Inhibition of gentamicin binding to rat renal brush-border membrane by megalin ligands and basic peptides

Journal of Controlled Release 112 (2006) 43 – 50 www.elsevier.com/locate/jconrel

Inhibition of gentamicin binding to rat renal brush-border membrane by megalin ligands and basic peptides Junya Nagai, Masaki Saito, Yoshinori Adachi, Ryoko Yumoto, Mikihisa Takano ⁎ Department of Pharmaceutics and Therapeutics, Graduate School of Biomedical Sciences, Hiroshima University, 1-2-3 Kasumi, Minami-ku, Hiroshima 734-8553, Japan Received 8 April 2005; accepted 13 January 2006 Available online 20 February 2006

Abstract Our previous studies showed that coadministration of cytochrome c and a 20-residue basic peptide, N-WASP181–200 (NISHTKEKKKGKAKKKRLTK, pI = 10.87) inhibits renal accumulation of gentamicin. In this study, we examined effects of ligands of megalin, an endocytic receptor involved in renal uptake of gentamicin, and basic peptides including N-WASP180–200 and its mutant peptides on gentamicin binding to isolated rat renal brush-border membrane (BBM). Gentamicin binding to BBM was inhibited by megalin ligands, basic peptide fragments of cytochrome c, and N-WASP181–200 in a concentration-dependent manner. Klotz plot analysis showed that N-WASP181–200 inhibited the binding of gentamicin in a competitive manner. By substituting glycines for lysines in N-WASP181–200 at positions 9 and 15, the inhibitory effect on gentamicin binding to BBM was reduced, which may be related to a decrease in the α-helix content in the peptide. Gentamicin binding to BBM treated with trypsin, in which megalin completely disappeared, was significantly but not completely decreased compared with the native BBM. In addition, treatment of BBM with trypsin led to a decrease in the inhibitory effect of N-WASP181–200 on gentamicin binding. These observations support that megalin ligands and basic peptides including N-WASP181–200 decrease renal accumulation of gentamicin by inhibiting its binding to BBM of proximal tubule cells, partly interacting with megalin. In addition, the α-helix conformation may play an important role in the inhibitory effect of N-WASP181-200 on the binding of gentamicin to BBM. © 2006 Elsevier B.V. All rights reserved. Keywords: Aminoglycosides; Multiligand endocytic receptor; Basic peptides; Renal brush-border membrane; Nephrotoxicity

1. Introduction Aminoglycoside antibiotics such as gentamicin and amikacin are widely used in the treatment of Gram-negative infections, including therapy for neonatal sepsis, meningitis, intraabdominal sepsis and urinary tract infections. However, its therapeutic usefulness is limited by its potential nephrotoxicity and ototoxicity. It is likely that aminoglycoside-induced nephrotoxicity is directly related to the concentrated accumulation of aminoglycosides in the renal proximal tubular cells. About 10% of the intravenously administered dose is selectively accumulated in the renal cortex, whereas little distribution of aminoglycosides to other tissues is observed. The concentration of aminoglycoside accumulated in the renal cortex amounts to ⁎ Corresponding author. Tel.: +81 82 257 5315; fax: +81 82 257 5319. E-mail address: [email protected] (M. Takano). 0168-3659/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2006.01.003

several times higher than in the plasma. Aminoglycoside taken up by the renal proximal tubular cells stays there with a longhalf life, leading to renal damage such as structural changes and functional impairments [1,2]. A considerable number of studies have been performed to clarify the molecular mechanisms responsible for aminoglycoside uptake in the renal proximal tubular cells [3,4]. Electron microscopic analysis revealed that aminoglycosides are localized in clathrin-coated pits, endocytic compartments and lysosomes in the proximal tubular cells, showing the involvement of an endocytic pathway in the renal accumulation of aminoglycosides. By examining renal accumulation of gentamicin in filtering and non-filtering perfused rat kidney, it was shown that renal tubular uptake of gentamicin occurs by reabsorption from the luminal surface of the proximal tubular cells [5]. Since the accumulation of the injected aminoglycosides in the renal cortex is saturable [6], a receptor-mediated

