Beyond Cas9: The New CRISPR Editors Reshaping In Vivo Gene Therapy in 2026

For more than a decade, CRISPR-Cas9 has been the undisputed workhorse of gene editing. It is precise, programmable, and remarkably versatile. But there is an uncomfortable truth that the field has been dancing around since the technology first moved toward the clinic: Cas9 is simply too big to deliver where it matters most.

Imagine trying to mail a basketball through a letter slot. That is, in essence, the delivery problem facing every team that wants to use SpCas9 for in vivo gene therapy — editing genes directly inside a patient's body rather than in a dish. The enzyme clocks in at 1,368 amino acids, and when you add a guide RNA and the regulatory sequences needed to express the system inside human cells, the total genetic payload exceeds what a single adeno-associated virus (AAV) vector can carry.

AAV, the gold-standard delivery vehicle for targeted in vivo therapies, has a strict packaging limit of roughly 4.7 kilobases (kb). SpCas9's coding sequence alone runs to about 4.2 kb — leaving almost no room for the promoter, guide RNA cassette, or polyadenylation signal that every functional construct needs.

This single constraint has kept the overwhelming majority of approved CRISPR therapies confined to ex vivo approaches, where cells are removed from the patient, edited in the laboratory, and then transplanted back. That works brilliantly for blood disorders like sickle cell disease. It does not work for the liver, the brain, the eye, or the muscle — tissues you cannot simply scoop out, edit, and return.

In 2026, that bottleneck is finally breaking. A new generation of compact CRISPR editors — naturally smaller enzymes discovered through systematic mining of microbial genomes, then engineered for human-cell performance — is arriving with the right combination of small size and high efficiency. From the NIH-funded Al3Cas12f RKK variant achieving up to 90% editing in human cells, to the radically different CRISPR-Cas3 system that shreds target DNA with zero detectable off-target indels, the toolbox is expanding fast. Here is what you need to know.

The Size Problem: Why Cas9 Can't Go Everywhere

The AAV bottleneck

Adeno-associated viruses are small, non-pathogenic viruses that have become the delivery vehicle of choice for in vivo gene therapy. They are well-tolerated by the immune system (relative to other viral vectors), they can be engineered to target specific tissues through different serotypes (AAV8 for liver, AAV9 for the central nervous system, AAVrh10 for muscle, and so on), and several AAV-based gene therapies have already received regulatory approval — including Luxturna for inherited retinal dystrophy and Zolgensma for spinal muscular atrophy.

But AAV has a hard packaging limit. The single-stranded DNA genome that AAV can carry tops out at approximately 4.7 kb. That is a firm physical constraint; overstuffing the capsid leads to truncated genomes, reduced titers, and dramatically lower transduction efficiency.

SpCas9 — the Streptococcus pyogenes Cas9 that dominates the field — has a coding sequence of roughly 4.2 kb. By the time you add a compact promoter (~0.2-0.5 kb), a guide RNA expression cassette (~0.4 kb), and a polyadenylation signal (~0.2 kb), you are well over 5 kb. The basketball does not fit through the letter slot.

Current workarounds and their costs

The field has not been idle. Several strategies have been developed to get around the size problem, but each comes with significant trade-offs:

• Dual-AAV systems: Split the Cas9 gene across two separate AAV vectors that must both infect the same cell. The two halves are then reassembled using split-intein protein splicing or overlapping recombination sequences. This works in principle, but in practice it cuts efficiency dramatically — both vectors need to reach the same cell, and reconstitution is never 100%. Manufacturing costs also double.

• Smaller Cas9 orthologs: Enzymes like Staphylococcus aureus Cas9 (SaCas9, 1,053 aa) and Campylobacter jejuni Cas9 (CjCas9, 984 aa) are smaller, but they come with more restrictive PAM requirements (the short DNA sequence adjacent to the target that the enzyme needs to recognize). SaCas9 requires an NNGRRT PAM, limiting the number of targetable sites in the genome.

• Lipid nanoparticles (LNPs): Deliver Cas9 as mRNA or ribonucleoprotein rather than DNA. LNPs have no strict size limit, but they overwhelmingly accumulate in the liver after systemic administration. Targeting other tissues with LNPs remains a major unsolved challenge, and repeated dosing can trigger immune responses against the lipid components.

