- Email: [email protected]
Significant differences in the disposition of cyclic prodrugs of opioid peptides in rats and guinea pigs following IV administration
Significant Differences in the Disposition of Cyclic Prodrugs of Opioid Peptides in Rats and Guinea Pigs Following IV Administration BIANCA M. LIEDERER,1 KIMTHOA T. PHAN,2 HUI OUYANG,1 RONALD T. BORCHARDT1 1
Department of Pharmaceutical Chemistry, The University of Kansas, Lawrence, Kansas 66047
2
Department of Chemistry, Creighton University, Omaha, Nebraska 68178
Received 17 March 2005; revised 14 July 2005; accepted 2 August 2005 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.20476
ABSTRACT: The stabilities of DADLE ([D-Ala2,D-Leu5]-Enk, H-Tyr-D-Ala-Gly-Phe-DLeu-OH), the capped derivative Ac-DADLE-NH2, and the oxymethyl-coumarinic acid (OMCA)-based cyclic prodrug of DADLE and [D-Ala2,Leu5]-Enk (H-Tyr-D-Ala-Gly-PheLeu-OH) were determined at 378C in rat and guinea pig liver microsomes in the presence and absence of paraoxon, an esterase B inhibitor, and ketoconazole, a CYP3A4 inhibitor. These studies showed that the order of stability in microsomes was: DADLE >> AcDADLE-NH2 > OMCA-DADLE ¼ OMCA-[D-Ala2,Leu5]-Enk. While paraoxon produced no significant effect on the stability of the studied compounds in liver microsomes, ketoconazole inhibited the metabolism, suggesting that the capped peptide and the cyclic prodrugs are substrates for cytochrome P450 enzymes. For pharmacokinetic studies, the cyclic prodrugs of DADLE and [D-Ala2,Leu5]-Enk were administered i.v. to rats and guinea pigs. Various biological fluids and tissue (brain, bile, and blood) were collected and analyzed for the free peptide and the prodrugs by high performance liquid chromatography with tandem mass spectrometric detection (LC-MS-MS). These studies showed that the conversion of the cyclic prodrugs to the respective linear peptides (i.e., DADLE and [D-Ala2,Leu5]-Enk) was rapid in rat and guinea pig. In terms of drug elimination, only trace amounts of OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk were recovered in guinea pig bile (3.3% and 0.82%, respectively), while significant amounts were recovered in rat bile (38.1% and 51.7%, respectively). Brain uptake of the cyclic prodrugs in guinea pigs compared to previously determined brain uptake of OMCA-DADLE in rats was also significantly different. For OMCA-DADLE, the brain levels of the cyclic prodrug and DADLE in guinea pigs were approximately 80 and 8.5 times greater, respectively, than the levels observed in rat brain. The brain-to-plasma prodrug concentration ratios in guinea pigs (0.6) were significantly higher than the ratio observed in rats (0.01). These species differences are most likely due to the different substrate specificities of the efflux transporters that facilitate liver clearance of these prodrugs and limit their permeation into the brain. ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 94:2676–2687, 2005
Keywords: opioid peptide; prodrugs; chirality; blood-brain barrier; membrane permeability; biliary excretion; cytochrome P450; efflux pumps; species differences
INTRODUCTION Correspondence to: R.T. Borchardt (Telephone: 785-8643427; Fax: 785-864-5736; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 94, 2676–2687 (2005) ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association
2676
The poor oral bioavailability and blood-brain barrier (BBB) permeation of a natural peptide (e.g., Leu5-enkephalin) result from its unfavorable physicochemical properties (e.g., charge,
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
hydrophilicity, high hydrogen-bonding potential) for transcellular permeation and its lack of metabolic stability.1–6 In addition, many nonmetabolic clearance pathways that exist in the kidney and liver involve transporters such as P-glycoprotein (P-gp, MDR1 in human) and multidrug resistance related protein 2 (MRP2).7,8 Some of the transporters (e.g., MDR1, MRP2) that facilitate this clearance in the kidney and the liver can also limit drug permeation across the intestinal mucosa and the BBB.8 Thus, despite the favorable CNS pharmacological properties of various natural and synthetic peptides and peptidomimetics, their clinical development in oral dosage forms remains a major challenge.9,10 The clinical applications of such brain-targeted peptide-based drugs would include the treatment of brain disorders such as Alzheimer disease and brain cancer or pain management.9,11,12 In an attempt to improve the oral bioavailability and BBB permeation of peptides and peptidomimetics, our laboratory has synthesized cyclic prodrugs of opioid peptides.13–16 Compared to the opioid peptides, these cyclic prodrugs exhibited improved stability to hydrolytic pathways of metabolism and better permeation across various cell membranes in the presence of inhibitors of efflux transporters.17,18 However, efflux transporters such as P-gp and MRP2 adversely affected the cell membrane permeation of the cyclic prodrugs by producing net basolateral-to-apical flux. In addition, these cyclic prodrugs were rapidly cleared by the liver into the bile after i.v. administration to rats.19 Again, biliary excretion is probably mediated by efflux transporters (e.g., P-gp, MRP2) that are present not only in the liver but also in the intestinal mucosa and BBB.7,8,20,21 Therefore, it appears that efflux transporters constitute a major barrier to the systemic availability of the cyclic prodrugs after oral administration and their access to their therapeutic target in the central nervous system (CNS) after i.v. administration. Another problem was that the bioconversion of the cyclic prodrugs in the brain versus peripheral fluids and tissue needed to be optimized.19 In an attempt to overcome these problems, our laboratory recently explored how modifications in the chirality of specific amino acid residues (Ala2 and Leu5) in the peptide portion of oxymethylmodified coumarinic acid (OMCA)-based cyclic prodrugs of the opioid peptide [Ala2,Leu5]-Enk (H-Tyr-Ala-Gly-Phe-Leu-OH) affected their membrane permeabilities and their bioconversion by esterases.22,23 Although we have observed dif-
2677
ferences in the solution structures of these cyclic prodrugs and their calculated physicochemical properties (e.g., polar surface area), no significant differences in the transport characteristics of these cyclic prodrugs were observed. Particularly significant was the observation that all compounds were substrates for efflux transporters such as Pgp and MRP2 and, thus, showed low Caco-2 cell permeation. In vitro stability studies of the cyclic prodrugs, demonstrated that the changes in the chirality of amino acid residues in the peptide portion of the cyclic prodrugs led to dramatic differences in their rates of bioconversion in plasma and tissue homogenates (brain and liver).23 In general, these molecules were more rapidly metabolized if an L-amino acid was placed at the C-terminus of the peptide (i.e., OMCA-[D-Ala2,Leu5]-Enk and OMCA-[Ala2,Leu5]-Enk). Moreover, these studies suggested that canine and guinea pig would be the most relevant animal models for the in vivo pharmacokinetic evaluations of cyclic prodrugs of opioid peptides because their enzymatic stability profiles in biological media (e.g., plasma, brain, liver) were similar to that of humans. The objective of this study was to conduct in vivo pharmacokinetic experiments of OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk (Fig. 1) in rats and guinea pigs to determine liver clearance, bioconversion efficiencies, and partitioning into the brain. OMCA-[D-Ala2,Leu5]-Enk, another opioid peptide, was chosen for these in vivo studies based on previous in vitro studies.23 These studies had shown more desirable prodrug bioconversion profiles for OMCA-[D-Ala2,Leu5]-Enk (e.g., longer half-lives in plasma than in brain) compared to OMCA-DADLE (e.g., longer half-lives in brain than in plasma) (Fig. 1). Very significant differences were observed in the disposition of these cyclic prodrugs in rats and guinea pigs. In an effort to explain these differences, in vitro stability studies were conducted using rat and guinea pig microsomes.
MATERIALS AND METHODS Materials The cyclic prodrugs OMCA-DADLE (>95%) and OMCA-[D-Ala2,Leu5]-Enk (>90%), their internal standard CA-[Leu5]-Enk (coumarinic acid-based cyclic prodrug of [Leu5]-enkephalin), and the capped peptide Ac-DADLE-NH2 were synthesized
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2678
LIEDERER ET AL.
nine dinucleotide phosphate (b-NADPH; reduced form), and Hank’s balanced salt solution (HBSS) (modified) were purchased from Sigma-Aldrich (St. Louis, MO). Potassium phosphate monobasic (KH2PO4) and potassium phosphate dibasic (K2HPO4) were obtained from Fisher Scientific (Pittsburgh, PA). Isoflurane, ketamine, xylaxine, lidocaine hydrochloride, Puralube1 Vet Ophthalmic Ointment, and atropine were obtained from Henry Schein (Melville, NY) through the Animal Care Unit at The University of Kansas (Lawrence, KS). All chemicals and solvents (HPLC grade) were used as received.
In Vivo Animal Studies Pharmacokinetic Studies
Figure 1. Proposed mechanism for the release of the opioid peptides [D-Ala2,Leu5]-Enk and DADLE from the respective oxymethyl-modified coumarinic acid-based cyclic prodrugs.
