and spacing of the sequences used (Pinkel et al., 1998; Solinas-Toldo et al., 1997). To compare conventional CGH and array CGH, we have hybridized DNA derived from a female renal cell carcinoma cell line (769P) onto the 1-Mb array described earlier. This cell line has previously been analyzed by conventional CGH and M-FISH. Microarray analysis reliably detected previously identified copy number changes, for example, a single copy deletion on the p-arm on chromosome 1 and a single copy gain on the q-arm on chromosome 1 (Fig. 7). In addition, we detected small copy number changes that had not been detected by any of the conventional methods, for example, a single copy loss on chromosome 9p between 18 and 25 Mb that failed to reach significance in conventional CGH (Fiegler et al., 2003a) (Table II). In addition to detecting single copy number changes, by hybridizing a female colorectal cell line, Colo320 HSR, we could also reliably detect a previously described highly amplified region on chromosome 8q24 containing CMYC (Fig. 8). Our results show an approximately 50-fold copy number increase of this region, which compares well with a previously described study that reported an amplification of approximately 40-fold using the same cell line (Wessendorf et al., 2002). DiVerences in the observed results could be explained using the cell line at a diVerent passage stage and diVerent arrays and hybridization conditions. We have now employed the 1-Mb array in several large-scale studies, one of which involved the screening of 22 bladder tumor-derived cell lines and one normal urothelium-derived cell line (Hurst et al., 2004). The array results were in concordance with numerous genetic changes previously identified by conventional CGH, M-FISH, or LOH analyses. In addition, we confirmed previously identified homozygous deletions harboring tumor suppressor genes such as CDKN2A on 9p21.3, DBCCR1 on 9q33.1, and PTEN on 10q in some of the cell lines used in this study. We also identified several potentially new homozygous deletions and high-level amplifications in this study. Further analysis is
777
33. Genomic Array Technology
Fig. 7 Chromosome 1 profile of ratios for DNA from the cell line 769P versus female DNA. (A) Conventional comparative genomic hybridization (CGH); (B) array CGH.
Table II Comparison of Conventional Comparative Genomic Hybridization (CGH) and Array CGH Analysis of Copy Number Changes in the Cell Line 769P Chromosome 1 3 5 8 9 11 14
Conventional CGH Deletion 1p35->1pter Gain 1q22->1qter Deletion 3p14->3pter Gain 5q31->5qter Gain 8q11.2->8qter Deletion 11q14->11qter Deletion of chromosome 14
Array CGH Single Single Single Single Single Single Single Single
copy copy copy copy copy copy copy copy
deletion 0.2–28.0 Mb gain 160–256 Mb deletion 0.5–72.0 Mb gain 137–184 Mb gain 52–145 Mb deletion 18–25 Mb deletion 90–141 Mb deletion 18–104 Mb
778
Heike Fiegler and Nigel P. Carter
Fig. 8
Chromosome 8 plot of ratios for DNA from the cell line Colo320 HSR versus male DNA.
necessary to study the role of candidate genes located in areas of deletion or amplification.
V. Array CGH for Cytogenetic Analyses Large insert clone arrays have proved to be of great value in the analysis of copy number changes in tumor-derived cell lines and tumor samples. In addition, they are being more widely used in cytogenetic analyses to define regions of deletion or amplification in the genome that might harbor genes contributing to cytogenetically defined syndromes. The most common of these is Down syndrome. Down syndrome is caused by trisomy of chromosome 21 and is characterized by, for example, cognitive impairment, hypotonia, specific phenotypic features such as flat facies and variations in digits, and ridge formation on hands and feet (Epstein, 2001). Patients with Down syndrome are also at risk of congenital heart disease, Hirschsprung’s disease, and other developmental abnormalities. The Down syndrome critical region responsible for the physical phenotype of the disease is thought to be located between 21q22.1 and 21q22.3; however, imbalances of other regions on chromosome 21 might also contribute to the phenotype (Epstein, 2001). The analysis of patients harboring copy number changes of parts of chromosome 21 is, therefore, extremely important because correlation of these partial imbalances with the expressed phenotypes will help identify candidate genes that play a key role in the aspects of Down syndrome (Korenberg et al., 1994). We have analyzed the DNA of a patient with an additional marker chromosome derived from chromosome 21 using the 1-Mb array. The patient did not show any of the dysmorphic features characteristic of Down syndrome but had learning disability with cognitive defects comparable to those seen in patients
33. Genomic Array Technology
779
with Down syndrome. The analysis revealed a partial tetrasomy of chromosome 21 with the location of the breakpoint within chromosome band 21q21.2. The partial tetrasomy did not involve the Down syndrome critical region (Rost et al., 2004). Thus, genes located in the proximal region of chromosome 21 might contribute to aspects of the learning disability but not the dysmorphic features associated with Down syndrome.
VI. Array Painting Although array CGH is becoming a more widespread technique in cytogenetic and tumor analyses to identify genome copy number changes, balanced translocations will not be detected by this methodology. To overcome this problem, we have developed a technique termed array painting, which uses the array technology with flow-sorted chromosome material to rapidly map the constitution and the breakpoints is aberrant chromosomes (Fiegler et al., 2003b). As in array CGH, the resolution is dependent only on the size and the density of the clones spotted onto the array. In array painting, each of the derivative chromosomes in the translocation is flow sorted, amplified, diVerentially labeled with two fluorochromes, and then hybridized to the array. Only clones containing sequences present in the flowsorted chromosomes will show fluorescence above background and the ratio determines from which derivative chromosome the hybridizing DNA has been derived. Intermediate values will be generated if a breakpoint spanning clone is present on the array because sequences present on both derivatives will hybridize to the same clone. For example, the DNA of a patient with a de novo 46,XY, t(17;22)(q21.1;q12.2) translocation was analyzed by array painting (Fiegler et al., 2003b). The two derivative chromosomes were flow sorted, amplified, and hybridized to the array. Hybridization signals were obtained only from clones representing chromosomes 17 and 22 with the exception of weak signals on chromosome 19. This can be explained by the fact that chromosome 19 is positioned very close to the derivative 17 in the flow karyotype and, therefore, will contaminate the derivative 17 sort. The profiles for both, chromosomes 17 and 22, obtained by plotting the fluorescence ratios against the position of the clones along the chromosomes, clearly show a transition from low to high ratio values on chromosome 17 and vice versa on chromosome 22, thereby identifying the breakpoints on both chromosomes within 1-Mb intervals (Fig. 9). Intermediate ratios that would identify breakpoint spanning clones could not be observed in this case (Fiegler et al., 2003b). To identify the breakpoint spanning region, the same derivatives were hybridized to a custom array containing overlapping clones within the previously identified 1-Mb breakpoint intervals. This hybridization identified one clone on chromosome 22 with an intermediate value, and subsequent FISH analysis onto chromosomes from the patient confirmed this clone as spanning the breakpoint. For the chromosome 17 breakpoint, three clones with consistent but diVerent
780
Heike Fiegler and Nigel P. Carter
Fig. 9 Array painting results for chromosomes 17 and 22 in the analysis of a t(17;22) patient. (A) Chromosome 17 profile; (B) chromosome 22 profile.
intermediate ratio values were identified, all three being confirmed as spanning the breakpoint. Fiber FISH analysis revealed that the diVerent absolute values of the intermediate ratios of the breakpoint spanning clones on the array were due to the position of the breakpoint within each clone (Fiegler et al., 2003b). The use of array painting to identify breakpoints in balanced translocations, particularly in more complex cases, is far less time consuming compared to conventional methods, which involve individual hybridizations of large insert clones onto metaphase chromosomes.
