Noncoding RNA and Alcohol Use Disorder: A Scoping Review of Current Research and Knowledge Gaps

Screening and Selection Criteria

A total of 3,358 studies were imported in Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia. Available at www.covidence.org). After removing 1,392 duplicates identified by the software and an additional 29 manually identified duplicates, 1,937 studies remained for title and abstract screening. During this screening, any additional duplicates, review articles, or studies with only abstracts were manually removed. Studies that only addressed ncRNAs without direct association with, or manipulation of, alcohol exposure were excluded. All experimental species (e.g., mouse, rhesus monkey, human) and disease conditions (e.g., hepatitis, inflammation, cancer) were included. Two reviewers independently screened each title and abstract in Covidence. Any disagreements between two reviewers were resolved by a third reviewer, whose decision was final. After the title and abstract screening phase, 400 studies remained and were moved to the full-text review stage, which used a similar approach as the title/abstract screening. During the full-text review, studies that did not meet the inclusion criteria established during the title and abstract screening phase, as mentioned above, were excluded. Further, the selection focused on organ systems that are the best documented targets of alcohol consumption, including liver, cardiovascular system, brain, and fetal development; studies of tissues of the reproductive system were excluded due to human resource and time constraints. Also excluded were citations with only an abstract but no full text available, studies that had no association between ncRNAs and alcohol in any disease condition, those that only screened for ncRNAs in any disease condition without a direct link to alcohol, studies that had no PubMed IDs, and any studies that were solely simulations or modelling based on secondary analysis of existing data curated in databases. Although all types of ncRNAs—including miRNAs, circRNAs, lncRNAs, eRNAs, tRNAs, and rRNAs—were included in the literature search, the subsequent discussion will focus on miRNAs, lncRNAs, and circRNAs because either no studies, or only a single study, addressed the other types of ncRNA or because, as in the case of eRNAs, publications were not uniformly captured due to the unsettled nature of the ncRNA terminology in the literature. This comprehensive approach resulted in the inclusion of 338 studies for final analysis. A list of the references for all the included studies is available upon request from the corresponding authors.

The PRISMA flow diagram of the scoping review (see Figure 1) was created using Covidence following the PRISMA guidelines.^21,22 The objective for this scoping review was to address the current research and identify knowledge gaps in ncRNA in association with AUD.

Image

Figure 1. The PRISMA flow diagram of the study selection process for alcohol use and noncoding RNAs.

Data Extraction, Characterization, and Analyses

Details extracted from each full text screen included the author and year of the publication as well as the species (humans, mice, rats, zebrafish, or cell), sex (male, female, or both), and sample type (e.g., brain, liver, blood, saliva, embryo tissue, or cell culture) that were studied. Also charted were sample processing methods, such as RNA extraction method (e.g., Trizol, miRNeasy kit, mirVana miRNA isolation kit, RNeasy kit, guanidinium thiocyanate/phenol/chloroform method); method of ncRNA analysis (e.g., RNA sequencing, quantitative reverse transcription polymerase chain reaction [qRT-PCR], reverse transcription PCR [RT-PCR], miRNA 4.0 Arrays, miRCURY LNA Microarray kit); and biomarker analysis, including reported “area under the curve” (AUC) value from “Receiver Operator Characteristic” (ROC) analysis, when available.

All characteristics were extracted using Microsoft Excel 2024. Studies were grouped by the types of ncRNAs (i.e., lncRNA, circRNA, miRNA, eRNA, rRNA). Appendices 1 to 5 summarize the characteristics listed above along with the key findings of each study. For lncRNAs and circRNAs, key findings from all relevant studies are presented in one table. However, because of the extensive literature, studies were categorized by specific disease conditions for miRNAs. The number of the studies in the extracted Excel data sheet and in Covidence were counted and matched, and studies that were missed while extracting the data were noted. Any identified duplicates and articles without PubMed IDs were manually removed from the study.

The NIH iCite bibliometrics analytic platform was used to assess the impact and relevance of the included studies.²³ Together with the data tables, this provided a comprehensive overview of the current state of research on ncRNAs in AUD and related pathologies, uncovered gaps in knowledge, and suggested directions for future studies. Sunburst figures that visually represent gaps in knowledge for the different types of ncRNA are also available from the authors.

