Metabolism of a New Local Anesthetic, Ropivacaine, by Human Hepatic Cytochrome P450 

The experimental protocol was approved by the Institutional Ethical Committee. Ropivacaine:(S)-enantiomer of 1-propyl-2′,6′-pipecoloxylidide, and its metabolites, 2′,6′-pipecoloxylidide (PPX), 3′-hydroxyropivacaine (3′-OH Rop), and 4′-hydroxyropivacaine (4′-OH Rop) were gifts from the Fujisawa Pharmaceuticals Co., Ltd (Osaka, Japan). Recombinant human P450s expressed in human lymphoblast cells were obtained from Gentest (Woburn, MA). These were supplied as microsomes. Dilauroylphosphatidylcholine and dioleoylphosphatidylcholine were obtained from Sigma Chemical Co. (St. Louis, Mo). Phosphatidylserine (bovine) was obtained from PL Biochemicals (Milwaukee, WI). Reduced nicotinamide adenine dinucleotide phosphate (NADPH) was obtained from the Oriental Yeast Co. (Tokyo, Japan). A reverse-phase octadecasilyl column (TSK gel ODS-1 20T, 4.6 x 250 mm) was obtained from the Tosoh Corp. (Tokyo, Japan). Other reagents and organic solvents were obtained from the Wako Pure Chemical Industries (Tokyo, Japan).

Human hepatic microsomes were prepared from 15 liver samples as described for the preparation of rat hepatic microsomes. [11] These liver samples were obtained from cancer patients undergoing liver resection. Tissues were selected from areas of the liver that were visually free of tumor, frozen within 60 min, and stored at -80 degrees Celsius until used. Samples that had a specific level of P450 above 0.1 nmol/mg of protein were used in this study. Rat hepatic microsomes were prepared, and rat P450s were purified as described. [11–13] NADPH-P450 reductase and cytochrome b⁵were purified as described. [14].

The ropivacaine monooxygenase activity was measured as we reported for lidocaine. [15] Ropivacaine (0.1 mM) was incubated with hepatic microsomes (200 micro gram of protein) from rats or humans. Metabolic activities of hepatic microsomes are proportional to 1.0 mg of microsomal protein. Purified rat hepatic P450 (30 pmol) and NADPH-P450 reductase (0.3 units) were reconstituted with dilauroylphosphatidylcholine (10 micro gram). The catalytic activity of each P450 is proportional to the hemoprotein concentration (0–60 pmol/0.5 ml). Microsomes from lymphoblast cells (500 micro gram of protein) were used to recombinant human P450 assay. CYP3A2** has low activity in the conventional reconstituted system described above, [17] so we used a modified reconstituted system containing a mixture of phospholipids (10 micro gram) consisting of dilauroylphosphatidylcholine, dioleoylphosphatidylcholine, and phosphatidylserine (1:1:1) and sodium cholate (100 micro gram). Microsomes and reconstituted P450 were reacted with ropivacaine (0.1 mM) in the presence of NADPH (0.4 mM) at 37 degrees Celsius 60 and 15 min respectively. The formation of ropivacaine metabolites was linear up to 90 min when rat and human hepatic microsomes were used. It was also linear up to at least 15 min when reconstituted P450 was used. The final volume of the reaction mixture was 0.5 ml in 0.1 M potassium phosphate (pH 7.4). The reaction was stopped by adding 1 N NaOH (50 micro liter). The metabolites were extracted with ethyl acetate (2 ml). The organic phase was evaporated in vacuo, and the residue was dissolved in elution buffer for high-performance liquid chromatography (200 micro liter). One hundred microliters of the solution was injected onto an high-performance liquid chromatography apparatus with a ODS-1 20T column and isocratically eluted with 0.1 M potassium phosphate buffer (pH 3.0) and acetonitrile (9:1, v/v). Chromatography was done at a flow rate of 1.5 ml/min at 60 degrees Celsius, and the metabolites were monitored at 214 nm.