• Virus-like particles (VLPs): Package Cas9 protein directly into retrovirus-derived particles. Still early-stage, with questions about scalability and tissue targeting.

Each workaround solves part of the problem while introducing new ones. The cleanest solution has always been obvious: find or engineer a CRISPR effector that is small enough to fit inside a single AAV alongside all the regulatory elements it needs.

Why the search took so long

Nature has been running CRISPR experiments for billions of years. Bacteria and archaea use CRISPR systems as adaptive immune defenses against phage infection, and the diversity of these systems is staggering. But for years, the field focused almost exclusively on Type II systems (Cas9) because they were the first to be characterized and the easiest to program.

It was not until researchers began systematically mining metagenomic databases — sequencing microbial communities from hot springs, ocean sediments, animal guts, and other extreme environments — that the true diversity of CRISPR effectors became apparent. Type V systems (Cas12 family), Type I systems (Cas3), and other exotic variants were hiding in plain sight, waiting to be discovered, characterized, and engineered.

Al3Cas12f RKK: The Tiny Editor With 90% Efficiency

Discovery and engineering

In April 2026, researchers funded by the National Institutes of Health (NIH) published results on what may be the most significant compact editor to date: an engineered variant of the naturally occurring Cas12f enzyme designated Al3Cas12f RKK.

Cas12f (formerly known as Cas14) enzymes are among the smallest CRISPR effectors ever identified. Where SpCas9 weighs in at 1,368 amino acids, Cas12f enzymes range from roughly 400 to 600 amino acids — less than half the size. Their coding sequences run to approximately 1.2-1.8 kb, leaving ample room inside a single AAV for promoters, guide RNAs, and even additional functional elements like fluorescent reporters or regulatory switches.

The challenge with wild-type Cas12f enzymes has always been efficiency. Earlier characterizations of natural Cas12f variants showed editing rates in human cells that were often below 10% — far too low for therapeutic applications. The enzymes had evolved to work inside bacteria, not inside the complex chromatin environment of a human nucleus.

The NIH-funded team solved this through rational protein engineering, introducing specific amino acid substitutions — the "RKK" mutations — that enhanced the enzyme's ability to access and cleave DNA within the context of human chromatin. The result was dramatic: Al3Cas12f RKK achieved editing efficiencies of up to 90% in human cells, rivaling or exceeding SpCas9 at many genomic loci.

What makes it special

Several features make Al3Cas12f RKK stand out:

• Size: At roughly 400-500 amino acids, the entire CRISPR system (enzyme + guide RNA + regulatory elements) fits comfortably within a single AAV vector with room to spare. This means single-vector delivery to any tissue that AAV can reach — liver, eye, muscle, brain, and beyond.

• Efficiency: The 90% editing efficiency figure is not a cherry-picked best case at one easy locus. The engineered variant showed robust performance across multiple target sites in human cells, suggesting broad applicability.

• T-rich PAM: Cas12f enzymes typically recognize a T-rich PAM sequence (5'-TTTV-3' or similar), which is complementary to the G-rich PAMs of SpCas9 (NGG). This means Cas12f can target genomic regions that are difficult or impossible for Cas9 to reach, expanding the total "editable" space of the human genome.

• Staggered cuts: Like other Cas12 family members, Cas12f generates staggered double-strand breaks with short overhangs, rather than the blunt cuts made by Cas9. This can influence repair outcomes and may be advantageous for certain editing strategies, including those that favor precise insertions via homology-directed repair.

How it compares to other compact editors

Al3Cas12f RKK enters a field that already includes several other compact CRISPR systems:

• CasMINI (~529 amino acids): Engineered from a different Cas12f ortholog by Stanley Qi's lab at Stanford, CasMINI was one of the first compact editors to demonstrate respectable activity in human cells. Published in 2021, it showed editing rates of approximately 30-70% depending on the target site — impressive for its time, but now surpassed by Al3Cas12f RKK's 90%.

• Un1Cas12f1: Another Cas12f variant that has been engineered for improved human-cell activity. Performance has been solid but generally below the levels achieved by Al3Cas12f RKK.

