Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Tolperisone, a centrally acting muscle relaxant (Pórszász et al., 1961), has been used for the treatment of painful muscle spasms in orthopaedic and rheumatologic diseases in several decades. Ono and colleagues have shown that tolperisone exerts a local anesthetic-like action both on motoneurons and primary afferents in vivo, as well as on peripheral nerves of rats in vitro (Ono et al., 1984). The effect of tolperisone is similar to that of lidocaine, which is known to inhibit voltage-dependent sodium currents (Oortgiesen et al., 1990). In a recent study it was shown that tolperisone was effective in patients suffering from painful reflex muscle spasms (Pratzel et al., 1996).

Although a few papers are available on the pharmacokinetics of tolperisone, no study has been performed to date on the in vitro metabolism of the compound. In early in vivo studies the drug concentration in the plasma was measured using gas chromatography and by mass fragmentographic techniques (Miyazaki et al., 1972, 1975). Miskolczi and coworkers in 1987 published a new and sensitive gas-liquid chromatographic method for the determination of tolperisone in human plasma and made a comparison of two different brands of tolperisone tablets, Mydeton and Mydocalm. In that study the elimination half-life of tolperisone was 1.55 ± 0.7 h, and the total body clearance was 140.8 ± 33.8 l/h. The oral bioavailability of the two different tolperisone products was 22.3 ± 6% and 16.7 ± 8.9%, respectively.

Metabolic profiling is a useful tool to predict the pharmacokinetic properties of drugs and drug candidates and to assess the possibility of drug-drug interaction risks. The genetic background and actual metabolic state of individuals can vary largely even in the same population group. Therefore, these differences can influence significantly the fate of ingested drugs (i.e., the AUC¹ can differ largely in different individuals). On one hand, the elevated plasma level of a drug can cause toxicity problems, especially those with narrow therapeutic range. On the other hand, high metabolic activity in certain population groups (extensive metabolizers) may prevent therapeutic efficacy of drugs. The goal of the present study was to identify by in vitro reaction phenotyping methods the microsomal enzymes that contribute to the biotransformation of tolperisone and to elucidate the metabolic profile of the compound.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. Tolperisone-hydrochloride and metabolite standards were synthesized in Gedeon Richter Ltd. (Budapest, Hungary). Quinidine was purchased from Merck (Darmstadt, Germany); ketoconazole, sulfaphenazole, 6-hydroxytestosterone, SKF-525A (proadifen-hydrochloride), glucose 6-phosphate, glucose-6-phosphate dehydrogenase, phenacetin, 4-acetamidophenol and 1-aminobenzotriazole were purchased from Sigma-Aldrich (St. Louis, MO). Furafylline and S-mephenytoin were from Ultrafine Chemicals Ltd. (Manchester, UK). Testosterone was from Fluka (Buchs, Switzerland); tolbutamide was obtained from Sigma/RBI (Natick, MA). Methyl p-tolyl sulfide and sulfoxide were from Aldrich Chemical Co. (Milwaukee, WI). 1-benzylimidazole was purchased from Sigma-Aldrich Chemie GmbH (Steinheim, Germany). Na-pyrophosphate, MgCl₂, NADPH-sodium salt was obtained from Reanal (Budapest, Hungary).

Incubation Conditions. All incubations were carried out in incubation mixtures containing 6 mM Na-pyrophosphate, 5 mM MgCl₂, 5 mM glucose 6-phosphate, 1 U/ml glucose-6-phosphate dehydrogenase in a final volume of 0.5 ml of 0.1 M Tris-HCl buffer, pH 7.4. Tolperisone concentration was 1 and 5 µM (Cl_int determination and inhibition experiments), or 0.5-1000 µM (enzyme kinetic measurements). In HLM incubations, the mixture contained 1 mg/ml of microsome pool and the reactions were initiated by the addition of 5 mM NADPH and conducted in shaking water bath at 37°C for 20 min. In incubations containing insect cell-expressed recombinant human P450s, the enzyme concentration was 20 pmol/ml (CYP1A2, CYP2A6, CYP2B6, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4) or 1000 pmol/ml (FMO3). The reactions were initiated by the addition of microsomes and incubated in water bath at 37°C, without shaking, according to the manufacturer's recommendation. Microsomes prepared from wild-type baculovirus-infected cells were used as control. The reaction was stopped with equal volume of ice-cold methanol, centrifuged for 10 min at 7500g, and 10 µl of supernatant was injected into HPLC.

