Easy-to-use CRISPR nuclease platforms have democratized genome editing, especially in the laboratory setting. But these platforms, as they exist today, are less precise than alternative genome editing systems, such as zinc finger nuclease (ZFN) systems and transcription activator-like effector nuclease (TALEN) systems. Indeed, the use of CRISPR platforms can result in off-target modifications, unwanted on-target modifications, and genomic rearrangements. Consequently, CRISPR platforms are being developed that will be more precise.
Even though CRISPR platforms are bound to improve, we shouldn’t be content to put all our genome editing eggs in one basket. We should be aware that the genome editing systems that preceded CRISPR—ZFN systems and TALEN systems, for example—are anything but static. They’ve been improving alongside CRISPR. Also, we should be aware of the newer systems that have been following CRISPR’s example.
Like CRISPR, these systems are adapted from naturally occurring systems. Whereas CRISPR originated in bacteria as a form of immune defense, a new system known as ARCUS consists of enzymes derived from I-CreI, a selfish genetic element that occurs in the algae Chlamydomonas reinhardtii. Other starting materials from nature include mobile genetic elements (MGEs), which have potential as “gene writing” tools that eschew double-strand breaks.
Yet another resource for the genome editing field is the family of Adenosine Deaminase Acting on RNA (ADAR) enzymes. ADAR is used to edit RNA rather than DNA, but like CRISPR and ARCUS, it has a natural origin—in this case, enzymes used by soft-bodied cephalopods to diversify protein expression, particularly in nervous system cells.
Genome editing systems of all kinds could stand improvement. For example, all need to support greater ease and specificity of delivery. Nonetheless, all—including the CRISPR-less systems—are demonstrating progress. For example, some therapies based on CRISPR-less systems are already in clinical trials.
Getting in touch with ZFNs
Using the ZFN platform, Sangamo Therapeutics was the first company to edit genes ex vivo and in vivo. The compact nature of the zinc finger binding domains provides flexibility for appending other molecules and expanding the platform, which today contains many forms of genome engineering tools including nucleases, base editors, and transcriptional repressors and activators.
“The ZFN platform is like a Swiss army knife,” says Sandy Macrae, MB, ChB, PhD, the CEO of Sangamo. “The bit that targets the DNA is the red handle, and to that you can attach a whole series of tools to match the right technology to the right disease.”
Libraries and archives of zinc finger molecules are extensively characterized. “If we decide to target a particular gene, we can take that sequence and design hundreds of proteins and rapidly generate those proteins for screening,” asserts Jason Fontenot, PhD, the senior vice president and CSO of Sangamo. “Then we can slightly alter the sequences to improve their specificity. That kind of tunability simply is not found in a lot of the other platforms.” A transition to in silico design industrializes the development process.
Sangamo is targeting the central nervous system (CNS) through its genome regulation programs and also developing chimeric antigen receptor (CAR)-engineered regulatory T cells (Tregs) that should address transplant rejection and autoimmune diseases such as multiple sclerosis and inflammatory bowel disease.
Genomic medicine is a combination of the genome engineering tool and the delivery method. Both have to be optimized and function together for effective therapies. The compact nature of zinc finger proteins makes them compatible with a range of delivery methods including adenovirus, adeno-associated virus (AAV), and lentivirus. An internal team exclusively focuses on AAV engineering to identify and develop new AAV capsids, especially those with enhanced tropism for cells of the CNS.
Through strategic partnerships, Sangamo can expand on its internal programs and familiarize itself with applications it might not have explored independently. Sangamo’s world-class team of structural and molecular biologists constantly innovate on the company’s platform, refining and elaborating on the technology.
“We are right at the cusp of solving delivery,” Macrae declares. “In 5 to 10 years, genomic medicines will depend upon how delivery has evolved. Editing will become the easiest part of the equation, and I hope people just talk about genome engineering and not the individual tools.”
Developing a talent for TALENs
“A genome editing technology must recognize a long stretch of DNA in a tight, precise way, and it must edit the DNA in a specific fashion,” says André Choulika, PhD, the CEO of Cellectis. “Technologies become obsolete. We are technology agnostic and use whatever can do the job best.”
Initially, Cellectis used naturally occurring meganucleases, but then the company started working with TALENs. “We compared TALENs to meganucleases,” recalls Choulika. “They worked amazingly and were very straightforward and easy to design.”
