The Medicines of Tomorrow: A primer on the gene editing landscape
CRISPR gene editing offers the potential to permanently treat genetic diseases with just a single dose of medicine. For this reason, CRISPR technology has received widespread attention from life scientists, biopharma companies, and the public in the nine years since its first reported use in editing human DNA. This concept of “one and done” has revolutionized not only how we think about treating disease; it has also revolutionized how therapies for many diseases are now being discovered and developed.
DNA is the blueprint of life. Genes are individual segments of DNA that contain the instructions for making the proteins and other molecules the cells in our bodies need to grow, survive and function. The complete set of genes in our bodies is called the genome.
Many genetic diseases are caused by inherited mutations in a single gene that render the corresponding protein non-functional or otherwise defective. Depending on the gene and its role in the human body, a person who inherits a mutant copy of the gene from one parent or both parents will develop the disease.
On a general level, many people grasp the potential of CRISPR-based gene editing to treat genetic diseases: it targets the mutated segment of DNA within the disease-related gene to “correct” it. But many people, including those in the general public, are keen to know more: how CRISPR works, which diseases it can and can’t treat, and what its future opportunities and challenges are.
We’ll take a look at all of these topics in this “primer”, which is the first article in a series about the gene editing landscape. In later articles, I’ll discuss how Life Edit’s gene editing technology fits into that landscape, and how our team is developing that technology into the medicines of tomorrow.
Before gene editing
CRISPR’s recent media attention may create the impression that gene editing is a brand-new idea, but it is not. The concept of treating a genetic disease by counteracting the underlying problem in the genome has a history that began with gene therapy, long before gene editing came along.
Traditional gene therapy medicines remedy the lack of functional protein by treating the patient with a normal copy of the gene. This is achieved by packaging the gene into a delivery vehicle, typically an adeno-associated viral (AAV) vector, which enters the cells most involved in the disease and then delivers the gene inside them. The normal gene remains in those cells for some period of time, treating the disease by expressing a functional version of the protein.
While it may seem straightforward, gene therapy faces several challenges.
Many genes are too large to fit into AAV vectors. Some diseases, such as muscular dystrophy, might be treatable with a truncated gene that produces a shortened, but still functional, form of the protein. But many other genetic diseases require delivery of a full-length gene for the therapy to be effective.
Additionally, because the therapeutic gene is not incorporated into the patient’s genome, its effects, while possibly long-lasting, have not yet proven to be permanent. Yet redosing with the gene therapy may not be possible, because the first dose can induce an immune response to the AAV vector that would prohibit a second dose of the therapy from effectively reaching the target cells.
Gene editing therapy aims to overcome the limitations of gene therapy by making a permanent change in the patient’s genome, so that only one dose is required to treat the disease. Depending on the genetic problem driving the disease, gene editing has the potential to remove, add, or alter DNA, which offers greater flexibility than gene therapy (which can only add) and the possibility of delivering treatments for diseases where gene therapy isn’t an option.
One example is spinocerebellar ataxia 1 (SCA1) where a mutant version of the ataxin 1 gene (ATXN1) contains an abnormally high number of repeated sections (called CAG repeats) in the DNA code and produces a mutant ATXN1 protein that damages several types of neurons. This problem can’t be addressed by adding a normal copy of ATXN1 with gene therapy, but may be improved by removing the excess CAG repeats with gene editing.
How gene editing works: Fingers, TALENs and nucleases
All gene editing technologies require two essential components: an enzyme capable of cutting DNA (a nuclease); and a method that directs the enzyme to the target sequence in the genome. As with gene therapy, these components are typically packaged into AAV vectors and delivered to the cells most involved in the disease. Once inside the cells, the nuclease finds the target DNA sequence, binds to it, and cuts both strands of DNA. Next, natural mechanisms found in all cells repair the DNA break, leaving behind the edited gene.
Three types of gene editing technology arose before CRISPR. One type utilizes a class of artificial enzymes called zinc finger nucleases (ZFNs), which were discovered in the 1980s and moved into therapeutic development in the early 2000s , with several products now in clinical trials. A second type is based on a class of naturally occurring enzymes called meganucleases (also known as homing endonucleases), which were discovered in the 1990s; they have since been tested in non-human primates , and the first meganuclease-based gene therapy entered clinical testing in 2019 . A third type also uses a class of engineered enzymes called transcription activator-like effector nucleases (TALENs); these were first described in the scientific literature in 2010 and quickly moved into clinical testing as well .
A strength of both ZFNs and TALENs is that they can edit any site in the genome. But one drawback is that these nucleases must be guided by proteins that specifically target the site. Designing and engineering such proteins involves a great deal of upstream R&D and yields a protein that can target only one place in the genome. Meganucleases have similar drawbacks: each naturally occurring enzyme targets a single genomic site. However, because the likelihood of finding a natural meganuclease specific for a given site is quite low, meganucleases for therapeutic applications are typically engineered, which also requires upstream R&D.