Inhibition Studies with Chemical Inhibitors. P450 isoform-specific chemical inhibitors or index substrates were used to inhibit tolperisone metabolism in HLM. Furafylline for CYP1A2 (Sesardic et al., 1990; Kunze and Trager, 1993; Tassaneeyakul et al., 1994; Newton et al., 1995), sulfaphenazole for CYP2C9 (Baldwin et al., 1995; Newton et al., 1995), quinidine for CYP2D6 (Newton et al., 1995), ketoconazole for CYP3A4 (Baldwin et al., 1995; Newton et al., 1995; Sai et al., 2000), and S-mephenytoin for CYP2C19 (Goldstein et al., 1994) and CYP2B6 (Küpfer and Preisig, 1984) were used in the concentration range of 1 to 100 µM (quinidine and ketoconazole), 2 to 200 µM (furafylline, sulfaphenazol, S-mephenytoin). Thiourea at 10 mM was used as a competitive inhibitor of FMO3 (Poulsen et al., 1979). SKF-525A (100-1000 µM; Schenkman et al., 1972), 1-aminobenzotriazole (ABT) (1000 µM; Ortiz de Montellano et al., 1984; Xu et al., 1994) and 1-benzylimidazole (BI) (1000 µM, Madan et al., 1993; Grothusen et al., 1996) were used as nonspecific P450 inhibitors to assess the P450-dependent metabolism of tolperisone. All inhibitors were dissolved in methanol and coincubated with tolperisone in the incubation mixture except for furafylline and ABT, which are mechanism-based inhibitors; therefore, those were preincubated with HLM in the presence of NADPH-regenerating system for 10 and 20 min, respectively. In these experiments, the reaction was initiated with tolperisone dissolved in methanol. The methanol concentration in the incubation mixture did not exceed 1% (v/v).

Inhibition of Index Reactions. Tolperisone was used as inhibitor in the concentration range of 0 to 90 µMexcept for dextromethorphan O-demethylase and MpTS oxidase, where tolperisone was 0 to 45 µM and 0 to 1500 µM, respectivelyand was coincubated with index substrates. The incubation mixture, quenching, and processing of samples for HPLC measurements were the same as described above. ¹⁴C-dextromethorphan O-demethylase assay was carried out as previously described (Rodrigues, 1996). Bufuralol hydroxylase assay described by Boobis et al. (1985) was also used as an index reaction of CYP2D6 with modifications. Briefly, 5 to 200 µM of bufuralol was coincubated with 10 to 90 µM tolperisone for 40 min in shaking water bath at 37°C in a total volume of 500 µl. Testosterone 6-hydroxylase assay was the modification of the method described by Anderson et al. (1995). Testosterone was used in a concentration range of 20 to 500 µM, and the mixture was incubated for 30 min. Tolbutamide hydroxylase was carried out as previously described (Miners et al., 1988) without extraction. Tolbutamide was used in 20 to 500 µM. Phenacetin O-deethylase was carried out as described previously (von Moltke et al., 1996), with slight modifications in the composition of incubation mixture and detection wavelength (249 nm). As index reaction of FMO3, methyl p-tolyl-sulfide (MpTS) oxidation was carried out (Pike et al., 1999). Since this reaction is known to be mediated not only by flavin-containing monooxygenases, insect cell-expressed human FMO3 was used. MpTS was added in the concentration range of 10 to 1500 µM. Incubations were performed for 15 min. The produced methyl p-tolyl sulfoxide (MpTSO) was monitored with HPLC.

Inhibition of Tolperisone Metabolism with Antibodies. Before addition to the mixture, pooled HLM were preincubated with the antibodies in triplicates either for 15 min on ice (CYP2D6- and CYP3A4-specific antibodies; 6 and 10 µl of antibodies/100 µg of microsomes, respectively) or for 30 min at room temperature (CYP2C8/9/19-specific and NADPH-P450 reductase-specific antisera; 20 µl and 7.5 µl of antisera/100 µg microsomes, respectively). Anti-reductase inhibition kit contained nonimmune goat serum as control. In experiments with CYP2C8/9/19-specific antiserum, normal goat serum from Invitrogen New Zealand Ltd. (Auckland, NZ) was used as control.

High Performance Liquid Chromatography. Analytical measurements were carried out using a Merck-Hitachi LaChrom HPLC system equipped with UV detector. Discovery C₁₈ 150 × 4.6 mm (5 µm) column with Supelguard 20 × 4 mm (5 µm) column (Supelco, Bellefonte, PA) was maintained at 40°C for all assays.

