NEW YORK – Vizgen, a Harvard University spinout, unveiled its Merscope platform this week, one of three startups that used the virtual Advances in Genome Biology and Technology conference as a launch pad for their spatial gene expression analysis platforms.
Rebus Biosystems, the latest incarnation of a firm commercializing specialized optics technology, launched its Esper platform while newcomer Veranome Biosystems surfaced to describe its Pisces platform. All three platforms employ multiplexed versions of single-molecule fluorescence in situ hybridization, long the standard for high-sensitivity gene expression detection. Vizgen and Rebus touted their platforms during sessions afforded to them as meeting sponsors, while Jinyue Liu of the Agency for Science, Technology, and Research (A*STAR) Genome Institute of Singapore shared data generated on Veranome's platform from human brain organoids during a Wednesday talk.
On Wednesday, Vizgen CEO Terry Lo gave conference attendees a peek at his electric guitar collection and the Merscope platform, which will be commercially available this summer on a limited basis. The platform, which builds on multiplexed error-robust FISH (MERFISH), developed by Harvard researcher Xiaowei Zhuang, offers gene expression analysis of hundreds of genes at a time in hundreds of thousands of cells at a time, he said, all at subcellular resolution.
"As we look at the marketplace, what customers seem most interested in is being able to home in on a certain number of targets based on what their hypotheses are," Lo said in an interview prior to AGBT. "You want high resolution, you want high multiplexing, and you want high sensitivity… This is the first time there is going to be a platform that is going to capture all of this."
Rebus' Esper platform allows automated detection of up to 30 custom genes at once with the company's high-fidelity assay for RNA analysis, Brett Cook, senior product manager, said in his presentation. He presented data from a study of mouse brain suggesting that only 17 cell type marker genes were needed to identify all major cell types. The company did not disclose pricing.
Veranome, a spinout from A*STAR, can detect hundreds of genes at a time with Pisces, said Chris Streck, the firm's director of applications. "We know we can go higher than that, but at this stage, that's where we're focused," he said. Veranome has been offering services for two years before coming out of stealth this week and has placed several early access instruments with researchers, though Streck did not say where.
"This is a breakthrough technology," said Rui Chen, a researcher at Baylor College of Medicine, and a Vizgen early-access program customer. "Once you've done single-cell RNA-seq, it's a natural step to do [spatial gene expression] next." His lab is new to spatial analysis and chose to buy a beta instrument from Vizgen because he wanted to try it out as soon as possible.
"You either hold off your experiments and wait for something bigger, better, cheaper, or you just jump in," he said. "There are definitely barriers and challenges, but I'm extremely excited about this technology."
The launches suggest that the interest in spatial genomics built up last year will only intensify as researchers get more ways to generate data, especially at higher resolution.
Along with single-cell RNA-seq, highly multiplexed methods like 10x Genomics' Visium and NanoString's GeoMx digital spatial profiler have allowed researchers to get broad views of gene expression patterns in their samples. But the drive to type single-cells has led the field to seek higher-resolution methods.
FISH-based methods have been the gold standard by which other gene expression is measured in terms of sensitivity and some technologies offer resolution near the physical limits of optical microscopes. BioTechne's ACD RNAScope, which detects up to 12 genes at once, has been used for years and has seen a surge in demand.
So, while NanoString and 10x have announced plans to offer new platforms with resolution at single-cell or subcellular resolution, it's perhaps not surprising that smFISH-based technologies are popping up first.
Vizgen, which popped up in 2019 to commercialize MERFISH, has already raised $14 million in Series A financing and grown to just over 20 employees, with plans to raise even more money and grow to about 60 by the end of the year.
Lo declined to disclose Vizgen's revenues but said the company has started multiple projects with three of the largest pharma companies, though he declined to disclose which ones.
Vizgen's Merscope includes several improvements to MERFISH, including a larger format camera that increases the area of analysis to 1 cm2 — thereby increasing throughput — and chemistry updates to reduce background autofluorescence, improve reagent stability, and make it applicable to a wider range of samples types.
