![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Molecular Medicine and Pathology (N.A.H.), Department of Pharmacology (M.A.G., M.H.Y.T., M.D.T.), and Auckland Cancer Society Research Centre (N.A.H., G.J.A, E.M.S., W.R.W.), Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand
(Received September 5, 2007; Accepted November 9, 2007)
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
p +0.78) to a hydroxylamine (p –0.32) or amine (p –0.66) metabolite; this "electronic switch" can be used to activate the latent nitrogen mustard moiety (Denny and Wilson, 1986
; Palmer et al., 1990; Siim et al., 1997) The 2,4-dinitrobenzamide-5-mustard class (2,4-DNBM) was investigated first and shown to provide cytotoxicity and DNA cross-linking selectively under hypoxic conditions (Palmer et al., 1992). The 2,4-DNBMs were subsequently shown to be substrates for the oxygen-insensitive nitroreductase nfsB from Escherichia coli (Anlezark et al., 1995; Wilson et al., 2002; Atwell et al., 2007), which is of interest for prodrug activation in the context of gene- or virus-directed enzyme prodrug therapy. We have shown that SN 23862 and other 2,4-DNBMs provide efficient bystander effects (killing of adjoining cells) when activated by either endogenous one-electron reductases or the nfsB nitroreductase through the diffusion of an activated metabolite in which the nitro group para to the mustard has been reduced to the corresponding amine (Wilson et al., 2002; Helsby et al., 2004; Wilson et al., 2007). In addition to its metabolic activation in tumor cells, the 2,4-DNBM SN 23862 is extensively metabolized in tumor-bearing mice, with nitroreduction of either the 4- or 2-nitro group as well as sequential oxidative N-dealkylation of the mustard moiety (Kestell et al., 2000). The latter route yields the nontoxic 2,4-dinitro-5 amine, but the dechloroethylation reactions also generate chloroacetaldehyde, which is a reactive cytotoxin (Sood and O'Brien, 1993). Formation of chloroacetaldehyde by oxidative dealkylation has been suggested to contribute to the toxicity of oxazaphosphorine mustards (Springate, 1997) and hence may lead to normal tissue toxicity of DNBM.
Evaluation of structure-activity relationships for DNBM as hypoxic cytotoxins has identified the 3,5-dinitrobenzamide-2-mustards (3,5-DNBM) as the preferred regioisomers; compounds in this class are also good substrates for the nfsB nitroreductase (Singleton et al., 2007). Low aqueous solubility of these prodrugs led to development of phosphate esters of 3,5-DNBM with hydroxylalkylcarboxamide side chains, with PR-104 possessing activity against hypoxic cells in human tumor xenografts (Patterson et al., 2007). PR-104 is hydrolyzed rapidly to the corresponding alcohol PR-104A (SN 27858; 1 in the present study) in vivo, which is reduced in hypoxic tumor cells to the DNA cross-linking para-hydroxylamine, PR-104H (1a in the present study) (Guise et al., 2007; Patterson et al., 2007). PR-104 is currently in a clinical trial as an antitumor agent (http://www.proacta.com).