VII. Application of Array CGH and Array Painting for Complete Cytogenetic Analyses After analyzing several balanced translocations by array painting, we found that apparently simple translocations are often more complicated than expected. Array painting in combination with array CGH has proven to be a powerful combination of methods to investigate these balanced translocations at a molecular level. We have previously described a case in which a patient had been referred for cytogenetic investigation for triangular facies and the failure to thrive. Conventional cytogenetic analysis revealed a simple balanced translocation with a 46,XX, t(11;12)(q21;p13.33) de novo chromosome constitution. Array painting using the 1-Mb array easily identified the breakpoint on chromosome 12. The most proximal clone mapping 1.02 Mb from the p-terminus of chromosome 12 showed an intermediate ratio (Fig. 10A) and subsequent FISH analysis confirmed that it spanned the breakpoint (Fiegler et al., 2003b).
33. Genomic Array Technology
781
Fig. 10 Array painting and array comparative genomic hybridization (CGH) results in the analysis of a t(11;12) patient. (A) Chromosome 12 array painting profile; (B) chromosome 11 array painting profile; (C) chromosome 11 array CGH profile.
However, the pattern for chromosome 11 was much more complicated. Array painting suggested the breakpoint on chromosome 11 to be between 87 and 101 Mb along the chromosome. For a balanced translocation, we would expect to see a transition of ratios around the breakpoint, but in this case, a complex pattern of ratios within a 14-Mb region, including a potential deletion and inversion, was observed (Fig. 10B). Array CGH confirmed a deletion of about 6 Mb within the breakpoint spanning region (Fig. 10C). Further analysis using FISH to confirm the position of the clones on the derivatives led to the hypothesis that initially one chromosome 11 experienced an inversion/deletion event. This was then followed by a translocation with chromosome 12. It is at this stage impossible to tell whether the deletion followed the inversion or vice versa or if both occurred as a result of the same event (Fiegler et al., 2003b). Cases such as the translocation described earlier demonstrate that the combination of array CGH and array painting for the analysis of balanced translocations
782
Heike Fiegler and Nigel P. Carter
not only identifies breakpoints but also helps reveal deletions and insertions associated with the translocation event. In addition, array CGH will identify intrachromosomal duplications within a single derivative chromosome that will not be detected by array painting, as well as screening the entire genome for unexpected copy number changes that might be associated with the phenotype of a patient.
VIII. ChIP on Genomic Clone Arrays An increasingly popular application for DNA microarrays is the study of DNA–protein interactions. DNA bound to DNA-binding proteins is cross-linked and then immunoprecipitated using an antibody against the protein of interest. DNA is then isolated and subjected to fluorescent labeling before hybridizing onto the arrays. We have used this application to study DNA damage checkpointmediated response in telomere-initiated senescence (Fagagna Fd et al., 2003). Senescence is defined as the exhaustion of proliferative potential and can be triggered by telomere erosion whereby proteins such as phosphorylated H2AX, which are usually involved in DNA double-stranded break repair, directly associate with uncapped telomeres. To prove the direct association between dysfunctional telomeres and DNA damage response in senescent cells, we performed chromatin immunoprecipitation using an antibody against phosphorylated H2AX and hybridized the immunoprecipitated DNA against the original input DNA onto the 1-Mb array (Fig. 11). By calculating the diVerence between the ratios (immunoprecipitated DNA vs input DNA) obtained in senescent and quiescent cells, we could show that phosphorylated H2AX accumulates at a subset of subtelomeric regions in senescent cells, with a preference toward chromosome ends known to harbor short telomeres. Performing the same hybridization on a chromosome 22q tiling path array consisting of overlapping large insert clones that cover the whole of chromosome 22q not only confirmed the association of phosphorylated H2AX to the subtelomeric region of chromosome 22q, but also revealed that phosphorylated H2AX spreads more than 270 kb inward from the chromosome end (Fig. 12).
IX. Conclusion We have described some of the methods and applications of genomic clone microarrays for the study of chromosome rearrangements. The analysis of genomic changes in this way is clearly becoming increasingly more widespread and the resolution provided by this approach is increasing. Large insert clone arrays are being produced with complete chromosome or even whole genome tiling path coverage. We predict that even higher resolution analysis will become possible with the development of quantitative genomic hybridization to arrays
33. Genomic Array Technology
783
Fig. 11 Principle of ChIP on ChIP analysis using genomic microarrays.
Fig. 12 ChIP analysis using an antibody against phosphorylated histone H2AX in senescent cells. Chromosome 22 profile using a 22q tiling path resolution array.
784
Heike Fiegler and Nigel P. Carter
consisting of short genomic fragments. These developments will allow increasingly subtle genomic changes to be identified and correlated with disease phenotypes. References Albertson, D. G., and Pinkel, D. (2003). Genomic microarrays in human genetic disease and cancer. Hum. Mol. Genet. 12(Spec. no. 2), R145–R152. Buckley, P. G., Mantripragada, K. K., Benetkiewicz, M., Tapia-Paez, I., Diaz De Stahl, T., Rosenquist, M., Ali, H., Jarbo, C., De Bustos, C., Hirvela, C., Sinder Wilen, B., Fransson, I., Thyr, C., Johnsson, B. I., Bruder, C. E., Menzel, U., Hergersberg, M., Mandahl, N., Blennow, E., Wedell, A., Beare, D. M., Collins, J. E., Dunham, I., Albertson, D., Pinkel, D., Bastian, B. C., Faruqi, A. F., Lasken, R. S., Ichimura, K., Collins, V. P., and Dumanski, J. P. (2002). A full-coverage, highresolution human chromosome 22 genomic microarray for clinical and research applications. Hum. Mol. Genet. 11, 3221–3229. Epstein, C. J. (2001). The metabolic and molecular bases of inherited disease. McGraw-Hill, New York. Fagagna Fd, F., Reaper, P. M., Clay-Farrace, L., Fiegler, H., Carr, P., Von Zglinicki, T., Saretzki, G., Carter, N. P., and Jackson, S. P. (2003a). A DNA damage checkpoint response in telomere-initiated senescence. Nature 426, 194–198. Fiegler, H., Carr, P., Douglas, E. J., Burford, D. C., Hunt, S., Scott, C. E., Smith, J., Vetrie, D., Gorman, P., Tomlinson, I. P., and Carter, N. P. (2003a). DNA microarrays for comparative genomic hybridization based on DOP-PCR amplification of BAC and PAC clones. Genes Chromosomes Cancer 36, 361–374. Fiegler, H., Gribble, S. M., Burford, D. C., Carr, P., Prigmore, E., Porter, K. M., Clegg, S., Crolla, J. A., Dennis, N. R., Jacobs, P., and Carter, N. P. (2003b). Array painting: A method for the rapid analysis of aberrant chromosomes using DNA microarrays. J. Med. Genet. 