0022-3565/01/2961-124 131$3.

00 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Copyright 2001 by The American Society for Pharmacology and Experimental Therapeutics JPET 296:124131, 2001

Vol. 296, No. 1 0/872061 Printed in U.S.A.

Enhanced Delivery of Doxorubicin into the Brain via a PeptideVector-Mediated Strategy: Saturation Kinetics and Specificity
CHRISTOPHE ROUSSELLE, MARIA SMIRNOVA, PHILIPPE CLAIR, JEANNE-MARIE LEFAUCONNIER, ALAIN CHAVANIEU, BERNARD CALAS, JEAN-MICHEL SCHERRMANN, and JAMAL TEMSAMANI
Institut National de la Sante et de la Recherche Medicale U26, Hopital Fernand Widal, Paris, France (C.R., M.S., J.-M.L., J.-M.S.); Synt:em, Parc Scientifique Georges Besse, Nmes, France (P.C., J.T.); and Center de Biochimie Structurale, Faculte de Pharmacie, Montpellier I, France (A.C., B.C.) Received July 6, 2000; accepted September 22, 2000 This paper is available online at http://jpet.aspetjournals.org Downloaded from jpet.aspetjournals.org at Universitaet Zuerich on January 31, 2012

ABSTRACT Doxorubicin delivery to the brain is often restricted because of the poor transport of this therapeutic molecule through the blood-brain barrier (BBB). To overcome this problem, we have recently developed a technology, Pep:trans, based on short natural-derived peptides that are able to cross efficiently the BBB without compromising its integrity. In this study, we have used the in situ mouse brain perfusion method to evaluate the brain uptake of free and vectorized doxorubicin. Doxorubicin was coupled covalently to small peptide vectors: L-SynB1 (18 amino acids), L-SynB3 (10 amino acids), and its enantio form D-SynB3. We first confirmed the very low brain uptake of free radiolabeled doxorubicin, which is most likely due to the efflux

activity of the P-glycoprotein at the level of the BBB. Vectorization with either L-SynB1, L-SynB3, or D-SynB3 significantly increased the brain uptake of doxorubicin (about 30-fold). We also investigated the mechanism of transport of vectorized doxorubicin. We show that vectorized doxorubicin uses a saturable transport mechanism to cross the BBB. The effect of poly(L-lysine) and protamine, endocytosis inhibitors, on the transport across the brain was also investigated. Both inhibitors reduced the brain uptake of vectorized doxorubicin in a dosedependent manner. These studies indicate that the transport of vectorized doxorubicin appears to occur via an adsorptivemediated endocytosis.

The pharmacological treatment of brain diseases is often complicated by the inability of potent drugs to pass across the BBB, which is formed by the tight endothelial cell junctions of capillaries within the brain. Malignant brain tumors respond poorly to chemotherapy because most of the anticancer agents cannot be delivered efficiently to the tumor site. For example the anthracycline doxorubicin, one of the most powerful anticancer agents, cannot be delivered systemically in effective concentrations to the tumor cells because of multidrug-resistance mechanisms (Ohnishi et al., 1995; MarbeufGueye et al., 1997). To overcome the limited access of drugs to the brain different methods have been developed that achieve BBB penetration (Temsamani et al., 2000). Some of these methods are characterized, for instance, by osmotic BBB opening or by the use of biologically active agents such as bradykinin (Kroll et al., 1998). Specific transport mechanisms have also been exploited, involving the activity of several independent transporters that mediate the influx of substances important for brain function, such as monocarboxylic acid, and amine and neutral amino acid transporters (Terasaki and Tsuji, 1994). A further strategy has
This study was supported partly by the Anvar Languedoc Roussillon.

been to incorporate the therapeutic drugs within delivering devices such as nanoparticles and liposomes (Huwyler et al., 1996; Gulyaev et al., 1999) or to conjugate the drug with a protein (Singh, 1999), a peptide vector (Pardridge et al., 1994; Schwarze et al., 1999), or with an antibody (Pardridge, 1999). This variety of strategies reflects the inherent difficulty in delivering drugs across the BBB. Recently, we have shown that small peptide vectors can be used to enhance brain uptake of therapeutic drugs (Rousselle et al., 2000). These peptide vectors cross cellular membranes efficiently and have been used to enhance penetration of a number of drugs into live cells (Derossi et al., 1998). The SynB vectors are derived from natural peptides called protegrins (Harwig et al., 1995; Mangoni et al., 1996). They possess an amphipathic structure in which the positively charged and hydrophobic residues are separated in the sequence. They are able to cross efficiently cell membranes without any cytolytic effect. We have shown recently (Rousselle et al., 2000) by in situ rat brain perfusion and by intravenous injection in mice that coupling of doxorubicin to SynB1 vector enhances significantly its brain uptake. Additionally, the coupling led to a significant decrease of doxorubicin concentrations in the heart.

ABBREVIATIONS: BBB, blood-brain barrier; HPLC, high performance liquid chromatography; TFA, trifluoroacetic acid; dox, doxorubicin; DMF, dimethylformamide; P-gp, P-glycoprotein. 124