• CasX/Cas12e (~980 amino acids): Discovered by Jill Banfield and Jennifer Doudna's groups at UC Berkeley, CasX is larger than the Cas12f enzymes but still small enough to fit in AAV (though with less headroom). It has been licensed exclusively to Scribe Therapeutics, which is developing it for clinical applications.

The 90% efficiency achieved by Al3Cas12f RKK is significant because it closes what had been a persistent gap: compact editors were small enough for AAV but not efficient enough for therapy. Al3Cas12f RKK appears to have crossed both thresholds simultaneously.

CRISPR-Cas3: The DNA Shredder for Large Deletions

A fundamentally different mechanism

While most of the compact-editor conversation has focused on making smaller versions of the Cas9-like single-protein effectors, a completely different branch of the CRISPR family tree offers an alternative approach: Type I CRISPR systems, which use the Cas3 enzyme.

Type I systems are actually the most abundant CRISPR systems in nature — far more common than the Type II systems that gave us Cas9. But they work in a fundamentally different way. Where Cas9 makes a single, precise double-strand break at its target site, Cas3 is a helicase-nuclease that processively degrades DNA. Once the target is recognized by a multi-protein complex called Cascade (CRISPR-associated complex for antiviral defense), Cas3 is recruited and begins chewing through the DNA in one direction, creating large deletions that can span tens of kilobases.

Think of Cas9 as a pair of surgical scissors making one snip. Cas3 is more like a paper shredder — once it grabs the DNA, it does not stop until it has removed a substantial chunk.

Mouse study results

In a landmark preclinical study, researchers demonstrated the therapeutic potential of CRISPR-Cas3 by targeting the transthyretin (TTR) gene in mice — the same gene targeted by Intellia Therapeutics' Cas9-based therapy NTLA-2001 for transthyretin amyloidosis (ATTR), providing a direct comparison.

The results were striking:

• A single LNP treatment delivered the Cas3 system to the liver
• 48% hepatic editing was achieved — meaning nearly half of all liver cells had their TTR gene disrupted
• This translated to an 80% reduction in circulating TTR protein levels
• Most remarkably, zero off-target indels were detected across the entire genome using unbiased detection methods

That last point deserves emphasis. Off-target editing — where the CRISPR system cuts at unintended locations in the genome — is the single biggest safety concern for any gene-editing therapy. Cas9, despite its precision, has a well-documented propensity for off-target cleavage, particularly at genomic sites with high sequence similarity to the intended target. The fact that Cas3 achieved robust on-target editing with no detectable off-target activity is a major safety advantage.

Why is Cas3 so specific?

The exceptional specificity of Cas3 likely stems from its multi-step target recognition mechanism. The Cascade complex that recognizes the target DNA uses a longer guide RNA (~32 nucleotides, compared to Cas9's 20) and requires a more extensive match before it recruits Cas3 for DNA degradation. This longer recognition sequence means there are fewer sites in the genome that could serve as off-target binding sites.

Additionally, the Cascade complex undergoes conformational changes upon target binding that serve as additional checkpoints — the system essentially double-checks its target identification before unleashing the nuclease. This built-in proofreading makes off-target activity inherently less likely.

Best use cases

CRISPR-Cas3 is not a replacement for Cas9 or compact Cas12 editors in all applications. Its strength is in gene knockout — permanently eliminating a gene's function by deleting large portions of it. The large deletions it creates (often 10-100 kb) make it essentially impossible for the cell to repair the gene, ensuring durable and complete loss of function.

This makes Cas3 ideal for:

• Gain-of-function diseases where a toxic protein must be eliminated (like TTR in ATTR amyloidosis)
• Viral gene disruption — destroying integrated viral genomes like HIV provirus or hepatitis B cccDNA
• Cancer immunotherapy — knocking out immune checkpoint genes in T cells
• Eliminating large regulatory elements that drive disease

However, Cas3 is not well-suited for precise gene correction — replacing a single disease-causing mutation with the correct sequence. For that, you still need single-cut editors (Cas9, Cas12) paired with a repair template, or base editors and prime editors that make targeted changes without double-strand breaks.

