CRISPR Genome Editing: Into the Second Decade

Abstract

GEN BiotechnologyVol. 1, No. 1 On The MoneyFree AccessCRISPR Genome Editing: Into the Second DecadeGeulah LivshitsGeulah Livshits*Address correspondence to: Geulah Livshits, Chardan, 17 State Street, New York, NY 10004, USA, E-mail Address: glivshits@chardan.comEquity Research Department, Chardan, New York, New York, USA.Search for more papers by this authorPublished Online:16 Feb 2022https://doi.org/10.1089/genbio.2022.29015.gliAboutSectionsPDF/EPUB ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail Approaching the 10th anniversary of the CRISPR “genetic scissors,” what are the successes, prospects and challenges facing clinical genome editing in 2022 and beyond?2a022 marks 10 years since the initial publications characterizing Cas9 as a programmable RNA-guided endonuclease.1,2 These findings have led to the emergence of powerful CRISPR-based platforms enabling gene disruption, insertion, correction, and deletion. Accordingly, for the past decade, genome editing has moved through multiple inflection points as CRISPR transitioned from basic science to near-ubiquitous research tool, and into a potentially transformative therapeutic modality.The astonishing rise of CRISPR has been accompanied by the establishment of a stable of biotech companies aiming to leverage the technology for therapeutic applications. We have started to see that therapeutic promise materializes with clinical proof of concept emerging in indications, including sickle-cell disease (SCD),3 a form of hereditary amyloidosis,4 and cancer.5 Publicly traded companies focused on therapeutic genome editing are now collectively valued at around $30 billion, and many big pharma companies have at least one collaboration related to genome editing, underscoring the broad commercial interest.Even as the first wave of programs moves forward through clinical studies, the technology side continues to advance at a rapid clip, with advances in activity, specificity, and the incorporation of novel functions positioned to further expand the “CRISPR toolbox” and the reach of therapeutic genome editing. Although various open questions remain regarding safety and activity, the outlook for the field is excellent. CRISPR technology is on track to become a pillar of cell and gene therapies in the coming decades.Platform PotencyWhat makes this gene-editing platform so powerful? At their core, CRISPR editing systems enable genetic manipulation of specific DNA sequences. This concept in and of itself is not new: earlier genome editing approaches, including meganucleases, zinc finger nucleases, and transcription activator-like effector nucleases (TALENs), laid the groundwork here.6 However, these other technologies rely on protein–DNA interactions to target enzymes to genomic sites, making them more complex to design. By contrast, Cas9 can be targeted to specific sites in the genome by programming its guide RNA sequence. This attribute has made CRISPR easier to design, easier to optimize through high-throughput screening, and more amenable to multiplex editing, contributing to its widespread adoption.Mechanistically, Cas9 triggers a double-strand break (DSB) in the target DNA sequence that is repaired by cellular pathways. How the break is repaired can dictate the editing outcome: the more efficient nonhomologous end joining (NHEJ) pathway typically introduces small insertions/deletions (indels); homology-directed repair (HDR) enables a more precise edit through use of a DNA template but is generally less efficient and largely limited to dividing cells. Gene disruption through NHEJ repair is often used to achieve permanent gene knockdown, whereas the HDR method can be used for targeted insertion, for example, of chimeric antigen receptor (CAR) constructs in allogeneic CAR-T cells.The past few years have yielded numerous insights into the “rules” governing guide RNA activity, empirical and computational methods to assess specificity, and the molecular mechanisms driving editing outcomes. These insights are being exploited in therapeutic settings: for example, therapeutic T cell editing programs typically show >90% disruption and >50% targeted insertion efficiencies.Into the ClinicCRISPR programs have been able to move quickly in the therapeutic arena by building on technological advances across other genetic medicine modalities, including gene therapy, RNA interference, and earlier genome editing platforms. In addition to the type of edit, therapeutic genome editing applications can be classified by where the editing takes place: in ex vivo editing cells edited in the laboratory are subsequently delivered into the patient, whereas in in vivo editing the editing takes place inside the body.Initial applications of ex vivo editing have largely concentrated in settings with existing cell transplantation paradigms (such as the hematopoietic system) and where clinical proof of concept has already been established using other engineered cell therapy approaches. For example, in SCD, a lentiviral gene-modified autologous hematopoietic stem cell (HSC) program demonstrated human engraftment and modification of disease phenotypes,7 paving the way for a gene-edited program that has similarly shown durable engraftment and phenotype modulation.3 Although the strategies differ in molecular details—the lentiviral program delivers a functional copy of the hemoglobin gene, whereas the gene editing strategy disrupts a regulator site to activate the fetal hemoglobin gene—the availability of good manufacturing practice workflows for HSC collection, culture, and infusion reduce the variables required for optimization.Getting on base. Base editing is one of the precision genome editing technologies poised to enter the clinic in 2022. (Left) Researchers at Beam Therapeutics