Application to a Cartilage Targeting Strategy: Synthesis and in Vivo

As part of a cartilage targeting program based on the affinity of the quaternary ammonium (QA) moiety for cartilage, QA derivatives of d-glucosamine (...
2 downloads 0 Views 108KB Size
Download PDF
212

Bioconjugate Chem. 2000, 11, 212−218

Application to a Cartilage Targeting Strategy: Synthesis and in Vivo Biodistribution of 14C-Labeled Quaternary Ammonium-Glucosamine Conjugates Isabelle Giraud, Maryse Rapp, Jean-Claude Maurizis, and Jean-Claude Madelmont* INSERM Unite´ 484, Rue Montalembert, BP 184, 63005 Clermont-Ferrand Cedex, France. Received October 1, 1999; Revised Manuscript Received November 29, 1999

As part of a cartilage targeting program based on the affinity of the quaternary ammonium (QA) moiety for cartilage, QA derivatives of D-glucosamine (DG), an antirheumatic drug exhibiting a natural tropism for cartilaginous tissues, were designed and evaluated by pharmacokinetic studies. Two QADG conjugates were synthesized and labeled with 14C by cross-linking the QA entity (trimethylammonium or pyridinium) to [14C]DG via an amide bond in a two-step procedure. After intravenous injection to male Sprague-Dawley rats, the two 14C-labeled conjugates exhibited similar pharmacokinetic profiles, but their behavior clearly differed from that of unconjugated DG in several ways. (i) The tissue distribution for the conjugates was more restricted, with a decreased radioactivity level for whole tissues except for kidney, cartilage, and skin. (ii) The radioactivity concentrated more rapidly and strongly in cartilage for the conjugates than for DG for the short times after injection; on the other hand, 1 h after administration, the radioactivity level in cartilage was higher for DG, this result being consistent with the tropism already observed for this compound. (iii) Both conjugates were eliminated predominantly by the urinary route (85%); the radioactivity level in urine for DG was lower (45% of the injected dose), and significant 14CO2 was found in expired air, indicating metabolization and utilization of DG for energy-consuming processes. (iv) Blood and plasma kinetics studies displayed an enterohepatic cycle for DG, whereas for the QA conjugates, a rapid disappearance was observed. (v) HPLC analyses of plasma and urine indicated a low degree of metabolization for the conjugates, most of the radioactivity recovered in urine and plasma corresponding to the unchanged molecule. This study demonstrates that the introduction of the QA moiety on DG modifies its biodistribution and lends it a greater specificity for cartilage, at least for short times after injection. These findings justify further work on QA derivatives of other antirheumatic agents.

INTRODUCTION

Current antirheumatic drugs including nonsteroidal antiinflammatory drugs (NSAIDs), corticosteroids, and disease-modifying antirheumatic drugs (DMARDs) often have severe adverse effects: gastrointestinal damage for NSAIDs, osteoporosis, and muscle weakness for corticosteroids and gastrointestinal distress, hepatic lesions, and lung injury for DMARDs (1-3). Safer agents are therefore being actively sought. An alternative way to reduce the adverse reactions is to carry the drugs selectively to their target tissue. On this basis, we are currently conducting research on the targeting of cartilage. This project has been undertaken following previous pharmacokinetic work on 14C-labeled acetylcholinesterase reactivators containing a quaternary ammonium (QA) moiety. This study clearly demonstrated a rapid and intensive concentration of these compounds in cartilaginous tissues after intramuscular injection (4, 5). We next demonstrated on cultured chondrocytes that the high specificity for cartilage resulted from ionic interactions between the QA entity of the reactivators and the anionic sites (carboxylate and ester sulfate groups) of proteoglycans, a major component of cartilage (6). This result was consistent with earlier findings concerning the binding of cations to cartilage (7-12). We * To whom correspondence should be addressed. Phone: 33 4 73 15 08 00. Fax: 33 4 73 15 08 01. E-mail: madelmont@ inserm484.u-clermont1.fr.

therefore hypothesized that the conjugation of the QA moiety with antirheumatic agents might strengthen their selectivity for cartilage, consequently minimizing their efficient doses and so reducing their side effects. To test this hypothesis, we developed a program to study several QA-antirheumatic drug conjugates (13). D-Glucosamine (DG) is one of the therapeutic agents we selected. DG has recently emerged as an alternative treatment for patients with osteoarthritis (14-17). It is an aminomonosaccharide occurring naturally in almost all human tissues, largely in proteoglycans of articular cartilage, as it is an intermediate substrate used in the synthesis of these macromolecules (18). The normal source of DG is endogenous biosynthesis from glucose, but exogenous DG, if available, becomes the preferential source for the proteoglycan biosynthesis (18). DG is consequently readily incorporated into proteoglycan molecules and shows a special tropism for cartilaginous tissues (18-20). Beneficial pharmacologic properties reported for DG include antiarthritic and antiinflammatory activities and an ability to stimulate proteoglycan synthesis through chondrocyte activation (21-24). DG was of particular interest for us, not owing to its toxicity, DG having a favorable safety profile, but because of its naturally high tropism for cartilage. Enhanced cartilage specificity of the QA-DG derivatives compared with the parent drug would support our cartilage targeting strategy. Here, we describe the preparation of two 14C-labeled QA-DG conjugates and the comparative pharmacokinetic studies in rats of [14C]DG and both 14C-conjugates.

