One CRISPR Infusion Kept ANGPTL3 Suppressed for Three Years in Primates

A single infusion of a CRISPR adenine base editor kept the liver gene ANGPTL3 largely switched off for three years in nonhuman primates, according to a peer-reviewed Nature study published on 6 November. The treatment cut circulating ANGPTL3 protein by more than 90% on average. In mouse and primate models of high blood lipids, associated reductions were about 40 to 50% for LDL cholesterol and 50 to 70% for triglycerides.

Key number: more than 90%. Mean circulating ANGPTL3 protein remained reduced by over 90% in treated nonhuman primates, with durability confirmed three years after a single dose.

Changing one DNA letter to stop a liver protein

ANGPTL3 is produced mainly in the liver and helps regulate the way the body processes cholesterol and triglyceride-rich particles. People who naturally carry loss-of-function variants in the gene tend to have lower blood lipids, making the gene a therapeutic target.

The experimental treatment, known in development as VERVE-201, carries two RNA components. One is messenger RNA encoding an adenine base editor. The other guides the editor to a splice site in ANGPTL3. The intended single-letter DNA change prevents the liver from making functional ANGPTL3 without cutting both strands of the DNA helix.

The RNA cargo sits inside a lipid nanoparticle decorated with GalNAc, a sugar ligand that binds a receptor abundant on liver cells. That route is designed to work even when LDL receptors are scarce, an important issue in homozygous familial hypercholesterolemia. NewTqnia has previously examined how targeted lipid nanoparticles can concentrate an RNA payload in selected cells, although that work addressed cancer immunity rather than inherited lipid disease.

What the experiments found

In cultured primary human liver cells, peak editing ranged from 84.7% to 93.7%. The researchers screened roughly 3,000 candidate genomic sites and reported no detectable off-target editing above their assay threshold, as well as no treatment-related structural variants in the analyses described by the paper.

In the long-term primate study, a single 3 milligram-per-kilogram dose produced 63.9% whole-liver editing and a 93.7% time-weighted average reduction in circulating ANGPTL3 over three years. The authors describe the treatment as well tolerated in the tested animals.

The lipid reductions came from hyperlipidaemia models rather than healthy human volunteers. Across mouse and nonhuman-primate models, ANGPTL3 inactivation was associated with approximately 40 to 50% lower LDL cholesterol and 50 to 70% lower triglycerides. In a small LDL-receptor-deficient primate experiment, four animals received the treatment and were followed for only 35 days. That experiment supports delivery without normal LDL-receptor activity, but it cannot establish long-term efficacy in that disease model.

Peer review adds evidence, not a clinical verdict

Some of the programme's nonclinical findings were presented publicly before the Nature paper. The new publication matters because it assembles the cell, mouse and primate evidence into a peer-reviewed record and extends the durability observation to three years. It also documents a broad laboratory search for unintended edits.

Neither result removes the central uncertainty. A test that does not detect off-target changes in cultured cells cannot prove that rare or delayed harms will never occur in patients. Permanent editing also changes the risk calculation because the intervention cannot simply be stopped if an unexpected effect appears.

The human test has begun

A first-in-human Phase 1b study called Pulse-1 is evaluating VERVE-201 in adults with refractory hypercholesterolaemia. Its initial job is to establish dose, tolerability and biological activity, not to prove that editing prevents heart attacks or strokes.

That distinction mirrors other cardiovascular screening and intervention research. A recent NewTqnia report on smartwatch-based hypertension detection stressed that a promising technical signal is not itself a clinical outcome. For ANGPTL3 editing, the decisive evidence will be durable lipid lowering in people, careful long-term surveillance and eventually proof that the treatment improves health outcomes compared with existing care.