Epigenetic and epitranscriptomic mechanisms of chromium carcinogenesis

Abstract

Hexavalent chromium [Cr(VI)], a Group I carcinogen classified by the International Agency for Research on Cancer (IARC), represents one of the most common occupational and environmental pollutants. The findings from human epidemiological and laboratory animal studies show that long-term exposure to Cr(VI) causes lung cancer and other cancer. Although Cr(VI) is a well-recognized carcinogen, the mechanism of Cr(VI) carcinogenesis has not been well understood. Due to the fact that Cr(VI) undergoes a series of metabolic reductions once entering cells to generate reactive Cr metabolites and reactive oxygen species (ROS) causing genotoxicity, Cr(VI) is generally considered as a genotoxic carcinogen. However, more and more studies have demonstrated that acute or chronic Cr(VI) exposure also causes epigenetic dysregulations including changing DNA methylation, histone posttranslational modifications and regulatory non-coding RNA (microRNA and long non-coding RNA) expressions. Moreover, emerging evidence shows that Cr(VI) exposure is also capable of altering cellular epitranscriptome. Given the increasingly recognized importance of epigenetic and epitranscriptomic dysregulations in cancer initiation and progression, it is believed that Cr(VI) exposure-caused epigenetic and epitranscriptomic changes could play important roles in Cr(VI) carcinogenesis. The goal of this chapter is to review the epigenetic and epitranscriptomic effects of Cr(VI) exposure and discuss their roles in Cr(VI) carcinogenesis. Better understanding the mechanism of Cr(VI) carcinogenesis may identify new molecular targets for more efficient prevention and treatment of cancer resulting from Cr(VI) exposure.

1. Introduction

Chromium (Cr) is a naturally occurring element found in rocks, soil, water, plants, and animals. Cr mainly exists in combination with other elements to form various compounds. The three main forms (valence) of Cr are: Cr(0), Cr(III), and Cr(VI). Hexavalent chromium [Cr(VI)] has been widely used in many manufacturing processes to produce various metal alloys such as stainless steel. Cr(VI) has also been widely used in making various consumer products such as chromate-treated wood, leather and Cr(VI)-containing metal-on-metal hip replacements. Due to its widespread industrial use, a large amount of Cr(VI) has been released into the environment. In addition, Cr(VI) is one of toxic chemicals in cigarette smoke. Thus, Cr(VI) is considered one of the most common environmental and occupational pollutants. It is estimated that over a thousand hazardous waste sites on U.S. National Priority List contain high levels of Cr(VI) (ATSDR, 2012, 2019). In fact, the U.S. Environmental Protection Agency (EPA) and the Agency for Toxic Substances and Disease Registry (ATSDR) list Cr(VI) as one of “Top 20 Hazardous Substances” in waste sites on the U.S. National Priority List (ATSDR, 2019). Occupational workers’ exposure to Cr(VI) in industries using or manufacturing Cr(VI) mainly occurs via inhalation. General population exposure to Cr(VI) mostly happens through living near a hazardous waste facility that contains Cr(VI), cigarette smoking and consuming Cr(VI)-contaminated drinking water.

Short- and long-term exposures to Cr(VI) cause serious adverse health effects. Acute exposure to high levels of Cr(VI) may cause dermal burns; eye, respiratory and gastrointestinal tract irritation; anemia; damages or failure of the liver, kidneys and the reproductive system. However, the most significant health concern of exposure to Cr(VI) is its carcinogenicity. Based on the findings from human epidemiological studies and animal experimental studies, the International Agency for Research on Cancer (IARC) has classified Cr(VI) as a Group I human carcinogen (IARC, 1990). Chronic Cr(VI) exposure causes lung cancer and other types of cancer. Although Cr(VI) is a well known carcinogen, the mechanism by which Cr(VI) exposure causes cancer has not been well understood.

The earlier studies on the mechanism of Cr(VI) carcinogenesis mostly focused on its genotoxic effects. This is mainly due to the fact that Cr(VI) undergoes a series of metabolic reductions inside cells to generate reactive Cr metabolites and reactive oxygen species (ROS), which produce various genotoxic effects and are thought playing important roles in Cr(VI) carcinogenesis ( Jomova & Valko, 2011; Nickens, Patierno, & Ceryak, 2010; Shi, Hudson, & Liu, 2004; Wise, Holmes, & Wise, 2008; Yao, Guo, Jiang, Luo, & Shi, 2008; Zhitkovich, 2005). However, subsequent studies showed that Cr(VI) exposure also causes non-genotoxic effects such as epigenetic and more recently epitranscriptomic effects (Brocato & Costa, 2013; Chen, Murphy, Sun, & Costa, 2019; Chervona, Arita, & Costa, 2012; Humphries, Wang, & Yang, 2016; Wang, Uddin, et al., 2022; Wang & Yang, 2019). Given the well-recognized critical roles of epigenetic dysregulations and the increasingly-recognized crucial roles of epitranscriptomic dysregulations in cancer initiation and progression (Avgustinova & Benitah, 2016; Banerjee, Smith, Eccles, Weeks, & Chatterjee, 2022; Butera, Melino, & Amelio, 2021; Dawson & Kouzarides, 2012; Nebbioso, Tambaro, Dell’Aversana, & Altucci, 2018; Shukla & Meeran, 2014; Skourti & Dhillon, 2022; Uddin, Wang, & Yang, 2021a; Zabransky, Jaffee, & Weeraratna, 2022), it is highly likely that Cr(VI) exposure-caused epigenetic and epitranscriptomic dysregulations could play important roles in Cr(VI) carcinogenesis (Rager, Suh, Chappell, Thompson, & Proctor, 2019; Wang & Yang, 2019). The goal of this chapter is to provide a critical review on the epigenetic and epitranscriptomic effects of Cr(VI) exposure and discuss their roles in Cr(VI) carcinogenesis.

