NEW YORK — Spatial proteomics grabbed a share of the spotlight at the American Association for Cancer Research annual meeting last week, as a number of firms launched or highlighted new platforms at the conference.
The new systems provide higher levels of multiplexing and throughput and point toward increasing use of spatial proteomics in clinical research and, ultimately, clinical trials.
Among the most notable releases was a new MALDI mass spectrometry imaging-based platform developed by Bruker and AmberGen that expands Bruker's ongoing push into spatial biology. Billerica, Massachusetts-based Bruker has made a strategic investment of an undisclosed amount as part of the two firms' work together.
Bruker subsidiary Canopy recently launched a next-generation version of its ChipCytometry technology, a fluorescence microscopy-based spatial proteomic system that uses rounds of antibody staining to measure dozens of proteins in tissue samples. By contrast, the approach developed with Watertown, Massachusetts-based AmberGen that the company detailed at AACR uses MALDI mass spectrometry to measure peptide-linked antibodies in tissue samples in a manner that is more analogous to the mass spec-based approaches employed by firms like Standard BioTools and Ionpath.
The approach combines AmberGen's Miralys reagents — antibodies linked to photocleavable mass tags — with Bruker's MALDI mass spec platforms, specifically its Rapiflex and timsTOF systems. Researchers stain a tissue sample of interest with the Miralys antibodies, then expose the tissue to UV light to cleave the mass tags, which can be detected via MALDI mass spec.
Unlike spatial proteomics platforms from companies including Akoya Biosciences, Standard BioTools, and Ionpath, the Bruker-AmberGen approach does not provide subcellular resolution, but it does enable high levels of multiplexing and high throughput. According to Kenneth Rothschild, cofounder and chairman of the board at AmberGen, researchers can multiplex assays to more than 100 proteins using the system and can scan a cm2 tissue sample in roughly the same amount of time it takes mass cytometry-based technologies offered by Standard BioTools and Ionpath to scan a mm2 sample, albeit at lower spatial resolution.
He suggested that the Bruker-AmberGen technology could prove a higher-throughput complement to these mass cytometry-based techniques.
"They are fantastic techniques, being used by many researchers, and they are exactly complementary to what we are offering," he said. "Because at this point we are not able to offer subcellular [resolution], but for them to be able to get an overview of a [full] slide, is very difficult."
As the leading provider of MALDI mass spec instrumentation for life science research and clinical work, Bruker is a major player in MALDI imaging, but much of that work has focused on small molecules like metabolites and glycans. A few academic researchers, like Richard Caprioli at Vanderbilt University, have developed methods for MALDI protein imaging, but it has remained a technically demanding and specialized approach.
The AmberGen collaboration and Miralys reagents allow Bruker to deploy its existing MALDI mass spec installed base as part of its move into spatial biology.
"Getting into protein imaging is something that the MALDI imaging community has wanted to do for a long time," said Michael Easterling, director of imaging for Bruker's life science mass spectrometry division, noting, like Rothschild, that the company believes the system's high multiplexing capacity and ability to analyze large tissue samples quickly will let it carve out a niche in the spatial proteomics space.
"The total addressable market [for spatial biology] is huge, and the total amount of saturation is low," he said. "So there's a lot of room for additional players in this space."
Easterling said that Bruker also expects the ability to measure proteins and small molecules in the same sample will prove attractive to customers.
"In one single tissue section you get the proteins and you get the metabolites, the glycans, the small molecule xenobiotics, the lipids," he said. "You don't have to use two machines, and you don't have that fuzzy relationship with serial sections — it all comes off the same tissue section."
He said the company anticipates strong demand from the pharma industry.
"For years, pharma has been addressing the druggable proteome, but the question is, if you are drugging the proteome, how do you know you are actually drugging the proteome?" he said. "Now you can see the [protein] targets and you can see their metabolic output all at the same time, and you can get a much better idea about the physiological interaction between your drugs and your targets."
Easterling said he expects immuno-oncology research — where spatial proteomics has seen significant uptake — to be another major source of interest.
Also at AACR, Marlborough, Massachusetts-based Akoya Biosciences presented data from its PhenoCycler-Fusion system in which researchers used the platform to analyze a panel of 100 proteins at single-cell resolution across whole-slide samples.
Previously, analyses from Akoya and other spatial proteomics firms have typically topped out at around the 40-protein range. The PhenoCycler-Fusion data along with the projected multiplexing capabilities of the Bruker-AmberGen system suggest, though, that the field is in the midst of a shift to higher plexes.
"I think things will increasingly trend towards the higher plex level," said Akoya CEO Brian McKelligon. He said that with the larger plexes, researchers are able to "really cover a wide berth of markers not just relevant to cancer but to immune response. So part of our strategy is to lay out these large, complete panels so that people can effectively have a one-stop shop for large, all-encompassing studies."
McKelligon noted that in the case of Akoya's technology, the move to higher plexes had been enabled by the increased speed of the PhenoCycler-Fusion compared to previous platforms as well as the development of data compression tools that makes such experiments more practical.
He added that while 100-plex experiments were theoretically possible on earlier versions of the company's technology, "there were sort of some self-governing [limitations] in terms of runtime and data file sizes."
McKelligon said he expects the field to hover for a time around the 100-plex threshold as researchers explore what they can learn from that level of multiplexing.
"I think it's probably a practical place to take a pause and see where customer demand goes," he said. "100-plex across an entire slide that contains 2 million to 4 million cells across 30 or 40 samples in a week or two is a very rich dataset. So I think we need to continue to invest in simplifying that workflow, making the panels ready to go, and improving the informatics, and then we'll decide after that whether going beyond 100 is practical."
Andrew Quong, CSO at Standard BioTools (formerly Fluidigm), likewise said his firm is seeing demand for increased multiplexing.
He said that while two years ago researchers were still getting used to the idea of spatial proteomics experiments looking at 30 to 40 proteins, today the company's customers are routinely bumping up against the top of this range.
"We are constantly asked when we are designing panels for our customers to kind of push the boundaries in terms of the number of markers that we can include," Quong said. "And so now that we are seeing more traction in this high complexity spatial biology arena, there is going to be a need and desire for additional markers."
For instance, he said, "if you are looking at three different types of cells within a tumor, and you're interested in the interplay on the functional level between, say, the stroma, the fibroblasts, the cancer, and a couple types of immune cells, you can very easily see needing significantly more markers than the 20 or 30 people have traditionally been looking at."
At AACR, South San Francisco, California-based Standard BioTools launched the latest version of its imaging CyTOF system, the Hyperion+, which the company said offers a 1.6-fold lower limit of detection than the previous system along with the ability to process 100-plus samples per week — double the throughput of the previous Hyperion system.
In addition to multiplexing, speed of analysis is a key metric for spatial proteomics firms, particularly as they position their instruments for clinical research and clinical trial work.
"Certainly if one is entering into the clinical research space, which we are, these studies require a larger cohort, and so being able to double the acquisition speed allows us to open up a much broader clinical research space," Quong said.
Akoya has likewise worked to boost its platforms' throughput. The PhenoCycler-Fusion is designed to run in several different modes to optimize either multiplexing or throughput. It can currently measure up to 50 proteins in parallel at a pace of around 10 samples per week, or up to six proteins per sample in around 100 samples per week. The company aims by the end of the year to enable a workflow capable of measuring 100 proteins per sample in 30 samples per week. It also sells its PhenoImager HT, which can measure up to six proteins per sample in around 300 samples per week.