Development of selective inhibitors for human aldehyde dehydrogenase 3A1 (ALDH3A1) for the enhancement of cyclophosphamide cytotoxicity

Aldehyde dehydrogenase 3A1 (ALDH3A1) plays an important role in many cellular oxidative processes, including cancer chemoresistance, by metabolizing activated forms of oxazaphosphorine drugs such as cyclophosphamide (CP) and its analogues, such as mafosfamide (MF), ifosfamide (IFM), and 4-hydroperoxycyclophosphamide (4-HPCP). Compounds that can selectively target ALDH3A1 could permit delineation of its roles in these processes and could restore chemosensitivity in cancer cells that express this isoenzyme. Here we report the detailed kinetic and structural characterization of an ALDH3A1-selective inhibitor, CB29, previously identified in a high-throughput screen. Kinetic and crystallographic studies demonstrate that CB29 binds within the aldehyde substrate-binding site of ALDH3A1. Cellular proliferation of ALDH3A1-expressing lung adenocarcinoma (A549) and glioblastoma (SF767) cell lines, as well as ALDH3A1 non-expressing lung fibroblast (CCD-13Lu) cells, is unaffected by treatment with CB29 and its analogues alone. However, sensitivity toward the anti-proliferative effects of mafosfamide is enhanced by treatment with CB29 and its analogue in the tumor cells. In contrast, the sensitivity of CCD-13Lu cells toward mafosfamide was unaffected by the addition of these same compounds. CB29 is chemically distinct from the previously reported small-molecule inhibitors of ALDH isoenzymes and does not inhibit ALDH1A1, ALDH1A2, ALDH1A3, ALDH1B1, or ALDH2 isoenzymes at concentrations up to 250 μM. Thus, CB29 is a novel small molecule inhibitor of ALDH3A1, which might be useful as a chemical tool to delineate the role of ALDH3A1 in numerous metabolic pathways, including sensitizing ALDH3A1-positive cancer cells to oxazaphosphorines.

Evaluation of the "steal" phenomenon on the efficacy of hypoxia activated prodrug TH-302 in pancreatic cancer

Pancreatic ductal adenocarcinomas are desmoplastic and hypoxic, both of which are associated with poor prognosis. Hypoxia-activated prodrugs (HAPs) are specifically activated in hypoxic environments to release cytotoxic or cytostatic effectors. TH-302 is a HAP that is currently being evaluated in a Phase III clinical trial in pancreatic cancer. Using animal models, we show that tumor hypoxia can be exacerbated using a vasodilator, hydralazine, improving TH-302 efficacy. Hydralazine reduces tumor blood flow through the "steal" phenomenon, in which atonal immature tumor vasculature fails to dilate in coordination with normal vasculature. We show that MIA PaCa-2 tumors exhibit a "steal" effect in response to hydralazine, resulting in decreased tumor blood flow and subsequent tumor pH reduction. The effect is not observed in SU.86.86 tumors with mature tumor vasculature, as measured by CD31 and smooth muscle actin (SMA) immunohistochemistry staining. Combination therapy of hydralazine and TH-302 resulted in a reduction in MIA PaCa-2 tumor volume growth after 18 days of treatment. These studies support a combination mechanism of action for TH-302 with a vasodilator that transiently increases tumor hypoxia.

[Identification of glufosfamide metabolites in rats]

AIM

To elucidate the metabolic pathway of glufosfamide in rats.

METHODS

In this study, a liquid chromatography-tandem mass spectrometric method was developed and applied to characterize the metabolites of glufosfamide in rat urine, after an i.v. administration of 50 mg x kg(-1). The analysis was performed under two ionization modes in two different chromatographic systems, separately. To make sure that the compounds detected in rat urine were metabolites or degradation products, the stability of glufosfamide, isophosphoramide mustard (M1), and the degradation products of M1 in urine were investigated.

RESULTS

In positive ionization mode, besides glufosfamide, two metabolites, isophosphoramide mustard and monoaziridinyl derivative of isophosphoramide mustard, were detected. In negative ionization mode, only glufosfamide itself was detected, while derivatives of isophosphoramide mustard have no response in such condition.

CONCLUSION

Glufosfamide was mainly unchanged excreted in urine, and two metabolites were detected as isophosphoramide mustard and monoaziridinyl derivative of isophosphoramide mustard.

Stability of glufosfamide in phosphate buffers and in biological samples

Glufosfamide is a new, potential chemotherapeutic agent currently under investigation. Stability of glufosfamide was investigated in sodium phosphate buffers with different pH and temperature and in biological samples. Glufosfamide and isophosphamide mustard were quantified simultaneously using a liquid chromatography-ion trap mass spectrometric method; precision and accuracy were within 15% for each analyte. Glufosfamide was stable in neutral buffers, but decomposed to form isophosphoramide mustard under acidic and basic conditions, which was pH- and temperature-dependent. The stability of glufosfamide varied in different biological samples. Results indicated that glufosfamide was unstable in some biological samples, such as the small intestine, smooth muscles, pancreas and urine, especially in the small intestine homogenate, with a half-life of 1.1 h. But the pH (<8) and beta-glucosidase of the tissue homogenate was found to have negligible contribution to the degradation of glufosfamide. The enzymatic inhibition experiment with the specific inhibitor, saccharo-1,4-lactone, demonstrated that it was glucuronidase that resulted in the degradation of glufosfamide in small intestine homogenate. Methanol was recommended to be used to homogenize the tissue in an ice water bath, and the container for urine collection should also be maintained in an ice water bath, and all the biological samples collected should be preserved in frozen condition until analysis.

Metabolism and transport of oxazaphosphorines and the clinical implications

The oxazaphosphorines including cyclophosphamide (CPA), ifosfamide (IFO), and trofosfamide represent an important group of therapeutic agents due to their substantial antitumor and immuno-modulating activity. CPA is widely used as an anticancer drug, an immunosuppressant, and for the mobilization of hematopoetic progenitor cells from the bone marrow into peripheral blood prior to bone marrow transplantation for aplastic anemia, leukemia, and other malignancies. New oxazaphosphorines derivatives have been developed in an attempt to improve selectivity and response with reduced toxicity. These derivatives include mafosfamide (NSC 345842), glufosfamide (D19575, beta-D-glucosylisophosphoramide mustard), NSC 612567 (aldophosphamide perhydrothiazine), and NSC 613060 (aldophosphamide thiazolidine). This review highlights the metabolism and transport of these oxazaphosphorines (mainly CPA and IFO, as these two oxazaphosphorine drugs are the most widely used alkylating agents) and the clinical implications. Both CPA and IFO are prodrugs that require activation by hepatic cytochrome P450 (CYP)-catalyzed 4-hydroxylation, yielding cytotoxic nitrogen mustards capable of reacting with DNA molecules to form crosslinks and lead to cell apoptosis and/or necrosis. Such prodrug activation can be enhanced within tumor cells by the CYP-based gene directed-enzyme prodrug therapy (GDEPT) approach. However, those newly synthesized oxazaphosphorine derivatives such as glufosfamide, NSC 612567 and NSC 613060, do not need hepatic activation. They are activated through other enzymatic and/or non-enzymatic pathways. For example, both NSC 612567 and NSC 613060 can be activated by plain phosphodiesterase (PDEs) in plasma and other tissues or by the high-affinity nuclear 3'-5' exonucleases associated with DNA polymerases, such as DNA polymerases and epsilon. The alternative CYP-catalyzed inactivation pathway by N-dechloroethylation generates the neurotoxic and nephrotoxic byproduct chloroacetaldehyde (CAA). Various aldehyde dehydrogenases (ALDHs) and glutathione S-transferases (GSTs) are involved in the detoxification of oxazaphosphorine metabolites. The metabolism of oxazaphosphorines is auto-inducible, with the activation of the orphan nuclear receptor pregnane X receptor (PXR) being the major mechanism. Oxazaphosphorine metabolism is affected by a number of factors associated with the drugs (e.g., dosage, route of administration, chirality, and drug combination) and patients (e.g., age, gender, renal and hepatic function). Several drug transporters, such as breast cancer resistance protein (BCRP), multidrug resistance associated proteins (MRP1, MRP2, and MRP4) are involved in the active uptake and efflux of parental oxazaphosphorines, their cytotoxic mustards and conjugates in hepatocytes and tumor cells. Oxazaphosphorine metabolism and transport have a major impact on pharmacokinetic variability, pharmacokinetic-pharmacodynamic relationship, toxicity, resistance, and drug interactions since the drug-metabolizing enzymes and drug transporters involved are key determinants of the pharmacokinetics and pharmacodynamics of oxazaphosphorines. A better understanding of the factors that affect the metabolism and transport of oxazaphosphorines is important for their optional use in cancer chemotherapy.

Pharmacokinetics of N-2-chloroethylaziridine, a volatile cytotoxic metabolite of cyclophosphamide, in the rat

OBJECTIVES

The objectives of this study were to characterize pharmacokinetics of N-2-chloroethylaziridine (CEA) in the rat model and assess the in vivo fraction of total clearance of phosphoramide mustard (PM) that furnished CEA to circulation.

METHODS

The disposition of CEA was investigated following separate intravenous (iv) administrations of PM, synthetic CEA, and their combination to the Sprague-Dawley rats. In addition, in rats receiving prodrug cyclophosphamide (CP), plasma concentrations of CP and its metabolites, 4-hydroxycyclophosphamide (HOCP), PM, and CEA, were simultaneously quantified using GC/MS and stable isotope dilution techniques.