in our laboratory following procedures reported elsewhere.13,22,24 DADLE, its internal standard [Leu5]-Enk, diethyl p-nitrophenyl phosphate (paraoxon, approx. 90%), dimethyl sulfoxide (DMSO, >99.5%), polyethylene glycol (average mol. wt. 300) (PEG 300), guanidine hydrochloride (>99%), ketoconazole, b-nicotinamide ade-
For pharmacokinetic studies, male Dunkin Hartley guinea pigs (200–250 g) from Harlan (Indianapolis, IN) were used. Male Dunkin Hartley guinea pigs (300–350 g), which were carotid artery, jugular vein, and/or bile duct cannulated, were obtained from Hilltop Lab Animals, Inc. (Scottdale, PA). Male Sprague Dawley rats (200–250 g) with a carotid artery or bile duct chronically implanted were obtained from Harlan. All animals were housed individually and fasted overnight before use. Water was allowed ad libitum. For each compound to be studied, three to six animals were given an i.v. dose of 1 mg/kg of the drug (200–400 mL). The solvent used for i.v. dosing consisted of 75% (v/v) saline, 20% (v/v) PEG 300, and 5% (v/v) DMSO. For plasma stability studies in rats, male Sprague Dawley rats with a cannula chronically implanted into the carotid artery were used. For the same experiment in guinea pigs, Dunkin Hartley guinea pigs, which were carotid artery and jugular vein cannulated, were used. Rats were preanesthetized with isoflurane prior to an i.m. injection of ketamine/xylazine followed by the i.v. injection of the cyclic opioid prodrugs. Moreover, a sterile ocular lubricant (Puralube1 Vet Ophthalmic Ointment) was used to avoid dryness and irritation of the eyes during anesthesia. First, a small incision was made on the medial surface of the hind leg and a small amount of lidocaine hydrochloride solution (for nerve block) applied. The injection was then made into the femoral vein, and the incision was closed with wound clips. The injection in guinea pigs was made through one of the chronically implanted cannulas, typically the jugular vein cannula. After i.v. administration of
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
OMCA-[D-Ala2,Leu5]-Enk in rats, blood samples (approx. 0.2 mL) were withdrawn from the carotid artery of each cannulated animal at 2, 4, 6, 10, 16, 20, and 25 min. After i.v. administration of OMCA[D-Ala2,Leu5]-Enk and OMCA-DADLE in guinea pigs, blood samples (approx. 0.2 mL) were withdrawn from the carotid artery or jugular vein cannula of each cannulated animal at 2, 5, 10, 15, 20, 30, 40, 60, 80, 100, and 120 min or at 2, 5, 10, 20, 30, 40, 60, 90, 120, 180, 240, 300, and 360 min, respectively. In between various sampling intervals, the cannulas were filled with saline containing heparin (10 U/mL) to maintain patency. The withdrawn blood was centrifuged immediately at 1900 g and 48C for approximately 5 min using a Thermo IEC Micromax microcentrifuge (Fisher Scientific) to separate the plasma from the erythrocytes. Plasma stability data for OMCADADLE in rats was determined previously in our laboratory and the results were used for comparison.19 Brain Permeation Studies Experiments were also conducted to determine the abilities of the cyclic prodrugs to partition into the brain after i.v. drug administration to guinea pigs. For these experiments, male Dunkin Hartley guinea pigs (350–400 g; Harlan) were used. For each compound studied, three guinea pigs were each given a 1 mg/kg i.v. dose of the prodrug into the saphenous vein. Prior to i.v. injections, the guinea pigs received a s.c. injection of atropine (0.05 mg/kg) and were anesthetized with isoflurane. At 10 min after i.v. administration, the animals were euthanized by decapitation while anesthetized with isoflurane, and the brains were rapidly removed and weighed. Each brain was homogenized (30 strokes, Wheaton) with 5 mL of ice-cold HBSS and split up into aliquots (approx. 1.5 mL). Approximate brainto-plasma concentration ratios for the cyclic prodrugs were calculated using the plasma concentrations (ng/mL) at 10 min (obtained from the pharmacokinetic studies) and 10 min postdose brain concentrations (ng/g tissue). The brain uptake for OMCA-DADLE in rats was determined previously in our laboratory and the results were used for species comparison.19 Biliary Clearance Studies To determine the extent of biliary excretion of the cyclic prodrugs, three to four male guinea pigs and rats were used, each having a chronically
2679
implanted bile duct continuous loop cannula. Additionally, the guinea pigs had a cannula chronically implanted into the carotid artery for drug administration. IV injections in rats were given as outlined above using the femoral vein. Cutting the continuous loop catheter resulted in a flow of bile and allowed the removal of bile at various time intervals through 2 h (5, 10, 15, 20, 30, 40, 50, 60, 80, 100, and 120 min) after drug administration. The volume of bile samples collected was dependent on the secretion rate of each animal and was determined gravimetrically. Following bile collection the catheter loop was reestablished. Processing of Biological Samples for Analysis To avoid degradation of the analytes, all biological samples (plasma, bile, and brain homogenates) were stored at 808C until pretreatment and analysis by high performance liquid chromatography with tandem mass spectrometric detection (LC-MS-MS). The pretreatment of the samples is extensively discussed elsewhere.19 Briefly, all samples were spiked with a solution mixture containing [Leu5]-Enk, an internal standard of the linear peptides, and CA-[Leu5]-Enk, an internal standard of the cyclic prodrugs, to correct for sample losses during sample cleanup. For brain homogenates, sample preparation included both protein precipitation with acetonitrile and solid phase extraction (SPE) using Waters Oasis HLB extraction cartridges (1 cc/30 mg, Waters) due to the complex nature of the tissue. For bile samples the concentration of the analytes had to be reduced by dilution but no further pretreatment was necessary. Plasma samples were treated with acetonitrile as the protein precipitation agent, followed by centrifugation. The supernatant was concentrated by solvent evaporation and reconstituted in 10% (v/v) acetonitrile. The solution was then centrifuged again and the supernatant removed for analysis. Pharmacokinetic parameters were obtained by fitting the plasma concentration-time data to a two-compartment model using SAAM II Version 1.2.1 software (Saam Institute, Inc., Seattle, WA). In Vitro Microsomal Stability Studies Pooled male Sprague Dawley rat liver microsomes were obtained from BD Biosciences (Woburn, MA) and pooled male Hartley albino guinea pig liver microsomes were obtained from Xenotech, LLC (Lenexa, KS).
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2680
LIEDERER ET AL.
For in vitro stability assessments in rat and guinea pig liver microsomes, DADLE, Ac-DADLENH2, and the cyclic prodrugs OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk (final conc. 2.5 mM) were each preincubated for 5 min at 378C (n ¼ 3) in 0.5 mL phosphate buffer (100 mM) solution at pH 7.4 with 1 mM b-NADPH in the presence and absence of 100 mM paraoxon, an esterase B inhibitor, and/or 5 mM ketoconazole, a CYP3A4 inhibitor. The enzymatic reaction was started by the addition of rat or guinea pig liver microsomes (approx. 400 nM cytochrome P450s) followed by 15 min of incubation at 378C. The reaction was quenched by adding 100 mL of methanol and 750 mL of acetonitrile to each sample. After vortex mixing, the solution was centrifuged at 13400 g for 10 min and 500 mL of supernatant was taken and evaporated (Centrivap concentrator, Labcono, Kansas City, MO). The residue was then reconstituted in 100 mL of 50% (v/v) acetonitrile and analyzed by LC-MS-MS. For control samples, 100 mL methanol was added to 0.5 mL phosphate buffer solution (100 mM, pH 7.4) containing 1 mM b-NADPH and 2.5 mM of the studied compound before the addition of liver microsomes. After quickly adding 750 mL acetonitrile, the solution was processed as outlined above. The percentage of analyte remaining was calculated based on the average analyte peak area divided by the average analyte peak area in the control sample. In vitro half-lives (t1/2) of the cyclic prodrugs were calculated from the disappearance rates in rat and guinea pig liver microsomes.25 ClInt, predicted and ClHep, predicted were calculated as follows:25 ClInt; predicted ¼
0:693 g of liver weight t1=2 ðminÞ kg of body weight mL incubation mg of microsomal protein 45 mg of microsomal protein g of liver weight
ClHep; predicted ¼ ðQ ClInt; predicted Þ=ðQ þ ClInt; predicted Þ
The hepatic blood flow (Q) used was 55.2 mL/ min/kg and 62.2 mL/min/kg for rat and guinea pig, respectively.26,27 Analytical Methods For in vitro and in vivo studies, LC-MS-MS was employed using a Quattro Micro triple quadrupole mass spectrometer (Micromass, Beverly, MA). The LC was conducted using a Waters 2690
HPLC system (Waters, Milford, MA). The analytical column, which was kept at 258C, was a Vydac (Hesperia, CA) C18 HPLC column (218TP5101, ˚ ,) with a guard column 50 1.0 mm i.d., 300 A (5 micron microbore cartridge, Vydac 218GK51). Sample separation was achieved utilizing a linear gradient from 5% to 100% B (Mobile phase: A, water with 0.1% FA (v/v); B, acetonitrile with 0.1% FA (v/v)), and a flow rate of 0.2 mL/min. The total run time was 12 min (5 min gradient 5% to 100% B; 2 min 100% B, 1 min reduction of 100% B to 5% B, 4 min equilibration with 5% B). The eluent from the first 2 min and the last 3 min was directed to waste. Samples and internal standards were recorded by multiple reaction monitoring (DADLE: 570.2 > 120.1; Ac-DADLENH2: 611.0 > 333.9; cyclic prodrugs: 728.3 > 120.1; [Leu5]-Enk (internal standard): 556.3 > 120.1; CA-[Leu5]-Enk (internal standard): 684.3 > 136.2). The intermediate(s) were indirectly determined as described elsewhere.19 The MS-MS conditions for the capped peptide Ac-DADLENH2 were: capillary voltage 4.0 kV; cone voltage 25 V, source temperature 1008C; desolvation gas flow 740 L/h; multiplier voltage 650 V; dwell time 0.5 s; inter-channel delay 0.03 s. The MS-MS conditions for the cyclic prodrugs were: capillary voltage 3.20 kV; cone voltage 30 V, source temperature 1208C; desolvation gas flow 740 L/h; multiplier voltage 650 V; dwell time 0.15 s; interchannel delay 0.03 s. Data acquisition and analysis were performed using Mass Lynx v3.5 and v4.0 software (Micromass). Further information on analytical methods developed in our laboratory for cyclic prodrugs can be found elsewhere.28
RESULTS In Vivo Animal Studies Pharmacokinetic Studies After i.v. administration of OMCA-[D-Ala2,Leu5]Enk and OMCA-DADLE in guinea pigs, the cyclic prodrugs disappeared rapidly from the systemic circulation following a two-phase exponential decay (Fig. 2A and B). Plasma concentrations could be detected for only up to 1 h. The respective intermediates and linear peptides (i.e., [D-Ala2, Leu5]-Enk and DADLE) generated by enzymatic and chemical breakdown of OMCA-[D-Ala2,Leu5]Enk and OMCA-DADLE were detected, however, at lower concentrations than the cyclic prodrugs.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
2681
Figure 2. Plasma concentration-time curves of OMCA-DADLE (squares) (A) and OMCA-[D-Ala2,Leu5]-Enk (squares) (B) and their respective intermediates (triangles) and linear peptides (circles) in guinea pigs (n ¼ 5–6) after i.v. administration (1 mg/kg) of each cyclic prodrug.
The plasma concentration time curves for OMCADADLE and OMCA-[D-Ala2,Leu5]-Enk are shown in Figure 2A and B, respectively. The plasma data were best described by a two-compartment model, and various pharmacokinetic parameters (AUC, clearance, Vss, k(0,1), t1/2) were estimated for the cyclic prodrugs (Tab. 1). The calculated parameters were similar and comparable to those obtained earlier in our laboratory from rat plasma data after i.v. administration of OMCA-DADLE to rats (Tab. 1).19 While the body clearance of OMCA-[D-Ala2,Leu5]-Enk (168 mL/min/kg) in guinea pigs and of OMCA-DADLE (81 mL/min/ kg) in rats exceeded hepatic blood flow the body clearance of OMCA-DADLE (61 mL/min/kg) in guinea pigs approximated the hepatic blood flow.
After i.v. administration of OMCA-[D-Ala2,Leu5]-Enk in rats the cyclic prodrug could not be quantified in plasma, because it was below the limits of quantitation of the assay (5 ng/mL). However, the conversion products (the intermediate(s)) and [D-Ala2,Leu5]-Enk) were observed indicating instantaneous breakdown of OMCA[D-Ala2,Leu5]-Enk (data not shown). It should be mentioned that due to the significant instability of OMCA-[D-Ala2,Leu5]-Enk in rat plasma additional degradation of the cyclic prodrug may have occurred during blood sample processing. Brain Permeation Studies Brain uptake of the cyclic prodrugs was determined 10 min after i.v. administration of
Table 1. Pharmacokinetic Parameters of OMCA-[D-Ala2,Leu5]-Enk after i.v. Administration (1 mg/kg) to Guinea Pigs and of OMCA-DADLE after i.v. Administration (1 mg/kg) to Guinea Pigs and Rats PK Parameters (Mean SE) AUC (ng min/mL) Clearance (mL/min/kg) Vss (mL/kg) k(0,1) (1/min) t1/2 (min)
OMCA- [D-Ala2,Leu5]-Enk (Guinea Pig) 7,570 2,080 168 29 437 178 0.3 0.1 2.7 0.4
OMCA- DADLE (Guinea Pig)
OMCA- DADLE (Rat)a
26,760 9,030 61 27 508 254 0.2 0.03 4.0 0.7
12,760 2,960 81 21 954 415 0.2 0.04 18 4
Pharmacokinetic parameters for OMCA-[D-Ala2,Leu5]-Enk in rats could not be determined because the cyclic prodrug could not be quantified in plasma, because it was below the limits of quantitation. a From Yang et al.19 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2682
LIEDERER ET AL.
