Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

A number of methods have been proposed and are currently under investigation for developing tissue-specific targeting vectors for several types of the drugs (Meijer and van der Sluijs, 1989; Kato and Sugiyama, 1997). Polypeptide ligands, antibodies, viruses, fatty acids, and sugar moieties have all been used as vector systems. Drug delivery targeted to "sugar-recognizing" receptors such as asialoglycoprotein and mannose receptors has been demonstrated to lead to the highly efficient delivery of various types of molecules, including low molecular weight therapeutic agents, proteins, and genes to the liver (Spanjer et al., 1985; Nishikawa et al., 1995; Takakura and Hashida, 1996; Wu et al., 1997; Zanta et al., 1997). In addition, several methods have been shown to deliver therapeutic agents to the kidney. These include a delivery system using reabsorption systems expressed on the brush-border membrane for macromolecular proteins and peptides after they have undergone glomerular filtration (Franssen et al., 1992; Haas et al., 1993, 1997). The prodrug system that is activated by kidney-specific enzymes has also been investigated (Elfarra et al., 1995; Kearney, 1996). However, compared with liver-targeting vectors, only a limited amount of information is available on systems that target the kidneys.

Suzuki et al. (1999b) recently found that arginine-vasopressin (AVP), when modified by linking it to a sugar via an octamethylene group, exhibits renal-selective and efficient association in rats. Further investigations revealed that 1) the alkylglucoside structure (Glc-S-C8-, Fig. 1) is necessary for recognition by the kidney, and n-octyl-thioglucoside (OTG, Fig. 1) is also distributed specifically to the kidney; 2) the affinity of an alkylglucoside vector for its specific binding site on the kidney membrane depends on the type of sugar moieties involved, the length of the alkyl chain, the structure of the peptide, and type of linkage between the sugar and alkyl chain; 3) the distribution of alkylglucoside derivatives to the kidney is much higher than can be explained by glomerular filtration alone, and saturation of such distribution occurs over a similar concentration range to that found for binding to the membrane, suggesting that specific binding site(s) are involved in the renal distribution (Suzuki et al., 1999a,c). Actually, the cross-linking of ¹²⁵I-labeled alkylglycoside to the renal plasma membrane revealed the presence of a binding protein with a molecular mass of 62 kDa (Watanabe et al., 2000); and 4) accumulation of this vector in the kidney takes place mainly at the proximal tubules (Suzuki et al., 1999b,c).

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Structures of the alkylglucoside vector and its derivatives.

Considering these results, this alkylglucoside structure seems to be a very useful kidney-targeting vector. However, only small molecules, such as AVP, oxytocin derivatives, tryptamine, and 4-nitrobenz-2-oxa-1,3-diazole, have been used as model ligands for derivatization with this vector to examine the biodistribution and specific binding to the kidney membranes (Suzuki et al., 1999b,c). Therefore, little information is available on the limitations of this vector in terms of the physicochemical properties of the ligands derivatized with alkylglucoside. The purpose of the present study is to characterize the targeting efficiency and limitations of this vector system. To examine the effect of the size of the ligands that are derivatized with alkylglucoside, we synthesized conjugates of Glc-S-C8- and acylated poly-L-lysine (APL) with three different molecular weights (mol. wt., Fig. 1) and investigated the distribution and specific binding in the kidney of these conjugates. To examine the effect of the charge on the ligands derivatized with alkylglucoside, Glc-S-C8- was conjugated with tyrosine derivatives having anionic, neutral, or cationic charges (Fig. 1).