44

J. Nagai et al. / Journal of Controlled Release 112 (2006) 43–50

endocytosis appears to be responsible for the concentrated accumulation of aminoglycosides. Concerning the aminoglycoside binding receptor, early research suggested that acidic phospholipids of the plasma membrane of the proximal tubular cells are the initial binding site [7,8]. By using an organic solvent-water partition system, several acidic phospholipids were shown to bind gentamicin in a saturable manner with Kd values of 3 to 12 μM, which are similar to that reported in renal brush-border membrane (BBM) binding assay [9]. On the other hand, other factors would be necessary for the specific accumulation of aminoglycosides in the renal proximal tubular cells, because acidic phospholipids are commonly distributed in various tissues. In 1995, Moestrup et al. [10] found that aminoglycosides including gentamicin and amikacin interact with megalin (alias LRP2; low-density lipoprotein receptor-related protein-2), a multiligand, endocytic receptor abundantly expressed in the renal proximal tubule. Subsequently, we showed that tissue distribution of amikacin was well correlated with the expression level of megalin in tissues [11]. In addition, in maleate-treated rats in which megalin was shed from the renal brush-border membrane, accumulation of amikacin was decreased and then recovered with time corresponding to megalin level in the renal cortex [11]. Furthermore, Schmitz et al. reported that there was little accumulation of gentamicin in the kidney in mice lacking megalin [12]. Thus, megalin as well as acidic phospholipids play an important role in uptake of aminoglycosides in the kidney. The above-mentioned findings suggest that aminoglycoside binding receptors such as megalin and/or acidic phospholipids are a potential target for preventing aminoglycoside-induced nephrotoxicity. In a previous report, we demonstrated that coadministration of cytochrome c, a ligand of megalin, decreased not only renal accumulation of gentamicin but also gentamicin-induced nephrotoxicity [13]. In addition, several basic peptide fragments with pI values from 10.30 to 11.72, which reportedly bind to phosphoinositides [14–16], inhibited renal accumulation of gentamicin injected intravenously. Among those peptide fragments examined, N-WASP181–200 (pI = 10.87) from neural Wiskott–Aldrich syndrome protein (NWASP) was the most effective antagonist [13]. However, the molecular mechanisms underlying the inhibitory effects of cytochrome c and N-WASP181–200 on gentamicin accumulation in the kidney have not been fully understood. In addition, more effective compounds should be explored because relatively high doses (N10 mg/kg) of cytochrome c and NWASP181–200 were needed to inhibit renal accumulation of gentamicin. For this purpose, it is important to obtain more detailed information concerning the molecular mechanisms underlying their inhibitory effects on renal gentamicin accumulation. In the present study, we have investigated the effects of megalin ligands such as cytochrome c as well as N-WASP181– 200 on binding of gentamicin to BBM, which is an initial step of renal accumulation of aminoglycosides. In addition, we compared the inhibitory effects of N-WASP181–200 and its mutant peptides on gentamicin binding to BBM, to analyze the role of

specific amino acid residues and/or their conformations in the inhibitory effect of N-WASP181–200. 2. Materials and methods 2.1. Materials [3H]Gentamicin sulfate (7.4 GBq/g) was obtained from American Radiolabeled Chemicals, Inc. (St. Louis, MO, USA). Lysozyme chloride from egg white and gentamicin sulfate were obtained from Nacalai Tesque (Kyoto, Japan). Cytochrome c from bovine heart was purchased from Sigma Chemical Co. (St. Louis, MO, USA). All other chemicals used for the experiments were of the highest purity available. 2.2. Preparation of peptide fragments Synthetic peptide fragments were produced with the peptide synthesizer (PSSM-8, Shimadzu, Kyoto, Japan). The peptide fragments were synthesized chemically and their amino acid sequences were as follows: Cyto22–41 (KGGKHKTGP NLHGLFGRKTG), Cyto70–89 (NPKKYIPGTKMIFA GIKKKG), Cyto79–88 (KMIFAGI KKK), N-WASP181–200 (NISHTKEKKKGKAKK KKRLTK), N-W(K9G, K15G) (NISHTKEKGKGKAKGKKRLTK), N-W(K9E, K15E) (NISHTKEKEKGKAKEKKRLTK), N-W(E7K) (NISHT KKKKKGKAKKK KRLTK). The peptide sequence of the fragments prepared by the synthesizer was confirmed by matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry (PerSeptive Biosystems, Tokyo, Japan). In Table 1, amino acid sequences and isoelectric points of synthesized peptides are listed. The pI values were

Table 1 Sequences and pI values of basic peptides and IC50 values of cytochrome c, lysozyme and the peptides Competitors Cytochrome c Lysozyme Cyto22–41 Cyto70–89 Cyto79–88 N-WASP181–200 N-W(K9G,K15G) N-W(K9E, K15E) N-W(E7K) N-WASP181–200 (trypsin-treated)

Sequences

KGGKHKTGPNL HGLFGRKTG NPKKYIPGTKMI FAGIKKKG KMIFAGIKKK NISHTKEKKKGK AKKKRLTK NISHTKEKGKGK AKGKRLTK NISHTKEKEKGK AKEKRLTK NISHTKKKKKGK AKKKRLTK

Calculated pI values

IC50 valuesa

11.33

3.20 ± 0.66 2.42 ± 0.92 0.312 ± 0.072

10.40

0.288 ± 0.101

10.48 10.87

0.240 ± 0.062 0.041 ± 0.0001

10.75

0.078 ± 0.009

10.12

0.043 ± 0.009

11.56

0.034 ± 0.003 0.435 ± 0.121

a The IC50 values (mM) ± standard errors of competitors were determined by fitting the data shown in Figs. 1, 2, 3 and 6 to the Hill equation. The details of the estimation are described in the text. r value for the linear regression in each IC50 determination was N0.95.