2. Epigenetics and epitranscriptomics

2.1. Epigenetics

Epigenetic modifications regulate gene expression without changing DNA sequences, rather being mediated by DNA methylation, histone post-translational modifications (acetylation, methylation, etc.), and non-coding RNAs such as microRNAs (miRNAs) and long non-coding RNAs (lncRNAs). Since the initial report showed the increased hypomethylation in genes of human tumor cells compared with their normal counterparts in 1983 (Feinberg & Vogelstein, 1983), many studies have demonstrated the importance of epigenetic dysregulations in cancer initiation and progression (Baylin & Jones, 2016; Butera et al., 2021; Dawson & Kouzarides, 2012).

Cancer is now considered as a stem cell disease and it has been proposed that cancer is initiated by cancer stem cells (CSCs) (Kreso & Dick, 2014; Nguyen, Vanner, Dirks, & Eaves, 2012). CSCs are a small population of cancer cells that have characteristics associated with normal stem cells, in particular, the capability of self-renewal and generation of different types of cells found in a tumor. Although it remains to be determined where and how CSCs are produced, studies showed that epigenetic dysregulations play important roles in generating CSCs or producing CSC-like cells (Avgustinova & Benitah, 2016; French & Pauklin, 2021; Shukla & Meeran, 2014; Wang & Yang, 2019).

While initial studies reported DNA global hypomethylation in human tumor tissues, subsequent studies identified DNA hypermethylation at specific genomic locus. In fact, increased DNA methylation of specific genomic sites that are usually not methylated in normal tissues is frequently observed in almost all tumor types (Feinberg & Tycko, 2004). Many studies revealed that increased DNA methylation usually happens at gene promoter CpG islands leading to transcriptional repression, reducing gene expression such as tumor suppressor genes, thus promoting tumor development and progression (Baylin & Jones, 2016). DNA methylation is dynamically regulated by DNA methyltransferases (DNMTs) and demethylases: DNMTs that deposit the methyl groups to CpG islands are called writers; and demethylases that remove methyl groups from CpG islands are called erasers (Biswas & Rao, 2018). The final functional outcome of DNA methylation is mediated by a group of proteins known as methyl CpG binding proteins (MBPs) defined as readers (Biswas & Rao, 2018). Abnormal expression levels or activities of DNA methylation writers, erasers or readers change DNA methylation levels and outcomes, playing important roles in cancer initiation and progression.

The epigenetic mechanism of histone posttranslational modifications (PTMs) refers to reversible modifications of nuclear core histone proteins that regulate gene expression by dynamically changing chromatin conformational structures (Millán-Zambrano, Burton, Bannister, & Schneider, 2022; Xu, Du, & Lau, 2014). About 20 types of histone PTMs are identified and histone methylation and acetylation are the most widely studied histone PTMs. Different histone PTMs exhibit different impacts on chromatin structures thus displaying different effects on gene expression. For example, acetylation of histones H3, H4 and methylation of H3 lysine 4 (H3K4) are usually associated with euchromatin structures and gene expression, but methylation of H3 lysine 9 (H3K9) and H3 lysine 27 (H3K27) cause the compaction of chromatin reducing gene expression through the recruitment of heterochromatin protein 1 and polycomb group proteins. Similarly, histone PTMs are also dynamically regulated by histone PTM writers and erasers and the functional outcomes of histone PTMs are mediated by the corresponding histone PTM reader proteins. Many studies have shown that abnormal expression levels or activities of histone PTM writers, erasers or readers dysregulate histone PTM levels and outcomes, playing important roles in cancer initiation and progression (Biswas & Rao, 2018; Neganova, Klochkov, Aleksandrova, & Aliev, 2022; Waldmann & Schneider, 2013).

Non-coding RNAs especially microRNAs (miRNAs) and long non-coding RNAs (lncRNAs) are additional epigenetic mechanisms regulating gene expression. By definition, non-coding RNAs refer to RNA transcripts that do not have significant protein coding capacities (Humphries et al., 2016; Wang, Wang, & Yang, 2021). The lncRNAs are non-coding RNAs longer than 200bps in length. Mechanistically, miRNA regulation of gene expression occurs mostly at the posttranscriptional level through the interaction of a miRNA’s seed sequence with the 3′-untranslated region (3′-UTR) of its target gene. In contrast, lncRNA regulation of gene expression could happen at transcriptional, posttranscriptional, translational and posttranslational levels. Functionally, miRNAs and lncRNAs are critically involved in regulating almost all important biological processes (Bartel, 2004; Kozomara, Birgaoanu, & Griffiths-Jones, 2019). As a result, abnormal expressions of miRNAs and lncRNAs contribute significantly to the development and progression of many diseases especially cancer (Statello, Guo, Chen, & Huarte, 2021; Tsagakis, Douka, Birds, & Aspden, 2020).

2.2. Epitranscriptomics

The epitranscriptome includes all forms of chemical modifications of all RNA transcripts (the transcriptome) inside a cell. The first RNA chemical modifications were reported in the 1950s and more than 150 types of chemical modifications in RNA molecules have now been identified (Boccaletto & Baginski, 2021; Boccaletto et al., 2018). Recent progresses in revealing the dynamic and reversible feature of RNA modifications and their important roles in a diverse of biological processes revolutionized our understanding on RNA biology and function (Shi, Wei, & He, 2019; Uddin et al., 2021a; Uddin, Wang, & Yang, 2021b; Zhao, Roundtree, & He, 2017). The recognition of critical biological functions of RNA modifications led to the birth of terms of “RNA epigenetics,” “Epitranscriptome” and “Epitranscriptomics” (Roundtree & He, 2016; Saletore et al., 2012).