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. Thioglucoside derivatives were synthesized by Meiji Seika (Kanagawa, Japan) as follows: the carboxyl residue of the D-tyrosine was conjugated with an amino residue of ethylenediamine, and the other amino residue of the conjugate was condensed with carboxyl residue of 9-(1-thio--D-glucopyranosyl)nonanoic acid synthesized by the reported method (Suzuki et al., 1999b) and the obtained conjugate was named compound X. The acylation of poly-L-lysines with mol. wt. of 1,000 to 4,000, 4,000 to 15,000, and 15,000 to 30,000 was performed by a reported method (Gonsho et al., 1994). The carboxyl terminal of each of the three kinds of APL was condensed with the amino residue of compound X and the obtained conjugates were named Glc-S-C8-APL4500, Glc-S-C8-APL17000, and Glc-S-C8-APL41000. The number in these abbreviations represents the mean mol. wt., assuming that all the lysine residues in the poly-L-lysines were acylated. Glc-S-C8-Tyr-Base, Glc-S-C8-Tyr-Neutral, and Glc-S-C8-Tyr-Acid were synthesized as follows: the carboxyl residue of 9-(1-thio--D-glucopyranosyl)nonanoic acid was conjugated with the amino residue of D-tyrosine (Glc-S-C8-Tyr). Glc-S-C8-Tyr-Base, Glc-S-C8-Tyr-Neutral, and Glc-S-C8-Tyr-Acid were obtained by conjugating carboxyl residue of Glc-S-C8-Tyr with the amino residue of ethylenediamine, propylamine, and -alanine, respectively. Glc-S-C8-Ala-VP and Glc-S-C8-AVP were synthesized as described previously (Suzuki et al., 1999b). Na[¹²⁵I] (98.5 mCi/ml) was purchased from Amersham Pharmacia Biotech UK, Ltd. (Little Chalfont, Buckinghamshire, UK). [¹⁴C]Inulin was purchased from PerkinElmer Life Science Products (Boston, MA). Soluene 350 and HIONIC Flour were purchased from Packard Instrument Co. (Douwners Grove, IL). Sep-Pak(C₈) was purchased from Waters (Milford, MA). OTG was purchased from Wako (Osaka, Japan). Centrifree and Microcon were purchased from Millipore Corporation (Bedford, MA).

Animals. Male Sprague-Dawley rats were purchased from Nihon Ikagaku (Tokyo, Japan) and used at 6 to 7 weeks of age. Food and water were available ad libitum. The studies reported in this article have been carried out in accordance with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health.

¹²⁵I-Labeling of Alkylglucoside Derivatives. Alkylglucoside derivatives were labeling by the chloramine-T method. Briefly, 20 µl of alkylglucoside derivative in PBS (50 nM) was mixed with 50 µl of 0.5 M phosphate buffer, pH 7.5 and then 1 µl of Na[¹²⁵I] was added. Following this, 10 µl of chloramine-T (2 mg/ml dissolved in PBS) was added to the reaction mixture and mixed using a vortex mixer for 30 s. After mixing, 50 µl of Na₂S₂O₅ (2.5 mg/ml in PBS) and 10 µl of potassium iodine (100 mg/ml in PBS) were added to stop the labeling. To exclude free Na[¹²⁵I], the reaction mixture was applied to a Sep-Pak(C₈) column and sequentially washed with 0.1% TFA and 15% CH₃CN in 0.1% TFA. Labeled ligand was eluted with 60% CH₃CN in 0.1% TFA and, after evaporation, the samples were stored at 20°C. The specific activity of the compounds obtained was 120,000 to 180,000 cpm/pmol.

Binding to Rat Kidney Membrane. Kidney membrane was prepared from 10 rats by the centrifugation method (Stassen et al., 1982), mixed with PBS, pH 7.4, containing ¹²⁵I-labeled ligand and 0.1% BSA, and incubated for 1 h on ice at a concentration of 1 mg of protein/ml (Suzuki et al., 1999b). After centrifugation (15,000g, 10 min at 4°C), the supernatant was removed and the radioactivity in the supernatant and precipitate was measured by gamma counter. Data were fitted to the following equation using the MULTI program (Yamaoka et al., 1981).

(1)

where C_b is the ligand bound to the membrane (pmol/mg of protein), C_f is the unbound ligand concentration (pmol/ml), B_max is the maximum binding capacity (pmol/mg of protein), K_d is the dissociation constant (nM), and  is the proportional constant of nonspecific binding.

