The significant roles that cytochrome P450 (P450) and UDPglucuronosyl transferase (UGT) enzymes play in drug discovery cannot be ignored, and these enzyme systems are commonly examined during drug optimization using liver microsomes or hepatocytes. At the same time, other drug-metabolizing enzymes have a role in the metabolism of drugs and can lead to challenges in drug optimization that could be mitigated if the contributions of these enzymes were better understood. We present examples (mostly from Genentech) of five different non-P450 and non-UGT enzymes that contribute to the metabolic clearance or bioactivation of drugs and drug candidates. Aldehyde oxidase mediates a unique amide hydrolysis of ,7-tetrahydro-1-benzothiophene-2-carboxamide), leading to high clearance of the drug. Likewise, the rodent-specific ribose conjugation by ADP-ribosyltransferase leads to high clearance of an interleukin-2-inducible T-cell kinase inhibitor. Metabolic reactions by flavin-containing monooxygenases (FMO) are easily mistaken for P450-mediated metabolism such as oxidative defluorination of 4-fluoro-N-methylaniline by FMO. Gamma-glutamyl transpeptidase is involved in the initial hydrolysis of glutathione metabolites, leading to formation of proximate toxins and nephrotoxicity, as is observed with cisplatin in the clinic, or renal toxicity, as is observed with efavirenz in rodents. Finally, cathepsin B is a lysosomal enzyme that is highly expressed in human tumors and has been targeted to release potent cytotoxins, as in the case of brentuximab vedotin. These examples of non-P450-and non-UGT-mediated metabolism show that a more complete understanding of drug metabolizing enzymes allows for better insight into the fate of drugs and improved design strategies of molecules in drug discovery.

A pulsed ultrafiltration-mass spectrometric screening assay has been developed to generate and identify electrophilic metabolites of xenobiotic compounds formed by hepatic cytochrome P450 enzymes. This assay would be suitable for the early identification of potentially toxic compounds during the initial phase of drug development. Rat liver microsomes were trapped by an ultrafiltration membrane in a stirred flow-through chamber, and substrates for microsomal cytochrome P450 including hydroxychavicol, 3-methylindole, cyproheptadine and 2-tert-butyl-4,6-dimethylphenol were flow-injected individually through the chamber along with the cofactors, NADPH and glutathione. Metabolites and glutathione conjugates were detected on-line using electrospray mass spectrometry. Alternatively, the ultrafiltrate was concentrated on a reversed phase HPLC column and analyzed using electrospray LC-MS or LC-MS-MS to separate and characterize isomeric metabolites and metabolites present at low concentration. Enzymatic activation of each xenobiotic substrate produced highly electrophilic metabolites such as quinones, quinone methides and imine methides that reacted with glutathione on-line to produce glutathione conjugates which were detected by using electrospray mass spectrometry. Although epoxides such as cyproheptadine epoxide were generated, it is likely that these compounds were insufficiently reactive to form glutathione conjugates in the absence of cytosolic glutathione S-transferases. Pulsed ultrafiltration-electrospray mass spectrometry offers an efficient method for in vitro formation and mass spectrometric characterization of activated microsomal drug metabolites and is suitable for use during the drug discovery process for the early identification and screening out of potentially toxic lead compounds.

The P450-catalyzed hydroxylation of tamoxifen to give alpha-hydroxytamoxifen [(E)-4-{4-[2-(dimethylamino)ethoxy]phenyl}-3,4-diphenyl-3-buten-2- ol] and subsequent formation of reactive sulfate esters which alkylate DNA has been proposed to be a potential carcinogenic pathway for tamoxifen. In the present study, the ability of alpha-hydroxytamoxifen analogs to form GSH and sulfate conjugates was investigated in order to understand the structural features influencing reactivity. The para oxo analogs 1 [1-(4-methoxyphenyl)-3-hydroxy-1-butene], 2 [1-(4-hydroxyphenyl)-3-hydroxy-1-butene], and 4 [1-(4-hydroxyphenyl)-1-phenyl-3-hydroxy-1-butene] reacted with GSH instantaneously under strong acidic conditions to yield GSH conjugates in greater than 90% yields. Interestingly, the meta phenolic analogs 3 [1-(3-hydroxyphenyl)-3-hydroxy-1-butene] and 5 [1-(3-hydroxyphenyl)-1-phenyl-3-hydroxy-1-butene] did not react with GSH to any significant extent under similar conditions. Characterization of the GSH conjugates with 1H-NMR, electrospray mass spectrometry, and UV showed that all of the conjugates resulted from attack of GSH at the alpha-position of the substrates with displacement of the hydroxyl group. The formation of a single pair of diastereomeric conjugates strongly supported adduct formation to proceed through a direct S(N)2 displacement mechanism and not through a quinone methide (4-alkyl-2,5-cyclohexadien-1-one) intermediate. At physiological pH and temperature only the para hydroxy analogs 2 and 4 gave GSH conjugates, a reaction which seems to be catalyzed by isoforms of glutathione S-transferase. Similar substituent effects were observed in the sulfotransferase-mediated formation of alpha-hydroxy sulfate esters in that only the para hydroxy analogs formed conjugates at the aliphatic hydroxyl group. Finally, the present investigation showed a remarkable difference in the reactivities of para and meta phenolic analogs of alpha-hydroxybutenylbenzenes toward GSH and sulfate conjugation reactions.