Oxidative Stress leads to Liver Cancer

Moderate.

The types of genotoxic oxidative DNA damage that may occur following exposure to ROS have been extensively reviewed previously (Dizdaroglu 2012, Dizdaroglu 2015). Briefly, ROS can react with nitrogenous bases to produce various adducts that may be converted into a mutation following DNA replication. Further, ROS can damage the sugar phosphate backbone of DNA leading to abasic sites and strand breaks. If DNA damage leads to mutations that increases the expression of oncogenes or decreases the expression of tumour suppressor or DNA damage repair genes, they will transform normal cells into malignant cells. It is generally thought that liver cancer results from an accumulation of mutations in key cancer-driving genes such as TP53 and CTNNB1 (Fujimoto, et al. 2016, Shibata and Aburatani 2014a) (http://atlasgeneticsoncology.org/Tumors/HepatoCarcinID5039.html). 

In addition to DNA damage, at the molecular level, chronic activation of the Nrf2 oxidative stress response has been linked to promoting malignant transformation in pre-cancerous cells. Persistent Nrf2 activation results in the long-term up-regulation of antioxidant genes (which protect cancer cells that are known to have elevated ROS) and phase II metabolism genes (which facilitate the rapid metabolism of chemotherapeutics) (Kansanen, et al. 2013) providing a favourable environment for growth of pre-cancerous cells. The connection between chronically activated Nrf2 and cancer has been extensively studied and reviewed, most recently by Furfaro et al. (2016) and Karin and Dhar (2016). Further, Nrf2 control over cellular proliferation and differentiation has also been studied; reviewed most recently by Murakami and Motohashi  (2015).   

Moderate. 

Carcinogens that cause cancer by ‘cytotoxicity and regenerative proliferation’ are generally accepted to be indirectly genotoxic. The most plausible source of indirect genotoxicity for these compounds are ROS. The following examples involve chemicals that are known rodent and/or human carcinogens. These data demonstrate that exposure leads to oxidative DNA damage, which may cause mutations required for malignant transformation. Overall, a variety of lines of evidence support the relationship between oxidative stress with the development and progression of hepatocellular carcinoma in humans. For example, reduced superoxide dismutase 2 (an antioxidant gene) mRNA and protein expression is correlated with mortality of hepatitis B virus-associated hepatocellular carcinoma patients in a mutant p53-dependent manner (Wang, et al. 2016b). This decrease in expression is accompanied by decreased copy number of the gene in tumours, supporting a genetic basis for the molecular phenotype. Plasma protein carbonyl content (biomarker of oxidative stress) is significantly higher, whereas plasma TAC (biomarker of antioxidant capacity) is significantly lower in hepatitis B virus-associated HCC patients with than healthy controls (Poungpairoj, et al. 2015). Exposure of human HepG2 cells to 2-Amino-9H-pyrido[2,3-b]indole (AαC) leads to increased levels of ROS and 8-Hydroxydesoxyguanosin (8-OHdG), and decreased GSH/GSSG ratio (Zhang, et al. 2015). Subsequent investigations showed that the concentration of 8-OHdG was positively related to DNA damage measured by the comet assay, indicating a role for genotoxicity in cancer for this compound.

Furan is not directly genotoxic; any observed genotoxicity following furan exposure is thought to be due to ROS-induced DNA damage. Oxidative DNA adducts (8-oxo-dG) have been observed following a three month exposure to a high dose of furan (30 mkd) (Hickling, et al. 2010). These 8-oxo-dG adducts were observed together with high levels of Cyp2E1 expression (the potential source of the ROS), apoptosis, necrosis, and proliferation in rat livers. Male mice exposed to a carcinogenic dose of furan (8 mkd) presented with increased levels of 8-OHdG DNA adducts, which could be prevented through co-treatment with apigenin (Wang, et al. 2014); global gene expression analysis of mice exposed at 8 mkd have also been shown to have increased expression of the Nrf2 oxidative stress response pathway (Jackson, et al. 2014). Rats exposed to furan (0, 2, 4, 8, 12 and 16 mkd) presented with dose-dependent increases in oxidized nitrogenous bases and double strand breaks, which were repaired in a time-dependent manner and in conjunction with an increase in expression of DNA repair genes (Ding, et al. 2012). Increases in DNA oxidative lesions, and supporting evidence from measures of gene expression changes, occur at doses that are aligned with those causing liver cancer in the same rodent models (furan-dependent liver cancer occurs at 2 mkd and up in rats and 4 mkd in mice) (Moser, et al. 2009, NTP 1993). Further, they occur at earlier time points than the corresponding carcinogenesis.