RESULTS

Following iv administration of synthetic CEA, concentrations of CEA declined biexponentially with the mean terminal half-life and total body clearance of 47.5 min and 167 ml/min/kg, respectively. Urinary excretion of unchanged CEA was 0.164% of the administered dose. CEA was found to be the major circulating metabolite after iv administration of precursor PM to rats. The fraction of total clearance of PM that furnished CEA to circulation was estimated to be 100%, indicating virtually complete availability of the metabolite to circulation once formed. In rats administered with CP, PM exhibited the highest plasma and urinary concentrations compared to HOCP and CEA.

CONCLUSIONS

For the first time, CEA was demonstrated to be an important in vivo metabolite of CP in the present study. In light of the poor permeability and in vivo stability of PM, the ultimate DNA alkylator, the findings obtained in this study suggested that CEA may contribute significantly to the overall antitumor activity of prodrug CP.

Hypersensitivity pneumonitis associated with the use of trofosfamide

Trofosfamide (Ixoten; Baxter Oncology, Germany) is an alkylating agent that, as with other oxazaphosphorine derivatives, has to be activated by hepatic cytochrome P450 oxidases. The bioavailability is nearly 100% after oral application, and the main metabolites are 4-hydroxytrofosfamide, and 4-hydroxyifosfamide. The main side-chain metabolites ifosfamide and cyclophosphamide can be further activated by oxidation and formation of their respective phosphoramide mustards. Oral continuous low-dose therapy has been the most widely used schedule. The toxicity profile consists mainly of dose-dependent hematotoxicity and, rarely, hemorrhagic cystitis. Nausea and vomiting are infrequently seen. Higher grades of nephrotoxicity or neurotoxicity--side-effects that typically limit the use of ifosfamide-have not been reported with low-dose continuous trofosfamide treatment. We report herein a case of a 83-year-old female patient with a disseminated malignant peripheral nerve sheath tumor treated with trofosfamide developing pulmonal toxicity. To our knowledge, this is the first reported case of exogenous allergic alveolitis after exposure to this drug.

1,3- vs 1,5-Intramolecular Alkylation Reactions in Isophosphoramide and Phosphoramide Mustards

It is well-established that at pH 7.4, intramolecular 1,3-N-alkylation reactions in isophosphoramide mustard (IPM) and phosphoramide mustard (PM) produce electrophilic alkylating agents with aziridinyl moieties. To investigate the role of 1,5-intramolecular cyclizations in the chemistry of IPM and PM, the five-membered ring phospholidine products of these reactions were independently synthesized and characterized by (31)P NMR. In 0.33 M BisTris, pH 7.4, 37 degrees C, the intramolecular O-alkylation product of IPM [2-(2-chloroethylamino)-2-tetrahydro-2H-1,3,2-oxazaphospholidine-2-oxide (11)] had a chemical shift of delta 33.0 and a half-life of 3.3 h. The O-alkylation product of PM [2-amino-3-(2-chloroethyl)tetrahydro-2H-1,3,2-oxazaphospholidine-2-oxide (12)] displayed a chemical shift of delta 30.6 and a half-life of 26.9 h. For both IPM and PM, 1,5-N-alkylation provides the same product [1-(2-chloroethyl)-2-hydroxy-tetrahydro-2H-1,3,2-diazaphospholidine-2-oxide (13)]. Because of its instability, 13 was generated in situ and was not isolated; however, the chemical shift (delta 33.0) and reactivity (half-life 0.3 h at 25 degrees C) of the species attributed to 13 were consistent with the assigned structure. Resonances with (31)P NMR chemical shifts indicative of 11 or 12 did not appear in reaction solutions of IPM or PM. The compound assigned as 13 gave hydrolysis products that were not found in reaction solutions of IPM or PM. The collective data supported the conclusion that intramolecular 1,5-alkylations do not contribute to the chemistry of IPM or PM in aqueous solutions at pH 7.4, 37 degrees C. Conversely, 11 and 12 were found to be the major if not exclusive products formed in DMSO solutions of the respective cyclohexylammonium salts of IPM and PM. Both 11 and 12 were relatively noncytotoxic against a series of cell lines, but there were differences in mutagenicities. Chinese hamster ovary cells were exposed to 11 or 12 for one half-life of each compound; 11 was nonmutagenic up to 500 microM, while 12 (500 microM) was mutagenic with 246 mutant colonies/10(6) surviving cells.

1,2-benzisoxazole phosphorodiamidates as novel anticancer prodrugs requiring bioreductive activation

Several 1,2-benzisoxazole phosphorodiamidates have been designed as prodrugs of phosphoramide mustard requiring bioreductive activation. Enzymatic reduction of 1,2-benziosoxazole moiety is expected to result in the formation of imine intermediate due to the cleavage of the N-O bond. The imine should then be spontaneously hydrolyzed to a ketone metabolite, thereby facilitating base-catalyzed beta-elimination of cytotoxic phosphoramide mustard. As expected, the proposed prodrugs 4, 9, and 12 were at least 3-5-fold more potent cytotoxins than control compounds 5 and 15, which lack in the phosphoramide mustard group. Upon incubation with phenobarb-induced rat liver S-9 fraction, compounds 4, 9, and 12 underwent extensive NADPH-dependent metabolism with concomitant generation of alkylating activity under both hypoxic and oxic conditions. Corresponding ketone metabolites were detected for 9 and 15. NADPH-dependent bioreduction of 15 to its ketone metabolite 16 was located in the microsomal fraction and inhibited by SKF-525A and pCMBA. Compared with phenobarb-induced rat liver microsomal fraction, incubation of 15 with rat or human p450 reductase microsomes showed moderate generation of 16. Microsomal cytochrome p450 and/or p450 reductase appear to be involved in the reductive metabolism of 1,2-benzisoxazole moiety under hypoxic as well as oxic conditions.

Glufosfamide (Baxter Oncology)

Glufosfamide is a sugar phosphamide alkylating agent under development by Baxter Oncology (formerly ASTA Medica) as a potential treatment for cancer. By April 2000, glufosfamide had commenced phase II trials, one of which involved intrathecal administration to patients with carcinomatous meningioma.

A mechanism-based pharmacokinetic model for the cytochrome P450 drug-drug interaction between cyclophosphamide and thioTEPA and the autoinduction of cyclophosphamide

Cyclophosphamide (CP) is widely used in high-dose chemotherapy regimens in combination with thioTEPA. CP is a prodrug and is activated by cytochrome P450 to 4-hydroxycyclophosphamide (HCP) which yields the final cytotoxic metabolite phosphoramide mustard (PM). The metabolism of CP into HCP exhibits autoinduction but is inhibited by thioTEPA. The aim of this study was to develop a population pharmacokinetic model for the bioactivation route of CP incorporating the phenomena of both autoinduction and the drug-drug interaction between CP and thioTEPA. Plasma samples were collected from 34 patients who received high-dose CP, thioTEPA and carboplatin in short infusions during 4 consecutive days. Elimination of CP was described by a noninducible route and an inducible route leading to HCP. The latter route was mediated by a hypothetical amount of enzyme. Autoinduction leads to a zero-order increase in amount of this enzyme during treatment. Inhibition by thioTEPA was modeled as a reversible, competitive, concentration-dependent inhibition. PM pharmacokinetics were described by first-order formation from HCP and first-order elimination. The final models for CP, HCP, and PM provided an adequate fit of the experimental data. The volume of distribution, noninducible and initial inducible clearances of CP were 31.0 L, 1.58 L/hr and 4.76 L/hr, respectively. The enzyme amount increased with a zero-order rate constant of 0.041 amount * hr-1. After each thioTEPA infusion, however, approximately 80% of the enzyme was inhibited. This inhibition was reversible with a half-life of 6.5 hr. The formation and elimination rate constants of PM were 1.58 and 0.338 hr-1, respectively. The developed model enabled the assessment of the complex pharmacokinetics of CP in combination with thio TEPA. This model provided an adequate description of enzyme induction and inhibition and can be used for treatment optimization in this combination.

DNA guanine-guanine crosslinking sequence specificity of isophosphoramide mustard, the alkylating metabolite of the clinical antitumor agent ifosfamide

PURPOSE

The purpose of this investigation was to determine the base sequence specificity of isophosphoramide mustard (IPM), the alkylating metabolite of ifosfamide, by crosslinking of designed DNA oligomers in comparison with the clinical alkylating agents mechlorethamine (ME) (nitrogen mustard) and phosphoramide mustard (PM), the alkylating metabolite of cyclophosphamide.

METHODS

IPM, as well as PM and ME were each reacted with three dodecameric duplexes, which were designed to detect interstrand crosslinking between guanines in 5'-GC-3' (I), 5'-GNC-3' (II) or 5'-GNNC-3' (III) sequences (N = A or T).

RESULTS

All three agents preferentially react with 5'-GNC-3' target sequences. The 5'-GNNC-3' target sequence is less reactive by a factor of approximately 2.5- to 10-fold, while 5'-GC-3' is of even lower reactivity.

CONCLUSION

These results indicate that all three agents show approximately equal preference for reaction with a 5'-GNC-3' target sequence in spite of the fact that IPM yields a 7-atom crosslink, while the other two agents yield 5-atom crosslinks.