40, 664–670. Foreman, P. K., and Davis, R. W. (2000). Real-time PCR-based method for assaying the purity of bacterial artificial chromosome preparations. Biotechniques 29, 410–412. Hodgson, G., Hager, J. H., Volik, S., Hariono, S., Wernick, M., Moore, D., Nowak, N., Albertson, D. G., Pinkel, D., Collins, C., Hanahan, D., and Gray, J. W. (2001). Genome scanning with array CGH delineates regional alterations in mouse islet carcinomas. Nat. Genet. 29, 459–464. Hurst, C. D., Fiegler, H., Carr, P., Williams, S., Carter, N. P., and Knowles, M. A. (2004). Highresolution analysis of genomic copy number alterations in bladder cancer by microarray-based comparative genomic hybridization. Oncogene. 23, 2250–2263. Korenberg, J. R., Chen, X., Schipper, R., Sun, Z., Gonsky, R., Gerwehr, S., Carpenter, N., Daumer, C., Dignan, P., Disteche, C., Graham Jr, J. M., Hudgins, L., McGillivray, B., Miyazaki, K., Ogasawara, N., Park, J. P., Pagon, R., Pueschel, S., Sack, G., Say, B., SchuVenhauer, S., Soukup, S., and Yamanaka, T. (1994). Down syndrome phenotypes: The consequences of chromosomal imbalance. Proc. Natl. Acad. Sci. USA 91, 4997–5001. Lander, E. S., Linton, L. M., Birren, B., Nusbaum, C., Zody, M. C., Baldwin, J., Devon, K., Dewar, K., Doyle, M., FitzHugh, W., Funke, R., Gage, D., Harris, K., Heaford, A., Howland, J., Kann, L., Lehoczky, J., LeVine, R., McEwan, P., McKernan, K., Meldrim, J., Mesirov, J. P., Miranda, C., Morris, W., Naylor, J., Raymond, C., Rosetti, M., Santos, R., Sheridan, A., Sougnez, C., Stange-Thomann, N., Stojanovic, N., Subramanian, A., Wyman, D., Rogers, J., Sulston, J., Ainscough, R., Beck, S., Bentley, D., Burton, J., Clee, C., Carter, N., Coulson, A., Deadman, R., Deloukas, P., Dunham, A., Dunham, I., Durbin, R., French, L., Grafham, D., Gregory, S., Hubbard, T., Humphray, S., Hunt, A., Jones, M., Lloyd, C., McMurray, A., Matthews, L., Mercer, S., Milne, S., Mullikin, J. C., Mungall, A., Plumb, R., Ross, M., Shownkeen, R., Sims, S., Waterston, R. H., Wilson, R. K., Hillier, L. W., McPherson, J. D., Marra, M. A., Mardis, E. R., Fulton, L. A., Chinwalla, A. T., Pepin, K. H., Gish, W. R., Chissoe, S. L., Wendl, M. C., Delehaunty, K. D., Miner, T. L., Delehaunty, A., Kramer, J. B., Cook, L. L., Fulton, R. S.,
33. Genomic Array Technology
785
Johnson, D. L., Minx, P. J., Clifton, S. W., Hawkins, T., Branscomb, E., Predki, P., Richardson, P., Wenning, S., Slezak, T., Doggett, N., Cheng, J. F., Olsen, A., Lucas, S., Elkin, C., Uberbacher, E., Frazier, M., Gibbs, R. A., Muzny, D. M., Scherer, S. E., Bouck, J. B., Sodergren, E. J., Worley, K. C., Rives, C. M., Gorrell, J. H., Metzker, M. L., Naylor, S. L., Kucherlapati, R. S., Nelson, D. L., Weinstock, G. M., Sakaki, Y., Fujiyama, A., Hattori, M., Yada, T., Toyoda, A., Itoh, T., Kawagoe, C., Watanabe, H., Totoki, Y., Taylor, T., Weissenbach, J., Heilig, R., Saurin, W., Artiguenave, F., Brottier, P., Bruls, T., Pelletier, E., Robert, C., Wincker, P., Smith, D. R., Doucette-Stamm, L., Rubenfield, M., Weinstock, K., Lee, H. M., Dubois, J., Rosenthal, A., Platzer, M., Nyakatura, G., Taudien, S., Rump, A., Yang, H., Yu, J., Wang, J., Huang, G., Gu, J., Hood, L., Rowen, L., Madan, A., Qin, S., Davis, R. W., Federspiel, N. A., Abola, A. P., Proctor, M. J., Myers, R. M., Schmutz, J., Dickson, M., Grimwood, J., Cox, D. R., Olson, M. V., Kaul, R., Raymond, C., Shimizu, N., Kawasaki, K., Minoshima, S., Evans, G. A., Athanasiou, M., Schultz, R., Roe, B. A., Chen, F., Pan, H., Ramser, J., Lehrach, H., Reinhardt, R., McCombie, W. R., de la Bastide, M., Dedhia, N., Blocker, H., Hornischer, K., Nordsiek, G., Agarwala, R., Aravind, L., Bailey, J. A., Bateman, A., Batzoglou, S., Birney, E., Bork, P., Brown, D. G., Burge, C. B., Cerutti, L., Chen, H. C., Church, D., Clamp, M., Copley, R. R., Doerks, T., Eddy, S. R., Eichler, E. E., Furey, T. S., Galagan, J., Gilbert, J. G., Harmon, C., Hayashizaki, Y., Haussler, D., Hermjakob, H., Hokamp, K., Jang, W., Johnson, L. S., Jones, T. A., Kasif, S., Kaspryzk, A., Kennedy, S., Kent, W. J., Kitts, P., Koonin, E. V., Korf, I., Kulp, D., Lancet, D., Lowe, T. M., McLysaght, A., Mikkelsen, T., Moran, J. V., Mulder, N., Pollara, V. J., Ponting, C. P., Schuler, G., Schultz, J., Slater, G., Smit, A. F., Stupka, E., Szustakowski, J., Thierry-Mieg, D., Thierry-Mieg, J., Wagner, L., Wallis, J., Wheeler, R., Williams, A., Wolf, Y. I., Wolfe, K. H., Yang, S. P., Yeh, R. F., Collins, F., Guyer, M. S., Peterson, J., Felsenfeld, A., Wetterstrand, K. A., Patrinos, A., Morgan, M. J., Szustakowki, J., de Jong, P., Catanese, J. J., Osoegawa, K., Shizuya, H., Choi, S., and Chen, Y. J. (2001). International Human Genome Sequencing Consortium. Initial sequencing and analysis of the human genome. Nature 409, 860–921. Lichter, P., Joos, S., Bentz, M., and Lampel, S. (2000). Comparative genomic hybridization: Uses and limitations. Semin. Hematol. 37, 348–357. Pinkel, D., Segraves, R., Sudar, D., Clark, S., Poole, I., Kowbel, D., Collins, C., Kuo, W. L., Chen, C., Zhai, Y., Dairkee, S. H., Ljung, B. M., Gray, J. W., and Albertson, D. G. (1998). High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat. Genet. 20, 207–211. Rost, I., Fiegler, H., Fauth, C., Carr, P., Bettecken, T., Kraus, J., Meyer, C., Enders, A., Wirtz, A., Meitinger, T., Carter, N. P., and Speicher, M. R. (2004). Tetrasomy 21pter–>q21.2 in a male infant without typical Down’s syndrome dysmorphic features but moderate mental retardation. J Med Genet. 41, e26. Snijders, A. M., Nowak, N., Segraves, R., Blackwood, S., Brown, N., Conroy, J., Hamilton, G., Hindle, A. K., Huey, B., Kimura, K., Law, S., Myambo, K., Palmer, J., Ylstra, B., Yue, J. P., Gray, J. W., Jain, A. N., Pinkel, D., and Albertson, D. G. (2001). Assembly of microarrays for genomewide measurement of DNA copy number. Nat. Genet. 29, 263–264. Solinas-Toldo, S., Lampel, S., Stilgenbauer, S., Nickolenko, J., Benner, A., Dohner, H., Cremer, T., and Lichter, P. (1997). Matrix-based comparative genomic hybridization: Biochips to screen for genomic imbalances. Genes Chromosomes Cancer 20, 399–407. Telenius, H., Carter, N. P., Bebb, C. E., Nordenskjold, M., Ponder, B. A., and TunnacliVe, A. (1992a). Degenerate oligonucleotide-primed PCR: General amplification of target DNA by a single degenerate primer. Genomics 13, 718–725. Telenius, H., Pelmear, A. H., TunnacliVe, A., Carter, N. P., Behmel, A., Ferguson-Smith, M. A., Nordenskjold, M., Pfragner, R., and Ponder, B. A. (1992b). Cytogenetic analysis by chromosome painting using DOP-PCR amplified flow-sorted chromosomes. Genes Chromosomes Cancer 4, 257–263. Wessendorf, S., Fritz, B., Wrobel, G., Nessling, M., Lampel, S., Goettel, D., Kuepper, M., Joos, S., Hopman, T., Kokocinski, F., Dohner, H., Bentz, M., Schwaenen, C., and Lichter, P. (2002). Automated screening for genomic imbalances using matrix-based comparative genomic hybridization. Lab. Invest. 82, 47–60.