Results

Selection of Sources of Evidence

The systematic search of the Embase, Medline, and Academic Search Ultimate databases yielded a total of 3,358 articles (see Figure 1). Specifically, the Embase search resulted in 1,440 articles, the Medline search resulted in 1,397 articles, and the Academic Search Ultimate search resulted in 521 articles. After removing duplicates, 1,937 unique articles remained for screening. Initial screening based on title and abstract exclusion criteria resulted in the exclusion of 1,537 studies. Subsequently, 400 full-text articles were assessed for eligibility. Following a detailed assessment against predefined inclusion and exclusion criteria, 338 studies were included in the study for qualitative analysis, and 3,020 initially captured studies were excluded.

Study Characteristics

The 338 studies selected for qualitative analysis included a broad range of topics relevant to the research question. These studies were published between 2007 and December 2023 (the end of the assessment period) and included experimental as well as observational studies across a range of AUD-relevant disease states. The data in these studies were obtained with multiple species, sample types, and extraction and assessment methodologies. The characteristics of all included articles are summarized in Table 3.

Based on the studies collected in this review, research on the role of ncRNA in the pathophysiology of alcohol use and misuse began to emerge in 2007. However, significant and vigorous investigations into this area only appeared in the literature around 2012, with more than 10 studies reported in that year alone. Based on the iCite analysis, studies in human populations are well represented in the literature, along with studies in animal models. The number of publications reported per year has been steadily growing, reflecting increased interest in this field of ncRNAs and AUD research. The impact of these publications has also been high. Approximately 21% of the publications (72 articles) had a relative citation ratio (RCR, a field- and time-normalized citation rate, benchmarked to a score of 1.0 for the typical NIH-funded paper in that field²³) of more than 2.0, with a maximum score of 14.9. This indicates that these publications were cited between two and more than 14 times more than other NIH-funded publications in the field. The median RCR for these 72 articles was 3.23. According to the iCite analysis, the selected papers collectively had an “approximate potential to translate” (APT) score of about 29%, and about 14% of the papers were cited at least once in a clinical paper, mainly in the field of alcohol-related liver disease or neurobiological basis of AUD, suggesting an emerging “bench-to-bedside” translational outcome.

Among the ncRNAs, miRNAs were most studied (86%) followed by lncRNAs (11%) and circRNAs (3.0%). Other types of ncRNAs each constitute a minimal percentage of studies and are therefore not discussed in this review. Regarding the sex distribution among the study subjects, males (42%) were more frequently studied than females (8%). A considerable proportion (31%) of studies involved both sexes, whereas a notable percentage (19%) of studies did not report sex-specific data. This highlights a gap in reporting standards across studies and emphasizes the importance of considering sex as a critical variable. In terms of model organisms and study subjects, mice (34%) were studied more often than rats (22%), humans (30%), or cells (7%). A few studies (about 6%) were conducted in other organisms, including invertebrates (drosophila), teleosts (zebrafish), other mammals (sheep), and non-human primates (macaques). Model diversity is important, since similarity of outcomes in diverse model organisms increases the translational relevance of these studies. Moreover, 36% of the studies used some combination of mice, rats, cells, and human tissues, which further emphasizes the cross-species translational potential of these studies. Finally, a wide range of samples were used for the analysis of ncRNAs, including blood/plasma/serum (14%), brain (27%), and liver/liver cells (20%). Other sample types included cells, embryos, and various tissues such as heart, lung, and intestinal epithelial cells.

