Epigenetic Mediators and Consequences of Excessive Alcohol Consumption

Alcohol has pleiotropic effects across multiple organ systems, including brain, cardiovascular, endocrine, immune, musculoskeletal, and gastrointestinal systems. Moreover, some effects, such as intoxication, can be brief, but others, such as the development of Alcohol Use Disorders (AUDs), cardiovascular disease, and liver damage, can persist for a lifetime. Effects resulting in encoding of endocrine and other dysfunctions can also persist across generations. This complexity creates a barrier to the creation of therapeutics and discovery of biomarkers. However, we know that environmental factors, which can include drugs of abuse such as alcohol, can have short- and long-term effects on gene expression through epigenetic mechanisms (Holliday, 2006; Shukla et al., 2008). Epigenetic mechanisms affect the transcription and translation of many genes simultaneously. Therefore, by understanding the mechanics of these epigenetic changes, we will have the ability to craft powerful new therapeutics to offset negative effects of alcohol exposure.

Epigenetic modifications commonly occur through three mechanisms: methylation of DNA, histone post-translational modifications, and the interactions of non-coding RNA with transcriptional and translational cellular machinery (Fig. 1). Modifications that open the tightly wound chromatin structure are thought to increase gene expression, while modifications that condense chromatin structure are thought to inhibit gene expression. DNA methylation, which usually occurs at groupings of cytosine and guanine nucleotides referred to as “CpG islands”, represses gene transcription. This repression occurs with the binding of methyl-CpG binding domain proteins, the subsequent recruitment transcription inhibitory complexes, and chromatin condensation (Cedar & Bergman, 2012). Post-translational modifications of histones, by covalent addition of functional groups to histone tails, can also regulate chromatin access. These modifications include commonly studied acetylation and methylation, as well as many other modifications, including ubiquitinylation, phosphorylation, and ADP ribosylation, to name a few (Bannister & Kouzarides, 2011; Kouzarides, 2007; Strahl & Allis, 2000). Histone acetylation relaxes chromatin, which facilitates gene transcription. Histone methylation can bidirectionallly affect gene expression depending on the amino acid location on the histone tail and quantity of methylation, i.e., whether singly, di-, or tri-methylated (Zhou, Goren, & Bernstein, 2011). Non-coding RNAs (ncRNAs) can affect both gene transcription and translation. For example, long non-coding RNAs (IncRNAs, >200 nucleotides in length) can interact with transcription machinery to trigger chromatin compaction at commonly imprinted regions, including the H19 locus and X chromosome (Nagano & Fraser, 2011; Rinn & Guttman, 2014). MicroRNAs (miRNAs, ∼21 nucleotides in length) affect translation by binding the 3′-untranslated region (3′-UTR) of mRNA transcripts and, along with the miRNA-induced silencing complex, inhibit translation and enhance mRNA degradation (Fabian & Sonenberg, 2012). miRNAs can bind to many related transcripts and regulate the expression of gene networks as a whole (reviewed in Miranda, 2014).

Fig. 1. Commonly Researched Epigenetic Modifications.

Commonly researched epigenetic modifications include covalently bonded groups attached to DNA and histones, as well as interactions with non-coding RNAs (ncRNAs). DNA methylation, via methyl-CpG binding domain proteins (MBDs), recruits repressor complexes to inhibit transcription and to encourage chromatin condensation. Histone acetylation opens chromatin to allow access to transcription machinery, while histone methylation can either condense or open chromatin, depending on the localization and degree of methyl group attachment. ncRNAs can interact at the chromatin level (e.g., long non-coding RNAs that act during imprinting to silence an allele) or at the translation level (e.g., microRNAs that bind to the 3′-UTR of mRNAs), along with the RNA-lnduced Silencing Complex (RISC), to trigger mRNA degradation.

DNA methylation, histone modification, and ncRNAs interact with each other to coordinate and regulate overall gene expression. The aim of this special edition of Alcohol is not only to provide background on what we currently know about epigenetic modifications in response to alcohol exposure, but also to further examine the impact of these changes throughout the lifespan and across generations. Additionally, we highlight, by targeting epigenetic mediators, how we can create novel therapeutics for both AUDs in adult populations and for Fetal Alcohol Spectrum Disorders (FASD). These studies provide evidence of the importance of continued research on the epigenomic alterations that occur in response to alcohol exposure, to combat the complex changes that occur across the body and across generations.

To facilitate navigation of the articles in this special issue of Alcohol, we have assembled a table that summarizes the epigenetic modifications and the timing of the alcohol exposures that are addressed in each paper (see Table 1).