Cas12a, CasX, CasMINI: The Supporting Cast

The compact editor landscape extends well beyond Al3Cas12f RKK and Cas3. Several other systems are in active development, each with distinct properties that make them suited to different applications.

Cas12a (Cpf1)

Cas12a, originally called Cpf1 (CRISPR from Prevotella and Francisella), was the first major alternative to Cas9 to gain widespread adoption. Discovered by Feng Zhang's group at the Broad Institute in 2015, Cas12a has several distinguishing features:

• Size: ~1,200-1,300 amino acids — smaller than SpCas9 but still too large for comfortable single-AAV delivery
• PAM: Recognizes a T-rich PAM (5'-TTTV-3'), complementary to Cas9's G-rich PAM
• Cut type: Generates staggered cuts with 4-5 nucleotide 5' overhangs, which some studies suggest favor more precise repair outcomes
• Guide RNA: Uses a shorter, simpler guide RNA (crRNA only, no tracrRNA needed), simplifying the system
• Multiplexing: Cas12a naturally processes its own guide RNA array, making it exceptionally well-suited for multiplexed editing — targeting multiple genes simultaneously with a single construct

Cas12a has become the editor of choice for applications requiring simultaneous knockout of multiple genes, such as engineering CAR-T cells with enhanced anti-tumor properties.

CasX (Cas12e)

CasX, also classified as Cas12e, was discovered in 2019 from metagenomic analysis of groundwater and sediment samples. At approximately 980 amino acids, it is significantly smaller than SpCas9 and just small enough for single-AAV delivery, though with limited room for additional elements.

Scribe Therapeutics, cofounded by Jennifer Doudna, has licensed CasX and is engineering improved variants for clinical applications. Their proprietary X-Editing (XE) technology platform has produced CasX variants with enhanced activity and specificity. Scribe's lead programs target neurological diseases and other conditions requiring CNS delivery.

Key properties of CasX:

• Size: ~980 amino acids — fits in AAV with careful construct design
• PAM: 5'-TTCN-3' (though engineered variants have expanded PAM compatibility)
• Origin: From non-pathogenic environmental bacteria, potentially reducing pre-existing immunity in human patients
• Specificity: Naturally high specificity, with low off-target activity in early studies

CasMINI

CasMINI deserves special mention as a pioneering example of what protein engineering can achieve. Developed by Stanley Qi's laboratory at Stanford and published in Nature Biotechnology in 2021, CasMINI was engineered from a naturally occurring Cas12f ortholog (Un1Cas12f1) through a combination of rational design and directed evolution.

At ~529 amino acids, CasMINI was among the first compact editors to demonstrate meaningful editing activity in mammalian cells. It showed that the Cas12f family, previously dismissed as too weak for therapeutic use, could be engineered into viable tools.

Comparison table

Editor	Type	Size (aa)	PAM	Cut Type	Efficiency	Fits AAV?	Best Use Case
SpCas9	II	1,368	NGG	Blunt DSB	70-95%	No (dual only)	Precise editing, HDR
SaCas9	II	1,053	NNGRRT	Blunt DSB	50-80%	Tight fit	In vivo editing (limited sites)
Cas12a	V-A	~1,250	TTTV	Staggered DSB	50-90%	No (tight)	Multiplexed editing
CasX	V-E	~980	TTCN	Staggered DSB	40-70%	Yes (tight)	CNS, in vivo editing
CasMINI	V-F	~529	TTTV	Staggered DSB	30-70%	Yes (ample room)	In vivo, compact constructs
Al3Cas12f RKK	V-F	~400-500	TTTV	Staggered DSB	Up to 90%	Yes (ample room)	In vivo, broad applications
Cas3	I	Multi-subunit	Variable	Processive deletion	48% (in vivo)	LNP delivery	Gene knockout, large deletions

From Lab to Clinic: Who's Using Compact Editors?