10.1021/bc990128+ CCC: $19.00 © 2000 American Chemical Society Published on Web 02/17/2000

Biodistribution of Quaternary Ammonium−Glucosamine Conjugates EXPERIMENTAL PROCEDURES

Syntheses. General. All chemicals were from commercial suppliers and used as received. The radiolabeled starting material, D-[1-14C]glucosamine hydrochloride ([14C]DG hydrochloride), was purchased from NEN Life Science Products (Boston, MA). Proton nuclear magnetic resonance (1H NMR) and carbon nuclear magnetic resonance (13C NMR) spectra were recorded at, respectively, 200.1 and 50.3 MHz on a Bruker AM 200 (4.5 T) spectrometer. Chemical shifts are reported in parts per million relative to the internal tetramethylsilane standard for 1H NMR and the solvent for 13C NMR (DMSOd6, δ ) 39.5 ppm). Coupling constants (J values) are given in hertz. Infrared (IR) spectra were recorded on a Bruker Vector 22 FTIR system. Electrospray ionization mass spectra (ESI-MS) and elemental analyses were performed by the CNRS Service Central d’Analyse (Vernaison, France). Elemental analyses for carbon, hydrogen, and nitrogen were within (0.4% of theory for the formulas given. Melting points (mp) were determined on an Electrothermal digital apparatus and are uncorrected. Thin-layer chromatography (TLC) was conducted on precoated silica gel plates (Merck, 60 F254, 0.25 mm thick). Radiochemical purity was evaluated by scanning the TLC plates with an Ambis 4000 detector (B. Braun Sciencetec) and by high-performance liquid chromatography (HPLC) studies for final compounds. HPLC analyses were performed on an HP 100 system (Hewlet Packard, D76337 Waldbronn, Germany) equipped with a Partisil SCX analytical column (10 µm, 250 × 4.6 mm; Whatman, Clifton) and coupled with a 500TR flow scintillation analyzer (Packard Instrument S. A., Rungis, France). A solution of 0.2 M KH2PO4 acidified to pH 2.8 with H3PO4 was used as the mobile phase, with a flow rate of 1 mL/min. Specific activity of synthesized compounds was measured in a Wallac Winspectral 1414 liquid scintillation spectrometer (EGG Instruments, Evry, France), using Packard Ultima Gold liquid scintillation cocktail. N-Chloroacetyl-R,β-D-[1-14C]glucosamine. To a solution of [14C]DG hydrochloride [1.95 g, 9.02 mmol, 1502 MBq (40.6 mCi), specific activity: 167 MBq/mmol (4.5 mCi/ mmol)] and potassium carbonate (1.87 g, 13.53 mmol) in water (10 mL) was added chloroethanoyl choride (1.43 mL, 18 mmol) with stirring at 0 °C. After 1 h, the solution was evaporated under reduced pressure. The residue was washed with ethanol (3 × 50 mL). The ethanol solutions were then concentrated under reduced pressure until a white precipitate appeared. Filtration and recrystallization in ethanol afforded the N-chloroacetyl intermediate [1.31 g, 855 MBq (23.1 mCi), 57%] as a white solid. Specific activity: 167 MBq/mmol (4.5 mCi/mmol). Radiochemical purity was 97% as determined by TLC analysis (Rf ) 0.54, dichloromethane/methanol/ammonium hydroxide 50/48/2). Analytical data for the unlabeled compound: mp 183-185 °C (dec); 1H NMR (DMSO-d6) δ 3.06-3.18 (m, 1H, H-4), 3.43-3.62 (m, 5H, H-2, H-3, H-5, H-6), 4.05 [s, 0.1H, CH2(β)Cl], 4.11 [s, 1.9H, CH2(R)Cl], 4.42-4.48 (m, 1.05H, H-1β, OH-6; after exchange with D2O, d, J ) 7.6, 0.05H, H-1β), 4.78 (d, J ) 5.1, 0.95H, OH-3R), 4.92-4.98 (m, 2H, H-1R, OH-3β, OH-4; after exchange with D2O, d, J ) 3.1, 0.95H, H-1R), 6.57 (d, J ) 4.4, 1H, OH-1), 7.98 (d, J ) 7.7, 1H, NH); 13C NMR (DMSO-d6) δ 42.77 [CH2(R)Cl], 43.10 [CH2(β)Cl], 54.77 (C-2R), 57.54 (C-2β), 61.07 (C-6Rβ), 70.47, 70.98 (C-3R C-4R), 70.79 (C-4β), 72.16 (C-5R), 74.03 (C-3β), 76.80 (C5β), 90.37 (C-1R), 95.03 (C-1β), 165.79 (CdOβ), 165.92 (CdOR); ESI-MS m/z 257 and 259 [M(35Cl,37Cl) + H]+; IR (KBr) ν (cm-1) 1645 (CdO).

