This story has been updated to correct the resolution of 10x Genomics' Visium platform.
NEW YORK – Vizgen, a Harvard University spinout seeking to commercialize an imaging-based, single-cell spatial transcriptomics technology, is launching the first of two early-access programs for the standalone instrument it is developing.
To start, the firm is targeting between four and eight "sophisticated" academic research centers and biotech companies to implement its proprietary version of multiplex error-robust fluorescence in situ hybridization (MERFISH) technology. MERFISH uses multiple hybridization probes for each transcript to create a quantitative signal that is easily reconstructed and tolerant of errors, leading to multiplexed detection of RNAs in the tissue environment and even within particular cells. Vizgen cofounder Xiaowei Zhuang developed the method in her Harvard lab.
Vizgen noted it is in active talks about early access with five organizations but declined to disclose which ones. One partner already in the program is the Broad Institute. "Imaging transcriptomics is key to unraveling pressing questions in basic biology," Sami Farhi, a senior research scientist who leads the optical profiling platform in Aviv Regev's lab at the Broad, said in a statement.
The Broad's Klarman Cell Observatory and Cell Circuits Program are also using the Vizgen technology. "The commercialization of these methods with robust kits and equipment is a long-awaited step to expand access to the broader research community," Farhi said. (Regev announced in May that she will be joining Genentech as head of research and early development, but her lab will continue at the Broad for at least another year, a spokesperson said.)
An established platform will help grow the method, which, in its current state is extremely difficult to adopt, according to Roy Wollman, a researcher at the University of California, Los Angeles. Wollman's lab had been exploring its own method for combinatorial RNA labeling but dropped that in favor of MERFISH when it came out in Science in 2015. "It took us almost four years to get it working," Wollman said. "It's one of the most challenging things we've done. Where MERFISH is right now, it's not in the place where a lab without a lot of microscopy expertise can pick it up and get it working." But the payoff, measured in new collaborations, has been big.
"There's so many labs in UCLA that want to collaborate with us," Wollman said. "I think it's a very exciting technology."
Launched in 2019, Vizgen has already raised $14 million in Series A financing, led by Arch Venture Partners and Northpond Ventures. At the moment, the firm has only 10 employees at a single location in Cambridge, Massachusetts; however, the firm said it has plans to double, even triple headcount by the end of the year.
Vizgen is the latest entrant into the rapidly growing spatial transcriptomics field and claims to offer subcellular resolution that its competitors aren't even close to reaching. MERFISH can reach resolution of individual RNA molecules down to 100 nanometers. Other platforms, including NanoString's GeoMx Digital Spatial Profiling and 10x Genomics' Visium Spatial Gene Expression have publicized resolution limits of about 10 micrometers and 50 micrometers, respectively.
"This is sort of like the pre-Hubble telescope days of spatial transcriptomics," said David Walt, a Vizgen cofounder and professor at the Wyss Institute at Harvard. "We're bringing Hubble-quality images. Instead of looking at blobs, you're looking at individual stars," he said. Walt is also a cofounder of Illumina and single-molecule detection firm Quanterix. Jeffrey Moffit, a former member of the Zhuang lab now at Boston Children's Hospital and Harvard Medical School, is also a Vizgen cofounder. Peidong Wang, formerly director of advanced technology at Thermo Fisher Scientific, is CEO and VP of Engineering.
And while sequencing-based spatial genomics methods require great sequencing depth to detect rare transcripts, just a single transcript "will light up like a supernova" on the Vizgen platform, Walt said. The method offers breadth as well. "Typically, you can measure 80 percent of all mRNA in a sample and you can quantify them," he said. Single-cell RNA sequencing methods appear to have detection efficiency of only about 40 percent.
Ultimately, Vizgen wants to offer customers a turnkey instrument with integrated fluidic handling and fluorescence microscopy. Walt said Vizgen has contracted a microscopy firm to make that instrument, but for now the systems it is offering partners are built internally from available components. The firm did not say how much these early-access instruments will cost.
