Following Roche's disclosure last week that it will shut down 454 Life Sciences and stop supporting 454 sequencing instruments by 2016, customers are making plans to move their sequencing over to other platforms, if they have not done so already.
While Illumina, Life Tech's Ion Torrent, and Pacific Biosciences are eager to step in to fill the void, some customers say aspects of 454's technology cannot be replaced by other platforms at this point. Also, those customers who have started to use 454 for routine clinical applications need to revalidate their assays on a new platform.
Roche said last week that it will close down 454, which is based in Branford, Conn., and lay off about 100 employees over the next three years (GWDN 10/15/2013). By mid-2016, it will stop supporting the 454 sequencing platforms, the GS FLX+ and the GS Junior.
Roche's decision to pull the plug on 454 came to light less than a month after the company announced a deal with Pacific Biosciences, worth up to $75 million, to develop a sequencing system and assays for clinical diagnostics using PacBio's single-molecule real-time sequencing technology (IS 10/1/2013).
Earlier this year, Roche had already eliminated 60 positions at 454 as part of a reorganization that combined the 454 business in Branford and the NimbleGen business in Madison, Wis., into a new sequencing unit (IS 4/23/2013). That move was part of a larger restructuring that dissolved Roche Applied Science, the life science unit of Roche Diagnostics, and integrated its products into other units. At the same time, Roche cut its sequencing technology development collaborations with DNA Electronics and with IBM.
Dan Zabrowski, head of Roche's sequencing unit, told In Sequence last week that all 454 products, such as instruments, parts, reagents, consumables, and services for those products, will continue to be available to customers until mid-2016. These include the GS FLX and GS Junior Titanium reagent series, the XLR70 and the XL+ sequencing kits, and subkits for the GS FLX and FLX+ instruments. Also included will be existing and soon-to-be-launched GS Junior/+ sequencing kits and subkits.
Roche still plans to launch a long read upgrade, new oncology tests, and an automated front-end emPCR workflow for the GS Junior.
In addition to the partnership with PacBio, Roche's sequencing unit "will continue to seek both internal and external opportunities to ensure we provide our customers with next-generation sequencing products and services that could differentiate us from our competitors," he said.
Roche's sequencing unit will relocate to Pleasanton, Calif., in order to be "closer to our clinical diagnostics experts in Molecular Diagnostics and [at] Ventana," Zabrowski said. He did not disclose how many employees the unit will retain but said that its staff "will grow over time as we increase the number of projects in the pipeline," although Roche will "look for synergies with our clinical diagnostics experts to avoid duplication as much as possible."
Roche's NimbleGen target enrichment business will remain in Madison, and Roche "will provide and develop our existing and new NimbleGen sequence capture and target enrichment products for our sequencing customers to support their research activities with essential next-generation sequencing workflow components," he said.
While 454 has lost much of the next-gen sequencing market to lower-cost, higher-throughput platforms from Illumina and Life Tech's Ion Torrent over the last few years, as well as, more recently, to Pacific Biosciences, it still has a number of active customers who value its long, high-quality reads.
"It's a bummer that they did not push it more over the last few years" by developing a higher-throughput system and lowering reagent prices, said Stephan Schuster, a professor at Penn State University and Nanyang Technological University in Singapore. "It clearly has interesting features because its sequencing technology and chemistry are so different from other platforms."
Three years ago, his team published the genomes of two African individuals, one of them a de novo assembly from 454 GS FLX data only (IS 2/23/2010). That data included contigs and scaffolds that did not map against the human reference genome and were undetectable by other platforms. "There are parts of the genome that sequence only with 454," Schuster told IS last week. "It is not clear why these parts are missing from the other sequencing technologies."
Schuster, who has been using the 454 technology for a number of high-profile projects, including the genome of the woolly mammoth (GWDN 11/19/2008), currently still has two GS FLX and two GS FLX+ instruments in his laboratory, using them mostly for plant genomes and "all projects that need high-quality assemblies." He said the platform is "still the best for ancient DNA and covering the complete genome," adding that "we get draft assemblies very fast [by] adding 454 data."
