BOSTON – Following the recent release of a preprint on the cause of Duchenne muscular dystrophy (DMD) patient Terry Horgan’s death in a closely watched gene therapy trial—plausibly partially the result of an immune reaction to the high dose of adeno-associated virus (AAV) gene delivery vector—there is renewed scrutiny as to the long-term safety of virally-delivered therapeutics. So, it is timely that my first two sit-down interviews at BIO 2023 are about making genetic medicines safe.
Dietrich Stephan, PhD, CEO at NeuBase Therapeutics in Pittsburgh, echoes my concerns. “How many more viral gene therapies do we have to give to people before we know they’re one and done, and that some people are going be primed to have an acute response against them?” he asked.
And Tuyen Ong, MD, the CEO of Ring Therapeutics (and a partner CEO with Flagship Pioneering), reiterated the importance of derisking his company’s viral delivery platform.
I asked both of them essentially: what would a genie (or an AI platform) give back if you wished for the perfect genetic medicine?
Nuclease-free gene editing
Stephan thinks that the answer to safe genetic medicines lies in overcoming hurdles of delivery, tolerability, selectivity, manufacturability, durability, and scalability. Just a couple of weeks ago at the American Society of Gene and Cell Therapy (ASGCT) conference, NeuBase revealed their stealth editors and demonstrated preclinical data that this non-viral, nuclease-free platform does not elicit a cell-based immunity and could be a non-immunogenic in vivo solution.
The “stealth editor” system builds off NeuBase’s Peptide-Nucleic Acid Antisense Oligonucleobase (PATrOL) platform, which uses peptide nucleic acids (PNAs)—small molecules consisting of modified DNA oligos of about 20 nucleotides with peptide-based backbones. By throwing in modifications to the PNAs that enable Hoogstein base pairing (instead of the traditional Watson-Crick method), this system can distort the DNA double helix and disrupt the binding of polymerase. Either of these events sends a molecular “flare” that triggers molecular cut-and-patch machinery called global genome nucleotide excision repair (GG-NER).
Stephan elaborated: “Suddenly this enzyme complex descends down on the locus and snips that PNA-DNA event out so much and throws it away. Now, you’ve got your single-stranded break, and if we add a donor oligonucleotide with a correct base in the middle of it that will slide in where that single strand of break is, it will install the correction. If you’ve got a gain-of-function mutation, you target the mutant sequence, block RNA polymerase from transcribing it, and you’re done. You can do the inverse of that and open up promoters so that they’re more transcriptionally active and boost genome output, which becomes interesting for haploinsufficiencies.”
As of now, NeuBase can inject the naked oligo system, but Stephan said that they are working on encapsulating the PNAs and DNAs in clinically validated lipid nanoparticles (LNPs). With these LNPs, NeuBase is going to aim for in vivo gene editing, starting with the liver and then the bone marrow, joining the race to safely treat hemoglobinopathies, such as sickle cell disease, which have sent a few companies spiraling.
“The field is getting crowded, and it’s a really bad macroeconomic climate for early-stage biotech companies,” said Stephan. “But I think it’s worth continuing to push.” (Neubase’s stock price peaked at more than $11/share in February 2021, but has plummeted since, currently trading around $0.20/share.)
In another bold move, Stephan said that NeuBase has a gene silencing program based on turning the PNA into a molecule that “can cross the cell membrane without the need for a cell penetrating peptide (CPP), like Sarepta or others use, and also engineers out the toxicity that’s associated with CPPs by one or two orders of magnitude.”
Peter Nielsen, who developed this technology in Copenhagen, published an initial report in Science in 1991 demonstrating the molecules’ exquisite sequence selectivity and stability in biological fluids. The next major innovation was conducted at Carnegie Mellon University, which made the molecules soluble and bioavailable. That’s where NeuBase, which was founded in mid-2019, jumped in and licensed the technology.
“There’s a lot you can do with these backbones,” Stephan said. “In fact, there’s almost infinite complexity.”
The commensal viruses inside us all
Instead of sidestepping viral delivery altogether, Ong is looking at the human commensal virome at Ring Therapeutics. When Ong learned about commensal viruses within us, a lightbulb went off in his mind on how to address the safety and delivery of programmable genetic medicines.
“We’ve approached [programmable genetic medicines] in a number of ways with viral, typically AAV, and non-viral delivery systems, such as LNPs, which are still lacking in terms of tropism,” said Ong.
By looking at the viral equivalent of the gut microbiome, Ong and colleagues at Flagship found anelloviruses—named after the Latin word for circular or ring (hence the company’s name). “The virus has circular single-stranded DNA,” said Ong. “We didn’t just find some random virus and ‘Frankenstein’ it into a vector! It was actually more subtle than that. And there’s no known disease associated with it.”
Scientists at Ring, which was founded in 2017, have uncovered a huge amount of genetic diversity in anelloviruses that is driving phenotypic diversity. “These viruses have, for some reason or another, figured out a way out of residing in different tissues of our bodies— they’re tropic for different tissues,” said Ong. “They have somewhat evaded our immune response as well. So far, we’ve collected a library of about 5,000 sequences from anelloviruses from different tissues of the body. Our hypothesis is that the sequence that Mother Nature has given us allows us to have those properties.”
From there, Ring has incorporated machine learning to create a platform for creating tissue-specific vectors. “Nobody else is working on anelloviruses, so we’re taking a very coordinated approach and trying to de-risk the scientific understanding in our development plan,” said Ong.
One limitation of these anelloviruses is that they can package only a few kilobases of nucleic acids. But Ong believes that given the challenges already around gene editing and delivery, it’s better to address the delivery challenges and tropism. “People haven’t been able to move away from the gravitational pull of the liver, and I think we have the ability to do that.”
Ong said that about 100 employees at Ring are working hard toward getting the first patient dosed. “In terms of our preclinical data, we’ve generated anelloviruses from different tissues, vectorized them, and shown that we can repeat doses and target tissues of interest,” said Ong. “We’re starting to build that profile of going into the clinic in a very deliberate fashion. We’ve identified indications where we can differentiate from a delivery standpoint.”
One small step for man, one large step for mankind
NeuBase and Ring are two examples of how to bring genetic medicines to fruition by facilitating human-specific attributes to create safer therapeutics. On the one hand, Ring Therapeutics is harnessing the harmless viruses within us. On the other hand, NeuBase is facilitating a DNA repair mechanism used by human cells as opposed to the CRISPR-Cas system, which is from a bacterial immune system. Prokaryotic and eukaryotic cells diverged over two billion years ago, with eukaryotic species becoming increasingly complex in terms of their genomes and the machinery that they had to evolve to maintain those genomes in a high-fidelity state.
“The family of DNA repair proteins—dozens and dozens of different proteins that act in different ways with different mutations—in aggregate clean up about a million mutations per cell per day and do it across trillions of cells pretty darn well,” said Stephan. “I think that gives you an intuitive sense of the exquisite fidelity of those machines that operate constantly versus some reports of CRISPR-Cas base- and prime-editors in terms of 0.1% error rates versus 10-6 to 10-9 error rates.”
It is a critical time for gene therapy to not fall into another dark age, like the one that followed Jessie Gelinger’s death in 1999, especially with so much invested. In the field of base editing, for example, David Liu noted at ASGCT that it takes just six years from the first publication to a successful clinical intervention, which is a tribute to the 3,000 labs and more than 4,000 publications in the field building on that core technology.
NeuBase and Ring are just two of the many efforts in genetic medicines inching closer to the clinic. The more these approaches can be derisked, the better the chance to make a positive impact on patients’ lives.
