NEW YORK – Data released at the recent American Society for Mass Spectrometry annual conference indicates that new instrumentation and new workflows have dramatically increased the depth and throughput of mass spec-based plasma proteomics experiments.
These new capabilities could make mass spec a more useful tool for a variety of applications including plasma protein biomarker development as well as the large-scale population studies that have to date been dominated by affinity-based platforms from Olink and SomaLogic.
Plasma is an easily accessible and commonly collected sample source, especially for clinical work and population studies. However, it has traditionally been a challenging sample source for mass spec-based proteomics due to factors like its large amounts of high-abundance proteins and its large dynamic range. While mass spec workflows in cell lysate commonly measure 8,000 to 12,000 proteins, similar workflow in plasma may only measure 500 to 1,000 proteins. This can be improved by using depletion to eliminate high-abundance proteins or extensive fractionation, but this comes at the expense of throughput.
Last year, Swiss proteomics firm Biognosys published a study in the Journal of Proteome Research in which they used a two-hour LC gradient on a Thermo Fisher Scientific Orbitrap Exploris 480 to measure 2,732 proteins across 180 depleted plasma samples, marking what was at the time the deepest analysis of the plasma proteome done without fractionation.
Recently, proteomics firm Seer brought a new plasma proteomics solution to market. The company's Proteograph system uses sets of nanoparticles to enrich plasma proteins, which they can then identify and quantify using technologies like mass spec. Compared to traditional plasma proteomic approaches, the Seer system improves both depth of coverage and throughput. In a study detailed in an April BioRxiv preprint, a team led by researchers at Weill Cornell Medicine-Qatar analyzed 345 plasma samples, measuring a total of roughly 3,000 proteins across those samples and with LC-MS run times that allowed for analysis of around 10 samples per day.
At those depths of coverage, both the Biognosys analyses and the Seer system coupled to mass spectrometry approached that of Olink's Explore platform, which measures around 3,000 proteins in plasma, though they remained well behind SomaLogic's SomaScan platform, which measures around 7,000 proteins in plasma. At around 70 samples per week, the Biognosys and Seer systems' throughput still lagged that of the Olink and SomaLogic platforms, which, respectively, can run up to 1,000 and 340 samples per week.
At ASMS, Thermo Fisher Scientific presented data using Seer's recently released Proteograph XT kits on its new Orbitrap Astral instrument to measure roughly 6,000 proteins at a rate of around 30 plasma samples per day. Those figures mark a substantial advance for plasma proteomic workflows and indicate that mass spec combined with plasma enrichment technologies like Seer's Proteograph may now be competitive with affinity-based platforms for large-scale plasma studies.
"We hadn't seen this massive leap coming, to be honest," said Maik Pietzner, a bioinformatician at the MRC Epidemiology Unit at the University of Cambridge School of Clinical Medicine who has used SomaLogic's SomaScan and Olink's Explore in large-scale proteogenomic studies. "We obviously more than welcome this development."
Pietzner said that for the proteogenomic experiments he and his colleagues are pursuing, they need sample cohorts of 1,000 or more, which, he noted, based on the data presented at ASMS, "seems now to become feasible."
Michael MacCoss, professor of genome sciences at the University of Washington, also suggested that this level of coverage and throughput positioned mass spec as a useful tool for large-scale population studies.
"These large cohorts from like the UK Biobank or the Framingham Heart Study … the value of those samples is huge, and [researchers] want to be able to get the most out of those [samples] with the least amount of resources, and a lot of those experiments have used either Olink or SomaLogic," he said.
If mass spec is able to reliably deliver the depth of coverage and throughput demonstrated in the ASMS presentations, it could prove a powerful complement and, perhaps, competitor, to the affinity-based platforms. Many proteins exist in multiple forms, or proteoforms, exhibiting alterations like amino acid variants or truncations or post-translational modifications, which can impact their function, and it is often unclear or unknown with affinity platforms which particular form of a protein they are measuring. Mass spec-based approaches are better equipped to tease out these different proteoforms.
Olink President Carl Raimond said that he saw mass spec and affinity-based approaches as "absolutely complementary," adding that it "is great to see innovation in the protein analysis space."
He said, however, that he does not view it as a near-term competitor for the types of large-scale population studies where Olink has carved out a leading position and questioned whether the impressive numbers presented at ASMS would hold up when the technology is put into wider use.
"The devil is in the details," he said. "It's easy to make a claim. It's a very different thing to be able to deliver or to ask the questions about what is underneath that claim."
Raimond added that while mass spec technologies continue to improve, affinity-based platforms will also advance. Olink is in the process of expanding its Explore platform to around 5,000 protein targets, while SomaLogic aims to expand the SomaScan platform to cover around 10,000 proteins by the end of the year.
Pietzner likewise said that while the numbers released at ASMS were exciting, he and his colleagues look forward to seeing more extensive data, not only on the overall coverage of proteins but also on how consistently and reproducibly different proteins and peptides are detected across samples.
"Affinity-based methods have now been measured in cohorts of 50,000-plus, with stunning insights emerging," he said. "We need head-to-head comparisons to evaluate whether these new mass spec technologies can be scaled to a similar extent."
MacCoss said that companies or researchers aiming to conduct this kind of research using mass spec will need to provide data showing that they are able to consistently and reproducibly measure a core set of proteins in every sample.
"When someone goes and orders Olink, they have a list of proteins they know will be measured every time," he said. "We still need to do that. We still need to say, 'this is the list of proteins that will be returned as a quantitative value for every experiment … and where we have known analytical figures of merit for those measurements.'"