J. Nagai et al. / Journal of Controlled Release 112 (2006) 43–50

45

calculated by a WWW-accessible program ProtParam (http://us. expasy.org/tools/protparam.html).

the linear regression lines using the double reciprocal analysis based on Klotz equation,

2.3. Preparation of BBM

1=ðB−B0 Þ ¼ ðKd =Bmax Þð1=FÞ þ 1=Bmax

Experiments with animals were performed in accordance with the Guide for Animal Experimentation, Hiroshima University, and the Committee of Research Facilities for Laboratory Animal Sciences, Graduate School of Biomedical Sciences, Hiroshima University. BBM were isolated from the renal cortex of male Wistar albino rats (260–310 g). The isolation procedure of BBM was based on the Mg/EGTA precipitation method as described previously [17,18]. The isolated membranes were suspended in a buffer comprising 100 mM mannitol and 10 mM HEPES/Tris (pH 7.5), and stored in liquid nitrogen until use (usually less than 1 month). The purity of the BBM purification was confirmed by measuring marker enzymes for BBM [aminopeptidase (EC 3.1.3.1) and alkaline phosphatase (EC 3.4.11.2)] and basolateral membranes [Na+/ K+-ATPase (EC 3.6.1.3)]. These enzymes were assayed as described previously [18]. The specific activities of aminopeptidase and alkaline phosphatase in the BBM were about 17-fold higher than those in the homogenate of renal cortex (enrichment), whereas the enrichment of Na+/K+-ATPase was less than two-fold. 2.4. Binding studies using BBM [3H]Gentamicin binding to BBM was measured by a rapid filtration technique. In brief, membrane vesicles (20 μL) were suspended in a buffer comprising 300 mM mannitol and 10 mM HEPES, adjusted to pH 7.5 with KOH (buffer A). The reaction was initiated by the addition of buffer A (20 μL) containing 20 μM [3H]gentamicin without or with competitor into the membrane suspension at 4 °C. At 60 min, incubation was stopped by diluting the reaction mixture with 1 mL of ice-cold buffer A. The contents of the tube were immediately poured onto Millipore filters (HAWP, 0.45 μm, 2.5 cm diameter), and then the filters were washed once with 5 mL of the ice-cold buffer A. The radioactivity on the filter was determined by liquid scintillation counting. Non-specific binding was estimated by the addition of unlabeled gentamicin at 1000-fold concentration (10 mM) of [ 3 H]gentamicin. Then, specific binding of gentamicin was determined by subtracting the non-specific binding from the total binding. The half-maximal inhibitory concentration (IC50) values of competitors were determined by the following Hill equation, B−B0 ¼ ðB V−B0 Þ=½1−ð½I=IC50 Þn  where [I] is the inhibitor concentration, B is the amount of bound [3H]gentamicin, B′ is the bound [3H]gentamicin in the absence of an inhibitor, and B0 is the amount of [3H]gentamicin in the presence of 10 mM unlabeled gentamicin (non-specific binding). The values of the dissociation constant (Kd) and the maximum amounts of the binding (Bmax) were estimated from

where F is the concentration of free [3H]gentamicin in the incubation buffer. The x- and y-axis intercepts of the linear regression line equal − 1 / Kd and 1 / Bmax, respectively. The KaleidaGraph™ program (Version 3.08, Synergy Software, PA, USA) was used for the curve-fitting. From the curve fitting to the above-mentioned equations, Kd, Bmax, and the IC50 values of each competitor were obtained. 2.5. Treatment of BBM with trypsin Aliquots of BBM were incubated without or with trypsin (final 1 mg/mL) for 15 min at 37 °C. Then, the reaction was stopped by adding 5 mL of ice-cold buffer A. The suspensions were centrifuged at 24,000 ×g for 30 min. The pellet was suspended in buffer A and centrifuged at 24,000×g for 30 min. The final pellet was resuspended with buffer A and the treated BBM were used for binding assay and immunoblotting. 2.6. Western blot analysis Immunoblot analysis of megalin was performed as described previously [19]. Briefly, BBM were heated for 5 min at 95 °C in a loading buffer containing 5% 2-mercaptoethanol. The samples were subjected to SDS-polyacrylamide gel electrophoresis with 6% polyacrylamide gels, and the proteins were transferred for 75 min to PVDF membrane at 4 °C. The membrane was blocked in 5% non-fat dry milk in Tris-buffered saline (150 mM NaCl, 20 mM Tris/HCl, pH 7.5) including 0.05% Tween 20 (TBS-T) overnight at 4 °C. The membranes were washed three times for 10 min in TBS-T, and were incubated with the anti-rat megalin rabbit antiserum (1 : 4000 dilution). The membranes were washed three times in TBS-T, and were incubated with the horseradish peroxidase-conjugated anti-rabbit IgG antibody (1 : 2000 dilution), washed 3 times in TBS-T, and visualized with enhanced chemiluminescence (ECL; Amersham Pharmacia Biotech). 2.7. Circular dichroism spectroscopy The peptide was dissolved in methanol with a concentration of 25 μM. An aliquot of the sample solution was applied to a Jasco J-720 spectropolarimeter (Japan Spectroscopy Co. Ltd., Tokyo) to perform circular dichroism (CD) measurements. Experiments were carried out in the far-UV region (200 to 250 nm) at room temperature. 2.8. Analytical methods Protein concentration was measured by the method of Bradford [20] with bovine serum albumin as a standard. Phospholipid content of BBM was determined spectrophotometrically using a commercially available kit, Phospholipid Test