Unlike the effects of DNA and histone core protein modifications on gene expression that occur mainly at transcriptional levels, the effects of RNA modifications on gene expression happen at posttranscriptional levels. RNA modifications may affect RNA splicing, RNA stability, and RNA translation (Uddin et al., 2021a). As a result, RNA modifications could up-regulate or down-regulate gene expressions. On the other hand, similar to DNA methylation and histone PTMs, the functional consequences of RNA modifications are also determined by three groups of proteins also known as the RNA modification machinery consisting of “writers” (the enzymes that deposit RNA modifications), “erasers” (the enzymes that remove RNA modifications) and “readers” (the proteins that interact with RNA modifications and mediate the functions of modifications) (Shi et al., 2019).

Among more than 150 types of RNA modifications, the N⁶-methyladenosine (m⁶A) modification is now recognized as the most common internal modification in eukaryotic messenger RNAs (mRNAs) (Uddin, Wang, & Yang, 2020; Uddin et al., 2021a). The deposit of m⁶A in RNA molecules is achieved by a multicomponent methyltransferase complex known as the “m⁶A writer,” which catalyzes the transfer of a methyl group from S-adenosylmethionine to the N-6 position of adenosine (Shi et al., 2019). The m⁶A writer complex consists of several methyltransferases including methyltransferase like-3 (METTL3) and METTL14, Wilms’ tumor 1-associating protein (WTAP) and other components. METTL3 is recognized as the major component playing the catalytic function in the writer complex. However, the other components of the complex are also essential for achieving the m⁶A installation in RNA molecules. Two erasers (demethylases) that remove the m⁶A from RNA molecules are fat mass and obesity-associated protein (FTO) and AlkB homolog 5 (ALKBH5) ( Jia et al., 2011; Zheng et al., 2013). The readers that mediate the functional outcomes of the m⁶A modification include the YTH domain-containing family proteins1–3 (YTHDF1–3), YTH domain-containing protein1–2 (YTHDC1–2) and some others (Uddin et al., 2021a; Yang, Hsu, Chen, & Yang, 2018; Zaccara, Ries, & Jaffrey, 2019).

With the development of the specific anti-m⁶A antibody and advances in high throughput sequencing technologies, it is now known that m⁶A modification is mostly distributed in 3′UTRs (3′untranslated regions) and near stop codons (Meyer et al., 2012). Many studies have shown that RNA m⁶A modification plays important roles in normal development and diseases as well (Uddin et al., 2021b). In particular, abnormal levels of m⁶A modification in RNA molecules have been shown to contribute significantly to cancer initiation, metastasis and therapeutic responses (Uddin et al., 2020, 2021a). Although the importance of abnormal DNA and protein modifications in environmental carcinogenesis has been well appreciated, little is known about the role of dysregulated RNA modifications (such as the m⁶A modification) in environmental carcinogenesis especially metal carcinogenesis (Yang, 2020).

3. Epigenetic mechanisms of hexavalent chromium carcinogenesis

3.1. Effects of hexavalent chromium exposure on DNA methylation

The first epigenetic effect of Cr(VI) was demonstrated by its capacity of inducing DNA methylation. Using a V79-derived mammalian cell line (G12) containing a bacterial gpt reporter gene in its DNA, Klein et al. reported for the first time that exposure to 20–50μM of potassium chromate for 2h increases the methylation of the gpt transgene and reduces the gpt expression (Klein, Su, Bowser, & Leszczynska, 2002). It was further determined that Cr(VI)-induced methylation variants are subject to 5-azacytidine–mediated reversion of gpt transgene silencing (Klein et al., 2002). This study provided the first evidence demonstrating that Cr(VI) is also capable of causing epigenetic dysregulations although it has been generally considered as a genotoxic carcinogen. A subsequent study showed that exposure to 10–200mg/L of potassium chromate for 3days also caused a genome-wide DNA hypermethylation in Brassica napus L. plants in a dose-dependent manner (Labra et al., 2004).

While these initial two studies demonstrated the capability of Cr(VI) exposure causing DNA methylation, its significance in Cr(VI) carcinogenesis is not clear. One mechanism of DNA hypermethylation contributing to cancer initiation and progression is to silence tumor suppressor gene expression. Studies from Dr. Kazuya Kondo’s group using human lung tumor samples from workers exposed to chromium revealed that DNA methylation levels are significantly increased in the promoter regions of several tumor suppressor genes (Ali et al., 2011; Kondo et al., 2006; Takahashi et al., 2005; Tsuboi et al., 2020). Using a combined bisulfite restriction analysis (COBRA, a sensitive method that allows infrequent cells bearing methylated alleles to be detected among a majority of cells with unmethylated alleles), Takahashi et al. reported that 5 out of 8 chromate lung cancers display methylation of the human mutL homolog 1 (hMLH1) gene promoter. In addition, four of five chromate lung cancers with methylation of the hMLH1 gene promoter showed repression of hMLH1 protein levels (Takahashi et al., 2005). Using a methylation-specific PCR method, Kondo et al. found that none of lung cancer patients with chromate exposure of less than 15years have p16 gene promoter methylation; in contrast, about 40% of lung cancer patients with chromate exposure more than 15years have p16 gene promoter methylation (Kondo et al., 2006). It was further determined that it was chromate exposure but not cigarette smoking that caused p16 gene promoter methylation in chromate lung cancer. Moreover, the majority of the chromate lung cancers with p16 methylation (85.7%) displayed repression of the p16 protein levels as determined by p16 immunohistochemistry staining (Kondo et al., 2006). These two studies provided initial evidence suggesting that Cr(VI) exposure may increase DNA methylation in human lungs. A later cell culture study also reported that human bronchial epithelial cells (16HBE) treated with 1.2–20μM of dichromate (Cr₂O₇²⁻) display significantly increased DNA methylation in p16 gene promoter (Hu et al., 2016).