Three different stable human breast adenocarcinoma sublines that overexpress ALDH3A1 and certain other enzymes, apparently as a consequence of constitutively upregulated gene transcription mediated by transactivated EpREs (electrophile responsive elements) present in the 5'-upstream regions of these genes

ALDH3A1 catalyzes the detoxification of cyclophosphamide, mafosfamide, 4-hydroperoxycyclophosphamide and other oxazaphosphorines. Constitutive ALDH3A1 levels, as well as those of certain other drug-metabolizing enzymes, e.g. NQO1 and CYP1A1, are relatively low in cultured, relatively oxazaphosphorine-sensitive, human breast adenocarcinoma MCF-7 cells. However, transient cellular insensitivity to the oxazaphosphorines can be brought about in these cells by transiently elevating ALDH3A1 levels in them as a consequence of transient exposure to: (1) electrophiles such as catechol that induce the transcription of a battery of genes, e.g. ALDH3A1 and NQO1, having in common an electrophile responsive element (EpRE) in their 5'-upstream regions; or (2) Ah-receptor agonists, e.g. indole-3-carbinol and polycyclic aromatic hydrocarbons such as 3-methylcholanthrene, that induce the transcription of a battery of genes, e.g. ALDH3A1, NQO1 and CYP1A1, having in common a xenobiotic responsive element (XRE) in their 5'-upstream regions. Further, MCF-7 sublines that are constitutively, i.e. when grown in the absence of the original selecting pressure, relatively oxazaphosphorine-insensitive as a consequence of constitutively relatively elevated cellular ALDH3A1 levels evolved when MCF-7 cells were: (1) continuously exposed for several months to gradually increasing concentrations of 4-hydroperoxycyclophosphamide or benz(a)pyrene; or (2) briefly exposed (once for 30 min) to a high concentration (1 mM) of mafosfamide. Each of these three stable sublines is constitutively relatively cross-insensitive to benz(a)pyrene and other polycyclic aromatic hydrocarbons. Cellular levels of NQO1, but not of CYP1A1, are also constitutively relatively elevated in each of the three sublines. RT-PCR-based experiments established that ALDH3A1 mRNA levels are constitutively elevated ( approximately 5- to 8-fold) in each of the three sublines. The elevated ALDH3A1 mRNA levels are not the consequence of gene amplification, hypomethylation of a relevant regulatory element, or ALDH3A1 mRNA stabilization. Collectively, these observations suggest that constitutively elevated levels of ALDH3A1 and certain other enzymes in the three stable sublines are probably the consequence of a constitutive change in the cellular concentration of a key component of the EpRE signaling pathway, such that the cellular concentration of the relevant ultimate transactivating factor is constitutively elevated, i.e. gene transcription promoted by transactivated EpREs is constitutively upregulated. Further, constitutively upregulated gene transcription mediated by transactivated EpREs can be relatively easily induced, whereas that mediated by transactivated XREs cannot, at least in MCF-7 cells. Still further, the three sublines may facilitate study of the signaling pathway that leads to transactivation of the EpREs present in the 5'-upstream regions of ALDH3A1, NQO1 and other gene loci.

Inactivation of aldophosphamide by human aldehyde dehydrogenase isozyme 3

Tumors resistant to chemotherapeutic oxazaphosphorines such as cyclophosphamide often overexpress aldehyde dehydrogenase (ALDH), some isozymes of which catalyze the oxidization of aldophosphamide, an intermediate of cyclophosphamide activation, with formation of inert carboxyphosphamide. Since resistance to oxazaphosphorines can be produced in mammalian cells by transfecting them with the gene for human ALDH isozyme 3 (hALDH3), it seems possible that patients receiving therapy for solid tumors with cyclophosphamide might be protected from myelosuppression by their prior transplantation with autologous bone marrow that has been transduced with a retroviral vector causing overexpression of hALDH3. We investigated whether retroviral introduction of hALDH3 into a human leukemia cell line confers resistance to oxazaphosphorines. This was examined in the polyclonal transduced population, that is, without selecting out high expression clones. hALDH3 activity was 0.016 IU/mg protein in the transduced cells (compared with 2x10(-5) IU/mg in untransduced cells), but there was no detectable resistance to aldophosphamide-generating compounds (mafosfamide or 4-hydroperoxycyclophosphamide). The lack of protection was due, in part, to low catalytic activity of hALDH3 towards aldophosphamide, since, with NAD as cofactor, the catalytic efficiency of homogeneous, recombinant hALDH3 for aldophosphamide oxidation was shown to be about seven times lower than that of recombinant hALDH1. The two polymorphic forms of hALDH3 had identical kinetics with either benzaldehyde or aldophosphamide as substrate. Results of initial velocity measurements were consistent with an ordered sequential mechanism for ALDH1 but not for hALDH3; a kinetic mechanism for the latter is proposed, and the corresponding rate equation is presented.

Pharmacokinetics of cyclophosphamide (CP) and 4-OH-CP/aldophosphamide in systemic vasculitis

Cyclophosphamide given in association with corticosteroids has markedly improved the prognosis of systemic vasculitis. Little information has been reported on cyclophosphamide pharmacokinetics in these diseases and data evaluating its metabolite, 4-hydroxycyclophosphamide/aldophosphamide, pharmacokinetics and concentrations are lacking. Cyclophosphamide was administered as a 1-h intravenous infusion every 3 weeks for six cycles to ten vasculitis patients. Serum cyclophosphamide and 4-hydroxycyclophosphamide/aldophosphamide concentrations were assayed on the first cycle of the treatment by reversed-phase high-pressure liquid chromatography with ultraviolet detection. The mean (+/- SD) 4-hydroxycyclophosphamide/aldophosphamide and cyclophosphamide areas under the serum concentration-time curves were, respectively, 1.86 +/- 1.12 and 154.1 +/- 62.7 mg/L x h with a ratio of 1.30 +/- 0.76%. The mean maximum serum 4-hydroxycyclophosphamide/aldophosphamide was reached 2.3 h after cyclophosphamide administration. The mean (+/- SD) cyclophosphamide and 4-hydroxycyclophosphamide/aldophosphamide half-lives were, respectively, 5.5 +/- 3.1 and 7.6 +/- 2.3 h. The results are consistent with those obtained for cancer patients, in spite of a wide interpatient variability of concentrations and pharmacokinetic parameters.

Biodegradability of antineoplastic compounds in screening tests: influence of glucosidation and of stereochemistry

Some pharmaceuticals such as antineoplastics are carcinogenic, mutagenic, teratogenic and fetotoxic. Antineoplastics and their metabolites are excreted by patients into waste water. In laboratory testing the frequently used isomeric anti-tumour agents cyclophosphamide (CP) and ifosfamide (IF) were shown to be not biodegradable. They are not eliminated in municipal sewage treatment plants and therefore detected in their effluents. Structural related compounds are beta-D-glucosylisophosphoramidmustard (beta-D-Glc-IPM; INN = glufosfamide) and beta-L-glucosylisophosphoramidmustard (beta-L-Glc-IPM). beta-L-Glc-IPM has no antineoplastic effects whereas beta-D-Glc-IPM is active against tumours. In contrast to IF and CP and almost all other investigated antineoplastics beta-D-Glc-IPM is inherently biodegradable. Improved biodegradability of beta-D-Glc-IPM compared to IF shows that reducing the impact of pharmaceuticals on the aquatic environment is feasible by changing the chemical structure of a given compound exerting a similar mode of action and therapeutic activity. Stereochemistry may be crucial for pharmaceutical activity of the compounds as well as for its biodegradability in the environment.

Inhibition of carboxyethylphosphoramide mustard formation from 4-hydroxycyclophosphamide by carmustine

It has been reported that the toxicity of carmustine (BCNU)/cyclophosphamide (CY)/etoposide regimen (when BCNU is split into 4 doses) is less than that of BCNU/CY/cisplatin regimen (when the same amount of BCNU is administered as a single dose). We hypothesized that this might in part be due to the inhibition of aldehyde dehydrogenase 1 (ALDH1) by BCNU or its degradation product, 2-chloroethyl isocyanate, which is likely to be more pronounced at the higher BCNU dose. The effects of BCNU and 2-chloroethyl isocyanate on the formation of carboxyethylphosphoramide mustard (CEPM) from 4-hydroxycyclophosphamide (HCY) was evaluated in human liver cytosol incubations. We found that CEPM formation from HCY was inhibited strongly by BCNU and weakly by 2-chloroethyl isocyanate. The mechanism of inhibition of ALDH1 activity by BCNU was elucidated using indole-3-acetaldehyde (IAL) as the probe substrate in ALDH1 prepared from human erythrocytes. BCNU was a competitive inhibitor of ALDH1 activity with a K(i) of 1.95 microM. The inhibition was independent of preincubation time and reversible by dialysis. The calculated %inhibition of ALDH1 activity by acrolein and BCNU in patients receiving BCNU in 4 split doses with CY was 81%, and it increased to 92% in single dose BCNU regimen. Thus, the calculation indicates that residual operating ALDH1 activity is halved in the presence of single-dose BCNU compared to split-dose BCNU. The inhibition of ALDH1 may contribute to the observed lower incidence of toxicity when BCNU was split into 4 doses compared with single dose and coadministered with CY although dose-dependent effects of BCNU on glutathione and glutathione reductase are also likely to contribute.

