The cholesterol derivative of a triantennary galactoside with high affinity for hepatic asialoglycoprotein receptor: a potent cholesterol lowering agent

Cholesterol-derivatized galactosides have been devised in order to induce liver uptake of lipoproteins via the galactose-recognizing asialoglycoprotein receptor in the liver. In this study we describe the derivatization of a newly developed triantennary cluster galactoside having high affinity for the asialoglycoprotein receptor, N-[[tris-O-(3,6,9-trioxaundecanyl-beta-D-galactopyranosyl)metho xym ethyl] -N alpha-[1-(6-methyladipyl)]glycinamide (TG(20A)) with cholesterol. Hereto, TG(20A) was coupled to glycine-(5-cholesten-3 beta-yl ester) in the presence of (benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexafluorophosphate, affording N-[[tris-O-(3,6,9-trioxaundecanyl-beta-D- galactopyranosyl)methoxymethyl]methyl]-N alpha-[1-(6-(5-cholesten-3 beta-yloxy)glycyl)adipyl]glycinamide (TG(20A)C) in 46% yield. This compound is an amphiphilic, water-soluble compound. In aqueous solution it readily formed small micelles (4.9 +/- 1.2 nm) consisting of approximately 20 molecules. Upon incubation with human serum, TG(20A)C spontaneously incorporated into the most prominent serum lipoproteins, i.e., low-density lipoprotein (LDL) and high-density lipoprotein (HDL), thereby inducing an increase in buoyant density of these lipoproteins. The integrity of HDL and LDL, as judged from particle size analysis of both lipoproteins, was not altered by incubation with up to 0.33% of TG(20A)C (w/v). Following intravenous bolus injection into rats, TG(20A)C induced a dose-dependent decrease in the serum cholesterol content of maximally 44%, at a dose of 1.9 mg kg-1. This makes TG(20A)C at least 30-fold more effective than the previously developed N-[[tris-O-(beta-D-galactopyranosyl)methyl]methyl]-N alpha-[4-(5- cholesten-3 beta-yloxy)succinyl]glycinamide (TG(4A)C), provided with a cluster galactoside that displayed a 2000-fold lower affinity for the asialoglycoprotein receptor than TG(20A). In conclusion, the hypocholesterolemic activity of a cholesterylated galactoside can be strongly enhanced by using a cluster galactoside with higher affinity for the asialoglycoprotein receptor.

Synthesis of cluster galactosides with high affinity for the hepatic asialoglycoprotein receptor

High-affinity ligands for the asialoglycoprotein receptor, which is uniquely localized on the parenchymal liver cell and recognizes oligoantennary galactosides, might be utilized as homing device to specifically target drugs or genes to parenchymal liver cells. In the present study, the synthesis of galactose-terminated triantennary glycosides, provided with various spacers between the beta-galactopyranosyl moieties and the branching point of the dendrite, is described. N-[Tris[[(methylthio)methoxy]methyl]methyl]-N alpha-[1-(6- methyladipy)]glycinamide (3b) was glycosylated with monogalactosyl derivatives, containing propanediol or ethylene glycol units as hydrophilic spacer moieties, to yield the corresponding cluster galactosides. To determine the affinity of the cluster galactosides for the asialoglycoprotein receptor, we have performed competition studies of [125I]ASOR binding, a specific ligand for the asialoglycoprotein receptor, to isolated parenchymal cells. The affinity for the asialoglycoprotein receptor significantly increased with increasing spacer length. N-[[[Tris-O-(beta-D-galactopyranosyl)-3,6,9-trioxaunde- canoxy]methoxy]methyl]-N-alpha-[1-(6-methyladipyl)]glycinami de (4e), a cluster galactoside provided with a 20 A spacer, possessed an at least 2000-fold higher affinity for the receptor than N-[[tris-O-(beta-D-galactopyranosyl)methyl]methyl]-N alpha-[1-(6- methyladipyl)]glycinamide (4a), a cluster galactoside lacking the spacer. It is concluded that vicinal galactosyl moieties within a cluster galactoside are more optimal recognized by the galactose binding sites of the asialoglycoprotein receptor upon proper spacing. The most potent galactoside, TG(20A), may constitute an attractive targeting device for the specific delivery of drugs and/or genes to the parenchymal liver cell.