Scribe Therapeutics — CasX/XE platform

Scribe Therapeutics is arguably the furthest along in clinical development with a compact editor platform. The company, based in Emeryville, California, has built its entire pipeline around engineered CasX variants. Key programs include:

• Neurological diseases: Leveraging AAV9 delivery to the CNS, Scribe is developing CasX-based therapies for conditions including amyotrophic lateral sclerosis (ALS) and Huntington's disease, where in vivo editing of neurons is the only viable approach
• Immunology: CasX-based engineering of immune cells for next-generation cell therapies
• Collaboration with major pharma: Scribe entered a multi-billion-dollar collaboration with Eli Lilly in 2023 to develop in vivo gene-editing therapies for neurological conditions, one of the largest deals in the compact editor space

Arbor Biotechnologies — Cas12f and beyond

Arbor Biotechnologies, cofounded by Feng Zhang, has been systematically mining microbial diversity for novel CRISPR systems since 2016. The company holds intellectual property on multiple compact editors, including Cas12f variants and other ultracompact systems discovered through their proprietary metagenomics pipeline.

Arbor's approach is to build a broad portfolio of editing tools, matching each enzyme's strengths to specific therapeutic applications rather than relying on a single platform.

Locus Biosciences — CRISPR-Cas3

Locus Biosciences has been a pioneer in developing CRISPR-Cas3 technology, initially for antimicrobial applications (using engineered phage to deliver Cas3 systems that shred bacterial genomes) and more recently for mammalian gene therapy.

Their antimicrobial programs — targeting drug-resistant urinary tract infections and other bacterial infections — provided early clinical proof that Cas3-based systems could be delivered and function in a therapeutic setting, albeit in bacteria rather than human cells.

Academic-to-industry pipeline

The NIH-funded work on Al3Cas12f RKK is still in the academic stage, but given the 90% efficiency results, it is virtually certain that commercial licensing and company formation will follow. The pattern is well established: landmark academic papers on new CRISPR systems are typically followed within 12-18 months by startup formation, exclusive licensing deals, or acquisition by established gene-therapy companies.

Delivery strategies evolving in parallel

The compact editors are arriving at the same time that delivery technologies are advancing:

• Engineered AAV capsids: Companies like Dyno Therapeutics (now part of Astellas) and Capsida Biotherapeutics are using machine learning to design AAV capsids with improved tissue targeting, reduced immunogenicity, and enhanced transduction efficiency. Pairing engineered capsids with compact editors creates a powerful combination.

• Tissue-targeted LNPs: While standard LNPs preferentially accumulate in the liver, new formulations incorporating targeting ligands or ionizable lipids with different biodistribution profiles are enabling LNP delivery to the lung, spleen, muscle, and even the brain. For larger CRISPR systems like Cas3 that do not fit in AAV, these next-generation LNPs are essential.

• Engineered virus-like particles (eVLPs): Developed by David Liu's group at the Broad Institute, eVLPs package CRISPR ribonucleoproteins (pre-formed protein-RNA complexes) directly into delivery particles, enabling transient editing without any DNA delivery. This eliminates the risk of insertional mutagenesis and reduces the window for off-target activity.

The Road Ahead

Remaining challenges

The arrival of compact editors with high efficiency does not mean in vivo gene therapy is a solved problem. Significant challenges remain:

Immunogenicity: All CRISPR proteins are bacterial in origin, and the human immune system may recognize them as foreign. Pre-existing immunity (from prior bacterial exposure) and adaptive immune responses (triggered by the therapy itself) could reduce efficacy or cause safety problems. This concern applies equally to Cas9 and compact editors, though enzymes derived from rare environmental bacteria (like CasX) may have lower rates of pre-existing immunity in human populations.

Tissue targeting: AAV serotypes provide preferential but not exclusive tropism for specific tissues. A therapy intended for the liver will also transduce some heart, muscle, and other cells. For gene-knockout applications, off-tissue editing could have unintended consequences. Improving the specificity of delivery — getting the editor only to the cells that need it — remains a major engineering challenge.

Durability: AAV delivers DNA that persists as an episome (non-integrating circular DNA) in the nucleus of transduced cells. In non-dividing cells like hepatocytes and neurons, this episome can persist for years, providing durable expression. But in dividing cells, the episome is gradually diluted with each cell division. For diseases affecting proliferating tissues, single-dose durability cannot be assumed.

Manufacturing scale: AAV manufacturing is notoriously expensive and difficult to scale. Current costs for AAV-based gene therapies often exceed $1-2 million per patient, driven largely by manufacturing challenges. Compact editors do not change this equation directly, though single-vector delivery (instead of dual-vector) does cut manufacturing complexity in half.