(2)

Tissue Distribution of Alkylglucoside Derivatives. The femoral artery was cannulated under light anesthesia with diethyl ether. ¹²⁵I-Labeled ligand was dissolved at a concentration of 1 nmol/ml in PBS alone or PBS containing OTG (final concentration 2.5, 5, 10, or µmol/ml) and administered via the tail vein at a dose of 1 ml/kg. Blood was collected at 0.5, 1, 1.5, 2, 3, 4, and 5 min and then quickly centrifuged to obtain plasma. Rats were sacrificed at 5 min and all or a portion of the tissues (liver, kidney, small intestine, skeletal muscle, lung, and spleen) was removed and rinsed with saline. The radioactivity was determined by gamma counter. The tissue uptake of [¹⁴C]inulin (2 µmol/kg) was also determined in the other groups of rats where plasma and kidney (0.2 g) were solubilized with 1 ml of Soluene 350 and added to 10 ml of HIONIC-Flour before measuring the radioactivity in a liquid scintillation counter. The tissue uptake clearance (CL_uptake) was calculated with the following equation:

(3)

where X_{(5 min)} (pmol/g of tissue) and AUC_{(0-5 min)} (pmol · min/ml) are the amount of ligand in the tissue at 5 min and the area under the plasma concentration-time curve from 0 to 5 min, respectively.

Mol. Wt. Distribution of Radioactivity Associated with Kidney after Glc-S-C8-APL Administration. ¹²⁵I-Glc-S-C8-APL17000 or ¹²⁵I-Glc-S-C8-APL41000 (1 nmol/kg) dissolved in PBS or PBS containing OTG (10 µmol/kg) was administered via the tail vein. The kidney was removed at 5 or 90 min and homogenized after adding a 3-fold volume of PBS. This 25% homogenate was mixed with a 4-fold volume of CH₃CN and, after centrifugation, the supernatant was evaporated and the residue dissolved in mobile phase (45% CH₃CN/10 mM phosphate buffer at pH 6.8). After filtration through a 0.22-µm filter, the sample was subjected to HPLC, and the radioactivity in each fraction collected at 30-s intervals was determined. The recovery of ¹²⁵I-Glc-S-C8-APL17000 and ¹²⁵I-Glc-S-C8-APL41000 was 87.3 and 92.3%, respectively. The HPLC column was G3000SW (TOSO, Tokyo, Japan) and the flow rate was 0.75 ml/min. The mol. wt. was calibrated using marker compounds, polyethylene glycol (Wako) and polyethylene oxide (TOSO) with a mean mol. wt. of 1.0 × 10³, 7.0 × 10³, 1.8 × 10⁴, 1.0 × 10⁵, and 3.0 × 10⁵.

Plasma Protein Binding. Blank plasma (500 µl) was incubated at 37°C for 30 min with each ligand and then applied to Centrifree or Microcon (Millipore Corporation) with a molecular mass limit of 50 and 100 kDa. All binding was normalized with respect to the filter blank.

Autoradiography. ¹²⁵I-Glc-S-C8-APL4500, ¹²⁵I-Glc-S-C8-APL41000, or ¹²⁵I-Glc-S-C8-AVP (1 nmol/kg) dissolved in PBS or PBS containing OTG (10 µmol/ml) was administered via the tail vein. At 5 min after administration, rats were sacrificed and kidney was quickly excised and immersed in liquid nitrogen. Sections prepared by microtome were dried under vacuum. Imaging plates (Fuji film) were placed in close contact with the sections and the density of radioactivity in the autoradiographic images was measured using an image analyzer (BAS2000, Fuji film). To quantitate the localization of radioactivity at a microscopic level, sections were mounted on glass slides and dipped into nuclear track emulsion and exposed. After developing and fixing the preparations, they were stained with hematoxylin and eosin.

Binding of Glc-S-C8-AVP to Basolateral (BLMV) and Brush-Border Membrane Vesicles (BBMV). BBMV and BLMV were prepared from rat kidney cortex by reported methods (Walter, 1975; Grassl and Aronson, 1986). Protein concentrations were determined by the method of Lowry et al. (1951). ¹²⁵I-Glc-S-C8-AVP (2 nM) and 12.5 µg of BBMV or BLMV were incubated for 1 h at 4°C in 0.1% BSA/PBS, pH 7.4. The incubation mixture was then applied to a membrane filter (pore size 0.45 µm; Millipore Corporation), rapidly filtered, and washed twice with 4 ml of ice-cold 0.1% BSA/PBS. The radioactivity in the filter was measured by gamma counter.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Specific Binding of Glc-S-C8-APLs to Rat Kidney Membrane. To characterize the mol. wt. dependence of the targeting efficiency of the alkylglycoside system for the kidney, we first synthesized Glc-S-C8-APLs from poly-L-lysines with three different mol. wt. distributions (Fig. 1). Gel filtration HPLC analysis revealed that the mean mol. wt. of Glc-S-C8-APL4500, Glc-S-C8-APL17000, and Glc-S-C8-APL41000 assessed using polyethylene oxide as a mol. wt. marker was 6,920, 11,200, and 28,200, respectively.

