Poor ole’ Sisyphus. Everyday he was forced to roll that big boulder up the hill. And everyday, the boulder rolled back down – doomed to repeat the task for eternity. Even the ancient Greeks could recognize the inability to complete this was a weird form of torture.
For the last few centuries, we’ve accepted this same approach for our medicines. The vast majority of our therapies work by putting a band-aid on the problem and reducing the effects of the disease, but only if you remain on the drug indefinitely. These drugs don’t fix the underlying problem – much like Sisyphus rolling that boulder up the hill each day. Fortunately, there’s a new form of medicine that gives us superpowers – one that will keep that boulder at the top of the hill.
Cell and gene therapies are rapidly being adopted to recode the genetics of our own cells, endowing them with new abilities that fight or reverse disease. They work by inserting new genetic material into our cells and come with the promise of curing everything from rare diseases to cancer.
These genetic “payloads” can add a healthy copy of a gene if you were born with a mutated version. For example, a mutation in the RPE65 gene leads to blindness but adding a non-mutated version to retina cells with gene therapy restores sight. In other cases, these payloads give cells a gene not found in nature. T cells, for example, can be pulled from a cancer patient and engineered with a chimeric antigen receptor (CAR) gene. These cells are then reintroduced to the patient as a cell therapy where the CAR directs the T cells to hunt and kill cancer.
To get these payloads into cells, we have repurposed nature’s best genetic engineers – viruses. Through decades of research, we have improved upon these microscopic gene delivery vectors, making them safe for use in humans, engineering them to target specific tissues in the body, and reworking them so that we choose what genetic payload they deliver.
The issue is that we haven’t gotten much better at producing a large quantity of these viral vectors. To make these delivery vehicles currently, we take human cell cultures and shove DNA in them, giving them the instruction manual on how to produce our vector. Over the course of a few days, the cells secrete these vectors into the culture media, where it can be isolated and used.
Unfortunately, the cells’ innate wiring make scaling-up this process extremely difficult – we’re asking the cells to go against eons of anti-viral evolutionary programming to make parasitic viruses for our benefit. Because cells are ill-prepared to make viruses, we end up with low yield. Often, this is simply because the cells throw in the towel before they make enough. Often, though, we simply didn’t give the cells the optimal instructions, leading to the creation of malformed and non-functional vectors.
This means manufacturing of cell and gene therapies is extremely bottlenecked. Use of these inefficient cells creates two manufacturing problems: capacity and cost.
First, manufacturing capacity with current paradigms is insufficient to cure even diseases that affect only a handful of people. To treat an ultra-rare disease that affects only 0.001% of people, we would need to double our current capacity. To treat a rare disease – one that affects less than 200,000 people – we would need to 200X our current manufacturing capacity.
In other words, our capacity to make vectors for cell and gene therapies is orders of magnitudes lower than it needs to be to help even a small subset of patients. To make matters worse, the problem can’t be solved by outsourcing to specialized manufacturers – every manufacturer uses the same inefficient cell lines and non-optimized instructions. So with a growing market, manufacturers have a 2-year waitlist.
But even if we could produce enough vector, manufacturing cost is prohibitively expensive. That's why gene and cell therapies currently cost people hundreds of thousands if not millions of dollars. To manufacture enough viral vector for more common diseases like diabetes with current efficiencies, costs would exceed the entire U.S. government spend in 2020!
Thankfully, the yield problem can be solved with synthetic biology. It’s well established that randomly or rationally mutating either the vector-producing cells or the vector instruction manuals can optimize yield. Doing so, though, forces the vector-producing cells to do something they’re wired to avoid, so helpful mutations are exceedingly rare. To make matters worse, a cell will likely need multiple mutations to optimize production of a single therapy.
This blows up the space that needs to be checked to find the perfect cell line. Conservatively, if we need 7 mutations in a single cell, that's 1,030 possibilities to screen for a winner! Checking for these mutations one at a time would be impossible given the size of the space. Even worse, each new therapy would require re-running this process.
