Search

Genes can be targeted to hepatocytes in vitro and in vivo by the use of asialoorosomucoid-polylysine conjugates. After systemic application, this nonviral vector is recognized by highly selective asialoglycoprotein (AsGP) receptors on the sinusoidal liver cell membrane and is taken up via receptor-mediated endocytosis. As most of the DNA is rapidly transferred to lysosomes where it is degraded, transfection efficiency is low and gene expression transient. To address this problem, we incorporated a pH-dependent synthetic hemolytic peptide derived of the G-protein of Vesicular Stomatitis Virus (VSV) into the gene transfer system, to increase endosomal escape of internalized DNA. The multicomponent carrier binds DNA in a nondamaging way, is still recognized by the AsGP receptor, and is targeted to the liver in vivo. Injection of DNA complexes containing a luciferase marker gene resulted in luciferase expression of 29 000 pg/g liver which corresponded to an increase of a factor of 10(3) overexpression after injection of DNA complexes without endosomolytic peptide. Furthermore, the amount of intact transgene within isolated liver cell nuclei was increased by a factor of 10(1)-10(2) by the use of the multicomponent carriers. These results demonstrate that incorporation of a hemolytic peptide into a nonviral vector can greatly increase gene expression while retaining cell type targetability in vivo.

The application of gene therapy to liver disease is contingent on the development of an effective gene delivery vehicle. Receptor-mediated endocytosis can be exploited as a means of selective and efficient targeting of gene therapy vectors to hepatocytes. DNA-binding conjugates have been directed to the liver by the attachment of asialoglycoproteins or other ligands for receptors expressed on hepatocytes. Recent studies suggest refinements in this approach through which high transduction rates in vitro may be reproduced in vivo. The intrinsic liver tropism of viral vectors and liposomes can be augmented by the addition of targeting features, as demonstrated in animal models. With further modification, such as the incorporation of hepatotropic elements of the hepatitis viruses or lipoproteins, the next generation of delivery systems may achieve efficient, persistent expression of therapeutic genes in a safe and cell type-specific manner.

DNA delivered to the liver by asialoglycoprotein receptor-mediated endocytosis is degraded in lysosomes within 48 h. To test the hypothesis that microtubular disruption should promote transgene persistence by interrupting endosomal translocation to lysosomes, plasmids containing bacterial chloramphenicol acetyltransferase (pSV2-CAT) or human bilirubin-UDP-glucuronosyltransferase-1 (pSVK3-hBUGT1) genes were complexed with asialoglycoprotein-polylysine conjugates, and 1 mg of the complexed DNA was injected intravenously into bilirubin-UDP-glucuronosyltransferase-deficient Gunn rats. 30 min before DNA injection, one group received 0.75 mg of colchicine/kg of body weight intraperitoneally, which was shown by immunofluorescent confocal microscopy to disrupt the microtubular network. Control rats received normal saline. In colchicine-pretreated rats receiving pSV2-CAT, hepatic chloramphenicol acetyltransferase activity persisted for 9-14 weeks, whereas in the saline-pretreated group the activity was detectable for 48 h only. In colchicine-pretreated Gunn rats receiving pSVK3-hBUGT1, the DNA persisted in liver for 10 weeks, bilirubin glucuronides were excreted in bile, and serum bilirubin levels declined by 25-35% in 2-4 weeks and remained reduced for 8 weeks. Without colchicine pretreatment, the DNA was detectable in liver for 2 days only, and serum bilirubin levels were not reduced. Thus, microtubular disruption provides a noninvasive method for prolonging the effect of liver-targeted gene therapy.

Asialoglycoproteins (ASG) are internalized by hepatocytes by ASG receptor (ASGR)-mediated endocytosis. We have shown previously that when a plasmid DNA, pAlb(9-12)CAT (expressing chloramphenicol acetyltransferase driven by an albumin promoter enhancer), was complexed with an ASG-polylysine conjugate and injected intravenously in rats, 80% of the DNA was internalized by the liver. In normal recipient rats, over 95% of the internalized DNA was degraded in 4 h; the plasmid was undetectable after 48 h. In contrast, when 66% hepatectomy was performed 20 min after DNA administration, the internalized DNA persisted for several weeks in cytoplasmic vesicles (Chowdhury, N. R., Wu, C. H., Wu, B. Y., Yerneni, P. C., Bommineni, V. R., and Chowdhury, J. R. (1993) J. Biol. Chem. 268, 11265-11271). Since microtubules are required for the translocation of ligand-containing endosomes to lysosomes, the site of ligand degradation, we hypothesized that persistence of the endocytosed DNA might be related to changes in microtubular structure and function. To test this hypothesis, we examined hepatocellular microtubules by immunofluorescence confocal microscopy. Liver from untreated rats or sham-operated controls showed a network of fibrillar microtubules throughout the cytoplasm. The extent of the microtubular network was substantially reduced 3-6 h after 66% hepatectomy. By 24 h, microtubules had regenerated. Intraportal infusion of cycloheximide (250 mg/kg body weight) 15 min before 66% hepatectomy, prevented microtubular disruption, indicating that protein synthesis is required for this process. Immunotransblot analysis showed that hepatic alpha-tubulin concentration remained unchanged through microtubular disassembly and subsequent reassembly, which is consistent with conservation and reutilization of tubulin released by depolymerization of microtubules.