"We can work with any type of sample, pretty much any type of tissue in mouse and human," Lo said. The only limitation comes from working with RNA itself. "Degraded RNA is problematic. It comes down to the RNA quality within sample itself."
Vizgen launched its early-access program last summer, but Chen, the Baylor researcher, has only had his instrument for about six weeks, he said. So far, his lab has only been able to run a validation experiment in mouse brain, targeting less than 200 genes, but "the results were pretty good," he said. "We're pretty confident the machine performs well." Soon he hopes to get spatial maps of cell types in eye tissues.
In addition to resolution, Chen is excited about Merscope's cellular throughput, which he said makes it a more affordable gateway to single-cell analysis than droplet-based sequencing. "It's much cheaper per cell, probably ten times cheaper or more," he said, mostly because it does not add the cost of next-generation sequencing.
Chen said he paid about $170,000 for his instrument, an alpha version that will not resemble the one available at commercial launch. Vizgen hasn't set its final pricing, but setting up Merscope will cost less than $300,000, Lo said, including any necessary computing infrastructure. "It's not just a core-lab platform, it's something that's in the price range of other capital equipment for research labs," he said, noting that per-sample pricing depends on the size of the panel, but that pricing would be similar to single-cell sequencing runs.
Rebus, fresh off a $20 million Series B financing round in November 2020, may have finally found the killer application for its optics technology. The company, formerly known as LightSpeed Genomics, had previously tried to tackle next-generation sequencing.
Using lasers and algorithms, the company is able to create a high-resolution image with a broad field of view, as much as 4 cm2, enabling it to analyze hundreds of thousands of cells at a time, over two days, Cook said in his AGBT presentation.
In an AGBT poster, researchers from Rebus and the University of California, San Francisco noted that "unique primary probes are synthesized to each mRNA target with between 33 [and] 96 primary probes hybridizing to each target transcript, depending on the length of the mRNA exon." Secondary fluorescent readout probes can bind each primary probe and are used to reveal the gene locations.
Veranome draws on expertise from the semiconductor industry to take high-resolution images at speed, Streck said. The Pisces platform includes custom panels and a software suite that enables researchers to determine cell types based on the RNA data, then recapitulate the cell populations and tissue architecture back onto the images. It has been shown to work with mouse and human tissues, organoids, spheroids, and some cell cultures for established and primary cell lines, Streck said.
Liu said in her talk that she was able to analyze up to 557 genes in an area as large as 15 mm2, which contained more than 1.1 million cells. Streck noted that Veranome has not settled on how big it would make its maximum area of analysis.
Streck declined to disclose per-sample costs, which vary depending on the project, but said the company aims to offer lower cost per sample than single-cell sequencing.
All three platforms will be useful for multi-omics analyses, with Veranome's platform already able to analyze proteins.
"A big advantage with MERFISH is that we can project other multiomic datasets onto the spatial transcriptomics map (e.g., RNA and ATAC data) using cell type anchor genes," said Benjamin Humphreys, a researcher at Washington University in St. Louis and a customer of Vizgen's lab services business.
Cook said Rebus is adding immunofluorescent detection of proteins to its high-fidelity assay in Q2 and will release a kit to process formalin-fixed paraffin-embedded samples later this year. DNA will join RNA as an analyte next year, he said. Further down the pipe, Rebus hopes to have assays that can detect tens of proteins and thousands of genes simultaneously.
For Vizgen, even higher multiplexing is an immediate focus. "At release, we'll be comfortable in the 500-gene range," Lo said, with up to 1,000 genes in certain panels, but he wants to go further. Improving the sample throughput will be important for unlocking the clinical or translational research markets, he added. Imaging a 500-gene panel currently takes a little over a day on the instrument. "Whether clinical studies would need the same amount of gene targets, imaging area, and high resolution still needs to be defined by the market and research community," Lo noted.
Of course, there's still the Merscope product launch to prepare for. "We're building out and expanding facilities right now to meet capacity for commercial release," Lo said.