Transport Mechanism of Coupled Doxorubicin through the BBB

125

The mechanism by which these peptides cross the BBB is not known. Various peptides and proteins with a high permeability have been found to cross the BBB by several saturable mechanisms, including carrier-mediated transport of small peptides with an N-terminal tyrosine (Banks et al., 1986), receptor-mediated transcytosis of insulin (Duffy and Pardridge, 1987) or transferrin (Fishman et al., 1987), and adsorptive-mediated transcytosis of positively charged proteins, e.g., cationized albumin and -endorphin-cationized albumin chimeric peptide (Smith and Borchardt, 1989). The aim of this study was to confirm the efficacy of the SynB vectors to deliver doxorubicin through the mouse BBB by the use of an in situ mouse brain perfusion model that we have recently developed (Dagenais et al., 2000) and to gain insight into the mechanism of transport. Complementary techniques were associated to it to measure the fraction of doxorubicin trapped into microvessel cells or present in brain parenchyma. We have used [14C]doxorubicin coupled to SynB1 (18 amino acids) as well as to a truncated derivative of SynB1: SynB3 (10 amino acids) and its enantio form D-SynB3. We show that coupling of doxorubicin to SynB vectors improves its penetration across the BBB with the same efficiency for all the SynB vectors used. We also demonstrate that the mechanism of transport of coupled doxorubicin involves a saturable system, which may operate via an adsorptive-mediated endocytosis.

purchased from Sigma (St-Quentin, France). Peptide synthesis chemicals were obtained from Novabiochem (Laufelfingen, Switzer land). All other chemical and reagents were commercial products of reagent grade.

Preparation and Characterization of Peptide Conjugates

Peptide Synthesis. The peptides were assembled by conventional solid phase chemistry using a 9-fluorenylmethoxycarbonyl/tertiobutyl protection scheme (Atherton and Sheppard, 1989) and purified on preparative C18 reverse phase HPLC after trifluoroacetic acid (TFA) cleavage/deprotection. Purity of the lyophilized products was assessed by C18 reverse phase analytic HPLC and their molecular weight checked by matrix-assisted laser desorption-ionization time-of-flight mass spectrometry. The peptides sequences were SynB1 (RGGRLSYSRRRFSTSTGR, molecular mass, 2099 Da) L-SynB3 (RRLSYSRRRF, molecular mass, 1395 Da), and D-SynB3 (RRLSYSRRRF, the amino acids are in D form, molecular mass, 1395 Da). Dox-SynB Synthesis (Fig. 1). Doxorubicin hydrochloride was suspended in dimethylformamide (DMF) containing diisopropylethylamine. Succinic anhydride (1 M equivalent; Fluka, Buchs, Switzerland) dissolved in DMF was added and incubated for 20 min. The resulting doxorubicin hemisuccinate was then activated by addition of benzotriazol-1-yl-oxopyrrolidinephosphonium hexafluorophosphate (1.1 M equivalent; Novabiochem, Laufelfingen, Switzerland) dissolved in DMF. The peptide was then added to the reaction mixture after 5 min of activation and left for another 20 min for coupling. Further processing and purity check of the conjugate was performed as described above. The molecular mass was found to be 2723 Da for dox-SynB1 and 2019 Da for both dox-SynB3 and dox-DSynB3. Radiolabeling of Dox-SynB. Preparations were performed as described above, except that [14C]doxorubicin was kept limiting by raising the stoichiometry of peptide, linkers, and activators to 1.3 M equivalent in the coupling reactions. The specific activity of doxSynB1, dox-SynB3, and dox-D-SynB3 was 55 mCi/mmol, 2.04 Tbq/ mol and the molar ratio of doxorubicin/peptide was 1:1. The radiochemical purity was estimated to be 98% according to the 480-nm chromatograms.

Downloaded from jpet.aspetjournals.org at Universitaet Zuerich on January 31, 2012

Materials and Methods

Chemicals
[14C]Doxorubicin (specific activity 55 Ci/mol) was purchased from Amersham Pharmacia Biotech (Orsay, France). [3H]Sucrose (specific activity 10.20 Ci/mmol) was obtained from DuPont New England Nuclear (Paris, France). Doxorubicin hydrochloride, protamine, and poly(L-lysine) hydrobromide (up to mol. wt. 5000) were

Fig. 1. Structure of dox-SynB1 and dox-SynB3.

126

Rousselle et al. marker that does not cross the BBB significantly. In some experiments, the brain uptake of [14C]dox-SynB was evaluated in the presence of various concentrations of unlabeled dox-SynB (0 100 M). In other experiments we have studied the influence of polycationic peptides: poly(L-lysine) (0 25 M) and protamine (25 M) on the distribution of dox-SynB. Determinations of BBB Transport Constants in the Right Mice Brain Hemisphere. Briefly, calculations were accomplished as previously described by Smith (1996). The integrity of the BBB was determined in each animal by the brain vascular volume (Vv, l g 1) estimated by the tissue distribution of [3H]sucrose from the following relationship:

Distribution Coefficient Determination. The lipophilicity of the radiolabeled free and vectorized doxorubicin was estimated by measuring their partitioning between the perfusion buffer (pH 7.4) and octan-1-ol. Distribution coefficients (Doctanol/buffer pH 7.4) were determined at volume ratios of 1:1 by vigorously shaking the two phases together. The samples were then incubated at 37C for 30 min to facilitate phase separation. One sample of each phase was weighed and the radioactivity counted in a gamma counter. Doctanol/ buffer pH 7.4 was calculated as ([dpm/ml] in the octanol phase)/([dpm/ ml] in the buffered-saline phase). Experiments were done in triplicate and the mean of the log Doctanol/buffer pH 7.4 was calculated. Stability in Vitro. Dox-SynB solution (1 ml, 2 mg/ml) was mixed with 4 ml of mouse plasma (obtained from Iffa-Credo, LArbresle, France). At various times (0, 15, 25, 30, 40, 45, 60, 120, 180, 210 min), 250- l aliquots were withdrawn and quenched in 1 ml of acid mixture (H2O/TFA 0.1%). The vectorized doxorubicin and metabolites were then extracted from plasma by applying the sample on a C18 SPE cartridge and eluting in 500 l of acetonitrile/isopropanol/H2O/ TFA (50:20:30:5, v/v) solution. The samples were then analyzed by HPLC at 418 nm on C18 column using acetonitrile/water gradient. The percentage of nondegraded vectorized doxorubicin and released doxorubicin was calculated.

