Abstract
Irinotecan, or CPT-11 (7-ethyl-10-[4-(1-piperidino)-1-piperidino]carbonyloxycamptothecine), is a water-soluble derivative of camptothecine with promising activity against several types of malignancies. In addition to 7-ethyl-10-hydroxycamptothecine (SN-38), its active metabolite, we were able to identify several metabolites in the plasma of patients treated with this drug, especially an oxidative metabolite, 7-ethyl-10[4-N-(5-aminopentanoic acid)-1-piperidino] carbonyloxy-camptothecine. During our study of the biosynthesis of 7-ethyl-10[4-N-(5-aminopentanoic acid)-1-piperidino] carbonyloxy-camptothecine from CPT-11 by human liver microsomes, we were able to detect another quantitatively important polar metabolite, which was also present in the plasma and urine of patients treated with CPT-11. On the basis of preliminary experiments, the structure of this compound was postulated to be 7-ethyl-10-(4-amino-1-piperidino)carbonyloxycamptothecine, and this structure was synthesized by Rhône-Poulenc Rorer. Urine samples and human liver microsomal extracts were studied by high-performance liquid chromatography/atmospheric pressure chemical ionization/tandem mass spectrometry to identify its structure formally. The identification of the metabolite was supported by identical retention time, mass-to-charge ratio and tandem mass spectrometry fragmentation as a synthetic standard. Like irinotecan, 7-ethyl-10-(4-amino-1-piperidino) carbonyloxycamptothecine was a weak inhibitor of cell growth of P388 cells in culture (IC50 = 3.4 μg/ml vs. 2.8 μg/ml for irinotecan and 0.001 μg/ml for SN-38). It was also a poor inducer of topoisomerase I-DNA cleavable complexes (100-fold less potent than SN-38). However, unlike 7-ethyl-10[4-N-(5-aminopentanoic acid)-1-piperidino] carbonyloxy-camptothecine, this new metabolite could be hydrolyzed to SN-38 by human liver microsomes and purified human liver carboxylesterase, though to a lesser extent than irinotecan. This compound can therefore contribute to the activity and toxicity profile of irinotecan in vivo.
Irinotecan, or CPT-11, is a water-soluble derivative of camptothecine (Sawadaet al., 1991) with promising activity against several types of malignancies, especially colorectal cancer (Wiseman and Markham, 1996). Camptothecines interact with the nuclear enzyme topo I. Topo I releases the torsional strain in DNA, arising from DNA processing, through the formation of reversible single-stranded DNA breaks that make possible strand rotation. Camptothecines prevent the religation of these breaks, leading to the formation of double-stranded breaks during replication. CPT-11 itself has little if any activity in vitro and is thought to exert its anticancer activity after biotransformation into SN-38 (fig. 1) by carboxylesterases (Kawato et al., 1991).
Recently, we identified a major plasma metabolite of CPT-11 known as APC or RPR121056 (Rivory et al., 1996b). APC is formed by ring-opening oxidation of the distal piperidine ring of CPT-11 mediated by CYP 3A enzymes (Haaz et al., 1998). The presence of significant plasma concentrations of APC raises the possibility that oxidation of the bipiperidino side chain of CPT-11 is an important pathway of metabolism for CPT-11.
During the course of in vitro studies of CPT-11 metabolism, we detected significant production of other polar metabolites of CPT-11 when it was incubated with human liver microsomes in the presence of NADPH (Haaz et al., 1998). In particular, one of these appeared to be produced extensively and could account for an important step in the oxidative metabolism of CPT-11. Preliminary experiments with HPLC coupled to MS indicated that this compound could be a dealkylated metabolite of APC (Rivory et al., 1996b)—that is, 7-ethyl-10-(4-amino-1-piperidino)carbonyloxycamptothecine (fig. 1). Additionally, it was found to correspond with a peak also present in the plasma of patients treated with CPT-11 (Rivory et al.,1997). In order for us to continue investigating the metabolic pattern of CPT-11, this compound was synthesized by Rhône-Poulenc Rorer (RPR 132595A), as described in a separate report (Tetrahedron Letters, submitted for publication). In this study, we describe the use of HPLC/APCI/MS/MS to identify formally its presence in biological fluids as well as in extracts from incubations of human liver microsomes with CPT-11. In addition, we have evaluated its in vitro activity in terms of cell growth inhibition and induction of topo I-mediated DNA cleavage. Finally, we studied its ability to be hydrolyzed to SN-38, the active metabolite of CPT-11.
