CRISPR Gene Editing: How It Works in 2026

In December 2023, the FDA approved the first DNA sequences in living cells with letter-level accuracy.">CRISPR-based therapy ever cleared for human use. Casgevy, developed by Vertex Pharmaceuticals and CRISPR Therapeutics, offered a functional cure for sickle cell disease and beta-thalassemia — conditions that had plagued millions of people for generations. What made that moment remarkable was not just the drug itself, but the distance traveled: from an obscure bacterial defense mechanism observed in the 1980s to a medicine capable of rewriting a patient's DNA in a single treatment.

This is the story of CRISPR — how it works, why it matters, and where it is heading in 2026.

The Bacterial Immune System That Started It All

CRISPR stands for Clustered Regularly Interspaced Short Palindromic Repeats. The name describes a distinctive pattern first noticed in 1987 by Japanese microbiologist Yoshizumi Ishino in the genome of E. coli: short repeated DNA sequences, separated by unique spacer sequences of roughly equal length. At the time, no one knew what they did.

The function became clear over the next two decades. Bacteria live in constant war with bacteriophages — viruses that hijack bacterial cells to replicate. CRISPR is one of bacteria's most sophisticated countermeasures. When a phage infects a bacterium and the bacterium survives, it can snip a fragment of the viral DNA and insert it between CRISPR repeats in its own genome. This creates a molecular memory — a genetic record of past infections.

When the same phage attacks again, the bacterium transcribes those spacer sequences into short RNA molecules called crRNA (CRISPR RNA). These crRNAs pair with a structural RNA called tracrRNA and then associate with a Cas (CRISPR-associated) protein, most famously Cas9. The resulting complex patrols the cell, scanning incoming DNA. When it finds a sequence that matches the stored crRNA — a sequence from the original viral attack — Cas9 cuts both strands of the viral DNA, destroying it.

The system is both a surveillance camera and a pair of scissors. It recognizes specific sequences with high precision and eliminates them on contact.

The 2012 Breakthrough: Repurposing a Bacterial Weapon

Microbiologists had been studying CRISPR's natural role in bacteria for years. What transformed the field was the insight that this machinery could be redirected to edit any DNA sequence — not just viral DNA in bacteria, but any genome, in any organism.

In June 2012, Jennifer Doudna at UC Berkeley and Emmanuelle Charpentier, then at the University of Vienna, published a landmark paper in Science demonstrating that Cas9 could be programmed with a synthetic single guide RNA (sgRNA) — a simplified version that fused crRNA and tracrRNA into one molecule — to cut any DNA sequence specified by the researcher. Point Cas9 at a new target by changing 20 nucleotides in the guide RNA, and it cuts there.

The implication was enormous: genome editing had previously required custom-engineered proteins (zinc finger nucleases, TALENs) that took months to design and synthesize per target. With CRISPR-Cas9, the same protein could be redirected to any genomic address in days, simply by changing the RNA guide sequence. Doudna and Charpentier received the 2020 Nobel Prize in Chemistry for this discovery.

Within months of publication, Feng Zhang's laboratory at the Broad Institute adapted the system to work in eukaryotic cells — cells with a nucleus, including human cells — publishing that result in January 2013. CRISPR-Cas9 had become a practical tool for human genome editing.

How CRISPR-Cas9 Cuts and Repairs DNA

The core mechanism of CRISPR-Cas9 editing involves three steps:

1. Target recognition. The sgRNA contains a ~20-nucleotide sequence complementary to the genomic target. Cas9 scans DNA until it finds a matching sequence adjacent to a short motif called the PAM (Protospacer Adjacent Motif) — for Streptococcus pyogenes Cas9, this is the sequence NGG. The PAM requirement adds a layer of specificity and limits the possible target sites to roughly every 8–12 base pairs across the human genome.

2. Double-strand break (DSB). Once bound, Cas9 unwinds the DNA helix and checks the guide RNA for complementarity. If it matches, two nuclease domains within Cas9 — RuvC and HNH — each cut one strand of the DNA, creating a double-strand break at a precise location.