Table 1. Summary of papers included in this special issue of Alcohol.

Authors	Epigenetic modification				Exposure period			Longevity of Effects
	Histone	DNA	ncRNA	Δ Gene Expression	Adult	Adolescent	Preconception (✘) Prenatal (✓)	Exposure Period/Effects in	Cross-Generational
	Reviews
Palmisano and Pandey	✓	✓		✓	✓	✓		Adolesence/Adulthood
Tulisiak et al.		✓		✓	✓		✓
Mayfield			✓		✓
Chater-Diehl et al.	✓			✓			✓	Prenatal/Adulthood
Chastain and Sarkar	✓	✓	✓	✓	✓		✓		✓
Laufer et al.		✓		✓			✘ ✓	Preconception Prenatal/Adulthood	✓
Research – Adult Use
Restrepo et al.	✓			✓	✓
Hashimoto et al.	✓	✓		✓	✓
Ponomarev et al.	✓	✓		✓	✓
Cervera-Juanes et al.		✓		✓	✓	✓
Hitzemann et al.				✓	✓
Research – Prenatal and Preconception Exposure/Cross-generational Effects
Veazey et al.	✓			✓			✓
Öztürk et al.		✓					✓
Burrowes et al.		✓	✓				✓
Balaraman et al.			✓				✓	Developmental/Pre-pubertal
Rompala et al.				✓	✓		✘	Adult/F1 offspring	✓
Asimes et al.		✓		✓		✓	✘	Adult/F1 offspring	✓
Popoola et al.				✓			✘✓	Prenatal/F1, F2 offspring	✓

Alcohol-Use Disorders – Adulthood

In the adult, drinking behavior can become classified as an AUD when there is evidence of dependence or withdrawal (American Psychiatric Association, 2013). It is estimated that about 30% of adults will have an AUD during their lifetime (Grant et al., 2015). Previous research has shown the presence of epigenetic modifications in alcoholic patients, including: DNA hypermethylation at specific genetic loci in blood cells, which has been associated with alcohol craving and preference (Bonsch, Lenz, Kornhuber, & Bleich, 2005; Bonsch, Lenz, Reulbach, Kornhuber, & Bleich, 2004; Hillemacher et al., 2009; Muschler et al., 2010; reviewed in Tulisiak, Harris, & Ponomarev, this issue); tri-methylation at histone 3 lysine 4 (H3K4me3), which has been associated with increased gene expression in postmortem tissue (Ponomarev, Wang, Zhang, Harris, & Mayfield, 2012); and upregulation of miRNAs, which has been associated with the downregulation of mRNA transcripts in the prefrontal cortex of postmortem samples (Lewohl et al., 2011; Nunez & Mayfield, 2012; reviewed in Mayfield, this issue).

To treat and prevent AUDs, we need to understand how the differences seen in both postmortem human tissue as well as the peripheral tissues of patients may contribute to the development of heavy, addictive drinking. This can be further understood by examining how the epigenome changes across the three components of AUDs: intoxication, withdrawal, and craving (Koob & Volkow, 2010). In this special issue on epigenetics, Cervera-Juanes, Wilhelm, Park, Grant, & Ferguson (this issue) show that within the nucleus accumbens core, which is involved in neurocircuitry for intoxication and withdrawal, chronic low level drinking results in DNA hypomethylation, and high levels of drinking result in hypermethylation. In the prefrontal cortex, which is involved in neurocircuitry for craving, Hashimoto, Gavin, Wiren, Crabbe, and Guizzetti (this issue) show altered gene expression for proteins needed for histone and DNA epigenetic modification creation and maintenance, and they show that ethanol withdrawal has a larger effect on gene expression than ethanol exposure or abstinence. Interestingly, Hashimoto et al. show that covalent modifications may act in opposition, with altered transcript expression of epigenetic modifiers that lead to open and closed chromatin conformations.