3,5-DNBMs, such as PR-104, with hydroxyethylcarboxamide side chains represent a new pharmacophore for which routes of biotransformation in normal tissues are not yet well understood. A recent report has demonstrated metabolites resulting from oxidative dealkylation and thiol conjugation of the mustard and glucuronidation of the hydroxylethyl side chain in mouse plasma (Patel et al., 2007). The aim of the present study was to establish the chemical identity of metabolites of hydroxyethyl 3,5-DNBM in liver S9 preparations from livers of mice, rats, dogs, and humans under aerobic conditions. We have previously demonstrated the in vitro aerobic NAD(P)H-dependent nitroreduction of the 2,4-dinitrobenzamide-5-aziridine CB 1954 by human liver S9 (Tang et al., 2005), suggesting that this will be a useful model for evaluating the sensitivity of DNBM oxygen-insensitive nitroreduction. Liver S9 preparations were used for these studies rather than microsomes as we have shown previously that nitroreduction of the related dinitrobenzamide aziridine compound, CB 1954, involved a complex interplay of microsomal and cytosolic enzymes (Tang et al., 2005). The four compounds studied here, SN 27858 (PR-104A; 1), SN 29546 (2), SN 27686 (3), and SN 29893 (4), were chosen to clarify the effect of structural changes at the nitrogen mustard position on the potentially cytotoxic routes of metabolism, namely nitroreduction and N-dealkylation. We also compared the sensitivity of each compound to glucuronidation by liver microsomes from each species as this pathway has potential implications for metabolic clearance and host toxicity via regeneration of the parent drug in the gastrointestinal tract.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Aerobic S9 Metabolism of Dinitrobenzamide Analogs (1–4). In vitro hepatic metabolism of compounds 1 through 4 was studied using the 9000g postmitochondrial supernatant (S9) of human (HL5 and HL18), dog (beagle), rat (male and female, Sprague-Dawley) and mouse (pooled male and female CD-1 athymic nu/nu) livers, prepared as described previously (Tang et al., 2005). The human livers selected were chosen because we have shown previously that they are representative of drug metabolism across a diverse range of cancer chemotherapeutic drugs (Zhou et al., 2000; Lu et al., 2004; Tang et al., 2005). Hepatic S9 preparations (8 mg of protein/ml) and dinitrobenzamide prodrug substrate (250 µM) were incubated for 30 min at 37°C in a final volume of 0.5 ml of sodium-potassium phosphate buffer (67 mM, pH 7.4) under air in the presence of cofactor (NADPH or NADH, 1 mM). The reaction was terminated and metabolites were extracted by two sequential additions of 0.5 ml of ice-cold ethyl acetate, with centrifugation at 10,600g for 5 min after each extraction. The supernatant from the two sequential ethyl acetate extractions were combined and evaporated to dryness under vacuum. The samples were resuspended in 100 µl of mobile phase consisting of 80% ammonium formate (45 mM, pH 4.5) and 20% acetonitrile, and an aliquot (80 µl) was injected onto the HPLC column. All experiments were performed in quadruplicate, and incubations with boiled S9 preparations were used as controls. Mass spectrometry was used for identification of the metabolites, and metabolite formation was determined using peak area absorbance at 254 nm.
In Vitro Glucuronidation of Dinitrobenzamide Compounds. Microsomes (2 mg/ml) prepared from human (HL18), dog (beagle), rat (pooled male and female Sprague-Dawley), and mouse (pooled male and female CD-1 athymic nu/nu) livers were used to detect formation of glucuronide metabolites. Microsomes were incubated with UDPGA (5 mM) and drug substrate (250 µM) for 60 min at 37°C in the presence of Brij 58 (0.15 µg/incubation) in a final volume of 0.5 ml in Tris-HCl (50 mM) buffer, pH 7.4. After extraction, formation of glucuronide metabolites was determined by LC/MS analysis as above.
LC/MS Analysis. The HPLC system consisted of an Agilent 1100 HPLC system interfaced with a diode array detector. Separation was carried out on a 2.1 x 150 mm Alltima C8 5-µm column (Alltech Associates, Deerfield, IL). The mobile phase consisted of 80% v/v acetonitrile in water (A) and 45 mM ammonium formate buffer, pH 4.5 (B) at a flow rate of 0.3 ml/min. The gradient conditions were 0 to 4 min 20% A, 4 to 21 min 90% A, 21 to 22 min 90% A, and 22 to 26 min 20% A, with a total run time of 40 min. Absorbance detection was at 254 and 370 nm (bandwidth 4 nm, reference 550 nm).
Online mass spectra were obtained with a single quadruple mass spectrometer (Agilent LC/MSD model A). Products from metabolism of the analogs were identified using positive and negative mode atmospheric pressure electrospray ionization with nitrogen as the vaporizing and drying gas. The following parameters were used: fragmentor voltage 100 V, capillary spray voltage 4 kV, nebulizer pressure 45 psi, gas temperature 350°C, auxiliary drying gas flow 12 liters/min and the mass/charge (m/z) ratio was scanned from 100 to 1000. An Agilent LC/MSD trap-SL mass spectrometer equipped with an Agilent capillary HPLC system was also used for the further identification of some metabolic products. Chromatographic separation was as above with the exception that a 150 x 0.5 mm, 5-µm column at a flow rate of 15 µl was used. The electrospray ionization source was set at positive ionization mode with auto MS(n).
