Hepatocyte-specific distribution of catalase and its inhibitory effect on hepatic ischemia/reperfusion injury in mice

To explore the possibility of using catalase for the treatment of reactive oxygen species (ROS)-mediated injuries, the pharmacokinetics of bovine liver catalase (CAT) labeled with 111In was investigated in mice. At a dose of 0.1 mg/kg, more than 70% of 111In-CAT was recovered in the liver within 10 min after intravenous injection. In addition, 111In-CAT was predominantly recovered from the parenchymal cells (PC) in the liver. Increasing the dose retarded the hepatic uptake of 111In-CAT, suggesting saturation of the uptake process. This cell-specific uptake could not be inhibited by coadministration of various compounds which are known to be taken up by liver PC, indicating that the uptake mechanism of CAT by PC is very specific to this compound. The preventive effect of CAT on a hepatic ischemia/reperfusion injury was examined in mice by measuring the GOT and GPT levels in plasma. A bolus injection of CAT at 5 min prior to the reperfusion attenuated the increase in the levels of these indicators in a dose-dependent manner. These results suggest that catalase can be used for various hepatic injuries caused by ROS.

Characterization of plasmid DNA binding and uptake by peritoneal macrophages from class A scavenger receptor knockout mice

PURPOSE

Plasmid DNA (pDNA) has become an important class of macromolecular agent suitable for non-viral gene therapy as well as DNA vaccination. Our recent study has suggested that pDNA is taken up by mouse peritoneal macrophages via a specific mechanism mediated by a receptor similar to the scavenger receptor (SR). This study was designed to further characterize the pDNA uptake by macrophages in order to elucidate the mechanism.

METHODS

The binding and uptake of pDNA labeled with 32P or a fluorescent marker were studied in vitro using cultured Chinese hamster ovary (CHO) cells expressing the class A scavenger receptor (SRA) and peritoneal macrophages from SRA-knockout mice.

RESULTS

pDNA binding and uptake by CHO(SRA) cells were minimal and almost identical to that by wild-type CHO cells. Macrophages from the knockout mice showed pronounced pDNA binding and uptake as did the control macrophages. In both types of macrophage, pDNA binding was significantly inhibited by cold pDNA, polyinosinic acid and dextran sulfate but not by polycytidylic acid or Ac-LDL. These results provide direct evidence that SRA is not responsible for the significant binding and subsequent uptake of pDNA by mouse peritoneal macrophages. Further binding experiments revealed that, in addition to polyinosinic acid and dextran sulfate, heparin was a potent inhibitor among a variety of polyanionic compounds such as polynucleotides, anionic polysaccharides and modified proteins including Ox-LDL.

CONCLUSIONS

The present study suggest that pDNA binding and uptake by mouse peritoneal macrophages are mediated by a specific mechanism to some defined polyanions not by scavenger receptors. The finding would be an important basis for further studies to elucidate the mechanism(s) of pDNA uptake by macrophages.

Pharmacokinetics and disposition characteristics of recombinant decorin after intravenous injection into mice

The pharmacokinetics and disposition characteristics of recombinant decorin after intravenous administration were investigated in mice. Following bolus injection of 111In-labeled decorin at doses of 0.02 and 0.1 mg/kg, radioactivity rapidly disappeared from the circulation and approximately 70% of the dose accumulated in liver within 10 min. 111In-labeled decorin was preferentially localized in hepatic nonparenchymal cells. At a higher dose of 1 mg/kg, clearance from the circulation and hepatic uptake of [111In]decorin were slower than at lower doses. Both the accumulation in other tissues and urinary excretion of [111In]decorin were 5% or less. Pharmacokinetic analysis demonstrated that hepatic uptake clearance was large and accounted almost completely for total body clearance; in addition the clearance values decreased as the dose increased, suggesting that the hepatic uptake of decorin is mediated by a specific mechanism which becomes saturated at higher doses. In competitive inhibition experiments, hepatic uptake of 111In-labeled decorin was partially inhibited (about 20-30%) by several sulfated glycans such as glycosaminoglycans and dextran sulfate and by mannosylated bovine serum albumin (BSA), mannan and mannose to a lesser extent (about 10%). On the other hand, polyinosinic acid, polycytidylic acid and succinylated BSA were ineffective, suggesting that the scavenger receptor for polyanions in the liver is not involved in the hepatic uptake of decorin. A basic protein, protamine, and a ligand of the apoE receptor, lactoferrin, also had no effect. Taken together, the present results have demonstrated that recombinant decorin is rapidly eliminated from the blood circulation through extensive uptake by the liver, primarily by the nonparenchymal cells, following systemic administration. The sugar structure and mannose residue in decorin have also been suggested to play an important role in the hepatic uptake of decorin. These findings provide useful information for the development of decorin as a therapeutic agent.