VV

Q* tot/C* pf

(1)

where Q*tot is the amount of radiolabeled sucrose in the right brain hemisphere (dpm g 1) and C*pf is the perfusate concentration of sucrose (dpm l 1). Doxorubicin uptake was expressed as the volume of distribution (Vd) from the following relationships:

Downloaded from jpet.aspetjournals.org at Universitaet Zuerich on January 31, 2012

Vd

Q* br/C* pf

(2)

Animals
Adult OF1 mice (30 40 g, 6 8 weeks old) were obtained from Iffa-Credo. Animals were maintained under standard conditions of temperature and lighting and had free access to food and water. The research adhered to the ethical rules of the French Ministry of Agriculture for experimentation with laboratory animals (Law No. 87-848).

where Q*br is the calculated quantity of 14C-tracer per gram of right brain hemisphere and C*pf (dpm l 1) is the labeled tracer concentration measured in the perfusate. Vascular radioactivity was subtracted from tissue sample as follows:

Q* br

Q* tot

V vC* pf

(3)

In Situ Mouse Brain Perfusion Study

Surgical Procedure. The uptake of free or vectorized [14C]doxorubicin to the luminal side of mouse brain capillaries was measured using the in situ brain perfusion method previously adapted in our laboratory for the study of drug uptake in the mouse brain (Dagenais et al., 2000). Briefly, the right common carotid of ketamine/xylazine (140:8 mg/kg i.p.)-anesthetized mice was exposed and ligated at the heart side. The external carotid artery was ligated at the level of the bifurcation of the common carotid, rostral to the occipital artery. The common carotid was then catheterized rostrally with polyethylene tubing (0.30 mm i.d. 0.70 mm o.d.; Biotrol Diagnostic, Chennevrieres-les-Louvres, France) filled with heparin (25 U/ml) and mounted ` on a 26-gauge needle. The syringe containing the perfusion fluid was placed in an infusion pump (Harvard pump PHD 2000; Harvard Apparatus, Holliston, MA) and connected to the catheter. Immediately before the perfusion, the heart was stopped by severing the ventricles to eliminate contralateral blood flow contribution. Brains were perfused for 20 to 120 s at a flow rate of 2.5 ml/min. At the end of the perfusion time, the mouse was decapitated and the brain removed. The right hemisphere and samples of perfusion fluid were placed in preweighed scintillation vials and weighed. Brain and perfusion samples were then digested for 2 h in 1 ml of Solvable (Packard, Rungis, France) at 50C and mixed with 9 ml of Ultima Gold XR scintillation cocktail (Packard). Total [14C] and [3H] were determined simultaneously in a Packard Tri-Carb model 1900 TR liquid scintillation analyzer and activities were converted from counts per minute to disintegrations per minute with the use of internally stored quenching curves. Transport Studies of Radiolabeled [14C]Doxorubicin. The perfusate consisted of a Krebs-bicarbonate buffer: 128 mM NaCl, 24 mM NaHCO3, 4.2 mM KCl, 2.4 mM NaH2PO4, 1.5 mM CaCl2, 0.22 mM MgSO4, and 9 mM D-glucose added before infusion. The solution was gassed with 95% O2 and 5% CO2 for pH control ( 7.4) and warmed at 37C in a water bath. Tracers were added to perfusate at concentrations of 0.4 Ci/ml for free doxorubicin, 0.1 Ci/ml for vectorized doxorubicin, and 0.3 Ci/ml for [3H]sucrose, a vascular

where Q*tot (dpm g 1) is the total quantity of tissue tracer measured in the tissue sample (vascular extravascular). The combination of eqs. 2 and 3 gives the following:

Vd
The influx Jin ( mol s
1

Q* tot
g
1

V vC* pf /C* pf

(4)

) was calculated as follows:

J in

V d C pf /T

(5)

where Cpf is the total concentration (0.5100 M) of dox-SynB. The concentration dependence of dox-SynB uptake was analyzed using the following relationship:

J in

V maxC pf / K m

C pf

K DC pf

(6)

where Vmax is the maximal transport velocity ( mol s 1 g 1), Km is the concentration ( M) at half-maximal velocity, and KD is the constant of nonsaturable transfer ( l s 1 g 1). Distribution in Brain Microvascular and Parenchymal Compartments. The distribution of [14C]doxorubicin between brain microvascular and parenchymal compartments was assessed using the capillary depletion method of Triguero et al. (1990) with some modifications (Rousselle et al., 2000). For this set of experiments, we used a dual-syringe infusion pump (Harvard Apparatus, Les Ulis, France) with one syringe containing the bicarbonate-buffered physiological saline with the radiotracer (syringe A) and the other without radiotracer (syringe B). The carotid catheter was connected to a four-way valve (Hamilton, Bonnaduz, Switzerland). After the carotid cannulation was completed and the appropriate connections were made, syringe A was discharged at a rate of 2.5 ml/min for 60 s. Syringe A was switched off and syringe B was switched on simultaneously to initiate the washout of the capillary space. After 60 s the mouse was decapitated and the right cerebral hemisphere was removed, weighed, and homogenized in 1 ml of capillary buffer (10 mM Hepes, 141 mM NaCl, 4 mM KCl, 1 mM NaH2PO4, 2.8 mM CaCl2, 1 mM MgSO4, and 10 mM D-glucose, pH 7.4) on ice. After 15 strokes, 1 ml of a chilled 37% neutral dextran solution was added to obtain a final concentration of 18.5%. All homogenizations were performed at 4C in a very short time. After taking an aliquot of homogenate, the