Bioconjugate Chem., Vol. 11, No. 2, 2000 213

N-(Trimethylammonio)acetyl-R,β-D-[1-14C]glucosamine Chloride ([14C]TMG). A solution of the N-chloroacetyl derivative (639 mg, 2.35 mmol) and trimethylamine (5 mL of a 4.2 M solution in ethanol) under argon was warmed to 40 °C. A white precipitate appeared in a few hours. The reaction was complete after 3 days. The solution was then evaporated under reduced pressure. The resulting solid was washed with ethanol and then ether and dried to give [14C]TMG [651 mg, 340 MBq (9.2 mCi), 87%] as a white solid. Specific activity: 167 MBq/ mmol (4.5 mCi/mmol). Radiochemical purity was 96% as determined by HPLC analysis (retention time ) 5.3 min). Analytical data for the unlabeled compound: mp 240242 °C (dec); 1H NMR (DMSO-d6) δ 3.05-3.75 [m, 15H, H-2, H-3, H-4, H-5, H-6, N+(CH3)3], 4.06 [s, 0.3H, CH2(β)N+], 4.09 [s, 1.7H CH2(R)N+], 4.46-4.61 (m, 1.15H, H-1β, OH-6; after exchange with D2O, d, J ) 7.8, 0.15H, H-1β), 4.94-5.07 (m, 2.85H, H-1R, OH-3, OH-4; after exchange with D2O, d, J ) 3.1, 0.85H, H-1R), 6.66 (d, J ) 4.0, 0.85H, OH-1R), 6.77 (d, J ) 6.0, 0.15H, OH-1β), 8.51 (d, J ) 8.0, 0.85H, NHR), 8.71 (d, J ) 8.4 Hz, 0.15H, NHβ); 13C NMR (DMSO-d6) δ 53.35 [N+(CH3)3], 54.56 (C2R), 57.05 (C-2β), 60.90 (C-6Rβ), 64.47 [CH2(R)N+], 64.88 [CH2(β)N+], 70.37, 70.80 (C-3R, C-4R), 70.63 (C-4β), 72.14 (C-5R), 73.94 (C-3β), 76.89 (C-5β), 90.14 (C-1R), 94.90 (C1β), 162.80 (CdOβ), 163.04 (CdOR); IR (KBr) ν (cm-1) 1660 (CdO); ESI-MS m/z 279 (M)+; Anal. (C11H23ClN2O6‚ 0.2H2O) C, H, N. N-Pyridinioacetyl-R,β-D-[1-14C]glucosamine Chloride ([14C]PG). A solution of the N-chloroacetyl derivative (600 mg, 2.35 mmol) in pyridine (30 mL) under argon was warmed to 40 °C. A white precipitate appeared in a few hours. The reaction was complete after 3 days. Pyridine was then evaporated under reduced pressure. The resulting solid was washed with ethanol and then ether and dried to give [14C]PG [583 mg, 285 MBq (7.7 mCi), 73%] as a white solid. Specific activity: 167 MBq/mmol (4.5 mCi/mmol). Radiochemical purity was 96% as determined by HPLC analysis (retention time ) 7.2 min). Analytical data for the unlabeled compound: mp 223-225 °C (dec); 1 H NMR (DMSO-d6) δ 3.08-3.72 (m, 6H, H-2, H-3, H-4, H-5, H-6), 4.47-4.59 (m, 1.95H, H-1β, OH-6; after exchange with D2O, d, J ) 7.6, 0.95H, H-1β), 5.04-5.09 (m, 2.05H, H-1R, OH-3, OH-4; after exchange with D2O, d, J ) 2.9, 0.05H, H-1R), 5.45 (s, 2H, CH2N+), 6.71 (d, J ) 6.0, 1H, OH-1), 8.18 (t, J ) 7.6, 2H, aromatic protons), 8.61-8.69 (m, 2H, aromatic protons, NH; after exchange with D2O, t, J ) 7.7, 1H, aromatic proton), 8.97 (d, J ) 6.0, 2H, aromatic protons); 13C NMR (DMSO-d6) δ 55.21 (C-2R), 58.02 (C-2β), 61.01 (C-6Rβ), 61.78 (CH2N+), 70.31 (C-4β), 70.44 70.65 (C-3R C-4R), 72.12 (C-5R), 74.05 (C-3β), 76.86 (C-5β), 90.34 (C-1R), 94.83 (C-1β), 127.45, 146.00, 146.13 (aromatic carbons), 164.11 (CdOβ), 164.33 (CdOR); ESI-MS m/z 299 (M)+; IR (KBr) ν (cm-1) 1675 (CdO); Anal. (C13H19ClN2O6‚0.3H2O) C, H, N. Pharmacokinetic Studies. Tissue Distribution Study. Tissue distribution analyses were performed in 4-weekold male Sprague-Dawley rats weighing 80-100 g (IffaCredo, L’Arbresle, France) by injection via the tail vein of each 14C-labeled test compound [75 µmol g/kg, 1110 kBq (30 µCi) in 200 µL of physiological saline]. Animals were sacrified by ether inhalation 5, 15, and 30 min and 1, 2, 6, and 24 h after injection. Carcasses rapidly frozen by immersion in liquid nitrogen were then embedded in a 2% gel of carboxymethyl cellulose. The resulting carboxyl methylcellulose block was sagittally sectioned at -22 °C with a Reichert-Jung Cryopolycut cryomicrotome (Heidelberg, Germany). When section surfaces of interest appeared, the corresponding 40-µm-thick slices

214 Bioconjugate Chem., Vol. 11, No. 2, 2000

Giraud et al.