Pietzner said that he and his colleagues are currently working to expand their proteogenomic research to include mass spec. Christopher Whelan, director of neuroscience data science at the Janssen Pharmaceutical Companies of Johnson & Johnson and chair of the UK Biobank Pharma Proteomics Project (PPP), has also said that project, which is among the largest ongoing proteogenomic population studies, is in the process of implementing mass spec-based proteomics.
Seer this month announced the launch of its Seer Technology Access Center, which will employ its XT Assay Kit in combination with the Orbitrap Astral mass spectrometer to provide proteomic services to customers without access to mass spectrometry.
While it may be difficult to thoroughly assess the performance of Thermo Fisher's Orbitrap Astral and Seer's Proteograph XT based on the data released thus far, some early-access users say they are impressed with the results they have generated.
Jennifer Van Eyk, director of the Cedars-Sinai Precision Biomarker Laboratories, has been using the Orbitrap Astral for plasma proteomic analyses, where it also appears to provide a substantial boost in power over preexisting instrumentation. Van Eyk said that in experiments running 60 samples per day in undepleted plasma, the new instrument is able to measure between 2 and 2.5 times as many proteins as measured using the same workflow on Thermo Fisher's Exploris 480 instrument.
"We're pulling up not just more proteins but more quantifiable proteins, and proteins that are repeatable, meaning, if we run [a sample] five times, do we see it all five times," she said. "It is a big jump."
Perhaps most impressive is the throughput the instrument enables, Van Eyk said, noting that she and her colleagues have gotten good data and deep coverage running up to 180 undepleted plasma samples per day.
"At 180 samples [per day], all of a sudden you can start talking about running 10,000 samples, and then it becomes a population study," she said.
Van Eyk and her colleagues are currently trialing a Seer Proteograph system to "put it through its paces" and assess whether they want to employ it as part of their plasma proteomic workflows, though they did not use it to generate the data cited above.
Joshua Coon, professor of biomolecular chemistry and chemistry at the University of Wisconsin-Madison and an early-access user of the Orbitrap Astral, noted that his lab was able to measure roughly 1,500 proteins in undepleted, unenriched plasma using a 50-minute LC gradient and had developed a one-minute direct infusion method on the instrument capable of measuring around 200 proteins per sample.
Coon is also a user of Seer's Proteograph platform, though he has not yet used it in combination with the Orbitrap Astral. His lab has been using the Seer XT kits to analyze plasma samples from Alzheimer's patients as well as from individuals with long COVID.
He said that while his team has not yet started running the bulk of these samples, in its preliminary work the lab is consistently measuring around 3,000 proteins per sample, which he said is about fivefold what it gets without using the Seer system. He suggested that those numbers will go up when the researchers put the workflow on the Orbitrap Astral system.
Beyond depth of coverage, Coon said he had found the Proteograph useful for simplifying mass spec sample prep.
"I didn't fully appreciate the level of automation," he said. "It has been really handy. Right now, the primary users are a first- and second-year graduate student … and so they had to learn pretty quickly. And they have been having really good success processing samples, getting digests, and getting peptides. When you have new people or people who haven't been doing it for a long time, the sample prep part of doing a big proteomics study is half the battle or more, just dialing that in and being good at it."
While the depth of coverage enabled by Seer's Proteograph system is making mass spec-based plasma proteomics more competitive for certain applications with affinity-based platforms from Olink and SomaLogic, Seer itself is facing new competition in the plasma enrichment space.
At ASMS, proteomics sample prep firm PreOmics launched its ENRICH-ist Kits for enrichment of plasma and serum proteins. The kits use non-functionalized paramagnetic microbeads to enrich low abundance proteins and, according to the company, boost protein detection by 50 percent to 100 percent compared to undepleted, unenriched plasma.
PreOmics CEO Garwin Pichler said the microbeads in combination with the buffers work to enrich low-abundance proteins while depleting high-abundance proteins, boosting depth of coverage.
Also at ASMS, Biognosys launched a new bead-based plasma proteome enrichment assay as part of its TrueDiscovery service platform. According to the company, the assay enables the high-throughput quantification of roughly 4,000 proteins in human plasma.
Additionally, this month, a team led by researchers at the University of Washington published a BioRxiv preprint describing a plasma protein enrichment approach using magnetic microparticles from ReSyn Biosciences to improve depth of coverage by binding membrane-bound vesicles in plasma and analyzing the associated proteins.
UW's MacCoss, senior author on the preprint, said that he and his UW colleague Christine Wu, first author on the preprint and lead developer of the enrichment approach, are able to consistently quantify around 4,800 proteins in plasma using a 30-minute LC gradient on an Orbitrap Astral, allowing for throughput of around 40 samples per day. They are able to measure between 5,000 and 6,000 proteins using a one-hour LC gradient.
MacCoss said he believed that those were conservative figures. "We haven't been pushing it at all," he said.
MacCoss said that researchers are interested in alternatives to Seer for plasma proteomic enrichment in large part due to the high cost of the company's technology. "Seer has done a fantastic job of making those products, but the cost is a high barrier," he said.
Karl Mechtler, proteomics head at Vienna's Research Institute of Molecular Pathology, said that in his discussions with Seer, the quoted cost per sample was around $600.
"If I have 100 samples, that is a lot of money for a proteomics lab," he said, noting that for a typical proteomics lab, an affordable price point would be around $25 to $50 per sample.
Wu said that experiments using the UW enrichment approach cost less than $5 per sample.
PreOmics sells its ENRICH-ist Kits as part of a complete proteomics sample prep workflow that costs a total of $80 per sample.
Addressing the question of cost, Seer Chair and CEO Omid Farokhzad said he considers price "a matter of value exchange."
"Not all content is equal value," he said. "The question is, what is the value exchange in terms of what Seer offers and the alternative to it?"
Amidst the recent developments in plasma proteomics, the answer to that question would seem to be a moving target.