The route and mode of elimination of some drugs such as paracetamol has been studied in the saliva of type 1 and type 2 diabetic patients [1] and diflunisal in animal model of diabetes [2]. In most of the reports on drug elimination, the controversial issues are the kinetics, route of elimination, metabolite formation and the mechanism by which they are eliminated.

Oxidative N-demethylation is one of the basic metabolic reactions of both xenobiotics and endogenous compounds of vertebrates [3-5]. Removal of the methyl group decreases the lipophilicity of the compounds, thereby facilitating their excretion. A classical model of metabolic demethylation is methylamine which is generated endogenously [5]. The sources of methylamine include adrenaline, creatinine and sarcosine [5].

Metabolic studies on drugs and food components have dealt mainly with the identification and quantitative determination of the N-demethylated compounds [3-6]. Much less, if any attention has been focused on the fate of the molecular fragment derived from the methyl group. However, it is well known that formaldehyde is produced by N-demethylation and has been widely cited by many textbooks of pharmacology. The generated formaldehyde was easily identified when the N-methyl group was ¹⁴C radiolabelled; and certain N-demethylation reactions occurred through the homocystein-methionine cycle resulting in methylation of an endogenous compound [5].

Recent publications stated that formaldehyde, formed in situ, can directly take part in methylation reactions [5, 7-9]. Tyihák E, et al. [7] and Trezli et al[8] reported a spontaneous methylation of N-amino group of lysine by formaldehyde. The application of planar chromatography, especially two-dimensional thin-layer chromatography has given an easy and reliable tool to locate the formaldehyde after pre-chromatography reaction with dimedone [10-13]. Although, formaldehyde was eliminated in urine, further reaction also took place in which a basic amino acid (lysine) is the major target. The methylation of lysine results in N^ε-monomethyl-lysine, which was identified using thin-layer chromatography and high-performance liquid chromatography combined with mass spectrometry detection (HPLC-MS) [14-17]. Direct verification of the N-methyl transfer from the drug [(-)-deprenyl] to lysine was also done using ¹⁴C-methyl labeled deprenyl, and one part of the radioactivity was detected on the N-monomethyllysine.

Studies by Yu et al. [18,19] detailed the significance of semicarbazide-sensitive amine oxidase (SSAO) in the generation of formaldehyde from exogenous and endogenous compounds as well as the level of SSAO in Type 2 diabetes mellitus. Other investigators reported the association of serum monoamine oxidase level and retinopathy in diabetes mellitus [20].

The present report examined the effect of diabetes mellitus on the urinary excretion of radiolabelled deprenyl [(R)(R)].

A JASCO system (JASCO Corporation, Tokyo, Japan) consisting of a DG-208054 degasser, PU-1580 pump, AS-2057 plus automatic sample injector, and UV-1575 detector at 265 nm were used. Chromatographic data were stored and evaluated using the SRI Model 202 Peak Simple Chromatographic Data System (SRI Instruments, Torrance, CA, USA). The standard N-Monomethyl- lysine HCl and urine samples were separated after derivatization with Fmoc hydrochloride. HPLC was carried out using a 25 cm x 4.6 mm I.D. column packed with 6 (m Kovasil C18 endcapped particles (Chemie Uetikon, Uetikon, Switzerland) to separate the radioactive (¹⁴C-labelled monomethyllysine) peak of the urine using acetonitrile—water (2:1), which also contain 0.1% formic acid as the mobile phase. The HPLC column was kept at 28 ºC. The mobile phase flow rate was 1.5 mL min^-1, and each fraction was collected for 1 min.

Rat urine was collected following oral administration of ¹⁴C-(methyl)-labeled (-)-deprenyl, and the cumulative excretion of radioactivity was determined. Urinary elimination of radiolabelled-deprenyl was significantly (p < 0.05) faster in the first 8 hours after oral administration of ¹⁴C-(methyl)-labeled (-)-deprenyl, when compared to controls (Fig. 1). STZ-induced diabetes facilitated urinary elimination of the radioactivity. The rat urine (containing radiolabelled-dep-renyl and its metabolites) was subjected to two-dimensional thin-layer chromatography after reacting it with dimedon, and radioactive spots were located using an X-ray film (Fig. 2). The spots were scarped, and their ratios were compared. Fig. (3) showed the major metabolites of deprenyl. The ratio of methamphetamine, p-hydroxy-metham-phetamine, deprenyl and formaldehyde were 20:3:1:1 in the urine collected for 8 hours (Table 1). Urine eliminated radioactivity in N-monomethyllysine was determined using HPLC, its amount was about 10% of the free urinary formaldehyde.

Fig. (1). The time course of urinary excretion of radioactivity when the N-methyl group of (-)-deprenyl was 14C-labelled.