Using a nested-methylation-specific PCR assay, Ali et al. simultaneously analyzed methylation status of CpG islands in adenomatous polyposis coli (APC), O⁶-methylguanine-DNA methyltransferase (MGMT), and hMLH1 genes in 36 chromate lung cancers and 25 non-chromate lung cancers (Ali et al., 2011). It was found that a higher frequency of methylation in chromate lung cancers is detected at 86% for APC, 20% for MGMT, and 28% for hMLH1; however, a much lower frequency of methylation in non-chromate lung cancers is observed especially in APC (44%) and hMLH1 (0%) genes (Ali et al., 2011). Previous study showed that the frequency of methylation in p16 gene promoter in chromate lung cancer is slightly higher than the non-chromate lung cancer (33% vs 26%) (Kondo et al., 2006). Ali et al. found that the mean methylation index (a parameter reflecting the overall methylation status) in chromate lung cancers is significantly higher than non-chromate lung cancers (0.41 vs 0.21, P=0.001) (Ali et al., 2011). Moreover, the methylation index of multiple genes (particularly hMLH1, p16, and APC genes) in chromate lung cancer patients who had more than 15years of chromate exposure was significantly higher than chromate lung cancer patients who has less than 15years of chromate exposure (0.42 vs 0.19) (Ali et al., 2011). It was further determined that the reduced expression or loss of expression of p16 and hMLH1 proteins are significantly correlated with p16 and hMLH1 methylation (Ali et al., 2011). In addition, a recent study also reported that chromate workers in China display increased peripheral blood cell DNA methylation in the promoters of several DNA repair genes including MGMT, human 8-oxoguanine DNA glycosylase 1 (HOGG1), and RAD51 recombinase (RAD51) (Hu et al., 2018). It was also determined there is a significant negative association between DNA methylation levels in these DNA repair genes and their corresponding mRNA levels in Cr(VI)-treated16HBE cells, providing addition evidence supporting that Cr(VI) exposure could down-regulate DNA repair gene expressions by increasing their DNA methylation (Hu et al., 2018).

Recently, Dr. Kazuya Kondo’s group further quantitatively determined the methylation status of the hMLH1 gene promoter region using bisulfite pyrosequencing and paired tumorous and non-tumorous sample sets from chromate-exposed and non-chromate-exposed human lung tissues (Tsuboi et al., 2020). It was found that the methylation levels of hMLH1 promoter region are significantly higher in chromate lung cancer tissues than non-chromate lung cancer tissues. It was also found that the methylation levels of hMLH1 promoter in normal lung tissues in chromate workers are higher than non-chromate-exposed human normal lung tissues. Moreover, lung cancer tissues with reduced protein levels of hMLH1 had significantly higher methylation levels of hMLH1 promoter than lung cancer tissues with normal MLH1 protein levels (Tsuboi et al., 2020). In addition, a more recent study showed that the methylation levels in the promoter region of hedgehog-interacting protein (HHIP), a negative regulator of Hedgehog signaling, are significantly increased in Cr(VI)-transformed human bronchial epithelial cells (Li et al., 2021). By using various pharmacological and genetic approaches, the authors determined that increased DNA methylation and downregulation of hedgehog-interacting protein may play an important role in hexavalent chromium-induced malignant transformation of human bronchial epithelial cells (Li et al., 2021). Collectively, the findings from these cell culture and human lung tissue sample studies provided substantial evidence indicating that chromate exposure increases DNA methylation in several tumor suppressor and DNA repair gene promoters and reduces their expression. Given the importance of these tumor suppressor and DAN repair genes in regulating cell proliferation and genome stability, these findings imply that chromate-caused DNA methylation could play an important role in its carcinogenicity.

In contrast to studies discussed above, a recent study reported that chronic Cr(VI) exposure also causes hypomethylation in a gene promoter region. Ge et al. found that the CpG islands in the promoter region of C-X-C Motif Chemokine Ligand 5 (CXCL5) gene is hypomethylated, which contributes partially to the up-regulation of CXCL5 expression levels in Cr(VI)-transformed human bronchial epithelial cells (Ge et al., 2022).

In addition to the studies discussed above showing Cr(VI) exposure-caused gene-specific promoter methylation, several other studies reported that Cr(VI) exposure also changes genome wide global DNA methylation status as discussed in other reviews (Guo, Feng, Lemos, & Lou, 2020; Martin & Fry, 2018). In particular, global DNA hypomethylation of peripheral blood cells was observed in Cr(VI)-exposed workers and experimental animals (Wang et al., 2012, 2016). The significance of Cr(VI) exposure-caused global DNA hypomethylation in Cr(VI) carcinogenesis is currently unknown. Obviously, further studies on the effect of Cr(VI) exposure on DNA methylation, the underlying mechanism and its role in Cr(VI) carcinogenesis are needed.

