Array construction and application of genomic microarrays

Array construction using DOP PCR

Large insert clones such as cosmids, bacterial artificial chromosomes (BACs) and P1 artificial chromosomes (PACs) are used typically for the construction of genomic DNA microarrays. DNA derived from these clones was originally prepared from large-scale cultures (3, 4), which, when expanded to the number of clones that are required to construct an array with a resolution of 1 Mb (~3500 clones) or even a whole genome tiling path array (~37000 clones) becomes a costly and time-consuming procedure. Therefore, several PCR-based methods have been developed to remove the requirement of large-scale cultures. These include ligation-mediated PCR (5, 6), rolling circle PCR using Phi29 (7), or degenerate oligonucleotide primed

PCR (DOP PCR) using an amine-modified version of the standard DOP PCR primer 6MW (8).

We have also chosen a DOP PCR-based approach for the construction of a large insert clone DNA microarray. DOP PCR uses a mixture of primers, whereby each of the primers consists of a defined 5' sequence and six defined bases at the 3' end flanking random hexanucleotide sequence (9, 10). Thus, DOP PCR allows a general amplification of any target DNA. However, the reaction relies on the frequency with which bases in the target sequence match the six 3' bases of the primer. If the matches to the primer are infrequent in a particular target sequence, the DOP PCR product will be a poor representation of the target (9, 10). In order to increase the level of representation, we have designed three DOP PCR primers that would be efficient in amplifying human genomic DNA, but inefficient in the amplification of Escherichia coli DNA, a known contaminant of DNA preparations from clone cultures (11). Amplification of the contaminating E. coli DNA together with the clone insert will reduce the capacity of the spotted product to hybridize with the DNA of interest thus reducing the sensitivity of the array. We found that the use of these three DOP PCR primers in combination resulted in a significant increase in signal to background ratio, sensitivity and reproducibility. Following this strategy, we have constructed a large insert clone DNA microarray consisting of 3523 Golden Path sequencing clones spaced at approximately 1 Mb intervals across the human genome. Each of the clones was amplified in three separate PCR reactions using the three different DOP PCR primers. This was followed by a secondary PCR reaction with a 5' amine-modified primer designed such that the 3' end matched the 5' end of the DOP PCR primers to enable cova-lent attachment of the products to specially coated glass slides (12). In order to interpret and report copy number changes correctly across the genome it is essential to map the exact position of each clone along the chromosomes. Clone information is available through various genome browsers but often not easily accessible for large clone collections. We have therefore generated a specific view of the human genome within the Ensembl genome browser (Cytoview, that displays the 1 Mb clone set in relation to the Golden Path sequencing clones. In particular, Cytoview facilitates the downloading of clone lists (from specific regions of interest to whole chromosomes and the whole genome) together with their corresponding map position. Additional information, such as location by fluorescent in situ hybridization (FISH), BAC-end sequence data, genes or expressed sequence tags (ESTs) for any region of interest can be viewed within the context of the 1 Mb clone set. Ensembl also provides an automatic update of all this information with every new assembly of the human genome (12).

Array CGH in tumor biology

Many tumors are characterized by the presence of copy number alterations ranging from gains or losses of whole chromosomes to less than a megabase of DNA. In human cancers, regions of gain potentially harbor oncogenes, while tumor suppressor genes are likely to be located within regions of copy number loss. We have used the 1 Mb array described above in several large-

scale studies, one of which involved the screening of 22 bladder-tumor-derived cell lines. The array results confirmed numerous genetic changes previously identified by conventional CGH, M-FISH, or LOH analyses. The most frequent copy number alterations included a complete or partial loss of chromosome 4q and gain of chromosome 20q. In addition to previously identified homozygous deletions on chromosomes 9p21.3 (harboring CDKN2A), 9q33.1 (harboring DBCCR1) and 10q (harboring PTEN) in some of the cell lines used in this study, we could also identify several potentially new homozygous deletions and high-level amplifications with a previously reported amplification at 6p22.3 being the most frequent. Subsequent real time PCR analysis of genes in that region revealed a novel candidate gene (NM_017774) with consistent over-expression in all the cell lines displaying the 6p22.3 amplicon (13).

Another large-scale study used the 1 Mb array to investigate copy number changes in 48 colorectal cancer cell lines and 37 colorectal primary carcinomas. Colorectal cancer (CRC) is the second most common malignancy in the Western World and accounts for around 20 000 deaths in the UK per year. So far, only a few genes have been identified in which somatic mutations contribute to the pathogenesis of CRC. These include APC, SMAD4, p53, KRAS, and p-catenin. Conventional CGH analyses of CRCs revealed consistent copy number changes such as gain of chromosomes 20, 13 and 8q and loss of chromosomes 18q and 8p. These observations were confirmed by array CGH; however, due to the increased resolution we could also identify the most frequently altered regions within these large-scale gains or losses. The most frequently gained regions were detected on chromosomes 20q (harboring potential candidate genes such as LIVIN, HD54, EEF1A2 and PTK6), and 13q (harboring FLT1 and FLT3), and the most frequently lost region on chromosome 18q (harboring SMAD2 and SMAD4). In addition, we detected previously unreported copy number changes, such as a common region of amplification on chromosome 17q11.2-q12 (harboring AATF and TBC1D3) and a common deletion on chromosome 1q41 (harboring TGFp2). Interestingly, both, colorectal cancer cell lines and primary carcinomas revealed a strikingly similar pattern of copy number alterations across the genome (14).

