Despite the rapid development of genetic medicines, Simon Harnest, chief investment officer and SVP strategy at Metagenomi, said that attending the annual ASGCT meeting is a good way to keep tabs on the industry. “There are so many specialized groups in gene editing, some that focus on base editing and have great data, some that focus on prime editing or larger gene integration systems, and then other groups that discover novel nucleases.”

Yet Harnest and his colleagues at Metagenomi are unfazed, having developed a gene-editing toolbox that is driven by metagenomics. “I’m thrilled that Metagenomi is not just playing in these fields,” said Harnest. “Having this toolbox is what enables us to be so fast and to be at the forefront of the greatest new inventions.”

Metagenomi came to ASGCT with three abstracts (two oral and one poster), all of which center on the company’s gene editing systems. “We’re talking in some abstracts about this kind of modularity where we have validated modules that we already know from one construct, and we can quickly assemble them to create a chimera of a system that the wild type wasn’t able to target with high efficiency,” said Harnest. “But in one engineering step, we get to high efficiency rather than in eight rounds of directed evolution. So, having a metagenomics library really plays an advantage in the real world.”


The first application of a type V nuclease from Metagenomi’s metagenomics library was discussed in a presentation about a novel type V nuclease called MG29-1. According to Harnest, the spCas9 nuclease, a member of the type II family of CRISPR systems, is the basis for many gene-editing tools. However, the Metagenomi team discovered that certain type V CRISPR systems have a higher level of specificity and unique PAM sequences that allows them to access genomic target sites that Cas9 can’t reach.

In this talk, Metagenomi demonstrated the high efficiency (up to 55% of the whole liver and about 75% of hepatocytes) with which MG29-1 can edit genes in primary human cells in the lab, as well as in mice and non-human primates (NHP). This is one of the first reports of NHP editing using a type V nuclease, and it suggests that therapeutic gene knockdown can be done safely and effectively. Traditional methods, such as in silico screening and in vitro digestion, show no off-targeting with MG29-1.

According to Harnest, the Metagenomi systems have helped the company speed up these types of projects by allowing them to screen many more guides in combination with different nucleases. “When you think about editing efficiency and really getting to therapeutic levels of high editing efficiency, you need to screen a lot of different guides, and it’s a bit of trial and error to find the right guide-nuclease combination that gives you that very high efficiency that you want to move forward with,” said Harnest. “But the more systems you have, the more guides you can screen.”

Ready to paRTy

Whereas the first talk focused on a single, novel nuclease, the second discussed metagenomics as a platform and a library from which various gene editors can be derived. Metagenomi has been working on constructing a library of nucleases, as well as a library of deaminases and a library of reverse transcriptases (RT), all of which are essential for more advanced engineering systems that allow for larger integrations.

“We have shown that we can make a base editor out of a type II nuclease, and we’ve shown a lot of progress with our RNA-mediated integration systems, or what we call RIGS,” said Harnest. “We categorize them into little RIGS and big RIGS because our metagenomics platform allows us to go into the reverse transcriptase library and find RTs that are capable of larger gene corrections.” Harnest stated that initial RIGS performance has been very promising.

Taking off the CAST

The third poster abstract focused on CRISPR-associated transposases (CASTs), which have been hailed as a major advance in the field of gene editing due to their ability to integrate DNA payloads contained within the element at a precise position, with a specific orientation, and in a programmable manner. The integration efficiency of most transposases is low, Harnest said, and these naturally occurring bacterial biodefense mechanisms are difficult to engineer to make them work in human cells. 

“Some groups, like Sam Sternberg’s lab and others, have been trying to crack that code to engineer task systems,” said Harnest. “We were the first group, as far as we know, that’s got that to work in human cells. That was in the second half of last year, and we gave the first presentation on this earlier this year. It’s very early days with this type of technology, but we see the promise of replacing genes with fully functioning genes. That would be absolutely incredible to do that site-specifically.”

A plug-and-play pipeline

Sarah Noonberg, the new chief medical officer at Metagenomi, is enthusiastic about these presentations because she has spent her entire career focusing on genetically defined indications, most notably hemophilia at BioMarin. “With this library of tools, you can start to think about what indications you want to go after and pick the right tool, as opposed to what is the right indication for a particular tool,” Noonberg told GEN. “It’s a whole different perspective on pipeline development and buildout.”

The first gene editing procedure at Metagenomi will involve the use of double-strand breaks (DSBs) in the liver. “Our pipeline development is going to be a step-wise approach to leveraging what we learn in liver delivery with these DSBs,” said Noonberg. “We will take these well-characterized DSB-nucleases that we know how to make and go outside of the liver, leveraging what we’ve learned and minimizing the risk.”

Noonberg stated that Metagenomi is keen on focusing on the CNS, where knockdown and knockout techniques offer the possibility of treating serious conditions with an acceptable risk-to-benefit ratio and solid clinical validation. While this is going on, Metagenomi will investigate liver delivery of RIGS and CASTs on the premise that they have already mastered delivery to this organ. They will then spread to organs and tissues apart from the liver.

“We carefully choose any indications that we believe we can move the needle on,” said Noonberg. “There’s a lot we can learn now rather than waiting and optimizing again, not just on the nuclease tool system and delivery but also in the regulatory environment. Building out all of these capabilities will provide a lot of learning that will allow us to move even faster with our next generation’s editing tools.”

When I compare the current state of genetic medicine regulation to the Wild West, Harnest quickly and politely corrects me.

“We’re extremely mindful of how we push things forward, and when you see presentations where little kids are born with genetic defects in walking and talking and then suddenly they can play soccer and talk—to us, it’s really bringing it back to the human aspect of this,” said Harnest. “When you look at our presentations, you see a lot of different nucleases, but it all leads back to the patients. There are so many indications where life expectancy is two years because you’re diagnosed when you’re born, and there’s very limited time to do something. So, we have a sense of urgency to get things into the clinic to help the patient population out, but we’ll do it in a very thoughtful way and not move too many parameters.”

Metagenomi persists in its microbe-rich ecosystem mining in order to discover, collect, analyze, and translate the genetic code of hitherto undiscovered organisms, the results of four billion years of evolution. In a few years, which is a nanosecond in the grand scheme of things, we may find out if Metagenomi’s extensive library of CRISPR nucleases can be refined into a powerful precision gene editing toolbox capable of curing diseases and saving lives.