NEW YORK – Swedish startup Pixelgen Technologies aims to carve out a new niche in spatial proteomics with plans to commercialize a technology for mapping the spatial distribution of cell surface proteins at the single-cell level next year.
Founded in 2020, the Stockholm-based company, which currently counts around 20 employees, emerged from stealth mode this week, backed by a roughly $6 million seed funding round led by Swedish venture firm Navigare Ventures.
Pixelgen was cofounded by Simon Fredriksson, who is also CEO. Fredriksson was previously cofounder, CEO, and CSO at proteomics company Olink, where he was an inventor of that firm's proximity extension assay (PEA) technology. Pixelgen uses a twist on that technology in its spatial proteomics platform.
PEA uses pairs of antibodies linked to DNA strands that are brought into proximity when the antibodies bind to their target. The DNA is then extended by a polymerase that creates a new sequence that can be used as a surrogate marker for the target protein.
Pixelgen's approach, which it calls Molecular Pixelation, similarly uses antibodies tagged with DNA to detect and quantify target proteins but uses elongation of those DNA strands to localize those proteins on the cell surface.
The company starts by incubating fixed single cells with nucleic acid-tagged antibodies to cell surface proteins of interest. They then treat the cell with what Fredriksson called "DNA pixels," pieces of DNA with sequences designed to fuse with the DNA tag on the antibody. The final step of the assay uses elongation to link all of these DNA pixels to each other, forming "an informatics grid" that allows researchers to identify both what proteins are present on the cell surface and where on the cell they are located, he said.
"It forms like a kind of honeycomb [of DNA] around a spherical cell," he said. "You get PCR products that you send to your next-generation sequencer, and the sequence combinations are read out and fed into software we have developed that then figures out where all the proteins are in relation to each other."
Most spatial proteomics tools are designed to identify and localize particular proteins and cell types in tissues, letting researchers study, for instance, how the spatial arrangement of different immune cells in a tumor microenvironment might correlate with response to therapy. Pixelgen, on the other hand, aims to help researchers unravel how the localization of various proteins on a cell's surface impacts biology.
Cell activity "is of course governed by what protein markers are on a cell, but also by how they are organized in clusters, forming immune synapses and so on," Fredriksson said. "Protein clustering is really, really important for the activity of cells communicating with each other."
He noted that surface protein clustering has often been studied via microscopy but that this approach is limited in terms of multiplexing of protein targets and sample throughput. With its Molecular Pixelation technology, Pixelgen aims to enable highly multiplexed and high-throughput studies of cell surface protein clustering and localization.
To date, the company has multiplexed as many as 100 proteins in an experiment. Fredriksson said that they include proteins of particular biological interest — many of them classic immune markers — as well as housekeeping proteins to make sure the cell surface is tagged densely enough to enable the formation of a DNA sequence latticework around the entire cell.
He noted that higher multiplexes produce denser grids of DNA around the cell, which improves the quality of the spatial analysis.
"So the more you multiplex, the easier this gets, which is kind of counterintuitive with multiplexing," he said.
Fredriksson said the company's DNA-based readout isn't subject to the same multiplexing limits as fluorescence-based approaches, and that the availability of high-quality antibodies is the primary constraint.
"We use very highly validated antibody clones for this to make sure we are detecting the right proteins and verify them internally in our assay to make sure everything makes sense," he said. He added that the large amount of research on cell surface proteins and immune markers over the years meant that good antibodies to many of these targets are widely available.
Fredriksson declined to say what level of multiplexing or throughput Pixelgen's Molecular Pixelation product would offer upon launch but said that the company would release more information on the system prior to its planned release sometime next year.
Speaking of spatial proteomics more generally, he said he believes "there will be an unsatisfied thirst for more multiplexing as long as the underlying technology can deliver that multiplexing at reasonable cost and throughput."
Pixelgen's system won't involve an instrument but will consist of the reagents required to tag the proteins of interest and produce the nucleic acid sequences, Fredriksson said. Researchers can then send the nucleic acid output to a core laboratory or other facility for readout via next-generation sequencing. He said that the initial version of the product will be compatible with Illumina sequencers.
The company is currently working with academic collaborators to explore potential applications for the technology, he added.
"We see a lot of interest around basic research around immune cell activation, immune cell interactions, and for drug development," he said.
He cited engineered immune cell therapies as another area where he believes the technology could find users.
"All engineered T cells and other engineered cell therapies are dependent on the clustering of the target protein that you are going to attack," he said. "This isn't really studied to a great extent because there is a lack of tools. We want to fill that gap."
While Pixelgen's initial focus will be mapping cell surface proteins in single cells, Fredriksson said the company also plans to apply the technology to the analysis of intracellular proteins and tissue samples.