Each TALEN consists of a nonspecific DNA-cleaving nuclease fused to a DNA-binding domain that can be engineered for targeting. “TALENs recognize a long stretch of DNA and are ideal to induce precise and efficient gene repair/insertion [by homologous recombination] or gene inactivation [by knockout],” Choulika elaborates. “The RVDs [repeat-variable di-residues] recognize one base pair, and you can easily target any sequence in the genome.” (The TALEN® technology at Cellectis is exclusively licensed from the University of Minnesota and Iowa State University.)
Cellectis benchmarked CRISPR against other technologies in 2013 and found that the constraint posed by the protospacer adjacent motif sequence reduced precision. In addition, CRISPR makes it difficult to induce homologous recombination.
Since TALENs are vectorized using mRNA, Cellectis uses a very gentle electroporation technology—the company’s own PulseAgile technology—to deliver nuclease-encoding mRNA into T cells and hematopoietic stem cells, and to provide high yields. The use of nanoparticles potentially positions next-generation TALENs to edit genes in a wider range of cells.
Cellectis focuses on allogeneic therapies. Currently, autologous CAR T-cell therapies can address only certain malignancies. Autologous cells remain in the body for a long period of time, potentially resulting in side effects such as leukopenia. To develop allogeneic therapies, Cellectis edited out factors that could result in graft-versus-host disease.
“In allogeneic therapy, the cells are injected,” Choulika relates. “The cells attack the tumor, and then they are gone.” Allogeneic therapies can be expanded to address a wide range of disorders and are more broadly available to patients. Accordingly, these therapies are called universal CAR T or UCART therapies. Candidate UCART therapies are targeting CD19, CD123, CS1, CD22, and CD38 to address bone and blood marrow cancers.
Catching up with mobile genetic elements
According to Michael Holmes, PhD, CSO, Tessera Therapeutics, genome editing has been limited by the reliance on nucleases and the cellular DNA repair machinery. The error-prone DNA repair process can result in unintended mutations as well as increase the risk of driving genomic rearrangements.
To avoid the limitations of nuclease-dependent genome editing, Tessera creates synthetic MGEs. Tessera asserts that synthetic MGEs, like natural MGEs, can make “diverse alterations to the genome, both small and large, without breaking the genome or relying on DNA repair pathways.”
MGEs are the most abundant elements found in nature. Indeed, they make up to 50% of the DNA found in the human genome. Given their abundance and their efficiency in copying and writing themselves into the genome, Tessera is investigating the different classes of MGEs to see how they can be leveraged to improve genome editing.
“Using computational biology and digital systems, we scan for and identify tens of thousands of MGEs and then use high-throughput screening to pinpoint ones that would make good genetic medicines, where they have naturally evolved to function with a high degree of efficiency, specificity, and fidelity in human cells,” Holmes explains. “Machine learning, generative protein design, and protein engineering techniques are used to further engineer, enhance, and optimize the elements. This approach creates broad sets of platforms.”
DNA writers, such as transposons and recombinases, have evolved to use a DNA template. RNA writers, derived from retrotransposons, use a process called target-primed reverse transcription. These enzymes recognize and bind their RNA template and then target a specific genomic location and nick the DNA. The nicked DNA anneals to the RNA and then primes reverse transcription to copy the RNA into DNA and thereby replace the original genetic sequence with a new one.
RNA writers use RNA templates and can take advantage of nonviral delivery vehicles such as lipid nanoparticles. “For rewriting, we have engineered both the driver enzyme and the RNA template to insert specific and precise edits,” Holmes notes. “For example, we can change single nucleotides at defined locations or introduce insertions or deletions.”
RNA writers provide the potential for redosing and can be manufactured at large scale with reduced cost of goods. Tessera plans to have a GMP manufacturing facility in place by the end of 2022. “This broad platform is going to allow us to advance more programs than we could take forward internally,” Holmes relates. “It will create opportunities to have meaningful partnerships.”
ARCUS
Jeff Smith, PhD, and Derek Jantz, PhD, developed ARCUS while they were postdoctoral fellows at Duke University. ARCUS is based on a naturally occurring genome editing enzyme, I-CreI, that occurs in the algae Chlamydomonas reinhardtii. I-CreI can add DNA to a specific location in the genome via a double-strand break.