Scheme 1a

a

(i) ClCOCH2Cl, K2CO3, H2O, 0 °C; (ii) trimethylamine (4.2 M solution in ethanol), 40 °C; (iii) pyridine, 40 °C.

were taken using no. 810 Scotch band tape (3 M, Saint Paul, MN) and dried for 48 h at -22 °C. These selected slices (n ) 8 for each time studied) were subsequently analyzed with the Ambis 4000 detector (described in the chemical part), which allows visualization and quantification of the radioactivity distribution in whole-body sections (25). The results of these tissue distribution analyses were expressed as the percentage of the injected dose of radioactivity per gram of tissue (25). Blood and Plasma Kinetics Study. Blood and plasma kinetics experiments were performed as decribed above for tissue distribution studies. Rats were exsanguinated by a cardiac puncture (n ) 3 for each time studied). Plasma and red cells were separated by centrifugation (Rotanta RP, Hettish, Tuttlingen, Germany). The radioactivity in blood and plasma samples was counted in a Wallac Winspectral 1414 liquid scintillation spectrometer. The results were expressed as the percentage of the injected dose of radioactivity per gram of blood or plasma. Excretion Study. The excretion study was conducted in 6-week-old male Sprague-Dawley rats weighing 180200 g by intravenous injection of each 14C-labeled test compound [75 µmol g/kg, 555 kBq (15 µCi) in 200 µL of physiological saline]. Urine and feces were collected by housing animals (n ) 5 for each test compound) in a metabolic cage (Iffa-Credo, L’Arbresle, France) for 24, 48, and 72 h after administration. For the collection of expired [14C]CO2, animals (n ) 1 for each test compound) were housed in an apparatus designed by us to trap [14C]CO2 with a flask containing a solution of ethanolamine/ methanol (20:80 v/v). Radioactivity of urine and ethanolamine/methanol samples was directly measured after addition of Packard Ultima Gold scintillation cocktail in the Wallac Winspectral 1414 liquid scintillation spectrometer. Radioactivity of feces was counted after combustion in a Packard 306 Oxidizer (Packard Instrument SA, Rungis, France). The results are given as percentages of the injected dose. Metabolism Study. Plasma samples were deproteinized by treatment with methanol (1:7 v/v). After stirring and centrifugation, the supernatant was evaporated under reduced pressure. The resulting dry residue was dissolved in the HPLC eluent and analyzed by HPLC (conditions described in the chemical part). The urine samples were also analyzed by HPLC after filtration through Dynagard syringe filters (Spectrum Microgon, Laguna Hills, CA). Determination of Pharmacokinetic Parameters. The elimination half-life (t1/2) was calculated on an Apple Macintosh computer with the pharmacokinetic data

treatment software SIPHAR (Simed, Cre´teil, France), using the formula t1/2 ) ln 2/K, where K is the slope of the elimination phase. RESULTS 14C-Synthesis of the QA-DG Conjugates. The [14C]QA-DG conjugates were prepared according to the sequence described in Scheme 1. The linkage between DG and the QA carrier was achieved via an amide bond by a two-step procedure. First, condensation of [14C]DG hydrochloride [specific activity: 167 MBq/mmol (4.5 mCi/ mmol)] with chloroethanoyl chloride in the presence of potassium bicarbonate in aqueous medium at 0 °C provided the N-chloroacetyl 14C-derivative in 57% yield. Subsequent reaction of this intermediate with pyridine (reactant and solvant) or with trimethylamine (4.2 M solution in ethanol) at 40 °C for 3 days gave, respectively, the trimethylammonium ([14C]TMG) and the pyridinium ([14C]PG) conjugates with overall yields of 50% and 42%. The two final quaternary ammonium compounds were found by NMR and TLC to be identical to the unlabeled reference compounds. They were obtained with a specific activity of 167 MBq/mmol (4.5 mCi/mmol) and a radiochemical purity of 96% as determined by HPLC analysis. Their physical data and spectral characterizations are presented in the Experimental Procedures. Pharmacokinetic Studies. Tissue Distribution. Results of the in vivo biodistribution experiments with the two [14C]QA-DG conjugates and the parent molecule in male Sprague-Dawley rats are summarized in Table 1. These data were determined by the method of whole-body autoradiography coupled with Ambis 4000 counting (25). Representative two-dimensional images of whole-body sections obtained by this method are shown in Figure 1. Following intravenous administration of [14C]DG, a fast distribution of the radioactivity was observed in a great number of tissues and specifically in kidney, brain, liver, skin, and cartilaginous tissues (Table 1). In the first 2 h, the radioactivity was highly concentrated in liver and kidney. For the liver, a specific kinetic law was observed, with a first increase in the radioactivity level until 15 min after injection, followed by a slow decrease. The radioactivity distribution pattern obtained with [14C]TMG was similar to that found with [14C]PG. At the short times, a marked radioactivity retention was detected in kidney. Until 30 min after administration, the highest levels of radioactivity were observed in cartilaginous tissues, skin, and kidney. In contrast, low concen-