Methods of RNA extraction in the individual studies were also charted. The Trizol reagent was the most commonly used method (32%), followed by the miRNeasy kit (17%), and mirVana isolation kit (13%). The traditional Trizol (guanidium isothiocyanate-phenol) reagent method is historically well-validated and facilitates the separation of RNA, DNA, and protein from a single sample by phase separation methodologies; however, it may result in loss of small RNAs. In contrast, other methodologies that couple an organic extraction step common to Trizol with RNA recovery on silica/glass fiber substrates (miRNeasy and mirVana) result in better retention of small RNAs. Several other extraction methods were used less frequently, and 9% of the studies did not report their RNA extraction method. Overall, a diverse range of approaches were used in ncRNA analysis. A majority of studies used qRT-PCR assays (51%) and related variants; RT-PCR (12%); or a combination of techniques (13%) that reflects the usage of an integrative approach for more robust analyses. The remaining studies used a variety of analytic methods. However, more than 1% of the studies lacked documentation of techniques used to analyze ncRNA. This is concerning, because methodological details are essential for documenting the integrity and trustworthiness of research outcomes.

Analysis of ncRNA Classes

In the context of AUD, ncRNAs have gained attention for their ability to influence key pathways involved in brain function, alcohol metabolism, and immune responses. In the last 15 years, a substantial body of research has accumulated, demonstrating that ncRNAs are indeed sensitive to ethanol, and ethanol treatment induces changes on their transcript levels in vivo and in vitro, contributing to phenotypical variations that underlie the emergence of AUDs and their associated pathologies. Most of this research has focused on miRNAs, lncRNAs, and circRNAs; very little to no research has been conducted on other small ncRNAs, which will therefore not be discussed here.

miRNAs

The first miRNA, lineage-4 (lin-4), was discovered in the 1990s as a critical regulator of nematode development.^15-17 The discovery that all cells and tissues had the capacity for RNA interference (RNAi)—that is, inhibition of translation with small RNAs²⁴—and the fact that the miRNA Let-7/Lethal-7^25,26 was ubiquitously expressed in all eukaryotes were key drivers for research into miRNA biology. The next stage of discovery of miRNA research was the almost simultaneous discovery of machinery for miRNA processing. This includes the nuclear RNAse Drosha^27,28 along with the double-stranded RNA-binding protein Di-George Syndrome Critical Region-8 (DGCR8),^27,29 which processes kilobase-length primary miRNAs (pri-miRNA) into an about 70 nucleotide hairpin-structured pre-miRNA; Exportin-V, which transfers the pre-miRNA to the cytoplasm;³⁰ DICER, which processes the pre-miRNAs in the cytoplasm into mature miRNAs; ^31,32 and the argonaute (Ago) family^33,34 that act as the key chaperones of the RNA-induced silencing complex (RISC). Following DICER cleavage, which yields a short double-strand RNA molecule, one strand, termed the “sense” or “guide” strand (usually, but not always, the 5’ strand) is loaded onto Ago proteins in the RISC complex, while the antisense strand, or “passenger miRNA” (usually, but not always, the 3’ strand) is generally assumed to be degraded. Finally, rules for the predictable and orderly behavior of miRNAs in silencing gene expression—including the importance of 5’ seed sequence homology (six to eight nucleotides at the 5’ end of a miRNA that exhibit reverse complementarity to a target region within the 3’UTRs of mRNAs), a mid-region of non-homology, and 3’ sequence pairing—are important determinants of mRNA stability, translation, and other processes.^35-38 miRNAs also can be post-transcriptionally processed from lncRNAs. An example is miR-9, an evolutionarily ancient miRNA that was expressed in the first bilaterally symmetrical organisms and is important for brain development; in mammals, it is processed from one of three separate and unique lncRNAs (pri-miR-9-1, pri-miR-9-2, and pri-miR-9-3).³⁹ In contrast, other miRNAs, such as miR-335, are encoded in the intron of protein-coding genes (sometimes termed miRtrons⁴⁰) and have evolved more recently.

The first study implicating miRNAs in ethanol pathology was published in 2007.⁴¹ This scoping review identified an additional 289 papers published since then, comprising almost every aspect of AUD pathology, from brain circuits, organ toxicity, and cancer to pregnancy effects and developmental outcomes. This field of ncRNA biology has been the most thoroughly explored, comprising about 86% of all publications. Nevertheless, even here, significant gaps in our knowledge remain. The following sections summarize information and knowledge gaps regarding the role of miRNA in various conditions related to AUD.

