NEW YORK – As it once was for genomics, next-generation sequencing appears to be the future of proteomics, judging by a recent deal between protein biomarker firm SomaLogic and NGS behemoth Illumina.
While technical details are still to be figured out, the partnership solidifies sequencing as the readout of choice for label-based proteomics.
"Illumina's killing it. There's a huge rush to partner with them as new technology is being developed in the proteomics space," said Philip Lorenzi, director of the metabolomics core facility at MD Anderson Cancer Center, who uses SomaLogic's aptamer technology with microarray readout, called SomaScan, as well as existing NGS-based proteomics methods, such as the offering from Olink.
Illumina's assay, due out in 2024, will incorporate SomaLogic's 7,000 target aptamer library but will "move quickly to a 10,000-target panel," said Alex Aravanis, chief technology officer of Illumina. The assay will be designed for the firm's highest-throughput instruments: for now, NovaSeq, and whatever might succeed it.
Illumina said it expects to enable researchers to analyze tens to hundreds of samples per flow cell, implying costs as low as $100 per sample or less, depending on several factors. "We expect to deliver an application similar to barcode counting on NovaSeq flow cells with a price range per flow cell of $10,000 to $13,000, depending on the cycle counts that will be required," according to an Illumina spokesperson. Unbiased, label-free mass spectrometry-based methods, such as tandem mass tagging, would also be able to analyze that many targets. Lorenzi estimated costs between $500 and $1,000 per sample using mass spec, taking from two to four hours of instrument time — processed in series rather than in parallel.
Illumina also plans to create combined workflows that provide genomic and proteomic information in the same analysis. "A lot of existing NGS users are interested in adding proteomics to their science or in clinical applications," Aravanis said. "We had unsolicited interest for this product from customers, without them knowing this was coming."
"I was pleased to see they're combining forces," said Kari Stefansson, CEO and founder of Iceland's DeCode Genetics, a subsidiary of Amgen. "Illumina has a reputation of delivering very good instruments. I hope that by combining these two, it will lead to a very, very important advance."
"I'm absolutely convinced that use of technologies like this will lead to the next big avalanche of discoveries in the pathogenesis of disease," he said.
While certainly a boon for Illumina, the deal was "necessary to remain competitive" for SomaLogic, Lorenzi said. "We were counting SomaLogic out, honestly, because they weren’t doing NGS."
Boulder-based SomaLogic now joins a handful of proteomics assay developers who have adopted NGS-based readout. Olink of Sweden, for example, launched its oligo-tagged antibody-based proximity extension assays with sequencing readout in 2020. In addition, Encodia, a startup led by former Illumina executives, including Illumina cofounder Mark Chee, is building technology to convert peptide sequence information into DNA that can be analyzed by NGS. In January 2021, Encodia announced $75 million in Series C funding, led by Northpond Ventures and Deerfield Management.
But SomaLogic is the only one of those companies offering assays based on aptamers, single-stranded oligonucleotides that can bind to protein targets with great specificity. "What's so cool is the physical chemical space [aptamers] can cover," Lorenzi said. "I'm in the group that think there are up to 6 million proteoforms. That's enormously more complex than the genome. Aptamers provide a viable way to achieve broad coverage of that deep space." While existing SomaLogic assays using microarray readout have greater sensitivity than mass spec, NGS can make them even better. "NGS could yield 1,000, 10,000, even 100,000 times better sensitivity," he said. "Maybe even a millionfold better," particularly if the analyte has very poor ionization efficiency by mass spec.
Using sequencing to identify and count aptamers helps solve a challenge specific to proteomics: the fact that cells can have many copies of a protein, at levels that can be orders of magnitude different. "You want something that allows you to see all these different concentration ranges," Aravanis said. "One really nice thing about SomaScan is the ability to quantitate it over a large dynamic range."
Whether Illumina's assays will read out the aptamer sequence itself or an incorporated barcode sequence is yet to be determined, Aravanis said. While he doesn't know which company reached out to the other first, he said Illumina and SomaLogic have been in discussions for about a year.
The agreement provides upfront royalty payments and other revenue considerations for SomaLogic, CEO Roy Smythe said at last week's virtual JP Morgan Healthcare Conference. Aravanis declined to address pricing but said he would "like it to price it so that it can be done routinely and so it’s accessible to researchers and ultimately clinical labs." He noted that Illumina will commit R&D resources to the assay development, as it has with other products, such as the TruSight Oncology 500 assay or TruSeq library preparation kits.
"It's that level of investment: to develop assays, the chemistry, manufacturing, analysis, packaging, quality control, sales, and marketing. It's a full-blown prod development comparable to what we do in other areas." Illumina will package, market, and sell the assays as its own products, though "it will be clear that it includes SomaScan technology," he said. It will also develop bioinformatics tools for the assay, as it has for its sequencing products.
Among sequencing technology firms, Illumina has a huge head start in proteomics, but others are getting ready to enter the field. Last year, multiple research groups showed they could use nanopores to detect or sequence proteins, and in August, researchers at the University of Washington reported use of protein reporter tags that can be read on the MinIon system.
Both Stefansson and Lorenzi said they'd be interested to try out the Illumina assay once it is available, and both are current customers of SomaLogic as well as Olink.
"There's no question about it, both [SomaLogic and Olink] are very powerful, but they’re not flawless," Stefansson said. Working with SomaLogic, "we have to make sure aptamers are always targeting the proteins we think they're targeting," he said. "There is room to improve them, but in their current form they are spectacularly powerful." He added that his company is studying the differences between SomaScan and Olink's PEA and expects to publish results soon. Olink's assays detect "substantially fewer proteins than SomaLogic," he added. "They have their work cut out for them to increase numbers of proteins."
For Lorenzi, the main disadvantage to using the proposed Illumina assay is that it's a biased approach. "For example, if the assay menu of 10,000 proteins doesn't include any proteins known to be associated with Alzheimer’s disease, then the platform could be considered to be useless to Alzheimer’s researchers," he said. "Given that the human proteome is estimated to be at least 300,000 proteins and up to 6 million proteoforms, it seems likely that many important proteins will still be missing from the assay menu."
And while Illumina can count single copies of genes and RNA transcripts, "the latest information in the field is that they cannot yet count single copies of proteins," he said. "Several competing technologies that will be unveiled this year will claim to offer single-molecule or nearly single-molecule sensitivity and the potential to measure thousands of proteins per biological sample." These technologies could include Quantum-Si's single molecule analyzer and Encodia's ProteoCode platform, he added.
Illumina's initial focus will be on getting a research product out, but Illumina and SomaLogic are also setting up a path to the clinic. Aravanis said Illumina expects to develop either laboratory-developed tests or distributed in vitro diagnostics, or both, noting that the company would be responsible for any regulatory submissions.
"This becomes really interesting once you begin to use the level of proteins [in a sample] to develop algorithms that predict risk of disease," Stefansson said. "You're really not measuring risk anymore, you're documenting early steps in the pathogenesis of disease."