By Monica Heger
This article was originally published on December 23, 2009.
Researchers from the Wellcome Trust Sanger Institute reported last month in Nature that they have detected more than 2,000 chromosomal rearrangements in 24 different types of breast cancers on the Illumina platform.
Previous sequencing technologies were prohibitively expensive to detect chromosomal rearrangements, so cytogenetic methods such as karyotyping, fluorescence in situ hybridization, and comparative genomic hybridization arrays have been the main tools used to detect and study them to date. But even cytogenetic tools were not able to identify the sheer number of rearrangements that were detected in this study, according to the authors, suggesting that next-gen sequencing may be particularly useful in the study of oncology.
"The really great power of next-gen sequencing is the ability to identify genomic rearrangements," said Matthew Meyerson of the Dana Farber Cancer Institute, who was not involved with the research. "This is more evidence for the power of the technology," he added.
Michael Stratton, who heads Sanger's Cancer Genome Project and who led the study described in Nature, agreed. "Cytogenetics was the way we picked up rearrangements before, but cytogenetics is bad at detecting intra-chromosomal rearrangements, and by far, the large majority of the rearrangements [we found] were intra-chromosomal," Stratton said.
In addition, some of the rearrangements detected with the Illumina GA had a very low, or no, copy number change, which would have rendered them essentially invisible to CGH arrays, he added.
Meyerson added that the ability of next-gen technologies to sequence whole genomes at greater depth for a lower cost makes it possible to detect the rearrangements. He said that in this case, the ability to get about a physical 10-fold coverage made it possible to detect the rearrangements.
The researchers used the Illumina GA and a paired-end sequencing strategy. They created 500-base-pair sequence libraries and generated sequence reads of 37 base pairs. They then mapped the sequence back to the reference genome to look for rearrangements, which they identified as "discordant paired-end reads that did not map back to the reference human genome in the correct orientation with respect to each other and/or within ~500 base pairs of each other."
Those rearrangements were then compared to a normal genome from the same individual to confirm that they did not originate in the germline. In total they detected 2,166 rearrangements from the 24 different breast cancers.
Next, they used Sanger sequencing to sequence the rearrangements across the break point, and were able to sequence 1,821 of the confirmed rearrangements. Among the 24 different cancer lines, there was wide variety in the number of rearrangements. Some lines only had one or two, while one had over 200 rearrangements. Over 1,000 intra-chromosomal rearrangements were found and 239 inter-chromosomal rearrangements. Of the intra-chromosomal rearrangements, all different types were detected, including 739 tandem duplications, 357 deletions, and 215 with inverted orientation.
"This reveals the complexity of the rearrangement patterns that are there," said Stratton. "The number of rearrangements that are sometimes there indicates that we're going to have to search pretty carefully and document carefully in future large-scale cancer sequencing initiatives," he added.
Despite the number of rearrangements that they did detect, however, Stratton estimated that they still only found about half of the total number of rearrangements.
Stratton said that this study could change researchers' understanding of the structure of cancer genomes. For example, tandem duplications were the most common class of rearrangement detected in the study, yet these had not been documented very extensively previously.
"That's a particular class that has not been associated with breast cancer before," he said, adding that the tandem duplications imply that something is wrong with the DNA maintenance machinery, possibly a gene involved in DNA repair.
There was also an overrepresentation of breakages in the genic area of the DNA. Stratton said that he and his colleagues have not yet determined what would cause the breakages to occur more often in genic regions.
Earlier in December, the Sanger team reported in Nature that they sequenced melanoma and lung cancers on the Illumina and SOLiD platforms respectively (see In Sequence 9/22/2009), demonstrating that next-gen sequencing technology could be used to create a comprehensive map of somatic mutations. This study builds on that work, showing that not only can next-gen technology catalog mutations, but it can delve into the structure of cancer genomes at a detail and level that was previously impossible to achieve.
The Sanger group is planning more large-scale cancer studies, on breast cancer in particular. Stratton said they plan to sequence 1,500 breast cancers. For these projects he said they will use the Illumina platform and a paired-end sequencing approach.
While the Sanger Institute previously tested the SOLiD system, he said they were unable to support both platforms, and since they already had more Illumina systems, they chose to stick with that (see In Sequence 11/18/2009). He added that the SOLiD system performed well. "As far as we were concerned there was little to distinguish the two platforms," Stratton said.
Additionally, Stratton said that the Sanger team will be working to improve its informatics tools for the analysis of chromosomal rearrangements.
"It's clear that there are additional patterns of rearrangements in other cancers," he said. "Starting with paired-end sequencing is good, but there will be other strategies with the informatics that that we'll have to adopt in order to find most of these rearrangements and confirm that they exist."
Dana-Farber's Meyerson added that this study is a continuation of the direction that the field of genomics and sequencing is moving with regards to cancer research. "This is another piece of the progression towards sequencing lots of cancers and really being able to fully understand what's happening in the cancer genomes," he said.