But his lab has already started replacing 454 in some areas, such as metagenomic sequencing, with "stitched" 300 x 300 base pair Illumina MiSeq reads (IS 3/12/2013).
Bruce Roe, a professor emeritus at the University of Oklahoma and an early 454 customer, said his group is also still using the 454 technology, mainly for shotgun sequencing of BACs, which he has found to be "an extremely efficient way of obtaining closure of difficult regions," for example in the tomato genome. In addition, the researchers are using 454 for multiplexed EST/RNA-seq of expressed genes from a variety of organisms and tissues.
To some extent, his group has replaced 454 with PacBio's technology, obtaining PacBio data from outside groups to close gaps between contigs in genome sequences, such as the tomato genome, and to order and orient contigs.
"It is ironic that, now that we are getting 454 read lengths approaching those obtained using Sanger sequencing on the ABI 3730xl, Roche wants to phase the 454 out over the next few years," Roe told IS. "On the other hand, it is very exciting to see the new technologies that are on the horizon and/or at various stages of commercialization that can produce better data, quicker, and at a reduced cost."
"As it turned out, other instruments likely will readily fill any gap left by Roche's dropping of support for this instrument in a few years," he added.
Some 454 customers who made headlines during the heydays of the technology have already moved on. Svante Pääbo's group at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, published an analysis of the Neandertal's mitochondrial genome, based on 454 data, in 2008 (IS 8/12/2008) and a draft of the Neandertal genome, based on both 454 and Illumina data, two years later (IS 5/11/2010).
According to Barbara Höber, a researcher at the institute, the group has since replaced the 454 technology entirely with Illumina HiSeqs and MiSeqs, and has not used its two 454 sequencers for almost four years. "There is nothing that we're really missing," she told IS in an email message.
A number of groups who have implemented 454 sequencing in routine clinical use will need to look for alternatives now. Earlier this year, for example, the Red Cross Blood Transfusion Service of Upper Austria won accreditation for an HLA typing method on the GS Junior and has been testing clinical samples on the platform (CSN 6/3/2013).
According to Christian Gabriel, the service's medical director, the news that Roche will stop supporting 454 in 2016 "was a full blow."
"I am fed up," he told IS by email, and fears receiving faulty reagents and less support "sooner than later." He said there might not be enough time to switch protocols or to test other systems, though many customers will probably switch to Illumina, and his group has already brought in a MiSeq. "I am confident that [Roche] management will try to find a solution for those who run diagnostics on their systems," he added.
Gabriel said his group is currently using 454 routinely for high-resolution HLA typing in immunogenetics, as well as for cancer gene panel sequencing, hematology applications, mutation analysis in CML, and metabolomics in ecosystems.
He said he especially values 454's long reads, good bioinformatics and large amount of software developed for the system, and Roche's "very good service organization."
Jonathan Rothberg, who was in charge of developing and commercializing the 454 technology when it was part of Curagen, said that the technology's demise means one needs to "innovate or die."
"At 454 we were always first; first non-bacterial cloning, first commercialization, first next-gen individual human genome, Neanderthal, mammoth, deep sequencing, cancer sequencing, drug response studies, HIV, metagenomics, first drug target by whole genome sequencing, and many more firsts," he told IS via email. "Always innovating, always first."
That apparently changed when Roche acquired 454 in 2007, for $155 million in cash and stock. "When I woke up and found Roche had bought 454 without me, I had to restart," Rothberg said. "It cost three years. We had to invent a new scalable way to sequence — ion semiconductor sequencing — and establish a clear path towards both truly low-cost and mobile sequencing." He went on to found Ion Torrent, which was bought by Life Technologies in 2010 for $375 million in cash and stock, and another $350 million based on milestones.
Rothberg said that when Roche bought 454, the company was "two years ahead of everyone else," but after the purchase, "they lost that lead, no more firsts, no more innovation."
He said that two concepts 454 introduced — sample preparation by limited dilution and massively parallel sequencing — "remain the foundation for all next-generation sequencing, and created an industry."