Genomic Array Technology Heike Fiegler and Nigel P. Carter The Wellcome Trust Sanger Institute Wellcome Trust Genome Campus Hinxton, Cambridge CB10 1SA, United Kingdom
I. II. III. IV. V. VI. VII. VIII. IX.
Introduction Construction of Genomic Clone Arrays Using DOP-PCR Amplification Validation of Array Performance and Data Analysis Array CGH Array CGH for Cytogenetic Analyses Array Painting Application of Array CGH and Array Painting for Complete Cytogenetic Analyses ChIP on Genomic Clone Arrays Conclusion References
I. Introduction The sequencing of the human genome in the public domain has generated an extensive resource of mapped and sequenced clones that are revolutionizing molecular cytogenetics. This clone resource finds particular use in the construction of large insert clone DNA microarrays spotted onto glass microscope slides. Originally, these arrays were produced to replace metaphase chromosomes in comparative genomic hybridization (CGH) experiments to detect copy number changes along the genome at an increased resolution (Albertson and Pinkel, 2003). In this way, the resolution of CGH becomes limited only by the insert size and the density of the mapped sequences used. However, because of the increased sensitivity and resolution compared to conventional methodologies, genomic arrays are also being applied for cytogenetic studies of microdeletion and microduplication, as well as for rapid mapping of breakpoints (Fiegler et al., 2003b). METHODS IN CELL BIOLOGY, VOL. 75 Copyright 2004, Elsevier Inc. All rights reserved. 0091-679X/04 $35.00
769
770
Heike Fiegler and Nigel P. Carter
II. Construction of Genomic Clone Arrays Using DOP-PCR Amplification Cosmid, P1, P1 artificial chromosome (PAC), and bacterial artificial chromosome (BAC) clones are being widely used for the construction of genomic DNA microarrays. DNA was originally extracted from large-scale cultures (Pinkel et al., 1998; Solinas-Toldo et al., 1997), which becomes a costly and time-consuming process when expanded to the number of clones required to construct an array with a resolution of 1 Mb (3500 clones). To overcome this problem, several approaches have been applied to amplify the clone DNA enzymatically for spotting, thereby removing the requirement for large-scale DNA preparations. These have included methods such as linker adapter polymerase chain reaction (PCR) (Snijders et al., 2001), rolling circle PCR using Phi29 (Buckley et al., 2002), or degenerate oligonucleotide primed PCR (DOP-PCR) (Hodgson et al., 2001) using an amine-modified version of the standard DOP-PCR primer 6MW (Telenius et al., 1992a,b). Escherichia coli genomic DNA, however, is a common contaminant of DNA preparations of large insert clones. The degree of contamination has been estimated by real-time PCR to be between 6% and 26%, depending on the method used for purification (Foreman and Davis, 2000). This contaminating E. coli DNA will reduce the capacity of each probe spotted on the array to hybridize with the DNA of interest and may contribute to an increased nonspecific background signal. To overcome this problem, we have designed three new DOP-PCR primers (DOP1, DOP2, and DOP3) that were chosen to be eYcient in amplifying human genomic DNA but ineYcient in amplifying the contaminating E. coli DNA (Fiegler et al., 2003a). The sequence and the frequency of the six bases at the 30 end of the new primers in human and E. coli DNA compared to the standard DOP-PCR primer 6MW is shown in Table I. The new primers retain the ability to amplify human sequence but demonstrate poor amplification of the E. coli sequence compared to the standard DOP-PCR primer 6MW. Furthermore, the 30 ends of the primers DOP1 and DOP3 are the reverse complement of each other, which allows the amplification from the complementary strand in the opposite direction, thereby
Table I Degenerate Oligonucleotide Primed Polymerase Chain Reaction Primer for the Amplification of Clones Containing Large Human Inserts Primer
30 sequence
Escherichia coli
Human
DOP1 DOP2 DOP3 6MW
CTAGAA TAGGAG TTCTAG ATGTGG
0.02 0.05 0.02 0.40
0.63 0.49 0.63 0.65
33. Genomic Array Technology
771
providing diVerent but complementary representations of the template sequence. The use of these three new DOP-PCR primers, particularly in combination, revealed a significant increase in sensitivity and reproducibility in genomic hybridizations compared to arrays constructed with the standard DOP-PCR primer 6MW (Fiegler et al., 2003a). Following this strategy, we have constructed a large insert clone DNA microarray composed of 3523 sequencing clones selected from the Golden Path of the human genome (Lander et al., 2001) located at approximately 1-Mb intervals across the human genome. The clones were picked from libraries held at the Wellcome Trust Sanger Institute to cover each chromosome omitting the short arms of acrocentric chromosomes. Essential for the correct interpretation of the array CGH experiments is the exact mapped position of each clone along the chromosomes; although this information is available through various genome browsers on a clone-by-clone basis, it is not easily assembled for large clone collections. We have, therefore, generated a database (Cytoview) within the Ensembl genome browser (Cytoview, www.ensembl.org/homo_sapiens/cytoview) that displays the 1-Mb clone set in relation to the Golden Path sequencing clones. Cytoview also oVers the possibility to download not only lists of clones but also their mapping information, including fluorescence in situ hybridization (FISH) and BAC end data when available. In addition, Ensembl provides automatic updating of all this information with every new assembly of the human genome (Fiegler et al., 2003a).
III. Validation of Array Performance and Data Analysis It is of great importance to prove the reproducibility of the genomic hybridization and to establish characteristics of the clones used to construct the array before applying this method in large sample studies. We, therefore, performed a series of validation experiments including self–self and male–female hybridizations. For a self–self hybridization, the linear ratio of all clones on the array (autosomal clones and chromosome X/Y clones) is expected to be 1:1, which corresponds to a ratio of 0 on a log2 scale (test vs reference, see Fig. 1). In contrast, for a female–male hybridization, the expected linear ratio for chromosome X clones is 2:1, although there should be no copy number changes detectable on autosomal clones. However, although most of the autosomal clones reported the expected 1:1 ratio, we have found several clones showing copy number changes between the two individuals (Fig. 2). Further investigation of these clones by FISH analysis revealed that some clones had been mismapped and could be subsequently placed correctly onto the genome. Others, however, mapped to their correct position, thereby suggesting copy number polymorphism between the two individuals used in the study. Although we were able to detect an increase of ratios for the clones representing chromosome X, most of those clones did not report the expected ratio of 2 (ratio
772
Heike Fiegler and Nigel P. Carter
Fig. 1 Genome-wide plot of ratios for a self–self hybridization.
