Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Site-specific drug delivery is a very important strategy for the optimization of drug therapy in terms of efficacy and safety since the pharmacological action of the drug of interest can be targeted to a specific site in the body. Among a variety of site-specific drug delivery methods, the use of a carrier system that is recognized by specific receptors on the target cells is one of the most powerful tools for targeted delivery of a variety of therapeutic agents, including chemotherapeutic compounds, protein drugs, antisense oligonucleotides, and genes (Takakura and Hashida, 1996; Wang and Low, 1998; Smith and Wu, 1999).

Scavenger receptors (SRs), which can recognize anionic macromolecules with unusually broad but circumscribed ligand specificity, are expressed on liver nonparenchymal cells (endothelial and Kupffer cells) and various macrophages (Linehan et al., 2000). The ligands for the SRs involve negatively charged proteins, such as maleylated and succinylated albumins, modified low-density lipoproteins, and polynucleotides (Terpstra et al., 2000). In particular, negatively charged albumins have been used as drug carriers via SRs. Successful receptor-mediated delivery to macrophages in vitro has been achieved with low-molecular-weight antitumor agents conjugated with maleylated albumin (Mukhopadhyay et al., 1992, 1995; Basu et al., 1994). Recently, maleylated albumin has also been used as a targeting carrier for a photosensitizer (Nagae et al., 1998) and a macrophage-activating peptide (Srividya et al., 2000a,b) via SRs. In addition, SR-mediated endocytosis has been used to deliver proteins as antigens. Several studies have shown that the immunological characteristics of proteins can be modulated by maleylation (Abraham et al., 1995, 1997; Singh et al., 1998; Bansal et al., 1999; Nicoletti et al., 1999). In addition, succinylated or other negatively charged albumins are interesting compounds since they exhibit antiviral effects (Jansen et al., 1993; Kuipers et al., 1997). Moreover, we have recently demonstrated that targeted delivery to liver nonparenchymal cells and improved therapeutic effects of catalase can be achieved by direct succinylation of the enzyme (Yabe et al., 1999).

In spite of the therapeutic potential of the delivery approach based on an SR-mediated mechanism, the relationship between the physicochemical characteristics of negatively charged proteins, such as the molecular weight and number of negative charges, and their in vivo pharmacokinetic profiles are not yet fully understood. Therefore, the purpose of the present study was to clarify this relationship to establish a strategy for the rational design of negatively charged proteins as drug carriers and for therapeutic purposes.

Three kinds of model proteins, recombinant superoxide dismutase (SOD), bovine serum albumin (BSA), and bovine IgG were selected and used to prepare succinylated derivatives with different degrees of modification. The pharmacokinetic characteristics of these succinylated proteins were studied after systemic administration of various doses to mice. The in vivo distribution properties of the succinylated proteins were analyzed and discussed in relation to their molecular characteristics as negatively charged proteins.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Male ddY mice (25-27 g) were purchased from the Shizuoka Agricultural Cooperative Association for Laboratory Animals (Shizuoka, Japan). Animals were maintained under conventional housing conditions.

Chemicals. BSA, IgG, fluorescein isothiocyanate (FITC), polyinosinic acid (poly I) and polycytidylic acid (poly C) were purchased from Sigma Chemical (St. Louis, MO). Rabbit anti-human von Willebrand factor antisera and rhodamine conjugated donkey anti-rabbit IgG were purchased from DAKO Japan (Kyoto, Japan) and Chemicon International, Inc. (Temecula, CA), respectively. Recombinant human SOD was supplied by Asahi Kasei (Tokyo, Japan). Succinic anhydride was obtained from Nacalai Tesque (Kyoto, Japan). [¹¹¹In]Indium chloride was supplied by Nihon Medi-Physics (Takarazuka, Japan). All other chemicals were obtained commercially as reagent-grade products.