Regulatory pathways: Each new CRISPR enzyme is, from a regulatory perspective, a new drug substance. It requires its own toxicology studies, biodistribution analyses, and safety data package. The FDA and EMA have gained experience with Cas9, but compact editors will need to build their own safety dossiers largely from scratch.

Emerging technologies to watch

Several emerging approaches could amplify the impact of compact editors:

• Epigenome editing: Using catalytically dead versions of compact editors (dCas12f) fused to epigenetic modifiers to turn genes on or off without cutting DNA. This is inherently reversible and avoids the risks associated with permanent DNA changes. The small size of Cas12f makes it particularly attractive for these fusion constructs, as there is room within AAV for both the editor and the epigenetic effector domain.

• RNA editing: CRISPR systems that target RNA instead of DNA (like Cas13) could complement compact DNA editors by providing temporary, tunable gene knockdown. Combining permanent DNA edits (compact editors) with transient RNA edits (Cas13) could enable sophisticated therapeutic strategies.

• Prime editing miniaturization: Prime editors, which can make virtually any small edit without double-strand breaks, are currently even larger than Cas9 (the prime editor 2 construct is ~6.3 kb). If the protein engineering approaches that produced Al3Cas12f RKK could be applied to create compact prime editors, it would be transformative.

Could compact editors replace Cas9 entirely?

Probably not — at least not in the foreseeable future. SpCas9 has enormous advantages in terms of characterization depth, available tools (base editors, prime editors, CRISPRa/i, and dozens of other adaptations), and clinical experience. For ex vivo therapies, where delivery size is not a constraint, Cas9 remains an excellent choice.

But for in vivo gene therapy — the frontier that holds the greatest promise for treating the thousands of genetic diseases that affect tissues inaccessible to ex vivo approaches — compact editors are poised to become the primary tools. The combination of Al3Cas12f RKK's 90% efficiency in a ~500 amino acid package, Cas3's unmatched safety profile for gene knockout, and CasX's neurotropic delivery potential gives the field a diverse toolkit that Cas9 alone could never provide.

The analogy might be to think of Cas9 as the Swiss Army knife that opened up the entire field — versatile, reliable, indispensable for many tasks. Compact editors are the precision instruments you reach for when you need to work in tight spaces. Both belong in the toolbox.

Frequently Asked Questions

What are compact CRISPR editors and why do they matter?

Compact CRISPR editors are naturally smaller gene editing enzymes discovered by mining microbial genomes, then engineered for use in human cells. They matter because the standard SpCas9 enzyme (1,368 amino acids) is too large to fit inside a single adeno-associated virus (AAV) vector — the gold-standard delivery vehicle for in vivo gene therapy, which has a strict ~4.7 kb packaging limit. Compact editors like Cas12f (400-600 amino acids) solve this size problem.

Why is Cas9 too big for AAV delivery?

SpCas9's coding sequence alone runs to about 4.2 kb, and when you add the promoter, guide RNA cassette, and polyadenylation signal needed for a functional construct, the total payload exceeds AAV's 4.7 kb packaging limit. This has forced the field to rely on dual-AAV systems that require both vectors to reach the same cell — dramatically cutting efficiency — or to use lipid nanoparticles that primarily target the liver.

What is Al3Cas12f RKK and how efficient is it?

Al3Cas12f RKK is an engineered variant of the naturally occurring Cas12f enzyme, developed by NIH-funded researchers and published in April 2026. At roughly 400-500 amino acids, it fits comfortably in a single AAV vector with room to spare. Through rational protein engineering (the "RKK" mutations), it achieved editing efficiencies of up to 90% in human cells — rivaling or exceeding SpCas9 at many genomic loci.

How does CRISPR-Cas3 differ from Cas9?

Unlike Cas9, which makes a single precise double-strand break, Cas3 is a helicase-nuclease that processively degrades DNA, creating large deletions spanning tens of kilobases. In a mouse study targeting the TTR gene, a single LNP treatment achieved 48% hepatic editing, 80% reduction in circulating TTR protein, and — most remarkably — zero detectable off-target indels across the entire genome, giving it an extraordinary safety profile for gene knockout applications.