Fig. 2 Genome-wide plot of ratios for a female–male hybridization.
value of 1 on a log2 scale, see Fig. 3). It is, however, interesting to note that all published methods for human array CGH underestimate female/male ratios for chromosome X clones. The mean linear ratios for chromosome X clones in female–male hybridizations reported by diVerent groups were found to be between 1.65 and 1.73 (Fiegler et al., 2003a; Snijders et al., 2001). It has been suggested that this underestimation of the true ratio on chromosome X clones might be due to incomplete suppression of shared repeat sequences and homology between chromosomes X and Y. It has also been suggested that the inactivation of one chromosome X in female DNA might impair the labeling reaction (Fiegler et al., 2003a). DNA derived from tumor samples often consists of a mixed population of cells (tumor cells and infiltrating normal cells), which can interfere with the correct detection of copy number changes. To test whether our arrays would be sensitive enough to detect single copy number changes even in a mixed cell population, we performed experiments hybridizing female DNA against a mix of 50% male and 50% female DNA onto a small custom array consisting of fifteen chromosome 1 and nine chromosome X clones, which previously reported ratios close to 2:1 in female–male hybridizations (Fig. 4). Although the expected linear ratio for chromosome 1 clones should again be 1:1, the ratios for the chromosome X clones
773
33. Genomic Array Technology
Fig. 3 Chromosome X plot of ratios for a female–male hybridization.
Fig. 4 Hybridization of female DNA versus a mix of 50% male and 50% female DNA. Chromosome 1 clones (open diamonds) and chromosome X clones ( filled diamonds) are shown. The error bars are the standard deviation of triplicates.
should now only be 1.33:1 (instead of 2:1 female vs male). As shown in Fig. 4, all chromosome X clones report an approximate ratio of 1.3, demonstrating that the arrays are indeed sensitive enough to detect single copy number changes even when DNA isolated from mixed populations consists of 50% tumor cells and 50% normal infiltrating cells. For clones to accurately report copy number changes in genomic hybridizations, the nonrepetitive sequence of each clone has to be unique to its location.
774
Heike Fiegler and Nigel P. Carter
Regions of homology or duplication (gene families, paralogs, pseudogenes, segmental duplications, etc.) will contribute to the hybridization signal and could potentially mask copy number changes at the specific locus. Although the presence of duplications can be detected by, for example, FISH analysis onto metaphase chromosomes, only a small proportion of the Golden Path clones have been FISH mapped so far, and because this method is time consuming, it is not suited for the analysis of the whole 1-Mb clone set (3500 clones). As an alternative, we have developed a method that allows the analysis of the response characteristics for all clones directly by hybridization on the array. The basis of this method is to use PCR to amplify flow-sorted chromosomes and add copies of chromosomal material to a series of normal versus normal genomic hybridizations, thereby increasing the copy number of a specific chromosome in the test sample. The hybridization profile for each clone can then be generated by plotting a standard curve of ratio versus additional chromosomal material. By repeating these hybridizations for each chromosome in turn, clones that do not respond appropriately to the addition of the chromosome can be identified easily (Fig. 5). In addition, reduction in the slope of the response curve identifies clones that share homologous sequences with other sites of the genome. Unfortunately, intrachromosomal duplications will not be identified using this strategy (Fiegler et al., 2003a). Another important factor for the reliable detection of single copy number changes in genome-wide array CGH experiments is the successful suppression of repeat sequences present in both target and probe DNA. This is achieved by use of high-quality human Cot1 DNA. The Cot1 fraction of human genomic DNA
Fig. 5 Characteristics for chromosome 1 clones in response to additional copies of chromosome 1 in a normal versus normal hybridization.
33. Genomic Array Technology
775
consists largely of rapidly annealing sequences such as small interspersed nuclear elements (SINEs) and large interspersed nuclear elements (LINEs) repeats. These highly repetitive elements contain either repeat sequences consisting of less than 500-bp elements (SINEs, e.g., Alu repeats) or sequences consisting of more than 5-kb elements (LINEs). When diVerentially labeled male and female DNA was cohybridized in two independent experiments with Cot1 DNA derived from two batches, we observed diVerent hybridization profiles after analyzing the arrays (Fig. 6A). Although co-hybridization with Cot1 DNA derived from batch 1 showed a clear copy number gain on chromosome X clones (expected linear ratio, 2:1; observed mean ratio of all chromosome X clones, 1.72:1), only a slight change in copy number could be seen when batch 2 was used to suppress repeat sequences (observed mean ratio of all chromosome X clones, 1.43:1). Although the expected ratio of 2:1 could not be obtained with any of the Cot1 samples, the observed ratio of 1.72:1, however, compares well with previously reported values of 1.73:1 and 1.65:1, as mentioned previously. Agarose gel electrophoresis of the two batches revealed that the average size of the DNA fragments varied between 100 bp (batch 2) and 400 bp (batch 1) (Fig. 6B). This suggests that Cot1 DNA containing larger DNA fragments (up to 400 bp) might be more eYcient at annealing to the repeat sequences present in both probe and target sequences and in generating specific
Fig. 6 Performance of diVerent batches of human Cot1 DNA in a female versus male hybridization. (A) Ratio plot for Cot1 DNA batch 1; (B) ratio plot for Cot1 DNA batch 2; (C) agarose gel electrophoresis of diVerent samples of Cot1 DNA batches 1 and 2.
776
Heike Fiegler and Nigel P. Carter
hybridization signals. However, the fragment size alone is not the only indicator of Cot1 DNA quality because we have also seen similar variation in the ability of Cot1 DNA batches containing DNA fragments of the same size (data not shown) to suppress repeat sequences.
IV. Array CGH Tumor development and progression is often associated with dramatic copy number changes. Although regions of DNA amplification commonly harbor oncogenes, regions of deletion can potentially harbor tumor suppressor genes. Although considerable information about genome-wide copy number changes in tumors has been provided by conventional CGH, its resolution is limited to typically 10–20 Mb and at best to approximately 3–5 Mb (Lichter et al., 2000). However, as mentioned earlier, by replacing the metaphase chromosomes as the hybridization targets with spatially mapped sequences arrayed onto glass slides, the resolution becomes limited only by the size and spacing of the sequences used (Pinkel et al., 1998; Solinas-Toldo et al., 1997). To compare conventional CGH and array CGH, we have hybridized DNA derived from a female renal cell carcinoma cell line (769P) onto the 1-Mb array described earlier. This cell line has previously been analyzed by conventional CGH and M-FISH. Microarray analysis reliably detected previously identified copy number changes, for example, a single copy deletion on the p-arm on chromosome 1 and a single copy gain on the q-arm on chromosome 1 (Fig. 7). In addition, we detected small copy number changes that had not been detected by any of the conventional methods, for example, a single copy loss on chromosome 9p between 18 and 25 Mb that failed to reach significance in conventional CGH (Fiegler et al., 2003a) (Table II). In addition to detecting single copy number changes, by hybridizing a female colorectal cell line, Colo320 HSR, we could also reliably detect a previously described highly amplified region on chromosome 8q24 containing CMYC (Fig. 8). Our results show an approximately 50-fold copy number increase of this region, which compares well with a previously described study that reported an amplification of approximately 40-fold using the same cell line (Wessendorf et al., 2002). DiVerences in the observed results could be explained using the cell line at a diVerent passage stage and diVerent arrays and hybridization conditions. We have now employed the 1-Mb array in several large-scale studies, one of which involved the screening of 22 bladder tumor-derived cell lines and one normal urothelium-derived cell line (Hurst et al., 2004). The array results were in concordance with numerous genetic changes previously identified by conventional CGH, M-FISH, or LOH analyses. In addition, we confirmed previously identified homozygous deletions harboring tumor suppressor genes such as CDKN2A on 9p21.3, DBCCR1 on 9q33.1, and PTEN on 10q in some of the cell lines used in this study. We also identified several potentially new homozygous deletions and high-level amplifications in this study. Further analysis is
777
33. Genomic Array Technology
Fig. 7 Chromosome 1 profile of ratios for DNA from the cell line 769P versus female DNA. (A) Conventional comparative genomic hybridization (CGH); (B) array CGH.