In addition to providing microscopy, Vizgen will offer reagents so customers "can do it without getting into all the gory details," Walt said. In an email, a company spokesperson added that the firm has established new MERFISH reagents and protocols for a range of human and mouse tissues. "When one buys Vizgen reagents, they not only get the necessary components for preparing and imaging a sample, but they get the guarantee from Vizgen that they will generate high quality MERFISH data," they said.
A key reagent is the hybridization probes, which must be manufactured at scale, synthetically. Wollman said his lab uses oligo pools from CustomArray, acquired by GenScript in 2017. Similar offerings from Twist Bioscience and Integrated DNA Technologies are potential alternatives, he noted. Vizgen did not immediately respond to questions about who is making their probes.
On a per-sample basis, Walt only said that Vizgen's solution will be "competitively priced," but did not provide more detail. He added that most customers will not be doing full transcriptomes for each run and that cost per sample would fluctuate with the number of targeted transcripts. "They may want to do [whole-transcriptome profiling] once or twice to discover the transcripts of interest, but often that gets honed down for more targeted expression" on the order of 100 to 500 transcripts, he said, which requires fewer hybridizations and takes less time.
Vizgen is planning to start out by using open-source software, Walt said. The system's output is an image file that identifies the location of each RNA transcript. Wollman said his lab developed its own software, which was "not trivial."
"Basically, you're imaging the same molecule 18 to 24 times," Wollman said. "A big part of computation is the registration step; over all the rounds of imaging, you need to make sure you know where the same molecule is." Wollman's software takes several days to analyze an experiment, he said. But the scientifically interesting computational problems are still unsolved.
"Once you get the spatial position of RNAs, what do you do with that? How do you assign them to cells, define cell types, define disease state?" he said. "I think there's a lot of work bioinformaticians will have to do to deal with this type of data." He noted that the Chan-Zuckerberg Biohub has released Starfish, a software suite for dealing with several other imaging-based spatial transcriptomics technologies.
So far, MERFISH has mostly been used in neuroscience, though Walt suggested immuno-oncology and infectious disease research could also be markets for Vizgen. Walt and others hold high hopes that Vizgen will eventually mature into a clinically useful technology.
Diagnosis or prognosis of diseases, as well as evaluating response to treatment, are potential scenarios under which MERFISH could enter the clinic, said Steven Wang, a researcher at Yale School of Medicine who studies genome architecture in mammalian cells and tissues. Wang helped invent MERFISH while working in the Zhuang lab, but is not involved with Vizgen, he said. "Like sequencing, this is a readout assay. If we manage to find linkages between phenotypes that can be profiled, we can use this as a drug screening tool," he said.
Wang has his own home-built MERFISH setup for now, but is interested in acquiring a Vizgen system, as well as other spatial genomics platforms. "Our own system is capable, but I know Vizgen is vigorously pushing the throughput, in terms of how many cells or how large an area they can image in a unit time," he said. "I imagine their product will be better than my current product." Walt said Vizgen's current field of view is about 200 microns by 200 microns, which can be tiled.
Both Wang and Wollman said they see a place for Vizgen's instrument in core labs. "We have too many requests for collaboration," Wang said. "This is something each genomics facility should have at least one of… anything a sequencer can be applied to, this can be applied to."
Wollman said he doesn't think individual labs would be interested in a Vizgen platform once it comes out. But Wang predicted that wealthier labs focused on omics research would want one, in the same way some labs have their own sequencers.
"I think it's really going to be up to the user base" of which applications benefit from MERFISH, Walt said. "Really, the sky's the limit." And while the technology will help researchers determine how cells form tissues, there could be big discoveries to be had at its extreme limit of resolution.
"Our belief in the tech is there's going to be important information encoded in the spatial locations within cells, as much as there is spatial information between cells," Walt said.