There’s no more iconic shape in biology than the twisting strands of DNA. Yet that same double helix that makes the molecule instantly recognizable may also be the very thing holding back cheaper, safer, and more effective forms of gene therapy, according to new research.
But foreign DNA used as a treatment can also trigger the innate immune system’s alarm bells, an insurmountable hurdle for scientists trying to develop gene therapies that don’t rely on engineered viruses, which have their own cost, safety and efficacy issues.
A new
study
published Wednesday in
Nature
adds weight to an emerging idea that simply using a single-stranded form of DNA can help it slide past immune sensors when packaged in lipid nanoparticles — the most popular non-viral delivery vessel.
“It addresses one of the main blockers for non-viral genome writing,” Eugene Yeo, a biologist at UCSD who wasn’t involved in the study, told
Endpoints News
. “It’s a foundational enabling technology, albeit not yet therapeutic grade,” he added.
In the Nature paper, scientists from Mass General Brigham and the biotech startup Full Circles Therapeutics describe how they designed a circular form of single-stranded DNA. Their innovation interacts with an emerging class of gene editing enzymes known as recombinases that can plop large pieces of DNA at precise locations in the genome.
The research is part of a trend to develop
one-size-fits-all editing systems
that can replace whole genes — a counterpoint to efforts to create individualized therapies that
fix one patient’s unique mutation
. Shannon Miller, a gene editing scientist at Scripps Research Institute who was not involved in the study, called it an “important step towards generalized, mutation-agnostic gene editing.”
The approach, dubbed INSTALL, worked relatively well in cells in a dish, with the gene being integrated in upwards of 10% of cells. When infused into mice, the gene was integrated into about 0.5% to 1% of liver cells.
That’s too low for most therapeutic applications, but it was significantly higher than the group’s comparator experiments using double-stranded DNA. Many promising gene editing technologies that debuted with similarly modest efficacy have quickly improved after further tinkering.
The immune system’s toxic response to unknown DNA has been long-known. In 2013, scientists studying the innate immune system discovered that an enzyme called cGAS senses double-stranded DNA in the cytoplasm — which the cell interprets as evidence of an infection — to kickstart a chain reaction that leads to inflammation and cell death.
“Having the ability to evade this toxic immune response allows us now to think more seriously about large sequence insertion in primary cells and in animals,” Benjamin Kleinstiver, a gene editing scientist who led the study at Mass General Brigham, told Endpoints. “We’re not saying that everything is solved yet. There’s still some other challenges.”
There have been other attempts by companies and labs that convinced themselves they could get around the cGAS sensor with clever DNA or nanoparticle designs. But the allure of non-viral gene therapy has led to a graveyard of biotech startups and drug programs that never even reached clinical testing.
Rampart Bioscience
, which claimed to have “a unique DNA structure” packaged in lipid nanoparticles, recently shut down, Endpoints reported in January. SalioGen Therapeutics shut down after DNA-filled nanoparticles “disintegrated” cells in the mouse eye and blinded a monkey,
Endpoints reported last year
. Another company, Generation Bio, pivoted to RNA therapies after its non-viral DNA delivery
failed in monkeys
.
Kleinstiver’s lab also struggled to get non-viral DNA delivery to work, and thinks the immune sensor cGAS is key to the field’s failures. “No matter what we did, we found double-stranded DNA to be quite toxic to most primary human cell types and to mice,” he said.
The solution his group came up with was simple: remove one DNA strand to make a single-stranded molecule undetectable by cGAS. The problem is that many enzymes being developed as experimental tools for gene insertion, including recombinases and transposases, also don’t recognize single-stranded DNA.
So the scientists made a shorter piece of single-stranded DNA that sticks to the bigger strand to create a handle big enough for those editing enzymes to grab onto, but small enough to stay “below the sensing limit” of the immune system, Kleinstiver said.
Several tests confirmed that the approach dramatically reduced inflammation compared to standard DNA, and the results suggested that cGAS was unable to detect it.
“They’ve solved one of the key issues with non-viral delivery of DNA,” John Finn, chief scientific officer of Basecamp Research, told Endpoints. His company wasn’t involved in the study but is one of a few leading startups developing recombinase-based therapies.
Finn is excited about the approach, but doesn’t think it is close to replacing viral vectors. The study’s low levels of gene integration were primarily in neonatal mice, whose still-growing livers make gene insertion easier. When the approach was tested in adult mice whose liver cells divide far less frequently, editing levels fell further.
“It’s definitely a step in the right direction, but I don’t think this solves all the problems,” Finn said.
Making single-stranded DNA has historically been more difficult and expensive than regular DNA, although several companies are hoping to change that.
Kleinstiver’s lab worked with Full Circles. The company originally planned to make circles of single-stranded DNA and then snip them to make strands, according to Howard Wu, the startup’s CEO. But when comparing the linear and circular DNA in an experiment, the company found the circular form worked even better.
“It was very surprising,” Wu said.
Full Circles is getting ready to test its approach in an investigator-initiated trial in China. Wu said his company’s DNA is being used in an undisclosed partner’s gene-edited CAR-T cell therapy for liver cancer. He’s also fundraising to start his own programs.
Other early-stage companies are in the early stages of developing therapies based on single-stranded DNA, including the
MIT spinout Kano Therapeutics
, which focuses on gene insertion, and the German company CPTx, which is developing CAR-T therapies.
The emerging form of DNA could be important for several biotech startups in the early stages of developing recombinase-based therapies, including
Basecamp Research
,
Stylus Medicine
and
Seamless Therapeutics
, which in January struck a
partnership with Eli Lilly
that could be worth more than $1.1 billion.
Those three startups have all focused on the herculean task of engineering recombinases, which are widely found in viruses, to work in human cells. They’ve disclosed far less about how they plan to safely deliver the therapeutic genes that the recombinases will insert.
For many of the experiments, Kleinstiver’s lab used a shortcut with its recombinases, simply installing the bacterial DNA they target, rather than designing new integrases to target mouse or human genes. He thinks that improvements in both recombinases and DNA delivery need to come together for the technology in humans.
“We’re really focused on the academic science part of this right now, and moving the efficiencies higher, getting it to a point where we think it would be attractive to take the next steps in terms of commercialization,” Kleinstiver said.