Table II Comparison of Conventional Comparative Genomic Hybridization (CGH) and Array CGH Analysis of Copy Number Changes in the Cell Line 769P Chromosome 1 3 5 8 9 11 14
Conventional CGH Deletion 1p35->1pter Gain 1q22->1qter Deletion 3p14->3pter Gain 5q31->5qter Gain 8q11.2->8qter Deletion 11q14->11qter Deletion of chromosome 14
Array CGH Single Single Single Single Single Single Single Single
copy copy copy copy copy copy copy copy
deletion 0.2–28.0 Mb gain 160–256 Mb deletion 0.5–72.0 Mb gain 137–184 Mb gain 52–145 Mb deletion 18–25 Mb deletion 90–141 Mb deletion 18–104 Mb
778
Heike Fiegler and Nigel P. Carter
Fig. 8
Chromosome 8 plot of ratios for DNA from the cell line Colo320 HSR versus male DNA.
necessary to study the role of candidate genes located in areas of deletion or amplification.
V. Array CGH for Cytogenetic Analyses Large insert clone arrays have proved to be of great value in the analysis of copy number changes in tumor-derived cell lines and tumor samples. In addition, they are being more widely used in cytogenetic analyses to define regions of deletion or amplification in the genome that might harbor genes contributing to cytogenetically defined syndromes. The most common of these is Down syndrome. Down syndrome is caused by trisomy of chromosome 21 and is characterized by, for example, cognitive impairment, hypotonia, specific phenotypic features such as flat facies and variations in digits, and ridge formation on hands and feet (Epstein, 2001). Patients with Down syndrome are also at risk of congenital heart disease, Hirschsprung’s disease, and other developmental abnormalities. The Down syndrome critical region responsible for the physical phenotype of the disease is thought to be located between 21q22.1 and 21q22.3; however, imbalances of other regions on chromosome 21 might also contribute to the phenotype (Epstein, 2001). The analysis of patients harboring copy number changes of parts of chromosome 21 is, therefore, extremely important because correlation of these partial imbalances with the expressed phenotypes will help identify candidate genes that play a key role in the aspects of Down syndrome (Korenberg et al., 1994). We have analyzed the DNA of a patient with an additional marker chromosome derived from chromosome 21 using the 1-Mb array. The patient did not show any of the dysmorphic features characteristic of Down syndrome but had learning disability with cognitive defects comparable to those seen in patients
33. Genomic Array Technology
779
with Down syndrome. The analysis revealed a partial tetrasomy of chromosome 21 with the location of the breakpoint within chromosome band 21q21.2. The partial tetrasomy did not involve the Down syndrome critical region (Rost et al., 2004). Thus, genes located in the proximal region of chromosome 21 might contribute to aspects of the learning disability but not the dysmorphic features associated with Down syndrome.
VI. Array Painting Although array CGH is becoming a more widespread technique in cytogenetic and tumor analyses to identify genome copy number changes, balanced translocations will not be detected by this methodology. To overcome this problem, we have developed a technique termed array painting, which uses the array technology with flow-sorted chromosome material to rapidly map the constitution and the breakpoints is aberrant chromosomes (Fiegler et al., 2003b). As in array CGH, the resolution is dependent only on the size and the density of the clones spotted onto the array. In array painting, each of the derivative chromosomes in the translocation is flow sorted, amplified, diVerentially labeled with two fluorochromes, and then hybridized to the array. Only clones containing sequences present in the flowsorted chromosomes will show fluorescence above background and the ratio determines from which derivative chromosome the hybridizing DNA has been derived. Intermediate values will be generated if a breakpoint spanning clone is present on the array because sequences present on both derivatives will hybridize to the same clone. For example, the DNA of a patient with a de novo 46,XY, t(17;22)(q21.1;q12.2) translocation was analyzed by array painting (Fiegler et al., 2003b). The two derivative chromosomes were flow sorted, amplified, and hybridized to the array. Hybridization signals were obtained only from clones representing chromosomes 17 and 22 with the exception of weak signals on chromosome 19. This can be explained by the fact that chromosome 19 is positioned very close to the derivative 17 in the flow karyotype and, therefore, will contaminate the derivative 17 sort. The profiles for both, chromosomes 17 and 22, obtained by plotting the fluorescence ratios against the position of the clones along the chromosomes, clearly show a transition from low to high ratio values on chromosome 17 and vice versa on chromosome 22, thereby identifying the breakpoints on both chromosomes within 1-Mb intervals (Fig. 9). Intermediate ratios that would identify breakpoint spanning clones could not be observed in this case (Fiegler et al., 2003b). To identify the breakpoint spanning region, the same derivatives were hybridized to a custom array containing overlapping clones within the previously identified 1-Mb breakpoint intervals. This hybridization identified one clone on chromosome 22 with an intermediate value, and subsequent FISH analysis onto chromosomes from the patient confirmed this clone as spanning the breakpoint. For the chromosome 17 breakpoint, three clones with consistent but diVerent
780
Heike Fiegler and Nigel P. Carter
Fig. 9 Array painting results for chromosomes 17 and 22 in the analysis of a t(17;22) patient. (A) Chromosome 17 profile; (B) chromosome 22 profile.
intermediate ratio values were identified, all three being confirmed as spanning the breakpoint. Fiber FISH analysis revealed that the diVerent absolute values of the intermediate ratios of the breakpoint spanning clones on the array were due to the position of the breakpoint within each clone (Fiegler et al., 2003b). The use of array painting to identify breakpoints in balanced translocations, particularly in more complex cases, is far less time consuming compared to conventional methods, which involve individual hybridizations of large insert clones onto metaphase chromosomes.
VII. Application of Array CGH and Array Painting for Complete Cytogenetic Analyses After analyzing several balanced translocations by array painting, we found that apparently simple translocations are often more complicated than expected. Array painting in combination with array CGH has proven to be a powerful combination of methods to investigate these balanced translocations at a molecular level. We have previously described a case in which a patient had been referred for cytogenetic investigation for triangular facies and the failure to thrive. Conventional cytogenetic analysis revealed a simple balanced translocation with a 46,XX, t(11;12)(q21;p13.33) de novo chromosome constitution. Array painting using the 1-Mb array easily identified the breakpoint on chromosome 12. The most proximal clone mapping 1.02 Mb from the p-terminus of chromosome 12 showed an intermediate ratio (Fig. 10A) and subsequent FISH analysis confirmed that it spanned the breakpoint (Fiegler et al., 2003b).
33. Genomic Array Technology
781
Fig. 10 Array painting and array comparative genomic hybridization (CGH) results in the analysis of a t(11;12) patient. (A) Chromosome 12 array painting profile; (B) chromosome 11 array painting profile; (C) chromosome 11 array CGH profile.