Cancer

Due to their accessibility for assessment in body fluids and high biological stability, miRNAs have gained significant attention as important regulatory molecules and as biomarkers for tumor development, progression, and response to treatment in cancer biology.⁴² A review of studies investigating the role of miRNAs in different types of cancer associated with alcohol use showed that most of the studies were conducted in human samples, although a few were conducted in mice or cell cultures (see Appendix 1). Key findings are that specific miRNAs are indeed differentially up- or down-regulated in response to alcohol exposure in a variety of cancers, such as hepatocellular carcinoma, head and neck squamous cell carcinoma, and esophageal squamous cell carcinoma. Differential miRNA expression in cancers associated with alcohol use holds a promise that miRNAs can be used as biomarkers for early detection and diagnosis of alcohol-related cancers. However, vigorous further validation of these miRNAs as potential biomarkers is needed. Furthermore, miRNAs play a key regulatory role at both transcriptional and post-translational levels (see below), and these roles are not yet fully understood in cancer biology. Moreover, the mechanisms that control miRNA expression in cancer warrant further investigation. For instance, in oral squamous cell carcinomas, shorter survival was associated with miR-34b/c through promoter methylation,⁴³ highlighting the importance of studying epigenetic modifications and gene expressions in cancer associated with alcohol.

Musculoskeletal health

Several studies demonstrated that expression levels of specific miRNAs had an impact on adults with musculoskeletal conditions such as osteopenia, osteonecrosis, muscle atrophy and dysfunction, as well as fractures and healing (see Appendix 2A). These studies reported that alcohol exposure led to a decrease in miRNA levels such as miR-4286 and miR-136-3p, which in turn led to impaired osteogenic differentiation.^44,45 In contrast, miR-31 was highly expressed in ethanol-induced osteonecrosis tissues, and decrease of miR-31 induced by a tumor necrosis factor alpha (TNF-alpha) inhibitor could activate the SATB2/RUNX2 regulatory pathway, leading to increased osteogenic differentiation.⁴⁶ These expression patterns in bone disorders open an avenue to investigate miRNAs further as therapeutics for alcohol-induced musculoskeletal disorders. Agomirs—synthetic molecules that mimic miRNA—are a potential means for therapeutic interventions in alcohol-induced bone disorders. An agomir for miR-19a-3p was able to improve alcohol-impaired fracture healing, while a miR-136-3p agomir was able to ameliorate downregulation of miR-136-3p in alcohol-induced osteopenia, which suppressed osteogenic differentiation.^45,47

Furthermore, miRNAs such as miR-1, miR-127-3p, miR-483-5p, miR-483-3p, miR-628-3p, and miR-885-5p have been explored as diagnostic biomarkers for alcohol-induced osteonecrosis of the femoral head⁴⁸ and have shown promise for early detection and diagnosis of alcohol-induced musculoskeletal disorders. A key future goal will be to ascertain whether such signatures are unique to alcohol’s effects on bone or a shared signal with other organ responses to alcohol. Finally, in a zebrafish model, miR-140-3p and miR-146a were shown to associated with Notch signaling and muscle differentiation, advancing a role for specific signaling pathways as a mechanistic intermediary.⁴⁷ These studies highlight the crucial role of miRNAs in musculoskeletal health, offering insights into potential therapeutic targets and diagnostic markers; however, further studies can provide clinical validation, mechanistic insight, and functional interactions.

Cardiovascular disease

The interplay between alcohol exposure, cardiovascular disease, and miRNAs also was assessed (see Appendix 2B). Several studies showed differential miRNA expression with alcohol exposure in cardiovascular tissues. Specifically, miR-155-5p has been shown to play a role in alcohol-associated cardiovascular pathophysiology. miR-155 is a multifunctional miRNA that mediates several pathophysiological processes in cardiovascular diseases such as coronary artery disease, heart failure, and diabetic heart disease.⁴⁹ For example, studies found that miR-155-5p appeared to regulate ethanol-induced myocardial insulin resistance, affecting mammalian target of rapamycin (mTOR) pathways.^50,51 Moreover, miR-155-5p increased in serum following alcohol consumption, and this increase correlated with cellular apoptosis in ischemic rats as well as rats with chronic alcohol exposure.⁵² Finally, miR-155-5p played a significant role in inflammation mechanisms induced by alcohol that potentially affect cardiovascular health.⁵³ miRNAs have also been explored as biomarkers for cardiovascular diseases; however, this review did not find any studies reporting on miRNAs as biomarkers or therapeutics for alcohol-related cardiovascular disease and risks.

