The role of base excision repair genes OGG1, APN1 and APN2 in benzo[a]pyrene-7,8-dione induced p53 mutagenesis

Abstract

Lung cancer is primarily caused by exposure to tobacco smoke. Tobacco smoke contains numerous carcinogens, including polycyclic aromatic hydrocarbons (PAH). The most common PAH studied is benzo[a]pyrene (B[a]P). B[a]P is metabolically activated through multiple routes, one of which is catalyzed by aldo-keto reductase (AKR) to B[a]P-7,8-dione (BPQ). BPQ undergoes a futile redox cycle in the presence of NADPH to generate reactive oxygen species (ROS). ROS, in turn, damages DNA. Studies with a yeast p53 mutagenesis system found that the generation of ROS by PAH o-quinones may contribute to lung carcinogenesis because of similarities between the patterns (types of mutations) and spectra (location of mutations) and those seen in lung cancer. The patterns were dominated by G to T transversions, and the spectra in the experimental system have mutations at lung cancer hotspots. To address repair mechanisms that are responsible for BPQ induced damage we observed the effect of mutating two DNA repair genes OGG1 and APE1 (APN1 in yeast) and tested them in a yeast reporter system for p53 mutagenesis. There was an increase in both the mutant frequency and the number of G:C/T:A transversions in p53 treated with BPQ in ogg1 yeast but not in apn1 yeast. Knocking out APN2 increased mutagenesis in the apn1 cells. In addition, we did not find a strand bias on p53 treated with BPQ in ogg1 yeast. These studies suggest that Ogg1 is involved in repairing the oxidative damage caused by BPQ, Apn1 and Apn2 have redundant functions and that the stand bias seen in lung cancer may not be due to impaired repair of oxidative lesions.

Introduction

The major risk factor for lung cancer is exposure to tobacco smoke [1]. There are about 50 known carcinogens in tobacco smoke, but some of the most established are the polycyclic aromatic hydrocarbons (PAH). PAH are ubiquitous combustion products and are also found in charbroiled foods, coal smoke and car exhaust [2], [3]. The most common PAH found in tobacco smoke is benzo[a]pyrene (B[a]P). B[a]P does not react with DNA and must undergo metabolic activation to become mutagenic. There are three metabolic pathways that B[a]P can go through to become what are known as ultimate carcinogens, which are compounds that directly bind to and damage DNA. The first pathway involves the conversion of B[a]P to radical cations, which can form depurinating adducts by utilizing P450 peroxidases [4], [5]. The second and third pathways involve the formation of an intermediate product known as B[a]P-7,8-trans-dihydrodiol (Diol), through the combined action of cytochrome P450 1A1 (CYP1A1) and epoxide hydrolases [6], [7]. B[a]P-7,8-trans-dihydrodiols can either be metabolized to (±)anti-BPDE through the actions of CYP1A1 and CYP1B1 or metabolized by aldo-keto reductases (AKRs) to form another intermediate product, catechols [8]. (±)anti-BPDE is highly mutagenic and forms bulky adducts with DNA [9], [10]. Catechols can undergo two spontaneous oxidation reactions to form B[a]P-7,8-dione (BPQ) [8]. BPQ can form both stable and depurinating adducts [11], [12], however, based on measurements of oxidized macromolecules, the majority of the DNA damage that occurs from BPQ is through the production of reactive oxygen species (ROS), which is generated during a futile redox cycle in the presence of NADPH. Under redox cycling conditions, catechols spontaneously oxidize to o-quinones and are reduced back to catechols through enzymatic or non-enzymatic reduction in the presence of NADPH [8], [13], [14], [15].

Each pathway, diol epoxides, quinones and radical cations metabolizes PAH at comparable rates in cells [16]. The diol epoxide pathway is supported by studies showing that B[a]P-7,8-trans-dihydrodiol epoxide adducts are found in smokers lungs and the location of DNA adducts can be mapped to known hotspots on the tumor suppressor p53 [17]. Data showing that smoking causes oxidative stress, which can be measured by examining antioxidant levels [18] and elevated levels of the oxidative lesion 8-oxo-2′-deoxyguanosine (8-oxo-dGuo), support the PAH o-quinone pathway [19], [20], [21], [22], [23], [24]. There have also been reports that products of radical cation damage and depurinating adducts are present in PAH treated mice and cells [5], [25]. Lesions caused by both anti-BPDE adducts and ROS cause G to T transversions on p53 [26], the major mutation found on p53 in lung cancer [27], [28], while the radical cation pathway is less mutagenic [29].

