Relationship Between Thiol Adducts and Toxicity

A relationship is proposed to exist between the estimated total human in vivo thiol adduct burden (normalized for the daily dose) and drug-induced toxicity (Gan et al. 2009). In this study, dansyl-GSH was used to facilitate quantitation of adducts.

6.5.1.3 Cyanide Conjugates

Cyanide conjugates are suited for detection of iminium ions. To screen for cyanide conjugates use the following:

• A 27 Da neutral loss scan in the positive ion mode on a triple quadrupole mass spectrometer (Argoti et al. 2005). A mixture of CN~ and stable-labeled 13C15N~ enhances the selectivity for detection of cyanide conjugates.

Aldehyde

Figure 6.3. Mechanisms of trapping reactive intermediates by glutathione (GSH), cyanide (CN_), methoxylamine, and semicarbazide.

6.5.1.4 Methoxylamine and Semicarbazide Conjugates These conjugates are suited for trapping aldehydes. No straightforward fragmentation pattern exists for MS/MS methods in the scan mode for automated data acquisition.

6.5.2 Covalent Modification of Proteins

Determination of covalent binding of drugs to endogenous biomo-lecules, such as proteins, has been conducted for many years, and high-throughput methods have been established (Day et al. 2005). Some companies have integrated covalent binding assays into the drug discovery stage and, although no fixed cut-off values are broadly applicable, a value of >50 pmol/mg protein is proposed as the threshold beyond which the drug candidate should not be pursued further (Evans et al. 2004). When multiple radiolabeled compounds are available, structure-activity relationships can be evaluated, which can lead to development of drugs that reduce or eliminate the extent of covalent binding. However, no apparent quantitative link exists between covalent binding and toxicity (Obach et al. 2008).

6.5.3 Time-Dependent Inhibition of P450 Enzymes

Reactive metabolites can inactivate P450 enzymes by covalently modifying the heme group or the apoprotein of the enzyme. This topic is covered in greater detail in Chap. 5.

6.5.4 Bioactivation in the Context of Drug Discovery and Development

The steps a metabolism scientist should take when bioactivation (defined by formation of GSH or other adducts, covalent binding, or mechanism-based inactivation of P450 enzymes) is observed continues to be debated in the literature. It is clear that covalent binding can be a liability, but it is not necessarily a showstopper. The following aspects should be taken into consideration:

• The extent of adduct formation, covalent binding, or mechanism-based P450 enzyme inactivation.

• Not all in vitro observations translate into in vivo liabilities because alternative, detoxifying clearance pathways may exist in vivo. For example, in the case of raloxifene, in vitro incubation with human liver microsomes and nicotinamide adenine dinucleotide phosphate (NADPH) leads to formation of GSH conjugates and covalent binding. However, the main route of elimination for raloxifene is glucuronidation in the intestine and liver (Dalvie et al. 2008).

• A low dose and/or low systemic exposure greatly reduces the risk for serious toxicity.

• Duration of therapy - short term treatment (e.g., antibiotics) versus long term use in relatively healthy patients.

• Therapeutic area - life-threatening diseases versus drug lifestyle drugs (e.g., obesity).

• Target population - some patient populations may be more sensitive to bioactivation than others.

• If a prototype for a new target or compound aims for best in class classification.

• Species differences - not all species process drugs in the same way, and bioactivation may be unique to one species.

Nevertheless, bioactivation studies can provide valuable mechanistic insights that can be incorporated in the overall optimization of a drug structure.

Efavirenz, a non-nucleoside reverse transcriptase inhibitor for the treatment of HIV, presents a classic example of the role of bioactivation in drug development (Mutlib et al. 2000). Efavirenz causes severe renal tubular renal cell necrosis in rats, but not in monkeys, and scientists were greatly concerned that this condition would manifest itself in humans as well. However, detailed metabolism studies showed that the differences in toxicity between the species are due to differences in the production and processing of reactive metabolites. Rats produce a unique GSH adduct that hydrolyzes to a cysteinegly-cine conjugate that ultimately leads to toxicity (see Fig. 6.4). Indeed, in this case human metabolism is more like that in monkeys, and no renal toxicity was observed in humans.

Efavirenz rat, monkey

Efavirenz rat, monkey

Figure 6.4. Metabolism of efavirenz in rats, monkeys, and humans.

Cl^ ^ V Ncysteinylglycine ^O

ho3so h rat rat

Figure 6.4. Metabolism of efavirenz in rats, monkeys, and humans.