LC/MS Measurements. Merck-Hitachi LaChrom HPLC system was coupled with mass spectrometer (Merck-Hitachi M-8000) with electrospray ionization interface. The enzyme reaction was stopped with equal volume of ice-cold methanol. The samples were kept at 20°C for 30 min. After centrifugation, the supernatant was collected, and the pH was set to 9 and extracted with chlorobutane. Organic phase was evaporated to dryness and redissolved in ethanol, and 10 µl was subjected to HPLC. Eluent A consisted of 140 ml of methanol and 100 ml of water, supplemented with 100 µl 10% formic acid and 200 µl 10% ammonium hydroxide. Eluent B consisted of methanol. Gradient: 0. min: 100% A, 8. min: 60% A, 12. min: 100% A. Flow rate was 0.25 ml/min. Samples were monitored at 255 nm with UV detector. Prontosil C₁₈ ACE EPS 5 µm, 100-2 mm column (Bischoff, Germany) was used at 35°C.

Data Analysis. Enzyme kinetic data of tolperisone hydroxylation in pooled HLM and recombinant isozymes were fitted to standard Michaelis-Menten plot (reaction rate "V" against substrate concentration "S") and Eadie-Hofstee plot ("V/S" against "V"). Kinetic parameters (apparent K_m and V_max) were calculated using nonlinear regression analysis of GraphPad Prism 2.01 (GraphPad Software Inc., San Diego, CA). Standard deviation and p values of Student's t test were calculated with Microsoft Excel (Redmond, WA). For determination of K_i, "K_m,app" method was used (Kakkar et al., 1999). Cl_int in pooled HLM was calculated from initial reaction rate using the following equation:
where dc/dt is the initial reaction rate and c₀ is the initial concentration of the substrate.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Intrinsic Clearance of Tolperisone. The in vitro Cl_int of tolperisone was determined using individual microsomes from nine donors (three female and six male). Tolperisone had a Cl_int of 1.27 ± 0.16 ml/min/g of liver, determined as the mean of nine individual clearance values ± S.D. The female and the male pool resulted in 1.39 ± 0.22 and 1.15 ± 0.20 ml/min/g of liver, respectively (mean of three and six individual samples ± S.D., respectively). The sex difference in Cl_int of the compound was not statistically significant. The metabolism was NADPH-dependent, since the omission of this coenzyme completely abolished the consumption of tolperisone (data not shown).

Isoform-Specific Chemical Inhibitors/Index Substrates. P450 isoform-specific chemical inhibitors or index substrates (i.e., competitive inhibitors) were used to inhibit the consumption of tolperisone (5 µM) and the formation of M1 metabolite in pooled HLM. Furafylline (CYP1A2), sulfaphenazole (CYP2C9), S-mephenytoin (CYP2C19), quinidine (CYP2D6), and ketoconazole (CYP3A4) were tested as inhibitors. The summarized results are shown in Table 1. The only compound found to substantially inhibit tolperisone consumption and M1 formation was quinidine at a diagnostic 10 µM concentration. Ketoconazole at 10 µM inhibited the loss of parent compound and M1 formation to a lesser extent. S-mephenytoin and sulfaphenazole resulted in slight and concentration independent (data not shown) inhibition both in consumption and M1 formation. Thiourea was also used to assess the contribution of FMO3 to the metabolism of tolperisone. No inhibition was seen at 10 mM (data not shown).

Inhibitory Antibodies. Human CYP2D6 and human CYP3A4-specific monoclonal and human CYP2C8/9/19-specific polyclonal antibodies were used to test the contribution of respective P450 enzymes to the in vitro metabolism of tolperisone (5 and 50 µM) in pooled human liver microsomes. CYP2D6- and CYP2C-specific antibodies inhibited the clearance of tolperisone and M1 formation (Table 2); antibodies against CYP3A4 had no effect (data not shown). Anti-CYP2C-inhibited tolperisone clearance more effectively when 50 µM substrate was used, whereas the inhibitory effect of anti-CYP2D6 decreased. The same pattern was seen in M1 formation, the 70% inhibition of anti-CYP2D6 at 5 µM decreased to about 50% at 50 µM. These results point toward the decreasing relative contribution of CYP2D6 along with increasing tolperisone concentration. Although CYP2C-specific antibodies resulted in significant inhibition, these results were difficult to interpret since nonimmune goat serum had an intrinsic, yet unidentified metabolizing capacity.