46

J. Nagai et al. / Journal of Controlled Release 112 (2006) 43–50

Gentamicin binding (% of control)

Cytochrome c

100

Lysozyme

80 60 40 20 0 0.01

0.1 1 Concentration (mM)

10

Fig. 1. Effects of cytochrome c and lysozyme on [3H]gentamicin binding to rat renal brush-border membrane (BBM). BBM (20 μL), suspended in 100 mM mannitol and 10 mM HEPES (pH 7.5), were incubated for 60 min at 4 °C with the substrate mixture (pH 7.5) (20 μL) comprised of 100 mM mannitol, 10 mM HEPES, 20 μM [3H]gentamicin with various concentrations of cytochrome c (closed circle) or lysozyme (closed square). These data were fitted to the Hill equation (see Materials and methods). Each symbol represents the mean ± S.E. of three determinations.

WAKO™ (Wako Pure Chemicals, Osaka, Japan). Statistical analysis was performed by Student's t-test. A difference of P b 0.05 was considered statistically significant. 3. Results

Cyto22-41

100

Gentamicin binding (% of control)

Gentamicin binding (% of control)

To confirm whether [3H]gentamicin uptake by BBM reflects transport into the vesicles and/or surface binding, the effect of

Cyto70-89 Cyto79-88

80 60 40 20 0

0.01

0.1 1 Concentration (mM)

extravesicular osmolarity on [ 3 H]gentamicin uptake was measured. The uptake of [3H]gentamicin was not affected by increasing the extravesicular osmolarity with various concentrations of sucrose (50 to 200 mM) (data not shown). This observation indicates that gentamicin binds to the membrane, but is not transported into the vesicles. The effects of cytochrome c and lysozyme, ligands of megalin, on [3H]gentamicin binding to BBM were examined. These compounds inhibited the binding of gentamicin in a concentration-dependent manner (Fig. 1). The IC50 values of cytochrome c and lysozyme, determined by the Hill equation, were 3.20 and 2.42 mM, respectively. Next, the effects of three peptide fragments derived from cytochrome c; Cyto22–41, Cyto70–89 and Cyto79–88 (Table 1), were examined. As reported previously [13], these peptide fragments were designed on the basis of the calculated pI since basic amino acids reportedly play an important role in the binding of megalin ligands to its receptor [10]. The binding of gentamicin to BBM was inhibited by these three peptide fragments from cytochrome c in a concentration-dependent manner (Fig. 2). The IC50 values of Cyto22–41, Cyto70–89 and Cyto79–88 were 0.312, 0.288 and 0.240 mM, respectively, showing that these peptide fragments are more potent than cytochrome c (3.19 mM) in inhibiting gentamicin binding to BBM. Our previous study [13] showed that N-WASP181–200, which has been reported to bind phosphoinositides [16], inhibits gentamicin accumulation in OK kidney epithelial cells expressing megalin as well as in mouse kidney. As shown in Fig. 3, NWASP181–200 decreased gentamicin binding to BBM in a concentration-dependent manner with an IC50 value of 0.041 mM. To examine the nature of the inhibitory effect of NWASP181–200 on gentamicin binding to BBM, the binding

10

Fig. 2. Effects of peptide fragments from cytochrome c (Cyto22–41, Cyto70–89 and Cyto79–88) on [3H]gentamicin binding to BBM. [3H]Gentamicin binding to BBM was measured as described in Fig. 1 in the presence of various concentrations of Cyto22–41 (closed circle), Cyto70–89 (closed square) and Cyto79–88 (closed triangle). These data were fitted to the Hill equation (see Materials and methods). Each symbol represents the mean ± S.E. of three determinations.

100

N-WAP181-200 N-W(K9G, K15G) N-W(K9E, K15E)

80

N-W(E7K)

60

40

20

0

0.01

0.1 Concentration (mM)

1

Fig. 3. Effects of N-WASP181–200 and its mutant peptides on [3H]gentamicin binding to BBM. [3H]Gentamicin binding to BBM was measured as described in Fig. 1 in the presence of various concentrations of N-WASP181–200 (open circle), N-W(K9G, K15G) (closed square), N-W(K9E, K15E) (closed triangle) and N-W(E7K) (closed diamond). These data were fitted to the Hill equation (see Materials and methods). Each symbol represents the mean ± S.E. of three determinations.