Aldehyde dehydrogenase-mediated cellular relative insensitivity to the oxazaphosphorines

As judged by findings in preclinical models, determinants of cellular sensitivity to cyclophosphamide and other oxazaphosphorines include two cytosolic aldehyde dehydrogenases, viz., ALDH1A1 and ALDH3A1. Each catalyzes the detoxification of the oxazaphosphorines; thus, cellular sensitivity to these agents decreases as cellular levels of ALDH1A1 and/or ALDH3A1 increase. Of particular clinical relevance may be that stable sublines, relatively insensitive to the oxazaphosphorines due to elevated ALDH1A1 or ALDH3A1 levels, emerged when cultured human tumor cells were exposed only once to a high concentration of one of these agents for 30 to 60 minutes. Whether differences in cellular levels of either enzyme accounts for the clinically-encountered uneven therapeutic effectiveness of the oxazaphosphorines remains to be determined. However, it has already been established that measurable levels of these enzymes are found in some, but not all, tumor types, and that in those tumor types where measurable levels are present, e.g., infiltrating ductal carcinomas of the breast, they vary widely from patient to patient. Potentially useful clinical strategies that might be pursued if it turns out that ALDH1A1 and/or ALDH3A1 are, indeed, clinically operative determinants of cellular sensitivity to the oxazaphosphorines include 1) individualizing cancer chemotherapeutic regimens based, at least in part, on the levels of these enzymes in the malignancy of interest, and 2) sensitizing tumor cells that express relatively large amounts of ALDH1A1 and/or ALDH3A1 to the oxazaphosphorines by preventing the synthesis of these enzymes, e.g., with antisense RNA, or by introducing an agent that directly inhibits the catalytic action of the operative enzyme. Further, the fact that ALDH1A1 and ALDH3A1 are determinants of cellular sensitivity to the oxazaphosphorines provides the rationale for the investigation of two additional strategies with clinical potential, viz., decreasing the sensitivity of vulnerable and essential normal cells, e.g., pluripotent hematopoietic cells, to the oxazaphosphorines by selectively transferring into them the genetic information that encodes 1) ALDH1A1 or ALDH3A1, or 2) a signaling factor, the presence of which would directly or indirectly, stably upregulate the expression of these enzymes.

Chemical factors in the action of phosphoramidic mustard alkylating anticancer drugs: roles for computational chemistry

The nitrogen mustard based DNA alkylating agents were the first effective anticancer agents and remain important drugs against many forms of cancer. More than fifty years of research on the nitrogen mustards has yielded a broad range of therapeutically useful compounds and a detailed knowledge of the biochemical mechanism of these drugs. Nevertheless, there is much ongoing research on the phosphosphoramidic and other nitrogen mustards to increase their potency and reduce their toxic and mutagenic side effects. To understand the existing nitrogen mustards, and to design the next generation of these drugs, more knowledge is needed about the effects of chemical modifications on their activation and selectivity. Because of the existing knowledge of these drugs, atomic-level chemical modeling can play an important role in the understanding of the phosphoramidic mustard compounds; however, it has not proved straight forward to directly relate the activity of these mustards with simple chemical properties such as bond lengths or atomic charges. Instead, quantum chemical simulations will be required to simulate the activation and alkylation reactions of these compounds, which will require the newest generation of quantum chemical and solvent modeling methods. Additionally, molecular dynamics simulations of the adducted DNA can provide data on the factors favoring crosslinking and its structural consequences. This review summarizes the extensive literature on the metabolism, activation, and action of the phosphoramidic mustards, with an emphasis on the roles that chemical modeling has and will play in the development of this important class of drugs.

Enhanced expressions of glucose-6-phosphate dehydrogenase and cytosolic aldehyde dehydrogenase and elevation of reduced glutathione level in cyclophosphamide-resistant human leukemia cells

Elevation of activity and mRNA level of a cytosolic aldehyde dehydrogenase-1 (ALDH1), which oxidizes aldophosphamide, was previously observed in a cyclophosphamide-resistant murine leukemia cell line. However, changes in other enzyme(s) which may detoxify the drug or produce anti-alkylating agent(s), have not been examined. The human leukemia cell line, K562, was made 30-fold resistant against 4-hydroperoxycyclophosphamide (4HC) by exposing the cells to increasing concentrations of the drug. Resistance against cisplatin was also increased by about 3-fold. Activities of glucose-6-phosphate dehydrogenase (G6PD) and ALDH1 were elevated more than 7-fold in the resistant cells. The mRNA level of the two enzymes was also proportionally elevated. The concentration of reduced glutathione (GSH) was higher in the resistant cells (i.e., 21.1 versus 4.68 nmole per 10(6) cells), while activities of gamma-glutamylcysteine synthetase and glutathione synthetase, and the expressions of other human ALDH genes were not increased in the resistant cells. These findings suggest that the acquired resistance against 4HC is a consequence of transcriptional activation of two genes, i.e., one encoding the G6PD, a major enzyme regenerating anti-alkylating GSH, and the other encoding ALDH1, which has a high activity for oxidation of aldophosphamide derived from 4HC.

Protection by transfected rat or human class 3 aldehyde dehydrogenase against the cytotoxic effects of oxazaphosphorine alkylating agents in hamster V79 cell lines. Demonstration of aldophosphamide metabolism by the human cytosolic class 3 isozyme

Expression of class 3 aldehyde dehydrogenase (ALDH-3) has been associated with acquired or inherent resistance to oxazaphosphorine (OAP) antineoplastic alkylating agents (eg. cyclophosphamide). We previously demonstrated that expression of transfected rat ALDH-3 can confer OAP-specific resistance in human MCF-7 cells (Bunting, K. D., Lindahl, R., and Townsend, A. J. (1994) J. Biol. Chem. 269, 23197-23203). However, the aldophosphamide intermediate inactivated by human class 1 ALDH (hALDH-1) has not proven to be a good substrate for the purified hALDH-3. We have examined the ability of transfected human or rat ALDH-3 to confer OAP resistance in V79/SDl cells. Clones expressing elevated human (386-5938 milliunits/mg) or rat (4-597 milliunits/mg, benzaldehyde/NADP+ substrate) ALDH-3 activity were 1.3- to 12-fold resistant to mafosfamide relative to control cells (<1 milliunit/mg). Resistance was correlated with hALDH-3 activity, and was reversed by pretreatment with the ALDH inhibitor diethylaminobenzaldehyde. Transfectants were cross-resistant to 4-hydroperoxycyclophosphamide and 4-hydroperoxyifosfamide but not to phosphoramide mustard, ifosfamide mustard, melphalan, or acrolein. DNA interstrand cross-links were reduced commensurately with the fold resistance to mafosfamide in the highest activity clone. A key finding was the detection of a metabolite, most likely carboxyphosphamide, that is formed only by cytosols from cells expressing either class 3 or class 1 ALDH.

Synthesis of novel androgen-linked phosphoramide mustard prodrugs and growth-inhibitory activity in human breast cancer cells

Two steroid-linked phosphoramide mustard prodrugs, 7a and 7b, synthesized. The androgens testosterone and 19-nortestosterone were linked through the 17beta-position via an acetal bond to aldophosphamide (3). Proton-catalyzed, as well as cytochrome P450-mediated cleavage of the acetal bond resulted in the release of 3 which decays into the ultimate cytotoxic species, phosphoramide mustard. In a competitive cellular binding assay, the new prodrugs displayed approximately 10-12% affinity to androgen binding proteins in breast cancer cells, relative to testosterone (100%). In the sex hormone receptor-negative cell line MDA-MB231, the testosterone conjugate 7a and the 19-nortestosterone conjugate 7b have been found to be as effective as 4-hydroperoxycyclophosphamide (5). Both compounds were more active than 5 in receptor-positive cell lines. No significant differences in response were observed, however, between receptor-negative and receptor-positive cell lines.

Effect of stereochemistry on the oxidative metabolism of the cyclophosphamide metabolite aldophosphamide

31P NMR and cell perfusion techniques were used to investigate the conversion of the individual enantiomers of aldophosphamide (AP) to carboxyphosphamide (CBP) as catalyzed by aldehyde dehydrogenase in human erythroleukemia K562 cells. R- and S-cyclophosphamides (CPs) were treated with ozone and hydrogen peroxide to yield Rp- and Sp-cis-4-hydroperoxycyclophosphamides (Rp- and Sp-cis-4-HO2-CP); reduction of each hydroperoxide gave the corresponding enantiomer of AP [along with its tautomer 4-hydroxycyclophosphamide (4-HO-CP)]. In separate experiments, K562 cells embedded in agarose gel threads were perfused at pH 7.4, 21 +/- 1 degrees, with solutions of 1.4 mM Rp- and Sp-4-HO-CP/AP, both with and without added mesna (an acrolein scavenger). A comparison of the 31P NMR spectral data derived from the experiments revealed little statistical difference (+/- 10-20% error limits) in the normalized intensities of the CBP peaks arising from the individual AP enantiomers [with added mesna, the ratio Rp-CBP:Sp-CBP was 1.00:1.24 +/- 0.13 (average deviation); without mesna, the same ratio was 1.00:1.35]. Using conventional methods for evaluating the in vitro drug toxicities, CP-resistant L1210 cells were treated in separate experiments with Rp- and Sp-cis-4-HO2-CP; there were no significant differences between the toxicities exhibited by the stereoisomers.