Synthesis and Characterization of Succinylated Proteins. Succinylated proteins with various degrees of modification were synthesized by reacting different amounts of succinic anhydride with the -NH₂ group of the lysine residues of proteins according to a previously described method (Takakura et al., 1994). In brief, each protein was dissolved in 0.2 M Tris buffer, pH 8.65, and an appropriate amount of succinic anhydride was added. The mixture was stirred for 12 h at room temperature. The succinylated proteins were then washed by dialysis, concentrated by ultrafiltration, and lyophilized. The number of free amino acid groups was determined by trinitrobenzene sulfonic acid using glycine as a standard to estimate the degree of modification (Habeeb, 1966). The molecular masses of the succinylated proteins were estimated by SDS-PAGE. Myosin (mol. wt., 220,000), phosphorylase (mol. wt., 97,400), BSA (mol. wt., 66,000), ovalbumin (mol. wt., 46,000), carbonic anhydrase (mol. wt., 30,000), trypsin inhibitor (mol. wt., 21,500), and lysozyme (mol. wt., 14,300) were used as molecular weight markers. The homogeneity of synthesized proteins was ascertained from the results of SDS-PAGE. The bands of succinylated proteins were very sharp and shifted upward with increasing the degree of succinylation (data not shown). The SDS-PAGE analysis also showed that polymerization of synthesized proteins was negligible. The apparent surface density of the succinylated amino groups was determined by dividing the number of succinylated amino groups by the accessible surface area of the succinylated protein calculated from the following equation: (accessible surface area) = 6.3 (molecular mass)^0.73 (Miller et al., 1987). The electrophoretic mobility of the succinylated proteins was determined using a laser electrophoresis-zeta potential analyzer (LEZA-500T, Otsuka Electronics).

Radiolabeling and Fluorescein-Labeling of Succinylated Proteins. For the in vivo disposition experiments, proteins were radiolabeled with ¹¹¹In using the bifunctional chelating agent DTPA anhydride according to the method of Hnatowich et al. (1982). Briefly, each succinylated protein (5 mg) was dissolved in 1 ml of 0.1 M HEPES buffer, pH 7.0, and mixed with 2- or 3-fold molar amounts of DTPA anhydride in 20 µl of dimethyl sulfoxide. The mixture was stirred for 1 h at room temperature and purified by gel filtration, using a Sephadex G-25 column to remove the unreacted DTPA. The protein derivative fractions were collected and concentrated by ultrafiltration. Then, 20 µl of ¹¹¹InCl₃ solution (74 MBq/ml) was added to 20 µl of 1 M sodium acetate, and 60 µl of DTPA-coupled derivative solution was added to the mixture. After 30 min, the mixture was purified by gel filtration using a PD-10 column and by eluting with 0.1 M acetate buffer, pH 6.0. The derivative fractions were collected and concentrated by ultrafiltration. This radiolabeling method is suitable for examining the distribution phase of macromolecules from plasma to various tissues because any radioactive metabolites produced after cellular uptake are retained within the cells where the uptake takes place (Duncan and Welch, 1993; Arano et al., 1994). For confocal microscopic studies, Suc₅₄-BSA was labeled with FITC by the method of Monsigny et al. (1984). In brief, 1 µmol of Suc₅₄-BSA was dissolved in 10 ml of 0.1 M sodium carbonate buffer, pH 9.5, and then 3 µmol of FITC was added to the solution, followed by stirring for 5 h at room temperature. The mixture was purified by gel filtration using a PD-10 column, eluting with n-butanol/water (5:95), and dialysis and then concentrated by ultrafiltration and lyophilized.

In Vivo Disposition Experiments. ¹¹¹In-labeled succinylated proteins were injected into the tail vein of male ddY mice at doses of 0.1, 1, 10, and 20 mg/kg. At appropriate times after administration, blood was collected from the vena cava under ether anesthesia, and the mice were sacrificed. Heparin sulfate was used as an anticoagulant. Plasma was obtained from the blood by centrifugation. The kidney, spleen, liver, lung, and heart were removed, rinsed with saline, and weighed. The amount of ¹¹¹In radioactivity in urine was also determined by collecting the excreted urine and that remaining in the bladder. The radioactivity in each sample was counted using a well-type NaI-scintillation counter (ARC-500; Aloka, Tokyo, Japan). Contamination by plasma in each tissue sample was corrected using the distribution data for ¹¹¹In-BSA after intravenous injection, assuming that BSA was not taken up by the tissue during the 10-min period.

Confocal Microscopic Studies. To examine the cellular localization in the liver and kidney, FITC-Suc₅₄-BSA was administered intravenously to mice at a dose of 20 mg/kg. At 10 min after administration, mice were sacrificed; saline was infused via the portal vein to remove blood, and the liver and kidney were excised. Slices 5 µm thick were made, fixed with 20% formalin buffer. Slices were treated with RNase to avoid staining the cell components, except for the nucleus, and dyed with propidium iodide (PI) to visualize the nucleus. Liver and kidney endothelial cells were stained by a rabbit anti-human von Willebrand factor (vWF) antisera at 1:200, followed by a donkey anti-rabbit IgG rhodamine-conjugated second antibody. The slices were scanned with a confocal laser microscope (MRC-1024; Bio-Rad, Hercules, CA).