Materials and Methods
Chemicals and Reagents
Pure CPT-11, SN-38 and NPC (RPR 132595A) were supplied by Rhône-Poulenc Rorer (Vitry-sur-Seine, France). 20- (S) camptothecin was obtained from Sigma (St. Louis, MO). Solvents and reagents were of the highest grade available commercially.
Structural Identification of NPC
HPLC/APCI/MS/MS analysis.
Mass spectra were acquired using a modification of the method of Dodds et al. (1997). Briefly, a 616 HPLC pump and 600S controller (Waters, Milford, MA) were used to deliver a mobile phase of 0.075 M ammonium acetate (pH 5.3)/acetonitrile (77:23, v/v) isocratically at a flow rate of 1.5 ml/min to a Waters Radial-Pak column (8 × 150 mm, Nova-Pak C18, 4 μm). Samples were injected with an ISS200 sample processor (Perkin-Elmer, Danbury, CT). The bulk of the column eluent was directed to a Hitachi F1000 fluorometer (Scientific Instruments-Optical Sales, Brisbane, Australia) with excitation and emission wavelengths set at 355 and 515 nm, respectively. Peak data were collected and analyzed using Maxima software (Waters). The remaining fraction of the column outflow (1/6) was diverted to a Sciex API III triple quadrupole mass spectrometer (Thornhill, Ontario, Canada) equipped with an APCI interface in positive ionization mode. The temperature of the interface and the orifice potential were 500°C and 80 V, respectively, and the nebulizer gas pressure and auxiliary nitrogen flow were 4.8 atm and 2 ml/min, respectively.
An extract of urine from a patient treated with CPT-11 and an extract of an incubation of CPT-11 (100 μM) with human liver microsomes were prepared as detailed below. Full-scan mass spectra were acquired for the NPC standard and the microsomal and urine extracts by scanning Q1 from m/z 200 to 800 with a step size of 0.5 amu and a dwell time of 2 ms. The product ions were then analyzed by scanning Q3 from m/z 150 to 550. Collision-induced spectra were obtained using argon as the collision gas at a thickness of 3.0 × 1014 molecules cm−2.
The MS/MS data were analyzed with MacSpec software (Perkin-Elmer). The delays in the fluorescence detector and mass spectrometer were nearly identical, and the traces from both instruments could be overlayed directly.
Sample preparation.
A standard solution of NPC was prepared at 4.0 mM in dimethylsulfoxide and diluted with mobile phase to yield a final concentration equal to 200 μM, of which 10 μl was injected into the HPLC/APCI/MS/MS.
Incubation of CPT-11 (100 μM) with human liver microsomes (1 mg protein/ml) was performed as previously described (Haaz et al., 1998) for 1 hr in the presence of 1 mM NADPH. A sample (1.5 ml) of the incubation mixture was extracted with 4.5 ml of cold methanol/acetonitrile mixture 50:50 (v/v), centrifuged at 8000 ×g and evaporated to dryness under nitrogen. The extract was reconstituted in 200 μl of mobile phase to which 200 μl of 0.1 M citric acid was added, and 5 μl was injected into the HPLC/APCI/MS/MS.
