Monitoring abacavir bioactivation in humans: Screening for an aldehyde metabolite

Abstract

The anti-HIV drug abacavir is associated with idiosyncratic hypersensitivity reactions and cardiotoxicity. Although the mechanism underlying abacavir-toxicity is not fully understood, drug bioactivation to reactive metabolites may be involved. This work was aimed at identifying abacavir–protein adducts in the hemoglobin of HIV patients as biomarkers of abacavir bioactivation and protein modification. The protocol received prior approval from the Hospital Ethics Committee, patients gave their written informed consent and adherence was controlled through a questionnaire. Abacavir-derived Edman adducts with the N-terminal valine of hemoglobin were analyzed by an established liquid chromatography–electrospray ionization–tandem mass spectrometry method. Abacavir–valine adducts were detected in three out of ten patients. This work represents the first evidence of abacavir–protein adduct formation in humans. The data confirm the ability of abacavir to modify self-proteins and suggest that the molecular mechanism(s) of some abacavir-induced adverse reactions may require bioactivation.

Introduction

The anti-HIV drug abacavir (ABC; Scheme 1) is associated with hypersensitivity reactions (HSRs), characterized by systemic symptoms that include fever, rash, nausea, vomiting, abdominal pain, lethargy and malaise. This allergenic response occurs generally within the first 6 weeks of treatment and has been linked to HLA class I allele HLA-B*57:01 (Nolan, 2009). In addition, long-term abacavir exposure has been associated with an increased risk of myocardial infarction (Costagliola et al., 2010, Islam et al., 2012) that prompted the FDA to issue an alert for heart attack risk in March 2011 (FDA, 2011). The potential for cardiotoxicity raises further concerns about chronic use of the drug and calls for urgent and accurate risk/benefit estimations. Toward this goal, both a thorough understanding of the mechanisms of abacavir-induced toxicity and the development of reliable toxicity biomarkers are required.

Four different models/hypotheses have been considered to explain the mechanism of the immune-mediated adverse reactions induced by abacavir: (1) the hapten/prohapten hypothesis (Uetrecht, 2007); (2) the pharmacological interaction with immune receptors, the p–i hypothesis (Pichler et al., 2006); (3) the danger model, which is complementary of the hapten hypothesis, adding a secondary signal (e.g. an infection) (Pirmohamed et al., 2002), and (4) the altered peptide model (Illing et al., 2012, Mallal and Phillips, 2012, Norcross et al., 2012, Ostrov et al., 2012). The hapten hypothesis considers that drugs or their metabolites are too small to be immunogenic. Instead, they can bind irreversibly to proteins, generating covalent adducts, which act as antigens (Uetrecht, 2007). The p–i hypothesis advocates a non-covalent interaction between the drug and the HLA protein, implying the formation of neo-antigen determinants, without the involvement of the proteasome. The most recent evidence that HLA-B*5701 is able to lodge abacavir led to the proposal of the altered peptide model, suggesting that binding of the parent drug per se can influence the repertoire of self-peptide ligands to be presented and trigger abacavir-induced HSR (Chessman et al., 2008, Yang et al., 2009, Bharadwaj et al., 2012, Illing et al., 2012, Mallal and Phillips, 2012, Ostrov et al., 2012).

Regardless of the specific mechanism of abacavir-induced HSR, it is noteworthy that the immunogenic effects of the drug are dose-dependent (Adam et al., 2012, Norcross et al., 2012, Ostrov et al., 2012). This correlation suggests that abacavir metabolism could be involved. Therefore, a clarification of the drug's metabolic pathways is essential to fully comprehend the mechanisms of abacavir toxicity.

The molecular mechanisms of abacavir-induced cardiotoxicity are even less understood. Although an immune-mediated response can be envisioned, metabolic activation of abacavir to aldehyde intermediates might play a significant role in this context. This hypothesis is consistent with the fact that several aldehydes have been associated with the inception of cardiovascular pathologies. For instance, acetaldehyde, the primary ethanol metabolite, is thought to be involved in the onset of alcoholic cardiomyopathy (Guo and Ren, 2010a). The report of an exacerbated ethanol-induced myocardial contractile dysfunction following acute ethanol exposure, in a murine model with cardiac-specific overexpression of alcohol dehydrogenase (ADH) (Guo and Ren, 2010b), suggests that heart metabolism to aldehydes, and subsequent site-specific protein modification, may underlie xenobiotic-induced cardiotoxicity. Similarly, exposure to the highly reactive α, β-unsaturated aldehyde acrolein, an endogenous product of lipid peroxidation and ubiquitous environmental pollutant, has been shown to cause myocardial dysfunction (Luo et al., 2007).

Abacavir is a guanosine analog, which is converted intracellularly to the pharmacologically active metabolite, carbovir triphosphate (CBV-TP), via stepwise anabolism. Abacavir is initially phosphorylated to abacavir monophosphate (ABC-MP) by adenosine phosphotransferase, followed by deamination via a cytosolic enzyme to form carbovir monophosphate (CBV-MP). Then two subsequent phosphorylations occur, to the diphosphate by guanylate kinase, and finally to carbovir triphosphate (CBV-TP) by nucleoside diphosphate kinase (Scheme 1). The active metabolite competes with endogenous 2′-deoxyguanosine triphosphate for the nucleic acid chain and, upon incorporation, terminates the DNA chain extension (Faletto et al., 1997, Hervey and Perry, 2000). Concurrently with this activation process abacavir is extensively metabolized by the liver. Primary abacavir metabolism pathways comprise Phase II glucuronidation, mediated by uridine diphosphate glucuronyltransferase and yielding an inactive glucuronide metabolite (ABC-glucuronide), and Phase I oxidation, mediated by ADH and yielding a carboxylate (ABC-carboxylate). These metabolites are excreted primarily in the urine, where in combination they account for 66% of the dose; an additional 15% of the dose is converted into a number of minor metabolites (McDowell et al., 1999), but structural considerations suggest that none of those identified to date are potentially reactive intermediates. By contrast, abacavir metabolism to the carboxylate involves a two-step oxidation process, via a reactive aldehyde intermediate (Walsh et al., 2002). The formation of isomers of both the acid metabolite and the parent drug through ADH metabolism led Walsh et al. (2002) to propose a metabolic pathway involving double bond migration and epimerization processes upon formation of two putative aldehyde intermediates: an unconjugated aldehyde and a conjugated aldehyde (Scheme 1).

We have demonstrated previously that the conjugated aldehyde is capable of reacting with proteins in vitro (Charneira et al., 2011) and in rats (Charneira et al., 2012) but, so far, this metabolite has eluded detection in humans. The present work was aimed at identifying abacavir–hemoglobin adducts in HIV-infected patients as biomarkers of abacavir bioactivation to an aldehyde metabolite.