View larger version (31K): [in this window] [in a new window]   Fig. 2.   Scatchard plot representing the specific binding of Glc-S-C8-AVP and Glc-S-C8-APLs to rat kidney membrane. Various concentrations of 125I-Glc-S-C8-AVP (), Glc-S-C8-APL4500 (), 17000 (), and 41000 () were incubated with kidney membrane (1 mg/ml) at 4°C for 1 h. The straight line represents the fitted line. Each value represents the mean ± S.E. of triplicate experiments in two membrane preparations. The inset represents the enlarged scale.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Mutual inhibition between Glc-S-C8-AVP and Glc-S-C8-APLs in terms of specific binding to rat kidney membrane. A, inhibitory effect of Glc-S-C8-APL4500 (), 17000 (), and 41000 () on the binding of Glc-S-C8-AVP. Each value represents the mean ± S.E. of triplicate experiments in two membrane preparations. B, inhibitory effect of Glc-S-C8-AVP on the binding of Glc-S-C8-APL4500 (), 17000 (), and 41000 (). Each value represents the mean ± S.E. of triplicate experiments in two membrane preparations.

Tissue Distribution of Glc-S-C8-APLs. ¹²⁵I-Glc-S-C8-APL4500, ¹²⁵I-Glc-S-C8-APL17000, and ¹²⁵I-Glc-S-C8-APL41000 were administered intravenously to rats, and the CL_uptake in each tissue was examined (Fig. 4). The CL_uptake of each ligand was highest in the kidney compared with other tissues (Fig. 4). As an control experiment, the CL_uptake of [¹⁴C]inulin, an extracellular and glomerular filtration rate (GFR) marker, was also examined. The CL_uptake of [¹⁴C]inulin was 0.0148 ± 0.0025, 0.0658 ± 0.0011, 0.713 ± 0.122, 0.0334 ± 0.0065, 0.0201 ± 0.0060, and 0.0277 ± 0.0061 ml/min/g of tissue in muscle, small intestine, kidney, lung, spleen, and liver (mean ± S.E., n = 3). Thus, the CL_uptake of ¹²⁵I-Glc-S-C8-APLs in tissues other than the kidney was almost comparable with that of [¹⁴C]inulin.

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Tissue CLuptake of 125I-Glc-S-C8 derivatives. Alkylglucoside derivatives (1 nmol/kg) were administered intravenously, and the time-profile of the plasma concentrations was followed. At 5 min after administration, tissues were removed and counted. Each value represents the mean ± S.E. of four to eight rats.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Inhibitory effect of OTG on renal uptake of 125I-Glc-S-C8 derivatives. 125I-Glc-S-C8-AVP (A), 125I-Glc-S-C8-APL4500 (B), 17000 (C), or 41000 (D) at 1 nmol/kg were simultaneously administered with OTG at the different doses shown. The dotted line represents the contribution of glomerular filtration, which was assessed as the fuGFR. Each value represents the mean ± S.E. of four to eight rats.

Mol. Wt. Distribution of Radioactivity in Kidney after Injection of Glc-S-C8-APLs. When ¹²⁵I-Glc-S-C8-APL41000 was administered intravenously, the radioactivity in the kidney at 5 min after injection exhibited a smaller mol. wt. distribution than authentic ¹²⁵I-Glc-S-C8-APL41000 and ¹²⁵I-Glc-S-C8-APL17000 (Fig. 6A). The mol. wt. distribution of radioactivity at 90 min after injection was lower than that at 5 min (Fig. 6B). When OTG was coadministered, the radioactivity corresponding to the relatively smaller mol. wt. was reduced, compared with the radioactivity after injection of Glc-S-C8-APL41000 alone (Fig. 6B). A similar analysis was performed for ¹²⁵I-Glc-S-C8-APL17000, and an almost exactly comparable mol. wt. distribution was found between the radioactivity associated in the kidney and that of authentic ¹²⁵I-Glc-S-C8-APL17000 (Fig. 6C).