A sample of urine (5 ml) collected from a patient treated with CPT-11 (350 mg/m2 over 30 min) was clarified by the addition of an equal volume of cold methanol/acetonitrile mixture (50:50, v/v), vortexed and centrifuged at 8000 × g. The supernatant was removed and dried down under nitrogen. The compound was resuspended in 3 ml of 0.05 M citric acid and applied to a solid-phase extraction column (Sep-Pak Vac, 3 ml, C18, Waters, Milford, MA) that had been conditioned with 5 ml of acetonitrile, followed by 5 ml of H2O and finally 5 ml of 0.05 M citric acid. The loaded column was washed with 0.05 M citric acid and compounds of interest eluted with 3 ml of acetonitrile/0.1 N HCl (95:5, v/v). The eluent was evaporated to dryness under nitrogen, and the residue was reconstituted in mobile phase (50 μl), of which 5 μl was injected into the HPLC/APCI/MS/MS.
Pharmacological Activity of NPC
Cell culture and growth inhibition assay.
Murine P388 leukemia cells free of mycoplasms (American Type Culture Collection, Rockville, MD) were grown as suspensions in a humidified atmosphere (37°C, 5% CO2) in RPMI medium containing 2 mMl-glutamine, 200 I.U./ml penicillin, 200 μg/ml streptomycin and 10 μM 2-mercaptoethanol and supplemented with 10% (v/v) of heat-inactivated fetal calf serum. Cytotoxicity was determined with exponentially growing cells as described previously (Madelaineet al., 1993). Briefly, cells were seeded in 96-well microculture plates (105 cells/ml) in the presence of a range of drug concentrations. After a 96-hr incubation, the cells were incubated for a further 16 hr with 0.02% (w/v) neutral red and then were washed and lysed with 1% SDS (w/v). Dye retention, which reflects the viability of the cells, was evaluated spectrophotometrically for each well at 540 nm with a multiwell spectrophotometer (Dynatech). Each point was performed in quadruplicate, and the results were expressed as a percentage of cell growth relative to the untreated control cells. The drug concentration resulting in 50% inhibition of cell growth (IC50) was estimated from semilogarithmic plots of cell growth vs. drug concentration.
Topo I-mediated DNA cleavage.
The topo I-mediated DNA cleavage induced by NPC was evaluated with the linearizedEcoR1-HindIII fragment of pBR322 and purified calf thymus topo I as already described (Kjeldsen et al.,1988). The cleavage reaction mixture contained 20 μM Tris-HCl, pH 7.4, 60 mM KCl, 0.5 mM EDTA, 0.5 M dithiothreitol, 2 × 104 dpm of 3′-end-labeled DNA, and a range of drug concentrations. The reaction was initiated by the addition of 20 units (in a 20-μl final volume) of topo I purified from calf thymus (Riouet al., 1986) and was allowed to proceed for 10 min at 37°C. Reactions were stopped by adding SDS at a final concentration of 0.25% (w/v) and proteinase K at 250 μg/ml, followed by incubation for 30 min at 50°C. Samples were denatured by the addition of 10 μl of denaturing buffer consisting of 0.45 M NaOH, 30 mM EDTA, 15% (w/v) sucrose and 0.1% (w/v) bromocresol green and were loaded onto 1% agarose gels in Tris-borate-EDTA buffer containing 0.1% SDS. After electrophoresis at 2 V/cm overnight, the gels were dried between two sheets of Whatman 3MM paper and autoradiographed with MP Hyperfilms (Amersham, Les Ulis, France) for 1 day. The X-ray films were scanned with a laser densitometer (Molecular Dynamics, Sunnyvale, CA) and analyzed using Image Quant software (Molecular Dynamics).
Transformation of NPC into SN-38
Human liver microsomes.
Human liver microsomes were prepared according to standard subcellular fractionation procedures (Berthouet al., 1989) from human livers obtained with the approval of the relevant institutional ethics committees. A pool of preparations containing ca. 15 mg/ml microsomal proteins was used for the study of NPC biotransformation into SN-38. All microsomes were stored at −80°C, and never refrozen after use.
Incubation conditions.