Calculation of AUC and Clearances. The plasma ¹¹¹In radioactivity concentrations were normalized with respect to the percentage of the dose per milliliter and analyzed using the nonlinear least-square program MULTI (Yamaoka et al., 1981). The tissue distribution patterns were evaluated using tissue uptake clearances according to the integration plot analysis. The tissue accumulation at time t was proportional to the AUC_0-t. By dividing the tissue accumulation at time t (X_t) and the AUC_0-t by the plasma concentration (C_t), CL_tissue was obtained from the slope of the plot of X_t/C_t versus AUC_0-t/C_t.

Pharmacokinetic Analysis Based on a Physiological Model. The time-courses of the plasma concentrations and liver accumulations of ¹¹¹In-labeled succinylated proteins were analyzed based on the physiological model shown in Fig. 1 (Nishikawa et al., 1995). In this model, the body is represented by three compartments [i.e., the plasma pool (PP), the sinusoidal and Disse spaces in the liver (EC), and the intracellular space in the liver (IC)]. The PP and EC compartments have apparent volumes of distribution V_p and V_l, respectively. The PP compartment represents all the plasma spaces within the blood vessels of tissues, except for the liver; it is connected with the EC by hepatic plasma flow. The uptake of succinylated protein from EC to IC is expressed as a saturable process exhibiting Michaelis-Menten kinetics, with a maximum rate of uptake, V_max,l (nanomoles per hour), and a Michaelis constant, K_m,l (nanomolar). Extrahepatic elimination from PP is assumed to be a saturable process represented by V_max,p or K_m,p. At time 0, the injected substance is assumed to be distributed in the PP and EC compartments at the same concentration. Mass balance equations for the concentration of succinylated proteins in PP and EC and the amount of succinylated proteins in IC are expressed as:

(1)

(2)

(3)

where C_p and C_l are the concentrations of succinylated proteins in PP and EC, respectively, X_l is the amount accumulated in IC, and Q is the hepatic plasma flow rate. The values of V_p, V_l, and Q were assumed to be 1.5 ml, 0.15 ml, and 85 ml/h, respectively (Gerlowski and Jain 1983).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Physiological pharmacokinetic model for analyzing the in vivo disposition of succinylated proteins. Km,p and Km,l are the Michaelis constants for plasma pool and liver, respectively. Vmax,p and Vmax,l are the maximum rates of uptake for plasma pool and liver, respectively. Q is the hepatic plasma flow rate.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Physicochemical Characteristics of Succinylated Proteins. The physicochemical properties of the succinylated proteins are summarized in Table 1. In each protein, the number of succinylated amino groups depended on the amount of reacted succinic anhydride. Succinylation slightly increased the molecular weight determined by SDS-PAGE. The calculated surface density of the succinylated amino groups correlated with the electrical mobility determined by the laser electrophoresis--potential analyzer (r² = 0.7652).

Distribution of ¹¹¹In-Succinylated Proteins after Intravenous Injection. Figure 2 shows the time course of the plasma concentrations, liver accumulation, and kidney accumulation of ¹¹¹In-Suc-SODs (A-C), ¹¹¹In-Suc-BSAs (D-I), and ¹¹¹In-Suc-IgGs (J-L), with different degrees of succinylation after intravenous injection in mice together with those of unmodified proteins. ¹¹¹In-SOD, ¹¹¹In-Suc₉-SOD, and ¹¹¹In-Suc₂₂-SOD rapidly disappeared from the plasma circulation in a similar manner but showed no significant accumulation in the liver after administration (Fig. 2, A-C). These profiles were independent of the injected dose. However, a marked difference was observed in their renal excretion, a major elimination pathway for small proteins. Unmodified SOD and Suc₉-SOD exhibited significant accumulation in the kidney, which decreased with a concomitant increase in urinary excretion (data not shown) as the dose increased. However, Suc₂₂-SOD was mainly recovered in the urine (30% of dose) with minimal renal uptake (approximately 15% of the dose), regardless of the injected dose.