View larger version (30K): [in this window] [in a new window]   Fig. 6.   Mol. wt. distribution of radioactivity associated in rat kidney after intravenous administration of 125I-Glc-S-C8-APL41000. 125I-Glc-S-C8-APL41000 was injected intravenously into rats and kidneys were excised at 5 min () or 90 () min. Mol. wt. distribution of the radioactivity extracted from the kidney homogenate was determined by gel filtration HPLC. Data are expressed as the ratio (%) of radioactivity in each fraction to that applied to HPLC. B, effect of coadministration of OTG () was also examined. C, similar analysis was performed on the kidney 5 min after administration of Glc-S-C8-APL17000 (). For comparison, authentic 125I-Glc-S-C8-APL41000 () and 17000 () are also shown in each panel. Data represent the typical results from one rat, but the similar results were obtained from at least three rats.

Renal Distribution and Specific Binding to Kidney Membranes of Glc-S-C8-APLs Fractionated by Gel Filtration. To more directly investigate the mol. wt. dependence of the renal targeting of Glc-S-C8-APLs, authentic ¹²⁵I-Glc-S-C8-APL17000 or ¹²⁵I-Glc-S-C8-APL41000 was further fractionated by gel filtration HPLC, and the CL_uptake of each fraction was examined (Table 2). In each case, the CL_uptake of fractionated Glc-S-C8-APLs was higher for the lower mol. wt. fraction than for the higher mol. wt. fraction, whereas the f_u of each did not differ by much (Table 2). The contribution of binding and/or uptake from the blood side, assessed as the CL_uptake  f_uGFR, also depended on the mol. wt. of Glc-S-C8-APLs (Table 2).

Localization of Glc-S-C8-APLs in Kidney. Both semimicroautoradiography (Fig. 7; Table 4) and microautoradiography (Fig. 8) were performed for the analysis of intrarenal localization. Radioactivity was found in the kidney cortex after the injection of ¹²⁵I-Glc-S-C8-4500, ¹²⁵I-Glc-S-C8-41000, and ¹²⁵I-Glc-S-C8-AVP alone (Fig. 7, A, C, and E), whereas this radioactivity fell when OTG was coadministered (Fig. 7, B and D). The density of radioactivity was quantified and this is shown in Table 4. The accumulation of radioactivity was mainly found in the cortex after the injection of ¹²⁵I-Glc-S-C8-AVP, whereas, after injection of Glc-S-C8-4500 and Glc-S-C8-41000, it was also found in the inner and outer medulla to a lesser extent (Table 4). The reduction after coadministration of OTG was found mainly in the cortex, whereas the effect of OTG was not so obvious in the inner or outer medulla (Table 4).

View larger version (36K): [in this window] [in a new window]   Fig. 7.   Intrarenal distribution of radioactivity after administration of 125I-Glc-S-C8-APL4500 (A and B), 41000 (C and D) and Glc-S-C8-AVP (E). 125I-Labeled alkylglucoside derivatives (1 nmol/kg) were injected into rats without (A, C, and E) or with (B and D) OTG (10 µmol/kg). Five minutes after administration, kidneys were excised and freeze-dried sections prepared by microtome were placed in contact with an imaging plate for 1 h. The density of radioactivity in the autoradiographic images was analyzed by a Bio imaging analyzer and are shown in Table 4.

View larger version (88K): [in this window] [in a new window]   Fig. 8.   Microautoradiography for the distribution of 125I-labeled alkylglucoside derivatives to rat kidney cortex. Kidney slices at 5 min after intravenous administration of 125I-Glc-S-C8-APL4500 (A and B), 125I-Glc-S-C8-APL41000 (C and D) or 125I-Glc-S-C8-AVP (E) without (A, C, and E) or with (B and D) OTG (10 µmol/kg). RC, renal corpuscle; PT, proximal tube; DT, distal tube. Original magnification, 400×.

View larger version (18K): [in this window] [in a new window]   Fig. 9.   Binding of 125I-Glc-S-C8-AVP to BLMV and BBMV prepared from rat kidney cortex. BLMV and BBMV (12.5 µg) were incubated with 125I-Glc-S-C8-AVP (2 nM) in the absence and presence of 1.25 µM Glc-S-C8-AVP. Binding was determined by the rapid filtration technique. Each value represents the mean ± S.E. of triplicate experiments.

