I stared at the display of cell culture images on the monitor. Every one of them was dark.
Sometimes a breakthrough comes from what is not there.
The blank space on my computer monitor was our first concrete signal that we had solved the single most vexing obstacle standing in the way of our mission. It was thrilling. We were one step closer to a redosable, titratable and scalable gene therapy, with the potential to transform the lives of millions of people worldwide living with rare and prevalent diseases.
To appreciate the excitement of that moment, it helps to understand the limitations of present-day gene therapies – limitations that we are designing our technology to overcome. All currently approved gene therapies, and most of those in development, use a viral delivery vehicle (a vector) to deliver a corrective gene. These viral vectors are excellent at ferrying the corrective DNA into the nucleus of cells and have been a great therapeutic advancement. But the most commonly used vector, adeno-associated virus (AAV), has significant challenges.
The most problematic challenge is the reaction of our immune system to AAV, specifically to its capsid or protein shell.
A large number of people have pre-existing antibodies to the AAV capsid, meaning their immune system will neutralize the virus as soon as it is detected -- before it can deliver its payload. Those individuals are not eligible to receive AAV-based gene therapy. People without pre-existing antibodies can receive the AAV therapy, but they will develop antibodies to AAV after the initial dose and therefore can only be treated once a lifetime. This prevents redosing to extend benefit should expression wane over time. It also makes it extremely challenging to treat children, especially with liver-directed therapies, because their cells divide rapidly, so they will soon “outgrow” their single lifetime dose.
Given these limitations, the team at Generation Bio knew we had to develop a different approach to get our closed-ended DNA (ceDNA) constructs into patients’ cells. We turned to the most clinically validated non-viral delivery system, lipid nanoparticles (LNPs).
These tiny particles – orders of magnitude smaller in diameter than a human hair – have been used to deliver nucleic acid therapies for years. They’re relatively easy to manufacture. They’re effective at delivering cargo. They don't engage the immune system like AAV does, so patients can potentially receive as many doses as they need over a lifetime. Importantly, LNPs have also been in use so long that clinicians have accumulated the equivalent of thousands of years of patient data showing they are safe for therapeutic delivery.
However, the standard LNP approach has a very important limitation: as a class, LNPs are somewhat indiscriminate in their biodistribution; they deliver their cargo to multiple cell types, including to immune cells. Although nucleic acid cargo is meant to be therapeutic, immune cells perceive it as an invading snippet of foreign genetic material. They mobilize to clear and neutralize the threat; acting like a molecular vacuum cleaner, they suck up so many LNP particles so quickly that very few are left to reach the tissues you’re trying to target. In some cases, the immune system also sends up an inflammatory response that can damage tissues.
It’s possible to get around this challenge if you are utilizing LNPs to deliver RNA as the therapeutic payload, because RNA can be chemically modified to hide itself from the immune system. Specifically, modifications to uridine (for mRNA therapies) and the 2’ ribose (for siRNA therapies) can render RNA relatively “invisible” to immune sensors while maintaining pharmacological activity. So even though much of the dose is delivered off-target, the consequences are neutralized.
At Generation Bio, however, our goal is to deliver DNA, not RNA; we aim to transport corrective genes so the patient’s own cells can produce the proteins they need to stay healthy. Unfortunately, DNA can’t be modified to hide it from the immune system in the same way RNA can. To bring our LNPs directly to the tissues that needed them, we would have to find another way to bypass the immune system.
This had never been done before. And this was the immense challenge facing the Generation Bio team:
Could we be the first to engineer an LNP that would deliver DNA only to specific tissue, without triggering a damaging immune response?
The day we finally got it right was one of the most memorable of my career. We had set up two dishes of cells, side by side. One contained the LNPs with our latest immune-evading modification. The other contained LNPs without the modification. In both wells, the nanoparticles were tagged so that they would glow red if they were engulfed by immune cells.
As expected, the image of the wells with the unmodified LNPs were soon speckled with bursts of red. But the wells with our new modified LNPs remained black. That was our Eureka moment: Initial data indicating that we could solve this decades-old challenge. We could engineer an LNP delivery system that would not be destroyed by the immune system or provoke an inflammatory reaction.
I’ll never forget my elation when I looked at my computer monitor and saw the contrasting images.
My first thought? Game On!
After a celebration to recognize the team’s accomplishment, we shifted our focus to the next step: how to get our LNP system to deliver ceDNA to specific cell types, while limiting delivery to off-target cells. This was not nearly as challenging as avoiding the immune system. Over the years, scientists have identified and characterized receptors specific to a number of cell types, including liver cells, muscle cells and retinal cells. We also know which ligands bind to those receptors. In the liver, for example, a ligand called GalNAc is known to bind to a receptor known as ASGPr that’s present on many hepatocytes (liver cells). We took advantage of this well-validated biology by attaching the GalNAc ligand to our LNPs. Sure enough, it acted similarly to the address on an envelope, directing the LNPs to the liver, where they bound to the ASGPr receptors on hepatocytes. Once fused to the cells, they could deliver their therapeutic cargo.
The specificity of our LNP for liver is high, with 97% of the dose reaching the liver in rodent studies. And most importantly, it has good in vivo tolerability – confirming our prediction that on-target delivery to liver cells and avoidance of delivery to immune cells would eliminate the inflammatory response mediated by immune-sensors.
Another benefit of this approach is that it’s modular. We are now working on swapping out the liver-targeting GalNAc and affixing a retina-targeting ligand or a muscle-targeting ligand or even, potentially, a tumor-targeting ligand to the LNPs. Whichever you attach should serve as a new address, directing the particles to deliver their cargo directly to the tissue you aim to treat, which we will test experimentally.
We call our delivery system cell-targeted LNP, or ctLNP, because it incorporates this precision targeting system. We combine it with our proprietary ceDNA construct for a therapeutic approach that we believe could revolutionize gene therapy by enabling titratable, redosable therapies that can be efficiently manufactured at scale.
Our ultimate goal: To treat more patients, with more diseases, in more places around the world, seeking to bring the benefits of genetic medicine to many more families than what is currently possible.
We still have years of work ahead of us, but each successful experiment brings us closer to that dream. And every one of us on the Generation Bio team feels the same sense of excitement as we advance our programs steadily forward to the clinic. As I said to myself when I saw the dark wells on that monitor: Game On!