3.2. Effects of hexavalent chromium exposure on histone posttranslational modifications

The second epigenetic effect of Cr(VI) was demonstrated by its capacity of causing nuclear core histone protein posttranslational modifications (PTMs). Schnekenburger et al. first reported that Cr(VI) (K₂CrO_4, 50μM) treatment in mouse hepatoma Hepa-1c1c7 (Hepa-1) cells cross-links histone deacetylase 1-DNA methyltransferase 1 (HDAC1-DNMT1) complexes to cytochrome P450 family 1 subfamily A member 1 (CYP1a1) promoter altering various histone PTMs induced by AHR-mediated gene transactivation resulting from benzo(a)pyrene (BaP) treatment, including phosphorylation of H3 Ser-10, tri-methylation of H3 Lys-4 (H3K4me3), and various acetylation marks in histones H3 and H4 (Schnekenburger, Talaska, & Puga, 2007). These changes maintain a state of histone deacetylation and inhibit CYP1a1 transcription. Inhibiting HDAC1, DNMT1 or knocking down HDAC1 or DNMT1 reversed Cr(VI)-induced transcriptional repression of CYP1a1 by creating a state of histone acetylation through decreasing the interaction of HDAC1-DNMT1 with the Cyp1a1 promoter (Schnekenburger et al., 2007). Interestingly, Cr(VI) treatment significantly increased the formation of BaP DNA adducts although it reduced the expression of CYP1a1, the enzyme that metabolically activates BaP. These findings suggest that Cr(VI)-caused histone PTMs changes may enhance the carcinogenicity of BaP by increasing the formation of genotoxic BaP DNA adducts.

Subsequent studies from Dr. Max Costa’s group showed that Cr(VI) treatment (K₂CrO_4, 10μM, 24h) significantly increases H3K4 tri-methylation (H3K4me3) in human lung adenocarcinoma cells (A549) (Zhou, Li, Arita, Sun, & Costa, 2009). Similarly, a lower Cr(VI) concentration and a longer treatment time (K₂CrO_4, 0.5, 1.0μM, 7days) also increased H3K4me3 in A549 cells (Zhou et al., 2009). Immunofluorescence (IF) staining revealed that short time Cr(VI) treatment (K₂CrO_4, 5μM, 24h) also increases H3K9 di-methylation (H3K9me2) globally in A549 cells; however, the tri-methyl H3K4 (H3K4me3) and di-methyl H3K9 (H3K9me2) marks localized in different regions in the nucleus of the cell (Zhou et al., 2009). It is known that the presence of the H3K4me3 mark in the promoter region of genes is associated with gene activation and the presence of H3K9me2 mark in the promoter region of genes is generally associated with gene silencing. These findings suggest that Cr(VI) exposure could simultaneously cause active and repressive histone PTMs, which may have different impacts on different regions of chromatin and gene expression as well. Another study from this group showed that short term Cr(VI) treatment increases the enrichment of the H3K9me2 mark in hMLH1 gene promoter region in A549 cells, which was correlated with decreased hMLH1 mRNA expression (Sun, Zhou, Chen, Li, & Costa, 2009). Mechanistic studies showed that Cr(VI) treatment increases the expression of G9a, a histone methyltransferase that specifically methylates H3K9. It was further determined that supplementation with ascorbate, the primary reductant of Cr(VI) and also an essential cofactor for the histone demethylase activity, partially reduced Cr(VI) treatment-induced H3K9me2 level (Sun et al., 2009).

Additional studies showed that Cr(VI) exposure also changes nuclear core histone protein acetylation. Chen et al. found that Cr(VI) exposure increases the levels of H3K9 and H3K14 acetylation but reduces the level of H4K16 acetylation (Chen et al., 2016). It was further determined that Cr(VI) exposure reduces H4K16 acetylation level by nuclear protein 1 (Nupr1)-mediated down-regulation of the histone acetyltransferase MOF (male absent on the first) (Chen et al., 2016). Xia et al. reported that Cr(VI) down-regulates the expression of biotinidase in human bronchial epithelial cells by reducing histone acetylation (Xia et al., 2011). Ren et al. found that short-term and long-term Cr(VI) exposure reduce H3 acetylation (H3K18ac and H3K27ac) leading to reduced expression of tumor protein p53 binding protein 1 (53BP1) in human bronchial epithelial cells (Ren et al., 2019). A follow up study from this group showed that long term Cr(VI) exposure reduces H3K18ac and H3K27ac levels and 53BP1 expression probably by increasing the expression of a multifunctional protein SET (patient SE translocation) (Chen et al., 2021). In contrast, our recent study showed that several histone acetylation mark levels (H3K9ac, H3K27ac, H4Ac, H2Bac) are increased in chronic low dose Cr(VI) exposure-transformed human bronchial epithelial BEAS-2B cells (Clementino et al., 2020). It was determined that chronic Cr(VI) exposure up-regulates ATP citrate lyase (ACLY) expression to increase acetyl-CoA levels promoting histone acetylation. Moreover, the H3ac mark was enriched in the promoter of the protooncogene c-Myc and inhibition of ACLY reduced H3ac level and c-Myc expression (Clementino et al., 2020). Similarly, a recent study also showed increased levels of histone acetylation marks (ace-H3, H3K9ac, H3K14ac, H3K18ac and H3K27ac) in the promoter region of CXCL5 to increase CXCL5 expression in Cr(VI)-transformed BEAS-22B cells (Ge et al., 2022).