Gastrointestinal injury

Chronic alcohol ingestion alters the gut microbiome composition and increases intestinal permeability and inflammation—factors that could contribute to damage in other organs within the enteric portal circulation, such as the liver (for recent reviews, see Bajaj 2019⁵⁴ and DiVincenzo et al. 2024⁵⁵). miRNAs control a variety of cellular processes, including proliferation, cell death, and integration of enterocytes into a functional epithelial barrier.⁵⁶ They also affect gut inflammation and the composition of the microbiome.⁵⁷ A number of studies have explored the link between alcohol consumption, miRNAs, and intestinal barrier function (see Appendix 2C). Two studies by Tang et al.^58,59 focused on miR-212 as a mechanism underlying reduced ZO-1 protein and enhanced alcohol-induced gut leakiness. An elevation of miR-9-5p was found in alcohol-induced gastric ulcer, which was alleviated by treatment with glycopeptides from the fungus Paecilomyces sinensis (CPS-II) that reduced inflammation and restored mucosal integrity.⁶⁰ This shows that miRNAs may potentially be used as targets for therapeutic intervention to maintain intestinal homeostasis in alcohol-induced gut barrier dysfunction.

Other studies have highlighted the dysregulation of several miRNAs with alcohol exposure, such as miR-141, miR-21, miR-145, miR-155-5p, and miR-146a-5p. These miRNAs play roles in mediating inflammatory responses, apoptosis, oxidative stress, and barrier integrity in gastrointestinal and liver tissues.^61-64 A key consideration when assessing the potential therapeutic utility of these miRNAs is whether target-organ specificity can be achieved, since at least some of these gut-active miRNAs (e.g., miR-21 and miR-9-5p) also have been implicated in alcohol effects in other tissues and pathologies.

Pancreatic and metabolic disease

Several studies have explored the relationship between alcohol, pancreatic and metabolic diseases, and miRNAs (see Appendix 2D). Alcohol use is a risk factor for pancreatitis in both men and women.⁶⁵ A pilot study in human populations documented the differential expression of 150 miRNAs in serum obtained from patients with alcohol-associated chronic pancreatitis.⁶⁶ The authors of that study used informatics approaches to link differentially expressed miRNAs to inflammatory processes, which they suggested were part of the pathology of the condition. In a mouse model, alcohol was shown to upregulate the secretion of both connective tissue growth factor 2 and miR-21 by pancreatic stellate cells; moreover, the miRNA and growth factor were part of a positive feedback loop that promoted collagen production and pancreatic fibrosis.⁶⁷

In contrast with pancreatitis, the link between alcohol consumption and diabetes is complex: moderate levels of consumption have been associated with decreased risk for diabetes, whereas heavy binge-like consumption patterns (defined as five or more U.S. standard drinks per occasion⁶⁸) are associated with increased risk for diabetes.^69,70 One study profiled circulating miRNAs in patients with type 2 diabetes and reported that miR-330, which was elevated in diabetes, also was elevated in persons with AUD.⁷¹ Again, more research will help to better understand the mechanisms involved and to validate miRNAs as biomarkers for alcohol-associated pancreatitis and metabolic disease.