Cholesterol derivative of a new triantennary cluster galactoside lowers serum cholesterol levels and enhances secretion of bile acids in the rat

BACKGROUND

Previous studies have demonstrated that cholesterol-derivatized galactosides exert a hypocholesterolemic effect by inducing hepatic uptake of atherogenic lipoproteins by means of galactose-recognizing receptors in the liver. However, a prolonged infusion of high concentrations of these compounds was required for this effect, possibly because of low affinity for the galactose-recognizing asialoglycoprotein receptor on the parenchymal liver cell.

METHODS AND RESULTS

We have designed a new series of triantennary galactosides to optimize the affinity and specificity for this receptor. The affinity of a triantennary galactoside for the asialoglycoprotein receptor appeared to be dramatically enhanced by proper spacing of the three terminal galactose groups. In rats, a single injection of N-[tris-O-(3,6,9-trioxaundecanyl-beta-D-galacto- pyranosyl)methoxymethyl]methyl-N alpha-[1-(6-(5-cholesten-3 beta- yloxy)glycyl)adipyl]glycinamide [TG(20A)C], the cholesterol derivative of the most selective galactoside, causes a dose-dependent decrease of < or = 45% in the serum cholesterol concentration (P < .001). This decrease is mainly attributed to a decrease in the level of serum HDL (P = .0066) and, to a lesser extent, serum LDL (P = .036). In addition, TG(20A)C strongly enhances the bile-acid secretion in rats during the first 2 hours after administration, which indicates that TG(20A)C-induced clearance of cholesterol from the bloodstream is efficiently coupled to hepatic bile-acid secretion.

CONCLUSIONS

We conclude that TG(20A)C efficiently directs lipoproteins that contain cholesterol to the liver at a 30-fold-lower concentration than previously developed cholesterol-derived cluster galactosides. This newly developed approach to lower cholesterol levels may prove valuable for familial hypercholesterolemic patients or those with familial defective apolipoprotein B-100 who do not respond or who respond insufficiently, respectively, to conventional therapies.

Binding characteristics of scavenger receptors on liver endothelial and Kupffer cells for modified low-density lipoproteins

Previous studies showed that both endothelial and Kupffer cells contain specific recognition sites of oxidized low-density lipoprotein (OxLDL), in addition to recognition sites which recognize OxLDL and acetylated LDL (AcLDL). We have determined the binding characteristics of the recognition sites for OxLDL on Kupffer cells and endothelial cells (OxLDL-specific binding-site) in comparison to the recognition site for AcLDL on endothelial cells, which recognizes both AcLDL and OxLDL (Ac/OxLDL binding site). The capacity of Kupffer cells to bind OxLDL (Bmax. = 779 ng of 125I-OxLDL/mg of cell protein; Kd = 6 micrograms/ml) was comparable to the binding-capacity of endothelial cells (Bmax. = 803 ng of 125I-OxLDL/mg of cell protein; Kd = 5 micrograms/ml). The effect of net charge of modified LDL on its affinity for the recognition sites on Kupffer and endothelial cells was evaluated using competition studies. The affinity of AcLDL for the Ac/OxLDL binding site was greatly increased from 460 micrograms/ml to 4 micrograms/ml with increasing extent of modification and thus net charge. The Ac/OxLDL binding-site on endothelial cells also displayed an increased affinity towards LDL with an increasing degree of oxidation. The affinity of OxLDL for the Ac/OxLDL binding-site appeared to be about 4-fold higher than that of AcLDL with a similar extent of modification. At higher degrees of oxidation of LDL, the affinity for the OxLDL-specific site on endothelial and Kupffer cells was also strongly enhanced; the OxLDL-specific binding-site possesses a higher affinity for mildly oxidized LDL as compared with the Ac/OxLDL binding-site. It is concluded that recognition of OxLDL by both the OxLDL-specific binding-site and the Ac/OxLDL binding-site on liver endothelial and Kupffer cells depends on the net negative charge of modified LDL. The similarity in binding pattern of these binding sites makes it likely that the newly described 95 kD OxLDL binding protein on Kupffer cells [Y. B. De Rijke and Th. J. C. van Berkel, J. Biol. Chem. (1994), 269, 824-827] contains a recognition site with similar structural elements as described earlier for scavenger receptors.