View larger version (27K): [in this window] [in a new window]   Fig. 2.   Time-courses of plasma concentrations and liver and kidney accumulation of 111In-labeled SOD (A), Suc9-SOD (B), Suc22-SOD (C), BSA (D), Suc20-BSA (E), Suc28-BSA (F), Suc40-BSA (G), Suc46-BSA (H), Suc54-BSA (I), IgG (J), Suc22-IgG (K), and Suc50-IgG (L) in mice after intravenous injection of doses of 0.1 (), 1 (), 10 (), and 20 () mg/kg. Results are expressed as the mean ± S. D. of three mice.

Confocal Microscopic Studies. Figure 3 shows the confocal microscopic images of mouse liver (A and C) and kidney (B and D) 10 min after injection of FITC-Suc₅₄-BSA (20 mg/kg). The images in Fig. 3, A and B, are stained by PI, and the images shown in Fig. 3, C and D, are stained immunohistochemically by anti-vWF antibody. In Fig. 3A, FITC derived from Suc₅₄-BSA was localized in the cells with smaller nuclei along with the sinusoid not in the parenchymal cells with larger nuclei, indicating that FITC-Suc₅₄-BSA was selectively internalized by liver nonparenchymal cells, including endothelial and/or Kupffer cells. Immunohistochemical staining with endothelial cell-specific anti-vWF antibody showed that vWF positive cells are the major contributors for the uptake (Fig. 3C). On the other hand, vWF positive cells in the kidney were not responsible for the renal uptake of FITC-Suc₅₄-BSA (Fig. 3B and 3D). It seemed that strong fluorescein staining was observed in the luminal side of the proximal renal tubule, suggesting FITC-Suc₅₄-BSA underwent glomerular filtration and was taken up by the tubular cells, although the BSA derivatives cannot pass through the glomerular capillary wall due to size restriction.

View larger version (90K): [in this window] [in a new window]   Fig. 3.   Confocal microscopic images of liver (A and C) and kidney (B and D) sections of mice 10 min after intravenous injection of FITC-labeled Suc54-BSA at a dose of 20 mg/kg. Green signals are derived from FITC-Suc54-BSA and red ones from PI for nuclear staining (A and B) or from rhodamine for vascular endothelial cells (C and D). Yellow color indicates that FITC and rhodamine colocalize in the same site.

Competition of Hepatic and Renal Uptake of ¹¹¹In-labeled Suc₅₄-BSA by Preadministration of Poly I, Poly C, or Suc₅₄-BSA. To determine whether ¹¹¹In-labeled succinylated proteins are taken up by the liver and by the kidney via specific receptors, competition studies were performed using poly I, poly C, and Suc₅₄-BSA (Fig. 4). Hepatic uptake of ¹¹¹In-Suc₅₄-BSA was significantly inhibited by poly I and Suc₅₄-BSA but not by poly C (Fig. 4A), indicating that succinylated proteins could be taken up by the liver via SRs. On the other hand, renal uptake of ¹¹¹In-Suc₅₄-BSA was not obviously inhibited by poly I and poly C, and rather Suc₅₄-BSA increased the amount of Suc₅₄-BSA in the kidney (Fig. 4B). These results show that renal uptake of succinylated proteins was independent of scavenger receptor-mediated endocytosis.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Competition of hepatic and renal uptake of 111In-Suc54-BSA by preadministration of poly I, poly C, or Suc54-BSA after intravenous injection in mice. [111In]Suc54-BSA (0.1 mg/kg) was injected 1 min after injection of unlabeled competitive macromolecules (20 mg/kg). Results were expressed as mean ± S. D. of three mice. Statistically significant difference based on Student's t test compared with each control.