Renal Distribution and Specific Binding to Kidney Membranes of Glc-S-C8-Derivatives with Different Charge Moieties. To characterize the effect of the charge on the ligand derivatized with alkylglycoside vector, we examined the renal distribution and specific binding to kidney membranes of Glc-S-C8-Tyr-Base, Glc-S-C8-Tyr-Neutral, and Glc-S-C8-Tyr-Acid (Table 5; Fig. 10). We also examined Glc-S-C8-Ala-VP, which has a neutral amino acid, alanine, instead of the cationic one, arginine, in Glc-S-C8-AVP (Table 5; Fig. 10). The CL_uptake of Glc-S-C8-Tyr-Base and Glc-S-C8-Tyr-Neutral was higher in the kidney than in other organs (Fig. 10A). The order of the CL_uptake in the kidney was Glc-S-C8-Tyr-Acid < Glc-S-C8-Tyr-Base < Glc-S-C8-Tyr-Neutral (Fig. 10A). The CL_uptake in the kidney of Glc-S-C8-Ala-VP was higher than that of Glc-S-C8-AVP (Fig. 10B). The f_u of Glc-S-C8-Tyr-Base, Glc-S-C8-Tyr-Neutral, Glc-S-C8-Tyr-Acid, and Glc-S-C8-Ala-VP was 0.574 ± 0.08, 0.475 ± 0.07, 0.540 ± 0.05, and 0.272 ± 0.05, respectively. The CL_uptake in the kidney of Glc-S-C8-Tyr-Base, Glc-S-C8-Tyr-Neutral, and Glc-S-C8-Ala-VP was higher than the respective f_uGFR value (Fig. 10).

View larger version (18K): [in this window] [in a new window]   Fig. 10.   Tissue CLuptake of 125I-Glc-S-C8 derivatives. Alkylglucoside derivatives (1 nmol/kg) were administered intravenously and the time profile of the plasma concentration was followed. At 5 min after administration, tissues were removed and counted. Each value represents the mean ± S.E. of three to four rats. The dotted line represents the contribution of glomerular filtration, which was assessed as the fuGFR.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Kidney is one of the major organs involved in maintaining homeostasis in the body and, therefore, could be a target for various types of drugs. We previously found that the kidney-specific distribution of AVP could be modified by some sugar moieties via octamethylene (Suzuki et al., 1999c). Further studies revealed that the alkylglucoside structure is necessary for recognition by the kidney (Suzuki et al., 1999b). To develop this structural recognition as a kidney-targeting vector for therapeutic agents, it is important to clarify the target efficiency and limitations of the compounds derivatized with alkylglucosides. As far as these limitations were concerned, the present study focused on two aspects, molecular size and charge.

To examine the molecular size dependence in the distribution to the kidney, we synthesized Glc-S-C8-APLs from poly-L-lysines with three ranges of mol. wt. To minimize the effect of the positive charge on poly-L-lysines, these compounds were acylated before alkylglycoside conjugation. Because this vector system is recognized by the binding sites on renal membranes (Suzuki et al., 1999b,c), we first examined their specific binding to the kidney membrane fractions (Figs. 2 and 3; Table 1). The results obtained suggest that Glc-S-C8-APL4500, Glc-S-C8-APL17000, and Glc-S-C8-APL41000 share the same binding site with Glc-S-C8-AVP because 1) mutual inhibition was observed (Fig. 3), 2) the obtained K_i values for these four compounds were comparable with their respective K_d values (Table 1), and 3) the B_max value was within 4 times the difference between these compounds (Table 1). In addition, the CL_uptake of Glc-S-C8-APL4500, Glc-S-C8-APL17000, and Glc-S-C8-APL41000 was much larger in the kidney, whereas in other organs it was similar to that of inulin, suggesting the kidney-specific distribution of these compounds. Considering that 1) the CL_uptake in kidney was higher than the f_uGFR of the three compounds and 2) OTG inhibits such renal distribution of these compounds (Fig. 5), a mechanism other than GFR appears to be involved in such renal distribution of Glc-S-C8-APLs.