Specific targeting of the antiviral drug 5-iodo 2'-deoxyuridine to the parenchymal liver cell using lactosylated poly-L-lysine

In this study, we describe the development and characterization of lactosylated poly-L-lysine as a potential carrier for targeting anti-viral drugs to the parenchymal liver cell. Poly-L-lysine (M(r) 38,000) was modified with 2 to 130 lactose residues per molecule poly-L-lysine. In vitro competition studies for the asialoglycoprotein receptor on parenchymal liver cells using 125I-asialoorosomucoid as radioligand revealed that mild modification of poly-L-lysine with only five lactose residues was sufficient for high affinity competition. In vivo studies showed that, after injection of poly-L-lysine modified with at least five lactose residues, about 70-80% of the injected dose was taken up by the liver. Preinjection of N-acetyl galactosamine almost completely blocked the hepatic uptake of lactosylated poly-L-lysine, indicating that galactose-recognizing receptors are involved. At 10 min following injection, the contribution of the various liver cell types to the hepatic uptake of lactosylated poly-L-lysine was determined; the parenchymal cell appeared to be responsible for more than 98% of the total liver uptake. To assess the applicability of lactosylated poly-L-lysine as an anti-viral drug carrier, it was derivatized with 4 to 15 residues of the antiviral drug 5-iodo 2'-deoxyuridine, 5'-monophosphate per molecule poly-L-lysine (4-16% by weight) via an acid-labile phosphamide bond. Maximally 0.7% of the conjugated 5-iodo 2'-deoxyuridine 5'-monophosphate was released after 1 h incubation of the drug/carrier conjugate with serum at 37 degrees C, thus establishing the stability of the conjugate in serum. The drug-carrier conjugate was rapidly cleared from the bloodstream within 1 min. Approximately 90% of the injected dose could be recovered in the liver. The parenchymal liver cell was responsible for 97% of the hepatic uptake. In vitro studies on the kinetics of endocytosis of lactosylated poly-L-lysine, derivatized with 5-iodo 2'-deoxyuridine 5'-monophosphate, by parenchymal liver cells revealed that the ligand was immediately internalized and, after a 10-min lag phase, deacetylated. Internalization and degradation did not occur in the presence of 100 mM N-acetyl galactosamine. In conclusion, the bioavailability of 5-iodo 2'-deoxyuridine 5'-monophosphate to the parenchymal liver cell is dramatically enhanced as a result of the conjugation of the anti-viral drugs to lactosylated poly-L-lysine. Accordingly, lactosylated poly-L-lysine constitutes a suitable carrier for targeting anti-viral drugs to the parenchymal liver cell.(ABSTRACT TRUNCATED AT 400 WORDS)

Cholesterol derivative of a new triantennary cluster galactoside directs low- and high-density lipoproteins to the parenchymal liver cell

