Mechanisms of palmitic acid-conjugated antisense oligonucleotide distribution in mice

Abstract

Conjugation of antisense oligonucleotide (ASO) with a variety of distinct lipophilic moieties like fatty acids and cholesterol increases ASO accumulation and activity in multiple tissues. While lipid conjugation increases tissue exposure in mice and reduces excretion of ASO in urine, histological review of skeletal and cardiac muscle indicates that the increased tissue accumulation of lipid conjugated ASO is isolated to the interstitium. Administration of palmitic acid-conjugated ASO (Palm-ASO) in mice results in a rapid and substantial accumulation in the interstitium of muscle tissue followed by relatively rapid clearance and only slight increases in intracellular accumulation in myocytes. We propose a model whereby increased affinity for lipid particles, albumin, and other plasma proteins by lipid-conjugation facilitates ASO transport across endothelial barriers into tissue interstitium. However, this increased affinity for lipid particles and plasma proteins also facilitates the transport of ASO from the interstitium to the lymph and back into circulation. The cumulative effect is only a slight (∼2-fold) increase in tissue accumulation and similar increase in ASO activity. To support this proposal, we demonstrate that the activity of lipid conjugated ASO was reduced in two mouse models with defects in endothelial transport of macromolecules: caveolin-1 knockout (Cav1-/-) and FcRn knockout (FcRn-/-).

Figure 1.

Conjugation of ASO with palmitic acid potentiates target mRNA knockdown in several tissues. A 3–10–3 PS cEt ASO (GCATTCTAATAGCAGC) targeted to MALAT1 or the same ASO conjugated at the 5′ terminus with palmitic acid was administered subcutaneously to male C57Bl/6 mice at concentrations ranging from 0.3 to 30 μmol/kg. Mice were terminated 72 h after administration and MALAT1 mRNA was extracted from liver, heart and quadricep tissues and quantified by qPCR. The ED50 is lowered 3–4-fold in heart and quadriceps.

Figure 2.

Conjugation of ASO with palmitate increases plasma and tissue AUC. Palm-ASO or ASO was administered subcutaneously to BL/6 mice at 7.5 μmol/kg and tissues collected from 0.5 to 24 h. (A) Palm-conjugated ASO has slower clearance from plasma, less excretion in urine, and greater accumulation in many tissues compared with unconjugated ASO. (B) Sections of quadricep muscle were stained with eosin (blue) and immunostained with rabbit anti-PS-ASO serum (brown). From 0.5 to 4 h, ASO staining appears darker in tissues from animals dosed with Palm-ASO compared with ASO and is concentrated along intercellular boundaries, suggesting ASO primarily occupies interstitial space or plasma membrane. At 8–24 h, punctate foci within the cells suggests internalization of the ASO.

Figure 3.

AUC of Palm-ASO is significantly increased for Palm-ASO in both plasma and lymph compared with ASO. Rats were administered 7.5 μmol/kg ASO or Palm-ASO via tail vein injection. Plasma was collected by tail snips at various time points from 5 to 360 min. Lymph was collected via a cannulated thoracic lymph duct at the same time points.

Figure 4.

Palmitate conjugation increases ASO affinity for several plasma proteins and lipid complexes. Binding of ASO and Palm-ASO to a set of human plasma proteins: albumin, transferrin (TF), IgG, fibrinogen, α-2-macroglobulin (A2M) and histidine-rich glycoprotein (HRG); and lipid particles: HDL and LDL. (A) FP binding curves for human serum albumin, HDL and HRG. (B) Ratio of K_ds ASO/ Palm-ASO. (C) Table of K_ds. (D) ASO and Palm-ASO binding profiles in human and mouse plasma by SEC.

Figure 5.

Activity of ASO and Palm-ASO is attenuated in Cav1^−/− and FcRn^−/- mice, but not Alb^−/− mice. (A) Activity of both ASO and Palm-ASO is attenuated in Cav1^−/− mice. The ED50 for both ASO is shifted in the Cav1^−/− mice with a more pronounced effect on Palm-ASO. This suggests both ASO are trafficked, at least in part, through Caveolin-1 dependent pathways. (B) Activity of ASO and Palm-ASO is attenuated in FcRn^−/− mice, with a more pronounced effect on Palm-ASO. (C) Activities of both Palm-ASO and ASO are unchanged in quadriceps of Alb^−/− mice compared to BL6 and slightly potentiated in heart and liver.

Figure 6.

Increased serum or albumin decreases in vitro activity of Palm- ASO. (A) HepG2 cells were treated with 2 or 10 μM ASO or Palm-ASO for 24 h in complete media (10% FBS), 5% FBS, 2% FBS or serum-free media (SFM). ASO activity was unaffected by changes in FBS in the media while Palm-ASO activity was dramatically increased with lower serum levels. (B) Similarly, supplementing SFM with increasing concentrations of BSA reduces activity of Palm-ASO but has no effect on the activity of ASO in HepG2 cells.

Figure 7.

Model for the transport of Palm-ASO from circulation to target cells. Palm-ASO bound to albumin is transported across continuous endothelium through Cav-1 and FcRn-dependent transcytosis. Once in the interstitium, Palm-ASO can be endocytosed by target cells along with albumin; dissociate from albumin and be endocytosed independently; or transported through lymph, back to circulation.