Comments on “DNA-binding activities of compounds acting as enzyme inhibitors, ion channel blockers and receptor binders.”

Abstract

I read with interest the article “DNA-binding activities of compounds acting as enzyme inhibitors, ion channel blockers and receptor binders” recently published in Chemico-Biological Interactions. The authors suggested that acetaminophen, one of the most used drugs worldwide, alkylates DNA at therapeutic doses and is genotoxic. Given the implications of this statements for public health, it is important for the reader to hear a different perspective that is based on the entire literature on this subject. Everything considered, there is no credible evidence that acetaminophen is a genotoxic hazard or a carcinogen at therapeutic doses.

Dear Editor:

I read with interest the recent review by Muhamedejevs et al. [1] on DNA-binding activities of various drugs and chemicals. The authors discuss, among other drugs, acetaminophen (APAP, paracetamol) and give the reader the impression that based mainly on literature from the 1980s and 1990s, acetaminophen directly binds to DNA, is pronounced genotoxic and even alkylates DNA at therapeutic doses [1]. Citing a newer paper from my laboratory [2], the authors claim that the APAP-induced “DNA lesions are considered to be the main cause of the drug hepatotoxicity in humans” [1].

APAP is present in hundreds of prescription and over-the-counter medications available in virtually every country around the world and is one of the most widely consumed drugs. Thus, these statements by the authors, if correct, would be very concerning and would imply an increased cancer risk when taking even therapeutic doses of APAP. Given that tens of millions of people take APAP daily in one form or another, if APAP would be “pronounced genotoxic” at therapeutic doses as suggested by the authors, APAP could be considered a significant public health hazard. Fortunately, this topic has been recently quite thoroughly reviewed by a large panel of experts in various fields including liver toxicology, drug metabolism, genetic toxicology, chemical carcinogenesis and epidemiology [3–5, 7–10]. In these reviews it was extensively discussed how only an overdose of APAP that causes severe liver injury triggers through reactive metabolite formation a mitochondrial oxidant stress and mitochondrial dysfunction [3]. This causes the cytosolic release of mitochondrial intermembrane proteins like endonuclease G, which then translocate to the nucleus and cause DNA degradation as the final step of necrotic cell death [3]. Importantly, there is no evidence that the reactive metabolite of APAP, N-acetyl-p-benzoquinone imine (NAPQI), generated mainly in the endoplasmic reticulum or any oxidants such as reactive oxygen and peroxynitrite generated exclusively inside mitochondria can ever reach the nucleus to directly affect nuclear DNA [3]. These adverse effects are strictly dose-dependent in mice and in humans and are only relevant for toxic doses, i.e., when the affected cells die [3,4]. Although the reactive metabolite and protein adducts are also formed to a limited degree after therapeutic doses, the endogenous antioxidant defenses and autophagy effectively prevent any adverse effects at therapeutic doses and there is no chance that low levels of the reactive metabolite can ever reach the nucleus to alkylate DNA [3]. Consistent with these mechanistic studies, weight-of-evidence assessment of long-term rodent carcinogenicity and tumor initiation/promotion studies did not show an increased carcinogenic risk of APAP in animals [5]. Several studies published in the late 1980s and early 1990s suggested that APAP is genotoxic [6,7]. However, as early as 1996, Bergman et al. [6] pointed out that “paracetamol does not cause gene mutations, either in bacteria or in mammalian cells”. Although there are some reports indicating that APAP causes genotoxicity in mammalian systems, there is “convincing evidence that genotoxic effects of paracetamol appear only at dosages inducing pronounced liver and bone marrow toxicity and that the threshold level for genotoxicity is not reached at therapeutic dosage” [6]. This earlier analysis was recently expanded to the newest literature using a weight of evidence approach [7], where “studies are evaluated based on quality, reproducibility and consistency, significance of the genetic alteration, phylogenetic relevance to humans, type of test system (in vivo versus in vitro, cell type, p53 status etc.), and relevance of the route of administration for in vivo studies” [7]. Evaluating a total of 69 studies, which investigated the genotoxic potential of APAP to date, it was again evident that any genotoxic effect of APAP only occurs at dose levels that induce cell toxicity, and it was thus concluded that “it is extremely unlikely that APAP induces the stable, genetic damage that would be indicative of a clear genotoxic carcinogenic hazard in humans” [7]. In addition to these observations and analysis of the literature, no epidemiological association between APAP and cancer was observed [8,9]. Thus, when all scientific evidence generated during the last 50 years was considered, it was concluded that “acetaminophen is not a carcinogenic hazard at any dose level” [10]. Therefore, the authors’ statement that APAP is “pronounced genotoxic” [1] has to be put in context, i.e., the genotoxicity of APAP is only observed with cytotoxic doses, and therefore, does not represent a carcinogenic hazard for humans at therapeutic doses. Likewise, the comment that APAP “even alkylates DNA at therapeutic doses” is a misinterpretation of the original study [11]. Rogers et al. [11] never showed alkylation of DNA by APAP or its metabolite. What they showed was that after administration of tritium labeled APAP, there was a small amount of radioactivity detectable in the nucleus. However, given that a tritium label can easily move to other molecules through a hydrogen exchange, radioactivity in the nucleus is not proof that APAP bound to the nucleus [7]. The authors never identified or characterized the hypothesized DNA adducts [11]. Furthermore, in a separate experiment, the authors clearly demonstrated that when nuclei were exposed to radioactive NAPQI, the label was more than 90% on the nuclear proteins and not on the DNA [11], which is consistent with the well-known reactivity of NAPQI with cysteine residues of proteins [12].

The claim that “DNA lesions are considered to be the main cause of the drug hepatotoxicity in humans [2]” is also a misinterpretation of our study. What is really shown is that in humans [2], similar to mice [13], DNA fragmentation leading to karyolysis occurs after an APAP overdose in response to irreversible mitochondrial dysfunction and presents the point of no-return to necrotic cell death because this severe DNA degradation cannot be repaired. Thus, the DNA fragmentation is a consequence of the mitochondrial dysfunction and damage, which is the central cause of cell death [3, 13].

I have focused on APAP because of its clinical importance and the available large body of work related to mechanisms of its toxicity and assessment of its genotoxic potential. Given the potential consequences when making claims of DNA alkylation or genotoxicity about any chemical, especially for such a widely used drug as APAP, it is imperative to cautiously interpret the entire body of literature on the subject and not rely only on a few selected highly questionable studies.