We have developed a new triantennary galactoside, in which the terminal galactose moieties are connected to the branching point of the cluster galactoside via a 20 A (2 nm) spacer [TG(20A)]. In vitro binding studies have demonstrated that introduction of a 20 A spacer resulted in avid and specific binding of the triantennary galactoside to the asialoglycoprotein receptor on the parenchymal liver cell. Derivatization of this galactoside with a cholesterol moiety afforded a compound [TG(20A)C] that lowered the serum cholesterol concentration when injected into rats. In the present study we have evaluated the direct effect of TG(20A)C on the in vivo fate of high-density lipoprotein (HDL) and low-density lipoprotein (LDL). A direct association of TG(20A)C with HDL and LDL was observed on mixing these components. Incorporation of TG(20A)C into 125I-HDL and 125I-LDL significantly accelerated the serum decay and concomitantly stimulated the hepatic uptake of these lipoproteins in rats. The liver uptake of TG(20A)C-loaded 125I-HDL or 125I-LDL could be inhibited by 81% and 82% respectively by preinjection of 150 mg of N-acetylgalactosamine, indicating that the enhanced liver uptake proceeded via galactose-specific receptors. More than 96% of the hepatic uptake of TG(20A)C-loaded 125I-HDL could be attributed to the parenchymal cell. Surprisingly, the parenchymal cell also accounted for 93% of the liver association of TG(20A)C-loaded 125I-LDL, suggesting that TG(20A)C stimulates the uptake and processing of both lipoproteins by the asialoglycoprotein receptor on the parenchymal liver cell. This contrasts with earlier data indicating that a triantennary cluster galactoside provided with a 4 A spacer between the terminal galactose moieties and the branching point of the dendrite stimulated hepatic uptake of LDL via the Kupffer cells. The parenchymal cell is the only liver cell type that is capable of irreversibly removing cholesterol from the body in the form of bile acids. The above results imply that administration of TG(20A)C not only facilitates the hepatic uptake of lipoprotein-derived cholesterol (esters) but also their elimination from the body. In addition, it might be possible to utilize TG(20A)C as a targeting device to selectively deliver large drug carriers and possibly genes to the parenchymal liver cell.

Characterization of the interaction of galactose-exposing particles with rat Kupffer cells

The characteristics of the recognition system involved in the binding of galactose-exposing particles to freshly isolated rat Kupffer cells were determined. For this purpose we used iodinated lactosylated low-density lipoprotein (125I-Lac-LDL) as a ligand for the galactose receptor on Kupffer cells. The affinity of the binding of 125I-Lac-LDL to Kupffer cells was saturable (23,500 galactose-specific binding sites per cell) and of high affinity (2.4 +/- 0.3 nM). The order of potency of various carbohydrates in inhibiting the association of 125I-Lac-LDL with Kupffer cells was as follows: N-acetylgalactosamine > L-fucose >> N-acetylglucosamine/mannan. Association of 125I-Lac-LDL with Kupffer cells in the absence of Ca2+ was at the same level as in the presence of 50 mM N-acetylgalactosamine. A polyclonal antibody raised against the rat asialoglycoprotein receptor inhibited the binding of 125I-Lac-LDL to Kupffer cells and reacted in a Western blot with two proteins (molecular mass 88 and 77 kDa), which correspond to the molecular mass of the fucose receptor [Lehrman, Haltiwanger and Hill (1986) J. Biol. Chem. 261, 7426-7432]. Furthermore, the ability of fucosylated neoglycoproteins to displace 125I-Lac-LDL from Kupffer cells was equally dependent on the extent of fucosylation as previously reported for the fucose receptor. We conclude that the fucose receptor and not the C-reactive protein, as recently proposed [Kempka, Roos and Kolb-Bachofen (1990) J. Immunol. 144, 1004-1009], functions as the galactose-particle receptor on the Kupffer cell. The binding of galactose-exposing particles to the fucose receptor is a previously unknown property of this receptor.