Calculation of AUC and Organ Uptake Clearances. For a quantitative comparison between the distribution profiles of the native and succinylated proteins, the clearance values for the total body (CL_total), liver (CL_liver), kidney (CL_kidney), and urine (CL_urine), as well as the AUC, were calculated based on the distribution data shown in Fig. 2 and summarized in Table 2. As far as Suc-SOD was concerned, the CL_total was almost independent of the injected dose, and CL_kidney and CL_urine clearances made a major contribution, although the CL_liver was slightly high in the case of Suc₂₂-SOD. These results suggested that glomerular filtration was a key pathway for the elimination of Suc-SOD after intravenous injection. The CL_kidney value of Suc₉-SOD was significantly higher than that of Suc₂₂-SOD, suggesting that Suc₉-SOD underwent more efficient tubular reabsorption. As the injected dose increased, the CL_kidney decreased and CL_urine increased, indicating that saturation of tubular uptake took place. On the other hand, in the case of Suc-BSA or Suc-IgG, the CL_total and CL_liver (the main contributor to the CL_total) varied depending on the degree of succinylation and the injected dose. At the lowest dose (0.1 mg/kg), Suc₄₆-BSA and Suc₅₄-BSA had a large CL_liver of about 70 ml/h, a value that is close to the hepatic plasma flow rate (85 ml/h) in mice, and the CL_liver value fell with a decrease in the number of succinylations per BSA molecule (48.0, 9.4, and 0.4 ml/h for Suc₄₀-BSA, Suc₂₈-BSA, and Suc₂₀-BSA, respectively), suggesting that the degree of succinylation determines the rate of hepatic uptake. On increasing the injected dose, the CL_liver value for each derivative significantly decreased. Regardless of their molecular weight being greater than the threshold of glomerular filtration, Suc-BSAs with a higher degree of succinylation showed relatively large CL_kidney values. At the higher doses (10 and 20 mg/kg), the CL_kidney values for Suc₄₆-BSA and Suc₅₄-BSA were comparable with the CL_liver, indicating that renal uptake plays an important role in the disposition of Suc-BSAs under these conditions. Although clearances were very small for Suc₂₂-IgG, the clearance values of Suc₅₀-IgG were similar to those of Suc₂₈-BSA.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Estimated surface density of succinylated amino groups × 103 (molecules/Å2). Hepatic (A) and renal (B) uptake clearances of 111In-labeled Suc-IgG (), Suc-BSA (), Suc-SOD (), and succinylated catalase () in mice after intravenous injection at a dose of 0.1 mg/kg.

Simulation Studies Based on a Physiological Model. Before determination of the pharmacokinetic parameters in the physiological model shown in Fig. 1, computer simulation studies were carried out to ascertain theoretically the effect of the pharmacokinetic parameters in the model on the relationship between the tissue uptake clearance and injected dose. As shown in Fig. 6, when the K_m value increased from 0.1 to 500 µg/ml, the absolute value of the tissue uptake clearance became lower, and the slope of the curve became more gentle with the threshold dose for rapidly decreasing clearance shifting to a lower dose. In contrast, when the V_max value increased from 1 to 200 µg/h, the absolute clearance value decreased, whereas the profile of the clearance as a function of the injected dose remained almost unchanged. These computer simulation results indicate that the dose dependence of the clearance profiles was significantly affected by K_m but not by V_max.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Simulation of the effect of pharmacokinetic parameters [i.e., Michaelis constant (A) and maximum rate (B)] on the relationship between the tissue uptake clearance and injected dose.

View larger version (23K): [in this window] [in a new window]   Fig. 7.   Hepatic and renal uptake clearances of 111In-labeled succinylated proteins after intravenous injection in mice at the doses of 0.1, 1, 10, and 20 mg/kg. A, IgG (), Suc22-IgG (), and Suc50-IgG (); B, BSA (), Suc20-BSA (), Suc28-BSA (), Suc40-BSA (), Suc46-BSA (), and Suc54-BSA ()

Pharmacokinetic Analysis of Distribution Profiles of Succinylated Proteins Based on a Physiological Model. Differential eqs. 1 to 3 were simultaneously fitted to the experimental data of the plasma concentrations and liver accumulation of each ¹¹¹In-Suc-BSA, and four parameters (K_m,l, V_max,l, K_m,p, and V_max,p) were estimated (Table 3). The estimated parameters properly describe the distribution profiles of Suc-BSAs (data not shown). Figure 8 illustrates the K_m,l and K_m,p values of all Suc-BSAs plotted against the surface density of the succinylated amino groups. These parameters, K_m,l and K_m,p, which correspond to the affinity of the succinylated proteins for hepatic uptake and extrahepatic elimination, respectively, correlated with the surface density of the succinylated amino groups (Fig. 8). With an increase in the surface density, K_m,l decreased from 2500 nM for Suc₂₀-BSA to 5.3 nM for the Suc₅₄-BSA, and K_m,p decreased from 24,000 nM to 230 nM. On the other hand, V_max,l and V_max,p remained relatively constant. The parameters of Suc₅₀-IgG were also plotted to examine the correlation for Suc-BSAs. These results suggested that the in vivo affinity of the succinylated proteins for SRs was closely related to the surface density of the succinylated amino groups on the modified proteins.