Table 2. Brain Concentrations and Brain-to-plasma Concentration Ratios of OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk at 10 min after i.v. Administration (1 mg/kg) to Rats and Guinea Pigs Brain Uptake (ng/g tissue, mean SD, n ¼ 3)
Species b
Rat Guinea Pig Guinea Pig
Prodrug
Linear Peptidea
Intermediate
Brain-to-Plasma Prodrug Concentration Ratio
3.30 0.33 263.6 114.8 170.9 119.5
1.34 1.50 11.5 2.6 10.9 3.9
NDc NDc NDc
0.01 0.6 0.9
Compound OMCA-DADLE OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk
a
DADLE or [D-Ala2,Leu5]-Enk. From Yang et al.19. c ND, not determined because it was below the limit of quantitation. b
OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE in guinea pigs and compared to previous in vivo rat data from our laboratory (Tab. 2). For OMCADADLE, the brain levels of the cyclic prodrug and DADLE were approximately 80 and 8.5 times greater, respectively, than the levels observed in rats (Tab. 2). Administration of OMCA-[DAla2,Leu5]-Enk in guinea pigs lead to similar high levels of the prodrug and [D-Ala2,Leu5]-Enk in brain (Tab. 2). No intermediates from either cyclic prodrug were detected in brain because they were below the limits of detection. The brainto-plasma concentration ratio of the cyclic prodrugs was significantly greater in guinea pigs than in rats (Tab. 2).
respective intermediates and the linear peptides of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE were present in rat and guinea pig bile, the main excretion products were the cyclic prodrugs. In rats, rapid clearance of OMCA-[D-Ala2,Leu5]-Enk into the bile occurred after i.v. administration. Approximately 52% of the administered dose of OMCA-[D-Ala2,Leu5]-Enk was recovered within 2 h as the intact prodrug. Very similar results were observed earlier by Yang et al. after i.v. administration of OMCA-DADLE to rats.19 In contrast, only small amounts of the administered dose (<4%) of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE were recovered in bile after i.v. administration to guinea pigs.
Biliary Clearance Studies
In Vitro Microsomal Stability Studies 2
5
Biliary excretion of OMCA-[D-Ala ,Leu ]-Enk in rats and in guinea pigs as well as biliary excretion of OMCA-DADLE in guinea pigs was examined to elucidate its contribution to first-pass clearance. The resulting data are displayed in Table 3 and Figure 3. Although detectable levels of the
One possible explanation for the significant differences observed in the disposition of the cyclic prodrugs in rat and guinea pig would be differences in their metabolism. Therefore, stability studies were conducted using rat and guinea pig liver microsomes. These studies showed that,
Table 3. Bile Recovery over a 2 h Time Period of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE after i.v. Administration to Rats and Guinea Pigs Bile Recovery of Dose (%, mean SE, n ¼ 3–4)
Species b
Rat Rat Guinea Pig Guinea Pig
Compound
Prodrug
Linear Peptidea
Intermediate
OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk
38.1 2.1 51.7 5.8 3.3 0.4 0.82 0.1
3.3 0.4 6.0 0.7 0.11 0.01 0.05 0.01
7.3 0.7 6.0 0.7 00 00
a
DADLE or [D-Ala2,Leu5]-Enk. From Yang et al.19.
b
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
Figure 3. Cumulative biliary excretion of OMCA-[DAla2,Leu5]-Enk in rats (squares) and guinea pigs (circles) and of OMCA-DADLE in rats (diamonds)19 and guinea pigs (triangles) (n ¼ 3–4).
under the experimental conditions employed, the linear peptide DADLE was stable in rat and guinea pig liver microsomes. In contrast, the capped peptide Ac-DADLE-NH2 and the cyclic prodrugs OMCA-DADLE and OMCA-[D-Ala2, Leu5]-Enk were metabolized in this biological medium (Fig. 4A and B). For both the capped peptide and the cyclic prodrugs, the metabolism was more pronounced in guinea pig microsomes than in rat microsomes (Fig. 4A and B, Tab. 4). The capped peptide was noticeably more stable in rat and guinea pig liver microsomes than were the cyclic prodrugs. While paraoxon showed no significant effect on the stability of the capped peptide or the cyclic prodrugs in liver microsomes, ketoconazole substantially inhibited metabolism of Ac-DADLE-NH2 and OMCA-DADLE. Ketoconazole alone did not significantly inhibit the metabolism of OMCA-[D-Ala2,Leu5]-Enk in guinea pig liver microsomes. Interestingly, OMCA-[D-Ala2, Leu5]-Enk could be stabilized in guinea pig liver microsomes in the presence of both paraoxon and ketoconazole. In both species the predicted hepatic clearance was approximately equal to hepatic blood flow and the predicted intrinsic clearance exceeded hepatic blood flow (Tab. 4).
DISCUSSION Previous in vivo studies in rats showed high biliary clearance and low brain penetration of OMCA-DADLE.18,19 Additionally, in vitro studies
2683
demonstrated that while the conversion of this prodrug to DADLE in the target organ brain was too slow, its conversion in plasma was too fast.19 Therefore, prodrugs of diastereomers of DADLE such as OMCA-[D-Ala2,Leu5]-Enk were synthesized.22 These prodrugs exhibited more favorable prodrug conversion profiles than did OMCADADLE since their conversion in brain was now faster than in plasma.23 The in vitro studies also showed that guinea pigs resembled the enzymatic stability profiles (e.g., t1/2 in various biological media such as plasma) of humans better than did rats. Therefore, guinea pigs appeared to be a more appropriate animal model than rats for in vivo pharmacokinetic evaluations of cyclic prodrugs of opioid peptides. Based on these observations, OMCA-[D-Ala2,Leu5]-Enk and OMCADADLE were administered i.v. to guinea pigs. Additionally, OMCA-[D-Ala2,Leu5]-Enk was administered i.v. to rats for in vivo species comparison. After i.v. administration of OMCA-[D-Ala2, Leu5]-Enk and OMCA-DADLE in guinea pigs and rats, the cyclic prodrugs disappeared rapidly from the systemic circulation. The estimated pharmacokinetic parameters from the plasma data for OMCA-[D-Ala2,Leu5]-Enk and OMCADADLE in guinea pigs were similar (Tab. 1). In addition, the parameters of OMCA-DADLE in guinea pigs were comparable to those obtained in previous studies of OMCA-DADLE in rats (Tab. 1).19 More pronounced species differences were observed for OMCA-[D-Ala2,Leu5]-Enk. As mentioned earlier, in contrast to OMCA-[D-Ala2,Leu5]-Enk in guinea pigs, OMCA-[D-Ala2,Leu5]Enk in rats could not be quantified in plasma. Therefore, OMCA-[D-Ala2,Leu5]-Enk appeared to be significantly less stable in rats than in guinea pigs. This observation is consistent with previous in vitro data from our laboratory where the half-life of OMCA-[D-Ala2,Leu5]-Enk in rat plasma (t1/2 (-Paraoxon) < 2 min) was much shorter than in guinea pig plasma (t1/2 (-Paraoxon) 23 min).23 The body clearance of OMCA-[D-Ala2,Leu5]-Enk (168 mL/min/kg) in guinea pigs and of OMCADADLE in rats (81 mL/min/kg)19 exceeded hepatic blood flow in both species suggesting extrahepatic metabolism (e.g., in blood or kidney), nonmetabolic clearance processes, and/or extensive partitioning of a compound into red blood cells. Previous studies have shown that urinary excretion of cyclic prodrugs of DADLE in rats is minor and thus does not significantly contribute to total body clearance.19 In addition, cyclic prodrugs of opioid peptides appear to be chemically stable (e.g.,
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2684
LIEDERER ET AL.
Figure 4. Stability of DADLE, Ac-DADLE-NH2, and the cyclic prodrugs OMCADADLE and OMCA-[D-Ala2,Leu5]-Enk in rat (A) and guinea pig (B) liver microsomes at 378C in the absence of inhibitors (no shading), in the presence of paraoxon (complete shading), an esterase B inhibitor, in the presence of ketoconazole (diagonal shading), a CYP3A4 inhibitor, and in the presence of paraoxon and ketoconazole (vertical shading). The incubation time was 15 min. Data are presented as mean S.D. (n ¼ 3).
Table 4. In Vitro Half-Lives of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE in Rat and Guinea Pig Plasma and Microsomes at 378C and Predicted Clearance Values t1/2 (min, mean SD) Species Rat Rat Guinea Pig Guinea Pig
Compound OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk
Plasmaa
Microsomes
10 1 <2 128 5 23 1
11 2 5.7 0.6 4.2 0.3 4.7 0.2
See Materials and Methods for experimental details and calculations. a From Liederer et al.23. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
Cl (mL/min/kg, mean SD) ClInt,
predicted
499 67 986 98 1706 119 1518 70
ClHep,
predicted
50 1 52 1 60 1 60 1
DISPOSITION OF CYCLIC PRODRUGS
t1/2 (OMCA-DADLE) approximately 390 min and 290 min at pH 4 and pH 7, respectively).17However, previous in vitro stability studies of the cyclic prodrugs in rat and guinea pig biological media suggested that metabolism in plasma contributes to extrahepatic clearance for both cyclic prodrugs in rats and for OMCA-[D-Ala2,Leu5]-Enk in guinea pigs (Tab. 4). OMCA-DADLE was relatively stable in guinea pig plasma (Tab. 4, t1/2 &128 min). Thus, its contribution to extrahepatic clearance should be less significant. The predicted in vivo intrinsic clearance overestimated the values observed in vivo and thus showed poor in vitro–in vivo correlation. Brain uptake studies of OMCA-[D-Ala2,Leu5]Enk and OMCA-DADLE in guinea pigs showed accumulation of the cyclic prodrugs and their respective linear peptides in the brain. Interestingly, the observed brain levels and the brain-toplasma prodrug concentration ratio of these compounds were dramatically higher in guinea pigs than the levels and ratio for OMCA-DADLE in rats (Tab. 2). The limited ability of OMCA-DADLE to cross the rat BBB should not have been caused by plasma protein binding, because a large fraction of the drug (approx. 40%) was found to be unbound in rat plasma.19 In addition, even though OMCADADLE is less stable in rat than in guinea pig biological media (plasma, liver, and brain homogenate),23 which can impact brain uptake negatively because less prodrug would be able reach the brain, the differences did not appear significant enough to affect the brain uptake to the extent observed. In vivo studies revealed extensive bile excretion of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE in rats (Tab. 3, Fig. 3). Significantly lower hepatic clearance was observed for the linear peptides and the intermediates. The majority of all molecules (cyclic prodrugs, linear peptides, and intermediates) excreted over a 2 h time period were recovered rapidly (within approx. 20 min) after drug administration. In contrast to these findings in rats, bile excretion of the cyclic prodrugs, their respective intermediates and linear peptides in guinea pigs was minor. While less than approximately 4% of the total dose was recovered in guinea pig bile, more than approximately 40% was recovered in rat bile (Tab. 3). In an attempt to explain these dramatic species differences, stability experiments were conducted in rat and guinea pig liver microsomes. Both cyclic prodrugs were more actively metabolized in guinea pig liver microsomes (t1/2 (OMCA-DADLE)
2685
&4 min, t1/2 (OMCA-[D-Ala2,Leu5]-Enk) &5 min) than in rat liver microsomes (t1/2 (OMCA-DADLE) &11 min, t1/2 (OMCA-[D-Ala2,Leu5]-Enk) &6 min), illustrating species differences. Increased metabolism in guinea pig liver microsomes would negatively influence brain uptake because lower concentrations of the cyclic prodrugs would reach the brain. Therefore, these data did not explain the increased brain uptake in guinea pigs. However, the increased phase I metabolism in guinea pigs could at least partially account for the reduced bile clearance in guinea pigs versus rats. In both rats and guinea pigs the predicted in vitro hepatic clearance for OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk approximately equaled hepatic blood flow and thus did not indicate any species differences but suggested extensive metabolism of the cyclic prodrugs in the liver in both species. Thus, it appears more reasonable that the observed species differences in brain uptake and bile excretion are mainly due to the different substrate specificities of efflux transporters that are expressed in rat and guinea pig BBB and liver. These transporter systems may have greater affinity for the cyclic prodrugs in rats than in guinea pigs. Thus, they are more effective in limiting partitioning of the compounds into the brain. As mentioned earlier, efflux transporters that limit permeation of molecules into the brain also facilitate liver clearance.8 A higher affinity of the cyclic prodrugs for efflux transporters could, therefore, not only reduce CNS entry but also enhance liver clearance or vice versa. Knockout animal models have been used to determine the impact of efflux mechanisms on the disposition of various compounds both in brain and liver.29–31 These studies have shown that the deficiency of efflux transporters can increase brain uptake and decrease hepatic clearance for some compounds and thus support our hypothesis. Studies using Caco-2 cells, an in vitro model of the intestinal mucosa, demonstrated that the cyclic prodrugs exhibited strong substrate activity for P-gp and MRP2.22 The identification of the efflux protein(s) involved in brain uptake and biliary clearance processes of these cyclic prodrugs remains to be examined but might involve the same protein(s). In an attempt to elucidate the role of the chemical linker and/or the induced solution conformations by the linker on the microsomal stability of the cyclic prodrugs, the stability of the linear peptide DADLE and the capped peptide AcDADLE-NH2 were determined in rat and guinea pig microsomes.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2686
LIEDERER ET AL.