Transport Mechanism of Coupled Doxorubicin through the BBB solution was centrifuged at 3800g for 25 min at 4C in a swingingbucket rotor. The pellet and supernatant were carefully separated and counted in the liquid scintillation counter. The pellet was composed mainly of brain capillary and the supernatant reflected brain parenchyma. Doxorubicin distribution in brain endothelial cells and parenchyma was expressed as partition percentages defined as follows:

127

Q* tis/Q* hg
14

(7)

where Q*tis is the measured quantity of [ C]doxorubicin in brain parenchyma or endothelial cells (total dpm per compartment/brain tissue weight) and Q*hg the quantity of [14C]doxorubicin in brain homogenate. Statistical Analysis. All experiments were performed on three to six mice. Data are expressed for the right cerebral hemisphere. Statistical comparisons conducted herein were accomplished by Students unpaired t test or ANOVA. Bonferronis multiple comparison test was used post hoc only when ANOVA results were significant. Statistical difference was accepted at the P 0.05 significance level. Data are the mean S.E.M. Estimates of the dox-SynB transport parameters (Km and Vmax) were obtained by fitting a MichaelisMenten-type equation to the uptake rate versus perfusate dox-SynB concentration data by nonlinear least-square regression using the SYSTAT software (SYSTAT, Inc., Evanston, IL).

Results
The radiolabeled model dox-peptides were shown to have a purity of at least 98%. To rule out that any degradation of the peptide vectors took place during the perfusion, we assessed the stability of the vectorized doxorubicin in mice plasma in vitro. Our results show that dox-SynB1 and dox-L-SynB3 have a degradation half-life of about 15 to 20 min in mice plasma, whereas for dox-D-SynB3, no significant degradation took place even after 48-h incubation in plasma (data not shown). Therefore, we assumed that no significant degradation of the compounds had taken place after 60 s of brain perfusion. First, the brain uptake of [14C]doxorubicin in OF1 mice was studied by the in situ mouse brain perfusion method (Fig. 2). To choose a reliable perfusion time that allows a

sufficient accumulation in the brain tissue, a time course study of brain distribution of doxorubicin was performed. The distribution volume (Vd) of doxorubicin to the brain was limited over 20 to 120 s but a sufficient accumulation of doxorubicin in the right brain hemisphere was reached after 60 s of perfusion, which was consequently the perfusion time selected for the following studies. For each mouse perfused, we have checked the integrity of the BBB by measuring the distribution volume of [3H]sucrose. Sucrose is commonly used as a marker of the integrity of the BBB because it does not measurably penetrate the BBB during short perfusion time. No differences in brain vascular volumes were observed between all perfusion times and the values measured were similar to those previously reported (Fig. 2). We then compared the brain uptake of free and coupled radiolabeled doxorubicin by measuring the total radioactivity in the right hemisphere of OF1 mice after 60 s of brain perfusion (Fig. 3A). When doxorubicin was coupled to SynB1, L-SynB3 (the amino acids are in the L form), or D-SynB3 (the amino acids are in the D form), a significant increase in brain uptake was obtained. An average of 30-fold increase in [14C]doxorubicin brain uptake was observed after its conjugation with either SynB1, L-SynB3, or D-SynB3 (Fig. 3A). It is noteworthy that no difference in brain uptake was observed between dox-L-SynB3 and dox-D-SynB3. This suggests that the mechanism of brain uptake does not involve a chiral receptor. To verify that this enhancement in brain uptake

Downloaded from jpet.aspetjournals.org at Universitaet Zuerich on January 31, 2012

Fig. 2. The time-dependant uptake of [14C]doxorubicin ( ) and [3H]sucrose (F) in the right brain hemisphere, expressed as apparent brain distribution volume measured with the in situ mouse brain perfusion method. Data are presented as means S.E.M. (n 4 6 animals). For [14C]doxorubicin values have been corrected for [3H]sucrose/vascular space.

Fig. 3. Distribution volumes of [14C]doxorubicin, [14C]dox-L-SynB1, [14C]dox-L-SynB3, and [14C]dox-D-SynB3 (A) and [3H]sucrose (B) in the right hemisphere of OF1 mice. The mice were perfused for 60 s with 3 M doxorubicin, 0.5 M dox-SynB1, dox-L-SynB3, and dox-D-SynB3. Each column represents a mean S.E.M. for n 4 to 6 mice. Statistical significance is indicated by ***P 0.001 using ANOVA followed by Bonferronis test with the sucrose space subtracted.

128

Rousselle et al.

does not result from a loss in the integrity of the BBB, we simultaneously measured the distribution volume of [3H]sucrose (Fig. 3B). No differences in the sucrose vascular volumes were observed between all the tested compounds. To evaluate whether coupled doxorubicin has actually crossed the BBB or is simply trapped within brain endothelial cells, a washing procedure followed by the capillary depletion method of Triguero et al. (1990), which separates the whole brain into endothelial-enriched (pellet) and -depleted (supernatant) fractions, was performed. This procedure distinguishes between compounds remaining in the endothelial cells from those having crossed the abluminal endothelial membrane to enter the brain parenchyma. By this method, we observed that about half of the vectorized doxorubicinderived radioactivity was associated with the parenchymal fraction after 60 s of perfusion followed by 60 s of washout (60.6 7.7 and 44.3 4.5% for dox-L-SynB1 and dox-LSynB3, respectively) (Fig. 4). To explore the mechanism by which these peptide vectors cross the BBB, the uptake of [14C]dox-SynB was studied under conditions in which the brain was perfused with increasing concentrations of dox-SynB (0 100 M). Figure 5 shows the brain uptake of [14C]dox-SynB1, [14C]dox-SynB3, and [14C]dox-D-SynB3 in the presence of various concentrations of unlabeled dox-SynB1, dox-SynB3, and dox-D-SynB3, respectively. Brain uptake of the three compounds after 60 s of perfusion was shown to be saturable. The Km and Vmax values are represented in Table 1. For example, the brain uptake of dox-D-SynB3 had a Km of 9.0 M and a Vmax of 134 mol s 1 g 1. To further shed light on the mechanism of transport, we used various inhibitors of endocytosis. Figure 6 shows the effect of increasing concentrations of poly(L-lysine) (0 25 M) on the brain uptake of [14C]dox-D-SynB3. The concentration of dox-D-SynB3 used in this study was 0.5 M and is lower that the half-saturation constant of the peptide transporter (Km 9.0 M), which was thus not saturated. Addition of increasing concentrations of poly(L-lysine) led to a significant inhibition of [14C]dox-D-SynB3 brain uptake in a dose-dependent manner. The blood-brain transport of [14C]dox-D-SynB3 was also dramatically decreased by 25 M protamine (Table 2). Similar results were obtained for [14C]dox-SynB1 and for [14C]dox-L-SynB3 (Table 2). During