3. DNA repair. A double-strand break is a serious cellular emergency. The cell responds with one of two main repair pathways:

• NHEJ (Non-Homologous End Joining): The fast, error-prone pathway. The cell glues the broken ends back together, but frequently introduces small insertions or deletions (indels). If the break falls within a gene's coding sequence, these indels typically disrupt the reading frame, knocking out gene function. This is useful for silencing a disease-causing gene.

• HDR (Homology-Directed Repair): The precise pathway. If a DNA template with the desired sequence is provided alongside Cas9, the cell's repair machinery can use it to introduce a specific edit — correcting a mutation, inserting a new sequence, or making any targeted change. HDR is more accurate but less efficient, and it only works in dividing cells.

Delivery Systems: Getting CRISPR Into Cells

The guide RNA and Cas9 protein must reach the target cells. This is one of the central engineering challenges of therapeutic CRISPR.

Adeno-Associated Virus (AAV): The most clinically mature delivery vehicle. AAV is a small, non-pathogenic virus that efficiently infects many cell types. Its main limitation for CRISPR: Cas9 is a large protein, and the full CRISPR machinery often exceeds AAV's packaging capacity (~4.7 kb), requiring split-AAV strategies or smaller Cas variants like SaCas9.

Lipid Nanoparticles (LNPs): Small fatty spheres that encapsulate mRNA or ribonucleoprotein complexes and fuse with cell membranes. LNPs deliver transiently — Cas9 is expressed briefly from mRNA, then degraded — reducing off-target risk. Intellia Therapeutics' in vivo CRISPR program for transthyretin amyloidosis uses LNP delivery to the liver, with Phase 2 results in 2024 showing >90% reduction in the disease-causing protein after a single dose.

Ex vivo editing: For blood and immune cell diseases, cells are extracted from the patient, edited in the lab, and reinfused. Casgevy uses this approach — patient stem cells are edited outside the body and transplanted back after chemotherapy conditioning.

CRISPR-Cas9 Limitations

Despite its power, first-generation CRISPR-Cas9 has real constraints:

• Off-target cuts: Cas9 can bind and cleave DNA at sequences that partially match the guide RNA. High-fidelity Cas9 variants (eSpCas9, HiFi Cas9) reduce this substantially, but off-target editing remains a safety concern in clinical settings.
• Double-strand breaks: DSBs can cause chromosomal rearrangements, deletions, and translocations — especially in rapidly dividing cells. Any repair pathway that creates genomic instability is a cancer risk in principle.
• PAM requirement: Standard SpCas9 requires an NGG PAM, restricting targetable sites. Engineered variants (SpRY, SpCas9-NG) have expanded the PAM repertoire.
• Delivery size: Cas9 from S. pyogenes is ~4.2 kb just as protein-coding sequence, straining AAV capacity.

Next-Generation Tools: Beyond the Scissors

The limitations of DSB-based editing have driven innovation toward tools that make changes without cutting both strands.

Base Editing

Developed by David Liu's laboratory at the Broad Institute starting in 2016, base editors fuse a catalytically impaired Cas9 (which can bind DNA but not cut) with a deaminase enzyme that chemically converts one DNA base into another. No DSB is required.

• Cytosine Base Editors (CBEs): Convert C•G base pairs to T•A.
• Adenine Base Editors (ABEs): Convert A•T base pairs to G•C.

Together, these four conversions cover the four possible transition mutations. Since approximately 30–40% of known pathogenic point mutations in humans involve transition changes, base editing can potentially correct a substantial fraction of single-nucleotide genetic diseases. In 2023, Beam Therapeutics reported early clinical data showing base editing corrected sickle cell disease mutations in patient cells with high efficiency.

Prime Editing

Also from the Liu lab, prime editing was introduced in 2019 and described as a "search and replace" function for the genome. A prime editor consists of:

1. A nickase Cas9 (cuts only one strand).
2. A reverse transcriptase fused to Cas9.
3. A pegRNA (prime editing guide RNA) that both specifies the target and encodes the desired edit as an RNA template.