View larger version (23K): [in this window] [in a new window]   Fig. 8.   Relationship between Km,l (A), Km,p (B), Vmax,l (C), and Vmax,p (D) of 111In-labeled succinylated proteins and the degree of succinylation. Parameters of 111In-labeled Suc-BSA () and 111In-labeled Suc-IgG () were plotted against the estimated surface density of the succinylated amino groups.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In recent years, several new members of the SR family have been cloned on the basis of their ability to recognize modified lipoproteins, and the family has been divided into six classes (Terpstra et al., 2000). The distinct, but partly overlapping, binding properties of the SR classes represent a complication in defining their respective activity in terms of ligand uptake. Most SRs can bind a variety of polyanionic ligands, including negatively charged albumins. Since many of the SRs are expressed in sinusoidal endothelial cells and Kupffer cells in the liver, plural SRs may be responsible for the hepatic uptake of succinylated proteins. We and others have shown that highly succinylated albumins are extensively taken up by mouse and rat liver (Franssen et al., 1993; Jansen et al., 1993; Takakura et al., 1994; Furitsu et al., 1997; Kuipers et al., 1997). Although both Kupffer cells and liver endothelial cells were involved in the uptake, the endothelial cells made the greatest contribution. The hepatic uptake of Suc-BSA was significantly inhibited by maleylated BSA, dextran sulfate, and poly I but not by poly C. This inhibition profile was similar to that observed in class AI/AII SRs (SRA), the most well characterized SRs (Krieger, 1992; Gough and Gordon, 2000). For a rational design of succinylated proteins for SR-mediated targeting, it is necessary to understand the quantitative relationship between the physicochemical characteristics and the nonlinear in vivo pharmacokinetics. Although our previous studies suggested the importance of the molecular weight of proteins (Takakura et al., 1994; Furitsu et al., 1997), a detailed analysis of this remained to be carried out. In the present study, a systematic pharmacokinetic study was performed using a series of succinylated proteins with various molecular weights and degree of modification to achieve this aim.

In this study, we used ¹¹¹In-labeled succinylated proteins that are more suitable than those labeled with radioiodine to estimate the tissue distribution since the radioactivity remains within the lysosomes and is only slowly released from cells after receptor-mediated endocytosis. This property enables us to quantitatively determine the tissue distribution by counting radioactivity. However, the amount of ¹¹¹In radioactivity in tissues gradually decreased in some cases, especially the hepatic accumulation of highly succinylated proteins at lower doses (Fig. 2). Therefore, we used the distribution data showing no obvious decrease in radioactivity to calculate the organ uptake clearances and to fit the differential equations to the experimental data.

The present study has demonstrated that succinylated proteins are selectively taken up by liver nonparenchymal cells via SRs according to the degree of succinylation and the injected dose. Confocal microscopic studies, including immunohistochemical studies, showed a significant contribution of the liver nonparenchymal cells, especially endothelial cells, to the hepatic uptake of Suc-BSA (Fig. 3). Furthermore, involvement of the SR was confirmed by the competition experiments using poly I, a typical ligand for the receptor (Fig. 4). The hepatic uptake clearance correlated well with the estimated surface density of the succinylated amino groups, suggesting that a negative charge density is a critical factor for recognition by SRs. Further pharmacokinetic analysis showed that the affinity corresponded well to the negative charge density. Suc₂₂-SOD, however, failed to be taken up by the liver regardless of its high negative-charge density. Taken together, these results suggest that not only the negative-charge density, but also the molecular weight or size is important. These findings provide useful guidelines for the development of targeted delivery systems using succinylated proteins. To design a molecule for efficient SR-mediated hepatic targeting, a protein larger than BSA, having a succinylated amino group density of about 1.5 × 10³ molecules/Ų, should be used. Higher densities than this will not dramatically increase targeting efficacy. This information will be important when succinylation is applied to biologically active proteins, such as enzymes. Since unnecessary chemical modification sometimes impairs the activity, an appropriate minimum degree of modification should be selected. Retrospectively, these considerations are supported by our previous successful approach using Suc-catalase (Fig. 5A) (molecular mass, 250 kDa; 1.5 × 10³ succinylated molecules/Ų), with a relatively high remaining enzymatic activity. This compound has been shown to be effectively targeted to liver nonparenchymal cells and has important therapeutic potential in the treatment of hepatic injuries induced by ischemia/reperfusion (Yabe et al., 1999).