J. Nagai et al. / Journal of Controlled Release 112 (2006) 43–50

was examined at various concentrations of gentamicin without or with 50 μM WASP181–200. Double reciprocal plots (Klotz plots) of the data (Fig. 4) revealed that two regression lines intersected at almost the same point on y-axis, indicating that NWASP181–200 inhibits gentamicin binding to BBM in a competitive manner. The maximum amounts of gentamicin binding (Bmax) in the absence and presence of N-WASP181– 200 were estimated to be 12.5 and 10.2 nmol/mg protein, respectively. The calculated dissociation constants (Kd) in the absence and presence of N-WASP181–200 were 42.3 and 87.1 μM, respectively. Next, we investigated the role of lysine residues of NWASP181–200 in the inhibitory effect on gentamicin binding to BBM, since there are 9 lysines in the 20-amino acid sequence of N-WASP181–200. First, the effect of N-W(K9G, K15G), in which two lysines at positions 9 and 15 were substituted for glycines, on gentamicin binding to BBM was compared to that of N-WASP181–200 (Fig. 3). Like N-WASP181–200, N-W (K9G, K15G) inhibited gentamicin binding to BBM in a concentration-dependent manner. However, as shown by its IC50 value of 0.078 mM, the concentration–response curve of N-W(K9G, K15G) was shifted to the right, compared to that of N-WASP181–200. On the other hand, the IC50 value (0.043 mM) of N-W(K9E, K15E), in which two lysines at positions 9 and 15 were replaced by glutamic acids, was almost the same as that of N-WASP181–200. Furthermore, we investigated the effect of N-W(E7K), in which glutamic acid at position 7 was substituted by lysine, on the binding of gentamicin to BBM. The IC50 value of N-W(E7K) (0.034 mM) was slightly lower than that of N-WASP181–200.

(μg/lane)

5

10

20

5

10

20

Megalin

Fig. 5. Immunoblotting of BBM without or with trypsin treatment using antimegalin antibody. BBM was treated with trypsin as described in Materials and methods. Five, 10 and 20 μg of BBM without or with trypsin treatment were separated in 6% SDS-polyacrylamide gel electrophoresis, transferred to a polyvinylidene difluoride membrane. Megalin was detected by Western blotting with anti-megalin antibody.

So far, acidic phospholipids and megalin have been suggested to be involved in binding of aminoglycosides in the apical membrane of renal proximal tubular cells. However, the role of megalin in aminoglycoside binding to BBM remains to be clarified. As shown in Fig. 5, megalin disappeared from BBM by the treatment with trypsin. In such trypsin-treated BBM, the binding of gentamicin, expressed as gentamicin binding per mg protein of BBM, was slightly but significantly decreased (9.3%) compared to the native BBM (Control, 1.51 ± 0.01; trypsin-treated, 1.37 ± 0.03 pmol·mg protein− 1, n = 3, means ± S.E., P b 0.05). When the amount of gentamicin binding was corrected by phospholipid content of BBM, a greater decrease (21.8%) in gentamicin binding by the trypsin treatment was observed (Control, 2.66 ± 0.02; trypsin-treated 2.08 ± 0.03 pmol·mg phospholipid − 1 , n = 3, means ± S.E., P b 0.05). Furthermore, the inhibitory effect of N-WASP181– 200 on gentamicin binding to trypsin-treated BBM was compared to that to the native BBM (Fig. 6). Trypsin treatment dramatically decreased the inhibitory effect of N-WASP181– 200 on gentamicin binding to BBM, as shown by an IC50 value

Control

1.0

Control

N-WASP181-200

0.8 0.6 0.4 0.2

-50

Trypsin-treated BBM

Native BBM

Gentamicin binding (% of control)

1/B (nmol-1 .mg protein)

1.2

47

0

50

100

150

200

-1

1/F (mM ) Fig. 4. Double reciprocal plots (Klotz plots) of gentamicin binding to BBM in the absence or presence of N-WASP181–200. The BBM binding of various concentrations of [3H]gentamicin in the absence (open circle) or presence (closed circle) of 50 μM N-WASP181–200 was measured as described in Fig. 1. Each symbol represents the mean ± S.E. of three determinations. Abbreviations are: B, the amount of bound [3H]gentamicin to BBM; F, the concentration of free [3H]gentamicin in the incubation buffer.

Trypsin-treated

100 80 60 40 20 0

0.01

0.1 Concentration (mM)

1

Fig. 6. Effects of trypsin treatment on inhibition of gentamicin binding by NWASP181–200. [3H]Gentamicin binding to trypsin-treated BBM (closed circle) and the native BBM (open circle) in the presence of various concentrations of NWASP181–200 was measured as described in Fig. 1. These data were fitted to the Hill equation. Each symbol represents the mean ± S.E. of three determinations.