The pegRNA's 3' extension hybridizes to the nicked strand and acts as a template for the reverse transcriptase, which writes the new sequence directly into the genome. Prime editing can install all 12 types of point mutations (all transitions and transversions), small insertions, and small deletions without a DSB. A 2023 study in Nature Biotechnology demonstrated prime editing corrections in mouse models of several genetic diseases, including Duchenne muscular dystrophy.

CRISPR-Cas13: Editing RNA, Not DNA

Cas13 is a distinct family of CRISPR proteins that targets RNA rather than DNA. This creates a fundamentally different mode of action: edits are transient (the RNA is degraded naturally), the genome is untouched, and there is no risk of permanent off-target genomic alterations.

Cas13-based tools include:

• HADS (Highly Accessible Diagnostic System): Rapid pathogen detection — the basis for several SARS-CoV-2 diagnostics during the pandemic.
• RNA knockdown: Degrading disease-associated RNA transcripts.
• RNA base editing (RESCUE, LEAPER): Deaminase-fused Cas13 proteins that convert A-to-I or C-to-U in RNA, transiently altering protein sequence or splicing.

Cas13 is also notably smaller than Cas9, fitting more easily into AAV vectors.

Comparison of CRISPR Editing Modalities

Feature	CRISPR-Cas9	Base Editing	Prime Editing	CRISPR-Cas13
Target	DNA	DNA	DNA	RNA
Mechanism	Double-strand break	Chemical base conversion	Reverse transcription into genome	RNA cleavage / RNA base conversion
DSB required?	Yes	No	No (single nick only)	No
Edit types	Knockouts, large insertions via HDR	4 transition mutations	All point mutations, small indels/insertions	Transient knockdown or RNA base edits
Off-target risk	Moderate (DNA)	Low-moderate	Low	Low (RNA, transient)
Delivery challenge	Large size (SpCas9 ~4.2 kb)	Moderate (editor + sgRNA)	Higher (pegRNA + editor)	Smaller, fits AAV easily
Clinical stage	Approved (Casgevy, 2023)	Phase 1/2 trials	Preclinical/early Phase 1	Research / diagnostics
Best suited for	Gene knockouts; ex vivo editing	Point mutation correction	Versatile precision edits	Antiviral; transient knockdown

Clinical Milestones: CRISPR in Human Medicine

Casgevy (exa-cel), 2023: The first CRISPR therapy approved by the FDA (December 2023) and the UK's MHRA (November 2023). Developed by CRISPR Therapeutics and Vertex, it targets sickle cell disease and transfusion-dependent beta-thalassemia. The therapy edits patients' hematopoietic stem cells to reactivate fetal hemoglobin (HbF) production — compensating for the defective adult hemoglobin. In the pivotal trial, 28 of 29 sickle cell patients were free of vaso-occlusive crises for at least 12 months after treatment.

NTLA-2001 (Intellia Therapeutics): An in vivo CRISPR therapy for transthyretin (ATTR) amyloidosis, a progressive and fatal condition caused by misfolded TTR protein. Delivered via LNPs directly to the liver, Phase 1/2 data published in 2023–2024 showed a single dose reduced serum TTR levels by over 90%, with durable effects maintained at 2-year follow-up. This represents a major proof-of-concept for in vivo CRISPR editing in humans.

Cancer immunotherapy: Multiple groups are using CRISPR to engineer T cells for cancer treatment. Approaches include knocking out PD-1 (an immune checkpoint) in T cells, disrupting the T cell receptor for universal CAR-T therapies, and multiplex editing to enhance tumor-killing capacity. The University of Pennsylvania and Peking University both published Phase 1 safety data for CRISPR-edited T cell therapies in 2020–2022, demonstrating feasibility. Several larger trials are active in 2026.

Ethical Dimensions: The Shadow of He Jiankui

CRISPR's power forced ethics to catch up with science. The most dramatic illustration came in November 2018, when Chinese researcher He Jiankui announced the birth of twin girls whose embryos he had CRISPR-edited to disrupt the CCR5 gene — a co-receptor used by HIV to enter cells. He claimed this would confer HIV resistance.