However, it should be noted that the affinity (1/K_d) of Glc-S-C8-APLs was 20 to 50 times lower than that of Glc-S-C8-AVP (Table 1), and that the difference between the CL_uptake and the f_uGFR was not as marked for Glc-S-C8-APL41000 (Fig. 4). These results suggest that the greater molecular size of Glc-S-C8-APLs may, at least to some extent, hinder the association of these compounds in the kidney. In addition, because these Glc-S-C8-APLs consists of a variety molecular sizes due to the crudeness of the poly-L-lysines used to synthesize Glc-S-C8-APLs, it may be that the apparent specific binding and/or renal distribution represents that of only the smaller size fraction of the Glc-S-C8-APLs. Therefore, we next attempted to identify the mol. wt. size distribution of the radioactivity actually associated in the kidney after the injection of ¹²⁵I-Glc-S-C8-APLs (Fig. 6). The size distribution at 5 min after the injection of Glc-S-C8-APL41000 exhibited a lower mol. wt. compared with authentic Glc-S-C8-APL41000 (Fig. 6A), indicating that the Glc-S-C8-APLs actually distributed to the kidney are smaller molecules compared with those injected. In addition, a similar mol. wt. distribution was found in case of Glc-S-C8-APL17000 (Fig. 6C). Therefore, the smaller molecules could be preferentially targeted to the kidney, although we cannot rule out the possibility that the larger molecules may be rapidly degraded into smaller ones during the 5-min period after injection. The mol. wt. distribution of radioactivity at 90 min was different from that at 5 min (Fig. 6B), indicating the possible fragmentation of Glc-S-C8-APL41000. To estimate the involvement of alkylglycoside binding sites on the kidney membranes in such renal distribution, the effect of OTG was also examined (Fig. 6B). When OTG was coadministered, there was relatively less radioactivity compared with the radioactivity after the injection of ¹²⁵I-Glc-S-C8-APL41000 alone (Fig. 6B). This was compatible with the hypothesis that the smaller sized molecules are preferentially recognized by the binding sites.

To more directly examine the mol. wt. dependence in the kidney targeting efficiency, ¹²⁵I-Glc-S-C8-APL41000 and ¹²⁵I-Glc-S-C8-APL17000 were further fractionated into ¹²⁵I-Glc-S-C8-APLs with a more limited size distribution (Tables 2 and 3). Both the specific portion of the CL_uptake (Table 2) and the specific binding to the kidney membrane (Table 3) were relatively smaller for the fractionated ¹²⁵I-Glc-S-C8-APLs of a larger size, suggesting that the kidney targeting is highly dependent on the molecular size of the derivatives containing the alkylglycoside moieties. However, even after fractionation, the ¹²⁵I-Glc-S-C8-APLs will still exhibit a degree of variation in their mol. wt. Therefore, the present analysis cannot conclusively determine the exact size limitation of kidney targeting. Considering the CL_uptake and binding of fractionated ¹²⁵I-Glc-S-C8-APLs (Tables 2 and 3), the renal distribution of Glc-S-C8-APLs with a mol. wt. higher than 10,000 seemed to be greatly impaired, compared with those with a lower mol. wt.