48

J. Nagai et al. / Journal of Controlled Release 112 (2006) 43–50

6

N-WASP181-200 N-W(K9G, K15G) N-W(K9E, K15E) N-W(E7K)

[θ] x 10-3 (deg . cm2 . dmol-1)

4 2 0 -2 -4 -6 -8 200

210 220 230 240 Wavelength (nm)

250

Fig. 7. Circular dichroism (CD) spectra of N-WASP181–200 and its mutant peptides. CD spectra of N-WASP181–200 (open circle), N-W(K9G, K15G) (open square), N-W(K9E, K15E) (open triangle) and N-W(E7K) (open diamond) (25 μM each) in methanol were measured in the far-UV range (200 to 250 nm).

of 0.435 mM, which was approximately 10-fold greater than that to the native BBM. We summarized the IC50 values of megalin ligands and peptides used in this study for the binding of gentamicin to BBM (Table 1). To examine whether the molecular conformation of NWASP181–200 plays a role in the inhibition of gentamicin binding to BBM, we measured circular dichroism (CD) spectra of N-WASP181–200, N-W(K9G, K15G), N-W(K9E, K15E), and N-W(E7K) (Fig. 7). N-WASP181–200 showed a typical spectrum of α-helix. Values of − [θ] at 222 nm for NWASP181–200, N-W(K9G, K15G), N-W(K9E, K15E), and N-W(E7K) were 4.39 × 10 3 , 1.77 × 10 3 , 3.47 × 10 3 , and 4.74 × 103 (deg·cm2·dmol− 1), respectively. These results suggest that the α-helix content of N-W(K9G, K15G) was less than those of the other three peptides. 4. Discussion Since earlier studies showed that aminoglycosides interact with phosphoinositides such as phsophatidylinositol-4-monophospahte (PIP) and phosphatidylinositol-4,5-bisphosphate (PIP2), much attention has been focused on acidic phospholipids as the initial binding site for the uptake of aminoglycosides in the renal proximal tubule. Sastransinh et al. [9] showed several acidic, but not neutral, phospholipids including PIP2 and phosphatidylserine bound gentamicin in a saturable manner. Furthermore, they demonstrated that binding of gentamicin to BBM increased according to elevations of phosphatidylinositol (PI) content of BBM. Sundin et al. [21] found that ischemiainduced increase in apical gentamicin binding was due to increased apical PI content. On one hand, Moestrup et al. [10]

showed that aminoglycosides bind to megalin, a multiligand endocytic receptor which is abundantly expressed in the endocytic apparatus including the brush-border membrane of the renal proximal tubular cells. In megalin knockout mice, a dramatic reduction in renal accumulation of gentamicin was observed, suggesting megalin-mediated endocytosis is the only pathway for renal accumulation aminoglycosides [12]. However, it has not been determined how megalin cooperates with acidic phospholipids in the process of renal accumulation of aminoglycosides. In this study, we investigated involvement of megalin in binding of gentamicin to renal brush-border membranes, which is the initial step in the endocytic pathway. First, we showed that megalin ligands, cytochrome c and lysozyme inhibit gentamicin binding to BBM in a concentration-dependent manner. This observation may indicate that megalin is, at least in part, involved in gentamicin binding to BBM. On one hand, the shapes of the inhibition curves of cytochrome c and lysozyme in Fig. 1 are different from those of basic peptides in Figs. 2 and 3, and therefore the inhibition by cytochrome c and lysozyme may not be competitive. Though the reasons for such differences are not clear at present, it may be related to their molecular sizes, which may affect their accessibilities to the binding site(s) of gentamicin in BBM. Next, we tested effect of trypsin treatment on gentamicin binding to BBM. Western blot analysis showed that megalin almost completely disappeared in trypsin-treated BBM. In the trypsin-treated BBM with little or no megalin, gentamicin binding was significantly decreased, indicating that megalin might be involved in the binding of gentamicin to BBM. On the other hand, it was reported that proteolytic enzymes such as trypsin did not decrease gentamicin binding to BBM [9]. The apparent inconsistency between the previous and present studies might be due to the method for data analysis; whether the binding activity of gentamicin to BBM is corrected by the amount of phospholipid or protein. Actually, we observed much smaller decrease in gentamicin binding when corrected by protein (9.3% decrease) than that corrected by phospholipid (21.8% decrease). Estimation of gentamicin binding to trypsintreated BBM based on the amount of phospholipid would be more appropriate than that based on protein, since trypsin treatment could lead to a decrease in the amount of megalin as well as other proteins. The treatment of BBM with trypsin, however, resulted in a partial decrease in gentamicin binding in spite of the absence of megalin. In addition, N-WASP181–200 still inhibited gentamicin binding to trypsin-treated BBM, though the inhibitory potency was dramatically attenuated as shown in Fig. 6. Therefore, other factors like acidic phospholipids may have to be taken into consideration in the initial interaction of gentamicin with BBM. Since it was reported that little [3H]gentamicin (0.6% of dose) was accumulated in the kidney 24 h after the intraperitoneal injection to megalin knockout mice [12], megalin may play a more important role in subsequent endocytosis after binding to the apical cell surface of the renal proximal tubules. Basic amino acids have been reported to be essential for the binding of megalin ligands to megalin [10]. Therefore, in the former study [13], we designed peptide fragments derived from