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
; Patterson et al., 2007). ortho-Nitroreduction of compound 1 in tumor cell lines under hypoxia results in intramolecular alkylation to form tetrahydroquinoxaline products (Patterson et al., 2007), which have a distinctive UV spectrum (
max 400 nm). However, there was no evidence of formation of the tetrahydroquinoxaline products of ortho-nitroreduction of compound 1 in human liver S9. In addition there was no detectable N-dealkylation of compound 1. However, a more lipophilic metabolite with a molecular mass of 540 (X + 42 amu) was observed (1j), which was tentatively identified as a product of O-acetylation of the alcohol side chain. Nucleophilic displacement of the bromine by a chloride ion from the incubation buffer resulted in the formation of the product 1k (X = CH2Cl and Y = CH2OSO2Me) with a molecular mass of 454. Other unidentified peaks were also observed in boiled S9 and are likely to be hydrolysis products of the parent compound. No new metabolites were observed after incubation with NADH as cofactor and metabolite levels were lower with this cofactor than with NADPH (data not shown).
|
|
In Vitro Hepatic S9 Metabolism of SN 29546 (2). Human liver S9 metabolism of the asymmetric chloromesylate compound (2) also resulted in the formation of three metabolites (Fig. 1b). The inferred structures are shown in Fig. 2, where X = CH2Cl and Y = CH2OSO2Me. The metabolites were identified as the hydroxylamine (2a; [M + H]+ 440.9 m/z) and the amine metabolite (2b; [M + H]+ 424.9 m/z). These products have similar retention times and UV spectra as 1a and 1b and hence are expected to be the products of reduction of the nitro group para to the mustard. A lipophilic metabolite with an increase in molecular mass of X + 42 amu was observed and tentatively identified as an O-acetylated metabolite 2j. Minor products presumably due to nucleophilic displacement of the mesylate moiety by the hydroxyl ion (Rt 14.7; [M + H]+ 377.0 m/z) or chloride ion (2k; Rt 16.7 min) were observed. In addition, the hydrolysis product of the para-nitroreduced metabolite (2b) was also tentatively identified (Rt 12.5 min). As with the asymmetric bromomesylate compound (1), the chloromesylate (2) did not undergo any detectable N-dealkylation or ortho-nitroreduction. Similar to compound 1 no new metabolites were observed after incubation with NADH as cofactor.
In Vitro Hepatic S9 Metabolism of SN 27686 (3). Incubation of the dibromo mustard (3) with human liver S9 plus NADPH resulted in the formation of a complex set of products (Fig. 1c), of which seven were identified (the inferred structures are shown in Fig. 2, where X = Y = CH2Br). The major metabolites were products of nitroreduction, 3b, 3d, and 3e and the N-dealkylated product (3f). Metabolite 3b was identified as an amine metabolite because of the decrease in mass (X–30 amu). This product had a molecular ion of [M + H]+ 453 m/z with an adduct (+44) in the negative mode. The presence of two bromine molecules was apparent from the clear signature pattern of the satellite peaks. Although the intermediate ortho-hydroxylamine and ortho-amine metabolites were not observed, the products of intramolecular alkylation, 3d and 3e, were detected and characterized by the distinctive UV spectra (max 400 nm) and the presence of a single bromine satellite peak signature pattern and molecular mass (388 and 372, respectively). The metabolite 3f (Rt 15.9 min) had a distinctive UV spectrum (max 365 nm) and a molecular mass of 375.5 with a satellite peak +2 amu with an intensity 98% of the base peak, indicative of a single bromine in the molecule. This metabolite was identified as the product of N-dealkylation and confirmed by independent synthesis. Peak 3h (Rt 15.2 min, molecular mass 418) also had this distinctive UV spectrum as well as the presence of a single bromine satellite peak in the mass spectrum and an increase in molecular mass of X + 42 amu indicating that it may be an O-acetylated derivative of 3f.