In our assay, internal standard was not used for measuring ropivacaine metabolic activity. The amount of ropivacaine metabolites was calculated by comparison of their peak area as calculated by a data processor with those of authentic ropivacaine metabolites such as PPX, 3′-OH Rop, and 4′-OH Rop. The same analytical method was used in our previous reports for measuring metabolites of lidocaine and aminopyrine. [14,15] Standard curve for PPX was linear between 0–5 micro Meter, standard curves for 3′-OH Rop and 4′-OH Rop were also linear between 0 and 2 micro Meter. Coefficients of correlation of these standard curves were more than 0.999, and limit of quantification was 0.01 nanomole of products per minute per nanomole of P450.

Antibodies against rat CYP1A2, 2B1, 2C11, 2D1, and 3A2 were raised in a rabbit and immunoglobulin G (IgG) as reported. [18] P450 was measured by immunoblotting as described previously. [18] Antibodies (20, 40, 60, 80, and 100 micro gram of IgG) were incubated with the microsomes (200 micro gram of protein) for 20 min at room temperature.

Rates of metabolism are expressed as nanomoles of product formed per minute per milligram of protein, or per nanomole of P450. The relationships between ropivacaine metabolic activity and the content of P450 enzymes were examined by the linear least-squares correlation analysis.

The high-performance liquid chromatography profile of the ropivacaine metabolites produced by human hepatic microsomes is shown in Figure 1(A). Peaks of ropivacaine metabolites were identified by comparison with those of authentic PPX, 3′-OH Rop, and 4′-OH Rop. The peak areas were increased in proportion to the amounts of hepatic microsomes present. These metabolites did not appear when ropivacaine was incubated with human hepatic microsomes without NADPH (Figure 1(B)), suggesting that those metabolites were produced by NADPH-dependent P450 monooxygenase (Figure 2). These ropivacaine metabolites also were formed when ropivacaine was intravenously administered into humans. **Formation rates of PPX, 4′-OH Rop, and 3′-OH Rop by human hepatic microsomes from one liver sample were 0.560, 0.030, and 0.047 nmol *symbol* min sup -1 *symbol* mg of protein sup -1, respectively.

The ropivacaine metabolic activity of purified rat hepatic P450s is shown in Table 1. CYP2C11 is male-specific, [19] and it formed high levels of PPX. The formation of PPX by CYP3A2, another major constitutive P450 isozyme in microsomes from male rats, [11] was also high in the modified reconstituted system. The ropivacaine N-dealkylation activity of CYP1A2, 2A2, 2B1, and 2D1 was low and undetectable using CYP2B2 nor 2E1.

CYP1A2 had low activity toward ropivacaine 4′-hydroxylation, but the other P450 isozymes did not. CYP1A2 and 2D1 had high activity toward ropivacaine 3′-hydroxylation. CYP2B1, 2C11, and 2E1 had slight ropivacaine 3′-hydroxylation activity.

We performed an immunoinhibition study using antibodies against rat CYP1A2, 2C11, 2D1, and 3A2, because these P450 isozymes had high ropivacaine metabolic activities. Ropivacaine N-dealkylation was more than 80% inhibited by an antibody against CYP3A2 (Figure 3). However, this activity was not inhibited by an antibody against CYP2C11, although CYP2C11 had high ropivacaine N-dealkylation activity (Table 1). We showed that CYP2C11 had high lidocaine N-dealkylation activity, whereas its antibody did not inhibit it, [20] although anti CYP2C11 antibody can inhibit testosterone 2 alpha- and 16 alpha-hydroxylation activities, which are catalyzed by CYP2C11 in microsomes from untreated rat liver. [18] These results suggested that CYP2C11 was not involved in ropivacaine N-dealkylation in rat hepatic microsomes. Ropivacaine 4′-hydroxylation was about 70% inhibited by anti CYP3A2 antibody. It was also 40% inhibited by antibodies against CYP1A2 and 2D1 (Figure 3), although neither purified CYP2D1 nor 3A2 had ropivacaine 4′-hydroxylation activity. CYP2D1 and 3A2, as well as CYP1A2, may be involved in ropivacaine 4′-hydroxylation in rat hepatic microsomes. The ropivacaine 4′-hydroxylation activity of rat hepatic microsomes was very low, and those of CYP2D1 and 3A2 could be undetectable under our reaction conditions (Table 1). Ropivacaine 3′-hydroxylation was almost completely inhibited by antibodies against CYP2D1 as well as CYP1A2 (Figure 3), suggesting that both of those P450 isozymes are involved in ropivacaine 3′-hydroxylation. These results are consistent with those obtained using purified rat hepatic P450 (Table 1).