The data showed that DADLE was stable in contrast to the capped peptide and the cyclic prodrugs (Fig. 2A and B). The order of stability was DADLE > Ac-DADLE-NH2 > cyclic prodrugs. While paraoxon tended to show no significant effects on the microsomal stability of the cyclic prodrugs or the capped peptide, ketoconazole tended to inhibit the metabolism, suggesting that the capped peptides and cyclic prodrugs are substrates for cytochrome P450 enzymes (Fig. 4A and B). From these results, it appears that the cyclic prodrugs are substrates for phase I enzymes because of the chemical linker itself and/or the effect of the linker on the solution structure. Preliminary studies in our laboratory have identified, by LCMS-MS, oxidative metabolism of the aromatic residues (e.g., Phe hydroxylation) when the cyclic prodrugs were exposed to cytochrome P450s (unpublished data; Tarra Fuchs et al.). Thus it appears that the chemical linker is not the major metabolically labile site. Based on these data, we hypothesize that the chemical linker mediates oxidation reactions by inducing conformational constraints and a solution structure that is more favorable for interactions with cytochrome P450 enzymes than the solution structure present in the linear and capped molecule. In an attempt to block this observed oxidative metabolism, our laboratory is currently synthesizing a derivative of the coumarin-based cyclic prodrug of DADLE with appropriate chemical modifications in the peptide portion of the molecule.
CONCLUSIONS The pharmacokinetic characteristics of OMCADADLE and OMCA-[D-Ala2,Leu5]-Enk, particularly brain uptake and biliary excretion, were highly species-dependent. Therefore, it would be difficult to select an animal species for preclinical studies that would best predict the pharmacokinetic properties of these cyclic prodrugs in human. The observed species differences of the cyclic prodrugs were most likely due to the different substrate specificities of efflux transporters in rats and guinea pigs.
ACKNOWLEDGMENTS The authors thank Dr. Tarra Fuchs for helpful discussions and technical support and Dr. Michelle McIntosh for valuable help in the determination of the pharmacokinetic parameters of the cyclic
prodrugs. This research was supported by a grant from NIH/NIDA (DA-03315).
REFERENCES 1. Atkinson F, Cole S, Green C, van de Waterbeemd H. 2002. Lipophilicity and other parameters affecting brain penetration. Curr Med ChemCentral Nervous System Agents 2:229–240. 2. Veber DF, Johnson SR, Cheng HY, Smith BR, Ward KW, Kopple KD. 2002. Molecular properties that influence the oral bioavailability of drug candidates. J Med Chem 45:2615–2623. 3. Pauletti GM, Gangwar S, Knipp GT, Nerkurkar MM, Okumu FW, Tamura K, Siahaan TJ, Borchardt RT. 1996. Structural requirements for intestinal absorption of peptide drugs. J Control Release 41:3–17. 4. Wacher VJ, Silverman JA, Zhang Y, Benet LZ. 1998. Role of P-glycoprotein and cytochrome P450 3A in limiting oral absorption of peptides and peptidomimetics. J Pharm Sci 87:1322–1330. 5. Adessi C, Soto C. 2002. Converting a peptide into a drug: Strategies to improve stability and bioavailability. Curr Med Chem 9:963–978. 6. Humphrey MJ, Ringrose PS. 1986. Peptides and related drugs: A review of their absorption, metabolism, and excretion. Drug Metab Rev 17:283– 310. 7. Chandra P, Brouwer KL. 2004. The complexities of hepatic drug transport: Current knowledge and emerging concepts. Pharm Res 21:719–735. 8. Schinkel AH, Jonker JW. 2003. Mammalian drug efflux transporters of the ATP binding cassette (ABC) family: An overview. Adv Drug Deliv Rev 55:3–29. 9. Pardridge WM. 2003. Blood-brain barrier drug targeting: The future of brain drug development. Mol Interv 3:90–105. 10. Lien S, Lowman HB. 2003. Therapeutic peptides. Trends Biotechnol 21:556–562. 11. Gentilucci L. 2004. New trends in the development of opioid peptide analogues as advanced remedies for pain relief. Curr Top Med Chem 4: 19 –38. 12. Okada Y, Tsuda Y, Bryant SD, Lazarus LH. 2002. Endomorphins and related opioid peptides. In: Vitamins and hormones, Vol. 65. Elsevier Science, pp 257–279. 13. Ouyang H, Vander Velde DG, Borchardt RT, Siahaan TJ. 2002. Synthesis and conformational analysis of a coumarinic acid-based cyclic prodrug of an opioid peptide with modified sensitivity to esterase-catalyzed bioconversion. J Pept Res 59: 183–195. 14. Gangwar S, Pauletti GM, Siahaan TJ, Stella VJ, Borchardt RT. 1997. Synthesis of a novel esterasesensitive cyclic prodrug of a hexapeptide using an
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
(acyloxy)alkoxy promoiety. J Org Chem 62:1356– 1362. Wang B, Gangwar S, Pauletti GM, Siahaan TJ, Borchardt RT. 1997. Synthesis of a novel esterasesensitive cyclic prodrug system for peptides that utilize a ‘‘trimethyl lock’’-facilitated lactonization reaction. J Org Chem 62:1363–1367. Wang W, Camenisch G, Sane DC, Zhang H, Hugger E, Wheeler GL, Borchardt RT, Wang B. 2000. A coumarin-based prodrug strategy to improve the oral absorption of RGD peptidomimetics. J Control Release 65:245–251. Ouyang H, Tang F, Siahaan TJ, Borchardt RT. 2002. A modified coumarinic acid-based cyclic prodrug of an opioid peptide: Its enzymatic and chemical stability and cell permeation characteristics. Pharm Res 19:794–801. Chen W, Yang JZ, Andersen R, Nielsen LH, Borchardt RT. 2002. Evaluation of the permeation characteristics of a model opioid peptide, H-Tyr-DAla-Gly-Phe-D-Leu-OH (DADLE), and its cyclic prodrugs across the blood-brain barrier using an in situ perfused rat brain model. J Pharmacol Exp Ther 303:849–857. Yang JZ, Chen W, Borchardt RT. 2002. In vitro stability and in vivo pharmacokinetic studies of a model opioid peptide, H-Tyr-D-Ala-Gly-Phe-D-LeuOH (DADLE), and its cyclic prodrugs. J Pharmacol Exp Ther 303:840–848. Sun H, Dai H, Shaik N, Elmquist WF. 2003. Drug efflux transporters in the CNS. Adv Drug Deliv Rev 55:83–105. Chan LM, Lowes S, Hirst BH. 2004. The ABCs of drug transport in intestine and liver: Efflux proteins limiting drug absorption and bioavailability. Eur J Pharm Sci 21:25–51. Liederer BM, Fuchs T, Vander Velde D, Siahaan TJ, Borchardt RT. 2005. Effects of amino acid chirality and the chemical linker on the cell permeation characteristics of cyclic prodrugs of opioid peptides. Submitted. Liederer BM, Borchardt RT. 2005. Stability of oxymethyl-modified coumarinic acid cyclic prodrugs of diastereomeric opioid peptides in biological media from various animal species including human. J Pharm Sci 94:2198–2206. Wang B, Nimkar K, Wang W, Zhang H, Shan D, Gudmundsson O, Gangwar S, Siahaan T,
25.
26.
27.
28.
29.
30.
31.
