Chemicals, reagent grade for synthesis and analytical grade for biological studies, were from Sigma-Aldrich (St. Louis, MO, U.S.A), E. Merck KG (Darmstadt, Germany), Fluka Chemika (Buchs, Switzerland), Acros (NJ, U.S.A) and Wako (Richmond, VA, U.S.A) companies. Acid-washed alumina was purchased from RiedaL-de Haen Company (Spring Valley, CA, U.S.A.). Melting points were determined in Buchi (Flawil, Switzerland) 510 capillary melting point apparatus and were uncorrected. IR spectra were carried out on a Perkin-Elmer (Shelton, CT, U.S.A) 1760 FT-IR instrument. ¹H NMR spectra were determined on a Bruker (Wissem-bourg, France) 80 MHz or Bruker 400 MHz spectrometer with chemical shifts recorded in parts per million relative to tetramethylsilane. Mass and high-resolution mass (HRMS) were measured on Finnigan (San Jose, CA. U.S.A.) MAT 4510 and JEOL (Boston, MA, U.S.A.) JNS-D300 spectrometer respectively. Branson (Danbury, CT. U.S.A.) Sonifier 450 sonicator, Kubota (Tokyo, Japan) 2010 or Eppendorf AG (Hamberg, Germany) 5415C centrifuge Model 905 incubator (Cherng Huei Instrument Co., Tainan, Taiwan) and Ystral (Ballrechten-Dottingen, Germany) Laboratory series × 10/20 homogenizer were used in the preparation of intestinal mucosal suspension. Osmolarity of test solutions was determined with Wescor 5500 vapor pressure osmometer (Wescor Company, Logan, UT, U.S.A.). Male Wistar rats (300 - 350 g) from the Animal center of National Taiwan University were used in preparing intestinal mucosal suspension, BBMV and in perfusion studies. The same species of rats weighing 180 - 200 g were used in rotational behaviour studies. Male Sprague-Dawley rats (280 - 320 g) were used for determining brain dopamine. Animal studies were in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals.

The intestinal cross-membrane transport of d-phenylglycine-l-dopa and l-dopa was investigated using simulated intestinal brush-boarder membrane vesicle [23,24]. BBMV was prepared using magnesium precipitation method [25]. Protein content was determined. The purity of BBMV was determined by measuring the activity of the marker enzymes, alkaline phosphatase and aminopeptidase. Generally, these two enzymes were enriched 8 - 21 folds in the preparation. The activity of Na⁺, K⁺-ATPase, the marker enzyme of basolateral membrane, was very small. Normal function of BBMV was confirmed by measuring the uptake of d-glucose. In the presence of Na⁺ gradient ([Na⁺]_in < [Na⁺]_out), an overshoot phenomenon of glucose uptake with peak values of 9-11 times the equilibrium was routinely observed. The membrane vesicles were preloaded in the buffer solution containing 300 mM mannitol and 16 mM HEPES/Tris (pH 7.4) before the experiment. The uptake of test compounds in BBMV was measured by rapid filtration.

Mucosa suspension was prepared from the intestine of male Wistar rats according to established method [26] and was stored in an ice bath before use.

Literature procedure was followed for the preparation of perfusion solutions and the jejunal segments [27]. To maximize the absorption and to prevent the test compounds from being oxidized during perfusion, the experiments were performed at pH 6.0 with 0.02% (w/v) ascorbic acid added as antioxidant and nitrogen gas was bubbled through for 10 min before each experiment. Perfusion solution was pumped through the jejunal segment at a flow rate of 0.2 ml/min by a syringe pump (Stoelting, KD Scientific, U.S.A.). The jejunal segment was pre-washed with drug-free buffer for 10 min before the drug solution was pumped in. Outlet tubing samples were collected every 10 min for 6 collection periods after water and solute transport reached steady-state. The dimensionless membrane permeability Pm* [28] was measured as indications for the disappearance rate of test compound from the intestine. Plasma samples were withdrawn from carotid artery.