Pharmacokinetic evaluation of biodistribution data obtained with radiolabeled proteins in mice

Radiolabeling of proteins is a widely used approach to study their in vivo disposition patterns. However, the obtained results may largely depend on the radiolabeling method used. The purpose of the present study was to investigate the effect of the radiolabeling method on the pharmacokinetic analysis of liver targeted protein in mice. Galactosylated bovine serum albumin (Gal-BSA) was labeled with 125I or 111In, using diethylenetriamine-pentaacetic dianhydride (cDTPA) or 1-(4-isothiocyanobenzyl)ethylenediaminetetraacetic acid (SCN-Bz-EDTA) as bifunctional chelating agents. The Gal-BSA was then injected in mice by a bolus intravenous injection. Samples of plasma, urine, liver, kidney, intestine and feces were collected at various time intervals and their radioactivity was measured. In none of the samples examined was there any significant difference in radioactivity distribution originating from the radiolabeling methods within 5 min after administration. After this period, 125I radioactivity in the liver started to decrease significantly faster than that of 111In, which would indicate the intracellular degradation of the protein. Consequently, the reappearance of trichloracetic acid (TCA) soluble 125I radioactivity in the plasma occurred. But whereas the hepatic uptake clearance (CLliver) of [111In]DTPA-Gal-BSA remained constant during 8 h postinjection, the CLliver of [125I]Gal-BSA at 30 min represented only one eighth of its initial values. The CLliver of [111In]SCN-Bz-EDTA-Gal-BSA resembled that of [111In]DTPA-Gal-BSA within 1 h of the experiment but it started to decline after this interval. The observed discrepancies most probably resulted from the formation of different radiolabeled metabolites in the hepatocytes and their different capability of crossing biological membranes. Our findings indicate that among the three methods employed, [111In]DTPA radiolabeling of Gal-BSA is the most appropriate method to study its tissue disposition.

Disposition characteristics of glycosylated poly(amino acids) as liver cell-specific drug carrier

Potentials of glycosylated poly(amino acids) as carriers of drugs and/or polynucleotides to the liver were studied in mice in detail. Poly-L-glutamic acid (PLGA) and poly-L-lysine (PLL) were selected as carrier backbones and modified with 2-imino-2-methoxyethyl 1-thiogalactoside or mannoside to obtain galactosylated and mannosylated derivatives. After intravenous injection in mice at a dose of 1 mg/kg, Gal-PLGA and Man-PLGA were selectively taken up by the liver parenchymal cells (PC) and liver nonparenchymal cells (NPC), respectively. Moreover, the uptake of Gal-PLGA and Man-PLGA by the liver were significantly inhibited by the presence of Gal-BSA and Man-BSA, respectively. On the other hand, PLL was targeted to the liver without glycosylation. However, Gal-PLL and Man-PLL showed higher accumulation in the liver than unmodified PLL. Although Gal-PLL and Man-PLL showed different distribution between PC and NPC from that of PLL, the effect of the modifications was less pronounced than the cases of PLGA derivatives. In addition, these glycosylated poly(amino acids) investigated, regardless of the type of amino acid or the sugar grafted, were rapidly degraded to be eluted in the low molecular weight fractions in a gel filtration chromatography. These results suggest that glycosylated PLGAs can be useful carriers of low molecular drugs to the liver cells through conjugation, while glycosylated PLLs could be targetable carriers to the cells after the reduction of their cationic charge by complex formation with polynucleotides.