Previous studies have demonstrated that a charged collagen-like domain containing a lysine cluster of the SRA forms a positively charged groove that specifically interacts with negatively charged ligands (Doi et al., 1993). It has also been suggested that the spatial distribution of the negatively charged residues or the negative charge density of ligands plays an important role in electrostatic interactions (Pearson et al., 1993). Recently, Suzuki et al. (1999) proposed a hypothesis that ligand binding to SRs, sufficient to allow cellular uptake, requires not only a high density of negative charges but also an increase in the apparent affinity by numerous interactions between one ligand and multiple SR molecules. It is likely that larger succinylated proteins offer a better chance of multiple binding compared with smaller ones with a higher curvature. The low affinity of Suc-SODs for SRs might be supported by this hypothesis, that both a negative charge density and multiple binding would be a prerequisite for efficient recognition in vivo, assuming that SRA or other receptors with similar characteristics play a major role. The detailed molecular mechanisms await further investigation.

Suc-SOD significantly accumulated in the kidney. This, however, should be primarily ascribed to tubular reabsorption after efficient glomerular filtration of this small protein. Interestingly, the degree of succinylation significantly affected the renal handling of SOD. Efficient and saturable renal uptake was observed for Suc₉-SOD, whereas the renal accumulation of Suc₂₂-SOD was significantly lower. This finding, suggesting the importance of the free amino groups on protein derivatives in the uptake by renal tubular epithelial cells (Sumpio and Maack, 1982; Christensen et al., 1983), is in good agreement with our previous observations involving various chemically modified small proteins (Mihara et al., 1994).

Unexpectedly, we found marked renal accumulation of highly succinylated BSA, which is not susceptible to glomerular filtration due to their size under a normal physiological condition (Fig. 2), and our confocal microscopic studies revealed that Suc₅₄-BSA localized predominantly in the luminal side of the proximal renal tubule in the kidney (Fig. 3). This phenomenon was clearly observed only at higher doses in which the uptake via hepatic SRs was saturated and plasma concentrations were maintained for long periods. These results suggested that, following intravenous injection, the BSA derivative underwent glomerular filtration to a significant extent regardless of its large size, and subsequent uptake by the renal tubular epithelial cells (i.e., reabsorption) occurred in a similar manner to smaller proteins like SOD. It is reasonable that neither poly I nor poly C showed obvious inhibitory effects on renal accumulation of ¹¹¹In-Suc₅₄-BSA in the competition experiments (Fig. 4), assuming that the accumulation was ascribed mainly to protein reabsorption. Increased renal uptake of ¹¹¹In-Suc₅₄-BSA after an excess dosing of cold Suc-BSA in the same experiment also can be explained.

Although Kuipers et al. (1997) also reported that a certain amount of radioactivity was located in the kidney, the mechanism is unknown. It is postulated that glomerular permeability might be enhanced through an unknown action of highly succinylated protein in the kidney. Although the mechanism for this phenomenon is not clear, we speculate that mesangial cells might play an important role. It has been reported that negatively charged BSA enhances production of nitric oxide (NO) from macrophages (Alford et al., 1998), and following that, glomerular mesangial cell relaxation is enhanced by NO (Stockand and Sansom, 1997). Although further studies are needed, NO may be involved in the phenomenon that causes Suc-BSA to be passed through glomeruli and accumulated in the proximal renal tubule.

In conclusion, the present study has demonstrated that the hepatic uptake of succinylated proteins is determined by the affinity for SRs expressed on the liver nonparenchymal cells, and the affinity depends on the molecular size of the protein and the surface density of the succinylated amino groups of the protein. Furthermore, we have shown that highly succinylated proteins are also accumulated in the kidney probably, due to altered glomerular permeability. Thus, the present study has provided useful basic information for therapeutic strategies and the molecular design of succinylated proteins for use as drug carriers and therapeutic agents for SR-mediated targeting in vivo. Based on the finding, we are currently developing the targeted delivery system of antigen proteins through the SR-mediated endocytosis. To control antigen-specific immune responses by effective delivery of antigen, direct succinylation of the antigen and conjugation of epitope peptide derived from the antigen to Suc-BSA have been used.