When will compact CRISPR editors be used in patients?

Scribe Therapeutics is arguably the furthest along, developing engineered CasX variants for neurological diseases including ALS and Huntington's disease through a multi-billion-dollar collaboration with Eli Lilly. The NIH-funded Al3Cas12f RKK work is still academic, but given the 90% efficiency results, commercial licensing and company formation are expected within 12-18 months. First clinical trials using compact editors could begin within the next 2-3 years.

The Bottom Line

The story of CRISPR gene therapy has always had two chapters: the science and the delivery. The science matured first — we have known how to program Cas9 to cut nearly any gene since 2012. The delivery side has taken far longer, and the central problem has been size.

In 2026, that problem is cracking open. Al3Cas12f RKK proves that a CRISPR editor small enough to fit in a single AAV can match or exceed Cas9's cutting efficiency. CRISPR-Cas3 demonstrates that a completely different approach — processive DNA degradation — can achieve potent gene knockout with an extraordinary safety profile. And a growing roster of other compact editors (CasX, CasMINI, engineered Cas12a variants) ensures that the field is not dependent on any single system.

The practical implications are profound. Diseases that were previously out of reach for gene therapy — neurodegenerative conditions, muscular dystrophies, inherited eye diseases, liver disorders, and many more — are now accessible targets. The therapy design equation changes from "How do we split this system across two vectors and hope both reach the same cell?" to "Which compact editor and which AAV serotype give us the best combination for this tissue and this disease?"

There are still hard problems to solve. Immunogenicity, tissue targeting, manufacturing costs, and regulatory pathways will keep the field busy for years. But the fundamental constraint — the size mismatch between the editing tool and the delivery vehicle — is being resolved. The basketball has been shrunk to fit through the letter slot.

The next wave of CRISPR therapies will look very different from the first. They will be delivered in a single injection, they will reach tissues that ex vivo approaches cannot, and they will use editors whose names most people have not yet learned. Cas12f. CasX. Cas3. Remember them.

Sources & Further Reading

1. NIH/NHGRI announcement on Al3Cas12f RKK — National Human Genome Research Institute, April 2026. Details on the engineered Cas12f variant achieving 90% editing efficiency in human cells.

2. CRISPR-Cas3 mouse study for TTR knockdown — Demonstrated 48% hepatic editing, 80% TTR reduction, and zero off-target indels with a single LNP treatment. Preclinical data supporting Type I CRISPR systems for therapeutic gene knockout.

3. Xu, X., Chemparathy, A., Zeng, N. et al. "Engineered miniature CRISPR-Cas system for mammalian genome regulation and editing." Molecular Cell, 2021. Original CasMINI publication.

4. Liu, J.J., Orlova, N., Oakes, B.L. et al. "CasX enzymes comprise a distinct family of RNA-guided genome editors." Nature, 2019. Discovery and characterization of CasX/Cas12e.

5. Zetsche, B., Gootenberg, J.S., Abudayyeh, O.O. et al. "Cpf1 is a single RNA-guided endonuclease of a class 2 CRISPR-Cas system." Cell, 2015. Original Cas12a/Cpf1 discovery paper.

6. Dolan, A.E., Hou, Z., Xiao, Y. et al. "Introducing a spectrum of long-range genomic deletions in human embryonic stem cells using Type I CRISPR-Cas." Molecular Cell, 2019. Early demonstration of Cas3 for large deletions in human cells.

7. Scribe Therapeutics corporate pipeline — scribetx.com. Information on CasX/XE platform clinical programs and Eli Lilly collaboration.

8. Wang, J.Y., Doudna, J.A. "CRISPR technology: A decade of genome editing is only the beginning." Science, 2023. Comprehensive review of the CRISPR editing landscape.

9. Locus Biosciences antimicrobial CRISPR-Cas3 programs — lfrx.com. Clinical-stage Cas3-based antimicrobial development.

10. Kannan, S., Altae-Tran, H., Jin, X. et al. "Compact RNA editors with small Cas13 proteins." Nature Biotechnology, 2022. Compact RNA-editing systems as complementary tools.