“We reengineer the natural I-CreI protein to direct it to new DNA sequences while preserving the properties of this natural system that make it ideal for therapeutic gene editing,” says Jantz, who is today the CSO of Precision BioSciences, a company he co-founded with Smith, the chief technology officer. “Only in the presence of its target DNA is an ARCUS nuclease designed to activate and perform its edit. Then the enzyme returns to its default inactive ‘off’ form.
“This allows us to express an ARCUS nuclease for an extended period of time without worrying about the accumulation of off-target edits. This is critical for therapeutic gene editing in organs other than the liver for which transient gene delivery vehicles like lipid nanoparticles are not an option.”
Long-term nonhuman primate data from Precision BioSciences’ preclinical familial hypercholesteremia program has demonstrated that ARCUS, after delivery via AAV, can achieve sustained reduction of low-density lipoprotein and stable low-frequency off-target editing. “We have demonstrated that ARCUS can target the addition of transgenes to a defined location in the genome in nonhuman primates with very high efficiency,” Jantz asserts. This is due to the unusual nature of the DNA break introduced by ARCUS in which highly recombinogenic 3¢ “sticky ends” are produced at the cut site. These DNA ends promote homology-directed repair to facilitate gene addition.
ARCUS recognizes its DNA target directly through interactions between the enzyme and the DNA bases. This gives it the ability to discriminate between DNA target sites that differ by as little as a single base pair.
“We know which amino acids are responsible for the enzyme’s catalytic efficiency, binding specificity, and binding affinity,” Jantz points out. “Those three parameters—efficiency, specificity, and affinity—can be fine-tuned independently of one another to optimize the nuclease for its particular function. Each ARCUS nuclease is highly customized for its intended use.”
Precision BioSciences has three allogeneic CAR T-cell therapeutic candidates in the clinic and expects to advance three in vivo gene editing programs to IND/CTA in the next three years. In partnership with Eli Lilly & Co., the company is also developing three in vivo gene editing candidates.
Korro Bio Wraps Its Arms around RNA Editing
Cephalopods have very high levels of plasticity in protein expression, mostly in nervous system cells. One of the ways they sustain this plasticity is through the use of Adenosine Deaminase Acting on RNA (ADAR) enzymes. Josh Rosenthal, PhD, a neurobiologist at the Marine Biological Laboratory and a co-founder of Korro Bio, discovered that ADAR could make an edit in mRNA using a chemical moiety. ADAR specifically catalyzes an adenosine to an inosine on RNA that gets read as a guanidine when translated to protein, providing transient base editing.
This RNA editing discovery in cephalopods led Korro Bio to see if it could co-opt the ADAR system to explore a new therapeutic approach. The company developed an RNA editing platform called Oligonucleotide Promoted Editing of RNA (OPERA).
The platform has four components. The first component is ADAR. The second is the ability to design and develop the drug product, which is a synthetic oligonucleotide that binds complementarily to the mRNA of choice and recruits ADAR to achieve the edit. The third is the ability to unravel the vast combinatorial space to identify targets. The fourth is “the ability to understand how proteins function,” says Ram Aiyar, PhD, CEO, Korro Bio. “An A-to-G edit translates to changing 12 different amino acid sequences on a protein structure.”
A-to-G changes are responsible for 30% of genetic diseases. Cases of chronic disease, especially in people over 40, do not necessarily warrant a permanent change to the DNA. For example, a slowly progressing, rare disease called α1-antitrypsin deficiency has a single point DNA mutation and is generally not detected until patients are over 20.
“The opportunity goes beyond rare diseases,” Aiyar states. “We believe that our transient approach will offer functional cures to prevalent chronic patient populations, and that patients will be able to take an RNA-editing drug, like any other chronic disease drug, for the rest of their lives.”
Aiyar believes in minimizing risk when moving toward the clinic. Approved antisense oligonucleotides have demonstrated the therapeutic profile of synthetic oligonucleotides. There is also precedent for the use of delivery components such as N-acetylgalactosamine in liver-mediated diseases.
“The way an oligo reacts with mRNA and ADAR is a very complex process,” Aiyar explains. “We layer together ADAR biology, chemistry, and data science to find the best compound for a specific target in a specific disease setting.
“ADAR is in every tissue type, but we are targeting liver and CNS-mediated diseases initially because delivery has been worked out. It takes a village to develop drugs. We want to partner with pharma companies, patient advocacy groups, and foundations to determine how to get these potential drugs to patients in an expedited fashion.”