Array CGH for cytogenetic analyses

Array CGH is being increasingly applied to the identification and analysis of sub-microscopic deletions (microdeletions) or gains (microduplications) in patients with constitutional genomic rearrangements in order to identify genes contributing to the patient's phenotype. Array CGH-based approaches are particularly suited for the analysis of patients with learning disability and dysmorphology (15-17). In a study of 50 patients with cyto-genetically normal karyotypes but with learning disability and dysmorphic features, we identified 12 patients (24% in total) harboring subtle genomic copy number changes. These copy number aberrations ranged in size from those involving only a single clone to regions as large as 14 Mb. Interestingly, none of the rearrangements coincided with previously reported cases from similar patient groups. We detected seven different microdeletions of which six were de novo and one deletion inherited from a phenotypically normal parent, and five different microduplications of which one was de novo and four inherited. While the de novo rearrangements are likely to account for the phenotype of the patient, the pathogenic significance, if any, of the inherited copy number changes is unknown (15).

Array CGH is also proving valuable in the more detailed analysis of genomic regions contributing to cytogenetically defined syndromes. We have used array CGH to study a patient with a chromosome 21-derived marker chromosome who displayed some features similar to Down syndrome. Down syndrome is usually caused by trisomy of chromosome 21 and is characterized by, for example, cognitive impairment, hypotonia and specific phenotypic features such as flat faces and ridge formation on hands and feet. The Down syndrome critical region, which is thought to be responsible for the physical phenotype of the patients, has been mapped to 21q22.1-21q22.3. However, it is not clear whether other regions on chromosome 21 contribute to the complex phenotype of Down syndrome. The patient with the chromosome 21-derived marker chromosome did not show any of the characteristic dysmorphic features of Down syndrome, but presented with learning disability and cognitive defects typical of Down syndrome. The array CGH analysis revealed a partial tetrasomy of chromosome 21 that did not involve the Down syndrome critical region. We suggest that the genes located within the amplified region may contribute to aspects of learning disability and cognitive impairment, but do not play a role in the typical dysmorphic features associated with Down syndrome (18).

Array painting

Although array CGH is able to identify genomic copy number alterations, chromosome rearrangements which do not result in genomic imbalance, such as reciprocal, balanced translocations, cannot be detected by this method. We have therefore developed a technique ('array painting') that utilizes flow-sorted derivative chromosomes in combination with the array technology to analyze the constitution and the breakpoints in balanced translocations. Briefly, each derivative chromosome involved in the translocation is flow-sorted, amplified by DOP PCR, differentially labeled and hybridized to the arrays. Signal intensities above background will only be obtained from clones that contain sequences present in the flow-sorted derivative chromosomes. The ratio of the intensities determines from which derivative chromosome the hybridizing DNA sequence has been derived. Breakpoint-spanning clones are identified by intermediate ratio values when sequences present on both derivatives hybridize to the same clone (19). For example, we have analyzed the DNA of a patient with a de novo 46,XY,t(17;22)(q21.1;q12.2) translocation by array painting. The derivative chromosomes 17 and 22 were flow-sorted, differentially labeled and hybridized to the 1 Mb array. Only clones representing chromosomes 17 and 22 showed strong signals above background. We also obtained weak signals on chromosome 19. However, chromosome 19 is close to the derivative chromosome 17 on the flow karyotype and inevitably contaminated the derivative chromosome 17 isolation during the sort. Plotting the fluorescent intensities against the position of the clones along the chromosomes clearly showed a transition from low to high ratios (chromosome 17) or vice versa (chromosome 22), but intermediate ratios that would identify breakpoint spanning clones were not detected using the 1 Mb array (Figure 12.1). However, by constructing a custom array consisting of tiling path clones within the previously identified 1 Mb intervals and hybridizing the same derivative DNAs, breakpoint-spanning clones were identified and could be confirmed by FISH analysis (19).

der22 der17

der22 der17


Chromosome 17 clones

Chromosome 22 clones o to

0 10 20 30 40 50 60 70 80 90 Distance in Mb o to

20 30

Distance in Mb

Figure 12.1.

Array painting results for chromosomes 17 and 22 in the analysis of a t(17;22) patient. The flow-sorted derivative chromosomes were differentially labeled with Cy3 (der17) and Cy5 (der22) and hybridized to the arrays. Only clones that correspond to sequences present in the derivative chromosomes will show signal intensities above background. The fluorescent ratio of the hybridizing clones will either be high or low depending to which derivative chromosome the sequence corresponds (e.g. clones mapping to the chromosome 17 sequences on the der17 will generate low ratios whilst clones mapping to the chromosome 17 sequences on the der22 will give high ratios). Breakpoint-spanning clones are identified by an intermediate ratio as both derivative chromosomes will hybridize to the same clone. Taken from Journal of Medical Genetics, Vol. 40, pages 664-670. Copyright (2003), with permission from the BMJ Publishing Group.

While array painting is particularly efficient in the analysis of the breakpoints of apparently balanced translocations, only the derivative chromosomes are screened. A more complete analysis of the patient's karyotype can be achieved by combining array painting of clearly rearranged chromosomes with the more general array CGH approach for the detection of genomic imbalances. In a study of 10 patients with apparently balanced chromosome translocations, we found a surprisingly high level of breakpoint complexity and genomic imbalance. Six of the patients studied showed complex rearrangements involving deletions, inversions and insertions at or near a breakpoint, the involvement of additional chromosomes in the translocation process and previously undetected microdeletions or microduplications unrelated to the translocation (20).

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook

Post a comment