However, the pattern for chromosome 11 was much more complicated. Array painting suggested the breakpoint on chromosome 11 to be between 87 and 101 Mb along the chromosome. For a balanced translocation, we would expect to see a transition of ratios around the breakpoint, but in this case, a complex pattern of ratios within a 14-Mb region, including a potential deletion and inversion, was observed (Fig. 10B). Array CGH confirmed a deletion of about 6 Mb within the breakpoint spanning region (Fig. 10C). Further analysis using FISH to confirm the position of the clones on the derivatives led to the hypothesis that initially one chromosome 11 experienced an inversion/deletion event. This was then followed by a translocation with chromosome 12. It is at this stage impossible to tell whether the deletion followed the inversion or vice versa or if both occurred as a result of the same event (Fiegler et al., 2003b). Cases such as the translocation described earlier demonstrate that the combination of array CGH and array painting for the analysis of balanced translocations
782
Heike Fiegler and Nigel P. Carter
not only identifies breakpoints but also helps reveal deletions and insertions associated with the translocation event. In addition, array CGH will identify intrachromosomal duplications within a single derivative chromosome that will not be detected by array painting, as well as screening the entire genome for unexpected copy number changes that might be associated with the phenotype of a patient.
VIII. ChIP on Genomic Clone Arrays An increasingly popular application for DNA microarrays is the study of DNA–protein interactions. DNA bound to DNA-binding proteins is cross-linked and then immunoprecipitated using an antibody against the protein of interest. DNA is then isolated and subjected to fluorescent labeling before hybridizing onto the arrays. We have used this application to study DNA damage checkpointmediated response in telomere-initiated senescence (Fagagna Fd et al., 2003). Senescence is defined as the exhaustion of proliferative potential and can be triggered by telomere erosion whereby proteins such as phosphorylated H2AX, which are usually involved in DNA double-stranded break repair, directly associate with uncapped telomeres. To prove the direct association between dysfunctional telomeres and DNA damage response in senescent cells, we performed chromatin immunoprecipitation using an antibody against phosphorylated H2AX and hybridized the immunoprecipitated DNA against the original input DNA onto the 1-Mb array (Fig. 11). By calculating the diVerence between the ratios (immunoprecipitated DNA vs input DNA) obtained in senescent and quiescent cells, we could show that phosphorylated H2AX accumulates at a subset of subtelomeric regions in senescent cells, with a preference toward chromosome ends known to harbor short telomeres. Performing the same hybridization on a chromosome 22q tiling path array consisting of overlapping large insert clones that cover the whole of chromosome 22q not only confirmed the association of phosphorylated H2AX to the subtelomeric region of chromosome 22q, but also revealed that phosphorylated H2AX spreads more than 270 kb inward from the chromosome end (Fig. 12).
IX. Conclusion We have described some of the methods and applications of genomic clone microarrays for the study of chromosome rearrangements. The analysis of genomic changes in this way is clearly becoming increasingly more widespread and the resolution provided by this approach is increasing. Large insert clone arrays are being produced with complete chromosome or even whole genome tiling path coverage. We predict that even higher resolution analysis will become possible with the development of quantitative genomic hybridization to arrays
33. Genomic Array Technology
783
Fig. 11 Principle of ChIP on ChIP analysis using genomic microarrays.
Fig. 12 ChIP analysis using an antibody against phosphorylated histone H2AX in senescent cells. Chromosome 22 profile using a 22q tiling path resolution array.
784
Heike Fiegler and Nigel P. Carter
consisting of short genomic fragments. These developments will allow increasingly subtle genomic changes to be identified and correlated with disease phenotypes. References Albertson, D. G., and Pinkel, D. (2003). Genomic microarrays in human genetic disease and cancer. Hum. Mol. Genet. 12(Spec. no. 2), R145–R152. Buckley, P. G., Mantripragada, K. K., Benetkiewicz, M., Tapia-Paez, I., Diaz De Stahl, T., Rosenquist, M., Ali, H., Jarbo, C., De Bustos, C., Hirvela, C., Sinder Wilen, B., Fransson, I., Thyr, C., Johnsson, B. I., Bruder, C. E., Menzel, U., Hergersberg, M., Mandahl, N., Blennow, E., Wedell, A., Beare, D. M., Collins, J. E., Dunham, I., Albertson, D., Pinkel, D., Bastian, B. C., Faruqi, A. F., Lasken, R. S., Ichimura, K., Collins, V. P., and Dumanski, J. P. (2002). A full-coverage, highresolution human chromosome 22 genomic microarray for clinical and research applications. Hum. Mol. Genet. 11, 3221–3229. Epstein, C. J. (2001). The metabolic and molecular bases of inherited disease. McGraw-Hill, New York. Fagagna Fd, F., Reaper, P. M., Clay-Farrace, L., Fiegler, H., Carr, P., Von Zglinicki, T., Saretzki, G., Carter, N. P., and Jackson, S. P. (2003a). A DNA damage checkpoint response in telomere-initiated senescence. Nature 426, 194–198. Fiegler, H., Carr, P., Douglas, E. J., Burford, D. C., Hunt, S., Scott, C. E., Smith, J., Vetrie, D., Gorman, P., Tomlinson, I. P., and Carter, N. P. (2003a). DNA microarrays for comparative genomic hybridization based on DOP-PCR amplification of BAC and PAC clones. Genes Chromosomes Cancer 36, 361–374. Fiegler, H., Gribble, S. M., Burford, D. C., Carr, P., Prigmore, E., Porter, K. M., Clegg, S., Crolla, J. A., Dennis, N. R., Jacobs, P., and Carter, N. P. (2003b). Array painting: A method for the rapid analysis of aberrant chromosomes using DNA microarrays. J. Med. Genet. 40, 664–670. Foreman, P. K., and Davis, R. W. (2000). Real-time PCR-based method for assaying the purity of bacterial artificial chromosome preparations. Biotechniques 29, 410–412. Hodgson, G., Hager, J. H., Volik, S., Hariono, S., Wernick, M., Moore, D., Nowak, N., Albertson, D. G., Pinkel, D., Collins, C., Hanahan, D., and Gray, J. W. (2001). Genome scanning with array CGH delineates regional alterations in mouse islet carcinomas. Nat. Genet. 