By-products of lipid peroxidation (Malondialdehyde (MDA), 4-Hydroxynonenal (4-HNE)) react with DNA to form etheno-DNA adducts (particularly, εdA and εdC), these adducts have been observed in vitro and in human patients, and are correlated with hepatocarcinogenesis (Linhart, et al. 2014, Wang, et al. 2009, Winczura, et al. 2012). Male Wistar rats exposed for one month to 35% ethanol had abasic sites, Ogg1-sensitive sites, and increased expression of base excision repair (BER) genes in liver DNA; this ROS-dependent DNA damage occurs at earlier time points than the corresponding carcinogenesis.

Chloroform and carbon tetrachloride cause GSH depletion in human HepG2 cells, which leads to ROS and subsequent DNA strand breaks, 8-oxodeoxyguanosine (8-oxodG) DNA adducts, malondialde hydedeoxyguanosine (M1dG) DNA adducts, and lipid peroxidation (Beddowes, et al. 2003). Using cultured hepatocytes from female Wistar rats, the same study reported dose-dependent increases in glutathione depletion and lipid peroxidation (LPO) following treatment with chloroform (0, 0.8, 2.4, 4, 8, 20 mM; results reported 2 hours following exposure) or carbon tetrachloride (0, 0.25, 1, 4 mM; results reported 2 hours following exposure).  Further, the chloroform-exposed cells had dose-dependent increases in M1dG adducts and DNA strand breakage; and the carbon tetrachloride-exposed cells had dose-dependent increases in M1dG adducts, DNA strand breaks, and 8-oxodG adducts. Dose concordance of chloroform was demonstrated and significant changes occurred at the following doses: LPO (4, 8, 20 mM), decreased cell viability (20 mM), GSH depletion (20 mM), COMET (8, 20 mM), M1dG levels (4, 8, 20 mM). Further, GSH depletion was measured over a 24 hour period and was significantly decreased at 4, 8 and 24 hours following exposure to chloroform. Similar results were reported for carbon tetrachloride; however the data was not shown. Significant changes were reported for: LPO (4mM mM), decreased cell viability (4 mM), COMET (1 and 4 mM), M1dG levels (1 and 4 mM), and 8-oxo-dG (4 mM) following exposure to carbon tetrachloride (Beddowes, et al. 2003). Carbon tetrachloride-dependent oxidative adducts have also been reported elsewhere (Takahashi, et al. 1998, Wacker, et al. 2001). These studies demonstrate that oxidative damage to DNA occurs before liver cancer.

The transcriptional actions of chronic Nrf2 activation provide a molecular environment that promotes malignant transformation (Furfaro, et al. 2016, Jaramillo and Zhang 2013, Xiang, et al. 2014). Most recently, this relationship was outlined in an award-winning article that was published in the journal Carcinogenesis: The 2016 Carcinogenesis Award Winners “Liver carcinogenesis: from naughty chemicals to soothing fat and the surprising role of NRF2” (Karin and Dhar 2016). Following the analysis of the empirical data, the authors conclude that NRF2 is required for hepatocellular carcinoma precursor cells (HcPC) to progress to become hepatocellular carcinoma cells; in fact, without NRF2 excess ROS accumulate and kill the HcPC cells. Importantly, sequencing of human hepatocellular carcinoma patients has demonstrated NRF2 activation (Schulze, et al. 2015, Shibata and Aburatani 2014b, Totoki, et al. 2014). Following furan exposure at carcinogenic doses (4, 8 mkd), the NRF2 oxidative stress pathway was the most highly implicated molecular pathway following microarray analysis of gene expression in mice exposed for three weeks (Jackson, et al. 2014), which demonstrates that the NRF2 oxidative stress pathway becomes active after chemical exposure and before the development of hepatocellular carcinoma. Further, the level of induction of the NRF2 pathway (i.e., the number of implicated genes) increased with dose, which demonstrates dose dependence of the response, which is concordant with the dose-dependent increases in hepatocellular adenoma and carcinoma. While the mechanistic details remain under investigation, there is clearly an important interplay between the two branches of this response (i.e., ROS-dependent mutations and the NRF2 oxidative stress response).

Not all agents that cause ROS in the liver cause liver cancer. Thus, there are additional modulating factors that must be considered when determining whether a ROS-producing chemical will cause liver cancer.

Overall, ROS-dependent DNA damage causing harmful mutations is known to occur. However, the specific mechanism and the quantitative relationships by which these mutations promote malignant transformation are incompletely understood.

Increase in NRF2 expression is associated with occurrence and recurrence of hepatocellular carcinoma; however, the mechanism is incompletely understood.