After intravenous injection, DNA complexed with asialoglycoprotein-polylysine conjugates is endocytosed by hepatocytes via asialoglycoprotein receptors and is expressed transiently. Long term persistence and expression occurs when partial hepatectomy is performed after gene delivery. To determine the intracellular location of the persisting DNA, we transferred a plasmid expressing bacterial chloramphenicol acetyltransferase into the liver of rats in vivo by asialoglycoprotein receptor-mediated endocytosis. The internalized DNA was measured by Southern blot. Twenty min after administration, 80-85% of the plasmid appeared in the liver, 80% of which was within hepatocytes (12,000-18,000 copies/hepatocyte). In sham-operated control rats, the transgene concentration decreased to 8-12 and 2-4% of the initial levels in 4 and 24 h, respectively, and became undetectable at 7 days. In rats subjected to 66% hepatectomy 20 min after DNA administration, 20, 9, and 7% of the plasmid in the residual liver persisted at 4 h, 24 h, and 7 days, respectively. Liver homogenates were fractionated by differential centrifugation and Percoll gradient centrifugation. In 66% hepatectomized rats, the plasmid persisted in an undegraded, transfection-competent form in plasma membrane/endosome-enriched fractions throughout the duration of the experiment (7 days), indicating that cytoplasmic vesicles are the main site of persistence of the endocytosed DNA.

Procollagen propeptides have been previously shown to be taken up by cultured fibroblasts and to inhibit collagen production at translational and transcriptional levels. In hepatic fibrosis, Ito cells are thought to be major contributors to collagen overproduction. We wondered whether an impaired response to procollagen propeptides by Ito cells from fibrotic liver could be a mechanism involved in collagen overproduction in hepatic fibrosis. In the presence of type I procollagen carboxy-terminal propeptide (200 nM), collagen production by Ito cells from normal rats was significantly decreased to 62.0% of control, whereas collagen production by Ito cells from two models of hepatic fibrosis in the rat, bile duct ligation and carbon tetrachloride treatment, was not significantly changed (102.4 and 102.6% of controls, respectively). By measurement of type I procollagen mRNA levels, this difference in response to procollagen carboxy-terminal propeptide was determined to occur at a pretranslational level. The rates of uptake, transfer to the nuclear compartment, and degradation of procollagen propeptide were not significantly different in Ito cells from normal and fibrotic liver.

Infusion of hepatocyte-specific DNA-protein complexes into rats leads to transient recombinant gene expression in liver. The eventual deterioration of gene expression is due in part to instability of the targeted DNA. In a previous report, we noted retention of transgene sequences in liver and persistent recombinant gene expression when the animals were subjected to partial hepatectomy following in vivo gene transfer. In an attempt to define the mechanism(s) responsible for persistent gene expression following partial hepatectomy, we characterized the molecular state of the retained, liver-associated transgenes. Southern blot analysis of DNA from liver tissues harvested various times after in vivo gene transfer and partial hepatectomy (10 min to 11 weeks) demonstrated high levels of transgene DNA (100-10,000 copies/cell). The predominant form of this DNA appeared to be episomal based on analyses of uncut DNA or DNA restricted by an endonuclease with one site in the plasmid. Livers from several animals contained a small proportion of transgene sequences of unknown structure. The existence of episomal DNA in liver was confirmed in experiments in which intact plasmid was rescued from total hepatocyte DNA by transformation of bacteria. Both strands of DNA in the liver-associated plasmid retained a bacterial pattern of methylation suggesting that the plasmid had not replicated in the eukaryotic cell. These results are consistent with the hypothesis that the majority of transgene sequences are retained as stabilized plasmids. The specific form of DNA which is transcriptionally active was not identified in these studies. This represents a new mechanism for retaining foreign DNA in eukaryotic cells in vivo and has implications both for the development of somatic gene therapies and the pathogenesis of viral diseases.