Downloaded from jpet.aspetjournals.org at Universitaet Zuerich on January 31, 2012

Fig. 5. Rate of [14C]dox-SynB uptake in the right hemisphere as a function of dox-SynB concentration in perfusate after 60 s of perfusion. Values have been corrected for [3H]sucrose/vascular space. The solid line represents a fit of the Michaelis-Menten equation to the data by nonlinear least-square regression using the SYSTAT software (SYSTAT, Inc.). TABLE 1 Kinetic parameters for the brain uptake of dox-L-SynB1, dox-L-SynB3, and dox-D-SynB3
Km M Vmax mol/s/g 106

Dox-L-SynB1 Dox-L-SynB3 Dox-D-SynB3

5.5 4.1 9.0

76 85 134

Fig. 6. Effects of various concentration of poly(L-lysine) on the uptake of [14C]dox-D-SynB3 by mouse brain. The murine brains were perfused for 60 s with 0.5 M dox-D-SynB3 and increasing concentrations of poly(Llysine) (0 25 M). Values have been corrected for [3H]sucrose/vascular space. Each point represents the mean S.E.M. for n 4 to 6 mice.

all experiments, the vascular volumes were monitored and found to be similar to those previously reported (Table 2).

Discussion
The results of our experiments show that coupling doxorubicin with small peptide vectors results in delivery of this drug to the brain of experimental animals. Doxorubicin, despite a favorable partition coefficient (log Doctanol/buffer, pH 7.4 0.45 0.06) failed to cross the BBB to an appreciable extent. Its brain uptake measured by the in situ mouse brain perfusion method is indeed very low (Vd 28.2 4.2 l/s/g after 60 s of perfusion)

Fig. 4. Distribution rate of [14C]dox-L-SynB1 and [14C]dox-L-SynB3 in vascular pellet and supernatant fractions after washout and capillary depletion. Values are mean S.E.M. (n 5 mice).

Transport Mechanism of Coupled Doxorubicin through the BBB

TABLE 2 Effects of endocytosis inhibitors on the mouse brain uptake of dox-LSynB1, dox-L-SynB3, and dox-D-SynB3
Mouse brain entry of [14C]dox-L-SynB1, [14C]dox-L-SynB3, and [14C]dox-D-SynB3 (0.5 M) was determined in the presence of 100 M unlabeled conjugates and the polycations poly(L-lysine) (25 M) or protamine (25 M) after 60 s of perfusion. The distribution volumes of [3H]sucrose representing vascular spaces (Vvasc) are also shown. Means S.E.M. (n 4 8 animals) have been corrected for the vascular space. Vd l/g Vvasc

129

[14C]Dox-L-SynB1 (0.5 M) only Dox-L-SynB1 (100 M) Poly(L-lysine) (25 M) Protamine (25 M) [14C]Dox-L-SynB3 (0.5 M) only Dox-L-SynB3 (100 M) Poly(L-lysine) (25 M) Protamine (25 M) [14C]Dox-D-SynB3 (0.5 M) only Dox-D-SynB3 (100 M) Poly(L-lysine) (25 M) Protamine (25 M)