The late eluting peak (Rt 21.7 min) with a molecular mass of 524 was identified as an O-acetylated metabolite (3j) of the hydroxyethyl side chain, as described for compounds 1 and 2. Nucleophilic displacement of the bromo group by a chloride ion to form a less lipophilic product (3k) was also observed. Similar to compounds 1 and 2 no new metabolites were observed after incubation with NADH as cofactor, and metabolite levels were lower with this cofactor than with NADPH (data not shown).
In Vitro Hepatic S9 Metabolism of SN 29893 (4). Incubation of the diethylamine compound (4) with human liver S9 plus NADPH resulted in the formation of six metabolites (Fig. 1d); the inferred structures are shown in Fig. 2, where X = Y = CH3. The products of nitroreduction were identified by mass spectrometry as an amine (4b) ([M + H]+ 297 m/z and [M–H]– 295 m/z) and hydroxylamine (4c) ([M + H]+ 313 m/z and [M–H]– 311 m/z). Compound 4b (X–30 amu; Rt 14.1 min) was assumed to be formed by reduction of the nitro group para to the N-diethylamine moiety via the hydroxylamine intermediate 4a, which was not observed. The hydroxylamine metabolite (4c; X–14 amu; Rt 11.2 min) was assumed to be formed by nitro group reduction ortho to this moiety. The para-hydroxylamine (1a) and para-amine (1b) metabolites of 1 have identical UV spectra (Patterson et al., 2007) and retention times that differ by 0.6 min. In contrast, 4b and 4c have a relatively large difference in retention time of 2.9 min and also have different absorbance spectra (4b, max 250 nm; 4c, max 250 and 400 nm); for details of UV spectra, see Supplemental Data. Hence these metabolites have been classified as products of para- and ortho-nitroreduction, respectively.
The lipophilic metabolite 4j (Rt 18.5 min) had a higher mass ([M–H]– 367 m/z) than the parent compound 4 but a similar UV spectrum (Rt 15.5 min). Sequential fragmentation of this metabolite using ion trap mass spectrometry (positive mode) resulted in the following daughter ions: MS2 308 m/z, MS3 267 m/z, and MS4 253 m/z (corresponding to elimination of –C2H6O2, –C2H3N, and –O, respectively). This metabolite (4j) was identified as the O-acetylated metabolite of 4, which was confirmed by independent synthesis.
Metabolite 4f ([M + H]+ 299 m/z and [M–H]– 297 m/z) with a distinctive UV spectrum, max 365 nm (Supplemental Data) was identified as the product of N-dealkylation of the diethylamine moiety, which was confirmed by independent synthesis. An additional metabolite 4g, which was poorly resolved from 4f, had a molecular mass of 296 with an additional change in UV spectrum at 365 nm (Supplemental Data) and was tentatively identified as an indazolinone product. The primary amine resulting from further N-dealkylation of 4f, 2-amino-N-(2-hydroxyethyl)-3,5-dinitrobenzamide, was not observed. However, an additional less lipophilic metabolite 4i (Rt 9.4 min) with a mass of 292 and a distinctive UV spectrum with max at 250 and 320 nm was observed (Supplemental Data). Sequential fragmentation of this metabolite using ion trap mass spectrometry (positive mode) resulted in the following daughter ions: MS1 293, MS2 247, and MS3 222 m/z, corresponding to elimination of –C2H5O and –C2H2, respectively. We have tentatively identified this metabolite as 2,4-dinitro-6,7-dihydro-11H-pyridazino[1,2-a]indazole-8,11(9H)-dione (4i).
Similar metabolites were detected when liver S9 was fortified with NADH as cofactor, but most were present at 4- to 5-fold lower concentrations (data not shown). The only exception was that in the presence of NADH relatively more O-acetylated metabolite (4j) was detected. No metabolites were observed in the presence of boiled S9.
Thus, in summary, there are three main routes of NADPH-dependent metabolism of compounds 1 through 4 in human liver S9: 1) aerobic reduction of either nitro group, 2) N-dealkylation, and 3) O-acetylation of the hydroxyethyl side chain. The two livers selected for this study were chosen because they are representative of cancer drug metabolism in a larger liver bank (Zhou et al., 2000; Lu et al., 2004; Tang et al., 2005). However, the potential for interindividual variability in the metabolism of these compounds has not been addressed in these studies.
