A series of glycolipids have been prepared which contain a cluster galactoside moiety with high affinity for the hepatic asialoglycoprotein receptor and a bile acid ester moiety which mediates stable incorporation into liposomes. Loading of liposomes with these glycolipids at a ratio of 5% (w/w) resulted in efficient recognition and uptake of the liposomes by the liver. Preinjection with asialofetuin almost completely inhibited the uptake, establishing that the liposomes were selectively recognized and processed by the asialoglycoprotein receptor on liver parenchymal cells. In contrast, a glycolipid content of 50% (w/w) led to a liver uptake that could not be inhibited by preinjection with asialofetuin, indicating that the liposomes were now processed by the Gal/Fuc-recognizing receptor on liver macrophages. The results presented in this study are important for future targeting of water-soluble and amphiphilic drugs, enveloped in these glycolipid-laden liposomes, to parenchymal liver cells.

Several studies have shown improved efficacy of cholesteryl-conjugated phosphorothioate antisense oligodeoxynucleotides. To gain insight into the mechanisms of the improved efficacy in vivo, we investigated the disposition of ISIS-9388, the 3'-cholesterol analog of the ICAM-1-specific phosphorothioate oligodeoxynucleotide ISIS-3082, in rats. Intravenously injected [(3)H]ISIS-9388 was cleared from the circulation with a half-life of 49.9 +/- 2.2 min (ISIS-3082, 23.3 +/- 3.8 min). At 3 h after injection, the liver contained 63.7 +/- 3. 3% of the dose. Compared to ISIS-3082, the hepatic uptake of ISIS-9388 is approximately 2-fold higher. Endothelial, Kupffer and parenchymal cells accounted for 45.7 +/- 5.7, 33.0 +/- 5.9 and 21.3 +/- 2.6% of the liver uptake of [(3)H]ISIS-9388, respectively, and intracellular concentrations of approximately 2, 75 and 50 microM, respectively, could be reached in these cells (1 mg/kg dose). Preinjection with polyinosinic acid or poly-adenylic acid reduced the hepatic uptake of [(3)H]ISIS-9388, which suggests the involvement of (multiple) scavenger receptors. Size exclusion chromatography of mixtures of the oligonucleotides and rat plasma indicated that ISIS-9388 binds to a larger extent to high molecular weight proteins than ISIS-3082. Analysis by agarose gel electrophoresis indicated that ISIS-9388 binds more tightly to plasma proteins than ISIS-3082. The different interaction of the oligonucleotides with plasma proteins possibly explains their different dispositions. We conclude that cholesterol conjugation results in high accumulation of phosphorothioate oligodeoxynucleotides in various liver cell types, which is likely to be beneficial for antisense therapy of liver-associated diseases.

Anti-sense oligodeoxynucleotides (ODNs) hold great promise for correcting the biosynthesis of clinically relevant proteins. The potential of ODNs for modulating liver-specific genes might be increased by preventing untimely elimination and by improving the local bioavailability of ODNs in the target tissue. In the present study we have assessed whether the local ODN concentration can be enhanced by the targeted delivery of ODNs through conjugation to a ligand for the parenchymal liver cell-specific asialoglycoprotein receptor. A capped ODN (miscellaneous 20-mer sequence) was derivatized with a ligand with high affinity for this receptor, N2-[N2-(N2,N6-bis{N-[p-(beta-d-galactopyranosyloxy) anilino] thiocarbamyl}-L-lysyl)-N6-(N-{p-[beta-D -galactopyranosyloxy] anilino} thiocarbamyl)-L-lysyl]-N6-[N- (p-{beta-D-galactopyranosyloxy}anilino)thiocarbamyl]-L-lysine (L3G4) (Kd 6.5+/-0.2 nM, mean+/-S.D.). Both the uptake studies in vitro and the confocal laser scan microscopy studies demonstrated that L3G4-ODN was far more efficiently bound to and taken up by parenchymal liver cells than underivatized ODN. Studies in vivo in rats showed that hepatic uptake could be greatly enhanced from 19+/-1% to 77+/-6% of the injected dose after glycoconjugation. Importantly, specific ODN accumulation of ODN into parenchymal liver cells was improved almost 60-fold after derivatization with L3G4, and could be attributed to the asialoglycoprotein receptor. In conclusion, the scavenger receptor-mediated elimination pathway for miscellaneous ODN sequences can be circumvented by direct conjugation to a synthetic tag for the asialoglycoprotein receptor. In this manner a crucial requisite is met towards the application of ODNs in vivo to modulate the biosynthesis of parenchymal liver cell-specific genes such as those for apolipoprotein (a), cholesterol ester transfer protein and viral proteins.