Antibodies against rat hepatic P450s were useful for detecting P450 isozymes involved in ropivacaine metabolism. We used these P450 antibodies to identify the P450 isozymes involved in ropivacaine metabolism in human hepatic microsomes. Antibody against CYP3A2 inhibited both ropivacaine N-dealkylation and the 4′-hydroxylation activity of human hepatic microsomes by more than 80%(Figure 4). We showed that antibody against CYP3A2 specifically reacts with CYP3A4 in human hepatic microsomes and that CYP3A2 and 3A4 are immunochemically related. [15] Antibodies against CYP1A2, 2B1, 2C11, and 2D1 inhibited neither ropivacaine N-dealkylation nor 4′-hydroxylation. These results suggest that ropivacaine is selectively N-dealkylated and hydroxylated at position 4 by CYP3A4 in human hepatic microsomes. Antibody against CYP1A2 inhibited ropivacaine 3′-hydroxylation by more than 80%. Antibodies against CYP2B1, 2C11, 2D1, and 3A2 did not inhibit this reaction (Figure 4), suggesting that ropivacaine 3′-hydroxylation is catalyzed selectively by CYP1A2.

The ropivacaine metabolic activity of seven different human hepatic P450s expressed in human lymphoblasts is shown in Table 2. CYP3A4 catalyzed ropivacaine N-dealkylation, and CYP1A1 and 2B6 had much less activity. CYP1A1, 1A2, 2D6, and 3A4 had low ropivacaine 4′-hydroxylation activity. CYP1A2 had high ropivacaine 3′-hydroxylation activity. CYP1A1, 2B6, and 2D6 had ropivacaine 3′-hydroxylation activity, whereas the formation rate of 3′-OH Rop by those P450 isozymes was lower than that of CYP1A2. CYP2A6, 2E1, and 3A4 did not have this activity. Although CYP2D1, purified from rat hepatic microsomes, had ropivacaine 3′-hydroxylation activity, CYP2D6, an ortholog of CYP2D1 in human hepatic microsomes, had very low levels of this activity. This is consistent with the immunoinhibition results using the antibody against CYP2D1.

The catalytic and inhibition studies described above suggested that ropivacaine is selectively N-dealkylated by human CYP3A4, and hydroxylated at position 3 by human CYP1A2. To confirm this, the levels of ropivacaine N-dealkylation and 3′-hydroxylation activities in individual human hepatic microsomes were compared with the immunochemically estimated levels of CYP3A4 and 1A2, respectively. The concentration of CYP3A4 was determined using an antibody against rat CYP3A2 as described. [15] There were significant linear correlations between ropivacaine N-dealkylation activity and the CYP3A4 content (r = 0.88, P < 0.0001;Figure 5(A)) as well as between ropivacaine 3′-hydroxylation activity and CYP1A2 level (r = 0.88, P < 0.0001;Figure 5(B)). Ropivacaine 3′-hydroxylation activity did not correlate with the concentration of CYP3A4 (r = 0.22, P > 0.05; data not shown).

These regression lines pass near the origin, suggesting that no other enzymes in the microsomes are involved in either ropivacaine N-dealkylation or 3′-hydroxylation. These results confirm that CYP3A4 and 1A2 contribute to ropivacaine N-dealkylation and 3′-hydroxylation, respectively.