Rats were fasted for at least 18 h prior to drug administration. Anaesthesia was induced by i.p. injection of urethane (0.15 g/100 g body weight). The rats were put under a heating lamp to maintain body temperature.

The HPLC system used in the assay of biological samples consisted of an autosampler (AS950, Jasco, Tokyo, Japan), a Waters Model 600E solvent delivery pump (Millipore, Milford, MA, USA), a Model LC-4C electrochemical detector with a glassy-carbon electrode (Bioanalytical Systems, Inc., West Lafayette, IN, USA), and an integrator (Macintosh LC II with Macintegrator I). A Nucleosil^® 10 SA cationic ion-exchange column (10 μm, 300 × 4.0 mm, Macherey-Nagel, Düren, Germany) with a mobile phase comprising NaCl (50 mM) and Na₂-EDTA (1.0 mM) in 0.1 M ammonium phosphate buffer (pH 2.0) at a flow rate of 2.0 ml/min was used for the elution of the samples. The detection limits of d-phenylglycine-l-dopa and l-dopa were 50 ng/ml and 25 ng/ml, respectively. HPLC assay methods were validated by determining the precision and accuracy of intra-day and inter-day analysis of serum standards over a period of 6 days. The coefficients of variation for inter- and intraday assays were less than 15% for both compounds (n = 6).

According to the literature [29,30], the area under the plasma concentration-time profile (AUC) was calculated by log-linear trapezoidal rule. Plasma concentration after i.v. administration of drugs were also fitted to a non-compartment model using PCNONLIN and Akeike's Information Criteria, sum of squared residuals, residual plot and correlation coefficient were use for determination of the compartment model. The residual area after the last observed data point was calculated as C_last/k, where C_last is the last observed concentration, and k is the corresponding terminal rate constant. Terminal half-life (t_1/2) was estimated compartment model-independently. The fraction of absorption was calculated according to Equation 1.

(1) BA = AUC oral ⋅ k oral dose oral AUC iv ⋅ k iv dose iv × 100 %

Single dose d-phenylglycine-l-dopa (50 mg/kg in 2.5 mL of normal saline) was administered i. v. via femoral vein to anesthetized male Sprague-Dawley rats (280 - 320 g). The body temperature of the rats was maintained at 37°C with a heating pad throughout the experiment. The rat was immobilized in a stereotaxic frame (David Kopf Instruments, Tujunga, CA, USA), the skull was surgically exposed, and a hole was drilled with a trephine into the skull based on stereotaxic coordinates. The brain microdialysis system consisted of a CMA/100 microinjection pump (CMA, Stockholm, Sweden) and a microdialysis probe. The dialysis probes (3 mm in length) were made of silica capillary in a concentric design with their tips covered by dialysis membrane (Spectrum, 150 μm outer diameter with a cut-off at nominal molecular mass of 13000, Laguna Hills, CA, USA). The probe was placed into right striatum (0.2 mm anterior to bregma and 3.2 mm lateral to midline) and perfused with Ringer's solution (147 mM Na+; 2.2 mM Ca++; 4 mM K+; pH 7.0) at a flow-rate of 1 μl min^-1. The position of each probe was verified at the end of experiments. The dialysates was collected at 10 min intervals and aliquots of 10 μl was assayed by microbore HPLC.

The HPLC system consisted of a pump (BAS PM-80, West Lafayette, IN, USA) and an on-line injector (CMA 160, Stockholm, Sweden) equipped with a 10 μl sample loop, a reversed phase C18 microbore column (particle size 5 μm, 150 × 1 mm I.D.; Bioanalytical Systems, West Lafayette, IN, USA) and an EC detector (BAS-4C amperometric) coupled to a glassy carbon working electrode and referenced to a Ag/AgCl electrode at 750 mV with a range set at 50 nanoamper. Output data from the detector were integrated via an EZChrom chromatographic data system (Scientific Software, San Ramon, CA, USA). The mobile phase for analyzing striatal dopamine, eluted at a flow rate of 0.05 ml/min, comprised 80 ml acetonitrile, 2.2 mM sodium 1-octanesulfonate, 14.7 mM monosodium dihydrogen orthophosphate, 30 mM sodium citrate, 0.027 mM EDTA, and 1 ml diethylamine in one liter double distilled water, adjusted to pH 3.5 by orthosphoric acid (85%). The elute was filtered through a Millipore 0.22 μm filter and degassed prior to use.