reported progress in correcting the sickle-cell disease point mutation in 2021 in The CRISPR Journal. (Right) Susie Suh (Case Western Reserve) is applying base editing to correct a hereditary form of blindness.One might well ask: why bother with a gene editing approach if lenti works? In short, the former may have advantages with respect to costs and safety. Viral vectors are a cost driver for genetic medicines, which could lead to challenges relating to patient access or reimbursement; ex vivo genome editing approaches in the pipeline tend to use electroporation. On the safety front, lentiviral vectors integrate semirandomly into the genome. In theory, this runs the risk of insertional mutagenesis and cancer. Although the overall risk may be relatively low, some instances of myelodysplastic syndrome have been reported in lentiviral HSC programs, suggesting the risk is not purely theoretical.In principle, CRISPR approaches circumvent this issue through site-specific editing, but may carry their own risks as discussed below. In oncology, gene editing similarly builds on established cell-therapy strategies, including CAR-T and CAR-NK therapies. In this space, genome editing has multiple uses, including targeted insertion of the CAR and other functional constructs (instead of lentiviral delivery), as well as disruption of select cell surface markers to enable off-the-shelf dosing by preventing immune rejection.5 Although several groups have demonstrated efficient editing from a technical standpoint, the oncology cell therapy field remains less mature with respect to biology: parameters such as optimal cell source, culture conditions, and which combinations of edits enable sufficient allogeneic engraftment/activity remain open questions, and will likely be areas of active research in the coming years, particularly as the field aims to expand from blood cancers into solid tumors.For in vivo genome editing achieving efficient delivery to the target cell type is the key challenge. The two most advanced methods for in vivo nucleic acid delivery are lipid nanoparticles (LNPs) and adeno-associated virus (AAV) vectors. Although LNPs have garnered recent attention for their use in COVID-19 vaccines, they are also used in the first approved RNAi therapy, Onpattro, which treats transthyretin amyloidosis by silencing the TTR gene in the liver.8An analogous in vivo gene editing program reduces biology risk by going after the same genetic target, but with an LNP-delivered CRISPR disruption strategy that would enable a one-time therapy rather than the repeat dosing required with Onpattro. In June 2021, initial clinical data from the editing program showed clean safety and 87% TTR reduction, indicating potent activity and good translation from preclinical models into human studies.4 LNPs have thus emerged as a favored strategy for liver-directed delivery of mRNAs encoding CRISPR components, as they are also less complex and cheaper to produce compared with viral vectors. The LNP-mRNA strategy is attractive for genome editing applications as long-term nuclease expression is not needed once the edit has happened and the components are rapidly broken down. Delivery to other tissues, including lung, the hematopoietic system, and specific immune cell types, is area of active research but is presently less mature.AAV vectors have demonstrated clinical proof of concept for delivery to tissues, including the eye, central nervous system, liver, and muscle. Different AAV capsids have distinct tissue tropism patterns, and target-cell expression can be further refined using tissue-specific promoters. The constrained AAV packaging limit of 4.7 kb necessitates use of smaller CRISPR-Cas variants than the widely used SpCas9, and several companies are advancing smaller enzymes for in vivo applications.In addition to further refinement of AAV and LNP platforms for optimal targeting of various tissues, other technologies are being explored, including alternative viral vectors and virus-like particles (see the “Views & News” in this issue covering recent study in David Liu's Lab by Ross Wilson).9 For both ex vivo and in vivo editing applications, I expect collaborations with developers of complementary or enabling technologies, such as cell processing or in vivo delivery, to drive the field forward for the coming years.Safety Concerns and Balancing Risk/BenefitAlthough targeted genome editing reduces the risk of mutational mutagenesis, each genome modification technology carries its own risks that must be evaluated as part of the overall risk/benefit assessment for a particular indication, such that the right tool can be selected for the job. For CRISPR, these largely fall into two categories: off-target editing and unwanted on-target editing outcomes.In vivo editing carries additional risks associated with uptake by unintended tissues, including the germline. For the past several years, extensive efforts to characterize off-target editing at predicted sites and in unbiased genome-wide scale assays have now largely been incorporated into guide-RNA screening workflows. Although this does not entirely eliminate the risk of off-target edits, it reduces the risk of surprises, and validated off-target editing sites are typically analyzed with respect for potential functional impact.Regarding on-target editing, DSB repair is by nature variable. Risks associated with DSBs were extensively discussed even before the advent of CRISPR, and include large deletions, translocations, and other chromosomal rearrangements. Indeed, TALEN-edited CAR-T products initially showed detectable translocations between two target sites, and similar findings have been reported in other programs, although the clinical impact of these translocations is not yet clear. For patients with late-stage cancer and