Metabolic Profile in Nonclinical Species. To determine whether any species differences in hepatic S9 metabolic profile exist the compounds were incubated with S9 from three nonclinical species: mouse, rat, and dog (Table 1). Of the nonclinical species studied, nitroreduction of 1 was greatest in the mouse (3348 ± 1749 mAU · s) and lowest in the dog (589 ± 118 mAU · s). Nitroreduction was also higher in female rats than in male rats (1994 ± 818 versus 705 ± 377 mAU · s) (Table 1). Moreover, there was considerable variability in the amount of para-nitroreduced metabolites formed between the two human livers HL5 and HL18 (261 ± 9 and 3338 ± 1064 mAU · s, respectively), which may indicate potential interindividual variability in human nitroreductase enzyme(s) activity for this substrate. The chloromesylate (2) was a poor substrate for para-nitroreduction in dog liver S9, with 10-fold lower levels of 2a and 2b formed, compared with the other species studied (Table 1). The ortho-nitroreduced metabolites 3d and 3e of the dibromo compound 3 were not detected in rat liver S9 (either male or female liver). Additionally, relatively low levels of para-nitroreduction to form 3b were observed in rat liver S9, and there were relatively low levels of N-dealkylation of compound 3 by female rat liver compared with male rat liver and livers of other species (Table 1). For the diethylamine compound 4, the same metabolites were formed as were observed with human liver S9, with the exception of the CD-1 nu/nu mouse, which did not support the formation of the nitroreduced metabolite 4c.
|
In Vitro Glucuronidation. Compound 1, the alcohol metabolite of PR-104, has been reported to undergo in vivo glucuronidation in mice (Patel et al., 2007). To evaluate this pathway in vitro, the compounds (1-4) were incubated with liver microsomes fortified with the cofactor UDPGA and detergent. A single more polar metabolite was detected after incubation with each compound. In each case this metabolite had a higher mass than the parent compound (X + 176), with [M + H]+ 675, 631, 659, and 503 m/z for 1, 2, 3, and 4, respectively. The product of UDPGA-dependent metabolism of compound 1m was confirmed as an O-glucuronide of the alcohol side chain by comparison with authentic sample. Formation of the O-glucuronide metabolite for all four compounds was observed in human, mouse, rat, and dog microsomes, although the amount formed varied across species. In rat and mouse liver microsomes the dinitrobenzamide compounds were relatively poor substrates for glucuronidation compared with human or dog liver (Fig. 3). The most extensive glucuronidation (66.6 ± 5.2%) was observed for the dibromo compound 3 with dog liver.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
; Helsby et al., 2003) cannot occur. Moreover, this compound was used to elucidate the effect of changes in the N-mustard moiety on the metabolic routes of these compounds.
Four metabolic routes were observed: aerobic nitroreduction, N-dealkylation, O-acetylation, and O-glucuronidation. Attempts to optimize for every individual metabolic route proved unsuccessful (data not shown); thus, the substrate concentration chosen for this study was based on the reported plasma Cmax for compound 1 in mouse and rat (Patel et al., 2007; Patterson et al., 2007). Dinitrobenzamide mustard compounds are substrates for E. coli nitroreductase (nfsB) (Anlezark et al., 1995); in addition, the prototype DNBM, SN 23862, also undergoes anaerobic nitroreduction by cell lines that overexpress NAD(P)H P450 oxidoreductase (Wilson et al., 2007). However, this is the first report of nitroreduction of DNBM compounds by hepatic preparations under atmospheric oxygen conditions. Aerobic nitroreduction by liver S9 preparations has been observed previously with the aziridinyl-dinitrobenzamide compound, CB 1954 (Tang et al., 2005), and hence hepatic nitroreduction was not an unexpected finding for this class of compounds.