For the study of SN-38 formation from NPC by human liver microsomes, we used the same conditions as already described for the study of SN-38 formation from CPT-11 (Haaz et al., 1997). These conditions yielded linear CPT-11 biotransformation with respect to incubation time (15 min to 1 hr) and microsomal protein concentration (1 mg/ml final). The microsomes were treated with 0.5% Triton X-100 (v/v) and incubated on ice with gentle agitation for 10 min. Aliquot samples (20 μl, 0.3 mg protein) were added to 300 μl of 0.2 M Tris buffer (pH 6.9 at 37°C) in sialinized glass tubes and incubated in a shaking water bath at 37°C for a further 5 min. NPC was then added as either the lactone (stock diluted in 0.01 M citric acid) or the carboxylate (diluted in 0.1 M sodium carbonate) to yield the required final concentration (10 to 100 μM), and the contents of the tube were lightly vortex-mixed. As in our studies on SN-38 formation from CPT-11 (Rivory et al.,1996a, Haaz et al., 1997), we observed a significant “burst” production of SN-38 after the addition of either form of NPC to the microsome preparation, which indicated that biotransformation of NPC to SN-38 by carboxylesterases follows enzyme deacylation-limited kinetics. This “burst” was largely over by 10 min, and samples (50 μl) were taken from the reaction mixture at 20, 40 and 60 min to assess the reaction velocity at steady state. These samples were then assayed for total SN-38 by HPLC. Reaction velocities were evaluated by linear regression of the SN-38 concentrationsvs. time. Experiments were carried out in triplicate. The Michaelis-Menten parameters (apparent Km andVmax) were then determined by nonlinear regression (SigmaPlot, Jandel Scientific, Erkrath, Germany). Limited incubations of CPT-11 (25 μM) were also performed with the same microsomal pool for comparative purposes.
SN-38 formation from NPC was also studied with purified human liver carboxylesterase in the conditions already described (Rivory et al., 1996a). Incubations of NPC, APC and CPT-11 (50 μM) with purified human liver carboxylesterase were carried out in triplicate in PBS (pH 7.4) using 2 μM enzyme (E.C. 3.1.1.1) prepared according toKetterman et al. (1989). The substrates were always equilibrated in the reaction buffer at room temperature before initiation of the reaction, so they were present as a mixture of the lactone and carboxylate forms. Samples were withdrawn during the steady-state phase of the reaction and analyzed by HPLC.
HPLC.
Quantitative evaluation of CPT-11 metabolites was performed using a HPLC technique with fluorescence detection adapted from the technique of Rivory and Robert (1994). The samples taken from the incubation mixtures were added to 0.5-ml Eppendorf tubes containing 100 μl of ice-cold methanol/acetonitrile (50/50, v/v) and 50 ng of camptothecine (internal standard). The mixture was vortex-mixed and centrifuged at 8000 × g for 2 min. After acidification by addition of 7 μl of 1 N HCl to the supernatant and vigorous stirring, samples (10 or 20 μl) were injected onto the chromatograph and assayed for SN-38. Compounds were separated using a Radial-Pak C-18 reversed-phase column (Waters) and isocratic elution at 1.5 ml/min. The mobile phase was 80:20 (v/v) 0.075 M ammonium acetate buffer (pH 6.0) and acetonitrile to which PIC-A (Waters) was added to a final concentration of 5 mM.
Samples were processed via an automatic sampler (Spectra Series AS300, Thermo Separation Products, Les Ulis, France). Detection was carried out with a spectrofluorometer (Hitachi 1050F, Merck, Nogent-sur-Marne, France) with excitation and emission wavelengths set at 355 and 515 nm, respectively. Peak data were automatically recorded and concentrations calculated by reference to a standard calibration curve obtained before each series of injections, using the PC1000 software (Thermo Separation Products).
Results
Structural identification of NPC.
The mobile phase promoted the protonation of the NPC standard to form a stable [M + H]+ precursor ion ( m/z 519). This ion, which was also visible in both the microsomal and the urine extracts (figs. 2and 3), fragmented to produce two dominant ions with m/z values of 349 and 393. NPC also yielded the initial loss of CO2 (−44 amu), which is typical for this class of compounds in their lactone form (Rivory et al., 1996b). The first predominant peak m/z 393 indicates cleavage of the ester moiety of CPT-11, and a further loss of 44 amu suggests the loss of CO2 from this species.