Cyclophosphamide metabolism in children

The alkylating agent cyclophosphamide is a prodrug which is metabolized in vivo to produce both therapeutic and toxic effects. Cyclophosphamide metabolism was investigated in 36 children with various malignancies. Concentrations of cyclophosphamide and its principal metabolites were measured in plasma and urine using a quantitative high-performance TLC method. The results indicated a high degree of inter-patient variation in metabolism. In contrast to previous adult studies on urinary metabolites, plasma carboxyphosphamide concentrations did not support the existence of polymorphic metabolism. Plasma concentrations of dechlorethylcyclophosphamide and carboxyphosphamide were correlated in individual patients, suggesting that the activity of both aldehyde dehydrogenase and cytochrome P450 enzyme(s) determine carboxyphosphamide production in vivo. The presence of ketocyclophosphamide in plasma was strongly associated with dexamethasone pretreatment and was also accompanied by a high clearance of the parent drug. Interpatient differences in metabolism reflect individual levels of enzyme expression and may contribute to variation in clinical effect.

The interaction of glutathione with 4-hydroxycyclophosphamide and phosphoramide mustard, studied by 31P nuclear magnetic resonance spectroscopy

Development of resistance of cancer cells against cyclophosphamide (CP) is probably associated with an increased conjugation with glutathione. 31P NMR spectroscopy was used to monitor the time courses for the chemical conjugation with glutathione of the CP metabolites 4-hydroxycyclophosphamide (4-OHCP) and phosphoramide mustard (PM) at 24 degrees C. PM incubated with a 10-fold molar excess of glutathione showed a disappearance of the PM signal (t1/2 = 112 min), accompanied by an increase of two signals, attributed to the intermediate PM monoglutathione conjugate and the PM diglutathione conjugate. After 680 min, only a signal assigned to the PM diglutathione conjugate was found. This conjugate was relatively stable. The formation of the PM diglutathione conjugate was confirmed with fast atom bombardment mass spectrometry (FAB-MS). The rate constant for the disappearance of the PM signal in incubations with glutathione was 6.2 x 10(-3) min-1, and was 5.4 x 10(-3) min-1 in incubations without glutathione, indicating that the rate-limiting step in both reactions in the formation of aziridinium ions. When 4-OHCP was incubated with a 10-fold molar excess of glutathione, six signals was found which were not present in spectra of incubations without glutathione. In addition to the signals assigned to the mono- and diglutathionyl conjugates of PM, four signals were found of which the pattern of formation in time was identical. These four signals correspond to the four stereoisomers of 4-glutathionylcyclophosphamide (4-GSCP). The formation of 4-GSCP was confirmed with FAB-MS. Within 120 min after the start of the reaction no free 4-OHCP or aldophosphamide signals were found in the spectra. Free PM was detected in all spectra indicating that degradation of 4-GSCP gives rise to PM, the ultimate cytotoxic metabolite of CP, 4-GSCP therefore appears an important pool of phosphoramide mustard, which in turn can be deactivated by glutathione.

Involvement of human glutathione S-transferase isoenzymes in the conjugation of cyclophosphamide metabolites with glutathione

Alkylating agents can be detoxified by conjugation with glutathione (GSH). One of the physiological significances of this lies in the observation that cancer cells resistant to the cytotoxic effects of alkylating agents have higher levels of GSH and high glutathione S-transferase (GST) activity. However, little is known about the GSH-/GST-dependent biotransformation of alkylating agents, including cyclophosphamide. Cyclophosphamide becomes cytostatic after the enzymatic formation of 4-hydroxycyclophosphamide. The ultimate alkylating species formed from cyclophosphamide is phosphoramide mustard. In this paper we describe the involvement of purified human glutathione S-transferases isoenzymes GST A1-1, A2-2, M1a-1a, and P1-1 in the formation of two types of glutathionyl conjugates of cyclophosphamide, i.e., 4-glutathionylcyclophosphamide (4-GSCP) and monochloromonoglutathionylphosphoramide mustard. When 0.1 mM 4-hydroxycyclophosphamide and 1 mM GSH was incubated in the presence of 10 microM GST A1-1, A2-2, M1a-1a, and P1-1 the formation of 4-GSCP was 2-4-fold increased above the spontaneous level. Enzyme kinetic analysis demonstrated the lowest Km (0.35 mM) for GST A1-1. Km values for the other GST enzymes ranged from 1.0 to 1.9 mM. Glutathione S-transferase A1-1 (40 microM) also increased the conjugation of phosphoramide mustard and GSH (both 1 mM) 2-fold, while the other major human isoenzymes, A2-2, M1a-1a, and P1-1, did not influence the formation of monochloromonoglutathionylphosphoramide mustard. These results indicate that only one enzyme within the class of human GST alpha enzymes was able to catalyze the reaction of the aziridinium ion of phosphoramide mustard with glutathione. Thus increased levels of GST A1-1 in tumor cells can contribute to an enhanced detoxification of phosphoramide mustard and hence to the development of drug resistance. Since all of the human GSTs tested did catalyze the formation of 4-GSCP, the role of 4-GSCP either as a transport form of activated cyclophosphamide or as a detoxification product is discussed.

Detoxification of cyclophosphamide by human aldehyde dehydrogenase isozymes

In in vitro studies, no turnover of aldophosphamide and mafosfamide was observed with the tumor-specific aldehyde dehydrogenase 3 isozyme (ALDH3) isolated from human stomach mucosa as well as from lung (A549) and pharynx (UMSCC2) carcinoma cell lines. Only the human liver cytosolic ALDH preparation (ALDH1) showed any significant oxidation of aldophosphamide and mafosfamide.

Direct detection of the intracellular formation of carboxyphosphamides using nuclear magnetic resonance spectroscopy

31P nuclear magnetic resonance (NMR) spectroscopy was used in conjunction with cell perfusion techniques to monitor the intracellular chemistry of the cyclophosphamide (CP, CAS 6055-19-2) metabolites 4-hydroxy-cyclophosphamide (4-HO-CP) and aldophosphamide (AP) in U937 human histiocytic (CP-sensitive) and K562 human erythroleukemia (CP-resistant) cells. Similar experiments were carried out using the ifosfamide (IF, CAS3778-73-2) metabolites 4-hydroxyifosfamide (4-HO-IF) and aldoifosfamide (AIF). The hydroxy and aldehydic metabolites were generated by the triphenylphosphine reduction of 4-hydroperoxycyclophosphamide (4-HO2-CP) or 4-hydroperoxyifosfamide (4-HO2-IF) or by a spontaneous elimination/addition reaction involving water and 4-thiocyclophosphamide analogs 4-(2-hydroxyethyl) thiocyclophosphamide (4-ESCP) or mafosfamide. Cell death resulting from 4-HO-CP/AP perfusions was mimicked by perfusion with acrolein or an acrolein producing but non-alkylating, dechloro-CP analog. Acrolein toxicity was minimized by the presence of 2-mercaptoethanol or mesna (sodium 2-mercaptoethanesulfonate) in perfusion solutions as well as by fractional dose drug perfusions (sequential 2.5-3.0 h perfusions separated by cell washes with drug-free medium). The intracellular half-life for phosphoramide mustard (PM) at an intracellular pH value of 7.1 +/- 0.1 and an ambient probe temperature of 23 +/- 1 degree C in U937 cells was 2.1 h [k = (5.4 +/- 0.3) x 10(-3) min-1] and in K562 cells was 3.1 h [k = (3.7 +/- 0.4) x 10(-3) min-1]. Similar half-lives (2-4 h) were determined for intracellular isophosphoramide mustard (IPM). Fractional dose perfusion of U937 or K562 cells with 1.5 mmol/l 4-HO-CP/AP (generated from 4-HO2-CP) and 0.3 mmol/l mesna allowed for the observation of intracellular carboxyphosphamide (CBP); CBP was formed in higher concentrations in the CP-resistant K562 cells. Similar results were obtained using 4-ESCP and mafosfamide as sources of 4-HO-CP/AP. Identification of CBP was based on chemical shift, chemical stability, and membrane permeability studies of synthetic CBP. Concentrations of carboxyifosfamide (CBIF) formed in K562 cells were also greater than that in U937 cells.

NADPH-dependent enzyme-catalyzed reduction of aldophosphamide, the pivotal metabolite of cyclophosphamide

One of the metabolites found in the urine of mammals given the prodrug cyclophosphamide is alcophosphamide, an alcohol. It is most probably generated from cyclophosphamide via aldophosphamide, an aldehyde which otherwise can directly give rise to phosphoramide mustard; the latter effects the cytotoxic action of cyclophosphamide and other oxazaphosphorines. It has already been demonstrated that horse liver alcohol dehydrogenase catalyzes the reduction of aldophosphamide to alcophosphamide. Herein, we report that aldose reductase and aldehyde reductase purified from human placenta also catalyze this reaction. The Km values for aldose reductase- and aldehyde reductase-catalyzed reduction of aldophosphamide to alcophosphamide were 0.15 and 1.6 mM, respectively. Aldose reductase and aldehyde reductase accounted for 94 and 6%, respectively, of total placental pyridine nucleotide-dependent enzyme-catalyzed aldophosphamide (160 microM) reduction. Aldose reductase-catalyzed reduction of aldophosphamide appeared to be noncompetitively inhibited by sorbinil; the Ki value was 0.4 microM. The in vivo significance of these observations is uncertain but could be of some magnitude since alcophosphamide is known to be only weakly cytotoxic.

