GalNAc-Lipid nanoparticles enable non-LDLR dependent hepatic delivery of a CRISPR base editing therapy

Abstract

Lipid nanoparticles have demonstrated utility in hepatic delivery of a range of therapeutic modalities and typically deliver their cargo via low-density lipoprotein receptor-mediated endocytosis. For patients lacking sufficient low-density lipoprotein receptor activity, such as those with homozygous familial hypercholesterolemia, an alternate strategy is needed. Here we show the use of structure-guided rational design in a series of mouse and non-human primate studies to optimize a GalNAc-Lipid nanoparticle that allows for low-density lipoprotein receptor independent delivery. In low-density lipoprotein receptor-deficient non-human primates administered a CRISPR base editing therapy targeting the ANGPTL3 gene, the introduction of an optimized GalNAc-based asialoglycoprotein receptor ligand to the nanoparticle surface increased liver editing from 5% to 61% with minimal editing in nontargeted tissues. Similar editing was noted in wild-type monkeys, with durable blood ANGPTL3 protein reduction up to 89% six months post dosing. These results suggest that GalNAc-Lipid nanoparticles may effectively deliver to both patients with intact low-density lipoprotein receptor activity as well as those afflicted by homozygous familial hypercholesterolemia.

Fig. 1. Structures and initial screen of GalNAc-Lipids.

a Structure of ligand Designs 1 and 2. b Structure of R moiety for GL3, which uses ligand Design 1. c Structures of R moiety for GL5, GL6, GL7, and GL9. These structures utilize ligand Design 2 but differ in their lipid anchors and polyethylene glycol (PEG) spacer lengths. d GalNAc-LNPs constituted with 0.05 mol % GL3 (ligand Design 1 – Table S1, entry 2) or 0.05 mol % GL6 (ligand Design 2 – Table S1, entry 1) were prepared via the in-lipid mixing method and were administered to female 8-10 week old Ldlr ^–/– mice via injection into the retro-orbital sinus at a dose of 0.1 mg/kg. GL3 and GL6 differ only in the ligand design; the PEG spacer and lipid chain are the same. Ligand 2-based GL6 GalNAc-LNPs achieved higher Angptl3 liver editing when mice were administered GalNAc-LNPs encapsulating mouse-specific Angptl3 guide RNA and ABE8.8 mRNA. Data was analyzed with a two-tailed unpaired T test, p = 0.0086 (df=8, mean difference = 8.9%, 95% confidence interval 2.9–14.7%). e GalNAc-LNPs constituted with 0.5 mol % GL3 (Table S1, entry 4) or GL6 (Table S1, entry 3) were prepared via the post-addition method and were administered to female Ldlr ^–/– mice via injection into the retro-orbital sinus at a dose of 0.25 mg/kg. Ligand Design 2-based GL6 GalNAc-LNPs achieved higher Pcsk9 liver editing, when mice were administered GalNAc-LNPs encapsulating mouse-specific Pcsk9 guide RNA and ABE8.8 mRNA (N = 5). Data were analyzed with a two-tailed unpaired T test, p = 0.01 (df = 8, mean difference = 10.3%, 95% confidence interval 2.9–17.6%). f GalNAc-LNPs formulated with the longer PEG spacer of GL6 (Table S1, entry 6) achieved higher editing of Angptl3 than the GalNAc-LNPs with the shorter PEG spacer of GL5 (Table S1 entry 5) at 0.3 mg/kg in female Ldlr ^–/– mice. Data were analyzed with a two-tailed unpaired T test, p < 0.0001 (df = 7, mean difference = 38.5%, 95% confidence interval 34.7–42.2%). g Modulation of the lipid tail hydrophobicity in GL7 and GL9 (Table S1, entries 7 and 8) with cholesteryl (Chol) and arachidoyl (C20) moieties was unable to improve the editing efficiency of GalNAc-LNPs in female Ldlr ^–/– mice compared to the 1,2-O-dioctadecyl-sn-glyceryl (DSG)-based GL6 (Table S1, entry 6) at 0.3 mg/kg. Data are presented as mean values + /- standard deviation, and individual data points for each animal are displayed (n = 5). * denotes p < 0.05, ** denotes p value <0.01, **** denotes p < 0.0001. Source data are provided as a Source Data file.