Recombinant Enzymes. Insect cell-expressed recombinant human P450s (CYP1A2, CYP2A6, CYP2B6, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4) were used to study the contribution of these enzymes to the metabolism of tolperisone. The reaction mixture contained 20 pmol/ml of each P450 and 1 µM tolperisone. Under these conditions, CYP2D6 and CYP2C19 containing microsomes metabolized tolperisone substantially to the main M1 metabolite (Fig. 1). Recombinant CYP1A2 also catalyzed tolperisone hydroxylation at an order of magnitude lower rate than CYP2D6 and CYP2C19. CYP2B6 also contributes to the consumption of parent compound, however metabolite produced by this isozyme still needs to be identified.

View larger version (30K): [in this window] [in a new window]   Fig. 1.   Comparison of tolperisone-consuming and M1-producing activity of baculovirus-infected insect cell-expressed P450s. Twenty picomoles per milliliter of each enzyme was used at tolperisone concentration of 1 µM. After 20 min of incubation, the reaction was stopped, and unchanged parent compound and produced M1 were determined with HPLC. The amount of consumed tolperisone was calculated, and the data were converted into activity. Each bar represents the mean of a triplicate measurement ± S.D. A, comparison of tolperisone-metabolizing activity of expressed human P450s. B, comparison of M1 formation rates of expressed human P450s.

Kinetics of Tolperisone Hydroxylation. Enzyme kinetic analysis of M1 formation was carried out to study the main metabolic route of tolperisone biotransformation. In experiments using HLM tolperisone was used in the concentration range of 1 to 1000 µM. Data of M1 formation were fitted to Michaelis-Menten kinetics (Fig. 2). Calculated apparent kinetic values were summarized in Table 3. Kinetic analysis of M1 formation showed biphasic kinetics, suggesting the contribution of at least two populations of P450s. Eadie-Hofstee plot (Fig. 2B) of data points also revealed that there is a high K_m and a low K_m component of M1 formation. Fine characterization of M1 formation in expressed human P450s revealed CYP2D6 as the low K_m enzyme of methyl-hydroxylation (Table 4). CYP2C19 showed medium K_m and CYP1A2 turned out to be the high K_m component. The apparent V_max rank order was CYP2D6 < CYP1A2 < CYP2C19. The Cl_int (calculated as V_max/K_m) implied that CYP2D6 played the essential role in tolperisone metabolism.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Kinetic analysis of tolperisone metabolism in HLM. A, Michaelis-Menten plot of M1 formation reaction. The dotted line represents one-Km, the solid line represents two-Km kinetic calculation. The corresponding r2 values are 0.95 and 0.998, respectively. B, Eadie-Hofstee plot of M1 metabolite formation in HLM. V and S stand for the reaction velocity and tolperisone concentration, respectively. Tolperisone was used at 1 to 1000 µM.

Inhibition Studies for K_i Determination. To determine inhibitory potential of tolperisone toward the prominent isozyme, compound was tested to inhibit dextromethorphan O-demethylation and bufuralol hydroxylationwidely used index reactions of CYP2D6. Enzyme kinetic analysis of the data with the "K_m,app" method (Kakkar et al., 1999) revealed that tolperisone inhibited dextromethorphan O-demethylation and bufuralol-hydroxylation with the K_i of 17 and 30 µM, respectively.

Contribution of Microsomal Non-P450 Enzymes. To test whether non-P450, NADPH-dependent enzymes contribute to the in vitro metabolism of tolperisone in HLM, nonspecific chemical P450 inhibitors (SKF-525A, 1-aminobenzotriazole, 1-benzylimidazole) and anti-NADPH-P450 reductase polyclonal antibodies were examined to inhibit P450-dependent reactions. Results are shown in Table 1 (SKF-525A, ABT, and BI).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study the in vitro pathways of tolperisone metabolism was examined in pooled human liver microsomes and in recombinant enzymes. In previous pharmacokinetic studies it was found that tolperisone had a low bioavailability (20%, Miskolczi et al., 1987). Here we demonstrated that the rate of NADPH-dependent Cl_int of tolperisone could account for the low in vivo bioavailability.

Experiments studying metabolic stability of tolperisone also revealed that the compound is metabolized mainly to hydroxymethyl-tolperisone (M1), in vitro. Stochiometric calculations predicted the formation of other, substantial, non-UV active metabolites besides the main M1. LC/MS measurements showed that as well as M1 (m/z 261), the carbonyl-reduced form of this metabolite (M2, m/z 263), and another metabolite that is two mass units greater than parent compound was produced (m/z 247). Since several carbonyl-reduced metabolites were identified previously in rats (Miyazaki et al., 1972), and one of these was identified in HLM in the present study (M2), it was reasonable to assume that metabolite at m/z 247 was identical with the carbonyl-reduced parent compound. Nonspecific P450 inhibitors (SKF-525A, ABT, BI, and NADPH-P450 reductase-specific antibodies) potently inhibited M1 formation, however parent compound consumption was inhibited to a much lesser extent (40-60% residual microsomal activity at 100% P450 inhibition). This suggested the substantial contribution of P450-independent enzymes. Additionally, anti-reductase antibodies did not inhibit the formation of metabolite m/z 247, whereas the amount of M1 and M2 was decreased. These indirect evidences implied that besides P450s, a significant P450-independent route is involved in tolperisone metabolism that is assumed a microsomal carbonyl reductase. However, further supportive results are needed to identify the reductase and evaluate its role in tolperisone metabolism.

