Skip to main content
Premium Trial:

Request an Annual Quote

EEL FISH Study Shows Proof of Concept for Spatial Method Licensed to Rebus Biosystems

Premium

NEW YORK – Researchers from the Karolinska Institute have published a proof-of-concept study for enhanced electric fluorescence in situ hybridization (EEL FISH), a spatial transcriptomics method licensed to Rebus Biosystems.

The method uses electrophoresis to transfer RNAs to the surface of a tissue section, where it is then captured on a separate glass surface.

The method can decode 448 genes per color channel and an area of approximately 1 square cm. In the paper, the researchers demonstrate using two color channels, or the ability to detect nearly 900 genes, but Lars Borm, a graduate student at the Karolinska Institute and first author of the study, published last month in Nature Biotechnology, said the lab can now "routinely" do three colors. While single-molecule FISH is the gold-standard for transcript detection efficiency, this method only transfers about 20 percent of the RNA to the capture slide, representing total efficiency of 3 percent to 13 percent, Borm noted. By reducing the planes that need to be imaged per sample, it can take the time to process each slide down from two weeks to just two and a half days.

"Before, we couldn't look at large samples at this resolution" and throughput, Borm said. Using OSM FISH, the lab was only able to do 33 genes per sample in a process that took up to two weeks. "The other big thing is that it enables studying human samples at high resolution," he said. After the RNA is transferred to the slide, the rest of the sample is removed, and with it background noise. "Human tissues have ridiculously high autofluorescence, which makes it almost impossible to image anything that is small," Borm said. Removing most of that noise enables studying human samples.

The method has already garnered interest, as spatial genomics startup Rebus Biosystems acquired the intellectual property rights to the assay in December 2021. Simone Codeluppi, a former Karolinska Institute researcher and coauthor of the paper, is a bioinformatician at Rebus.

But how far along the firm is in commercializing the technology is unclear. Rebus and Codeluppi did not respond to a request for comment. Borm declined to comment on commercialization of the technology, suggesting that was better answered by Rebus. At the time it acquired the EEL FISH IP, Rebus said it would have a commercial launch in June.

Yu-Chun Wang, a technology implementation specialist at VIB's Tech Watch team, said his lab has expressed interest in Rebus' EEL FISH, as well as the Rebus Esper spatial genomics platform, but has not yet gotten their hands on either.

"It's very interesting technology," he said. "Once this can be productized or [offered under] early access, we're very interested to try it," he said, noting that Tech Watch would rather wait to try out the commercialized workflow rather than dive into the academic protocol outlined in the paper.

"It was a 'stupid' idea, like, 'Oh, that would be fun,'" Borm said of EEL FISH. His lab was trying to analyze RNA in large human tissue samples and the optics they were using required them to stack many images in the Z-axis, due to the shallow depth of field of the lens.

Wide-field microscopes always offer worse resolution in the Z-axis, he said, "so you can't do anything with that information," he said. "The image analysis pipeline is throwing all that information away. The team thought, 'If we're doing this, can't we just skip the whole thing?'" They came up with the idea to make an RNA blot at the surface of the tissue section.

The team acquired some indium tin oxide coated glass slides to try out on a Friday afternoon. "Luckily, we got good results in the first trial," Borm said. "That's how we were convinced to move forward."

EEL FISH's process of transferring molecules to a different slide not only offers a boost in signal-to-noise ratio, due to removing autofluorescence often seen in human tissue samples, but also may open up the technique to analysis by sequencing, not just optics. A sequencing-based approach to molecules transferred to a glass slide with electrophoresis "hasn't been tested," Wang said, but noted, "it would be interesting if they're doing that."

"It is an interesting suggestion, but it might not be straightforward," Borm said, because one would need to track the spatial origin of the RNA.

Wang also suggested that improving RNA transfer and capture efficiency would go a long way to improving the method in comparison to other spatial transcriptomics techniques. For example, Merscope, the version of multiplex, error-robust FISH (MERFISH) being commercialized by Vizgen, touts detection efficiency of 80 percent.

In addition to the wet lab process, the Karolinska Institute team built a new bioinformatics pipeline to process two terabytes of data generated per slide. "We use distributed computer architecture that spits out all the locations of RNA molecules," Borm said. Not only that, but the algorithm also assigns them to single cells.

In the paper, the researchers presented data from human visual cortex samples. Borm said that the method was already being used in three studies, including a consortium his lab is a part of, but declined to say more. "One is quite big and will be human-focused," he said.