J. Nagai et al. / Journal of Controlled Release 112 (2006) 43–50

cytochrome c on the basis of the calculated pI; Cyto22–41, Cyto70–89 and Cyto79–88. In this study, cytochrome c, Cyto22–41, Cyto70–89 and Cyto79–88 inhibited the binding of gentamicin to BBM in a concentration-dependent manner with IC50 values of 3.19, 0.312, 0.288 and 0.240 mM, respectively. This observation may indicate that these peptide fragments have higher affinities for aminoglycoside receptor(s) than the parent protein. On the other hand, in a previous study using OK kidney epithelial cells expressing megalin [13], IC50 values of cytochrome c and Cyto79–88 for gentamicin uptake were almost the same, which were 0.94 and 0.91 mM, respectively. Concerning Cyto22–41 and Cyto70–89, no significant inhibitory effects on gentamicin uptake were observed in OK cells at concentrations up to 0.5 mM. Thus, the IC50 values of cytochrome c and these peptide fragments for gentamicin binding to BBM did not necessarily correspond to those for gentamicin uptake in OK cells. The apparent discrepancy in the IC50 values between BBM and OK cell studies may be related to the methodological differences (binding to BBM and accumulation in OK cells), though further experiments will be required. In order to examine the molecular mechanisms underlying the inhibition of N-WASP181–200 on gentamicin binding, we compared the inhibitory effects of N-WASP181–200 and its mutant peptides on gentamicin binding to BBM. Our experimental design was as follows: megalin has low-density lipoprotein (LDL) receptor type A repeats constituting the ligand binding regions, each consists of about 40 amino acids containing 6 cysteine residues and the SDE (Ser-Asp-Glu) motif. Since the SDE motif in LDL receptor has been reported to be responsible for high-affinity binding of positively charged sequences in ligands [22,23], we focused on two triplets of lysine residue in the sequence of N-WASP181–200. Therefore, we designed N-W(K9G, K15G) and N-W(K9E, K15E), where two lysines at positions 9 and 15 in N-WASP181–200 were replaced by glycines and glutamic acids, respectively. The concentration–response curve of N-W(K9G, K15G) on gentamicin binding to BBM was shifted to the right, compared to that of N-WASP181–200. On the other hand, there was little difference between the inhibitory effects of N-WASP181–200 and N-W(K9E, K15E). In addition, the inhibitory effect of N-W (E7K), where a lysine at position 7 in N-WASP181–200 were replaced by glutamic acid, was examined. Though the net charge of N-W(E7K) was increased by two compared with that of N-WASP181–200, the replacement resulted in a small reduction in the IC50 value. Thus, the inhibitory potency of peptide on gentamicin binding to BBM cannot be simply explained by the total number of basic amino acids nor the net charge of the peptide. Circular dichroism measurements demonstrated that the α-helix content of N-W(K9G, K15G) was remarkably decreased compared to that of N-WASP181– 200 and two other mutant peptides (Fig. 7). Therefore, the decrease in inhibitory potency for gentamicin binding by N-W (K9G, K15G) may be due to a change in the molecular conformation of the peptide. Interestingly, the molecular conformation containing α-helix in the receptor binding region of apolipoprotein E, which is rich in basic amino acids, plays a