Finally, the view towards bioactivation is also influenced by the ease of addressing it. If multiple lead series are pursued and one or more series do not display bioactivation, these leads should be given serious consideration for further development. The most pragmatic approach may be to simply avoid compounds with certain structural characteristics that are associated with metabolic bioactivation (Kalgutkar and Soglia 2005).

6.6 BIOTRANSFORMATION/BIOACTIVATION OF MOIETIES

Here, we describe major biotransformation pathways by Phase I and Phase II DMEs (Figs. 6.5-6.15).

Oxidation of alkanes

O OH

FeV FeIV

FeIV Fe111

If there is a alpha-CH

FeIV-OH

ho^F

Oxidation of alkenes

C Oxidation of alkynes t>

oxirene

+h2o

if terminal acetylene

Ketene Carboxylic acid

Figure 6.5. Oxidation of alkanes, alkenes, and alkynes. Alkanes. P450-mediated oxidation of carbon atoms proceeds via hydrogen atom transfer a

H2O+FeMI

Figure 6.5 Continued (HAT; homolytic bond breakage) to form a carbon radical intermediate, followed by a hydroxyl radical rebound reaction. Selectivity of the site of oxidation. When the carbon radical intermediate mentioned above is stabilized, the activation energy required for the reaction to proceed is lowered. The carbon radical is usually most stable in the benzylic position, followed by the tertiary and secondary carbon positions (benzylic > allylic (branch > unbranched) > aliphatic (3 > 2 > 1)). This relationship, however, does not always hold true since other factors are involved in oxidation, such as binding orientation to the enzyme and the propensity of neighboring moieties to induce oxidation (allylic moieties for example). Alkenes. P450-mediated oxidation of alkenes proceeds via the introduction of an oxygen atom to form an epoxide. Most epoxides are not stable, but some epoxides, such as the 10,11-epoxide of carbamazepine, are stable. The stability of the epoxide depends on the electron density of the double bond; the higher the electron density, the more stable the epoxide. This can be followed by enzymatic (via epoxide hydrolases) or non-enzymatic opening of the epoxide to form a diol. The diol can loose a water molecule to form a ketone or aldehyde (via an enol intermediate). Alkynes. P450-mediated oxidation of alkynes proceeds by formation of an oxirene, which rearranges to a ketone and reacts with nucleophiles to form esters and amides, or with water to give an acid. Terminal acetylenes can also be oxidized to ketoradicals, which inactivate the P450 enzyme by binding to the heme nitrogen.

Figure 6.6. Aromatic ring oxidation. Several mechanisms have been proposed for the oxidation of aromatic rings. One mechanism begins with a SET, which results in the formation of a cation radical of the benzene ring. This step is followed by nucleophilic attack of FeIV-O—. The other mechanism involves formation of an epoxide, similar to what was described for alkene oxidation. The arene oxide formed can rearrange in two ways. In one mechanism, an NIH shift is observed in which X migrates to the neighboring carbon (X can be CH3, Cl, Br, F). The arene oxide can also react with GSH (catalyzed by GST) or water (catalyzed by microsomal expoxide hydrolase).

Figure 6.6. Aromatic ring oxidation. Several mechanisms have been proposed for the oxidation of aromatic rings. One mechanism begins with a SET, which results in the formation of a cation radical of the benzene ring. This step is followed by nucleophilic attack of FeIV-O—. The other mechanism involves formation of an epoxide, similar to what was described for alkene oxidation. The arene oxide formed can rearrange in two ways. In one mechanism, an NIH shift is observed in which X migrates to the neighboring carbon (X can be CH3, Cl, Br, F). The arene oxide can also react with GSH (catalyzed by GST) or water (catalyzed by microsomal expoxide hydrolase).

Amine

retro-reduction

R R' R" Carbinolamine

H Aldehyde

Nitrone or oxime

Iminiumion

CN R"

Amine

H^ "R' "R" Amide migration of double bond to naighboring carbon to form enamine

c d=Ar-dealkylation

Figure 6.7. Aliphatic nitrogen metabolism. Amines can be oxidized to several different types of metabolites, including a carbinolamine intermediate that leads to N-dealkylated compounds, N-oxides, nitrones, oximes, iminium ions, amides, and enamines. For P450-mediated oxidation of amines and amides, the first step of the reaction is generally thought to be via hydrogen atom transfer (HAT) from the alpha carbon followed by a hydroxyl radical rebound.

An NIH shift is a chemical rearrangement in which a chemical group undergoes intramolecular migration during the process of oxidation (see Fig. 6.6).

Ether oxidation

Was this article helpful?

0 0

Post a comment