The advent of CRISPR in 2012 transformed the gene editing landscape by streamlining the method to direct the nuclease to a target sequence in the genome. Instead of being directed to the target sequence by a protein, CRISPR is directed by a guide RNA molecule. These guide molecules can be easily designed and generated for each target and paired with the same nuclease protein instead of relying on engineering new ZFN or TALEN proteins or a meganuclease for every target.
The best-known CRISPR system uses an RNA-guided nuclease (RGN) called CRISPR-associated protein 9 (Cas9), which is derived from Staphylococcus pyogenes bacteria (SpCas9). Instead of requiring a guide protein, Cas9 relies on a short RNA molecule called a guide RNA (gRNA), whose sequence is complementary to the target site in the genome. The gRNA must be designed specifically to target that site, but it’s much easier to develop a gRNA than to engineer a guide protein.
This type of guide molecule gives CRISPR a modular aspect: the same RGN can be targeted to different sites in the genome simply by swapping one gRNA for another.
That said, a single RGN can’t target everywhere in the genome; it can only bind DNA segments that are located close to a short nucleotide sequence called a protospacer adjacent motif (PAM).
The therapeutic potential of CRISPR
The ease of design and genomic range of CRISPR raises many interesting questions: What can it do therapeutically? What diseases could it treat and how are target genes chosen? How many copies of the gene must be edited to achieve a permanent fix – all of them?
As I mentioned above, genetic diseases caused by a mutation in a single gene, such as SCA1, are starting points for gene editing. CRISPR-based therapies would cut out the small segment of the gene where the mutation occurs.
Medical doctors and researchers continue to investigate the genes that cause diseases and mapping the specific mutations involved, in order to figure out which part of the gene to fix. But CRISPR can also target non-mutant genes that play roles in disease. A good example is the PCSK9 gene in hypercholesterolemia.
The PCSK9 protein controls the number of low-density lipoprotein (LDL) receptors on cell surfaces that regulate blood cholesterol levels. Studies have shown that people with naturally occurring loss-of-function mutations in the PCSK9 gene are otherwise healthy and have higher levels of LDL receptors, with resulting lower blood cholesterol levels. As a way to treat and reduce high cholesterol, CRISPR gene editing can be targeted to cut the PCSK9 gene to create a loss-of-function mutation. This underscores the importance of ongoing investigations into basic human biology to identify potential genetic targets that could be edited to treat diseases.
Notably, gene editing doesn’t need to correct every copy of the gene in a patient’s body to be effective. Rather, only the cell types and tissues where that gene plays a major functional role need to be targeted, such as muscle cells for ATXN1 and liver cells for PCSK9.
That brings us to another point: gene editing machinery is just one part of the equation; delivering the gene editing machinery to the target cells is the other part.
Engineered AAV vectors are currently the delivery workhorses for gene editing because AAV has many subtypes (serotypes), each of which targets a different tissue type, making them good choices for delivering many gene editing therapies. However, because wild-type (naturally occurring) AAVs can infect humans, many people have some level of pre-existing immunity to AAVs that could compromise the efficacy of a gene therapy delivered with an AAV vector. Additionally, while CRISPR packages are generally smaller than the full-length genes used in gene therapy, not all CRISPR packages will fit into AAV vectors. Because AAVs have these potential limitations, other delivery vehicles, such as lipid nanoparticles (LNPs), are under investigation to expand the toolbox of delivery options.
What’s next: Questions, challenges and avenues of exploration
In addition to new delivery methods, there are several aspects of gene editing we need to investigate in order to fully tap its potential.
A major concern is its ability to make off-target edits – that is, to alter one or more sites in the genome apart from the intended target. The technology’s success will depend on understanding where those off-target edits could occur and the ramifications they could have. For example, is the off-target gene critical to cell survival? Will editing the off-target gene lead to another disease, such as cancer? Of course, it will also be critical to design gene editing systems that have little or no potential to make off-target edits.
There are also questions about how “permanent” a given edit is. For example, because SpCas9 is derived from bacteria that infect humans, patients may have pre-existing immunity to the nuclease. Similarly, patients may also have pre-existing immunity to the AAV vectors, as noted above. If AAV remains in the individual’s system expressing SpCas9 for too long, the vector and/or the nuclease might be recognized and eliminated by the immune system before the gene editing process is complete, or edited cells that have prolonged AAV and/or SpCas9 expression may be cleared by the immune response.
But there are exciting new areas to explore as well. An expansion of potential genome targets for CRISPR editing is possible with the discovery of new nucleases that have different PAM requirements. Also, new editing strategies such as altering a single nucleotide in the genome for correction of disease-causing point mutations, is enabled with development of new gene editing technologies, such as base editors.
Life Edit is delving into all aspects of gene editing, from new nucleases and base editors, to delivery methods and off-target edit detection assays. In my next article, we’ll take a closer look at Life Edit’s technology, its unique potential, and why I am excited to be working on it.