Startup’s ‘escape artist’ approach could help therapeutic DNA evade cellular silencing – with future implications for neurodegeneration.
Gene therapies have transformed the outlook for a growing number of inherited diseases, yet one stubborn obstacle remains. Delivering therapeutic DNA into cells is only half the battle; persuading those cells not to switch it off can prove considerably harder. Long before a treatment has the chance to wear out, the body’s own defenses may have quietly shown it the door.
London-based Houdini Bio believes it has found a way to change that. Emerging from stealth with £1.5 million ($2m) in grant and pre-seed funding, the company has developed a machine learning-guided platform that redesigns therapeutic DNA to evade the HUSH complex – a key cellular mechanism that silences foreign genetic material. While the immediate opportunity lies in improving the durability, safety and economics of gene and cell therapies, the approach could ultimately have particular relevance for neurodegenerative diseases, where sustaining gene expression in long-lived brain cells may prove just as important as delivering the therapy in the first place.
Longevity.Technology: Gene therapy has spent much of the past decade solving one problem – getting therapeutic DNA into cells. Increasingly, however, the greater challenge is keeping it there. Cells have evolved sophisticated mechanisms to recognize and silence foreign genetic material, an entirely sensible defense against viruses that becomes rather less helpful when the DNA in question is intended as medicine. If Houdini Bio can indeed teach therapeutic sequences to evade that molecular lockdown, the implications could extend well beyond improving manufacturing economics or reducing doses.
Durable gene expression is likely to become an increasingly important consideration in age-related diseases, particularly neurodegenerative disorders, where long-lived neurons offer little margin for therapies that gradually fade from view. The company may have borrowed its name from history’s greatest escape artist, but the real trick here is not making DNA disappear – it is persuading it to remain, quietly and productively, in plain sight. To find out more about durable gene expression, neurodegeneration and why escaping cellular silencing could reshape the next generation of genetic medicines, we sat down with Jonathan Cohen-Gold, CEO and Cofounder of Houdini Bio.
Lower doses, longer-lasting therapies
One of the biggest challenges facing gene therapies for neurodegenerative disease is simply getting enough treatment into the brain without triggering a toxic immune response. Cohen-Gold believes Houdini’s platform addresses this roadblock not by changing how the drug is delivered, but by maximizing its impact once it arrives.
Cohen-Gold believes Houdini’s approach addresses that problem from an unexpected direction. Rather than improving delivery itself, the company focuses on making every copy of therapeutic DNA work harder once it has reached its destination.
“Due to the inflammatory state which accompanies neurodegeneration, toxic immune responses to gene therapy are a major concern,” he says. “In the case of CNS diseases specifically, as the field is pushing towards systemic delivery and fighting against the blood brain barrier, dose requirements are currently extremely high. By enhancing the productivity of each DNA copy once inside target cells, our technology enables significant dose reductions. This lessens dose-dependent toxicity and thereby enhances safety.”
The company says its anti-silencing modifications are agnostic to payload, delivery platform and target cell, meaning they can complement existing gene therapy approaches rather than replace them.
“Once inside cells, Houdini-engineered gene therapies are resistant to cellular shutdown by HUSH,” says Cohen-Gold. “This ensures the therapy can be durably active in patients, another bottleneck which currently limits long-term benefit across the wider field.”
Building for the next generation of CNS therapies
That platform-first philosophy also extends to vector choice. While many dementia programs rely on AAVs, the field is increasingly exploring lentiviral vectors, transposons and non-viral delivery systems, each with their own advantages and limitations.
According to Cohen-Gold, Houdini’s technology has already demonstrated compatibility across multiple delivery modalities.
“We’re extremely excited by how widely applicable our technology is across vectors and target cell types,” he says. “We’ve proven this across around 10 cell types for AAV, lentivirus, transposons and naked plasmid. We’re currently in the early phases of testing in neuronal and glial lines.”
The company is also beginning to look beyond proof-of-concept. While in vivo benchmarking studies are already underway, Cohen-Gold says commercial discussions are progressing with partners seeking to strengthen preclinical assets, including programs with a central nervous system focus.
“We are currently benchmarking a select few constructs against state-of-the-art comparators in vivo, but the real work starts early next year when we will begin to expand across indications and disease models,” he says. “We’re currently negotiating with commercial partners to improve their pre-IND assets, some of whom do have a CNS or CNS-adjacent focus.”
Engineering better economics
Gene therapies face biological challenges, but commercial ones too. Manufacturing complex viral vectors remains expensive, helping to explain why some approved therapies carry seven-figure price tags.
For Cohen-Gold, improving expression offers a practical route to reducing those costs because fewer vector particles are needed to achieve the same therapeutic effect.
“We estimate that a three-to-five-fold reduction in dosage could halve the cost of goods for AAV therapies, which routinely reach $1 million per dose, by allowing a single batch to be split into multiple doses,” he says.
The company is also applying the same anti-silencing strategy beyond gene therapy manufacture. “We are focusing on HEK cells and currently have ongoing partnership projects to assess our technology within industry-adopted systems,” he explains. “We’ve also validated our technology in CHO cells which produce complex antibody formats. In this setting, we can directly increase yield versus state-of-the-art approaches… we welcome the opportunity to partner with existing manufacturers to improve their yields, particularly as payloads increase in complexity and become more susceptible to HUSH silencing.”
The disappearing act
Much of gene therapy’s progress has been driven by increasingly sophisticated delivery systems. Houdini Bio is betting that the next gains may come after the DNA arrives, by helping therapies coexist more successfully with the cells they are designed to treat. Whether that proves to be an enabling technology for neurodegeneration remains to be demonstrated, but if durable expression becomes the next frontier for genetic medicines, escaping cellular silencing may turn out to be less of a magic trick than an essential piece of biological engineering.