49

critical role in its binding to the LDL receptor [24]. Thus, the αhelix conformation may be generally required for the ligand– receptor interactions in the members of the LDL receptor family including megalin. In summary, gentamicin binding to BBM was inhibited by megalin ligands such as cytochrome c and lysozyme, and by basic peptide fragments including N-WASP181–200. In addition, trypsin digestion assay suggested that megalin is, at least in part, involved in the binding of gentamicin to BBM. The α-helix conformation in N-WASP181–200 might be important for interaction with the aminoglycoside binding receptor(s) in BBM. Acknowledgements This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture in Japan. We also thank the Research Center for Molecular Medicine, Faculty of Medicine, Hiroshima University, for the use of their facilities. References [1] H.F. Chambers, The aminoglycosides, in: J.G. Hardman, L.E. Limbird, A.G. Gilman (Eds.), Goodman and Gilman's The Pharmacological Basis of Therapeutics, 10th ed, McGraw-Hill, New York, 2001, pp. 1219–1238. [2] M.P. Mingeot-Leclercq, P.M. Tulkens, Aminoglycoside nephrotoxicity, Antimicrob. Agents Chemother. 43 (1999) 1003–1012. [3] B.A. Molitoris, Cell biology of aminoglycoside nephrotoxicity: newer aspects, Curr. Opin. Nephrol. Hypertens. 6 (1997) 384–388. [4] J. Nagai, M. Takano, Targeted prevention of renal accumulation and toxicity of gentamicin by aminoglycoside binding receptor antagonists, Drug Metab. Pharmacokin. 19 (2004) 159–170. [5] V.U. Collier, P.S. Lietman, W.E. Mitch, Evidence for luminal uptake of gentamicin in the perfused rat kidney, J. Pharmacol. Exp. Ther. 210 (1979) 247–251. [6] R.A. Giuliano, G.A. Verpooten, L. Verbist, R.P. Wedeen, M.E. De Broe, In vivo uptake kinetics of aminoglycosides in the kidney cortex of rats, J. Pharmacol. Exp. Ther. 236 (1986) 470–475. [7] J. Schacht, Purification of polyphosphoinositides by chromatography on immobilized neomycin, J. Lipid Res. 19 (1978) 1063–1067. [8] J. Schacht, Isolation of an aminoglycoside receptor from guinea pig inner ear tissues and kidney, Arch. Oto-Rhino-Laryngol. 224 (1979) 129–134. [9] M. Sastrasinh, T.C. Knauss, J.M. Weinberg, H.D. Humes, Identification of the aminoglycoside binding site in rat renal brush border membranes, J. Pharmacol. Exp. Ther. 222 (1982) 350–358. [10] S.K. Moestrup, S. Cui, H. Vorum, C. Bregengård, S.E. Bjørn, K. Norris, J. Gliemann, E.I. Christensen, Evidence that epithelial glycoprotein 330/ megalin mediates uptake of polybasic drugs, J. Clin. Invest. 96 (1995) 1404–1413. [11] J. Nagai, H. Tanaka, N. Nakanishi, T. Murakami, M. Takano, Role of megalin in renal handling of aminoglycosides, Am. J. Physiol., Renal Fluid Electrolyte Physiol. 281 (2001) F337–F344. [12] C. Schmitz, J. Hilpert, C. Jacobsen, C. Boensch, E.I. Christensen, F.C. Luft, T.E. Willnow, Megalin deficiency offers protection from renal aminoglycoside accumulation, J. Biol. Chem. 277 (2002) 618–622. [13] A. Watanabe, J. Nagai, Y. Adachi, T. Katsube, Y. Kitahara, T. Murakami, M. Takano, Targeted prevention of renal accumulation and toxicity of gentamicin by aminoglycoside binding receptor antagonists, J. Control. Release 95 (2004) 423–433. [14] F.X. Yu, P.A. Johnston, T.C. Südhof, H.L. Yin, gCap39, a calcium ion- and polyphosphoinositide-regulated actin capping protein, Science 250 (1990) 1413–1415.

50

J. Nagai et al. / Journal of Controlled Release 112 (2006) 43–50

[15] P.A. Janmey, J. Lamb, P.G. Allen, P.T. Matsudaira, Phosphoinositidebinding peptides derived from the sequences of gelsolin and villin, J. Biol. Chem. 267 (1992) 11818–11823. [16] K.E. Prehoda, J.A. Scott, R.D. Mullins, W.A. Lim, Integration of multiple signals through cooperative regulation of the N-WASP-Arp2/3 complex, Science 290 (2000) 801–806. [17] M. Takano, J. Nagai, T. Okano, K. Hirozane, M. Yasuhara, R. Hori, Differential sensitivity of organic anion transporters in rat renal brushborder membrane to diethyl pyrocarbonate, Biochim. Biophys. Acta 1192 (1994) 190–196. [18] Y. Shiraishi, J. Nagai, T. Murakami, M. Takano, Effect of cisplatin on H+ transport by H+-ATPase and Na+/H+ exchanger in rat renal brush-border membrane, Life Sci. 67 (2000) 1047–1058. [19] M. Takano, N. Nakanishi, Y. Kitahara, Y. Sasaki, T. Murakami, J. Nagai, Cisplatin-induced inhibition of receptor-mediated endocytosis of protein in the kidney, Kidney Int. 62 (2002) 1707–1717.

[20] M.M. Bradford, A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding, Anal. Biochem. 72 (1976) 248–254. [21] D.P. Sundin, C. Meyer, A. Geerdes, R. Sandoval, B.A. Molitoris, Cellular mechanism of aminoglycoside tolerance in long-term gentamicin treatment, Am. J. Physiol. 272 (1997) C1308–C1318. [22] A. Saito, S. Pietromonaco, A.K. Loo, M.G. Farquhar, Complete cloning and sequencing of rat gp330/“megalin,” a distinctive member of the low density lipoprotein receptor gene family, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 9725–9729. [23] E.I. Christensen, H. Birn, Megalin and cubilin: multifunctional endocytic receptors, Nat. Rev., Mol. Cell Biol. 3 (2002) 256–266. [24] R.W. Mahley, Apolipoprotein E: cholesterol transport protein with expanding role in cell biology, Science 240 (1988) 622–630.