CRISPR Prime Editing Corrects Sickle Cell Mutation In Vivo

For the first time, researchers have corrected the sickle cell mutation directly inside living animals using prime editing — no donor template, no double-strand breaks, no ex vivo cell manufacturing. The work, published this week in Nature, demonstrates that a single intravenous dose of lipid nanoparticles carrying prime editing machinery can rewrite the pathogenic HBB allele in hematopoietic stem cells at clinically relevant efficiency.

Why This Changes the Delivery Paradigm

Current approved CRISPR therapies for sickle cell (Casgevy, Lyfgenia) require harvesting a patient's stem cells, editing them in a GMP facility, myeloablative conditioning, and reinfusion — a process costing $2-3M per patient and available only at specialized centers. Prime editing in vivo could collapse this to a single infusion.

The team engineered a dual-LNP system: one particle delivers prime editor mRNA (PE3b variant), the other carries pegRNA and nicking sgRNA. In humanized sickle cell mice, 42% of bone marrow HSCs showed precise A-to-T correction at the HbS mutation (E6V) after one dose. Corrected cells produced adult hemoglobin (HbA) at 38% of total hemoglobin — well above the 20% threshold for clinical benefit.

💡

TipPrime editing's 'search-and-replace' mechanism avoids the chromosomal translocations and large deletions seen with Cas9 nuclease editing at the BCL11A locus.

Key Technical Advances

Parameter	Ex Vivo Cas9 (Casgevy)	In Vivo Prime Editing
Editing target	BCL11A enhancer (indirect)	HBB E6V mutation (direct)
Delivery	Electroporation + AAV6	Dual LNP (IV)
Conditioning	Busulfan myeloablation	None required
Manufacturing	12-16 weeks centralized	Off-the-shelf LNPs
Estimated cost	$2.2M	Projected <$500K

The pegRNA design incorporated three critical optimizations: a 13-nt PBS for stability, a 16-nt RT template with silent PAM-disrupting mutations, and chemical modifications (2'-OMe, phosphorothioate) at both ends to resist serum nucleases. The nicking sgRNA was offset 40 bp downstream to favor flap resolution toward the edited strand.

"
We're not just disrupting a repressor anymore. We're spelling the correct codon back into the genome — in the patient's own bone marrow — without cutting both DNA strands.
— Dr. David Liu, Broad Institute

Safety Profile So Far

Whole-genome sequencing of edited mice at 6 months revealed zero off-target edits above 0.1% frequency at predicted sites. No clonal expansions or insertional oncogenesis detected. Liver transaminases transiently elevated (Grade 1-2) in 15% of animals, resolving within 72 hours — consistent with LNP uptake by hepatocytes.

⚠️

WarningHuman HSCs may tolerate prime editing differently. The field still lacks long-term genotoxicity data for in vivo prime editing at scale.

What's Next

IND-enabling studies in non-human primates begin Q1 2027. The team is also testing a single-LNP formulation co-encapsulating all components to simplify manufacturing. If primate data holds, a Phase 1 trial could start by late 2027 — initially for adults ineligible for myeloablative conditioning.

python

# pegRNA design parameters for HBB E6V correction
pegRNA = {
    'spacer': 'GCTGTACCGCCCTGTGG',
    'pbs_length': 13,
    'rt_template': 'TGGGCCACTCTGGGCCAC',  # includes silent PAM mutation
    'modifications': ['2-OMe-3PS', '5-PS']
}

The implications extend beyond sickle cell. This delivery platform could target any hematopoietic disorder with a defined point mutation — beta-thalassemia, chronic granulomatous disease, Wiskott-Aldrich syndrome — without the infrastructure burden of ex vivo manufacturing.

✦

For developers: the pegRNA design rules and LNP formulation details are deposited in Addgene (#215440) and the Prime Editing Guide RNA Database. Start benchmarking against your target locus now — the in vivo window is opening faster than most anticipated.