The correspondence of the NPC peak with the fluorescent metabolite of interest (fig. 2) indicates that NPC is a major metabolite produced during the incubation of CPT-11 with human liver microsomes. The identification of this compound is supported by identical retention time, m/z of parent ion and MS/MS fragmentation as compared with the synthetic standard during analysis with HPLC/APCI/MS/MS. NPC is also present in the urine of patients treated with CPT-11 (fig. 3) and in the plasma (data not shown), though in apparently lower concentrations than APC (Rivory et al., 1996b).
Pharmacological activity of NPC.
The cytotoxic potential of NPC was assessed by a cell growth inhibition assay using the P388 murine leukemia cell line. For comparative purposes, CPT-11 and SN-38 were evaluated in parallel experiments. Under the conditions used (4 days of continuous exposure), NPC was found to be a weak inhibitor of cell proliferation. The IC50 for NPC (mean ± S.D.;n = 4) was 3.4 ± 0.3 μg/ml, whereas the IC50 values for CPT-11 and SN-38 were 2.8 ± 0.4 and 0.0010 ± 0.0006 μg/ml, respectively.
The induction of topo I-mediated DNA cleavage by NPC was assessed with an end-labeled fragment of pBR322 DNA, and the cleavage products were separated by alkaline agarose gel electrophoresis. An autoradiogram of a gel from a typical experiment is shown in figure4. In the presence of drugs at the highest concentrations tested (CPT-11, 100 μg/ml, lane 4; SN-38, 1 μg/ml, lane 8; NPC, 100 μg/ml, lane 13), there was a strong enhancement of DNA cleavage compared with the enzyme alone, an enhancement that was not present in the absence of enzyme (data not shown). The cleavage resulted in reduction of the intensity of the band corresponding to the DNA substrate and the appearance of an increasing proportion of smaller fragments. The relative potency of the three compounds could be estimated by inspecting the autoradiogram and comparing the concentrations of drug required to cause a similar extent of cleavage as determined from the intensity of the band corresponding to the DNA substrate. NPC at 10 μg/ml (lane 14) produced cleavage intermediate between those resulting from CPT-11 at 100 and 10 μg/ml (lanes 4 and 5, respectively), which suggests that NPC was slightly more potent. In comparison, SN-38 yielded equivalent cleavage at a concentration of 0.1 μg/ml (lane 9), which indicates that NPC is approximately 100-fold less active than SN-38 in the stabilization of DNA topo I cleavable complexes. SN-38 was found to be a contaminant of the preparation of NPC (∼0.7%) to a greater extent than in CPT-11 (∼0.03%); this probably explains the slightly higher activity of NPC.
In vitro kinetics of the transformation of NPC into SN-38 by human liver microsomes.
The formation of total SN-38 from NPC lactone and carboxylate at different concentrations is shown in figure 5. Both the velocity and the values of the y-axis intercepts were greater after the incubation of lactone than of the corresponding carboxylate. The mean values (± S.D.) of Km andVmax obtained using the Michaelis-Menten equation were 66.2 ± 30.2 μM and 1.66 ± 0.38 pmol/min/mg protein for the lactone and 86.5 ± 60.0 μM and 0.98 ± 0.38 pmol/min/mg protein for the carboxylate form of NPC, respectively. At 25 μM, the velocities of SN-38 production for the lactone forms of CPT-11 and NPC were 0.65 ± 0.29 and 0.29 ± 0.09 pmol/min/mg, respectively.
We confirmed the possibility of transformation of NPC into SN-38 by incubating this metabolite with purified human liver carboxylesterase. Figure 6 shows the formation of SN-38 from NPC, CPT-11 and APC, each at a concentration of 50 μM. SN-38 production was 4-fold lower with NPC as a substrate than with CPT-11, but it was much higher than with APC, which itself yielded negligible production of SN-38 over 2 hr.