Familial hypercholesterolemia is an inherited disease in humans, caused by a deficiency of low density lipoprotein (LDL) receptors, that we have used as a model for developing liver-directed gene therapies. Our strategy is to reconstitute hepatic LDL receptor expression in vivo by administering a DNA-protein complex that is capable of targeting the delivery of functional LDL receptor genes to hepatocytes. Infusion of this DNA-protein complex into the peripheral circulation of a rabbit animal model for familial hypercholesterolemia resulted in hepatocyte-specific gene transfer and a temporary amelioration of hypercholesterolemia. This noninvasive approach to gene therapy should have applications in the treatment of a wide spectrum of human diseases.

Effects of extracellular pH, CO2, and HCO3- on ketone body production from octanoate were studied in perfused livers from fasted rats. pH was adjusted to 7.1-7.5 by varying perfusate [HCO3-] and [CO2], where brackets denote concentration. At constant 25 mM [HCO3-], total production of beta-hydroxybutyrate (beta-OHB) + acetoacetate (AcAc) was constant from pH 7.1 to 7.5. However, the [beta-OHB]/[AcAc] ratio decreased from 1.60 to 1.00 when pH decreased from 7.3 to 7.1; there was no change at pH 7.4. At constant [CO2], decreasing pH from 7.4 to 7.1 did not alter either total ketogenesis or the [beta-OHB]/[AcAc] ratio. This suggests that high [CO2] rather than low pH was responsible for the alteration in the redox ratio. At constant pH of 7.4, variations in [HCO3-] between 15 and 25 mM did not influence total ketogenesis or the [beta-OHB]/[AcAc] ratio. However, increasing [HCO3-] from 25 to 35 mM decreased the [beta-OHB]/[AcAc] ratio from 1.76 to approximately 1, again without affecting total ketogenesis. At constant 1.75 mM [CO2], increasing [HCO3-] from 25 to 35 mM also reduced the [beta-OHB]/[AcAc] ratio from 1.63 to approximately 1, suggesting that the effect of high [HCO3-] on this redox ratio can be ascribed to HCO3- itself. It is concluded that high [CO2] or [HCO3-] decreases the mitochondrial [NADH]/[NAD+] ratio in hepatocytes, resulting in a decreased [beta-OHB]/[AcAc] ratio.

We present an in vivo model for specific protection of normal hepatocytes from damage by the highly specific hepatotoxin galactosamine. The idea is based on the fact that normal, unlike malignant, hepatocytes possess unique cell-surface receptors that can bind and internalize galactose terminal (asialo)glycoproteins by receptor-mediated endocytosis. A targetable carrier-antagonist conjugate was formed by coupling asialofetuin to the galactosamine antagonist uridine monophosphate. Intravenous injection of the antagonist conjugate resulted in specific uptake by the liver. Rats treated with carrier-antagonist conjugate together with a toxic dose of galactosamine developed significantly less hepatotoxicity than did controls. We conclude that a galactosamine antagonist can be targeted to liver, resulting in specific protection of hepatocytes from galactosamine toxicity in vivo. Because hepatoma cells lack asialoglycoprotein receptor activity, this "targeted rescue" may be of value in the differential protection of normal cells in the treatment of hepatocellular carcinoma.

Hydroxylysine residues are important structural components of collagen as sites of glycosylation and cross-linking. Because hydroxylysine is formed from lysine by post-translational enzymatic hydroxylation, we examined the effects of incorporation of a lysine analogue that could interfere with hydroxylation. For this purpose, we synthesized 35S-labeled 2-aminoethyl cysteine. By addition of this compound to a smooth muscle cell culture, we obtained radiolabeled collagen which, upon acid hydrolysis, yielded the intact labeled analogue, thus demonstrating incorporation of the compound into collagen. Using a model hepatic collagen synthesizing system, freshly isolated granulomas derived from mouse livers after 8 weeks of infection with Schistosoma mansoni, incubation over a period of 7 days with [14C]proline to measure collagen production, demonstrated a reduction in acid-precipitable [14C]hydroxyproline per milligram DNA in granuloma cultures incubated with 2-aminoethyl cysteine compared to control cultures, and the effect was dose-related. That this decrease was selective for collagen accumulation over noncollagenous proteins was demonstrated by the finding of a significant reduction in the [14C]hydroxyproline/[14C] proline ratios compared to controls. A maximum incorporation of 10 residues of analogue/1000 residues was achieved. While the extent of hydroxylation of proline in collagen remained unchanged, hydroxylation of lysine decreased with increasing amounts of analogue incorporated. We conclude that in our system, 2-aminoethyl cysteine is incorporable into collagen. In a dose-related fashion, it selectively diminishes collagen accumulation compared to control.