2687
Borchardt RT. 1999. Synthesis and evaluation of the physicochemical properties of esterase-sensitive cyclic prodrugs of opioid peptides using coumarinic acid and phenylpropionic acid linkers. J Pept Res 53:370–382. Obach RS, Baxter JG, Liston TE, Silber BM, Jones BC, MacIntyre F, Rance DJ, Wastall P. 1997. The prediction of human pharmacokinetic parameters from preclinical and in vitro metabolism data. JPET 283:46–58. Boxenbaum H. 1980. Interspecies variation in liver weight, hepatic blood flow, and antipyrine intrinsic clearance: Extrapolation of data to benzodiazepines and phenytoin. J Pharmacokinet Biopharm 8:165– 176. Naritomi Y, Terashita S, Kimura S, Suzuki A, Kagayama A, Sugiyama Y. 2001. Prediction of human hepatic clearance from in vivo animal experiments and in vitro metabolic studies with liver microsomes from animals and humans. Drug Metab Dispos 29:1316–1324. Yang JZ, Bastian KC, Moore RD, Stobaugh JF, Borchardt RT. 2002. Quantitative analysis of a model opioid peptide and its cyclic prodrugs in rat plasma using high-performance liquid chromatography with fluorescence and tandem mass spectrometric detection. J Chromatogr B Analyt Technol Biomed Life Sci 780:269–281. Hoffmaster KA, Zamek-Gliszczynski MJ, Pollack GM, Brouwer KL. 2004. Hepatobiliary disposition of the metabolically stable opioid peptide [D-Pen2, D-Pen5]-enkephalin (DPDPE): Pharmacokinetic consequences of the interplay between multiple transport systems. J Pharmacol Exp Ther 311: 1203–1210. Wang JS, Ruan Y, Taylor RM, Donovan JL, Markowitz JS, DeVane CL. 2004. Brain penetration of methadone (R)- and (S)-enantiomers is greatly increased by P-glycoprotein deficiency in the blood-brain barrier of Abcb1a gene knockout mice. Psychopharmacology (Berl) 173:132– 138. Grauer MT, Uhr M. 2004. P-glycoprotein reduces the ability of amitriptyline metabolites to cross the blood brain barrier in mice after a 10-day administration of amitriptyline. J Psychopharmacol 18: 66–74.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
Department of Pharmaceutical Chemistry, The University of Kansas, Lawrence, Kansas 66047
2
Department of Chemistry, Creighton University, Omaha, Nebraska 68178
Received 17 March 2005; revised 14 July 2005; accepted 2 August 2005 Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.20476
ABSTRACT: The stabilities of DADLE ([D-Ala2,D-Leu5]-Enk, H-Tyr-D-Ala-Gly-Phe-DLeu-OH), the capped derivative Ac-DADLE-NH2, and the oxymethyl-coumarinic acid (OMCA)-based cyclic prodrug of DADLE and [D-Ala2,Leu5]-Enk (H-Tyr-D-Ala-Gly-PheLeu-OH) were determined at 378C in rat and guinea pig liver microsomes in the presence and absence of paraoxon, an esterase B inhibitor, and ketoconazole, a CYP3A4 inhibitor. These studies showed that the order of stability in microsomes was: DADLE >> AcDADLE-NH2 > OMCA-DADLE ¼ OMCA-[D-Ala2,Leu5]-Enk. While paraoxon produced no significant effect on the stability of the studied compounds in liver microsomes, ketoconazole inhibited the metabolism, suggesting that the capped peptide and the cyclic prodrugs are substrates for cytochrome P450 enzymes. For pharmacokinetic studies, the cyclic prodrugs of DADLE and [D-Ala2,Leu5]-Enk were administered i.v. to rats and guinea pigs. Various biological fluids and tissue (brain, bile, and blood) were collected and analyzed for the free peptide and the prodrugs by high performance liquid chromatography with tandem mass spectrometric detection (LC-MS-MS). These studies showed that the conversion of the cyclic prodrugs to the respective linear peptides (i.e., DADLE and [D-Ala2,Leu5]-Enk) was rapid in rat and guinea pig. In terms of drug elimination, only trace amounts of OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk were recovered in guinea pig bile (3.3% and 0.82%, respectively), while significant amounts were recovered in rat bile (38.1% and 51.7%, respectively). Brain uptake of the cyclic prodrugs in guinea pigs compared to previously determined brain uptake of OMCA-DADLE in rats was also significantly different. For OMCA-DADLE, the brain levels of the cyclic prodrug and DADLE in guinea pigs were approximately 80 and 8.5 times greater, respectively, than the levels observed in rat brain. The brain-to-plasma prodrug concentration ratios in guinea pigs (0.6) were significantly higher than the ratio observed in rats (0.01). These species differences are most likely due to the different substrate specificities of the efflux transporters that facilitate liver clearance of these prodrugs and limit their permeation into the brain. ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 94:2676–2687, 2005
Keywords: opioid peptide; prodrugs; chirality; blood-brain barrier; membrane permeability; biliary excretion; cytochrome P450; efflux pumps; species differences
INTRODUCTION Correspondence to: R.T. Borchardt (Telephone: 785-8643427; Fax: 785-864-5736; E-mail: [email protected]) Journal of Pharmaceutical Sciences, Vol. 94, 2676–2687 (2005) ß 2005 Wiley-Liss, Inc. and the American Pharmacists Association
2676
The poor oral bioavailability and blood-brain barrier (BBB) permeation of a natural peptide (e.g., Leu5-enkephalin) result from its unfavorable physicochemical properties (e.g., charge,
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
hydrophilicity, high hydrogen-bonding potential) for transcellular permeation and its lack of metabolic stability.1–6 In addition, many nonmetabolic clearance pathways that exist in the kidney and liver involve transporters such as P-glycoprotein (P-gp, MDR1 in human) and multidrug resistance related protein 2 (MRP2).7,8 Some of the transporters (e.g., MDR1, MRP2) that facilitate this clearance in the kidney and the liver can also limit drug permeation across the intestinal mucosa and the BBB.8 Thus, despite the favorable CNS pharmacological properties of various natural and synthetic peptides and peptidomimetics, their clinical development in oral dosage forms remains a major challenge.9,10 The clinical applications of such brain-targeted peptide-based drugs would include the treatment of brain disorders such as Alzheimer disease and brain cancer or pain management.9,11,12 In an attempt to improve the oral bioavailability and BBB permeation of peptides and peptidomimetics, our laboratory has synthesized cyclic prodrugs of opioid peptides.13–16 Compared to the opioid peptides, these cyclic prodrugs exhibited improved stability to hydrolytic pathways of metabolism and better permeation across various cell membranes in the presence of inhibitors of efflux transporters.17,18 However, efflux transporters such as P-gp and MRP2 adversely affected the cell membrane permeation of the cyclic prodrugs by producing net basolateral-to-apical flux. In addition, these cyclic prodrugs were rapidly cleared by the liver into the bile after i.v. administration to rats.19 Again, biliary excretion is probably mediated by efflux transporters (e.g., P-gp, MRP2) that are present not only in the liver but also in the intestinal mucosa and BBB.7,8,20,21 Therefore, it appears that efflux transporters constitute a major barrier to the systemic availability of the cyclic prodrugs after oral administration and their access to their therapeutic target in the central nervous system (CNS) after i.v. administration. Another problem was that the bioconversion of the cyclic prodrugs in the brain versus peripheral fluids and tissue needed to be optimized.19 In an attempt to overcome these problems, our laboratory recently explored how modifications in the chirality of specific amino acid residues (Ala2 and Leu5) in the peptide portion of oxymethylmodified coumarinic acid (OMCA)-based cyclic prodrugs of the opioid peptide [Ala2,Leu5]-Enk (H-Tyr-Ala-Gly-Phe-Leu-OH) affected their membrane permeabilities and their bioconversion by esterases.22,23 Although we have observed dif-
2677
ferences in the solution structures of these cyclic prodrugs and their calculated physicochemical properties (e.g., polar surface area), no significant differences in the transport characteristics of these cyclic prodrugs were observed. Particularly significant was the observation that all compounds were substrates for efflux transporters such as Pgp and MRP2 and, thus, showed low Caco-2 cell permeation. In vitro stability studies of the cyclic prodrugs, demonstrated that the changes in the chirality of amino acid residues in the peptide portion of the cyclic prodrugs led to dramatic differences in their rates of bioconversion in plasma and tissue homogenates (brain and liver).23 In general, these molecules were more rapidly metabolized if an L-amino acid was placed at the C-terminus of the peptide (i.e., OMCA-[D-Ala2,Leu5]-Enk and OMCA-[Ala2,Leu5]-Enk). Moreover, these studies suggested that canine and guinea pig would be the most relevant animal models for the in vivo pharmacokinetic evaluations of cyclic prodrugs of opioid peptides because their enzymatic stability profiles in biological media (e.g., plasma, brain, liver) were similar to that of humans. The objective of this study was to conduct in vivo pharmacokinetic experiments of OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk (Fig. 1) in rats and guinea pigs to determine liver clearance, bioconversion efficiencies, and partitioning into the brain. OMCA-[D-Ala2,Leu5]-Enk, another opioid peptide, was chosen for these in vivo studies based on previous in vitro studies.23 These studies had shown more desirable prodrug bioconversion profiles for OMCA-[D-Ala2,Leu5]-Enk (e.g., longer half-lives in plasma than in brain) compared to OMCA-DADLE (e.g., longer half-lives in brain than in plasma) (Fig. 1). Very significant differences were observed in the disposition of these cyclic prodrugs in rats and guinea pigs. In an effort to explain these differences, in vitro stability studies were conducted using rat and guinea pig microsomes.
MATERIALS AND METHODS Materials The cyclic prodrugs OMCA-DADLE (>95%) and OMCA-[D-Ala2,Leu5]-Enk (>90%), their internal standard CA-[Leu5]-Enk (coumarinic acid-based cyclic prodrug of [Leu5]-enkephalin), and the capped peptide Ac-DADLE-NH2 were synthesized
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2678
LIEDERER ET AL.
nine dinucleotide phosphate (b-NADPH; reduced form), and Hank’s balanced salt solution (HBSS) (modified) were purchased from Sigma-Aldrich (St. Louis, MO). Potassium phosphate monobasic (KH2PO4) and potassium phosphate dibasic (K2HPO4) were obtained from Fisher Scientific (Pittsburgh, PA). Isoflurane, ketamine, xylaxine, lidocaine hydrochloride, Puralube1 Vet Ophthalmic Ointment, and atropine were obtained from Henry Schein (Melville, NY) through the Animal Care Unit at The University of Kansas (Lawrence, KS). All chemicals and solvents (HPLC grade) were used as received.
In Vivo Animal Studies Pharmacokinetic Studies
Figure 1. Proposed mechanism for the release of the opioid peptides [D-Ala2,Leu5]-Enk and DADLE from the respective oxymethyl-modified coumarinic acid-based cyclic prodrugs.