While representative studies discussed above clearly demonstrated that short term and long term Cr(VI) exposure is capable of causing nuclear core histone protein PTM dysregulations, the significance of Cr(VI) exposure-caused histone PTM dysregulations in Cr(VI) carcinogenesis remains largely unknown. Our recent study showed that chronic exposure of two immortalized human bronchial epithelial cells (BEAS-2B and 16HBE) to a low dose of Cr(VI) (K₂Cr₂O_7, 0.25μM) for 20 and 40weeks caused cell malignant transformation (Wang et al., 2018). Further characterization of Cr(VI)-transformed BEAS-2B and 16HBE cells revealed that the Cr(VI)-transformed cells display increased levels of histone H3 repressive methylation marks (H3K9me2 and H3K27me3) and their related histone-lysing methyltransferases (HMTases) G9a, SUV39H1 (suppressor of variegation 3–9 homolog 1) and EZH2 (enhancer of zeste homolog 2). It was determined that pharmacological inhibition or knockdown of these HMTases reduces H3 repressive methylation marks and malignant phenotypes of Cr(VI)-transformed cells. Moreover, stable knockdown of these HMTases in parental non-transformed cells significantly reduced the capability of chronic Cr(VI) exposure to induce CSC-like property and cell transformation. Further mechanistic study showed that knockdown of HMTases reduce Cr(VI) exposure-caused DNA damage, suggesting an interaction between Cr(VI) exposure-caused epigenetic and genetic effects (Wang et al., 2018). Furthermore, the increased levels of HMTases SUV39H1 and EZH2 were also detected in chromate exposure-caused human lung cancer tissues (Tsuboi et al., 2020; Wang et al., 2018). Collectively, these findings indicate that chronic Cr(VI) exposure increases H3 repressive methylation marks by increasing the related HMTases expression, which plays an important role in Cr(VI)-induced CSC-like property and cell malignant transformation.

3.3. Effects of hexavalent chromium exposure on microRNA expressions

The effect of Cr(VI) exposure on microRNA (miRNA) expression and its role in Cr(VI)-induced cell transformation have also been explored (Humphries et al., 2016; Wang, Liu, & Jiang, 2021). He et al. first reported that the expression level of miR-143 in Cr(VI)-transformed human bronchial epithelial cells is reduced (He et al., 2013). The authors found that the conditioned medium prepared from Cr(VI)-transformed BEAS-2B cells increases the tube formation by human umbilical vein endothelial cells (HUVECs), suggesting Cr(VI)-transformed BEAS-2B cells is capable of stimulating angiogenesis. Interestingly, ectopic expression of miR-143 in Cr(VI)-transformed BEAS-2B cells reduced their angiogenic stimulating effect (He et al., 2013). Moreover, overexpressing miR-143 in Cr(VI)-transformed BEAs-2B cells also reduced their tumor growth and tumor angiogenesis in nude mice. Mechanistic studies revealed that miR-143 down-regulation promotes Cr(VI)-induced cell transformation and tumor angiogenesis likely through up-regulating insulin-like growth factor-1 receptor (IGF-I-R) and insulin receptor substrate-1 (IRS1) expression, leading to the activation of the extracellular regulated MAP kinase (ERK)/hypoxia-induced factor 1α (HIF1α)/NF-κB signaling pathway (He et al., 2013). A follow up study from this group found that the expression level of miR-143 is decreased in blood samples of Cr(VI)-exposed workers, compared with corresponding un-exposed workers (Wang, Qiu, et al., 2019). Another recent study from this group showed that chronic Cr(VI) exposure down-regulates the expression of miR-27a/b leading to the up-regulation of NF-E2-related factor 2 (Nrf2) to promote Cr(VI)-induced cell transformation and tumorigenesis (Wang, Bayanbold, et al., 2022).

Our recent study showed that the expression level of miR-494 in Cr(VI)-transformed human bronchial epithelial cells is significantly lower than the passage-matched control cells (Wang, Lin, et al., 2019). Stably overexpressing miR-494 in Cr(VI)-transformed BEAS-2B cells greatly decreased their transformed phenotypes as evidenced by their reduced CSC-like property and tumorigenicity. Mechanistic studies revealed that down-regulation of miR-494 by chronic Cr(VI) exposure leads to the up-regulation of the protooncogene c-Myc expression. It was also observed that the c-Myc protein level is significantly higher in chromate exposure-caused human lung cancer tissue than the adjacent normal lung tissue. Moreover, stable knockdown of c-Myc expression in Cr(VI)-transformed BEAS-2B cells significantly reduced their CSC-like property and tumorigenicity. Furthermore, stably overexpressing c-Myc in chronic Cr(VI)-exposed miR-494 stable overexpression cells reversed the inhibitory effect of miR-494 on Cr(VI)-induced CSC-like property, cell transformation and tumorigenesis (Wang, Lin, et al., 2019). These findings suggest that down-regulation of miR-494 and up-regulation of c-Myc may play important roles in Cr(VI) carcinogenesis.

In a recent study from Dr. John P. Wise’s group, Speer et al. reported that acute (24h) or prolonged (72 and 120h) exposure to 0.1, 0.2 and 0.3μg/cm² particulate zinc chromate alters global miRNA expressions in an immortalized, non-cancerous human lung cell line (WTHBF-6) (Speer et al., 2022). It was determined that the zinc chromate exposure exhibits concentration- and time-dependent effects on global miRNA expression. Further bioinformatics analysis showed that zinc chromate exposure-altered miRNAs regulate important signaling pathways involved in chromosome instability, inflammation, cell transformation, oxidative stress and escape from cell death. The authors proposed to further investigate how specific miRNAs altered by zinc chromate exposure are involved in the pathways of carcinogenesis and contributing to Cr(VI) carcinogenesis (Speer et al., 2022).