Sensitivity of aldehyde dehydrogenases in murine tumor and hematopoietic progenitor cells to inhibition by chloral hydrate as determined by the ability of chloral hydrate to potentiate the cytotoxic action of mafosfamide

Several murine aldehyde dehydrogenases, most notably AHD-2, are known to catalyze the detoxification of cyclophosphamide, mafosfamide, and other oxazaphosphorines. Thus, cellular sensitivity to these agents decreases as the relevant aldehyde dehydrogenase activity increases, and vice versa. Chloral hydrate is a sedative/hypnotic agent that is sometimes administered to patients being treated with cyclophosphamide. It is known to inhibit some, but not all, aldehyde dehydrogenases. Murine (CFU-S, CFU-GEMM and CFU-Mk) and human (CFU-Mix, CFU-GM, BFU-E and CFU-Mk) hematopoietic progenitor cells, as well as murine oxazaphosphorine-resistant (L1210/OAP and P388/CLA) tumor cells, are known to contain the relevant aldehyde dehydrogenase activity but the identity of the specific enzyme present in the normal cells is unknown and may be different than that, namely AHD-2, present in neoplastic cells. In that event, the potential exists to inhibit the detoxification of the oxazaphosphorines in tumor cells without inhibiting this event in normal cells; the net effect of such a selective inhibition would be to increase the margin of safety of the oxazaphosphorines. In ex vivo experiments, chloral hydrate markedly potentiated the antitumor activity of mafosfamide against oxazaphosphorine-resistant L1210/OAP and P388/CLA cells. It did not potentiate the cytotoxic action of mafosfamide against any of the murine or human hematopoietic cells tested, even at concentrations which fully restored the sensitivity of the resistant tumor cell lines to this agent. One explanation for these observations is that hematopoietic progenitor, and the resistant tumor, cells express different relevant aldehyde dehydrogenases and that these aldehyde dehydrogenases differ in their sensitivity to inhibition by chloral hydrate. Consistent with this notion were the observations that AHD-2 was exquisitely sensitive to inhibition by chloral hydrate, whereas two other aldehyde dehydrogenases that also catalyze the detoxification of aldophosphamide, namely AHD-12a, b and AHD-13, were relatively unaffected.

Role of glutathione in cellular resistance to alkylating agents

Both elevated glutathione levels and increased activity of the enzyme glutathione S-transferase have been associated with the resistance of cells to alkylating agents. We have demonstrated that one mechanism of this resistance is the inactivation of the alkylating agents by conjugation with glutathione. This conjugation can be catalyzed by glutathione S-transferase. For the nitrogen mustard agents we have studied, both the spontaneous and enzyme catalyzed reactions proceed through the aziridinium intermediates of the alkylating agents, and the alpha isoenzymes of GST are involved. In a study of cyclophosphamide resistant medulloblastoma cell lines elevated cellular concentrations of glutathione correlated well with the resistance of the cell lines.

High-performance tandem mass spectrometry in metabolism studies

1. High-performance tandem mass spectrometry provides unit resolution in both selection of precursor ions and analysis of fragment ions, and extensive and reproducible fragmentation through collisional activation at high energy. 2. Metabolites can be analysed that occur as minor components in h.p.l.c. peaks or other mixtures. Homogeneous isotopic species can be selected for unambiguous analysis of distributions of isotope labels. Fragmentation may be significantly enhanced to provide structural information. Overall, the signal to noise ratio is greatly improved and the spectrum is simplified. 3. These points are illustrated by isotope-labelling studies of the mechanisms of glutathione conjugation of the anti-tumour agent cyclophosphamide, the cytotoxic agent phosphoramide mustard and dimethylbilirubin, an analogue of bilirubin designed to be distinguishable from endogenous bilirubin. Analysis of isomeric mixed disulphides formed between glutathione and a peptide with an internal disulphide bond is discussed. 4. Reaction-induced decomposition is presented as an alternative to collisionally induced decomposition with more efficient energy transfer.

Identification of human liver aldehyde dehydrogenases that catalyze the oxidation of aldophosphamide and retinaldehyde

Biotransformation of the biologically and pharmacologically important aldehydes, retinaldehyde and aldophosphamide, is mediated, in part, by NAD(P)-dependent aldehyde dehydrogenases catalyze the oxidation of the aldehydes to their respective acids, retinoic acid and carboxyphosphamide. Not known at the onset of this investigation was which of the several known human aldehyde dehydrogenases (ALDHs) catalyze these reactions. Thus, human liver aldehyde dehydrogenases were chromatographically resolved and the ability of each to catalyze the oxidation of retinaldehyde and aldophosphamide was assessed. Only one, namely ALDH-1, catalyzed the oxidation of retinaldehyde; the Km value was 0.3 microM. Three, namely ALDH-1, ALDH-2 and succinic semialdehyde dehydrogenase, catalyzed the oxidation of aldophosphamide; Km values were 52, 1193, and 560 microM, respectively. ALDH-4, ALDH-5 and betaine aldehyde dehydrogenase did not catalyze the oxidation of either aldophosphamide or retinaldehyde. ALDH-1 and succinic semialdehyde dehydrogenase accounted for 64 and 30%, respectively, of the total hepatic aldehyde dehydrogenase-catalyzed aldophosphamide (160 microM) oxidation. ALDH-1-catalyzed oxidation of aldophosphamide was noncompetitively inhibited by chloral hydrate; the Ki value was 13 microM. ALDH-2- and succinic semialdehyde dehydrogenase-catalyzed oxidation of aldophosphamide was relatively insensitive to inhibition by chloral hydrate. These observations strongly suggest an important in vivo role for ALDH-1 in the catalysis of retinaldehyde and aldophosphamide biotransformation. Succinic semialdehyde dehydrogenase-catalyzed biotransformation of aldophosphamide may also be of some in vivo importance.

Synthesis and antitumor properties of activated cyclophosphamide analogues

A series of 5- and 6-substituted cyclophosphamide analogues has been prepared, and their 31P NMR kinetics of phosphoramide mustard (PDA) release and in vitro and in vivo cytotoxicity have been evaluated. cis-4-Hydroxy-5-methoxycyclophosphamide equilibrated very slowly and to a minor extent with the ring-opened aldophosphamide analogues in aqueous buffer; release of PDA was observed to a minor extent and only at high (1 M) buffer concentrations. This analogue was essentially inactive in vitro against L1210 and P388 leukemia cells. 6-Phenylcyclophosphamide and its 4-hydroperoxy derivative were potent inhibitors of blood acetylcholinesterase and were lethal at therapeutic doses in mice. In contrast, 4-hydroperoxy-6-(4-pyridyl)cyclophosphamide did not inhibit acetylcholinesterase and showed significant antitumor activity in vitro and in vivo against both wild-type and cyclophosphamide-resistant L1210 leukemia. The 4-hydroperoxy-6-arylcyclophosphamides were generally active in vitro against both wild-type and cyclophosphamide-resistant L1210 and P388 cells, and several analogues showed significant activity in vivo. Surprisingly, there was no correlation between antitumor activity in vitro and the rate of PDA release in aqueous buffer. Several compounds that showed essentially no release of PDA in aqueous buffer over several hours were highly cytotoxic to leukemia cells following a 1-h exposure in vitro. These results show that activated cyclophosphamide analogues substituted at the 6-position are not cross-resistant in these leukemia cell lines, and that a specific intracellular activation mechanism may be catalyzing PDA release in these analogues.

Potentiation of the cytotoxic action of mafosfamide by N-isopropyl-p-formylbenzamide, a metabolite of procarbazine

Several mouse aldehyde dehydrogenases catalyze the detoxification of aldophosphamide, the pivotal metabolite of the prodrugs cyclophosphamide, mafosfamide, and other oxazaphosphorines. N-Isopropyl-p-formylbenzamide, a major metabolite of procarbazine, was found to be an excellent substrate (Km = 0.84 microM) for at least one of these enzymes, namely, mouse aldehyde dehydrogenase-2. The Km for mouse aldehyde dehydrogenase-2-catalyzed detoxification of aldophosphamide is 16 microM. Thus, competition between N-isopropyl-p-formylbenzamide and aldophosphamide for the catalytic site on the enzyme should strongly favor the former, and the rate at which aldophosphamide is detoxified should be markedly retarded. Mouse L1210/OAP and P388/CLA leukemia cells are relatively insensitive to the oxazaphosphorines because they contain large amounts of mouse aldehyde dehydrogenase-2. As predicted, N-isopropyl-p-formylbenzamide markedly potentiated the cytotoxic action of mafosfamide against these cells. Mouse L1210/0 and P388/0 lack the enzyme. Again as expected, N-isopropyl-p-formylbenzamide essentially did not potentiate the cytotoxic action of mafosfamide against these cells. Certain mouse and human hematopoietic progenitor cells also contain an aldehyde dehydrogenase that catalyzes the detoxification of aldophosphamide, but the specific identity of this enzyme remains to be established. N-Isopropyl-p-formylbenzamide potentiated the cytotoxic action of mafosfamide against these cells as well. Clinically, procarbazine and the oxazaphosphorines are used to treat certain neoplastic diseases. Frequently, they are used in combination. Our findings demonstrate the potential for both desirable and undesirable drug interactions when these agents are used concurrently. Similar drug interactions can be expected when other substrates for, or inhibitors of, the relevant aldehyde dehydrogenases, e.g., chloramphenicol, chloral hydrate, and methyltetrazolethiol-containing cephalosporins, are co-administered with the oxazaphosphorines.