The complexity of these epigenetic changes leads to the question – What is the outcome for gene expression? Previous research has shown that changes in gene expression are not readily predictable (Ponomarev et al., 2012; Veazey, Carnahan, Muller, Miranda, & Golding, 2013; Veazey, Parnell, Miranda, & Golding, 2015). In a model system of mice predisposed to binge drinking, Hitzemann et al. (this issue) compared gene expression in the dorsal striatum, which contributes to habitual intoxication, of animals who drank to a high blood ethanol content (BEC), compared to those who drank to a lower BEC. They found a larger number of genes were expressed with a positive correlation to BEC levels than genes with a negative correlation to BEC. This effect is unexpected if there is an increased rate of DNA methylation in the dorsal striatum, as has been shown in other brain regions. Yet, these data corroborate evidence of potential maladaptive hypomethylation, specifically of Grin2b, in the dorsal striatum, contributing to increased alcohol consumption (Wang et al., 2010; Wong, Tauck, Fong, & Kendig, 1998). Additionally, epigenetic markers have been shown to vary between brain regions (Finegersh et al., 2015) and can be further complicated by comorbid conditions, including stress (Meyer, Long, Fanselow, & Spigelman, 2013; Moonat & Pandey, 2012; reviewed in Palmisano & Pandey, this issue).

While understanding how epigenetic modifications affect central control of alcohol use is critical to combating AUDs, the identification of peripheral biomarkers is also an invaluable tool and could shed light on the progression of AUDs along with associated disorders, such as alcoholic liver diseases. Epigenetic markers can serve as peripheral biomarkers of alcohol exposure, and while methylation is commonly examined (Andersen, Dogan, Beach, & Philibert, 2015), microRNAs have been indicated as biomarkers for prenatal alcohol exposure (Balaraman et al., 2014, 2016), and histone modifications have been explored as biomarkers in other fields, including cancer biology (Chervona & Costa, 2012). Restrepo, Lim, Korthuis, and Shukla (this issue) show that histone acetylation at histone 3 lysine 9, and upregulation of the protein PNPLA3, may be involved in fatty liver disease development, which can predispose patients to alcoholic liver disease. Previous studies have also suggested that the histone post-translational modification of H3K9me2 may also serve as a biomarker in FASD (Veazey et al., 2015). Further investigation into these potential peripheral markers could assist in the identification and treatment of alcohol-linked pathophysiology associated with AUDs, including the progression of alcoholic liver disease.

Prenatal and Preconception Alcohol Exposure

Drinking behavior can impact multiple generations. Alcohol is a known teratogen. As such, prenatal exposure can lead to an array of perturbations to neurological and behavioral development, resulting in Fetal Alcohol Spectrum Disorders (FASD, Hoyme et al., 2005; Jones, 2011). Prenatal alcohol exposure has been shown to affect DNA methylation, histone modification, and miRNA components (Miranda, 2012; also reviewed in Laufer, Chater-Diehl, Kapalanga, and Singh, this issue and Chater-Diehl, Laufer, & Singh, this issue). Direct fetal exposure is not the only mechanism of cross-generational effects of alcohol exposure. Recent evidence has shown that paternal drinking during the preconception period affects behavior, development, and gene expression in the following generations (Finegersh & Homanics, 2014; Jabbar et al., 2016; Knezovich & Ramsay, 2012). This cross-generational transmission likely occurs through epigenetic modifications encoded in the germline (Bohacek & Mansuy, 2015; reviewed in Chastain & Sarkar, this issue).

Ethanol exposure during the prenatal period, while toxic for neural progenitors and more developed neurons, does not kill neural stem cells (Cheema, West, & Miranda, 2000; Prock & Miranda, 2007). Neural stem cells, in response to ethanol exposure, increase proliferation and undergo premature differentiation, thus depleting the neural stem cell pool (Camarillo & Miranda, 2008; Santillano et al., 2005). These effects on neural stem cells and developing cortical neurons likely contribute to thinning of the cerebral cortex, a hallmark of FASD (Zhou, Lebel, et al., 2011). Moreover, correct neuronal and cortical development is dependent on specific DNA methylation programs (Zhou, 2012). Öztürk, Resendiz, Öztürk, and Zhou (this issue) detail the alterations to cortical development and DNA methylation within the developing cortex after prenatal alcohol exposure. miRNAs, particularly miR-9, also regulate cortical development and neurogenesis (Coolen, Katz, & Bally-Cuif, 2013). miR-9 expression is decreased in neural stem cells after ethanol exposure, and miR-9 knockdown mimics the effects of ethanol on craniofacial development in a zebrafish model (Pappalardo-Carter et al., 2013; Sathyan, Golden, & Miranda, 2007). Burrowes et al. (this issue) provide evidence that developmental chromatin remodeling, due to the maturation of the BAF (Brg/brm-associated factors) complex, can be adaptive and protective in response to ethanol exposure and act to preserve miR-9 expression. Prenatal alcohol exposure can also affect histone post-translational modifications in a dose-dependent manner in embryonic stem cells (Veazey et al., this issue). Veazey and colleagues found no correlations between histone modifications and expression levels of their genes of interest. This disconnect between epigenetic modifications and gene expression was also seen in AUDs, and shows that more work needs to be done to map the relationship between epigenetic modifications and gene expression, particularly since, as seen with miR-9 in neural stem cells, some adaptions may be protective.