Intracellular degradation of newly synthesized collagen is studied by incubating cells with (14C)proline and measuring the hydroxy(14C)proline in a low molecular weight fraction relative to total hydroxy-(14C)proline synthesized. This communication describes a fundamental limitation on the precision with which the measurement can be made; the limitation derives from an irreducible uncertainty in estimating the background radioactivity contributed by the isotope used for metabolic labeling. The problem is illustrated by experiments with cultured liver cells. In rat hepatocytes, the hydroxy-(14C)proline in the low molecular weight fraction was barely detectable above background; the degradation was 29%, however the uncertainty was +/- 28%. In contrast, significant amounts of hydroxy(14C)proline were detected in fractions from human and rat hepatoma cells; and while the levels of degradation were approximately the same as in rat hepatocytes, the measurements could be made with substantially greater precision.

We have taken advantage of the presence of hepatic receptors for galactose-terminal (asialo-)glycoproteins to achieve targeted rescue of differentiated hepatocytes from acetaminophen-induced toxicity in vitro. To accomplish this, a conjugate was formed by covalent coupling of N-acetylcysteine (an acetaminophen antagonist) to galactose-terminal (asialo-)fetuin. We used two human hepatocyte-derived cell lines to test our targeted-rescue method: Hep G2 cells are capable of receptor-mediated endocytosis of galactose-terminal glycoproteins and PLC/PRF/5 cells are not. In the presence of acetaminophen alone, both cell lines demonstrated a similar concentration-dependent sensitivity. Growth rates of both cell lines became normal when N-acetylcysteine was administered in equimolar quantities with acetaminophen indicating that both cell lines had the potential of responding to the antagonist. When asialofetuin-N-acetylcysteine conjugate was given to both cell lines in the presence of acetaminophen, PLC/PRF/5, receptor (-) cells failed to respond. However, Hep G2, receptor (+) cells treated with asialofetuin-N-acetylcysteine conjugate under identical conditions, increased their populations and eventually reached confluence. Control conjugate fetuin-N-acetylcysteine as well as asialofetuin alone had no effect on either cell line.

From liver of dogs injected iv with potassium dichromate (38 mg/kg body wt), a low-molecular-weight chromium-binding substance (LMCr) was purified into two subfractions, LMCr I and LMCr II, which differ in physical and chemical properties. LMCr I was identified to be an anionic, organic chromium compound with a molecular weight of 1500. It contained glutamic acid, glycine, and cysteine as the predominant amino acids and firmly bound chromium in a ratio of one chromium(III) to one molecule of LMCr I. LMCr II was isolated in crystalline form and demonstrated to be a water-soluble, inorganic chromium(III) complex consisting of Na2HPO4 . 7H2O and Na2HPO4 . 2H2O. Although its crystallization reduced the chromium content, it had a maximum chromium-binding capacity as much as one chromium per one phosphorus in water. The mixture of LMCr I and LMCr II as approximated to be the natural composition showed a lower acute toxicity as measured by lethality in mice and had higher rates of urinary excretion and renal clearance in rabbits, accompanied by lower rates of renal tubular reabsorption and retention in kidney and liver than potassium dichromate(VI) and chromium(III) chloride. Pretreatment with chromium-free LMCr II remarkably reduced the mortality rates of mice acutely poisoned with chromium chloride. These results indicate that LMCr plays an important role in the detoxification and excretion of chromium in mammals.

We have developed the methodology for evaluating the effects of pathophysiological conditions on the molecular mechanisms of hepatic protein synthesis and fibrogenesis in baboons and man. Total RNA was extracted from percutaneous liver biopsies of five baboons who were chronically fed an ethanol-rich liquid diet, their pair-fed controls and from humans with a variety of liver abnormalities. Chronic alcohol administration in baboons with liver fibrosis and normal serum albumin levels increased in vitro protein synthesis as measured by [35S]methionine incorporation, albumin mRNA content and Type I procollagen mRNA content. There was no difference in the beta-actin (a constitutive protein) mRNA content. In humans, serum albumin levels correlated with albumin mRNA content as indicated by the intensity of dot blot hybridization and Type I procollagen mRNA levels correlated with the activity of liver fibrosis. The use of RNA-DNA hybridization to investigate procollagen mRNA from human biopsies appears to be a valuable tool for evaluating the potential for collagen synthesis and the future course of liver disease. Besides the use of RNA-DNA hybridization, we describe other methodologies which are useful in delineating the levels of gene expression responsible for hepatic mRNA regulation in normal liver and disease states in man. The use of molecular techniques to evaluate human liver disease provides an opportunity to develop clinically relevant information while at the same time offering the additional advantage of providing fundamental knowledge about fibrogenesis.

