By Julia Karow
As high-throughput sequencing platforms are becoming routine tools in research labs, clinical laboratories are also exploring their use to help with the diagnosis of conditions such as genetic diseases and cancer.
This spring, for example, Münchner Leukämie Labor, a Munich-based reference laboratory that specializes in the diagnosis of hematological malignancies, started testing Roche's 454 GS FLX system. Earlier this month at the American Society for Hematology annual meeting in New Orleans, company researchers presented initial results from targeted resequencing studies of leukemia samples, one using the 454 platform in conjunction with NimbleGen capture arrays, and another that sequenced PCR amplicons.
While sequencing array-captured products enabled the researchers to detect a variety of cancer mutations in a single experiment, MLL found that the array workflow is currently too time-consuming, cumbersome, and costly for routine diagnostic applications, so it is looking into alternative enrichment methods.
Amplicon sequencing, on the other hand, might have more immediate applications in the lab, according to an MLL researcher.
In 2009, MLL performed diagnostic assays on about 32,000 patients, including cytomorphology, immunophenotyping, cytogenetics, fluorescent in situ hybridization, and PCR-based and sequencing-based tests. The company, which is in its fifth year and has almost 80 employees, also maintains an active program in genomics, including microarray-based gene expression and copy number analysis.
This spring, piqued by the potential for increased sensitivity and throughput and cost savings, compared to Sanger sequencing, MLL started to look into next-generation sequencing applications, according to Alexander Kohlmann, head of the firm's next-gen sequencing group.
He said that hematology labs have traditionally been early adopters of new molecular diagnostics, and several labs he is familiar with "can't wait to get their hands on next-gen [sequencing]," which he said will be a "hot topic for 2010."
Earlier this year, the MLL researchers considered the 454 GS FLX, the Illumina Genome Analyzer, and the Applied Biosystems SOLiD — the Helicos platform was not yet available, according to Kohlmann — and decided to go with the 454 because of the system's shorter run time and longer reads, which would be important to assay gene fusions with unknown partners. At present, the firm has two GS FLX platforms in operation.
MLL's data showed that the 454 system detects mutations present at 1 percent frequency, "so we are quite confident that 454 technology really delivers a very sensitive detection of mutations," Kohlmann said, adding that it also outperforms Sanger sequencing in terms of throughput.
He said the company has not yet performed a cost calculation but expects that "if we can move forward with the 454 into routine diagnostics, then we are pretty sure that the cost [compared to Sanger] would also come down." At the conference, he and his colleagues showed results from a project where they sequenced amplicons of oncogenic regions in 95 leukemia samples and myeloproliferative neoplasms.
Regarding the enrichment of DNA for targeted resequencing, MLL has so far focused on NimbleGen sequence capture arrays, for which it has all necessary equipment in house. In one study, company researchers used 385K arrays to enrich a 1.91-megabase region containing 95 cancer-related genes in six acute myeloid leukemia samples and sequenced them by 454 using the Titanium long-read chemistry.
In proof-of-principle studies, this combination allowed them to characterize several types of mutations in one swoop. "This is the only assay that, in addition to detecting point mutations and deletions or insertions would deliver, in one workflow, also the confirmation of translocation or inversion events," Kohlmann said. Currently, chromosomal translocations are detected by separate methods such as cytogenetics, which requires highly trained experts, or, in some cases, by FISH, he said.
[ pagebreak ]
In particular, using arrays coupled with 454 sequencing, the scientists were able to capture one partner of a fusion gene and sequence into the second, unknown partner. "That was, for us, the basic breakthrough, that we can enrich for one partner and then … we can also discover novel fusion partner genes," he said.
But for the moment, the assay will stay in the research realm. "It's way too complicated for routine procedures because of the many, many steps," Kohlmann said, as well as the time required to obtain a result. "Even if you are fast, [it takes] at least two weeks in the lab to generate data on a couple of patients," he said.
The assay is also still costly, compared to cytogenetics analysis, but Kohlmann said he expects prices will come down.
In order to speed up the assay, he and his colleagues are looking into NimbleGen's solution-based capture method to replace the arrays. Regarding alternatives to targeted enrichment, the company is also about to embark on a project that uses RainDance Technologies' microdroplet-based multiplexed PCR technology. In addition, it just started installing robots to automate a number of manual pipetting steps related to the 454 system.
Near term, Kohlmann said, he sees the most immediate diagnostic application of 454's technology at MLL in amplicon sequencing assays. The 400-base reads "would give us a nice coverage," he said, and enable them to detect both "the usual suspects" as well as novel mutations in genes. Amplicon sequencing might be applicable, for example, to sequence the p53 gene, which is frequently mutated in cancer and would be "an ideal candidate to switch over to a next-gen deep sequencing platform," Kohlmann said.
MLL is currently investigating such assays, he said, "and very early next year, we will get our hands on the first Titanium chemistry to look into the utility of amplicon sequencing on the 454."
Regarding the NimbleGen/454 assay to characterize several mutation types, he said that "it might find its niche" but "whether it's broadly applied, time will tell."
"If you wait too long, the genome will be sequenced anyway in its entirety by some other method, but for the moment, there is a chance that this type of assay will find its niche," he added.