Stimulation side-dependent asymmetrical secretion of poly I:poly C-induced interferon-beta from polarized epithelial cell lines

Mode of secretion of poly I:poly C-induced IFN was examined using epithelial cell lines in a bicameral culture system. Although the cell lines formed a tight cell sheet and produced IFN-beta following poly I:poly C treatment in spite of its application to the upper or lower compartment, IFN secretion differed between the apical and basolateral cell membranes. When poly I:poly C was applied to the upper compartment, IFN was secreted predominantly from the apical membrane. Inversely, poly I:poly C applied to the lower compartment caused preferential IFN secretion from the basolateral membrane. These results suggest that in epithelial cells poly I:poly C stimulation induces intracellular membrane traffic toward the stimulation side.

Synthesis and pharmacokinetics of a novel macromolecular prodrug of Tacrolimus (FK506), FK506-dextran conjugate

A novel macromolecular prodrug of Tacrolimus (FK506), FK506-dextran conjugate, was developed and its physico-chemical, biological and pharmacokinetic characteristics were studied. The conjugate was estimated to contain 0.45% of FK506 and the coupling molar ratio was approximately 1:1 (dextran-FK-506). Adsorption experiments using ion exchangers indicated that FK506-dextran conjugate acted as a weakly negatively charged macromolecule. Low molecular weight radioactive compound(s), which was eluted in the same fractions as [(3)H]FK506, was released from [(3)H]FK506-dextran conjugate by chemical hydrolysis with a half-life of 150 h in phosphate buffer. In vitro immunosuppressive activity of the conjugate, as assessed by the rat lymphocyte stimulation test, was almost comparable to that of free FK506, suggesting that biologically active FK506 could be liberated from the conjugate. In vitro biodistribution studies demonstrated that conjugation with the dextran derivative dramatically changed the pharmacokinetic properties of FK506 after intravenous injection in rats. AUC of the FK506-dextran conjugate was almost 2000 times higher than that of free FK506 and organ uptake clearances of the conjugate were significantly smaller than those of the free drug. Thus, the present study has demonstrated that the FK506-dextran conjugate behaves as a prodrug of FK506 with an extended blood circulating time and can be expected to have an improved therapeutic potency.

Gene expression and antitumor effects following direct interferon (IFN)-gamma gene transfer with naked plasmid DNA and DC-chol liposome complexes in mice

Gene expression was assessed in three types of mouse solid tumors after direct injection of naked plasmid DNA encoding the luciferase gene (pCMV-Luc) and its DC-chol liposome complexes. Intratumoral injection of 5 or 100 micrograms naked pCMV-Luc into subcutaneously inoculated mouse colon tumor (CT-26), fibrosarcoma (MCA-15) and bladder carcinoma (MBT-2) resulted in significant gene expression. A DC-chol liposome formulation (5 micrograms pCMV-Luc complexed with 25 micrograms DC-chol liposome) showed lower level of gene expression in the tumor models. Based on the results using the reporter gene, we examined the antitumor effect after direct interferon-gamma (IFN-gamma) gene transfer into CT-26 tumors. A significant IFN-gamma production and growth inhibition were obtained following direct intratumoral injection of IFN-gamma gene with naked plasmid DNA (pCMV-Mu gamma). Interestingly, pCMV-Mu gamma/DC-chol liposome complexes exhibited more pronounced growth inhibitory effect despite lower IFN-gamma production. Induction of CT-26 specific antitumor immunity by IFN-gamma gene transfer was confirmed by rejection of a CT-26 tumor challenge in the mice showing complete regression of CT-26 tumors after both treatments. Further analysis demonstrated that a significant cDNA-independent induction of IFN-beta and TNF-alpha occurred following injection with the liposome complexes, suggesting a nonspecific suppressive effect on CT-26 tumor growth by these cytokines. Thus, the present study has demonstrated that tumor tissue might be a promising target for direct IFN-gamma gene transfer with plasmid-based nonviral vectors. It is also suggested that immunomodulatory effects by various cytokines could be involved in antitumor effects after direct intratumoral injection of plasmid DNA formulations.