in our laboratory following procedures reported elsewhere.13,22,24 DADLE, its internal standard [Leu5]-Enk, diethyl p-nitrophenyl phosphate (paraoxon, approx. 90%), dimethyl sulfoxide (DMSO, >99.5%), polyethylene glycol (average mol. wt. 300) (PEG 300), guanidine hydrochloride (>99%), ketoconazole, b-nicotinamide ade-
For pharmacokinetic studies, male Dunkin Hartley guinea pigs (200–250 g) from Harlan (Indianapolis, IN) were used. Male Dunkin Hartley guinea pigs (300–350 g), which were carotid artery, jugular vein, and/or bile duct cannulated, were obtained from Hilltop Lab Animals, Inc. (Scottdale, PA). Male Sprague Dawley rats (200–250 g) with a carotid artery or bile duct chronically implanted were obtained from Harlan. All animals were housed individually and fasted overnight before use. Water was allowed ad libitum. For each compound to be studied, three to six animals were given an i.v. dose of 1 mg/kg of the drug (200–400 mL). The solvent used for i.v. dosing consisted of 75% (v/v) saline, 20% (v/v) PEG 300, and 5% (v/v) DMSO. For plasma stability studies in rats, male Sprague Dawley rats with a cannula chronically implanted into the carotid artery were used. For the same experiment in guinea pigs, Dunkin Hartley guinea pigs, which were carotid artery and jugular vein cannulated, were used. Rats were preanesthetized with isoflurane prior to an i.m. injection of ketamine/xylazine followed by the i.v. injection of the cyclic opioid prodrugs. Moreover, a sterile ocular lubricant (Puralube1 Vet Ophthalmic Ointment) was used to avoid dryness and irritation of the eyes during anesthesia. First, a small incision was made on the medial surface of the hind leg and a small amount of lidocaine hydrochloride solution (for nerve block) applied. The injection was then made into the femoral vein, and the incision was closed with wound clips. The injection in guinea pigs was made through one of the chronically implanted cannulas, typically the jugular vein cannula. After i.v. administration of
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
OMCA-[D-Ala2,Leu5]-Enk in rats, blood samples (approx. 0.2 mL) were withdrawn from the carotid artery of each cannulated animal at 2, 4, 6, 10, 16, 20, and 25 min. After i.v. administration of OMCA[D-Ala2,Leu5]-Enk and OMCA-DADLE in guinea pigs, blood samples (approx. 0.2 mL) were withdrawn from the carotid artery or jugular vein cannula of each cannulated animal at 2, 5, 10, 15, 20, 30, 40, 60, 80, 100, and 120 min or at 2, 5, 10, 20, 30, 40, 60, 90, 120, 180, 240, 300, and 360 min, respectively. In between various sampling intervals, the cannulas were filled with saline containing heparin (10 U/mL) to maintain patency. The withdrawn blood was centrifuged immediately at 1900 g and 48C for approximately 5 min using a Thermo IEC Micromax microcentrifuge (Fisher Scientific) to separate the plasma from the erythrocytes. Plasma stability data for OMCADADLE in rats was determined previously in our laboratory and the results were used for comparison.19 Brain Permeation Studies Experiments were also conducted to determine the abilities of the cyclic prodrugs to partition into the brain after i.v. drug administration to guinea pigs. For these experiments, male Dunkin Hartley guinea pigs (350–400 g; Harlan) were used. For each compound studied, three guinea pigs were each given a 1 mg/kg i.v. dose of the prodrug into the saphenous vein. Prior to i.v. injections, the guinea pigs received a s.c. injection of atropine (0.05 mg/kg) and were anesthetized with isoflurane. At 10 min after i.v. administration, the animals were euthanized by decapitation while anesthetized with isoflurane, and the brains were rapidly removed and weighed. Each brain was homogenized (30 strokes, Wheaton) with 5 mL of ice-cold HBSS and split up into aliquots (approx. 1.5 mL). Approximate brainto-plasma concentration ratios for the cyclic prodrugs were calculated using the plasma concentrations (ng/mL) at 10 min (obtained from the pharmacokinetic studies) and 10 min postdose brain concentrations (ng/g tissue). The brain uptake for OMCA-DADLE in rats was determined previously in our laboratory and the results were used for species comparison.19 Biliary Clearance Studies To determine the extent of biliary excretion of the cyclic prodrugs, three to four male guinea pigs and rats were used, each having a chronically
2679
implanted bile duct continuous loop cannula. Additionally, the guinea pigs had a cannula chronically implanted into the carotid artery for drug administration. IV injections in rats were given as outlined above using the femoral vein. Cutting the continuous loop catheter resulted in a flow of bile and allowed the removal of bile at various time intervals through 2 h (5, 10, 15, 20, 30, 40, 50, 60, 80, 100, and 120 min) after drug administration. The volume of bile samples collected was dependent on the secretion rate of each animal and was determined gravimetrically. Following bile collection the catheter loop was reestablished. Processing of Biological Samples for Analysis To avoid degradation of the analytes, all biological samples (plasma, bile, and brain homogenates) were stored at 808C until pretreatment and analysis by high performance liquid chromatography with tandem mass spectrometric detection (LC-MS-MS). The pretreatment of the samples is extensively discussed elsewhere.19 Briefly, all samples were spiked with a solution mixture containing [Leu5]-Enk, an internal standard of the linear peptides, and CA-[Leu5]-Enk, an internal standard of the cyclic prodrugs, to correct for sample losses during sample cleanup. For brain homogenates, sample preparation included both protein precipitation with acetonitrile and solid phase extraction (SPE) using Waters Oasis HLB extraction cartridges (1 cc/30 mg, Waters) due to the complex nature of the tissue. For bile samples the concentration of the analytes had to be reduced by dilution but no further pretreatment was necessary. Plasma samples were treated with acetonitrile as the protein precipitation agent, followed by centrifugation. The supernatant was concentrated by solvent evaporation and reconstituted in 10% (v/v) acetonitrile. The solution was then centrifuged again and the supernatant removed for analysis. Pharmacokinetic parameters were obtained by fitting the plasma concentration-time data to a two-compartment model using SAAM II Version 1.2.1 software (Saam Institute, Inc., Seattle, WA). In Vitro Microsomal Stability Studies Pooled male Sprague Dawley rat liver microsomes were obtained from BD Biosciences (Woburn, MA) and pooled male Hartley albino guinea pig liver microsomes were obtained from Xenotech, LLC (Lenexa, KS).
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2680
LIEDERER ET AL.
For in vitro stability assessments in rat and guinea pig liver microsomes, DADLE, Ac-DADLENH2, and the cyclic prodrugs OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk (final conc. 2.5 mM) were each preincubated for 5 min at 378C (n ¼ 3) in 0.5 mL phosphate buffer (100 mM) solution at pH 7.4 with 1 mM b-NADPH in the presence and absence of 100 mM paraoxon, an esterase B inhibitor, and/or 5 mM ketoconazole, a CYP3A4 inhibitor. The enzymatic reaction was started by the addition of rat or guinea pig liver microsomes (approx. 400 nM cytochrome P450s) followed by 15 min of incubation at 378C. The reaction was quenched by adding 100 mL of methanol and 750 mL of acetonitrile to each sample. After vortex mixing, the solution was centrifuged at 13400 g for 10 min and 500 mL of supernatant was taken and evaporated (Centrivap concentrator, Labcono, Kansas City, MO). The residue was then reconstituted in 100 mL of 50% (v/v) acetonitrile and analyzed by LC-MS-MS. For control samples, 100 mL methanol was added to 0.5 mL phosphate buffer solution (100 mM, pH 7.4) containing 1 mM b-NADPH and 2.5 mM of the studied compound before the addition of liver microsomes. After quickly adding 750 mL acetonitrile, the solution was processed as outlined above. The percentage of analyte remaining was calculated based on the average analyte peak area divided by the average analyte peak area in the control sample. In vitro half-lives (t1/2) of the cyclic prodrugs were calculated from the disappearance rates in rat and guinea pig liver microsomes.25 ClInt, predicted and ClHep, predicted were calculated as follows:25 ClInt; predicted ¼
0:693 g of liver weight t1=2 ðminÞ kg of body weight mL incubation mg of microsomal protein 45 mg of microsomal protein g of liver weight
ClHep; predicted ¼ ðQ ClInt; predicted Þ=ðQ þ ClInt; predicted Þ
The hepatic blood flow (Q) used was 55.2 mL/ min/kg and 62.2 mL/min/kg for rat and guinea pig, respectively.26,27 Analytical Methods For in vitro and in vivo studies, LC-MS-MS was employed using a Quattro Micro triple quadrupole mass spectrometer (Micromass, Beverly, MA). The LC was conducted using a Waters 2690
HPLC system (Waters, Milford, MA). The analytical column, which was kept at 258C, was a Vydac (Hesperia, CA) C18 HPLC column (218TP5101, ˚ ,) with a guard column 50 1.0 mm i.d., 300 A (5 micron microbore cartridge, Vydac 218GK51). Sample separation was achieved utilizing a linear gradient from 5% to 100% B (Mobile phase: A, water with 0.1% FA (v/v); B, acetonitrile with 0.1% FA (v/v)), and a flow rate of 0.2 mL/min. The total run time was 12 min (5 min gradient 5% to 100% B; 2 min 100% B, 1 min reduction of 100% B to 5% B, 4 min equilibration with 5% B). The eluent from the first 2 min and the last 3 min was directed to waste. Samples and internal standards were recorded by multiple reaction monitoring (DADLE: 570.2 > 120.1; Ac-DADLENH2: 611.0 > 333.9; cyclic prodrugs: 728.3 > 120.1; [Leu5]-Enk (internal standard): 556.3 > 120.1; CA-[Leu5]-Enk (internal standard): 684.3 > 136.2). The intermediate(s) were indirectly determined as described elsewhere.19 The MS-MS conditions for the capped peptide Ac-DADLENH2 were: capillary voltage 4.0 kV; cone voltage 25 V, source temperature 1008C; desolvation gas flow 740 L/h; multiplier voltage 650 V; dwell time 0.5 s; inter-channel delay 0.03 s. The MS-MS conditions for the cyclic prodrugs were: capillary voltage 3.20 kV; cone voltage 30 V, source temperature 1208C; desolvation gas flow 740 L/h; multiplier voltage 650 V; dwell time 0.15 s; interchannel delay 0.03 s. Data acquisition and analysis were performed using Mass Lynx v3.5 and v4.0 software (Micromass). Further information on analytical methods developed in our laboratory for cyclic prodrugs can be found elsewhere.28
RESULTS In Vivo Animal Studies Pharmacokinetic Studies After i.v. administration of OMCA-[D-Ala2,Leu5]Enk and OMCA-DADLE in guinea pigs, the cyclic prodrugs disappeared rapidly from the systemic circulation following a two-phase exponential decay (Fig. 2A and B). Plasma concentrations could be detected for only up to 1 h. The respective intermediates and linear peptides (i.e., [D-Ala2, Leu5]-Enk and DADLE) generated by enzymatic and chemical breakdown of OMCA-[D-Ala2,Leu5]Enk and OMCA-DADLE were detected, however, at lower concentrations than the cyclic prodrugs.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
2681
Figure 2. Plasma concentration-time curves of OMCA-DADLE (squares) (A) and OMCA-[D-Ala2,Leu5]-Enk (squares) (B) and their respective intermediates (triangles) and linear peptides (circles) in guinea pigs (n ¼ 5–6) after i.v. administration (1 mg/kg) of each cyclic prodrug.
The plasma concentration time curves for OMCADADLE and OMCA-[D-Ala2,Leu5]-Enk are shown in Figure 2A and B, respectively. The plasma data were best described by a two-compartment model, and various pharmacokinetic parameters (AUC, clearance, Vss, k(0,1), t1/2) were estimated for the cyclic prodrugs (Tab. 1). The calculated parameters were similar and comparable to those obtained earlier in our laboratory from rat plasma data after i.v. administration of OMCA-DADLE to rats (Tab. 1).19 While the body clearance of OMCA-[D-Ala2,Leu5]-Enk (168 mL/min/kg) in guinea pigs and of OMCA-DADLE (81 mL/min/ kg) in rats exceeded hepatic blood flow the body clearance of OMCA-DADLE (61 mL/min/kg) in guinea pigs approximated the hepatic blood flow.
After i.v. administration of OMCA-[D-Ala2,Leu5]-Enk in rats the cyclic prodrug could not be quantified in plasma, because it was below the limits of quantitation of the assay (5 ng/mL). However, the conversion products (the intermediate(s)) and [D-Ala2,Leu5]-Enk) were observed indicating instantaneous breakdown of OMCA[D-Ala2,Leu5]-Enk (data not shown). It should be mentioned that due to the significant instability of OMCA-[D-Ala2,Leu5]-Enk in rat plasma additional degradation of the cyclic prodrug may have occurred during blood sample processing. Brain Permeation Studies Brain uptake of the cyclic prodrugs was determined 10 min after i.v. administration of
Table 1. Pharmacokinetic Parameters of OMCA-[D-Ala2,Leu5]-Enk after i.v. Administration (1 mg/kg) to Guinea Pigs and of OMCA-DADLE after i.v. Administration (1 mg/kg) to Guinea Pigs and Rats PK Parameters (Mean SE) AUC (ng min/mL) Clearance (mL/min/kg) Vss (mL/kg) k(0,1) (1/min) t1/2 (min)
OMCA- [D-Ala2,Leu5]-Enk (Guinea Pig) 7,570 2,080 168 29 437 178 0.3 0.1 2.7 0.4
OMCA- DADLE (Guinea Pig)
OMCA- DADLE (Rat)a
26,760 9,030 61 27 508 254 0.2 0.03 4.0 0.7
12,760 2,960 81 21 954 415 0.2 0.04 18 4
Pharmacokinetic parameters for OMCA-[D-Ala2,Leu5]-Enk in rats could not be determined because the cyclic prodrug could not be quantified in plasma, because it was below the limits of quantitation. a From Yang et al.19 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2682
LIEDERER ET AL.