776.4 66.3 61.2 45.6 961.8 55.8 55.8 31.8 788.4 52.8 51 28.2

37.8 6.6 9.6 4.2 96.6 6.6 11.4 3.6 49.8 3.6 7.2 3

16.9 17.7 14.4 14.2 17.4 16.6 11.5 14.4 15.3 17.1 13.3 12.0

0.9 0.7 0.9 0.5 1.7 1.0 1.0 0.9 1.2 2.0 0.4 0.1

and was similar to that obtained previously in rats (Rousselle et al., 2000). This low penetration may be related to the 170-kDa ATP-dependent efflux pump P-gp, which has been shown to be present at the luminal site of the endothelial cells of the BBB (Van Asperen et al., 1999). Our experiments using P-gp-deficient mice confirmed the efflux of doxorubicin by P-gp at the BBB because a higher Kin was obtained in this strain compared with normal mice (S. Cisternino, C. Rousselle, C. Dagenais, J. M. Lefauconnier, and J. M. Scherrmann, submitted). However, the efflux activity of P-gp may not be sufficient to explain the poor penetration of doxorubicin in relation to its chemical properties, even in the deficient mice. It is possible that other efflux pumps, such as the multidrug resistance-associated protein, may be responsible for the poor penetration of doxorubicin through the BBB. During in vitro studies using cancer cells, doxorubicin has been shown to be excluded by multidrug resistance-associated protein (Abe et al., 1994; Marbeuf-Gueye et al., 1997). Coupling doxorubicin with SynB vectors resulted in a significant enhancement of its brain uptake although the doxSynB compounds are less lipophilic (log Doctanol/buffer, pH 7.4 1.44 for dox-SynB1, 1.27 for dox-L-SynB3, and 1.30 for dox-D-SynB3). An average of 30-fold increase was obtained after coupling the doxorubicin with SynB1, L-SynB3, or D-SynB3. These results confirmed the efficacy of SynB vectors in delivering doxorubicin into the brain without compromising the integrity of the tight junctions. It is noteworthy that SynB3, a truncated derivative of SynB1, gave similar enhancement of doxorubicin brain uptake as SynB1. To demonstrate that vectorized doxorubicin is not trapped inside the endothelial cells but has actually crossed the BBB, we carried out the wash-out procedure and the capillary depletion method. Our results indicate that about 50% of doxorubicin-derived radioactivity was associated with the parenchyma after 60 s of brain perfusion, suggesting the efficiency of these peptide vectors in delivering doxorubicin to the brain parenchyma. Because the amount of free doxorubicin crossing the BBB is very low, resulting in dpm close to the background of the detection method, the distribution between brain compartments could not be performed for this compound. However, we had previously shown in rats, using

a similar method, that about 50% of free doxorubicin was distributed in the parenchyma (Rousselle et al., 2000) and that coupling of doxorubicin with SynB1 led to a 20-fold increase in the amount of doxorubicin transported into brain parenchyma. We then explored the mechanism by which these doxpeptide complexes cross readily the BBB. The kinetic analysis of [14C]dox-SynB was determined by using increasing concentrations of the compound ranging from 0.5 to 100 M. The results of this study demonstrate that vectorized doxorubicin enters the brain via a saturable mechanism that can be described by Michaelis-Menten-type kinetics, exhibiting a relatively high affinity and a low capacity (Table 1). A diffusional component was not necessary to model the data because the amount of vectorized doxorubicin crossing the BBB by passive diffusion over 60 s is not significant. Some other peptides have been found to enter the central nervous system by several saturable mechanisms, including adsorptive-mediated endocytosis (Tamai et al., 1997), receptor-mediated transport (Pardridge et al., 1987), and carrier-mediated transport (Banks et al., 1986). The observed Km values (4.19 M) measured in this study are comparable to those for substrates reported to be taken up into brain endothelial cells via the adsorptive-mediated endocytosis mechanism, including ebiratide (15.9 M; Terasaki et al., 1992), E-2078 (4.62 M; Terasaki et al., 1989), histone (15.2 M; Pardridge et al., 1989), and cationized bovine serum albumin (0.8 M; Kumagai et al., 1987). The Kd values for receptor-mediated endocytosis reported for atrial natriuretic factor (0.4 nM; Smith et al., 1988), transferrin (5.6 nM; Pardridge et al., 1987), and insulin (2.3 nM; Frank et al., 1986) are several thousand times smaller than those for adsorptive-mediated endocytosis. Moreover, the fact that the D-SynB3 composed entirely of D-amino acids increased the brain uptake of doxorubucin with the same efficiency as L-SynB3, suggests that the mechanism of transport is nonstereospecific, as would be required for receptor-mediated transport. All these data indicate that the mechanism of transport of dox-SynB is unlikely to be via receptor-mediated transcytosis. In an attempt to further clarify the saturable mechanism by which SynB vectors transport doxorubicin across the BBB, we examined the brain uptake in the presence of polycationic substances. First, we assessed the effect of poly(L-lysine), a known endocytosis inhibitor. Poly(L-lysine) has been shown to inhibit the binding of E-2078 (Terasaki et al., 1989) and ebiratide (Terasaki et al., 1992). Intracarotid infusions of 5 mg of polycations [poly(L-lysine) and protamine] have also been reported to induce an extravasation of albumin, suggesting an opening of the BBB (Westergren and Johansson, 1993). However in this study, the sucrose space was systematically measured to assess the effect of polycations on the basal permeability of the BBB. The values measured for basal, poly(L-lysine)- (25 M), and protamine (25 M)-pretreated mice are well within the range of normal sucrose spaces reported previously (Dagenais et al., 2000). Therefore, in the conditions used in our studies, poly(L-lysine) and protamine have no significant effect on the basal permeability of the BBB during a 60-s exposure. The discrepancy observed between the studies of Westergren and Johansson (1993) and our own might be explained by the higher concentrations of polycations used in the former. Our results showed that the uptake of vectorized doxorubicin was inhibited in a dose-

Downloaded from jpet.aspetjournals.org at Universitaet Zuerich on January 31, 2012