Hepatic disposition characteristics of recombinant human interleukin-11 (rhIL-11) in the perfused rat liver

The hepatic disposition characteristics of recombinant human interleukin-11 (rhIL-11) were investigated in perfused rat liver to clarify the mechanism of hepatic clearance which is a major contributor to the rapid clearance of rhIL-11 in vivo. We analyzed the disposition characteristics of [(111)In]-labeled rhIL-11 using a single-pass constant infusion mode at different concentrations of rhIL-11. The venous outflow rapidly reached a steady-state condition at every concentration. Liver extraction ratio at steady-state (Ess) was decreased with increase in the concentration, suggesting that there is a saturable interaction between the liver cells and rhIL-11 molecule. Cellular localization experiments demonstrated that rhIL-11 was taken up by both liver parenchymal and nonparenchymal cells depending on their surface area, suggesting that this uptake was mediated by electrostatic interaction due to cationic charges in the cytokine.

Asialoglycoprotein receptor-mediated gene transfer using novel galactosylated cationic liposomes

We synthesized three novel galactosylated cholesterol derivatives, cholesten-5-yloxy-N-(4-((1-imino-c-beta-D-thiogalactosyl+ ++-ethyl)amino) butyl)formamide (Gal-C4-Chol) and its ethyl formamide and hexyl formamide analogues (Gal-C2-Chol, Gal-C6-Chol), to prepare liposomal gene carriers possessing the cationic charge necessary for plasmid DNA binding and galactose residues as a targetable ligand for liver parenchymal cells. Liposome/DNA complexes prepared with these lipids showed low cytotoxicity in human hepatoma HepG2 cells. Gal-C4-Chol/DC-Chol/DOPE(3:3:4) liposomes, consisting of 3:3:4 mixtures of Gal-C4-Chol, 3beta[N',N', N'-dimethylaminoethane)-carbamoyl] cholesterol (DC-Chol), and dioleoylphosphatidylethanolamine (DOPE), showed higher transfection activity and [32P]DNA uptake than DC-Chol/DOPE(6:4) liposomes. The presence of 20 mM galactose significantly inhibited both transfection efficiency and uptake of DNA of Gal-C4-Chol/DC-Chol/DOPE(3:3:4) and Gal-C4-Chol/DOPE(6:4) liposomes, but not those of DC-Chol/DOPE(6:4) liposomes. These results indicate that the liposome/DNA complexes prepared using novel galactosylated cholesterol derivatives are efficiently recognized by asialoglycoprotein receptors and internalized and lead to gene expression. In addition, we found that the galactosylated cholesterol derivative with a longer spacer showed higher transfection activity.

Long-term results of Watson-Jones tenodesis of the ankle. Clinical and radiographic findings after ten to eighteen years of follow-up

Thirty-seven chronically unstable ankles in thirty-six patients were operated on with use of a Watson-Jones tenodesis. Thirty-four ankles (thirty-three patients) were followed for a mean duration of thirteen years and eight months (range, ten to eighteen years) after the operation. There were nine male and twenty-four female patients. The mean age of the patients was thirty-one years (range, fourteen to fifty-seven years) at the time of the operation and forty-four years (range, twenty-eight to seventy years) at the time of the latest follow-up. At the time of the most recent follow-up evaluation, twenty-seven patients (twenty-eight ankles) were examined directly by one of us and twenty-five patients (twenty-six ankles) also were evaluated radiographically. The other six patients were interviewed, with use of a questionnaire, by telephone. Of the thirty-four ankles, nineteen had an excellent result (grade 1), eleven had a good result (grade 2), three had a fair result (grade 3), and one had a poor result (grade 4) according to the rating system of Good et al. The mean score (and standard deviation) on the ankle-hindfoot scale of the American Orthopaedic Foot and Ankle Society for the twenty-eight ankles that were examined directly by one of us was 90 +/- 9.3 points (range, 68 to 100 points). Progression of an exostosis at the edge of the joint was detected in eighteen (69 percent) of the twenty-six ankles that were examined radiographically, but narrowing of the joint space was not seen in any ankle. No relationship was detected between the clinical results and radiographic osteoarthrotic changes or the duration of follow-up. The results did not deteriorate over the long term.