Table 2. Brain Concentrations and Brain-to-plasma Concentration Ratios of OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk at 10 min after i.v. Administration (1 mg/kg) to Rats and Guinea Pigs Brain Uptake (ng/g tissue, mean SD, n ¼ 3)
Species b
Rat Guinea Pig Guinea Pig
Prodrug
Linear Peptidea
Intermediate
Brain-to-Plasma Prodrug Concentration Ratio
3.30 0.33 263.6 114.8 170.9 119.5
1.34 1.50 11.5 2.6 10.9 3.9
NDc NDc NDc
0.01 0.6 0.9
Compound OMCA-DADLE OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk
a
DADLE or [D-Ala2,Leu5]-Enk. From Yang et al.19. c ND, not determined because it was below the limit of quantitation. b
OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE in guinea pigs and compared to previous in vivo rat data from our laboratory (Tab. 2). For OMCADADLE, the brain levels of the cyclic prodrug and DADLE were approximately 80 and 8.5 times greater, respectively, than the levels observed in rats (Tab. 2). Administration of OMCA-[DAla2,Leu5]-Enk in guinea pigs lead to similar high levels of the prodrug and [D-Ala2,Leu5]-Enk in brain (Tab. 2). No intermediates from either cyclic prodrug were detected in brain because they were below the limits of detection. The brainto-plasma concentration ratio of the cyclic prodrugs was significantly greater in guinea pigs than in rats (Tab. 2).
respective intermediates and the linear peptides of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE were present in rat and guinea pig bile, the main excretion products were the cyclic prodrugs. In rats, rapid clearance of OMCA-[D-Ala2,Leu5]-Enk into the bile occurred after i.v. administration. Approximately 52% of the administered dose of OMCA-[D-Ala2,Leu5]-Enk was recovered within 2 h as the intact prodrug. Very similar results were observed earlier by Yang et al. after i.v. administration of OMCA-DADLE to rats.19 In contrast, only small amounts of the administered dose (<4%) of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE were recovered in bile after i.v. administration to guinea pigs.
Biliary Clearance Studies
In Vitro Microsomal Stability Studies 2
5
Biliary excretion of OMCA-[D-Ala ,Leu ]-Enk in rats and in guinea pigs as well as biliary excretion of OMCA-DADLE in guinea pigs was examined to elucidate its contribution to first-pass clearance. The resulting data are displayed in Table 3 and Figure 3. Although detectable levels of the
One possible explanation for the significant differences observed in the disposition of the cyclic prodrugs in rat and guinea pig would be differences in their metabolism. Therefore, stability studies were conducted using rat and guinea pig liver microsomes. These studies showed that,
Table 3. Bile Recovery over a 2 h Time Period of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE after i.v. Administration to Rats and Guinea Pigs Bile Recovery of Dose (%, mean SE, n ¼ 3–4)
Species b
Rat Rat Guinea Pig Guinea Pig
Compound
Prodrug
Linear Peptidea
Intermediate
OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk
38.1 2.1 51.7 5.8 3.3 0.4 0.82 0.1
3.3 0.4 6.0 0.7 0.11 0.01 0.05 0.01
7.3 0.7 6.0 0.7 00 00
a
DADLE or [D-Ala2,Leu5]-Enk. From Yang et al.19.
b
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
Figure 3. Cumulative biliary excretion of OMCA-[DAla2,Leu5]-Enk in rats (squares) and guinea pigs (circles) and of OMCA-DADLE in rats (diamonds)19 and guinea pigs (triangles) (n ¼ 3–4).
under the experimental conditions employed, the linear peptide DADLE was stable in rat and guinea pig liver microsomes. In contrast, the capped peptide Ac-DADLE-NH2 and the cyclic prodrugs OMCA-DADLE and OMCA-[D-Ala2, Leu5]-Enk were metabolized in this biological medium (Fig. 4A and B). For both the capped peptide and the cyclic prodrugs, the metabolism was more pronounced in guinea pig microsomes than in rat microsomes (Fig. 4A and B, Tab. 4). The capped peptide was noticeably more stable in rat and guinea pig liver microsomes than were the cyclic prodrugs. While paraoxon showed no significant effect on the stability of the capped peptide or the cyclic prodrugs in liver microsomes, ketoconazole substantially inhibited metabolism of Ac-DADLE-NH2 and OMCA-DADLE. Ketoconazole alone did not significantly inhibit the metabolism of OMCA-[D-Ala2,Leu5]-Enk in guinea pig liver microsomes. Interestingly, OMCA-[D-Ala2, Leu5]-Enk could be stabilized in guinea pig liver microsomes in the presence of both paraoxon and ketoconazole. In both species the predicted hepatic clearance was approximately equal to hepatic blood flow and the predicted intrinsic clearance exceeded hepatic blood flow (Tab. 4).
DISCUSSION Previous in vivo studies in rats showed high biliary clearance and low brain penetration of OMCA-DADLE.18,19 Additionally, in vitro studies
2683
demonstrated that while the conversion of this prodrug to DADLE in the target organ brain was too slow, its conversion in plasma was too fast.19 Therefore, prodrugs of diastereomers of DADLE such as OMCA-[D-Ala2,Leu5]-Enk were synthesized.22 These prodrugs exhibited more favorable prodrug conversion profiles than did OMCADADLE since their conversion in brain was now faster than in plasma.23 The in vitro studies also showed that guinea pigs resembled the enzymatic stability profiles (e.g., t1/2 in various biological media such as plasma) of humans better than did rats. Therefore, guinea pigs appeared to be a more appropriate animal model than rats for in vivo pharmacokinetic evaluations of cyclic prodrugs of opioid peptides. Based on these observations, OMCA-[D-Ala2,Leu5]-Enk and OMCADADLE were administered i.v. to guinea pigs. Additionally, OMCA-[D-Ala2,Leu5]-Enk was administered i.v. to rats for in vivo species comparison. After i.v. administration of OMCA-[D-Ala2, Leu5]-Enk and OMCA-DADLE in guinea pigs and rats, the cyclic prodrugs disappeared rapidly from the systemic circulation. The estimated pharmacokinetic parameters from the plasma data for OMCA-[D-Ala2,Leu5]-Enk and OMCADADLE in guinea pigs were similar (Tab. 1). In addition, the parameters of OMCA-DADLE in guinea pigs were comparable to those obtained in previous studies of OMCA-DADLE in rats (Tab. 1).19 More pronounced species differences were observed for OMCA-[D-Ala2,Leu5]-Enk. As mentioned earlier, in contrast to OMCA-[D-Ala2,Leu5]-Enk in guinea pigs, OMCA-[D-Ala2,Leu5]Enk in rats could not be quantified in plasma. Therefore, OMCA-[D-Ala2,Leu5]-Enk appeared to be significantly less stable in rats than in guinea pigs. This observation is consistent with previous in vitro data from our laboratory where the half-life of OMCA-[D-Ala2,Leu5]-Enk in rat plasma (t1/2 (-Paraoxon) < 2 min) was much shorter than in guinea pig plasma (t1/2 (-Paraoxon) 23 min).23 The body clearance of OMCA-[D-Ala2,Leu5]-Enk (168 mL/min/kg) in guinea pigs and of OMCADADLE in rats (81 mL/min/kg)19 exceeded hepatic blood flow in both species suggesting extrahepatic metabolism (e.g., in blood or kidney), nonmetabolic clearance processes, and/or extensive partitioning of a compound into red blood cells. Previous studies have shown that urinary excretion of cyclic prodrugs of DADLE in rats is minor and thus does not significantly contribute to total body clearance.19 In addition, cyclic prodrugs of opioid peptides appear to be chemically stable (e.g.,
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2684
LIEDERER ET AL.
Figure 4. Stability of DADLE, Ac-DADLE-NH2, and the cyclic prodrugs OMCADADLE and OMCA-[D-Ala2,Leu5]-Enk in rat (A) and guinea pig (B) liver microsomes at 378C in the absence of inhibitors (no shading), in the presence of paraoxon (complete shading), an esterase B inhibitor, in the presence of ketoconazole (diagonal shading), a CYP3A4 inhibitor, and in the presence of paraoxon and ketoconazole (vertical shading). The incubation time was 15 min. Data are presented as mean S.D. (n ¼ 3).
Table 4. In Vitro Half-Lives of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE in Rat and Guinea Pig Plasma and Microsomes at 378C and Predicted Clearance Values t1/2 (min, mean SD) Species Rat Rat Guinea Pig Guinea Pig
Compound OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk OMCA-DADLE OMCA-[D-Ala2,Leu5]-Enk
Plasmaa
Microsomes
10 1 <2 128 5 23 1
11 2 5.7 0.6 4.2 0.3 4.7 0.2
See Materials and Methods for experimental details and calculations. a From Liederer et al.23. JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
Cl (mL/min/kg, mean SD) ClInt,
predicted
499 67 986 98 1706 119 1518 70
ClHep,
predicted
50 1 52 1 60 1 60 1
DISPOSITION OF CYCLIC PRODRUGS
t1/2 (OMCA-DADLE) approximately 390 min and 290 min at pH 4 and pH 7, respectively).17However, previous in vitro stability studies of the cyclic prodrugs in rat and guinea pig biological media suggested that metabolism in plasma contributes to extrahepatic clearance for both cyclic prodrugs in rats and for OMCA-[D-Ala2,Leu5]-Enk in guinea pigs (Tab. 4). OMCA-DADLE was relatively stable in guinea pig plasma (Tab. 4, t1/2 &128 min). Thus, its contribution to extrahepatic clearance should be less significant. The predicted in vivo intrinsic clearance overestimated the values observed in vivo and thus showed poor in vitro–in vivo correlation. Brain uptake studies of OMCA-[D-Ala2,Leu5]Enk and OMCA-DADLE in guinea pigs showed accumulation of the cyclic prodrugs and their respective linear peptides in the brain. Interestingly, the observed brain levels and the brain-toplasma prodrug concentration ratio of these compounds were dramatically higher in guinea pigs than the levels and ratio for OMCA-DADLE in rats (Tab. 2). The limited ability of OMCA-DADLE to cross the rat BBB should not have been caused by plasma protein binding, because a large fraction of the drug (approx. 40%) was found to be unbound in rat plasma.19 In addition, even though OMCADADLE is less stable in rat than in guinea pig biological media (plasma, liver, and brain homogenate),23 which can impact brain uptake negatively because less prodrug would be able reach the brain, the differences did not appear significant enough to affect the brain uptake to the extent observed. In vivo studies revealed extensive bile excretion of OMCA-[D-Ala2,Leu5]-Enk and OMCA-DADLE in rats (Tab. 3, Fig. 3). Significantly lower hepatic clearance was observed for the linear peptides and the intermediates. The majority of all molecules (cyclic prodrugs, linear peptides, and intermediates) excreted over a 2 h time period were recovered rapidly (within approx. 20 min) after drug administration. In contrast to these findings in rats, bile excretion of the cyclic prodrugs, their respective intermediates and linear peptides in guinea pigs was minor. While less than approximately 4% of the total dose was recovered in guinea pig bile, more than approximately 40% was recovered in rat bile (Tab. 3). In an attempt to explain these dramatic species differences, stability experiments were conducted in rat and guinea pig liver microsomes. Both cyclic prodrugs were more actively metabolized in guinea pig liver microsomes (t1/2 (OMCA-DADLE)
2685
&4 min, t1/2 (OMCA-[D-Ala2,Leu5]-Enk) &5 min) than in rat liver microsomes (t1/2 (OMCA-DADLE) &11 min, t1/2 (OMCA-[D-Ala2,Leu5]-Enk) &6 min), illustrating species differences. Increased metabolism in guinea pig liver microsomes would negatively influence brain uptake because lower concentrations of the cyclic prodrugs would reach the brain. Therefore, these data did not explain the increased brain uptake in guinea pigs. However, the increased phase I metabolism in guinea pigs could at least partially account for the reduced bile clearance in guinea pigs versus rats. In both rats and guinea pigs the predicted in vitro hepatic clearance for OMCA-DADLE and OMCA-[D-Ala2,Leu5]-Enk approximately equaled hepatic blood flow and thus did not indicate any species differences but suggested extensive metabolism of the cyclic prodrugs in the liver in both species. Thus, it appears more reasonable that the observed species differences in brain uptake and bile excretion are mainly due to the different substrate specificities of efflux transporters that are expressed in rat and guinea pig BBB and liver. These transporter systems may have greater affinity for the cyclic prodrugs in rats than in guinea pigs. Thus, they are more effective in limiting partitioning of the compounds into the brain. As mentioned earlier, efflux transporters that limit permeation of molecules into the brain also facilitate liver clearance.8 A higher affinity of the cyclic prodrugs for efflux transporters could, therefore, not only reduce CNS entry but also enhance liver clearance or vice versa. Knockout animal models have been used to determine the impact of efflux mechanisms on the disposition of various compounds both in brain and liver.29–31 These studies have shown that the deficiency of efflux transporters can increase brain uptake and decrease hepatic clearance for some compounds and thus support our hypothesis. Studies using Caco-2 cells, an in vitro model of the intestinal mucosa, demonstrated that the cyclic prodrugs exhibited strong substrate activity for P-gp and MRP2.22 The identification of the efflux protein(s) involved in brain uptake and biliary clearance processes of these cyclic prodrugs remains to be examined but might involve the same protein(s). In an attempt to elucidate the role of the chemical linker and/or the induced solution conformations by the linker on the microsomal stability of the cyclic prodrugs, the stability of the linear peptide DADLE and the capped peptide AcDADLE-NH2 were determined in rat and guinea pig microsomes.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
2686
LIEDERER ET AL.
