NEW YORK — A team led by researchers at Sweden's SciLifeLab and KTH Royal Institute of Technology have demonstrated the feasibility of using highly multiplexed affinity assays to make proteomic measurements in dried blood spot (DBS) samples.
The work, detailed in a preprint published last month on MedRxiv, raises the possibility of more nimble and flexible approaches to population-scale proteomic studies, suggested Jochen Schwenk, professor of translational proteomics at KTH and senior author on the study.
In recent years, advances in plasma proteomics have made large-scale population studies possible. In a number of studies, researchers have used highly multiplexed affinity platforms from Olink (now part of Thermo Fisher Scientific) and SomaLogic (now part of Standard BioTools) to measure thousands of proteins in blood samples from thousands of individuals. More recently, researchers have begun exploring the potential of mass spec-based approaches — specifically workflows using technology from proteomics firm Seer — for similar work. In January, the UK Biobank said it has launched a new study in which it will use Olink's Explore HT platform to measure up to 5,400 proteins in 600,000 blood samples.
Such efforts have identified a number of gene-protein relationships linked to disease and other biological questions of interest. Schwenk noted, however, that projects like the UK Biobank require large amounts of resources to put together — far beyond what is typically available to an individual researcher or small group of researchers.
Such projects are also less amenable to longitudinal monitoring of participants, he suggested.
"It's a big ship that moves slowly," he said. "It carries a lot of information … but you will not be able to build a biobank on time-resolved samples."
Remote sampling tools like dried blood spots could allow researchers to more easily and cheaply put together study cohorts and take measurements more frequently, Schwenk said. "You can empower individuals to contribute and do it probably at a much higher frequency. You break apart a lot of the constraints of having to call people in and have this long and complicated pipeline."
That dried blood spots and other remote sampling techniques could offer such advantages is not a new idea. Within proteomics, researchers including Stanford University's Michael Snyder, SISCAPA Assay Technologies CEO Leigh Anderson, and Uppsala University professor Ulf Landegren (a cofounder of Olink) have all explored the use of dried blood spots or other remote sampling technologies for making protein measurements. Schwenk suggested, though, that depths of coverage well beyond what has been published to date are possible, noting that it is feasible to run in dried blood spot samples the full Olink Explore HT panel, which includes assays to roughly 5,000 protein targets.
Mathias Uhlén, professor of microbiology at KTH, is involved in a separate project using remote sampling devices from Swedish firm Capitainer, in which he is an investor and advisory board member, to conduct a longitudinal study in which researchers plan to sample individuals every three months. While the project has just started, he said his group has been able to get "very nice data" from these samples using the Olink platform.
Stanford's Snyder said that his team is likewise collecting data on the use of Olink's Explore HT in samples collected using microsampling devices from Torrance, California-based firm Neoteryx. While this work is not yet complete, Snyder said he believes it will be feasible to measure on the order of thousands of proteins in these samples using the Olink platform. He said he has also been in discussions with Standard BioTools about using the SomaScan platform for similar work.
In the study, Schwenk and his colleagues analyzed samples from 437 individuals who submitted blood samples in 2021 using Capitainer devices as part of a project studying the multimolecular profiles of a selection of the populations of Stockholm and Gothenburg. The researchers selected 1,000 random individuals in each city and mailed them home sampling kits along with health questionnaires and consent forms. After taking their samples, participants returned these items via regular mail.
The researchers performed serological assays measuring IgG antibodies against the SARS-Cov-2 nucleocapsid and spike proteins as well as assays measuring auto-reactive antibodies against 22 human interferons. Using the serological data they were able to classify subjects as either having been both vaccinated and infected naturally by SARS-CoV-2, only vaccinated, only infected naturally, or neither vaccinated nor infected naturally.
For their proteomic measurements, the SciLife-KTH team used the Olink Explore 384 Inflammation panel, which measures 368 proteins linked to inflammation. The researchers also sent samples to Alamar Biosciences, which ran them on its NULISAseq Inflammation Panel 250, which measures 250 proteins linked to immune response.
The authors noted that the proteomic measurements generally showed low levels of "well-known acute-phase proteins," indicating that "most samples were taken after the acute phase of the infection." They also grouped the proteomic data into five different clusters using Weighted Gene Correlation Network Analysis (WGCNA) and generated five protein-based phenotypes, or proteotypes, which they were able to correlate to the serological profiles.
The study also provides one of the first comparisons of data from the Olink and Alamar platforms. While Alamar does not currently compete in the higher plex plasma proteomics market (multiple thousands of proteins per experiment), it does compete with Olink's lower plex offerings (hundreds of proteins per experiment). Both companies use versions of proximity ligation assay (PLA) technology, and Alamar's NULISA (NUcleic acid-Linked Immuno-Sandwich Assay) workflow features additional target pull-down and wash steps that the company says boost the assay's sensitivity by improving its signal-to-noise ratio.
In 2023, Olink sued Alamar for patent infringement. The lawsuit is ongoing, and this week the US District Court for the District of Delaware partially granted Alamar's motion to dismiss Olink's complaint. Both companies have 30 days from Feb. 11 to file an amended complaint if they wish to pursue the case further.
The Olink and Alamar panels have 110 targets in common. Of those 110, measurements for 60 correlated across the two platforms with a rho of 0.5 or greater, with higher correlation for proteins where measurements on both platforms were above the limit of detection (LOD) in 50 percent or more samples. Consistent with Alamar's claims of higher sensitivity, the NULISA panel returned measurements above the LOD for more of the 110 common targets than did the Olink panel.
"There are some markers where Alamar outperformed Olink and vice-versa, so then it is a matter of, if these markers are of a higher priority to you, then probably one platform would be preferred to the other," Schwenk said.
He added that he had been particularly pleased with the Alamar data given the company's limited experience using its platform in dried blood spots and noted that this indicates the viability of the approach.