Cationic liposome-mediated efficient induction of type I interferons by a low dose of poly I:poly C in mouse cell lines

Double-stranded polyriboinosinic acid:polyribocytidylic acid (poly I:poly C) is a powerful inducer of type I interferons (IFNs). However, the dose of poly I:poly C required for efficient IFN induction is so high as occasionally to be cytotoxic. In this work, we examined the IFN-inducibility of poly I:poly C complexed with several cationic reagents in mouse fibroblast L cells and found that Lipofectin and LipofectACE can induce the production of a substantial amount of type I IFNs (mostly beta-type) even at a two-order lower dose compared with poly I:poly C alone. Such effects of poly I:poly C were optimal at 0.1 microgram/ml for 2-10 microgram/ml of Lipofectin and LipofectACE. These conditions caused no significant cytotoxicity in the recipient cells. Furthermore, a short treatment (less than 10 min) with the complexes was sufficient for the maximum induction. This IFN induction method was applicable to other cell types and other species including human. Hence, our observations may pave the way for clinical application of the IFN inducer.

Targeted delivery of plasmid DNA to hepatocytes in vivo: optimization of the pharmacokinetics of plasmid DNA/galactosylated poly(L-lysine) complexes by controlling their physicochemical properties

In vivo receptor-mediated targeting of plasmid DNA to hepatocytes was achieved through optimizing the physicochemical and pharmacokinetic properties of a plasmid DNA/carrier complex. Galactosylated poly(L-lysine) (Gal-PLL) was synthesized using PLL with a molecular weight of 1,800, 13,000 or 29,000 without loss of the cationic charge. Plasmid DNA encoding chloramphenicol acetyltransferase was complexed with each Gal-PLL. A larger amount of PLL1800 is required for the complex formation than with PLL13000 and PLL29000, and increasing the number of galactose units on Gal-PLL resulted in reduced binding to plasmid DNA. The particle size and zeta-potential of the complexes varied depending on the mixing ratio and Gal-PLL used. Then, plasmid DNA/Gal-PLL complexes having diameters of 200 nm or less and a weak negative charge were prepared. After i.v. injection of [32P]plasmid DNA/Gal13-PLL13000 and [32P]plasmid DNA/Gal26-PLL29000, almost 80% of the radioactivity rapidly accumulated in the liver, preferentially in the parenchymal cells. The hepatic uptake clearances (CLliver) were much greater than any of the other tissue uptake clearances. Compared with these complexes, [32P]plasmid DNA/Gal5-PLL1800 and [32P]plasmid DNA/Gal5-PLL13000 had a smaller CLliver, suggesting that both the molecular weight of PLL and the degree of galactose modification determine the hepatic targeting of plasmid DNA. In vitro and in vivo gene expression studies revealed that plasmid DNA/Gal13-PLL13000 and plasmid DNA/Gal26-PLL29000 complexes are superior to plasmid DNA/Gal5-PLL1800 complex for introducing DNA into cells. These results demonstrated that an optimal design of a DNA/carrier complex based on physicochemical properties and a pharmacokinetic analysis of the distribution properties leads to successful in vivo gene delivery.

Subtalar arthrography in acute injuries of the calcaneofibular ligament

We treated 43 acute tears of the calcaneofibular ligament by operation in 43 patients after subtalar arthrography. There were 22 men and 21 women with a mean age of 22.3 years (14 to 61). Anteroposterior (AP), lateral and oblique views were obtained with the foot in 45° of internal rotation and the ankle in the neutral position. Any communication or leakage to the ankle, tendon sheaths, subcutaneous tissue and sinus tarsi was recorded. We examined an oblique view of the microrecess along the interosseous ligament and an AP view of the lateral recess just under the distal end of the fibula. We also studied a control group of 27 patients with isolated injuries of the anterior talofibular ligament without rupture of the calcaneofibular ligament.