Biodistribution of Quaternary Ammonium−Glucosamine Conjugates

Bioconjugate Chem., Vol. 11, No. 2, 2000 215

Table 1. Comparative Biodistribution of [14C]DG, [14C]TMG, and [14C]PG in Male Sprague-Dawley Rats Following Intravenous Injectiona tissue/organ liver kidney brain bone marrow cartilage skin

DG TMG PG DG TMG PG DG TMG PG DG TMG PG DG TMG PG DG TMG PG

5 min

15 min

30 min

1h

2h

6h

24 h

3.36 ( 0.56 1.23 ( 0.03 0.76 ( 0.15 4.76 ( 0.11 8.06 ( 0.76 9.57 ( 6.55 0.58 ( 0.16 0.07 ( 0.04 0.07 ( 0.11 0.96 ( 0.01 0.83 ( 0.04 0.64 ( 0.02 1.77 ( 0.14 3.28 ( 0.31 2.51 ( 0.20 1.47 ( 0.44 2.37 ( 0.25 2.75 ( 0.29

5.46 ( 0.85 0.83 ( 0.04 0.88 ( 0.19 4.44 ( 0.16 9.54 ( 1.41 5.40 ( 2.66 1.08 ( 0.11 0.07 ( 0.07 0.15 ( 0.08 0.29 ( 0.01 0.59 ( 0.02 0.64 ( 0.03 1.56 ( 0.14 2.68 ( 0.27 2.30 ( 0.40 1.17 ( 0.16 2.00 ( 0.28 1.97 ( 0.52

4.94 ( 1.27 0.38 ( 0.04 0.49 ( 0.09 2.51 ( 0.38 2.83 ( 1.56 2.52 ( 0.71 0.95 ( 0.09 0.04 ( 0.01 0.10 ( 0.08 0.33 ( 0.08 0.21 ( 0.01 0.19 ( 0.09 0.95 ( 0.20 1.08 ( 0.16 1.49 ( 0.09 0.49 ( 0.11 0.65 ( 0.16 0.87 ( 0.14

4.91 ( 0.33 0.19 ( 0.04 0.30 ( 0.10 1.99 ( 0.26 1.24 ( 0.30 1.43 ( 0.45 0.92 ( 0.18 0.02 ( 0.01 0.07 ( 0.06 0.36 ( 0.04 b b 0.88 ( 0.25 0.37 ( 0.08 0.53 ( 0.05 0.32 ( 0.04 0.19 ( 0.06 0.26 ( 0.03

2.59 ( 0.35 0.12 ( 0.02 0.15 ( 0.02 1.79 ( 0.06 0.66 ( 0.12 0.70 ( 0.10 0.77 ( 0.11 0.02 ( 0.01 0.03 ( 0.01 b b b 0.88 ( 0.20 b b 0.45 ( 0.17 0.12 ( 0.06 0.20 ( 0.06

1.89 ( 0.20 0.15 ( 0.02 0.14 ( 0.02 0.98 ( 0.08 0.27 ( 0.11 b 0.66 ( 0.07 0.01 ( 0.01 b b b b 1.17 ( 0.17 b b 0.48 ( 0.03 0.12 ( 0.08 0.19 ( 0.07

0.74 ( 0.01 b b b b b 0.44 ( 0.03 b b b b b 0.76 ( 0.30 b b 0.51 ( 0.13 b b

a Results are expressed as the percentage of the injected dose per gram of tissue and correspond to the mean ( standard deviation for the quantification of eight whole-body sections of one animal for each point. b Could not be quantified.

Figure 1. Representative two-dimensional images obtained with Ambis 4000 detector of rat whole-body sections showing the tissue distribution of radioactivity 15 min following intravenous administration of each 14C-labeled test compound: [14C]DG, [14C]TMG, and [14C]PG. The radioactivity distribution pattern obtained with both congugates was more restricted than that found with DG, with a decreased radioactivity distribution for whole tissues except for kidney, cartilage and skin.

trations were found in brain and liver. The radioactivity levels quantified in cartilaginous tissues for both conjugates were higher than those detected with [14C]DG until 30 min after administration. From 1 h after injection, the cartilaginous radioactivity concentration became greater for [14C]DG, and the elimination rate was slower than that observed for TMG and PG. Blood and Plasma Kinetics. The radioactivity levels found in blood and plasma of animals after intravenous injection of [14C]DG, [14C]TMG, and [14C]PG are presented in Figure 2. For each test compound, the plasma kinetics

followed the blood kinetics. As observed for tissue distribution analyses, DG exhibited a different plasma behavior from its two conjugates. For DG, a fast decrease in the radioactivity was first detected, reaching a minimal value 15 min after injection. An increase until 2 h after administration followed by a slow disappearance phase were then observed. In contrast, for the QA-DG derivatives, a rapid plasma clearance was detected. Radioactivity Excretion. Table 2 presents the radioactivity excreted in urine, feces and in expired air (14CO2) after injection to rats of [14C]DG, [14C]TMG, and [14C]PG.