In this study, we found that human hepatic P450 specifically metabolized ropivacaine. PPX was the major metabolite commonly found in humans and rats, and it was formed predominantly by P450 in the same gene family, CYP3A4 and 3A2. CYP3A2 is ortholog to CYP3A4, and they are structurally and immunochemically related each other. [21] Antibodies against CYP3A2 also react with CYP3A4 in human hepatic microsomes and can be used to determine the level of CYP3A4 in human hepatic microsomes. [15].

3′-OH Rop and 4′-OH Rop also were formed in microsomes from human and rat liver, and P450 isozymes were involved in these reactions. We have determined the enzyme specificity for lidocaine metabolism, and CYP3A4 is the major P450 isozyme involved in lidocaine N-dealkylation in human hepatic microsomes. [15] CYP3A4 plays a major role in the oxidation of many drugs including nifedipine, alfentanil, midazolam, and quinidine. [21–24] These drugs may compete with ropivacaine for drug metabolism. In this study, ropivacaine also was metabolized by CYP3A4 in human hepatic microsomes to the N-dealkylated form, PPX. PPX has been detected in humans after mepivacaine and bupivacaine administration, [25] suggesting that the N-dealkylation of these anesthetics is catalyzed by CYP3A4. CYP3A4 is inducible by drugs such as barbiturate and dexamethasone. [21] When this P450 is induced, metabolism of ropivacaine will be increased.

CYP2D6 is present in human hepatic microsomes and contributes to genetic polymorphisms in debrisoquine metabolizing activities. [4] The ropivacaine 3′-hydroxylation activity of CYP2D6 was very low, and the formation of 3′-OH Rop was not inhibited by antibody against CYP2D1, suggesting that CYP2D6 is not involved in ropivacaine 3′-hydroxylation in human hepatic microsomes. These results contradict those obtained from rats, because CYP2D1 is involved in ropivacaine 3′-hydroxylation in rat hepatic microsomes. We showed that CYP2D1 is involved in lidocaine 3′-hydroxylation in rat hepatic microsomes. [26] However, lidocaine 3′-hydroxylation activity in human hepatic microsomes constituted less than one-tenth of that in rat hepatic microsomes, [15,20] and the plasma concentration of aromatic 3′-hydroxylated lidocaine in human is very low after the intravenous administration of lidocaine. [27] In this study, the formation rate of 3′-OH Rop was greater by rat than by human hepatic microsomes. Because 3′-hydroxylation activity of lidocaine and ropivacaine by human hepatic microsomes was low and plasma concentration of 3′-hydroxylated lidocaine is also low, we speculate that aromatic ring hydroxylation at position 3 of lidocaine and ropivacaine is not the major metabolic pathway in humans, and that CYP2D6 is not involved in the metabolism of these agents. CYP1A2 is involved in ropivacaine 3′-hydroxylation commonly in human and rat hepatic microsomes.

Although ropivacaine 4′-hydroxylation by human hepatic microsomes was approximately 70% inhibited by antibody against CYP3A2 (Figure 4), ropivacaine 4′-hydroxylation activity of CYP3A4 was as high as that of CYP1A1 and 2D6. These results suggest that CYP3A4 is responsible for a significant fraction of 4′-OH Rop, whereas other P450 isozymes also could be involved in this reaction.

In summary, we showed that ropivacaine is N-dealkylated and aromatic-ring hydroxylated at positions 3 and 4 to PPX, 3′-OH Rop, and 4′-OH Rop, respectively, in human hepatic microsomes and that PPX was the major metabolite. The specific P450 isozymes, CYP3A4 and 1A2, are involved in these metabolic pathways. CYP3A4 in human hepatic microsomes metabolizes many substrates, including nifedipine, alfentanil, midazolam, and quinidine. These agents commonly are administered during anesthesia. Thus, ropivacaine and other commonly administered drugs that are also metabolized by CYP3A4 may pharmacokinetically interact.

**Fujisawa Pharmaceuticals Co. Ltd.: Personal communication, 1993.

**The nomenclature used here for the P450 enzymes is that described by Nelson et al. [16].