We utilized the presence of hepatic receptors for galactose-terminal glycoproteins to achieve specific rescue of differentiated hepatocytes from methotrexate toxicity. This was accomplished by administration of a conjugate formed by covalent coupling of the methotrexate antagonist, folinic acid (leucovorin), to galactose-terminating (desialylated) bovine fetuin. Two cultured hepatocyte-derived cell lines were chosen to test rescue from methotrexate toxicity: Hep G2 cells are capable of receptor-mediated endocytosis of galactose-terminating glycoproteins; PLC/PRF/5 cells are not. In the presence of methotrexate alone, both cells lines were reduced in number to 20% of control populations by day 5. Growth rates of both lines returned to normal when free folinic acid was added to the medium containing methotrexate. When asialofetuin-folinic acid conjugate was given to both cell lines in the presence of methotrexate, Hep G2 cells doubled their numbers by day 4 and reached control (without methotrexate) populations by day 8. However, PLC/PRF/5 cells failed to respond to administered asialofetuin-folinic acid conjugate under identical conditions, and the growth curve was the same as that for cells which received methotrexate alone. Conjugates of asialofetuin bound to folic acid (folate derivative requiring dihydrofolate reductase activity for conversion to an active metabolite) and fetuin (non-galactose-terminal glycoprotein) linked to folinic acid also were made for control studies. These conjugates failed to rescue either cell line.

We present evidence that normal hepatocytes can be specifically protected from galactosamine toxicity in vitro by targeting an antagonist to these cells via receptor-mediated endocytosis. The strategy is based upon the following principles: 1) galactosamine is a highly selective hepatotoxin that causes a dose-dependent depletion of uridine intermediates; 2) galactosamine toxicity can be antagonized by supplemental administration of uridine; 3) normal hepatocytes possess unique cell-surface receptors that can internalize galactose terminal (asialo-)glycoproteins with subsequent degradation of the glycoprotein ligand. Based on these facts, we hypothesized that chemical coupling of a galactosamine antagonist to an asialoglycoprotein could result in cell-specific delivery and protection of normal hepatocytes by targeting the antagonist via asialoglycoprotein receptors. Using a model system consisting of freshly isolated rat hepatocytes (receptor (+)) and Morris 7777 rat hepatoma (receptor (-)) cells, sensitivity to galactosamine in vitro was determined and found to be similar for both types of cells. A targetable antagonist was synthesized by coupling uridine monophosphate to asialoorosomucoid in a molar ratio of 5 to 1. Exposure of Morris 7777 cells to the targetable antagonist in the presence of a toxic concentration of galactosamine did not protect these cells as evidenced by a steady decline in the number of viable cells in a fashion identical to cells treated with galactosamine alone. However, normal hepatocytes that received the conjugate in the presence of galactosamine were protected as their viable cell number remained the same as control (untreated) cells. Competition by an excess of asialoglycoprotein inhibited the protective effect of the conjugate, supporting the concept that the asialoglycoprotein component of the conjugate was responsible for the specific delivery of the antagonist to the target cells.

We present evidence that a foreign gene driven by natural mammalian regulatory elements can be targeted to hepatocytes and the resultant gene expression made to persist. This was accomplished using a soluble DNA carrier system consisting of two covalently linked components: 1) a polycation, poly-L-lysine, that can bind DNA in a strong but non-damaging interaction, and 2) an asialoglycoprotein which can be targeted specifically to hepatocytes by cell surface asialoglycoprotein receptors unique to this cell type. A plasmid, palb-CAT, containing the gene for chloramphenicol acetyltransferase (CAT) driven by mouse albumin regulatory sequences was complexed to the carrier system. Intravenous injection of palb-CAT DNA in the form of a complex resulted in the presence of CAT enzyme activity in liver homogenates 24 h after injection. The targeted gene expression, however, was transient, reaching a maximum of 10 units/g liver at 24 h but was not detectable by 96 h. However, partial hepatectomy 30 min after injection resulted in persistent high levels of hepatic CAT activity (11.3 units/g) through 11 weeks post-injection. Southern analysis of livers 11 weeks after partial hepatectomy demonstrated that some of the targeted DNA had been integrated into the host genome. We conclude that a foreign gene driven by natural mammalian regulatory elements can be delivered to hepatocytes by intravenous injection in vivo using a soluble DNA carrier system. Foreign gene expression targeted in this manner can be made to persist by stimulation of hepatocyte replication.