216 Bioconjugate Chem., Vol. 11, No. 2, 2000

Giraud et al.

Figure 2. A comparison of the blood (A) and plasma (B) kinetics of [14C]DG, [14C]TMG and [14C]PG in male Sprague-Dawley rats following intravenous administration. Results are expressed as the percentage of the injected dose per gram of blood or plasma and correspond to the mean ( standard deviation for three animals for each point. Table 2. Radioactivity Excreted in Urine Feces and Expired Air after Intravenous Administration of [14C]DG, [14C]TMG, and [14C]PG in Male Sprague-Dawley Ratsa DG TMG PG

time (h)

urine

feces

14CO

0-24 24-78 48-72 0-24 24-78 48-72 0-24 24-78 48-72

41.7 ( 1.2 1.3 ( 0.1 0.8 ( 0.2 85.3 ( 3.4 1.5 ( 0.5 0.3 ( 0.1 84.6 ( 4.9 1.4 ( 0.9 0.4 ( 0.1

4.5 ( 0.4 3.1 ( 0.2 0.8 ( 0.1 7.8 ( 0.7 2.7 ( 0.3 0.9 ( 0.1 7.1 ( 0.4 1.6 ( 0.1 0.4 ( 0.1

18.9 4.5 1.5 4.2 0.6 0.2 6.3 0.9 0.6

2

a Results are expressed as the percentage of the injected dose and correspond to the mean ( standard deviation for five animals for each point, except for 14CO2 excretion (one rat studied).

Whereas for the two conjugates more than 85% of the radioactivity was excreted by the urine route, only 44% of the injected dose was recovered in the urine for DG, and a significant level of radioactivity was detected in the expired air (25% of the injected dose was found in the expired CO2 72 h after injection). Metabolism Study. Plasma and urine samples were analyzed by HPLC to determine the percentage of unchanged molecule and to find potential metabolites. In our HPLC conditions, the retention times were 3.7, 5.3, and 7.2 min for [14C]DG, [14C]TMG, and [14C]PG, respectively. Chromatographic analyses of plasma samples were performed on deproteinized plasma. For [14C]DG, the radioactive concentration found in deproteinized plasma represented 93% of the radioactivity of whole plasma 5 min after injection. The percentage then decreased to 6% after 1 h. Simultaneously, the radioactivity incorporated into plasma proteins increased (from about 10% 5 min after administration to 91% after 1 h). The HPLC profiles of initial deproteinized plasmas showed that the radioactivity was mainly due to DG (more than 90%). For the [14C]QA-DG derivatives, the radioactivity measured in deproteinized plasmas represented more than 90% of the radioactivity of whole plasma, whatever the time studied. The HPLC analyses of this radioactivity indicated that 1 h after injection, for both conjugates, more than 90% of the radioactivity corresponded to the unchanged molecule. This percentage was still 83% after 2 h. From 6 h following administration, the plasma radioactivity

became too low to permit HPLC analysis. Thus, for the two QA-DG derivatives, the radioactivity found in plasma was identified as the unchanged molecule. This result enabled us to calculate the half-life time in the plasma of compounds, which was determined to be 15 min for TMG and 14 min for PG. HPLC study of urine samples for the study of [14C]DG indicated that, while only 42% of the injected dose was recovered in urine 24 h after injection, 87% of this radioactivity corresponded to the parent molecule. For both QA conjugates, more than 85% of the radioactivity detected in urine until 24 h after administration was identified as the unchanged compound. DISCUSSION

The objective of this work was to determine whether the DG-QA conjugates exhibited a higher tropism for cartilaginous tissues than DG. This involved conducting pharmacokinetic studies in animals of the QA derivatives, which consequently had to be radiolabeled. The preferred isotope for such experiments was 14C because of its long half-life time and convenient energy of decay. As [14C]DG was a commercially available compound, the radiolabeling of the conjugates was performed by functionalization of this 14C-labeled precursor. The QA moiety was attached on the amino group of the carbohydrate. We chose this position because (i) chemically, the amino group reacts more selectively and rapidly in condensation reactions than the hydroxyl groups, and (ii) biologically, DG is incorporated into proteoglycans with an N-acetylated form involving OH-1 and OH-3 or OH-4 groups. Accordingly, we thought a N-functionalization would reproduce this biological form. A trimethylammonium conjugate and a pyridinium derivative ([14C]TMG and [14C]PG) were synthesized to assess the possible influence of the QA structure on the biodistribution of these compounds. The conjugates were prepared from [14C]DG by a two-step procedure via N-chloroacetyl-DG as intermediate. Biodistribution patterns of the radioactivity in rats after intravenous injection of [14C]TMG and [14C]PG were compared with that of [14C]DG. Whatever the parameter studied, both conjugates were found to have similar behavior, while DG exhibited a completely different profile.