There were some controversial reports on the effect of Cr(VI) exposure on miR-21 expression and functions. Pratheeshkumar et al. reported that reactive oxygen species (ROS)-dependent up-regulation of miR-21 and down-regulation of its target gene programmed cell death 4 (PDCD4) play an important role in Cr(VI)-induced transformation of human bronchial epithelial cells (Pratheeshkumar, Son, Divya, Turcios, et al., 2016; Pratheeshkumar, Son, Divya, Wang, et al., 2016). In contrast, Zhang et al. recently reported that acute Cr(VI) exposure-generated ROS leads to down-regulation of miR-21-5p and up-regulation of its target gene PDCD4 in human L02 hepatocytes, which mediates Cr(VI)-induced apoptosis and cell proliferation inhibition (Zhang et al., 2020). These findings indicate that acute or chronic Cr(VI) exposure could have different impacts on the same miRNA expression and functions.

While the representative studies discussed above clearly demonstrated that short-term or long-term Cr(VI) exposure dysregulates miRNA expressions, the mechanisms of how Cr(VI) exposure alters miRNA expression remain largely unknown. In addition, some studies also showed that some miRNA expression levels are changed in peripheral blood samples of workers exposed to chromate ( Jia et al., 2020; Li et al., 2014; Wang, Qiu, et al., 2019). However, the significance of Cr(VI) exposure-caused human blood miRNA changes in Cr(VI) carcinogenesis has not been determined. Given the minimal invasive nature of collecting blood samples, further studies are needed to determine whether abnormally-expressed blood miRNAs in chromate workers could be a useful marker for early diagnosis of Cr(VI)-induced cancer, especially lung cancer.

3.4. Effects of hexavalent chromium exposure on long non-coding RNA expressions

The regulation of gene expression by long non-coding RNAs (lncRNAs) is a relatively newly-established epigenetic mechanism that regulates gene expression at multiple levels. As a result, the effect of Cr(VI) exposure on lncRNA expression and its role in Cr(VI) carcinogenesis remain largely unexplored (Wang, Wang, & Yang, 2021). Hu et al. reported that a 24h exposure of human bronchial epithelial cells (16HBE) to 10μM of Cr(VI) (K₂Cr₂O₇) causes DNA damage and reduces cell viability (Hu et al., 2019). It was found by using a lncRNA microarray analysis that this acute Cr(VI) exposure alters a large number of lncRNA expressions with 1868 lncRNAs being up-regulated and 2203 lncRNAs being down-regulated (Hu et al., 2019). The bioinformatics analysis suggested that the differentially expressing lncRNAs resulting from Cr(VI) exposure are involved in pathways regulating immune responses, cell cycle, and DNA damage repairs (Hu et al., 2019). It is likely that differentially-expressed lncRNAs may be involved in acute Cr(VI) exposure-caused DNA damage.

By using the lncRNA microarray profiling, we recently screened differentially expressed lncRNAs in chronic low dose Cr(VI) (0.25μM of K₂Cr₂O₇, 20weeks) exposure-transformed human bronchial epithelial BEAS-2B cells. It was found that the expression levels of a large number of lncRNAs are changed in Cr(VI)-transformed BEAS-2B cells. We are now in the process of determining the role of lncRNA dysregulation in chronic Cr(VI) exposure-induced cell transformation, CSC-like property and tumorigenesis.

4. Epitranscriptomic mechanisms of hexavalent chromium carcinogenesis

4.1. Effects of hexavalent chromium exposure on RNA m⁶A modification

With the discovery of the m⁶A two erasers (demethylases) FTO and ALKBH5 (Jia et al., 2011; Zheng et al., 2013), the role of the RNA m⁶A modification dysregulation in cancer has been actively investigated during the past decade (Uddin et al., 2021a). However, the role of the RNA m⁶A modification dysregulation in environmental carcinogenesis especially in metal carcinogenesis has been rarely studied (Yang, 2020). Lv et al. recently reported that exposure of mouse spermatogonial stem cells (SSCs) to a 10μM of Cr(VI) (Na₂CrO₄) for 1h reduces the total RNA m⁶A levels determined by using the m⁶A Dot-Blot Assay (Lv et al., 2021). By using the m⁶A-IP-qPCR analysis, it was found that 4h of 10μM Cr (VI) treatment decreases the m⁶A levels in mitochondrial fusion genes mitofusin 2 (Mfn2) and OPA1 mitochondrial dynamin like GTPase (Opa1) and in mitophagy genes BCL2 interacting protein 3 (Bnip3) and BCL2 interacting protein 3 like (Nix). Interestingly, the authors found that pretreatment with 50μM melatonin for 2h reverses subsequent 4h Cr(VI) exposure-caused decreases of m⁶A levels in mitochondrial fusion genes Mfn2 and Opa1 and in mitophagy genes Bnip3 and Nix, which is correlated with the effect of melatonin pretreatment attenuating Cr(VI)-caused mitochondrial disorders in mouse SSCs (Lv et al., 2021). It was further determined that acute Cr(VI) exposure reduces the RNA m⁶A levels likely through down-regulating the m⁶A writer METTL3 expression levels. Moreover, knockdown of METTL3 in mouse SSCs impaired the reversal effect of melatonin pretreatment on Cr(VI)-caused down-regulation of the m⁶A levels in mitochondrial genes, and diminished the protective effect of melatonin pretreatment on Cr(VI)-caused mitochondrial damages. The findings of this study showed that acute Cr(VI) exposure is capable of reducing RNA m⁶A modification levels.