The data showed that DADLE was stable in contrast to the capped peptide and the cyclic prodrugs (Fig. 2A and B). The order of stability was DADLE > Ac-DADLE-NH2 > cyclic prodrugs. While paraoxon tended to show no significant effects on the microsomal stability of the cyclic prodrugs or the capped peptide, ketoconazole tended to inhibit the metabolism, suggesting that the capped peptides and cyclic prodrugs are substrates for cytochrome P450 enzymes (Fig. 4A and B). From these results, it appears that the cyclic prodrugs are substrates for phase I enzymes because of the chemical linker itself and/or the effect of the linker on the solution structure. Preliminary studies in our laboratory have identified, by LCMS-MS, oxidative metabolism of the aromatic residues (e.g., Phe hydroxylation) when the cyclic prodrugs were exposed to cytochrome P450s (unpublished data; Tarra Fuchs et al.). Thus it appears that the chemical linker is not the major metabolically labile site. Based on these data, we hypothesize that the chemical linker mediates oxidation reactions by inducing conformational constraints and a solution structure that is more favorable for interactions with cytochrome P450 enzymes than the solution structure present in the linear and capped molecule. In an attempt to block this observed oxidative metabolism, our laboratory is currently synthesizing a derivative of the coumarin-based cyclic prodrug of DADLE with appropriate chemical modifications in the peptide portion of the molecule.
CONCLUSIONS The pharmacokinetic characteristics of OMCADADLE and OMCA-[D-Ala2,Leu5]-Enk, particularly brain uptake and biliary excretion, were highly species-dependent. Therefore, it would be difficult to select an animal species for preclinical studies that would best predict the pharmacokinetic properties of these cyclic prodrugs in human. The observed species differences of the cyclic prodrugs were most likely due to the different substrate specificities of efflux transporters in rats and guinea pigs.
ACKNOWLEDGMENTS The authors thank Dr. Tarra Fuchs for helpful discussions and technical support and Dr. Michelle McIntosh for valuable help in the determination of the pharmacokinetic parameters of the cyclic
prodrugs. This research was supported by a grant from NIH/NIDA (DA-03315).
REFERENCES 1. Atkinson F, Cole S, Green C, van de Waterbeemd H. 2002. Lipophilicity and other parameters affecting brain penetration. Curr Med ChemCentral Nervous System Agents 2:229–240. 2. Veber DF, Johnson SR, Cheng HY, Smith BR, Ward KW, Kopple KD. 2002. Molecular properties that influence the oral bioavailability of drug candidates. J Med Chem 45:2615–2623. 3. Pauletti GM, Gangwar S, Knipp GT, Nerkurkar MM, Okumu FW, Tamura K, Siahaan TJ, Borchardt RT. 1996. Structural requirements for intestinal absorption of peptide drugs. J Control Release 41:3–17. 4. Wacher VJ, Silverman JA, Zhang Y, Benet LZ. 1998. Role of P-glycoprotein and cytochrome P450 3A in limiting oral absorption of peptides and peptidomimetics. J Pharm Sci 87:1322–1330. 5. Adessi C, Soto C. 2002. Converting a peptide into a drug: Strategies to improve stability and bioavailability. Curr Med Chem 9:963–978. 6. Humphrey MJ, Ringrose PS. 1986. Peptides and related drugs: A review of their absorption, metabolism, and excretion. Drug Metab Rev 17:283– 310. 7. Chandra P, Brouwer KL. 2004. The complexities of hepatic drug transport: Current knowledge and emerging concepts. Pharm Res 21:719–735. 8. Schinkel AH, Jonker JW. 2003. Mammalian drug efflux transporters of the ATP binding cassette (ABC) family: An overview. Adv Drug Deliv Rev 55:3–29. 9. Pardridge WM. 2003. Blood-brain barrier drug targeting: The future of brain drug development. Mol Interv 3:90–105. 10. Lien S, Lowman HB. 2003. Therapeutic peptides. Trends Biotechnol 21:556–562. 11. Gentilucci L. 2004. New trends in the development of opioid peptide analogues as advanced remedies for pain relief. Curr Top Med Chem 4: 19 –38. 12. Okada Y, Tsuda Y, Bryant SD, Lazarus LH. 2002. Endomorphins and related opioid peptides. In: Vitamins and hormones, Vol. 65. Elsevier Science, pp 257–279. 13. Ouyang H, Vander Velde DG, Borchardt RT, Siahaan TJ. 2002. Synthesis and conformational analysis of a coumarinic acid-based cyclic prodrug of an opioid peptide with modified sensitivity to esterase-catalyzed bioconversion. J Pept Res 59: 183–195. 14. Gangwar S, Pauletti GM, Siahaan TJ, Stella VJ, Borchardt RT. 1997. Synthesis of a novel esterasesensitive cyclic prodrug of a hexapeptide using an
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005
DISPOSITION OF CYCLIC PRODRUGS
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
(acyloxy)alkoxy promoiety. J Org Chem 62:1356– 1362. Wang B, Gangwar S, Pauletti GM, Siahaan TJ, Borchardt RT. 1997. Synthesis of a novel esterasesensitive cyclic prodrug system for peptides that utilize a ‘‘trimethyl lock’’-facilitated lactonization reaction. J Org Chem 62:1363–1367. Wang W, Camenisch G, Sane DC, Zhang H, Hugger E, Wheeler GL, Borchardt RT, Wang B. 2000. A coumarin-based prodrug strategy to improve the oral absorption of RGD peptidomimetics. J Control Release 65:245–251. Ouyang H, Tang F, Siahaan TJ, Borchardt RT. 2002. A modified coumarinic acid-based cyclic prodrug of an opioid peptide: Its enzymatic and chemical stability and cell permeation characteristics. Pharm Res 19:794–801. Chen W, Yang JZ, Andersen R, Nielsen LH, Borchardt RT. 2002. Evaluation of the permeation characteristics of a model opioid peptide, H-Tyr-DAla-Gly-Phe-D-Leu-OH (DADLE), and its cyclic prodrugs across the blood-brain barrier using an in situ perfused rat brain model. J Pharmacol Exp Ther 303:849–857. Yang JZ, Chen W, Borchardt RT. 2002. In vitro stability and in vivo pharmacokinetic studies of a model opioid peptide, H-Tyr-D-Ala-Gly-Phe-D-LeuOH (DADLE), and its cyclic prodrugs. J Pharmacol Exp Ther 303:840–848. Sun H, Dai H, Shaik N, Elmquist WF. 2003. Drug efflux transporters in the CNS. Adv Drug Deliv Rev 55:83–105. Chan LM, Lowes S, Hirst BH. 2004. The ABCs of drug transport in intestine and liver: Efflux proteins limiting drug absorption and bioavailability. Eur J Pharm Sci 21:25–51. Liederer BM, Fuchs T, Vander Velde D, Siahaan TJ, Borchardt RT. 2005. Effects of amino acid chirality and the chemical linker on the cell permeation characteristics of cyclic prodrugs of opioid peptides. Submitted. Liederer BM, Borchardt RT. 2005. Stability of oxymethyl-modified coumarinic acid cyclic prodrugs of diastereomeric opioid peptides in biological media from various animal species including human. J Pharm Sci 94:2198–2206. Wang B, Nimkar K, Wang W, Zhang H, Shan D, Gudmundsson O, Gangwar S, Siahaan T,
25.
26.
27.
28.
29.
30.
31.
2687
Borchardt RT. 1999. Synthesis and evaluation of the physicochemical properties of esterase-sensitive cyclic prodrugs of opioid peptides using coumarinic acid and phenylpropionic acid linkers. J Pept Res 53:370–382. Obach RS, Baxter JG, Liston TE, Silber BM, Jones BC, MacIntyre F, Rance DJ, Wastall P. 1997. The prediction of human pharmacokinetic parameters from preclinical and in vitro metabolism data. JPET 283:46–58. Boxenbaum H. 1980. Interspecies variation in liver weight, hepatic blood flow, and antipyrine intrinsic clearance: Extrapolation of data to benzodiazepines and phenytoin. J Pharmacokinet Biopharm 8:165– 176. Naritomi Y, Terashita S, Kimura S, Suzuki A, Kagayama A, Sugiyama Y. 2001. Prediction of human hepatic clearance from in vivo animal experiments and in vitro metabolic studies with liver microsomes from animals and humans. Drug Metab Dispos 29:1316–1324. Yang JZ, Bastian KC, Moore RD, Stobaugh JF, Borchardt RT. 2002. Quantitative analysis of a model opioid peptide and its cyclic prodrugs in rat plasma using high-performance liquid chromatography with fluorescence and tandem mass spectrometric detection. J Chromatogr B Analyt Technol Biomed Life Sci 780:269–281. Hoffmaster KA, Zamek-Gliszczynski MJ, Pollack GM, Brouwer KL. 2004. Hepatobiliary disposition of the metabolically stable opioid peptide [D-Pen2, D-Pen5]-enkephalin (DPDPE): Pharmacokinetic consequences of the interplay between multiple transport systems. J Pharmacol Exp Ther 311: 1203–1210. Wang JS, Ruan Y, Taylor RM, Donovan JL, Markowitz JS, DeVane CL. 2004. Brain penetration of methadone (R)- and (S)-enantiomers is greatly increased by P-glycoprotein deficiency in the blood-brain barrier of Abcb1a gene knockout mice. Psychopharmacology (Berl) 173:132– 138. Grauer MT, Uhr M. 2004. P-glycoprotein reduces the ability of amitriptyline metabolites to cross the blood brain barrier in mice after a 10-day administration of amitriptyline. J Psychopharmacol 18: 66–74.
JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 94, NO. 12, DECEMBER 2005