Fig. 2. GalNAc-Lipid optimization in LDLR-deficient mouse models.

a GalNAc-Lipid GL6 comprising a PEG spacer and 1,2-di-O-octadecyl-sn-glyceryl lipid anchor; b titration of the surface density of GalNAc-Lipid demonstrated a low density near 0.05 mol % of GalNAc-Lipid optimally rescues liver editing in female 8–10 week old Ldlr ^–/– mice while preserving editing in wild type (WT) mice at an RNA dose of 0.1 mg/kg (Table S1, entries 9–14); c LNPs constituted with 0.05 mol % GL6 maintained Angptl3 base editing in female WT and Ldlr ^+/– mice and rescued base editing in Ldlr ^–/– mice in vivo at 0.25 mg/kg; d demonstration of near-identical dose response of liver Angptl3 editing using the optimized GalNAc-LNPs (constituted with 0.05 mol % GL6) in three genotypes: WT, Ldlr ^+/–, and Ldlr ^–/–. Data are presented as mean values + /- standard deviation, and individual data points for each animal are displayed (n = 5). Source data are provided as a Source Data file.

Fig. 3. Demonstration of adenine base editing by GalNAc-LNPs targeting ANGPTL3 in the liver of a somatic LDLR-deficient NHP model.

a Schematic detailing the creation of the somatic low-density lipoprotein receptor (LDLR) deficient/knockout (KO) model in non-human primates (NHPs) using CRISPR-Cas9 lipid nanoparticles (LNPs). 2–3-year-old wild-type (WT) male cynomolgus NHPs were treated with 2 mg/kg of SpCas9 dual-gRNA LNPs targeting the LDLR gene, editing and disrupting LDLR in the liver. b Liver biopsy demonstrated editing of 68% of all LDLR alleles in a liver biopsy. c LDLR protein levels assayed by ELISA on a second liver biopsy were markedly reduced by 95%, and (d) blood low-density lipoprotein cholesterol (LDL-C) increased from ~50 mg/dL to ~300 mg/dL. LNPs without GalNAc-Lipid were not effective in LDLR-deficient NHPs, yielding (e) 4.5% ANGPTL3 editing and (f) minimal or no ANGPTL3 protein reductions at a 2 mg/kg dose. GalNAc-LNPs at a 2 mg/kg total RNA dose and with 0.05 mol % GL6 resulted in rescue of (e) high-efficiency liver ANGPTL3 editing and (f) durable 89% reduction in blood ANGPTL3 protein out to three months. In panel (e), each column represents an NHP, and editing results reflect liver biopsy (1-2) or necropsy (8) samples. In the homozygous familial hypercholesterolemia HoFH NHP model with markedly elevated baseline LDL-C levels of ~300 mg/dL, editing of ANGPTL3 with GalNAc-LNPs (g) lowered LDL-C (~35%, or ~100 mg/dL in absolute terms) out to three months. h Comparison of blood ANGPTL3 reductions out to six months in WT NHPs dosed with either standard LNPs or GalNAc-LNPs at 2 mg/kg. Data are presented as mean values + /- standard deviation (where error bars are present) and individual data points for each animal are displayed in panels (b–g). Panels (d), (f), and (g) have the means plotted as a line along with the individual points. Panel (h) averages three ANGPTL3 protein values from three distinct NHPs and the data are presented as mean values + /- standard deviation. Source data are provided as a Source Data file.

Fig. 4. Hepatic and extra-hepatic editing of ANGPTL3 following treatment with a GalNAc-LNP or standard LNP.

Targeted amplicon sequencing of the ANGPTL3 target site was performed in tissue samples collected at necropsy following dosing with standard lipid nanoparticles (LNPs) (N = 3) and GalNAc-LNPs (N = 3) at a 2 mg/kg dose. Biodistribution and liver editing of GalNAc-LNPs and standard LNPs is similar in wild type (WT) male cynomolgus NHPs, with little editing seen outside the liver for both LNPs. Each point represents results from an individual animal. LN denotes lymph node. Data are presented as mean values + /- standard deviation. Source data are provided as a Source Data file.