Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Ifosfamide (IFO)¹ is an oxazaphosphorine alkylating agent effective against a variety of cancers (Sladek, 1988; Sechant et al., 1991). It is a prodrug which requires metabolic activation by cytochrome P450, specifically CYP3A4 (Walker et al., 1994) to produce the active alkylating agent isophosphoramide mustard. In addition to the production of isophosphoramide mustard, N-dechloroethylation of the parent compound yields the inactive metabolites 2- and 3-dechloroethylifosfamide and chloroacetaldehyde (CAA). CAA has been associated with neurotoxicity (Goren et al., 1986) and suggested to be responsible for inducing IFO related CNS toxicity (Kurowski et al., 1991), although the underlying mechanism remains poorly understood. Recently, CAA was shown to inhibit mitochondrial function (Visarius et al., 1997), but little attention has been given to the study of potential mitochondrial dysfunction induced by products of CAA biotransformation. CAA and its associated oxidation/reduction products, 2-chloroacetic acid and 2-chloroethanol, respectively, may react with endogenous cysteine or glutathione forming S-conjugates which may in turn be metabolized to S-carboxymethyl-L-cysteine (Hoffman et al., 1991). Further biotransformation of CMC, by oxidative desamination and decarboxylation yields TDGA (Hoffman et al., 1991), a 3-thia dicarboxylic acid and thus a potential inhibitor of mitochondrial function.

In addition to its identification as a major human biotransformation product of the mucolytic agent CMC, TDGA has also been found as a metabolite of the industrially important compounds vinyl chloride (Draminski and Trojanowska, 1981; Chen et al., 1983; Sharma et al., 1980; Jedrychowski et al., 1984), ethylene dichloride (Payan et al., 1993; Cheever et al., 1990), vinylidine chloride (Reichert et al., 1979), acrylonitrile (Fennell et al., 1991; Kedderis et al., 1993), of 2,2'-bis-(chloroethyl)-ether (Müller et al., 1979; Lingg et al., 1982), and 2-(chloroethyl) nitrosoureas (Godeneche et al., 1993). The presence of TDGA in urine has been used as an index for monitoring occupational exposure to vinyl chloride (Heger et al., 1982); however, adverse effects of exposure to TDGA have not as yet been reported. Since TDGA is structurally similar to other known inhibitors of mitochondrial function (Tonsgard and Getz, 1985; Passi et al., 1984) and sulfur-substituted fatty acid analogues may inhibit -oxidation (Skrede et al., 1997), the purpose of this study was to investigate the effects of TDGA on mitochondrial function in vivo and in vitro and to quantify the importance of pathways leading to the production of TDGA in patients treated with IFO.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. [1-¹⁴C]PA was purchased from Amersham (Buckinghamshire, UK), [1-¹⁴C]OA, and [1,4-¹⁴C]SA from DuPont NEN (Boston, MA). IFO was obtained from ASTA Medica (Wangen, Switzerland), TDGA, PA, palmitoyl-L-carnitine (PLC), and SA were from Sigma (St. Louis, MO).

Patients. Thirteen patients (five female, eight male) with advanced soft-tissue sarcoma were studied during one to three chemotherapy cycles administered 3 months apart. In each cycle of 5 days' treatment, the patients received 12-16 g IFO/m²·day by continuous infusion together with adequate hydration (2 liters 5% glucose/day) and the uroprotective agent mesna which was administered intravenously at a dose of 1.9-2.8 g/m²·day. Urine was collected prior to the administration of IFO and during the infusion of IFO. Aliquots were stored frozen until analysis. TDGA was measured in the urines obtained during 25 chemotherapy cycles of 5 days' duration. All patients gave informed consent to participate in the study, which had been approved by the local ethics committee.

Analysis of TDGA. Patient urine (0.5ml) together with 75µl thiodipropionic acid (TDPA) (0.22mM), which was used as an internal standard, were lyophilized, and TDGA- and TDPA-dimethyl esters were then formed by reaction with borontrifluoride/methanol (BF₃ ~10% [~1.3M] in MeOH), for 20 min at 60°C. Dimethyl esters were extracted into ethylacetate and subsequently detected by gas chromatography. Quantitative analyses were performed on a Perkin Elmer 3920 gas chromatograph equipped with a sulfur-specific flame photometric detector on an OV-17, 3% packed glass column. Injector and detector temperatures of 200°C and isothermal oven temperature of 150°C allowed for the elution of TDGA and TDPA derivatives at 2.1 and 5.6 min, respectively. Calibration curves for the derivatized standards were linear over the range of concentrations studied (1.25µM - 2.5mM) after transformation to account for exponential response of the sulfur specific detector. Peak areas were calculated with a Spectra Physics 4290 integrator.