The findings in the two groups were significantly different when examined for leakage to the ankle (p = 0.0002), to the peroneal tendon sheaths (p = 0.0347) and to the subcutaneous tissue (p = 0.0222), absence of the microrecess (p = 0.0055) and presence of the lateral recess (p = 0.0012).

Many ankle sprains which involve tearing of the calcaneofibular ligament are accompanied by injuries of the subtalar joint. Combined injuries of the anterior talofibular ligament and calcaneofibular ligament, and isolated injury of the anterior talofibular ligament should be differentiated.

Subtalar arthrography in acute injuries of the calcaneofibular ligament

We treated 43 acute tears of the calcaneofibular ligament by operation in 43 patients after subtalar arthrography. There were 22 men and 21 women with a mean age of 22.3 years (14 to 61). Anteroposterior (AP), lateral and oblique views were obtained with the foot in 45 degrees of internal rotation and the ankle in the neutral position. Any communication or leakage to the ankle, tendon sheaths, subcutaneous tissue and sinus tarsi was recorded. We examined an oblique view of the microrecess along the interosseous ligament and an AP view of the lateral recess just under the distal end of the fibula. We also studied a control group of 27 patients with isolated injuries of the anterior talofibular ligament without rupture of the calcaneofibular ligament. The findings in the two groups were significantly different when examined for leakage to the ankle (p=0.0002), to the peroneal tendon sheaths (p=0.0347) and to the subcutaneous tissue (p=0.0222), absence of the microrecess (p=0.0055) and presence of the lateral recess (p=0.0012). Many ankle sprains which involve tearing of the calcaneofibular ligament are accompanied by injuries of the subtalar joint. Combined injuries of the anterior talofibular ligament and calcaneofibular ligament, and isolated injury of the anterior talofibular ligament should be differentiated.

Biological effects and cellular uptake of c-myc antisense oligonucleotides and their cationic liposome complexes

The biological effects and cellular uptake of human c-myc antisense oligonucleotides and their liposome complexes were investigated in vitro using human promonocytic leukemia U937 cells. Antisense phosphorothioate oligonucleotides (S-Oligo) significantly inhibited the growth of U937 cells in a dose-dependent manner. However, no significant effect on cell proliferation was observed with unmodified phosphodiester (P-Oligo) and partially phosphorothioated (PS3-Oligo) oligonucleotides with an antisense sequence and S-Oligo with sense and G-quartet control sequences. In cellular uptake experiments, radiolabeled S-Oligo was taken up by U937 cells more than P-Oligo and PS3-Oligo. Similar results were obtained in mouse peritoneal macrophages used for comparison. Confocal microscopic studies demonstrated a significant distribution of FITC-labeled oligonucleotides on the cell surface and in the cytoplasm in a punctate pattern, but not in the nucleus. When complexed with cationic liposomes, cellular uptake of FITC-labeled P-Oligo or S-Oligo was significantly increased and the fluorescence was located mainly in the nucleus, indicating that the uptake and intracellular pharmacokinetics of both oligonucleotides can be modified by complexation. An inhibitory effect of the complexes was observed at a dose which is ineffective in the case of the oligonucleotides alone. However, this effect was also associated with cytotoxicity of the cationic liposomes, suggesting that optimization of this formulation will be necessary to achieve a more efficient delivery of the oligonucleotides to U937 cells.