Our recent study investigated the effect of chronic low dose Cr(VI) exposure (0.25μM of K₂Cr₂O₇, 20–40weeks) on total RNA m⁶A levels and its role in Cr(VI)-induced cell transformation, CSC-like property and tumorigenesis (Wang, Uddin, et al., 2022). It was found that total RNA m⁶A levels in Cr(VI)-transformed human bronchial epithelial cells (BEAS-2B and 16HBE) are significantly higher than that in the passage-matched control cells as determined by using the m⁶A microarray and the m⁶A RNA modification ELISA-like colorimetric assay. Moreover, the total RNA m⁶A levels in the lungs of mice exposed to Cr(VI) (calcium chromate) for 26weeks are also significantly higher than that in the PBS-exposed control mouse lungs. Mechanistic studies showed that chronic Cr(VI) exposure increases the m⁶A levels by up-regulating the m⁶A writer METTL3 expression. Indeed, increased METTL3 protein levels were also observed in Cr(VI)-transformed cells, Cr(VI) exposure-caused mouse and human lung tumor tissues. Functional studies revealed that knockdown of METTL3 in Cr(VI)-transformed human bronchial epithelial cells (BEAS-2B) significantly decreased their malignant phenotypes as evidenced by their reduced capabilities in forming soft agar clones, suspension spheres and xenograft tumors in nude mice. Furthermore, stably knocking down the expression of METTL3 in parental non-transformed BEAS-2B cells significantly reduced the capability of chronic low dose Cr(VI) exposure to induce cell transformation and CSC-like property (Wang, Uddin, et al., 2022). Collectively, these findings show that long term exposure to a low dose of Cr(VI) is capable of increasing RNA m⁶A modification levels, which could play important roles in Cr(VI) carcinogenesis.

4.2. Effects of hexavalent chromium exposure on RNA other modifications

By using the LC-ESI-MS/MS analytical method, Chen et al. investigated the effects of Cr(VI) exposure (1 or 5μM of K2CrO4, 24h) on 14 kinds of modifications in mRNA of HEK293T cells (Chen, Xiong, Ding, Yuan, & Feng, 2019). It was found that the modification of inosine is the only one significantly decreased in 1 and 5μM of Cr(VI)-exposed HEK293T cells, indicating that the Cr(VI) may affect the A-to-I edition in mRNAs. It was further determined that the decline of the level of inosine in mRNAs is due to the reduced expression of the adenosine deaminase acting on RNA (ADAR1) resulting from Cr(VI) exposure.

5. Conclusion

Although Cr(VI) is generally considered as a genotoxic carcinogen, acute and chronic Cr(VI) exposures also cause various epigenetic dysregulations. A substantial number of studies demonstrate that Cr(VI) exposure is capable of causing DNA hyper- and hypo-methylation, changing nuclear core histone protein various PTMs, altering miRNA and lncRNA expressions, and dysregulating cellular epitranscriptome. Moreover, current evidence supports the notion that Cr(VI) exposure-caused epigenetic and epitranscriptomic dysregulations could play important roles in Cr(VI) carcinogenesis (Fig. 1).

Fig. 1.

A schematic description for the proposed epigenetic and epitranscriptomic mechanisms of hexavalent chromium Cr(VI) carcinogenesis. The solid black lines refer to the reported mechanisms discussed in the text. The yellow description and yellow dot lines refer to the speculated mechanisms. The genotoxic mechanism of Cr(VI) carcinogenesis is not discussed and not included here. lncRNAs: long non-coding RNAs; PTM: histone posttranslational modification.

Further studies are needed to better define the role of Cr(VI) exposure-caused epigenetic and epitranscriptomic dysregulations in Cr(VI) carcinogenesis. First, the majority of evidence showing the epigenetic and epitranscriptomic effects of Cr(VI) exposure are obtained from in vitro cell culture studies. More animal and human studies are needed to demonstrate chronic low dose Cr(VI) exposure-caused epigenetic and epitranscriptomic effects and their roles in Cr(VI) carcinogenesis. Second, while studies showed that Cr(VI) exposure causes epigenetic and epitranscriptomic changes, the underlying mechanisms by which Cr(VI) causes these changes remain largely unknown. Further studies are needed to better understand the mechanism of how Cr(VI) exposure alters epigenetics and epitranscriptome, which will lead to identify new molecular targets for prevention and treatment of Cr(VI)-caused cancer. Third, further studies are needed to determine whether Cr(VI) exposure-dysregulated epigenetic and epitranscriptomic marks could serve as biomarkers for early diagnosis of Cr(VI)-caused cancer.

Abbreviations

3′UTR

3′untranslated region

ALKBH5

AlkB homolog 5

APC

adenomatous polyposis coli

BaP

benzo(a)pyrene

Cr(VI)

hexavalent chromium

CSC

cancer stem cell

CXCL5

C-X-C motif chemokine ligand 5

CYP1a1

cytochrome P450 family 1 subfamily A member 1

DNMT

DNA methyltransferase

EZH2

enhancer of zeste homolog 2

FTO

fat mass and obesity-associated protein

HDAC

histone deacetylase

hMLH1

human mutL homolog 1

HMTase

histone-lysing methyltransferase

lncRNA

long non-coding RNA

m⁶A

N⁶-methyladenosine

MBP

methyl CpG binding protein

METTL3

methyltransferase like-3

MGMT

O⁶-methylguanine-DNA methyltransferase

miRNA

microRNAs

PTM

posttranslational modification

ROS

reactive oxygen species

SUV39H1

suppressor of variegation 3–9 homolog 1

WTAP

Wilms’ tumor 1-associating protein

YTHDC

YTH domain-containing protein

YTHDF

YTH domain-containing family protein