Male Wistar rats (180 - 200 g) were anesthetized with pentobarbital sodium (30 mg/kg body weight, i.p.) and the heads were fixed in a DaviD-Kopf steric taxic frame. A solution of 6-hydroxydopamine (6-OHDA, 2.00 mg/ml × 8 ml) in saline was infused using Paxinos and Watson coordinates (AP 5.3, L 2.0, H 7.8 mm, [34]) into the unilateral substantia nigra compacta (SNc) of brain with a syringe pump through a 30 gauge stainless steel needle at a flow rate of 2 μl/min. After two weeks of recovery period, the 6-OHDA treated rats were placed in a spherical bowl (radius 20 cm) and secured by a thoracic harness which was connected to a 486 PC computer for automatic recording of rotation induced by (+)-methamphetamine (MA). The rotational behavior of rats was recorded 10 min after MA treatment (MA in saline, 4.00 body weight mg/kg of rat, s.c.). The numbers of turns recorded were defined as the control value (T₀) for each individual animal. Only animals showing a T₀ greater than 400 were chosen for further experiments. After two weeks of a wash-out period the animals were subjected to drug treatment. Single dose (0.051 mmol) of each test compound was administered orally to rats 5 min prior to MA treatment (4.00 mg/kg body weight, s.c.). The rotation counted for a period of 110 min starting 10 min after MA treatment was recorded as T_d for each tested rat. The percentage of reduction in rotation for each animal was calculated and presented as (T_d - T₀)/T₀ × 100%.

Data analysis were performed on Visual dBase and SPSS/PC+ and were represented as mean ± SE for n experiments. Treatment differences were evaluated by paired-t test.

The uptake of d-phenylglycine-l-dopa in BBMV was measured. Amino acid l-Phe or l-dopa, dipeptide l-Gly-l-Pro or l-Gly-l-Phe, or cephradine was added for investigating the competition with d-phenylglycine-l-dopa in BBMV uptake (Figure 2).

The stability of d-phenylglycine-l-dopa in the intestine was determined prior to the intestinal absorption studies. In order to simulate intestinal microclimate pH, the compound was incubated with the intestinal mucosal suspension in a pH 6.5 isotonic buffer solution. l-Gly-l-Phe comprising essential amino acids degraded rapidly with only 50% of recovery after 2 min of incubation. d-Phenylglycine-l-dopa, on the other hand, was very stable with almost 100% of recovery after 90 min of incubation (Figure 3).

The absorption of d-phenylglycine-l-dopa and l-dopa was compared in rats by in situ single-pass jejunal perfusion experiments. Amidon's dimensionless cross-membrane permeability (P_m*) was determined as a parameter of intestinal absorption [28]. The steady-state plasma concentration was also determined (Table 1).

The mean plasma concentration-time profiles after single dose oral and i.v. administration of d-phenylglycine-l-dopa and l-dopa are depicted in Figure 4. The pharmacokinetic parameters calculated with the data of plasma concentration-time curves based on the non-compartmental model analysis were summarized in Table 2. The fraction of oral absorption (BA) was calculated according to Equation 1. The Striatal dopamine level after i.v. injection of d-phenylglycine-l-dopa (50 mg/kg) is depicted in Figure 5.

The in vivo anti-Parkinsonism effect was determined with conventional rotation model measured in 6-OHDA-treated unilateral striatal-lesioned rats elicited rotation with (+)-methamphetamine (MA) [32,35]. As shown in Table 3, d-phenylglycine-l-dopa as well as l-dopa demonstrated inhibition of MA-induced rotation of rats. With equal molar of test compound administered, the activity of d-phenylglycine-l-dopa in reducing the rotation of rats was significantly higher than that of l-dopa.