The global scientific community responded with near-unanimous condemnation. The editing had not been shown safe in human embryos. Embryonic edits are germline changes — they are heritable and will be passed to all future descendants. Informed consent from the parents was inadequate and potentially coercive. He was subsequently sentenced to three years in prison by Chinese courts.

The incident crystallized the line most bioethicists draw: somatic cell editing (in non-reproductive cells of one individual) is ethically analogous to other medical interventions. Germline editing — changing heritable DNA — is a different category entirely, touching questions of consent for future generations, genetic enhancement, and inequality. As of 2026, germline editing for reproduction remains prohibited in nearly all countries and deeply contested in international scientific bodies.

The 2026 Pipeline

As of mid-2026, the CRISPR pipeline has expanded substantially:

• In vivo liver targeting is now the most mature in vivo modality, with multiple programs in Phase 2/3 for hemophilia, hypercholesterolemia (PCSK9 knockdown), and alpha-1 antitrypsin deficiency.
• In vivo eye delivery (subretinal injection) is in Phase 2 for Leber congenital amaurosis and other inherited retinal diseases.
• Base editing has entered Phase 1/2 for sickle cell disease (Beam Therapeutics) and acute lymphoblastic leukemia (base-edited allogeneic CAR-T from Great Ormond Street Hospital).
• Prime editing is transitioning from preclinical to first-in-human studies, with programs in Friedreich's ataxia and certain lysosomal storage disorders.
• CRISPR diagnostics (SHERLOCK, DETECTR platforms) are commercially deployed for infectious disease testing.

The pace of translation from bench to clinic has been faster for CRISPR than for almost any prior biotechnology, compressing a 30-year arc into roughly 12 years from the 2012 mechanistic paper to the first approved therapy.

Why CRISPR Matters Beyond Medicine

CRISPR's impact reaches well beyond the clinic. In agriculture, CRISPR-edited crops with disease resistance, improved yield, and altered nutritional profiles have cleared regulatory pathways in the US and Japan. In basic research, CRISPR screens — systematically knocking out thousands of genes to find which ones affect a phenotype — have become a standard tool in cancer biology and drug discovery. The technology has made large-scale functional genomics tractable in a way that was impossible before 2013.

For a deeper look at how genetic information is encoded and expressed, see our article on ./dna-structure-and-function-complete-guide, and for the broader landscape of biotechnology tools, see ./synthetic-biology-explained.

FAQs

Q: What does CRISPR stand for and what is it? A: CRISPR stands for Clustered Regularly Interspaced Short Palindromic Repeats. It refers to a natural immune system found in bacteria that stores fragments of viral DNA to recognize future infections. Scientists have repurposed the associated Cas9 protein and guide RNA into a programmable tool that can cut and edit DNA at any specified location in any genome.

Q: How does Cas9 know where to cut DNA? A: Cas9 is guided by a short RNA molecule (the guide RNA or sgRNA) that contains a ~20-nucleotide sequence complementary to the target DNA. Cas9 scans the genome for matching sequences adjacent to a PAM motif (NGG for the most common SpCas9). When it finds a match, it unwinds the double helix and cleaves both strands of DNA at that location.

Q: What is the difference between NHEJ and HDR repair? A: After Cas9 cuts DNA, the cell repairs the break via one of two pathways. NHEJ (Non-Homologous End Joining) is fast but error-prone — it often introduces small insertions or deletions that disrupt gene function, making it useful for gene knockouts. HDR (Homology-Directed Repair) uses a supplied DNA template to introduce precise changes, enabling correction of specific mutations, but it is less efficient and only works in dividing cells.

Q: What was the first CRISPR therapy approved for human use? A: Casgevy (exa-cel), developed by CRISPR Therapeutics and Vertex Pharmaceuticals, became the first approved CRISPR therapy in November–December 2023. It treats sickle cell disease and transfusion-dependent beta-thalassemia by editing patients' hematopoietic stem cells to reactivate fetal hemoglobin production. The UK's MHRA approved it in November 2023, followed by the FDA in December 2023.