Biodistribution of Quaternary Ammonium−Glucosamine Conjugates

From the point of view of tissue distribution, a marked diminution of the affinity of the QA derivatives compared with DG was observed for whole tissues except for kidney, cartilage, and skin. The radioactivity accumulation observed in kidney reflected the major urinary excretion of these compounds. The radioactivity levels found in skin could be explained by the fact that skin is also a proteoglycan-containing tissue. Contrary to DG, low hepatic and cerebral radioactivity concentrations were detected, indicating that the conjugates do not cross the blood brain barrier and are not extensively metabolized. Consequently, this tissue distribution study of the conjugates suggested, through a more restricted distribution profile, a higher specificity for cartilage than DG. In the case of [14C]TMG or [14C]PG, cartilaginous tissues concentrated the radioactivity more rapidly and strongly than [14C]DG (until 30 min after administration). Such high radioactivity levels in cartilage for the short times following the injection were already observed in previous results obtained by us with other QA derivatives (4, 5, 10). From 1 h after injection, the radioactivity level became higher for DG. This last result is consistent with the literature: exogenous DG has a special tropism for cartilage as it constitutes the preferential substrate for the proteoglycan biosynthesis (18-20). The pattern of the radioactivity excretion after administration of the conjugates revealed a predominant urinary excretion. Moreover, these compounds were found to be eliminated unchanged with little metabolic degradation. For DG, the excretion profile was quite different: the radioactivity level detected in urine was lower and a significant amount was found in expired air, suggesting a substantial part of the administered DG was completely broken down and utilized for energy-consuming processes (19). The investigation of blood and plasma kinetics indicated a faster disappearance for the QA derivatives than for the parent drug. For DG, the plasma radioactivity decreased rapidly during the first 15 min. During this phase, DG was taken up in large amounts in the liver, where it was metabolized and incorporated into plasma proteins (17-19, 26, 27). After 15 min, the radioactivity increased in plasma, reaching a peak at 2 h. In this phase, the radioactivity originated from plasma proteins, in which the exogenous DG or fragments were incorporated (17-19, 26). Thus, this study provides evidence for the existence of an enterohepatic cycle for DG, which was also confirmed by the radioactivity pattern obtained for liver in the tissue distribution experiment. On the contrary, the conjugates exhibited a rapid clearance from the blood with a monoexponential elimination. Furthermore, they were found to be not hardly metabolized, most of the radioactivity detected in urine and plasma being identified as the unchanged molecule. To conclude, the work reported here demonstrates a striking difference in the pharmacokinetic profiles of DG and its two QA conjugates. On the other hand, the nature of the QA moiety does not seem to have a significant influence, as both conjugates exhibited similar behavior. Finally, in the context of our cartilage targeting strategy, two points must be underlined: the QA derivatives showed a more restricted tissue distribution than the parent DG and a higher cartilaginous concentration in the short times following intravenous administration. Therefore, this study shows that the introduction of a QA moiety on DG, a compound which already exhibits a special tropism for cartilage, allows the molecule to be carried more selectively and intensively to cartilaginous tissues soon after injection. These results clearly argue

Bioconjugate Chem., Vol. 11, No. 2, 2000 217

in favor of cartilage targeting. Following this encouraging work, we now intend to study the synthesis and evaluation of the QA derivatives of other antirheumatic agents. ACKNOWLEDGMENT