Histopathological study of nonosseous tarsal coalition

Histopathological analysis was performed on 55 feet in 48 patients with nonosseous tarsal coalitions. Histological findings were similar to those observed at the tendinous attachment site of Osgood-Schlatter disease, accessory navicular, and bipartite patellae. No nerve elements were observed in the fibrocartilaginous tissue at the coalition. Nerve elements were present only in periosteum and articular capsule surrounding the coalition. Pain in the tarsal coalition is not mediated by nerve elements at the coalition site itself. It is assumed that the pain is caused by mechanical abnormality that results from incomplete coalition. Incomplete coalition produces microfractures and remodelings on the boundaries between bone and the coalition, which then lead to degenerative changes. This mechanical abnormality seems to induce pain via free nerve endings in the periosteum and in the articular capsule surrounding the coalition.

Control of plasma cholesterol-lowering action of probucol with various lipid carrier systems

In order to explore the relationship between the pharmacokinetic properties and pharmacological actions of lipophilic drugs injected with lipid carrier systems, probucol was selected as a model drug with high lipophilicity, and the effect of disposition control on cholesterol-lowering activities was evaluated. Both large emulsion, with mean diameter of 280 nm, and long-circulating type small emulsion containing egg sphingomyelin with mean diameter of 100 nm, showed stable incorporation of probucol. The former produced rapid accumulation of probucol in the liver, while the latter demonstrated prolonged systemic circulation and gradual hepatic uptake. On the other hand, injection of a micellar solution with HCO-60 (polyoxyethylene hydrogenated castor oil) showed a rapid decrease in plasma concentration and a high hepatic uptake of probucol, similar to injections with serum, suggesting the rapid release of the drug from the micelles. However, probucol in a micellar solution showed higher cholesterol-lowering action than that in emulsion formulations. These results suggested that the pharmacological action of probucol in the liver might be affected by the uptake mode and sequential disposition in the organ, depending on the drug retention properties of the lipid carrier particles.

Targeted delivery of plasmid DNA complexed with galactosylated poly(L-lysine)

Galactose was introduced to poly(L-lysine) (PLL) with an average molecular weight of 13,000 to develop a hepatocyte-specific carrier for gene drugs. The pharmacokinetic characteristics of a model plasmid, pCAT (plasmid DNA encoding chloramphenicol acetyltransferase reporter gene), complexed with galactosylated PLL (Gal-PLL) was studied in mice in relation to its physicochemical properties. pCAT/Gal-PLL complex at a ratio of 1:0.6 (w/w) has a zeta potential of -20 mV and a mean particle size of about 180 nm. After intravenous injection, [32P]pCAT/Gal-PLL was rapidly eliminated from the circulation and preferentially taken up by the liver's parenchymal cells. The hepatic uptake of [32P]pCAT/Gal-PLL was significantly inhibited by prior administration of Gal-bovine serum albumin, suggesting that the uptake was mediated by the asialoglycoprotein receptors on hepatocytes. In vitro transfection experiments using a hepatoma cell line expressing the asialoglycoprotein receptor revealed that pCAT/Gal-PLL gave a high CAT gene expression whereas pCAT complexed with unmodified PLL failed to transfect the cells.

Involvement of specific mechanism in plasmid DNA uptake by mouse peritoneal macrophages

The binding and uptake of plasmid DNA encoding luciferase reporter gene (pCMV-Luc) were studied in vitro using cultured mouse peritoneal macrophages. A significant and time-dependent cellular association of [32P]pCMV-Luc with resident macrophages was observed at 37 degrees C and this decreased at 4 degrees C. The binding at 4 degrees C was saturable and a Scatchard plot gave a maximum binding capacity of 0.81 microgram/mg-protein and a dissociation constant of 0.30 microgram/ml. The binding of [32P]-pCMV-Luc was inhibited by polyinosinic acid, dextran sulfate and salmon sperm DNA, but not by polycytidylic acid, dextran and EDTA. A confocal microscopic study demonstrated that fluorescein-labeled pCMV-Luc was internalized at 37 degrees C while only cell surface binding occurred at 4 degrees C. No significant luciferase gene expression was obtained after incubation with a high concentration (100 micrograms/ml) of pCMV-Luc. These data suggest that plasmid DNA is taken up by macrophages via a mechanism mediated by a receptor like the macrophage scavenger receptor.