few therapeutic options, the risk/benefit may still favor treatment with an agent with known translocations or genomic variability, especially given that many widely used cancer therapies, including chemotherapy drugs and radiation, can also cause mutations in the DNA. Ongoing bioinformatics and machine-learning efforts are increasing the field's understanding of DSB repair outcome patterns, which can further aid in guide/target selection.Second- and Third-Gen ApproachesThe CRISPR toolbox is expanding along several exciting dimensions. First, we are seeing characterization of new Cas and Cas-like enzymes from metagenomic analyses10; these can expand DNA-targeting capabilities and can include smaller enzymes amenable to AAV delivery. Second, we are seeing engineering of existing enzymes through rational design and directed evolution to enhance therapeutic properties, including potency and specificity or to change DNA targeting/editing preferences.11 And third (but definitely not least), we are seeing extensive engineering of novel functionalities through enzyme fusions that leverage the programmable DNA targeting activity of Cas proteins but achieve a distinct effector function.12Each of these next-gen platforms harbors its own advantages but will also require its own characterization of specificity and unintended editing outcomes. One of the more advanced examples of this is base editing, which fuses a Cas protein to a DNA deaminase, resulting in enzymes that can directly introduce 4 out of the 12 possible single-base mutations without the need for DSBs or a DNA template.13 Base editors can be used to directly convert amenable mutations and to introduce stop codons, as another strategy for permanent gene silencing. Initially published in 2016,14 base editors are expected to enter the clinic in 2022 (in both in vivo and ex vivo settings). Although base editors are less prone to DSBs, they can elicit “bystander editing” (of A or C bases within the targeting window), as well as a low level of target-independent deamination. As with DSBs, the potential clinical impact of these is not yet clear, and likely varies greatly by target sequence and location. Engineering efforts in recent years have aimed to generate base editors with reduced unintended editing as well as ones with expanded targeting windows.15Even as base editing nears the clinic, prime editing is waiting in the wings. Also developed in the Liu Lab, prime editing can directly introduce all 12 types of point mutations as well as “write” small insertions and deletions at the targeted location.16 The approach leverages a Cas9 fused to an engineered reverse transcriptase enzyme, and similar to base editing, acts without the need for DSBs. Although bystander editing does not appear to be an issue for prime editing, the technology does introduce indel mutations at some frequency and reported editing efficiencies vary across cell types.As large fusion proteins, both prime and base editors would have additional constraints for in vivo delivery. Importantly, neither prime nor base editing offer a means to introduce a gene-size insert. From a therapeutic standpoint, this means the technologies are better suited to address genetic diseases where many patients harbor the same mutation (as a separate therapy would likely be needed for distinct mutations, adding logistical complexity) or in settings that can be addressed by disruption of a gene, regulatory element, or splice site. Efforts to address this gap are in the works, including fusion of CRISPR proteins to other DNA modifying enzymes including transposases, as well as a combination of prime editing with integrase technology to enable transgene insertion without DSBs.17 These efforts are in their infancy, and will likely require iterative optimization to achieve an activity, specificity, and deliverability profile suitable for therapeutic use, but offer yet another exciting avenue for genome editing for the coming decade.Looking beyond permanent DNA modification, CRISPR technologies are being leveraged for RNA editing and epigenome editing (control of gene expression). Such approaches may make sense for settings where the risk/benefit assessment does not favor permanent genome modification, in settings where AAV-based delivery of an editor enzyme to nondividing cells, where expression is expected to be durable, or in platforms where redosing is feasible.ConclusionCollectively, the expanding genome editing toolkit, in conjunction with advances in cell processing, gene delivery, and genomic analysis, is positioned to open numerous avenues for therapeutic development for the coming years. Further advances in understanding of editing mechanisms can expand applications from cancer and genetic disease to larger and more complex indications. At present, there is not a one-size-fits-all strategy. As the technology continues to mature, it will be important for therapeutic developers to exercise discipline in deploying each subset of the technology to indications with an appropriate benefit risk profile and unmet need.Author Disclosure StatementThe author currently works for Chardan Capital Markets (“Chardan”), which is a boutique investment bank focused on identifying companies that will generate exceptionally high long-term investment returns by creating shared value for society. The opinions in this report do not necessarily reflect the opinions of Chardan and are subject to change without notice, including opinions on company valuation methods.Geulah Livshits, PhD,is a senior research analyst with Chardan covering the gene editing/cell therapy field. She holds a PhD from Rockefeller University and conducted postdoctoral research in epigenetics and CRISPR at Memorial Sloan Kettering Cancer Center before joining Chardan. Contact: glivshits@chardan.com.