Fig. (2). Thin-layer chromatography of radiolabelled deprenyl metabolites.

Fig. (3). Major metabolic routes and metabolites of (-)-deprenyl. (The number in bracket corresponds to those in Table 1)

Table 1.

The Name,Chemical Structure, Presence of Radioactivity and logP Value of Deprenyl and its Major Metabolites

There are two major metabolites of (-)-deprenyl. The detection of these metabolites depends on the analytical method applied. Reynolds et al. [21] detected methamphetamine about 30 years ago, as the major deprenyl metabolite in man. They used GC-MS (gas chromatography combined with mass spectrometry), so their method did not make it possible to find the other major deprenyl metabolite, deprenyl-N-oxide that was later identified by Katagi et al. [22]. Unfortunately, deprenyl-N-oxide cannot be detected using thin-layer chromatography either. Other deprenyl metabolites include para-hydroxy-methamphetamine, nordeprenyl (also called desmethyldeprenyl), amphetamine, and phenylacetone. However, the radiolabelled N-methyl group is present only in para-hydroxy-methamphetamine. Moreover, Szatmári described 25 deprenyl metabolites with aromatic ring, as well as two more fragments containing the propargyl (N-methyl-propargylamine and propargylamine) as well as propiolaldehyde [23]. If we also calculate the conjugates of several deprenyl metabolites (mainly that of p-hydroxy-metabolites), the gross number of metabolic products can approach 40.

It is peculiar that previous studies on deprenyl metabolism neglected an evident metabolite of deprenyl, the formaldehyde. In fact, most of the basic textbooks on pharmacology show that formaldehyde is a byproduct of R-X-CH₃ splitting, where the R-X-CH₃ R-X-H + CH₂O reaction can be either N-demethylation, or O-demethylation or S-demethy-lation [3, 4, 6].

Our earlier papers have shown that ¹⁴C-labelled deprenyl is generated from ¹⁴C-labelled (-)-deprenyl which is eliminated in urine. Moreover, one of the radiolabelled peaks of the same rat urine was ¹⁴C-labelled-N^ε-monomethyl- lysine. Identification of N^ε-monomethyllysine in the urine as well as in isolated homogenous peak with radioactivity was done by the use of two different HPLC-MS techniques that used electrospray (ES) ionization [24] as well as atmospheric pressure chemical ionization (APCI) modes [17]. It is also worth noting that the generated compounds were present in the urine only in traces (formaldehyde, about 1 to 2% of radioactivity administered, and N^ε-monomethyllysine, about 0.1%).

The responsible enzyme of oxidative demethylation (or one of the responsible enzymes) is semicarbazide-sensitive amine oxidase (SSAO; EC.1.4.3.6), which shows an elevated level in the blood of diabetic patients [25, 26].

In contrast, hepatic cytochrome P450 activity of aminopyrine N-demethylase and ethoxyresorufin O-deethylase enzymes showed changes in opposite directions after the onset of STZ-induced diabetes. Aminopyrine N-demethylase activity significantly decreased [88+0.16 (control)2.88+0.16 (control)+0.46 (after treatment with 50 mg/kg STZ> 2.19+0.46 (after treatment with 50 mg/kg STZ p≤0.001) and 1.53+0.22 (after treatment with 70 mg/kg STZ p≤0.001) and 1.53+0.22 (after treatment with 70 mg/kg STZ p≤0.001)> p≤0.001)], while ethoxyresorufin O-deethylase activity increased [R49.6+5.0 (control)>49.6+5.0 (control)+12.6 (after treatment with 50 mg/kg STZ> 73.3+12.6 (after treatment with 50 mg/kg STZ p≤0.01) and 82.6+14.9 (after treatment with 70 mg/kg STZ> p≤0.01) and 82.6+14.9 (after treatment with 70 mg/kg STZ p≤0.01)> p≤0.01)].

Watanabe et al. [27] reported that the change in pharmacokinetics after STZ-induced diabetes t_max became shorter, however, C_max and the area under the time versus concentration curve (AUC) became higher in the case of diabetic rats compared to control rats. Watanabe et al. [27] also compared the parameters of pharmacokinetics in the spontaneously diabetic (GK-rats, NIDDM model) and Zucker rats (almost normal glycemic but abnormally glucose tolerant, hyperinsulinemic, increasingly obese, insulin-resistant rats). These rats showed even shorter t_½, but higher C_max and AUC parameters.

Urinary elimination of (-)-deprenyl and its metabolites was faster in STZ-induced diabetic rats compared to control. As the metabolites and their ration was not changed, more frequent treatment with lower dose, or the use of sustained release preparations can counterbalance the faster elimination of deprenyl and possibly other drugs used in diabetes.