The effects of ethanol are not limited to exposure during in utero development, but can also occur from alcohol exposure prior to conception. Work on the proopiomelanocortin (POMC) gene has shown that not only is gene expression downregulated in rats which were prenatally exposed to alcohol, but that the effect of lowered POMC expression can be transmitted through the paternal line to the F2 and F3 generations (Govorko, Bekdash, Zhang, & Sarkar, 2012). Studies have found additional genes with altered expression across generations (Asimes et al., this issue; Popoola, Nizhnikov, & Cameron, this issue; Przybycien-Szymanska, Rao, Prins, & Pak, 2014), as well as altered alcohol drinking-associated behaviors (Finegersh & Homanics, 2014; Nizhnikov, Popoola, & Cameron, 2016), and methylation programming (Asimes et al., this issue; Finegersh & Homanics, 2014). The cross-generational effects on behavior can be dependent on genetic background, as Popoola et al. show using rat strains, or conserved across strains, as Rompala, Finegersh, Slater, and Homanics show in mouse models (both in this issue). Cross-generational inheritance may also vary by maternal and paternal lineage. Asimes et al. (this issue) show that DNA methylation patterns in alcohol-naïve offspring can widely vary by which parent (or if both parents) had been exposed to binge-like drinking. More research needs to be done to understand the effects of genetic background, to better understand the contribution of maternal lineage verses paternal lineage, and how inheritance of these changes in gene expression and behavior varies based on the sex of the offspring (Finegersh & Homanics, 2014; Vassoler, Byrnes, & Pierce, 2014).

Epigenetic Therapeutics

There is an important question that remains – Can alterations to the epigenome serve as sites for therapeutic intervention? Previous research has shown that inhibition of DNA methylation (Barbier et al., 2015; Ponomarev et al., this issue; Warnault, Darcq, Levine, Barak, & Ron, 2013) and histone deacetylation (Sanchis-Segura, Lopez-Atalaya, & Barco, 2009; Simon-O'Brien et al., 2015; Warnault et al., 2013) can decrease drinking behavior. Interestingly, Ponomarev et al. (this issue) show that the efficacy of histone deacetylation inhibition may be dependent on the drinking model examined, indicating that further study is needed to determine the most efficacious use of these inhibitors. During prenatal exposure to alcohol, DNMT inhibition mimics alcohol's negative effects on adult neural stem cell dynamics (Zhou, Balaraman, et al., 2011), and methyl supplementation through diet can ameliorate some effects of prenatal alcohol exposure (Downing et al., 2011). Choline, which can act as a methyl donor (Otero, Thomas, Saski, Xia, & Kelly, 2012), has neurological benefits within the hippocampus after third-trimester equivalent alcohol exposure (Monk, Leslie, & Thomas, 2012; Thomas, Garrison, & O'Neill, 2004). The benefits of choline may be due to a combination of actions, including the ability to restore and augment miRNA expression within the hippocampus after prenatal alcohol exposure (Balaraman, Idrus, Miranda, & Thomas, this issue).

Future Directions

We are just beginning to understand the effects of alcohol on the epigenome. The studies outlined here show how alcohol exposure can affect the epigenome, from AUDs, to prenatal alcohol exposure, to cross-generational effects. More work is needed to understand the interaction of all the methods of epigenetic modifications, including less commonly explored factors such as histone phosphorylation, arginine methylation (see Hashimoto et al., this issue), ubiquitination, sumoylation, citrullination, ADP-ribosolyation, as well as the dynamics of DNA methylation removal through 5-hydroxymethyl cysteine. Additionally, the interactions of long non-coding RNAs in normal neuronal development in disease are just beginning to be discovered and will provide an extra layer of complexity to gene expression regulation (Fatica & Bozzoni, 2014; Wapinski & Chang, 2011). With further investigation, it may be possible to develop novel therapeutics and discover new biomarkers to combat the negative effects of alcohol exposure throughout the lifespan.

Highlights.

• Epigenetic modifications change global gene expression in response to environmental factors.

• Alcohol consumption can affect the epigenome across the lifespan and generations.

• Further research into epigenetic changes may allow for the identification of biomarkers and the creation of novel therapeutics.