Glutathione conjugation with phosphoramide mustard and cyclophosphamide. A mechanistic study using tandem mass spectrometry

The conjugations of cyclophosphamide and of phosphoramide mustard with glutathione are shown to be catalyzed by hepatic cytosolic glutathione-S-transferases. Cyclophosphamide conjugation is also catalyzed by microsomal glutathione-S-transferases, both in intact microsomes and after solubilization and immobilization. Deuterium isotope labels are used to test whether chloride is directly displaced by glutathione in the enzyme-catalyzed conjugations, or whether conjugation takes place via symmetrical cyclic aziridinium ions. Tandem mass spectrometry with high energy collisional activation is shown to provide reliable analysis of the isotope-labeling patterns in the conjugated products. This experiment leads to the conclusion that the aziridinium ion is opened in the conjugation of phosphoramide mustard in both the enzyme-catalyzed and the chemical reactions. Cyclophosphamide, on the other hand, is shown to be conjugated through direct displacement of chloride.

Pharmacokinetics of 4-hydroxycyclophosphamide and metabolites in the rat

Pharmacokinetics of 4-hydroxycyclophosphamide (4-OHCP), the major active microsomal metabolite of cyclophosphamide (CP), were investigated in the Sprague-Dawley rat following separate iv bolus administrations of CP, synthetic 4-OHCP, and their combination. CP, 4-OHCP, and other metabolites such as phosphoramide mustard, alcophosphamide, and 3-(2-chloroethyl)-1,3-oxazolidin-2-one in rat plasma were simultaneously analyzed using GC/MS and stable isotope dilution techniques. Following iv administrations of 4-OHCP to rats at doses of 10 mg/kg, 20 mg/kg, and 40 mg/kg, phosphoramide mustard was found to be the major circulating metabolite followed by alcophosphamide and 3-(2-chloroethyl)-1,3-oxazolidin-2-one. No appreciable amount of nor-nitrogen mustard was detected in circulation. The mean excretion of unchanged 4-OHCP in two rats was 4.1 +/- 0.2% of the administered dose in 24-hr urine. Plasma concentration-time profiles of 4-OHCP declined monoexponentially with a mean half-life and total plasma clearance values of 8.1 min and 81.2 ml/min.kg, respectively. No dose-dependent kinetics were apparent between the 10 and 40 mg/kg doses used. The short half-life of 4-OHCP may be due partly to its degradation in plasma, which was found to be a first-order process in vitro with a half-life of 5.2 min. On the other hand, when CP was administrated to eight separate rats at 20 mg/kg each, the mean apparent elimination half-life for the derived 4-OHCP was 55.4 min as compared to 58.2 min, the mean elimination half-life for the parent drug.(ABSTRACT TRUNCATED AT 250 WORDS)

Identification of the mouse aldehyde dehydrogenases important in aldophosphamide detoxification

Aldophosphamide, the penultimate cytotoxic metabolite of cyclophosphamide, can be detoxified by an oxidation reaction catalyzed by certain aldehyde dehydrogenases. The selective toxicity of cyclophosphamide is due, at least in part, to a greater expression of the relevant aldehyde dehydrogenase activity in normal cells relative to that expressed in certain tumor cells. Not known at the onset of this investigation was which of the several known mouse aldehyde dehydrogenases catalyze this reaction. Twelve enzymes that catalyze the NAD(P)-linked oxidation of aldophosphamide, acetaldehyde, benzaldehyde, and/or octanal were chromatographically resolved from mouse liver. Four of these appear to be novel; four others were determined to be betaine aldehyde dehydrogenase, succinic semialdehyde dehydrogenase, glutamic gamma-semialdehyde dehydrogenase, and xanthine oxidase (dehydrogenase). An additional aldehyde dehydrogenase, namely AHD-4, was semipurified from stomach. The stomach enzyme and nine of the hepatic enzymes catalyze the oxidation of aldophosphamide. Km values for these reactions range from 16 microM to 2.5 mM. The relevant aldehyde dehydrogenase of major importance varies with the tissue. In the liver, the major cytosolic aldehyde dehydrogenase, namely AHD-2, accounts for greater than 60% of total hepatic aldehyde dehydrogenase-catalyzed aldophosphamide (160 microM) detoxification. Succinic semialdehyde dehydrogenase (AHD-12) and three of the novel hepatic aldehyde dehydrogenases, namely AHD-8, AHD-10, and AHD-13, also contribute significantly to total hepatic aldehyde dehydrogenase-catalyzed aldophosphamide detoxification. In the stomach, AHD-4 and AHD-8 account for approximately 86% of total aldehyde dehydrogenase-catalyzed aldophosphamide (160 microM) detoxification. AHD-2 was not found in this tissue. Of all the aldehyde dehydrogenases examined, AHD-2 and AHD-8 were estimated to be the most efficient catalysts of aldophosphamide oxidation. Thus, these enzymes would seem most likely to be operative when tumor cells acquire aldehyde dehydrogenase-mediated cyclophosphamide resistance.

Formation of a phosphoramide mustard-nucleotide adduct that is not by alkylation at the N7 position of guanine

The reaction of 2'-deoxyguanosine 3'-monophosphate with phosphoramide mustard resulted in the formation of several adducts. One of these adducts was formed by linking phosphoramide mustard to the phosphate group of 2'-deoxyguanosine 3'-monophosphate rather than by the generally accepted mechanism involving alkylation at the N7 position of guanine. This adduct served as an acceptor for the transfer of 32p from [gamma 32P]ATP by polynucleotide kinase and thus could be detected by the sensitive 32p-postlabeling assay.

Proton-mediated liberation of aldophosphamide from a nontoxic prodrug: a strategy for tumor-selective activation of cytocidal drugs

Based on the findings that the pH in malignant tumors can be preferentially decreased by stimulation of their aerobic glycolysis, acid-sensible prodrugs, which are nearly nontoxic at physiological pH, were synthesized. At lower pH, however, these compounds are cleaved with liberation of a cytotoxic species. The prototypic drug compound 2-hexenopyranoside of aldophosphamide was prepared, which releases aldophosphamide by acid-catalyzed hydrolysis. Exposure of cultured M1R rat mammary carcinoma cells to this agent at pH 7.4 only resulted in slight toxicity. However, when drug treatment was performed at pH 6.2, the mean pH in malignant tumors of hyperglycemic hosts, the colony-forming fraction of M1R cells decreased to 0.05 and 0.0001 of controls treated at pH 7.4 after exposure for 24 h and 48 h, respectively. The synthesis of the 2-hexenopyranoside of aldophosphamide is described in detail.

Rapid development of enhanced clearance after high-dose cyclophosphamide

We have studied the disposition of cyclophosphamide, its major cytotoxic metabolite phosphoramide mustard, and the synthetic glucocorticoid dexamethasone in nine patients receiving high-dose cyclophosphamide daily for 2 days before bone marrow transplantation. The total body clearance of cyclophosphamide was observed to increase from 93 +/- 27 ml/min on the first day to 178 +/- 83 ml/min on the second day. This was associated with an increase in the clearance of dexamethasone from 369 +/- 104 ml/min to 526 +/- 123 ml/min. An increased rate of formation of phosphoramide mustard with higher peak concentrations was also seen. Simulation studies show that these changes are most likely the result of an increase in the hepatic metabolism of cyclophosphamide. These results show that high-dose cyclophosphamide causes an increase in its own clearance and that of dexamethasone through an apparent induction of hepatic-metabolizing enzymes detectable 24 hours after initial exposure to cyclophosphamide.

Kinetic characterization of the catalysis of "activated" cyclophosphamide (4-hydroxycyclophosphamide/aldophosphamide) oxidation to carboxyphosphamide by mouse hepatic aldehyde dehydrogenases

A spectrophotometric assay was developed and utilized to directly characterize aldehyde dehydrogenase-catalyzed oxidation of aldophosphamide to carboxyphosphamide by soluble and solubilized particulate fractions prepared from mouse liver homogenates. Vmax values of 3310 and 1170 nmol/min/g liver were obtained for the soluble and solubilized particulate fractions respectively. Km values were 22 and 84 microM respectively. Alkaline pH optimums were observed in each case. Aldehyde dehydrogenase-catalyzed oxidation of aldophosphamide by the soluble fraction was markedly more temperature responsive. Catalysis of aldophosphamide and acetaldehyde or benzaldehyde oxidation was apparently by the same isozyme(s) in the soluble fraction. Similarly, low Km (acetaldehyde/benzaldehyde) and high Km (acetaldehyde/benzaldehyde) isozymes each apparently catalyzed the oxidation of aldophosphamide in the solubilized particulate fraction. Our findings suggest that (1) oxidation of aldophosphamide to carboxyphosphamide by mouse liver is catalyzed largely by the predominant aldehyde dehydrogenase isozyme present in the soluble fraction (cytosol) of this tissue, and (2) isozymes that catalyze aldophosphamide oxidation are not different from those that catalyze the oxidation of acetaldehyde and benzaldehyde, though the relative contribution of each isozyme within the solubilized particulate fraction to the catalysis of aldophosphamide oxidation remains to be determined.