The P450-dependent metabolic processes were characterized in details. In our hands, the most potent chemical inhibitor of tolperisone consumption and M1 metabolite formation was quinidine that implied the main role of CYP2D6. This result was in concert with results of recombinant P450s, where the most active isoform (i.e., the enzyme with the highest turnover number) was CYP2D6 both in parent loss and M1 formation. More supportive data of the key role of CYP2D6 came from the experiments with specific inhibitory antibodies; 70% inhibition of M1 formation meant that the majority of the P450-dependent biotransformation is mediated by CYP2D6. S-mephenytoin inhibited parent loss about to the same extent as quinidine (22%), however the effect to M1 formation was weak (16%). On the other hand, the activity of recombinant CYP2C19 revealed 60% of rCYP2D6 activity in both compound loss and M1 formation placing recombinant CYP2C19 to the second most active P450 in tolperisone metabolism. Despite this high activity in expressed enzyme system, we concluded that at "physiological" tolperisone concentration the role of CYP2C19 is less essential than that of CYP2D6. The results given by inhibitory antibodies are controversial, since an unidentified interaction of tolperisone with goat serum was seen. However, it was clearly shown that at a higher concentration, the relative contribution of CYP2D6 decreased. This finding was supported with the results of enzyme kinetic characterization of tolperisone hydroxylation. Recombinant CYP2D6 revealed the low K_m enzyme with low apparent V_max indicating that at low concentration CYP2D6 plays the prominent role, whereas increasing the concentration of tolperisone increased the role of CYP2C19 as a medium K_m-high V_max enzyme. CYP1A2 turned out to be the highest K_m counterpart; however, because of the relative abundance of this isozyme its role is not negligible. Moreover, the role of a nonpolymorphic enzyme is important mainly in a subject of poor metabolizer phenotype both for CYP2D6 and CYP2C19. Calculation of Cl_int (V_max/K_m) revealed a rank order (CYP2D6 > CYP2C19 > CYP1A2) that assumed to be a valid order of activity for microsomal metabolism, however accurate scaling from heterologously expressed enzymes to human liver microsomes requires additional factors to consider (Venkatakrishnan et al., 2000; Nakajima et al., 2002).

Furafylline, a potent mechanism-based inactivator of CYP1A2 (Kunze and Trager, 1993) did not inhibit strongly tolperisone consumption, however the 25% inhibition in M1 formation may represent its ratio in the contribution to hydroxylation reaction as demonstrated by recombinant P450s. The inhibition measured using ketoconazole and sulfaphenazole was not strengthened by other methods; experiments with recombinant P450s and inhibitory antibodies did not support the contribution of CYP3A4 and CYP2C9. The effect is probably due to the fact that in this concentration range ketoconazole and sulfaphenazole are no longer selective for CYP3A4 and CYP2C9, respectively (Newton et al., 1995; Moody et al., 1999).

In our hands, tolperisone revealed to be a weak inhibitor of FMO3 index reaction in recombinant enzyme. The measured K_i, the small turnover number, and the inability of thiourea to inhibit tolperisone metabolism in HLM implied that this enzyme is not likely to contribute substantially to the metabolism of tolperisone, in vivo.

In the present study, it has been demonstrated that both P450-dependent and P450-independent microsomal biotransformations are involved in tolperisone metabolism, in vitro. Proposed metabolic routes are summarized in Fig. 3. Hydroxymethyl metabolite formation revealed to be the main P450-mediated metabolic pathway. CYP2D6 was identified as the key enzyme in metabolism through M1 formation and involvement of CYP2C19 and CYP1A2 were also shown. It was evidenced that P450-independent metabolism was mediated to a small extent by FMO3. Metabolites detected and indirect evidences from inhibition studies pointed toward the substantial involvement of presumable microsomal carbonyl reductase in the metabolism of tolperisone.

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Proposed in vitro metabolic pathways of tolperisone. Dashed arrows indicate the possible ways of M2 formation. Gray letters are used for the assumed enzymatic processes and metabolites.