Animals. Male Wistar rats (200-250 g) were bred and housed in the University of Berne vivarium, allowed access to food and water ad libitum, and kept on a 12-hr light/dark cycle. Rats anesthetized with pentobarbital (50mg/kg, ip), whose body temperatures were regularly monitored and maintained at 38°C, were used for all experiments.

In Vivo Oxidation of Labeled Substrates. To collect breath of anesthetized rats, a cylindrical vessel attached to a vacuum pump was placed over the rodent's head. TDGA (50mg/kg, ip), dissolved in sterile saline (0.9%), was applied 5 min preceding the ip administration of [1-¹⁴C]PA (3µCi/kg, 55.0mCi/mmol), [1-¹⁴C]OA (0.3µCi/kg, 55.0mCi/mmol), or [1,4-¹⁴C]SA (0.3µCi/kg, 59mCi/mmol). [1-¹⁴C]PA and [1-¹⁴C]OA were prepared in commercially available thistle oil, and [1,4-¹⁴C]SA in sterile saline (0.9%). Control animals received an equivalent volume of saline without TDGA (ip) 5 min before receiving either [1-¹⁴C]PA, [1-¹⁴C]OA, or [1,4-¹⁴C]SA. ¹⁴CO₂ produced from the oxidation of labeled substrates was collected from a constant air stream while an applied vacuum (200 ml/min) pulled exhaled breath through successive solutions of ethanol, to dry exhaled breath, and ethanolamine (4M, in ethanol), to trap exhaled ¹⁴CO₂. The ¹⁴CO₂ was then quantified by scintillation spectroscopy.

Isolation and Incubation of Mitochondria. Mitochondria from the livers of male Wistar rats (200-250 g) were isolated as previously described (Johnson and Lardy, 1967). Briefly, livers were quickly removed from decapitated rats and immediately transferred to ice-cold mannitol (0.25 M)-sucrose (0.07 M) buffer (MSB). Livers were finely minced and EDTA (2 mmol/L) was added prior to homogenization. Mitochondria were isolated by means of differential centrifugation, suspended in MSB at a concentration of 1g original liver weight per ml, and stored on ice. Protein concentrations in mitochondrial preparations were determined with the biuret method using bovine serum albumin as standard (Cornall et al., 1949).

Statistical Analysis. Results are presented as mean ± SEM. The urinary excretion of TDGA was evaluated by repeat measure one-way analysis of variance using Bonferroni correction for multiple comparisons. The results of the animal and in vitro experiments were assessed by Student-t test. Two-tailed probabilities of less than 0.05 were considered to be significant.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Prior to the infusion of IFO, TDGA was found in the urine samples of two patients who excreted 22 and 4 µmol TDGA/24 hr. During the infusion of IFO the excretion of TDGA gradually increased to 0.10 ± 0.02 mmol on day 1, 1.44 ± 0.15 mmol on day 2, 2.87 ± 0.21 mmol on day 3, 3.27 ± 0.15 mmol on day 4, and 2.84 ± 0.14 on the last day of IFO infusion. Amount of TDGA excreted on days 2-5 was significantly different (p < 0.01) from the excretion of TDGA prior to IFO infusion. The fraction of the administered dose of IFO recovered as TDGA in urine ranged from 0.23 ± 0.05% on the first day of IFO infusion to 8.49 ± 0.36% on the day of maximal TDGA excretion (fig. 1). This amounted to 5.4 ± 0.2% of the administered dose of IFO appearing as TDGA during the 5 days' continuous IFO infusion.

View larger version (40K): [in this window] [in a new window]   Fig. 1.   Excretion of TDGA in urine by patients receiving IFO. Data represent 25 IFO cycles in 13 patients who received 12-16g IFO/m2·day (mean ± SEM).

As shown in fig. 2, TDGA resulted a 55% inhibition of the oxidation of PA. To localize the inhibitory effect to the carnitine-dependent transport of long chain fatty acids into the mitochondria, -oxidation, or the Kreb's cycle, the effects of TDGA on the oxidation of OA and SA were also studied. Neither the oxidation of OA, a substrate for -oxidation which does not require carnitine for its import into the mitochondrial matrix, nor the oxidation of SA, a substrate for the Kreb's cycle, were inhibited by TDGA (fig. 2 and table 1).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Effect of TDGA on the exhalation of 14CO2 formed from labeled palmitic, octanoic and succinic acid, respectively. Filled symbols mark the exhalation of 14CO2 from the oxidations of: (a) [1-14C]palmitic acid, (b) [1-14C]octanoic acid, and (c) [1,4-14C]succinic acid, after the administration of of TDGA (50mg/kg). Open symbols represent 14CO2 produced by control animals which received the respective labeled substrates (mean ± SEM, N = 3).