Discussion
Because SN-38 is likely to play an important role in the cytotoxic activity of CPT-11, characterization of the metabolism of CPT-11 and its effect on SN-38 pharmacokinetics is of fundamental importance. Recently, we reported on the inter-relationships among the pharmacokinetics of CPT-11, SN-38, SN-38 β-glucuronide and APC in patients receiving CPT-11 (Rivory et al., 1997). An interesting observation was that patients in whom plasma concentrations of APC were high also tended to have high concentrations of SN-38 in spite of the fact that APC is not appreciably converted to SN-38 (Rivory et al., 1996b). We postulated that oxidative metabolism of the bipiperidino side-chain of CPT-11 might lead to the formation of metabolites that might be converted to SN-38 at a greater rate than CPT-11. Because the conversion of CPT-11 to SN-38 by human liver carboxylesterase is relatively low (Rivory et al.,1996a; Haaz et al., 1997), such a finding could explain the observation noted above. Therefore, we set about identifying the major products of the oxidative metabolism of CPT-11. One major peak was observed after the incubation of CPT-11 with human liver microsomes in the presence of NADPH. Its structure was confirmed as being that of NPC by its correspondence, in terms of both retention time and fragmentation, with a synthetic standard. Interestingly, NPC appeared to be a minor metabolite in plasma samples in this and other studies (Rivory et al., 1997). Quantitative studies will be required to evaluate the production of NPC relative to that of APC; such studies are currently underway. Incubations of NCP with both human liver microsomes and human liver microsomal carboxylesterase revealed that it, unlike APC, is a possible precursor of SN-38. Although the conversion of NPC to SN-38 was lower than the corresponding form of CPT-11 in experiments carried out at fixed concentrations (25 μM, microsomes; 50 μM, pure enzyme) of the two substrates, the appreciable conversion of NPC confirms the potential of oxidative metabolites of CPT-11 taking part in the production of SN-38.
As has been observed with both CPT-11 and APC (Rivory et al., 1996b), the presence of the bulky carbamate substitution at the C-10 position of camptothecine precludes significant induction of topo I-mediated DNA cleavage relative to SN-38.
In conclusion, a major metabolite detected after the incubation of CPT-11 with human liver microsomes was identified as being NPC. Because NPC is converted to SN-38, it may contribute to the activity and toxicity of CPT-11 in vivo.
Acknowledgments
Helen Dodds is the recipient of a Dora Lush Fellowship, and Marie-Christine Haaz is a recipient of a fellowship from the Ligue Nationale Française contre le Cancer. We thank Mrs. C. Garcia for skillful technical assistance.
Footnotes
-
Send reprint requests to: Jacques Robert, Dr. Med., Dr. Sci., Institut Bergonié, 180 rue de Saint-Genès, 33076 Bordeaux-cedex, France.
-
↵1 This study was supported by grants from the Ligue Nationale Française contre le Cancer, the Rhône-Poulenc Rorer company and the Princess Alexandra Hospital Research Foundation.
-
↵2 Present address: Cancer Pharmacology, Sydney Cancer Centre, Royal Prince Alfred Hospital, Sydney, Australia.
- Abbreviations:
- APCI
- atmospheric pressure chemical ionization
- MS
- mass spectrometry
- CPT-11
- 7-ethyl-10-[4-(1-piperidino)-1-piperidino] carbonyloxycamptothecine
- SN-38
- 7-ethyl-10-hydroxycamptothecine
- APC
- 7-ethyl-10-[4-N-(5-aminopentanoic acid)-1-piperidino]carbonyloxy-camptothecine
- NPC
- 7-ethyl-10-(4-amino-1-piperidino)carbonyloxycamptothecine
- CYP
- cytochrome P-450
- m/z
- mass-to-charge ratio
- SDS
- sodium dodecyl sulfate
- PBS
- phosphate-buffered saline
- topo I
- topoisomerase I
- Received January 7, 1998.
- Accepted March 30, 1998.
- The American Society for Pharmacology and Experimental Therapeutics