This work was supported by ADIR, Servier (France). LITERATURE CITED (1) Blackburn, W. D. (1996) Management of osteoarthritis and rheumatoid arthritis: prospects and possibilities. Am. J. Med. 100 (Suppl. 2A), 24S-30S. (2) Schiff, M. (1997) Emerging treatments for rheumatoid arthritis. Am. J. Med. 102 (Suppl. 1A), 11S-15S. (3) Brooks, P. M. (1998) Treatment of rheumatoid arthritis: from symptomatic relief to potential cure. Br. J. Rheumatol. 37, 1265-1271. (4) Garrigue, H., Maurizis, J. C., Nicolas, C., Madelmont, J. C., Godeneche, D., Hulot, T., Morge, X., Demerseman, P., Sentenac-Roumanou, H., and Veyre, A. (1990) Disposition and metabolism of two acetylcholinesterase reactivators Pyrimidoxime and HI6 in rats submitted to organophosphate poisoning. Xenobiotica 20, 699-709. (5) Garrigue, H., Maurizis, J. C., Madelmont, J. C., Nicolas, C., Meyniel, J. M., Louvel, A., Demerseman, P., SentenacRoumanou, H., and Veyre, A. (1991) Disposition and metabolism of acetylcholinesterase reactivators 2PAM-I, TMB4 and R665 in rats submitted to organophosphate poisoning. Xenobiotica 21, 583-595. (6) Maurizis, J. C., Ollier, M., Nicolas, C., Madelmont, J. C., Garrigue, H., and Veyre, A. (1992) In vitro binding of oxime acetylcholinesterase reactivators to proteoglycans synthesized by cultured chondrocytes and fibroblasts. Biochem. Pharmacol. 44, 1927-1933. (7) Asghar, K., and Roth, L. J. (1971) Distribution of hexamethonium and other quaternary ammonium compounds in cartilage. J. Pharmacol. Exp. Ther. 176, 83-92. (8) Olsen, G. D., Chan, E. M., and Riker, W. K. (1975) Binding of d-tubocurarine di(methyl-14C)ether iodide and other amines to cartilage, chondroitin sulfate and human plasma proteins. J. Pharmacol. Exp. Ther. 195, 242-250. (9) Van Ootegham, S. P., Smith, R. G., Daves, G. D., El-Matazi, A., Chan, E. M., Olsen, G. D., and Riker, W. K. (1976) Iodobis(quaternary ammonium) Salts. Potential Cartilage-Selective X-ray Contrast Agents. J. Med. Chem. 19, 1349-1352. (10) Larsson, B., Nilsson, M., and Tja¨lve, H. (1981) The binding of inorganic and organic cations and H+ to cartilage in vitro. Biochem. Pharmacol. 30, 2963-2970. (11) Yu, W. K. S., Bartlett, J. M., Van Sickle, D. C., and Mock, B. H. (1988) Biodistribution of bis-[β-(N,N,N-trimethylamino)ethyl]selenide-75Se diiodide, a potential articular imaging agent. Nucl. Med. Biol. 15, 229-230. (12) Yu, W. K. S., Shaw, S. M., Bartlett, J. M., Van Sickle, D. C., and Mock, B. H. (1989) The biodistribution of bis-[β(N,N,N-trimethylamino)ethyl]selenide-75Se diiodide in adult guinea pigs. Nucl. Med. Biol. 16, 255-259. (13) Madelmont, J.-C., Giraud, I., Nicolas, C., Maurizis, J.-C., Rapp, M., Ollier, M., Renard, P., and Caignard, D.-H. French Patent 99,08020, June 23, 1999, 29 pp. (14) MacCarthy, M. F. (1994) The neglect of Glucosamine as a treatment for osteoarthritis-a personal perspective. Med. Hypotheses 42, 323-327. (15) Barclay, T. S., Tsourounis, C., and McCart, G. M. (1998) Glucosamine. Ann. Pharmacother. 32, 574-579. (16) da Camara, C. C., and Dowless, G. V. (1998) Glucosamine sulfate for osteoarthritis. Ann. Pharmacother. 32, 580-587. (17) Kelly, G. S. (1998) The role of glucosamine sulfate and chondroitine sulfate in the treatment of degenerative joint disease. Altern. Med. Rev. 3 (1), 27-39. (18) Rode´n, L. (1956) Effect of hexosamines on the synthesis of chondroitin sulphuric acid in vitro. Ark. Kemi. 10, 345-352. (19) Setnikar, I., Giachetti, C., and Zanolo, G. (1984) Absorption, distribution and excretion of radioactivity after a single intravenous or oral administration of 14Cglucosamine in the rat. Pharmatherapeutica 3, 538-550.

218 Bioconjugate Chem., Vol. 11, No. 2, 2000 (20) Setnikar, I., Giacchetti, C. and Zanolo, G. (1986) Pharmacokinetics of glucosamine in the dog and in man. Arzneim. Forsch./Drug Res. 36 (I), 729-735. (21) Bassleer, C., Henrotin, Y., and Franchimont, P. (1992) In vitro evaluation of drugs proposed as chondroprotective agents. Int. J. Tissue React. 14, 231-241. (22) Setnikar, I. (1992) Antireactive properties of “chondroprotective” drugs. Int. J. Tissue React. 14, 253-261. (23) Karzel, K., and Domenjoz, R. (1971) Effects of hexosamine derivatives and uronic derivatives on glycosaminoglycan metabolism of fibroblast cultures. Pharmacology 5, 337. (24) Vidal, Y., Plana, R. R., Bizarri, D., and Rovati, A. L. (1978) Articular cartilage pharmacology. I. In vitro studies on glucosamine and non steroidal antiinflammatory drugs. Pharmacol. Res. Commun. 10, 557.

Giraud et al. (25) Labarre, P., Papon, J., Moreau, M.-F., Madelmont, J.-C., and Veyre, A. (1998) A new quantitative method to evaluate the biodistribution of a radiolabeled tracer for melanoma using whole-body cryosectioning and a gaseous detector: comparison with conventional tissue combustion technology. Eur. J. Nucl. Med. 25, 109-114. (26) Akamastsu, N., Nakajima, H., and Miyata, S. (1976) Glucosamine metabolism in regenerating rat liver. Biochem. J. 158, 589-592. (27) Lo, C.-H., and Marsh, J. B. (1970) Biosynthesis of plasma lipoproteins. J. Biol. Chem. 245, 5001-5006.

BC990128+