29, 459–464. Hurst, C. D., Fiegler, H., Carr, P., Williams, S., Carter, N. P., and Knowles, M. A. (2004). Highresolution analysis of genomic copy number alterations in bladder cancer by microarray-based comparative genomic hybridization. Oncogene. 23, 2250–2263. Korenberg, J. R., Chen, X., Schipper, R., Sun, Z., Gonsky, R., Gerwehr, S., Carpenter, N., Daumer, C., Dignan, P., Disteche, C., Graham Jr, J. M., Hudgins, L., McGillivray, B., Miyazaki, K., Ogasawara, N., Park, J. P., Pagon, R., Pueschel, S., Sack, G., Say, B., SchuVenhauer, S., Soukup, S., and Yamanaka, T. (1994). Down syndrome phenotypes: The consequences of chromosomal imbalance. Proc. Natl. Acad. Sci. USA 91, 4997–5001. Lander, E. S., Linton, L. M., Birren, B., Nusbaum, C., Zody, M. C., Baldwin, J., Devon, K., Dewar, K., Doyle, M., FitzHugh, W., Funke, R., Gage, D., Harris, K., Heaford, A., Howland, J., Kann, L., Lehoczky, J., LeVine, R., McEwan, P., McKernan, K., Meldrim, J., Mesirov, J. P., Miranda, C., Morris, W., Naylor, J., Raymond, C., Rosetti, M., Santos, R., Sheridan, A., Sougnez, C., Stange-Thomann, N., Stojanovic, N., Subramanian, A., Wyman, D., Rogers, J., Sulston, J., Ainscough, R., Beck, S., Bentley, D., Burton, J., Clee, C., Carter, N., Coulson, A., Deadman, R., Deloukas, P., Dunham, A., Dunham, I., Durbin, R., French, L., Grafham, D., Gregory, S., Hubbard, T., Humphray, S., Hunt, A., Jones, M., Lloyd, C., McMurray, A., Matthews, L., Mercer, S., Milne, S., Mullikin, J. C., Mungall, A., Plumb, R., Ross, M., Shownkeen, R., Sims, S., Waterston, R. H., Wilson, R. K., Hillier, L. W., McPherson, J. D., Marra, M. A., Mardis, E. R., Fulton, L. A., Chinwalla, A. T., Pepin, K. H., Gish, W. R., Chissoe, S. L., Wendl, M. C., Delehaunty, K. D., Miner, T. L., Delehaunty, A., Kramer, J. B., Cook, L. L., Fulton, R. S.,
33. Genomic Array Technology
785
Johnson, D. L., Minx, P. J., Clifton, S. W., Hawkins, T., Branscomb, E., Predki, P., Richardson, P., Wenning, S., Slezak, T., Doggett, N., Cheng, J. F., Olsen, A., Lucas, S., Elkin, C., Uberbacher, E., Frazier, M., Gibbs, R. A., Muzny, D. M., Scherer, S. E., Bouck, J. B., Sodergren, E. J., Worley, K. C., Rives, C. M., Gorrell, J. H., Metzker, M. L., Naylor, S. L., Kucherlapati, R. S., Nelson, D. L., Weinstock, G. M., Sakaki, Y., Fujiyama, A., Hattori, M., Yada, T., Toyoda, A., Itoh, T., Kawagoe, C., Watanabe, H., Totoki, Y., Taylor, T., Weissenbach, J., Heilig, R., Saurin, W., Artiguenave, F., Brottier, P., Bruls, T., Pelletier, E., Robert, C., Wincker, P., Smith, D. R., Doucette-Stamm, L., Rubenfield, M., Weinstock, K., Lee, H. M., Dubois, J., Rosenthal, A., Platzer, M., Nyakatura, G., Taudien, S., Rump, A., Yang, H., Yu, J., Wang, J., Huang, G., Gu, J., Hood, L., Rowen, L., Madan, A., Qin, S., Davis, R. W., Federspiel, N. A., Abola, A. P., Proctor, M. J., Myers, R. M., Schmutz, J., Dickson, M., Grimwood, J., Cox, D. R., Olson, M. V., Kaul, R., Raymond, C., Shimizu, N., Kawasaki, K., Minoshima, S., Evans, G. A., Athanasiou, M., Schultz, R., Roe, B. A., Chen, F., Pan, H., Ramser, J., Lehrach, H., Reinhardt, R., McCombie, W. R., de la Bastide, M., Dedhia, N., Blocker, H., Hornischer, K., Nordsiek, G., Agarwala, R., Aravind, L., Bailey, J. A., Bateman, A., Batzoglou, S., Birney, E., Bork, P., Brown, D. G., Burge, C. B., Cerutti, L., Chen, H. C., Church, D., Clamp, M., Copley, R. R., Doerks, T., Eddy, S. R., Eichler, E. E., Furey, T. S., Galagan, J., Gilbert, J. G., Harmon, C., Hayashizaki, Y., Haussler, D., Hermjakob, H., Hokamp, K., Jang, W., Johnson, L. S., Jones, T. A., Kasif, S., Kaspryzk, A., Kennedy, S., Kent, W. J., Kitts, P., Koonin, E. V., Korf, I., Kulp, D., Lancet, D., Lowe, T. M., McLysaght, A., Mikkelsen, T., Moran, J. V., Mulder, N., Pollara, V. J., Ponting, C. P., Schuler, G., Schultz, J., Slater, G., Smit, A. F., Stupka, E., Szustakowski, J., Thierry-Mieg, D., Thierry-Mieg, J., Wagner, L., Wallis, J., Wheeler, R., Williams, A., Wolf, Y. I., Wolfe, K. H., Yang, S. P., Yeh, R. F., Collins, F., Guyer, M. S., Peterson, J., Felsenfeld, A., Wetterstrand, K. A., Patrinos, A., Morgan, M. J., Szustakowki, J., de Jong, P., Catanese, J. J., Osoegawa, K., Shizuya, H., Choi, S., and Chen, Y. J. (2001). International Human Genome Sequencing Consortium. Initial sequencing and analysis of the human genome. Nature 409, 860–921. Lichter, P., Joos, S., Bentz, M., and Lampel, S. (2000). Comparative genomic hybridization: Uses and limitations. Semin. Hematol. 37, 348–357. Pinkel, D., Segraves, R., Sudar, D., Clark, S., Poole, I., Kowbel, D., Collins, C., Kuo, W. L., Chen, C., Zhai, Y., Dairkee, S. H., Ljung, B. M., Gray, J. W., and Albertson, D. G. (1998). High resolution analysis of DNA copy number variation using comparative genomic hybridization to microarrays. Nat. Genet. 20, 207–211. Rost, I., Fiegler, H., Fauth, C., Carr, P., Bettecken, T., Kraus, J., Meyer, C., Enders, A., Wirtz, A., Meitinger, T., Carter, N. P., and Speicher, M. R. (2004). Tetrasomy 21pter–>q21.2 in a male infant without typical Down’s syndrome dysmorphic features but moderate mental retardation. J Med Genet. 41, e26. Snijders, A. M., Nowak, N., Segraves, R., Blackwood, S., Brown, N., Conroy, J., Hamilton, G., Hindle, A. K., Huey, B., Kimura, K., Law, S., Myambo, K., Palmer, J., Ylstra, B., Yue, J. P., Gray, J. W., Jain, A. N., Pinkel, D., and Albertson, D. G. (2001). Assembly of microarrays for genomewide measurement of DNA copy number. Nat. Genet. 29, 263–264. Solinas-Toldo, S., Lampel, S., Stilgenbauer, S., Nickolenko, J., Benner, A., Dohner, H., Cremer, T., and Lichter, P. (1997). Matrix-based comparative genomic hybridization: Biochips to screen for genomic imbalances. Genes Chromosomes Cancer 20, 399–407. Telenius, H., Carter, N. P., Bebb, C. E., Nordenskjold, M., Ponder, B. A., and TunnacliVe, A. (1992a). Degenerate oligonucleotide-primed PCR: General amplification of target DNA by a single degenerate primer. Genomics 13, 718–725. Telenius, H., Pelmear, A. H., TunnacliVe, A., Carter, N. P., Behmel, A., Ferguson-Smith, M. A., Nordenskjold, M., Pfragner, R., and Ponder, B. A. (1992b). Cytogenetic analysis by chromosome painting using DOP-PCR amplified flow-sorted chromosomes. Genes Chromosomes Cancer 4, 257–263. Wessendorf, S., Fritz, B., Wrobel, G., Nessling, M., Lampel, S., Goettel, D., Kuepper, M., Joos, S., Hopman, T., Kokocinski, F., Dohner, H., Bentz, M., Schwaenen, C., and Lichter, P. (2002). Automated screening for genomic imbalances using matrix-based comparative genomic hybridization. Lab. Invest. 82, 47–60.