Q: What is base editing and how is it different from standard CRISPR? A: Base editing uses a modified Cas9 that can bind DNA without cutting both strands, fused to a deaminase enzyme that chemically converts one DNA base into another. Cytosine base editors change C to T; adenine base editors change A to G. Because no double-strand break is created, base editing has a lower risk of chromosomal rearrangements and higher precision for correcting single-letter mutations.

Q: What is prime editing? A: Prime editing is a "search and replace" genome editing technology developed by David Liu's lab at the Broad Institute. A prime editor uses a nickase Cas9 (which cuts only one strand) fused to a reverse transcriptase, guided by a pegRNA that both specifies the target and encodes the desired edit. The reverse transcriptase writes the new sequence directly into the genome. Prime editing can make all 12 types of point mutations plus small insertions and deletions, without requiring a double-strand break or a separate DNA template.

Q: Why is germline CRISPR editing controversial? A: Germline editing — editing embryos, sperm, or eggs — creates heritable changes that are passed to all future descendants. This raises profound ethical concerns: future people cannot consent to genetic changes made before their birth; off-target edits could introduce new inherited diseases; and the technology could be used for non-therapeutic enhancement, potentially widening social inequality. The 2018 case of He Jiankui, who created the first CRISPR-edited babies without adequate ethical oversight, demonstrated these risks and resulted in international calls for a moratorium on heritable genome editing.

Q: What are CRISPR-Cas13 and its potential applications? A: CRISPR-Cas13 is a variant that targets RNA rather than DNA. Because RNA is naturally degraded and replaced, Cas13-based edits are transient — they do not permanently alter the genome. This makes Cas13 attractive for applications where permanent editing is unnecessary or undesirable: knocking down disease-causing RNA transcripts, antiviral applications (disrupting viral RNA), RNA base editing for transient protein changes, and highly sensitive diagnostics. Several SARS-CoV-2 rapid tests used Cas13-based detection platforms during the pandemic.

Sources and Further Reading

• Jinek M, Chylinski K, Fonfara I, Hauer M, Doudna JA, Charpentier E. (2012). "A programmable dual-RNA-guided DNA endonuclease in adaptive bacterial immunity." Science, 337(6096), 816–821.
• Cong L, Ran FA, Cox D, et al. (2013). "Multiplex genome engineering using CRISPR/Cas systems." Science, 339(6121), 819–823. [Feng Zhang lab, eukaryotic cells]
• Komor AC, Kim YB, Packer MS, Zuris JA, Liu DR. (2016). "Programmable editing of a target base in genomic DNA without double-stranded DNA cleavage." Nature, 533, 420–424. [Original base editing paper]
• Anzalone AV, Randolph PB, Davis JR, et al. (2019). "Search-and-replace genome editing without double-strand breaks or donor DNA." Nature, 576, 149–157. [Prime editing]
• Frangoul H, Altshuler D, Cappellini MD, et al. (2021). "CRISPR-Cas9 gene editing for sickle cell disease and β-thalassemia." New England Journal of Medicine, 384, 252–260. [Casgevy pivotal trial]
• Gillmore JD, Gane E, Taubel J, et al. (2021). "CRISPR-Cas9 in vivo gene editing for transthyretin amyloidosis." New England Journal of Medicine, 385, 493–502. [NTLA-2001 Phase 1]
• Vertex Pharmaceuticals / CRISPR Therapeutics. (2023). FDA Approval Announcement for Casgevy. December 8, 2023.
• Intellia Therapeutics. (2024). NTLA-2001 Phase 2 two-year follow-up data, presented at American Heart Association Scientific Sessions 2024.
• National Academies of Sciences, Engineering, and Medicine. (2020). Heritable Human Genome Editing. Washington, DC: The National Academies Press.

---

Read next:

• CRISPR Gene Editing Explained: How It Actually Works
• CRISPR Gene Editing in 2026 — What's Been Cured, What's Coming, What's Real vs Hype
• CRISPR Gene Editing: Precise Genetic Surgery Explained