For mutations to lead to cancer, they must occur in key driver genes. The most commonly mutated gene in lung cancer is the tumor suppressor p53 [30]. p53 is a transcription factor responsible regulating cell cycle progression and apoptosis. Mutations in p53 result in unregulated cell cycle progression and may lead to carcinogenesis. Although p53 is mutated in many cancers, there are three features in lung cancer that result in a “signature” [27]. The first feature is that most of the mutations are G to T transversions. G to T transversions are rare in most other cancers. The second feature is that there is a strand bias seen on p53 in lung cancers. A strand bias occurs when there are more mutations on the coding strand compared to the transcribed strand. Specifically, there are more guanines that are mutated on the coding strand compared to the transcribed strand of p53 in lung cancers [31]. This is reflected in the observation that there are more G to T transversions than the reciprocal transversions C to A. The third feature of the p53 signature is that there are hotspot codons, in that about 23 codons account for about 50% of all mutations. The main hotspot codons include, but are not limited to, codon 157, 158, and 248 and the majority of these mutations are G to T transversions. However, the hotspot codons on p53 are also mutated in other cancers so this attribute is not unique to lung cancers [30].

Reactive oxygen species may play a key role in the induction of lung cancer by generating 8-oxo-dGuo. 8-oxo-dGuo is usually repaired by the base excision repair (BER) pathway. Two DNA repair genes in the BER pathway that repair oxidative lesions are OGG1 and APE1 (APN1 in yeast). Ogg1 is a bi-functional glycosylase, in that it has both AP lyase and DNA glycosylase activity and one of its functions is to excise and remove 8-oxo-dGuo from DNA [32], [33]. It has been reported that there is a loss of heterozygosity of OGG1 in small cell lung cancers, and low levels of Ogg1 activity are associated with an increased risk of cancer [34], [35]. This suggests that not only is there an increase in ROS and oxidative damage during the induction of lung cancer, but also that reduced efficiency of oxidative damage repair is a contributing factor in carcinogenesis.

The other BER gene, APE1, is a type II apurinic/apyrimidinic (AP) endonuclease. Ape1 cleaves the 5′ end of an AP site after a damaged base is removed by a mono-functional glycosylase [36]. Ape1 works in conjunction with Ogg1, a bi-functional glycosylase, and other BER pathway enzymes to remove oxidative lesions such as 8-oxo-dGuo. Ape1, through its 3′ phosphodiesterase activity, removes the 3′ blocking end of the AP site, which is formed after Ogg1 removes the 8-oxo-dGuo formed during oxidative damage [37], [38]. Ape1 can enhance the glycosylase activity of Ogg1 suggesting the two genes work together to repair the oxidative damaged caused by ROS [39], [40].

In this study we addressed the role of the BER enzymes, Ogg1, Apn1 (yeast homolog of Ape1) and Apn2 in BPQ induced p53 mutagenesis. Using a yeast system that utilizes a red/white selection paradigm to test for mutations in p53 cDNA, we determined if knocking out OGG1, APN1 or APN1 and APN2 affected the mutant frequency, the mutation pattern and mutation spectrums of BPQ treated p53. To do this, we treated p53 cDNA with BPQ under redox cycling conditions and measured the mutant frequencies of p53 in wild type (yIG397), ogg1, apn1 and apn1apn2 yeast strains. We then isolated and sequenced the mutant p53 to determine the mutation patterns and spectrums of the mutations. The loss of Ogg1, but not Apn1, increased the mutant frequency and the incidence of G to T transversions in BPQ mutagenesis of p53. Although the loss of APN1 did not result in increase in mutant frequency, the loss of APN1 and APN2 increased mutant frequency of p53 approximately threefold compared to wild type. This suggests that Ogg1 plays a major role in repairing BPQ induced DNA damage while Apn1 and Apn2 have a redundant role in repairing oxidative damage caused by BPQ.