130

Rousselle et al.
Acknowledgments

dependent manner by poly(L-lysine), suggesting an endocytosis mechanism. Infusion of polycationic molecules usually results in neutralization of the negative surface charges (Nagy et al., 1983). It is known that the endothelial cell membranes of the BBB have a net negative charge (Vorbrodt, 1989) originating from the sialic acid or heparan sulfate residues on the surface of endothelial cells. The peptides used in this study are positively charged (six positive charges for SynB1 and five positive charges for both L- and D-SynB3), this net positive charge is likely to play a major role in the adsorptive-mediated endocytosis. This is confirmed by the fact that when doxorubicin was coupled to a peptide vector in which positively charged residues have been replaced by neutral amino acids, no brain uptake was seen (data not shown). The inhibitory effect on brain penetration of dox-DSynB3 observed with polycationic compounds strongly suggests that electrostatic interactions of the peptide vector with the surface of endothelial cells play an important role in the surface binding and subsequent internalization of the vectorized doxorubicin into the brain capillaries. This kind of electrostatic interactions between cationic proteins and negative charges mediate the adsorptive endocytosis (Vorbrodt, 1989). All these results indicate that the vectorized doxorubicin might be transported by an adsorptive-mediated endocytosis system. For the transcytosis of peptides through the BBB, three steps have been proposed: 1) binding and internalization at the luminal side of endothelial cell membrane, 2) diffusion through the cytoplasm of endothelial cells, and 3) externalization at the abluminal side of endothelial cell membrane (Bar et al., 1983). Our results suggest that adsorptive-mediated endocytosis occurs at least at the luminal side of brain capillaries. The similarity in behavior observed for the three peptide vectors used in this study suggests that the externalization at the abluminal side of endothelial cells may also not be via a receptor-mediated mechanism. However, because the endocytosis inhibitors have only been tested at the luminal side of the endothelial cells, one cannot exclude that a different mechanism may be involved in the externalization step. Few strategies have been described that result in an improved uptake of doxorubicin into cells. Doxorubicin has been given in combination with P-gp inhibitors (Kusunoki et al., 1998) but if such combinations are effective in vitro, results of studies in patients with solid tumors have been until now somewhat disappointing (Fisher and Sikic, 1995). Arap et al. (1998) have identified peptides that specifically ferry doxorubicin to tumors in nude mice implanted with breast cancer xenographts but to our knowledge, no data regarding the brain uptake have been disclosed. Gulyaev et al. (1999) have observed high brain concentrations of doxorubicin bound to nanoparticles overcoated with surfactant polysorbate 80 but this effect may be related to the toxicity of the carrier against the BBB (Olivier et al., 1999). Our studies indicate that coupling doxorubicin with SynB vectors enhances its brain uptake. The vectorized doxorubicin bypasses P-gp at the BBB and is most likely transported via adsorptive-mediated transcytosis. Our results suggest the possibility of using this system for the treatment of brain cancer and other central nervous system diseases. Toward this goal, future studies are in progress to explore the biological activity of vectorized drugs in animal models.

We thank Dr. Michel Kaczorek and Professor Anthony Rees for helpful advice and criticism, Dr. Pierre Vidal for the in vitro stability work, and Marcel Debray for help in the statistical analysis.
References
Abe T, Hasegawa S, Taniguchi K, Yokomizo A, Kuwano T, Ono M, Mori T, Hori S, Kohno K and Kuwano M (1994) Possible involvement of multidrug-resistanceassociated protein (MRP) gene expression in spontaneous drug resistance to vincristine, etoposide and adriamycin in human glioma cells. Int J Cancer 58:860 864. Arap W, Pasqualini R and Ruoslahti E (1998) Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science (Wash DC) 279:377380. Atherton E and Sheppard RC (1989) Solid Phase Peptide Synthesis, A Practical Approach. IRL Press, Oxford. Banks WA, Kastin AJ, Fischman AJ, Coy DH and Strauss SL (1986) Carriermediated transport of enkephalins and N-tyr-MIF-1 across blood-brain barrier. Am J Physiol 251:E477E482. Bar RS, DeRose A, Sandra A, Peacock ML and Owen WG (1983) Insulin binding to microvascular endothelium of intact heart: A kinetic and morphometric analysis. Am J Physiol 244:E447E442. Dagenais C, Rousselle C, Pollack GM and Scherrmann JM (2000) Development of an in situ mouse brain perfusion model and its application to mdr1a P-glycoproteindeficient mice. J Cereb Blood Flow Metab 20:381386. Derossi D, Chassaing G and Prochiantz A (1998) Trojan peptides: The penetratin system for intracellular delivery. Trends Cell Biol 8:84 87. Duffy KR and Pardridge WM (1987) Blood-brain barrier transcytosis of insulin in developing rabbits. Brain Res 420:3238. Fisher GA and Sikic BI (1995) Clinical studies with modulators of multidrug resistance. Hematol Oncol Clin North Am 9:363382. Fishman JB, Rubin JB, Handrahan JV, Connor JR and Fine RE (1987) Receptormediated transcytosis of transferrin across the blood-brain barrier. J Neurosci Res 18:299 304. Frank HJ, Pardridge WM, Morris WL, Rosenfeld RG and Choi TB (1986) Binding and internalization of insulin and insulin-like growth factors by isolated brain microvessels. Diabetes 35:654 661. Gulyaev AE, Gelparina SE, Skidan IN, Antropov AS, Kivman GY and Kreuter J (1999) Significant transport of doxorubicin into the brain with polysorbate 80coated nanoparticles. Pharm Res 16:1564 1569. Harwig SS, Swiderek KM, Lee TD and Lehrer RI (1995) Determination of disulphide bridges in PG-2, an antimicrobial peptide from porcine leukocytes. J Pept Sci 1:207215. Huwyler J, Wu D and Padridge WM (1996) Brain drug delivery of small molecules using immunoliposomes. Proc Natl Acad Sci USA 93:14164 14169. Kroll RA, Pagel MA, Muldoon LL, Roman-Goldstein S, Fiamengo SA and Neuwelt EA (1998) Improving drug delivery to intracerebral tumor and surrounding brain in a rodent model: A comparison of osmotic versus bradykinin modification of the blood-brain and/or blood-tumor barriers. Neurosurgery 43:879 889. Kumagai AK, Eisenberg JB and Pardridge WM (1987) Adsorptive-mediated endocytosis of cationized albumin and a beta-endorphin-cationized albumin chimeric peptide by isolated brain capillaries. Model system of blood-brain barrier transport. J Biol Chem 262:15214 15219. Kusunoki N, Takara K, Tanigawara Y, Yamauchi A, Ueda K, Komada F, Ku Y, Kuroda Y, Saitoh Y and Okumura K (1998) Inhibitory effects of a cyclosporin derivative, SDZ PSC 833, on transport of doxorubicin and vinblastine via human P-glycoprotein. Jpn J Cancer Res 89:1220 1228. Mangoni ME, Aumelas A, Charnet P, Roumestand C, Chiche L, Despaux E, Grassy G, Calas B and Chavanieu A (1996) Change in membrane permeability induced by protegrin 1: Implication of disulphide bridges for pore formation. FEBS Lett 383:9398. Marbeuf-Gueye C, Broxterman HJ, Dubru F, Priebe W and Garnier-Suillerot A (1997) Kinetics of anthracycline efflux from multidrug resistance proteinexpressing cancer cells compared with P-glycoprotein-expressing cancer cells. Mol Pharmacol 53:141147. Nagy Z, Peters H and Huttner I (1983) Charge-related alterations of the cerebral endothelium. Lab Invest 49:662 671. Ohnishi T, Tamai I, Sakanaka K, Sakata A, Yamashima T, Yamashita J and Tsuji A (1995) In vivo and in vitro evidence for ATP-dependency of P-glycoproteinmediated efflux of doxorubicin at the blood-brain barrier. Biochem Pharmacol 49:15411544. Olivier JC, Fenart L, Chauvet R, Pariat C, Cecchelli R and Couet W (1999) Indirect evidence that drug brain targeting using polysorbate 80-coated polybutylcyanoacrylate nanoparticles is related to toxicity. Pharm Res 16:1836 1842. Pardridge WM (1999) Non-invasive drug delivery to the human brain using endogenous blood-brain barrier transport systems. Pharm Sci Technol Today 2:49 59. Pardridge WM, Eisenberg J and Yang J (1987) Human blood-brain barrier transferrin receptor. Metabolism 36:892 895. Pardridge WM, Kang YS and Buciak JL (1994) Transport of human recombinant brain-derived neurotrophic factor (BDNF) through the rat blood-brain barrier in vivo using vector-mediated peptide drug delivery. Pharm Res 11:738 746. Pardridge WM, Triguero D and Buciak J (1989) Transport of histone through the blood-brain barrier. J Pharmacol Exp Ther 251:821 826. Rousselle C, Clair P, Lefauconnier JM, Kaczorek M, Scherrmann JM and Temsamani J (2000) New advances in the transport of doxorubicin through the bloodbrain barrier by a peptide-vector mediated strategy. Mol Pharmacol 57:679 686. Schwarze SR, Ho A, Vocero-Akbani A and Dowdy SF (1999) In vivo protein transduction: Delivery of a biologically active protein into the mouse. Science (Wash DC) 285:1569 1572.