To demonstrate that Glc-S-C8-APLs can be targeted to the alkylglycoside vector binding sites, we also attempted to examine their intrarenal distribution because the location of the alkylglucoside targeting site, assessed as the radioactivity after injection of Glc-S-C8-AVP, is the cortex, especially the proximal tubules (Suzuki et al., 1999b). Similar results were obtained for ¹²⁵I-Glc-S-C8-AVP (Fig. 8E; Table 4). In addition, the location of the radioactivity, which can be reduced by OTG, was also found in the same region after the injection of ¹²⁵I-Glc-S-C8-APL4500 and ¹²⁵I-Glc-S-C8-APL41000 (Fig. 7; Table 4), and the density of radioactivity in the kidney cortex is higher after the injection of ¹²⁵I-Glc-S-C8-APL4500 than ¹²⁵I-Glc-S-C8-APL41000 (Table 4). This supports the hypothesis that Glc-S-C8-APLs share the same binding sites with Glc-S-C8-AVP with higher affinity for Glc-S-C8-APLs with a lower mol. wt. Considering that the CL_uptake of Glc-S-C8-AVP and Glc-S-C8-APLs was higher than the respective f_uGFR (Fig. 4), and that the binding of ¹²⁵I-Glc-S-C8-AVP is much higher in BLMV than BBMV (Fig. 9), the binding sites for alkylglycoside should be localized on the basolateral side of kidney proximal tubules. However, it was noted that there was clear selective distribution to the cortex for ¹²⁵I-Glc-S-C8-AVP, whereas ¹²⁵I-Glc-S-C8-APL4500 and ¹²⁵I-Glc-S-C8-APL41000 were also distributed to the outer and inner medulla (Table 4). Considering that the contribution of GFR to the renal distribution seemed to be higher for Glc-S-C8-APLs (Fig. 4), such distribution to the medulla may reflect the radioactivity that undergoes glomerular filtration. The microautoradiography also showed that Glc-S-C8-APL4500 and Glc-S-C8-41000 in the kidney cortex were also localized in the distal tubules, and such distribution was more clearly defined in Glc-S-C8-APL41000 (Fig. 8). Because the affinity of Glc-S-C8-APL4500 and Glc-S-C8-41000 for the kidney membrane is much lower than that of Glc-S-C8-AVP (Table 1), nonspecific distribution, including glomerular filtration may be marked in the case of Glc-S-C8-APL4500 and Glc-S-C8-41000 compared with Glc-S-C8-AVP.

Of the five compounds examined in Table 5, only the acidic compound Glc-S-C8-Tyr-Acid exhibited a much lower affinity for kidney membrane than the others. As in the case of Glc-S-C8-APLs, mutual inhibition among Glc-S-C8-AVP and the other four compounds was found, and the K_i values obtained were comparable with the K_d of the binding of inhibitors, suggesting that these five compounds also share the same binding sites on the kidney membrane. Therefore, the acidic moiety may hinder the binding of alkylglycoside derivatives. This hypothesis is compatible with the present finding that the CL_uptake of Glc-S-C8-Tyr-Acid in the kidney was lower than that of Glc-S-C8-Tyr-Base and Glc-S-C8-Tyr-Neutral (Fig. 10A). Additionally, the CL_uptake of Glc-S-C8-Tyr-Base was smaller than that of Glc-S-C8-Tyr-Neutral (Fig. 10A), and that of Glc-S-C8-AVP was smaller than that of Glc-S-C8-Ala-VP (Fig. 10B). Such a difference (Glc-S-C8-Tyr-Base < Glc-S-C8-Tyr-Neutral, Glc-S-C8-AVP < Glc-S-C8-Ala-VP) was also found in the specific binding to the kidney membrane (Table 5). These results suggest that the basic moiety also hinders, at least to some extent, the renal targeting of the alkylglycoside, although further studies are needed before a final conclusion can be drawn. As far as tissues other than the kidney are concerned, the CL_uptake in the liver was relatively higher for Glc-S-C8-Tyr-Base, Glc-S-C8-Tyr-Acid, and Glc-S-C8-Ala-VP. We previously found a greater CL_uptake for Glc-S-C8-oxytocin, a neutral peptide, both in the liver and small intestine (Suzuki et al., 1999b). Although the mechanism for such extrarenal distribution is still unclear, some nonspecific mechanism affected by moieties other than the alkylglycoside are likely to be involved.

The binding site for the alkylglucoside vector is located on the basolateral membrane (Fig. 9), whereas the Na⁺/glucose cotransporter, which is inhibited by OTG, is expressed on the brush-border membrane (Haase et al., 1990; Thorens, 1996). Therefore, involvement of this cotransporter in the renal targeting is unlikely. The kidney lectin that recognizes acidic sugars is expressed in many other organs and may not explain the kidney-specific distribution of alkylglucosides (Kojima et al., 1996). Watanabe et al. (2000) identified a 62-kDa protein expressed on the kidney membranes. Additional studies are needed to clarify the molecular mechanism involved in the renal distribution of the alkylglucoside vector.

In conclusion, the efficiency of kidney targeting by using the alkylglucoside vector depends both on the size and charge of the molecules derivatized with this vector. The delivery system with this vector seems to be more suitable for therapeutic agents with a lower mol. wt. and neutral charge.