Ligand size is a major determinant of high-affinity binding of fucose- and galactose-exposing (lipo)proteins by the hepatic fucose receptor

Previous in vivo studies have demonstrated that small galactose-exposing particles are preferentially internalized by the asialoglycoprotein receptor on the parenchymal liver cell and large particles by the galactose-particle receptor on the Kupffer cell. In this study, we have investigated using in vitro binding studies whether the affinity for either receptor is affected by the ligand size. The asialoglycoprotein receptor appeared to bind and process lactosylated proteins irrespective of their size. In contrast, recognition of galactose-exposing proteins by the galactose-particle receptor on the Kupffer cell was strongly dependent on size. The affinity increased 3000-fold with protein sizes increasing from 5 to 15 nm, reaching its maximum at approx. 1 nM for ligands larger than 15 nm. Apparently, the preferential in vivo uptake of large galactose-exposing ligands by Kupffer cells does not result from an inability of the parenchymal liver cells to internalize these ligands, but from the high affinity of large ligands for the galactose-particle receptor and the strategic anatomical localization of the Kupffer cells in the liver. In the preceding paper [Kuiper, Bakkeren, Biessen and Van Berkel (1994) Biochem. J. 299, 285-290] the galactose-particle receptor on the Kupffer cell was suggested to be identical with the fucose receptor. 125I-Lac-LDL-binding studies clearly showed that the galactose-particle receptor exhibited high-affinity binding of fucose-exposing proteins also. The affinity of fucosylated proteins for the galactose-particle receptor was greatly affected by ligand size. The above data strongly support the hypothesis that the galactose-particle receptor is identical with the fucose receptor. The size of neoglycoproteins can be appreciated as a new major determinant of affinity for the fucose receptor.

Partial purification of the 5-hydroxytryptamine-reuptake system from human blood platelets using a citalopram-derived affinity resin [corrected]

This paper describes a procedure for the synthesis and application of a citalopram-derived affinity resin in purifying the 5HT-reuptake system from human blood platelets. A two-step scheme has been developed for partial purification, based on wheat germ agglutinin-lectin (WGA) affinity and citalopram affinity chromatographies. Upon solubilization of the carrier with 1% digitonin, a 50-70-fold increase in specific [3H]imipramine binding activity with a 70% recovery could be accomplished through WGA-lectin chromatography. The WGA pool was then subjected to affinity chromatography on citalopram-agarose. At least 90% of the binding capacity adsorbed to the column. Specific elution using 10 microM citalopram resulted in a 22% recovery of binding activity. A 10,000-fold overall purification was obtained by using this two-step procedure. Analysis of the fractions on SDS-PAGE after 125I labeling revealed specific elution of 78- and 55-kDa proteins concomitant with the appearance of [3H]imipramine binding activity. The pharmacological profile of the partially purified reuptake system correlated well with that derived from the crude membrane-bound reuptake system, suggesting a copurification of the 5HT binding activity and [3H]imipramine binding activity.

Evidence for the existence of at least two different binding sites for 5HT-reuptake inhibitors within the 5HT-reuptake system from human platelets

Chemical modification procedures have been used to study the interaction of tricyclic and non-tricyclic 5HT-reuptake inhibitors with the [3H]imipramine binding site (IBS). N-Ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline (EEDQ) induced a pronounced loss in [3H]imipramine binding due to a reduction in Bmax. Preincubation with reuptake inhibitors and subsequent inactivation by EEDQ revealed that imipramine and 5HT prevented the EEDQ-induced inhibition, but citalopram and fluoxetine did not. Thiol modification studies demonstrated that reduction by dithiothreitol (DTT) enhanced the binding of [3H]imipramine by increasing the Bmax. The thioselective reagents 1,1-diazobis- (N,N-dimethylformamide) (diamide), phenyl-arsineoxide (PAO) and N-ethylmaleimide (NEM) attenuated the binding capacity by lowering the Bmax. PAO, a reversible thiol reagent, prevented NEM alkylation indicating that dithiols are involved in the NEM-induced inactivation. Binding of tricyclics or non-tricyclics prior to PAO inactivation revealed that tricyclics provide complete protection against thiol modification, while the non-tricyclics do not. The results support the hypothesis that the 5HT-reuptake system of human platelets possesses at least two distinguishable binding sites.