The least complex analog studied, the diethylamine (4), underwent extensive nitroreduction by human, rat, and dog liver S9. Two nitroreduced metabolites were formed, the hydroxylamine and the amine. These metabolites were assumed to be the products of ortho- and para-nitroreduction, respectively, as para-hydroxylamine and para-amine metabolites of 1 have identical UV spectra (Patterson et al., 2007), whereas the hydroxylamine metabolite of 4 has a spectral shift compared with the amine metabolite. In addition, a spectral shift after ortho-nitroreduction of related DNBM compounds has been observed previously (Wheeler, 1999); however, the identity of 4c as the ortho-hydroxylamine requires further confirmation by nuclear magnetic resonance. In contrast, the more complex dibromo mustard analog (3) was a substrate for ortho-nitroreduction, with characteristic intramolecular alkylation to form a tetrahydroquinoxaline (3d and 3e), in all species except the rat. This finding may indicate species-dependent enzyme-substrate interactions for the reductase(s) involved in this route. More importantly, the increased complexity of the bromomesylate (1) and chloromesylate moiety (2) appeared to entirely abrogate ortho-nitroreduction in all the species examined. There was no evidence of formation of these metabolites either in the UV chromatogram or by interrogation of the mass chromatogram for the expected m/z values of the tetrahydroquinoxaline metabolites of compounds 1 and 2.
Previous studies have indicated that the steric bulk can influence the ortho-nitroreduction of dinitrobenzamides by E. coli nitroreductase (Wilson et al., 2002; Helsby et al., 2004) as ortho-nitroreduction of the dichloro DNBM compound, SN 23862, is precluded by a limited binding orientation in the E. coli nitroreductase crystal structure attributable to the steric bulk of the mustard moiety compared with the smaller aziridinyl moiety (Johansson et al., 2003). Ortho-nitroreduction of DNBM results in intramolecular alkylation and formation of "half-mustard" compounds and is considered to be an inactivation pathway for these prodrugs. However, extensive metabolic clearance via this route would be detrimental to the systemic availability of the parent prodrug at the site of action, the tumor.
In contrast, reduction of the nitro group para to the mustard moiety is a bioactivation pathway, which results in the formation of known DNA cross-linking cytotoxins. Aerobic para-nitro reduction was observed for all four compounds in all species studied. Both compounds 3 and 4 underwent six-electron nitroreduction to the para-amine derivative (3b and 4b), indicating extensive nitroreduction of these compounds. In contrast, both the four-electron para-hydroxylamine (1a and 2a) and the six-electron para-amine (1b and 2b) metabolites of compounds 1 and 2 were observed, which may indicate a relatively slower rate of nitroreduction for these compounds. However, the amine 1b is reported to undergo slow autooxidation to the hydroxylamine 1a (Patterson et al., 2007); hence, detection of the products of four-electron reduction of compounds 1 and 2 may indicate the relative aerobic instability of the para-amine metabolite of these compounds.
Although there was some variability in the amount of para-nitroreduced metabolites formed across species, in general increasing the steric bulk of the mustard group had little effect on the ability of these analogs to undergo para-nitroreduction. Thus, under aerobic conditions the DNBM analogs can undergo hepatic metabolism to cytotoxic alkylating metabolites, and this process may result in adducts at the site of formation, i.e., the liver. Importantly, these metabolites may not be readily detectable systemically (in plasma) and hence the bioactivation of these compounds in normal (oxic) tissue may be underestimated.
A further pathway of metabolism was observed for the DNBM analogs, namely, N-dealkylation. In vivo N-dealkylation of the prototype DNBM compound, SN 23862, has been reported previously (Kestell et al., 2000), and this metabolic route results in the formation of an inactive half-mustard, which cannot cross-link DNA and is less cytotoxic. However, the N-dealkylation of the DNBM analogs may also release reactive acetaldehydes, such as bromoacetaldehyde, which are potent cytotoxins (Khan et al., 1996). An alternative mechanism for N-dealkylation of the DNBM is possible via oxidation to the N-oxide of the nitrogen mustard followed by nucleophilic attack by the hydroxyl ion and elimination of formaldehyde (Tercel et al., 1995).