Experiments with isolated mitochondria allowed for the comparison between PA and PLC oxidation and, thus, mechanistic evaluation of long chain fatty acid oxidation. In uncoupled mitochondria, TDGA markedly inhibited PA oxidation but did not affect PLC metabolism (table 1).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The results depicted indicate that large quantities of TDGA are excreted by patients treated with IFO. TDGA may be produced from the pathways indicated in fig. 3. The key metabolite of IFO resulting in the formation of TDGA is most likely CAA. It is possible that an appreciable fraction of the IFO dose is converted into TDGA as each IFO molecule possesses two chloroethyl side chains which may be metabolized to CAA through N-dealkylation (Sechant et al., 1991; Walker et al., 1994) or by enzymatic conversion of IFO-produced chloroethylamine (Aeschlimann et al., 1996). CAA is present in plasma of patients treated with IFO in concentrations up to 210 µM (Cerny and Küpfer, 1989). It may directly react with glutathione to produce equal amounts of S-(carboxymethyl)glutathione and S-(2-hydroxyethyl)glutathione (Jean and Reed, 1992). Alternatively, it can be oxidized or reduced to 2-chloroacetic acid or 2-chloroethanol, respectively, which will react with cysteine and glutathione, although markedly slower than CAA itself (Johnson, 1967), to form CMC. CMC in turn may undergo deamination and decarboxylation to form TDGA. When CMC, which is used as a mucolytic agent, is administered to patients, approximately 20% of the administered dose is excreted as TDGA (Hoffman et al., 1991). Assuming a similar rate of conversion of CMC to TDGA in patients treated with IFO, the formation of CMC from IFO can be estimated at approximately 30%. The gradual increase in the excretion of TDGA during a continuous infusion of IFO suggests time-dependent changes in the metabolism of IFO or TDGA. One possible explanation for this phenomenon is an auto-induction of cytochrome P450 during the IFO infusion resulting in a higher rate of N-dechloroethylation and subsequently reflected in an increase in CAA production. A time-dependent increase in CAA generation could explain observed decreases in total cysteine and glutathione reported in plasma during IFO therapy (Lauterburg et al., 1994) and result in increasing amounts of TDGA excreted in urine.

View larger version (11K): [in this window] [in a new window]   Fig. 3.   Proposed scheme of TDGA generation from IFO. For clarity, pathways indicated here represent only those discussed in this work. For more about IFO metabolism see: Sechant et al. (1991) Drugs 42:428-467.

Side effects of IFO therapy have been attributed to mitochondrial dysfunction (Küpfer et al., 1996). Because TDGA, in analogy to other dicarboxylic acids and 3-thia fatty acids, has the potential to impair mitochondrial function, we studied its effect on mitochondrial oxidation of the long chain fatty acid PA, the medium chain fatty acid OA, and SA. The oxidation of PA, but not of OA or SA, was markedly inhibited in vivo. Contrary to previously studied dicarboxylic acids of chain length C₈ to C₁₃ (Passi et al., 1984), our results suggest that -oxidation, tricarboxylic acid cycle function, and the mitochondrial respiratory chain are not affected by TDGA. Specifically, normal oxidation of SA in the presence of TDGA indicates intact Kreb's cycle function and an undisturbed mitochondrial respiratory chain, while normal oxidation of OA allows for the conclusion that depletion or sequestration of CoA cannot account for the inhibitory effect of TDGA observed with PA. Rather, TDGA affects either the formation of palmitoyl-carnitine ester or carnitine translocase. Since -oxidation of PA and OA is almost exclusively initiated in mitochondria (Jacobs et al., 1994), it is unlikely that peroxisomal metabolism of fatty acids influenced the present results.

To support our in vivo findings and further investigate the mechanism of TDGA induced inhibition of long chain fatty acid oxidation, the effect of TDGA on isolated mitochondria was studied. In vitro decreases in oxidative metabolism of PA but not PLC in the presence of TDGA indicate that TDGA effects either the activity of carnitine palmitoyl-transferase I or sequesters carnitine. Normal oxidation of PLC in the presence of TDGA demonstrates intact function of carnitine translocase, carnitine palmitoyltransferase II, -oxidation, the Kreb's cycle, and the respiratory chain.

Taken together, this study has identified and quantified TDGA as a metabolite of the cytostatic drug IFO in humans and demonstrated its inhibitory effects on mitochondrial long chain fatty acid oxidation in vitro and in vivo. The appearance of TDGA in urine, whether it is IFO derived or otherwise, may thus indicate compromised mitochondrial function and signal potentially elevated concentrations of circulating long chain fatty acids. In view of these findings, further investigations on the severity of mitochondrial dysfunction in humans receiving IFO chemotherapy or those chronically exposed to known TDGA producing compounds seems to be important.