Accelerated decomposition of 4-hydroxycyclophosphamide by human serum albumin

Cyclophosphamide, a widely used anticancer agent, requires initial metabolic activation to 4-hydroxycyclophosphamide (4-OHCP) to elicit its activity. The rate of decomposition of cis-4-OHCP was much faster in plasma than in buffer at pH 7.4. This plasma activity was not affected by treatment with acid (pH 1.3) or heat (60 degrees C for 30 min). The activity was retained in the macromolecular fraction (greater than 10,000) but not in the filtrate. Serum albumin was identified as the catalyst for the elimination step that generates phosphoramide mustard from aldophosphamide; albumin had no effect on the rate of ring opening of cis-4-OHCP to aldophosphamide. This catalytic activity was dependent on serum albumin concentration and independent of pH over the range of 6.5 to 7.5, in contrast to the buffer-catalyzed reaction. The catalytic rate constants kcat (pH 7.4, 37 degrees C) for phosphate buffer, human serum albumin, and bovine serum albumin were 1.13, 285, and 83 M-1 min-1, respectively. Pretreatment of cis-4-OHCP with serum albumin resulted in a time-dependent decrease in cytotoxic activity against L1210 tumor cells in vitro. These data suggest that the albumin-catalyzed reaction of cis-4-OHCP in plasma represents an important pathway for the transformation of cyclophosphamide metabolites and further emphasize the importance of considering phosphoramide mustard generated extracellularly versus intracellularly and the respective contributions of extracellular and intracellular phosphoramide mustard to cyclophosphamide cytotoxicity in vivo.

Binding of metabolites of cyclophosphamide to DNA in a rat liver microsomal system and in vivo in mice

The stability of phosphoramide mustard, a metabolite of cyclophosphamide was studied at pH 7.2 and 37 degrees C using 31P nuclear magnetic resonance. The phosphorus signal of phosphoramide mustard disappeared with a half-life of 8 min indicating rapid conversion to other species. The final product, inorganic phosphate, appeared with a half-life of 105 min indicating that phosphoramide mustard was easily dephosphoramidated. A rat liver microsomal system was used to study the binding of [chloroethyl-3H]cyclophosphamide to DNA. DNA was hydrolyzed in 0.1 N HCl:0.5 N NaCl at 80 degrees C for 20 min, conditions known to convert phosphoramide mustard to nornitrogen mustard with liberation of the phosphoramide residue. After such treatment three adducts were detected by high-performance liquid chromatography using several elution systems. They were all 7-substituted guanine adducts of nornitrogen mustard; two were monoalkylation products with an intact [N-(2-chloroethyl)-N-[2-(7-guaninyl)ethyl]amine] or an hydroxylated mustard arm [N-(2-hydroxyethyl)-N-[2-(7-guaninyl)ethyl]amine]; the third adduct was a cross-linked product [N,N-bis [2-(7-guaninyl)ethyl]-amine]. The relative abundance of these adducts depended on the length of the microsomal incubation. After 2 h, N-(2-chloroethyl)-N-[2-(guaninyl)ethyl]amine was the main product but after 6 h N-(2-hydroxyethyl)-N-[2-(7-guaninyl)ethyl]amine was most abundant, and at this time the cross-linked product represented 12% of the total adducts. The adducts in DNA depurinated readily and after 24 h at pH 7.0 and 37 degrees C 70% of them had been liberated. The rate of depurination was decreased in the presence of 0.5 N NaCl. After short-term depurination in 0.1 N HCl at 25 degrees C the primary alkylating species was phosphoramide mustard rather than nornitrogen mustard. In in vivo studies mice were given injections i.p. of 100 microCi of cyclophosphamide. Maximal levels of radioactivity had been incorporated into DNA between 2-7 h after injection; the specific activity of DNA from the kidney and lung exceeded that from the liver. While the level of radioactivity found in kidney DNA was rapidly reduced the rate of fall was lower in the lung. Between 24 and 72 h the specific activity of lung DNA exceeded that of kidney and liver DNA by a factor of 3:8. Lung is the principal target tissue for tumor formation in mice after an i.p. injection.

Half-life of oxazaphosphorines in biological fluids

Plasma AUC and half-life values for cyclophosphamide were determined in rats manipulated to hydroxylate cyclophosphamide at different rates; plasma AUC and apparent half-life values for two pharmacologically important metabolites of cyclophosphamide, viz. 4-hydroxycyclophosphamide/aldophosphamide and phosphoramide mustard, were also determined in these animals. Apparent plasma half-life values for 4-hydroxycyclophosphamide/aldophosphamide and phosphoramide mustard increased with an increase in plasma half-life values for cyclophosphamide. AUC values for cyclophosphamide increased approximately linearly with an increase in its plasma half-life but AUC values for 4-hydroxycyclophosphamide/aldophosphamide and phosphoramide mustard remained approximately constant with an increase in their respective apparent plasma half-life values. Given that the cytotoxic effects of cyclophosphamide are directly proportional to AUC values for 4-hydroxycyclophosphamide/aldophosphamide and/or phosphoramide mustard, we conclude that changes in the rate of cyclophosphamide hydroxylation will not alter the systemic toxic and therapeutic responses to a given dose of cyclophosphamide. Actual half-life values for 4-hydroxycyclophosphamide/aldophosphamide and phosphoramide mustard after the iv infusion of these agents were also determined. A comparison of the actual plasma half-life values for cyclophosphamide (29 min), 4-hydroxycyclophosphamide/aldophosphamide (14 min), and phosphoramide mustard (14 min) with the apparent plasma half-life values obtained for 4-hydroxycyclophosphamide/aldophosphamide (34 min) and phosphoramide mustard (55 min) following cyclophosphamide administration suggests that the major determinant with regard to the apparent plasma half-life of 4-hydroxycyclophosphamide/aldophosphamide is its rate of formation whereas in the case of phosphoramide mustard, an additional determinant, perhaps efflux from the cell, is operative.(ABSTRACT TRUNCATED AT 250 WORDS)

NMR spectroscopic studies of intermediary metabolites of cyclophosphamide. A comprehensive kinetic analysis of the interconversion of cis- and trans-4-hydroxycyclophosphamide with aldophosphamide and the concomitant partitioning of aldophosphamide between irreversible fragmentation and reversible conjugation pathways

Multinuclear (31P, 13C, 2H, and 1H) Fourier-transform NMR spectroscopy, with and without isotopically enriched materials, was used to identify and quantify, as a function of time, the following intermediary (short-lived) metabolites of the anticancer prodrug cyclophosphamide (1, Scheme I): cis-4-hydroxycyclophosphamide (cis-2), its trans isomer (trans-2), aldophosphamide (3), and its aldehyde-hydrate (5). Under a standard set of reaction conditions (1 M 2,6-dimethylpyridine buffer, pH 7.4, 37 degrees C), the stereospecific deoxygenation of synthetic cis-4-hydroperoxycyclophosphamide (cis-12, 20 mM) with 4 equiv of sodium thiosulfate (Na2S2O3) afforded, after approximately 20 min, a "pseudoequilibrium" distribution of cis-2, 3, 5, and trans-2, i.e., the relative proportions of these reactants (57:4:9:30, respectively) remained constant during their continual disappearance. NMR absorption signals indicative of "iminophosphamide" (8) and enol 6 were not detected (less than 0.5-1% of the synthetic metabolite mixture). A computerized least-squares fitting procedure was applied to the individual 31P NMR derived time courses for conversion of cis-2, 3 plus 5 (i.e., "3"), and trans-2 into acrolein and phosphoramide mustard (4), the latter of which gave an expected array of thiosulfate S-alkylation products (e.g., 16) and other phosphorus-containing materials derived from secondary decomposition reactions. This kinetic analysis gave the individual forward and reverse rate constants for the apparent tautomerization processes, viz., cis-2 in equilibrium "3" in equilibrium trans-2, as well as the rate constant (k3) for the irreversible fragmentation of 3. The values of k3 at pH 6.3, 7.4, and 7.8 were equal to 0.030 +/- 0.004, 0.090 +/- 0.008, and 0.169 +/- 0.006 min-1, respectively. Replacement of the HC(O)CH2 moiety n 3 with HC(O)CD2 led to a primary kinetic isotope effect (kH/kD = 5.6 +/- 0.4) for k3. The apparent half-lives (tau 1/2) for cis-2, "3", and trans-2 under the standard reaction conditions, at "pseudoequilibrium" (constant ratio of cis-2/"3"/trans-2), were each equal to approximately 38 min, which is considerably shorter than the widely cited colorimetrically derived half-lives reported by earlier investigators. The values of tau 1/2 for cis-2, "3", and trans-2 were affected by pH in the same manner as that found for k3 but were relatively insensitive to the presence of either K+, Na+, Ca2+, or Mg2+. The presence of certain primary amines led to marked decreases in tau 1/2 and, in some cases, the formation of acyclic adducts of aldehyde 3.(ABSTRACT TRUNCATED AT 400 WORDS)

Interaction of cyclophosphamide with DNA in isolated rat liver cell nuclei

Isolated rat liver cell nuclei were incubated with the alkylating cytostatic drug cyclophosphamide (CPA) in the presence of a microsomal activation system. Digestion of the 3H-CPA-treated nuclei with DNase I and micrococcal nuclease, respectively, showed that the CPA-modified DNA apparently has become resistant against such enzymatic attack. For analysis of the DNA-CPA reaction products, the DNA was isolated under mild conditions and degraded enzymatically. In cell nuclei whose DNA had been prelabeled with 32P in vivo, a predominant binding of 3H-CPA (about 50%) to the DNA phosphate groups was observed. Terminal phosphate groups apparently play an important role in this reaction. Neutral heating of DNA from 3H-CPA-treated nuclei liberated up to 60% of the initially bound 3H-radioactivity. The results are discussed in relation to the recent data concerning the route of decomposition of activated CPA to phosphoramide mustard and acrolein.