The high expression level of receptors for low-density lipoprotein (LDL) on tumor cells makes LDL an attractive carrier for selective delivery of drugs to these cells. The aim of this study is to allow incorporation of oncogene-directed antisense oligodeoxynucleotides (ODNs) into the lipid moiety of LDL. Therefore, ODNs were conjugated with oleic acid, cholesterol, and several other steroid lipids. These latter steroid lipids were synthesized starting from bile acids and were varied in lipophilicity by attaching oleic acid ester structures. The lipid structures, activated as pentafluorophenyl esters, were conjugated in solution phase to 3'-amino-tailed ODNs. The ODNs conjugated with lithocholic acid, oleic acid, and cholesterol could easily be purified by reversed phase (RP)-HPLC. However, the ODNs conjugated with the oleoyl steroid ester structures irreversibly bound to the column material. These highly lipidic ODNs were separated from the nonconjugated ODN by electrophoresis in a 1% low-melting agarose gel containing 0.1% Tween 20. This method was found to be very effective in isolating the ODNs conjugated to the oleoyl steroid ester structures. The ODNs conjugated with cholesterol and the oleoyl esters of lithocholic and cholenic acid associated readily and nearly completely with LDL. However, the less lipidic ODNs and the ODN conjugated with the dioleoyl ester of chenodeoxycholic acid did not and did incompletely associate, respectively. Lithocholic acid and oleic acid are probably not sufficiently lipophilic to induce association with LDL, whereas the dioleoyl ester structure is probably too bulky and extended to allow partitioning into the lipid moiety of LDL. We conclude that several of the lipid-ODNs can associate readily with LDL, enabling delivery of oncogene-directed antisense ODNs via the LDL receptor pathway.

No abstract

Our aim is to selectively deliver 9-(2-phosphonylmethoxyethyl)adenine (PMEA) to parenchymal liver cells, the primary site of hepatitis B virus (HBV) infection. Selective delivery is necessary because PMEA, which is effective against HBV in vitro, is hardly taken up by the liver in vivo. Lactosylated reconstituted high-density lipoprotein (LacNeoHDL), a lipid particle that is specifically internalized by parenchymal liver cells via the asialoglycoprotein receptor, was used as the carrier. PMEA could be incorporated into the lipid moiety of LacNeoHDL by attaching, via an acid-labile bond, lithocholic acid-3␣-oleate to the drug. The uptake of the lipophilic prodrug (PMEA-LO) by the liver was substantially increased after incorporation into LacNeoHDL. Thirty minutes after injection of [3 H]PMEA-LO-loaded LacNeoHDL into rats, the liver contained 68.9% ؎ 7.7% of the dose (free [ 3 H]PMEA, <5%). Concomitantly, the uptake by the kidney was reduced to <2% of the dose (free [ 3 H]PMEA, >45%). The hepatic uptake of PMEA-LO-loaded LacNeoHDL occurred mainly by parenchymal cells (88.5% ؎ 8.2% of the hepatic uptake). Moreover, asialofetuin inhibited the liver association by >75%, indicating uptake via the asialoglycoprotein receptor. The acid-labile linkage in PMEA-LO, designed to release PMEA during lysosomal processing of the prodrug-loaded carrier, was stable at physiological pH but was hydrolyzed at lysosomal pH (half-life, 60 to 70 min). Finally, subcellular fractionation indicates that the released PMEA is translocated to the cytosol, where it is converted into its active diphosphorylated metabolite. In conclusion, lipophilic modification and incorporation of PMEA into LacNeoHDL improves the biological fate of the drug and may lead to an enhanced therapeutic efficacy against chronic hepatitis B.

Anti-sense oligodeoxynucleotides (ODNs) hold great promise for correcting the biosynthesis of clinically relevant proteins. The potential of ODNs for modulating liver-specific genes might be increased by preventing untimely elimination and by improving the local bioavailability of ODNs in the target tissue. In the present study we have assessed whether the local ODN concentration can be enhanced by the targeted delivery of ODNs through conjugation to a ligand for the parenchymal liver cell-specific asialoglycoprotein receptor. A capped ODN (miscellaneous 20-mer sequence) was derivatized with a ligand with high affinity for this receptor, N2-[N2-(N2,N6-bis{N-[p-(beta-d-galactopyranosyloxy) anilino] thiocarbamyl}-L-lysyl)-N6-(N-{p-[beta-D -galactopyranosyloxy] anilino} thiocarbamyl)-L-lysyl]-N6-[N- (p-{beta-D-galactopyranosyloxy}anilino)thiocarbamyl]-L-lysine (L3G4) (Kd 6.5+/-0.2 nM, mean+/-S.D.). Both the uptake studies in vitro and the confocal laser scan microscopy studies demonstrated that L3G4-ODN was far more efficiently bound to and taken up by parenchymal liver cells than underivatized ODN. Studies in vivo in rats showed that hepatic uptake could be greatly enhanced from 19+/-1% to 77+/-6% of the injected dose after glycoconjugation. Importantly, specific ODN accumulation of ODN into parenchymal liver cells was improved almost 60-fold after derivatization with L3G4, and could be attributed to the asialoglycoprotein receptor. In conclusion, the scavenger receptor-mediated elimination pathway for miscellaneous ODN sequences can be circumvented by direct conjugation to a synthetic tag for the asialoglycoprotein receptor. In this manner a crucial requisite is met towards the application of ODNs in vivo to modulate the biosynthesis of parenchymal liver cell-specific genes such as those for apolipoprotein (a), cholesterol ester transfer protein and viral proteins.

The coupling of a cholesterol ester analog to daunorubicin results in a lipophilic prodrug that can be firmly anchored into lipidic carries. LAD-loaded emulsions and liposomes provided with recombinant apoE will be tested in the near future for their ability to deliver LAD to tumor tissue in vivo via the LDL receptor.