Thus, it is important to minimize metabolic clearance of DNBM via this route, first to avoid toxicity in normal tissue and second to avoid decreased systemic availability of the prodrug. Altering the structure of the mustard group had a dramatic effect on N-dealkylation, decreasing the amount of metabolite observed by up to 10-fold for the dibromo compound (3) compared with the diethylamine (4) and entirely abrogating this route for the more complex bromomesylate (1) and chloromesylate (2) compounds. Hence, the in vitro metabolism data would predict that that both 1 and 2 would undergo minimal dealkylation in vivo. However, the half-mustard metabolite of 1 resulting from oxidative debromoethylation (1f) has been identified in plasma after administration of 1 to mice, although it coelutes with the O-glucuronide metabolite (1m) (Patel et al., 2007). Hence, further studies to determine the role of extrahepatic dealkylation of DNBM analogs may be warranted.
The O-acetylation of the hydroxyethyl side chain by liver S9 preparations was a relatively minor route compared with oxidative metabolism. In addition, as the compounds were also O-glucuronidated at this position the formation of O-acetylated metabolites may be precluded in fully competent hepatocytes in vivo. Notably, O-glucuronidation was particularly extensive in dog liver compared with livers of the other species. A high rate of glucuronidation by dog liver compared with livers of other species has been reported previously for the human immunodeficiency virus drug, bevirimat (Wen et al., 2007). Further studies to elucidate the relative importance of O-glucuronidation of the hydroxyethyl side chain in vivo are in progress.
In conclusion, increasing the steric bulk of the N-mustard moiety abrogated the ortho-nitroreduction and N-dealkylation routes. This increase improves the in vitro metabolic profile of DNBM compounds with the prediction of a relatively high metabolic stability for compounds 1 and 2. Indeed, mouse S9 metabolic stability data (t = 30 min) indicated <10% loss of prodrug for compounds 1 and 2, compared with 75% loss of compound 3. However, structural changes at the mustard moiety had no effect on the aerobic para-nitroreduction of these compounds. Although compounds 1, 2, and 3 undergo bioactivation to form metabolites that are expected to be cytotoxic DNA alkylators, it is postulated that para-nitroreduction may be a relatively minor route of metabolism for these compounds compared with O-glucuronidation in vivo. Further studies to determine the in vivo metabolic profile of these compounds are in progress.
|
Footnotes |
|---|
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
ABBREVIATIONS: DNBM, dinitrobenzamide mustard; 2,4-DNBM, 2,4-dinitrobenzamide 5-mustard; SN 23862, bischloroethyl-2,4-dinitrobenzamide; 3,5-DNBM, 3,5-dinitrobenzamide 2-mustard; PR-104, 2-((2-bromoethyl)-2-{[(2-hydroxyethyl)amino]carbonyl}-4,6-dinitroanilino)ethyl methanesulfonate phosphate ester; CB 1954, 5-(aziridin-1-yl)-2,4-dinitrobenzamide; SN 27858, 2-((2-bromoethyl)-2-{[(2-hydroxyethyl)amino]carbonyl}-4,6-dinitroanilino)ethyl methanesulfonate, PR-104A, 1; SN 29546, 2-((2-chloroethyl)-2-{[(2-hydroxyethyl)amino]carbonyl}-4,6-dinitroanilino)ethyl methanesulfonate, 2; SN 27686, 2-[bis(2-bromoethyl)amino]-N-(2-hydroxyethyl)-3,5-dinitrobenzamide, 3; SN 29893, 2-(diethylamino)-N-(2-hydroxyethyl)-3,5-dinitrobenzamide, 4; LC, liquid chromatography; MS, mass spectrometry; HPLC, high-performance liquid chromatography; UDPGA, UDP-glucuronic acid; Rt, retention time; amu, atomic mass units.
The online version of this article (available at http://dmd.aspetjournals.org) contains supplemental material. 
Address correspondence to: Dr. Nuala A. Helsby, Department of Molecular Medicine and Pathology, Faculty of Medical and Health Sciences, University of Auckland, Private Bag 92019, Auckland, New Zealand. E-mail: n.helsby{at}auckland.ac.nz
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  Y. Gu, A. V. Patterson, G. J. Atwell, S. B. Chernikova, J. M. Brown, L. H. Thompson, and W. R. Wilson Roles of DNA repair and reductase activity in the cytotoxicity of the hypoxia-activated dinitrobenzamide mustard PR-104A Mol. Cancer Ther., June 1, 2009; 8(6): 1714 - 1723. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