Downloaded from jpet.aspetjournals.org at Universitaet Zuerich on January 31, 2012

Transport Mechanism of Coupled Doxorubicin through the BBB

Singh M (1999) Transferrin as a targeting ligand for liposomes and anticancer drugs. Curr Pharm Des 5:443 451. Smith QR (1996) Brain perfusion systems for studies of drug uptake and metabolism in the central nervous system. Pharm Biotechnol 8:285307. Smith KR and Borchardt RT (1989) Permeability and mechanism of albumin, cationized albumin, and glycosylated albumin transcellular transport across monolayers of cultured bovine brain capillary endothelial cells. Pharm Res 6:466 473. Smith KR, Kato A and Borchardt RT (1988) Characterisation of specific receptors for atrial natriuretic factor on cultured bovine brain capillary endothelial cells. Biochem Biophys Res Commun 157:308 314. Tamai I, Sai Y, Kobayashi H, Kamata M, Wakamiya T and Tsuji A (1997) Structureinternalization relationship for adsorptive-mediated endocytosis of basic peptides at the blood-brain barrier. J Pharmacol Exp Ther 280:410 415. Temsamani J, Scherrmann JM, Rees AR and Kaczorek M (2000) Brain drug delivery technologies: Novel approaches for transporting therapeutics. Pharm Sci Technol Today 3:155162. Terasaki T, Hirai K, Sato H, Kang YS and Tsuji A (1989) Absorptive-mediated endocytosis of a dynorphin-like analgesic peptide, E-2078, into the blood-brain barrier. J Pharmacol Exp Ther 251:351357. Terasaki T, Takakuwa S, Saheki A, Moritani S, Shimura T, Tabata S and Tsuji A (1992) Adsorptive-mediated endocytosis of an adrenocorticotropic hormone

131

(ACTH) analogue, ebiratide, into the blood-brain barrier: Studies with monolayers of primary cultured bovine brain capillary endothelial cells. Pharm Res 9:529 534. Terasaki T and Tsuji A (1994) Drug delivery to the brain utilizing blood-brain barrier transport systems. J Control Release 29:163169. Triguero D, Buciak J and Pardridge WM (1990) Capillary depletion method for quantification of blood brain barrier transport of circulating peptides and plasma proteins. J Neurochem 54:18821888. Van Asperen J, van Tellingen O, Tijssen F, Schinkel AH and Beijnen JH (1999) Increased accumulation of doxorubicin and doxorubicinol in cardiac tissue of mice lacking mdr1a P-glycoprotein. Br J Cancer 79:108 113. Vorbrodt AW (1989) Ultracytochemical characterization of anionic sites in the wall of brain capillaries. J Neurocytol 18:359 368. Westergren I and Johansson BB (1993) Altering the blood-brain barrier in the rat by intracarotid infusion of polycations: A comparison between protamine, poly-Llysine and poly-L-arginine. Acta Physiol Scand 149:99 104.

Send reprint requests to: Jamal Temsamani, Synt:em, Parc Scientifique Georges Besse, 30000 Nmes, France. E-mail: jtemsamani@syntem.com

Downloaded from jpet.aspetjournals.org at Universitaet Zuerich on January 31, 2012