Enter 64x who have built a first-of-its-kind solution to overcome this problem. Realizing the intractability of searching this space with traditional methods, 64x generated a new approach that enables them to screen millions of optimizations in a single culture. Instead of editing, growing, and measuring cell-factories one at a time, 64x makes millions of edits to a pool of cells, manufactures vectors, and uses sequencing and computation to pinpoint which edit was optimal.
Yet even at this vastly improved throughput, it would still take a long time to empirically measure the whole space of 1,030 possibilities. If it took just one second to check each possibility, it would take longer than the entire history of the universe. So, 64x takes their proprietary dataset – the largest amount of edit-to-yield information ever coalesced and feeds it into CellMap, a model that whittles down the enormous possibility space into a subset that can actually be screened.
With this, Dr. Lex Rovner, co-founder and CEO, leverages her years spent in Professor George Church’s lab at Harvard University pioneering the use of recoded cells to give them better functionality. At 64x, she combined integrated DNA library synthesis, next generation sequencing, and machine learning to create VectorSelect – an ultra-high-throughput approach that feeds 64x’s AI, enabling them to rationally design optimal vector manufacturing cell lines.
Using VectorSelect, millions of cell or instruction manual variants create vectors in a single reactor. The key innovation involves a proprietary genetic barcoding method that connects information on viral vector productivity back to the parent cell which produced it, enabling massively parallelized genetic screens of millions of candidate production cell lines simultaneously – a feat which has been impossible until now.
There are a lot of advantages:
VectorSelect enables millions of variants to be tested in a single assay relative to traditional methods. By doing everything at once in a pooled format instead of checking each variant one at a time in a multi-welled plate, 64x can screen orders of magnitude more variants.
Every VectorSelect assay creates data on each variant’s specific yield given a particular payload. This generates valuable information that is fed into computational models, refining 64x’s variant libraries and creating higher and higher yields for their customers.
The VectorSelect process is run at scale in a bioreactor, meaning 64x is finding optimal variants under the same conditions in which the vector is actually produced. This is in stark contrast to current standards which are checked in a much smaller multi-welled plate format and provide no data on whether the cells or instruction manual will scale well.
VectorSelect is designed to work with any vector generation process. Whether you’re working with the tried and true vectors generally used in the field, a weird subtype that has a particularly advantageous property for your therapy, or even a niche vector that is only beginning to be used, 64x can adapt to make sure the maximum vector is generated.
VectorSelect only captures well formed, fully functional vectors. Other methods optimize on the production of vector particles, meaning they count malformed and non-functional vectors, leading to a lower quality cell or gene therapy. 64x only takes into consideration those particles that are operational and ready to provide benefit to a patient.
64x will be at the forefront of the gene and cell therapy explosion, enabling the viral vectors that underpin them to quickly and cheaply scale-up. They will not only enable cures for ultra-rare diseases, but also for diseases with huge populations like those related to aging. Even more excitingly, they will massively enable the design-build-test cycle of future vector-based therapies that have shown promise in redesigning immune cells, exploding tumors, and vaccinating against infectious diseases.
At Fifty Years, our sweet spot is supporting founders at the earliest stages building deep tech companies that can generate huge financial outcomes and create massive positive impact.
Deep tech: Lex has leveraged her enormous synthetic biology skills and deep understanding of how to recode organisms to make a proprietary platform that optimizes virus production faster, cheaper, and easier than any one else can.
$1B yearly revenue potential: Cell and gene therapies are exploding into medicine and have the potential to touch every disease we could imagine. 64x is poised to become the de facto scale-up company that enables these therapies to move from bench to bedside.
Massive positive societal impact: Therapies that rely on viral vectors have huge value but are only able to touch the lives of very few people. 64x’s platform technology not only allows more people to be treated with the medicines we have, but also opens the possibilities for novel ones to be created that are currently too difficult to even imagine.
Inspired by their technology and vision for its future, Fifty Years was excited to co-lead 64x’s seed round with our friends at First Round Capital. We’re grateful to deepen the partnership in their $55M Series A, led by Lifeforce Capital, with support from Northpond Ventures, First Round Capital, Alix Ventures, and Refactor Capital. 64x is a model for how brilliant inventors can move their work from academia to industry and completely transform a niche into the de facto standard; a narrative that inspires us here at Fifty Years.