Liver disease

Among the studies included in this scoping review that investigated the role of miRNAs in AUD and associated pathologies across organs, the majority (38%) focused on alcohol-associated liver disease (ALD) in adult human or animal models. ALD is considered the significant contributor to chronic liver diseases, including alcohol-associated fatty liver disease/hepatic steatosis, which can progress to more severe conditions such as alcohol-associated steatohepatitis (ASH), alcohol-associated hepatitis (AH), fibrosis, and cirrhosis. As seen in other pathological conditions, several miRNAs were found to be differentially expressed in patients with ALD or animal models, including upregulation of miR-122,^72-74 miR-155,^73,75,76 miR-21,^77,78 and miR-34a^79,80 (see Appendix 2E). miR-122 is a liver-specific miRNA that is highly expressed in the liver and plays a crucial role in metabolism and cell growth and death.⁸¹ miR-122 along with miR-155 is suggested to be a potential biomarker for liver damage and inflammation in ALD.⁷³ Furthermore, miR-21 and miR-34a are involved in regulating pathways related to cell proliferation, apoptosis, and survival, impacting the progression of ALD.^78,79 Circulating miRNAs, such as miR-513-3p, miR-571, and miR-652, which are found in bodily fluids and serve as stable, non-invasive biomarkers for various physiological and pathological processes, were also differentially expressed in patients with chronic liver disease and liver cirrhosis, demonstrating high diagnostic accuracy for cirrhosis progression.⁸² Additionally, miR-203 showed a protective role in alcohol-associated fatty liver disease;⁸³ miR-26a protected against ethanol-induced hepatic steatosis and liver injury;⁸⁴ and miR-205 reduced ALD progression.⁸⁵ This suggests that these miRNAs could potentially be used as therapeutic targets for ALD.

Numerous studies have examined the role of miRNAs in different stages of ALD using both humans and animal models (see Appendix 2E). miR-30e,⁸⁶ miR-192, miR-122, and miR-30a,⁸⁷ and miR-182⁸⁸ were dysregulated in AH and were related to the progression and severity of AH. miR-21 regulated the nuclear factor kappa-B (NF-kappa-B) pathway, thereby playing a role in modulating hepatic inflammation and reducing inflammatory cytokine release during ALD.⁸⁹ In people with ASH, an increase in miR-432 and sodium-coupled neutral amino acid transporter 1 (SLC38A1) gene expression in the liver correlated with poor survival outcomes, suggesting a potential role of miR-432 in the pathogenesis of ASH.⁹⁰ Furthermore, ethanol exposure decreased miR-192-5p in ASH, which negatively regulated expression of fibronectin type III domain containing 3B (FNDC3B), leading to inactivation of adenosine monophosphate-activated protein kinase (AMPK) and contributing to iron overload and iron-dependent cell death (ferroptosis).⁹¹ A common feature in alcohol-associated fatty liver disease is a dysfunction in lipid metabolism. miR-181b-5p was reported to target protein arginine N-methyltransferase 1 (PRMT1), which is associated with various cellular processes, and miR-378b was shown to target calcium/calmodulin-dependent protein kinase kinase 2 (CaMKK2), which is involved in activating AMPK, to influence lipid metabolism.^92,93

Other studies examined the role of miRNAs and their molecular mechanisms in alcohol-induced liver injury and liver fibrosis in human and mouse samples (see Appendix 2E). The studies found that miR-122 and miR-214 were consistently altered in response to alcohol, affecting antioxidant enzyme levels and fibrotic gene expression.^94-98 Other miRNAs such as miR-145 miR-34a also were implicated as inducers of fibrosis via transforming growth factor-beta (TGF-beta)/SMAD signaling and modulation of macrophage polarization, respectively,^99-101 underscoring the relevance of miRNAs as potential targets for therapeutic interventions in ALD.

Alcohol use disorder/alcohol dependence

Long-term use of alcohol may induce changes in miRNA expression. Studies have identified altered miRNA expression associated with AUD or alcohol dependence across various species and sample types (see Appendix 3A). Animal models have contributed significantly to our knowledge of key biological mechanisms mediated by miRNAs. For instance, the earliest animal study in this area linked miR-9 to the development of acute tolerance to alcohol, which resulted from the miRNA’s ability to change the balance of transcripts with alternate 3’UTRs for the BK potassium-activated sodium channel.¹⁰² In other animal models, key miRNAs, such as miR-30a-5p and miR-206-3p, were upregulated,¹⁰³ with miR-206 playing a role in increased alcohol self-administration and reduction of brain-derived neurotropic factor (BDNF) expression.¹⁰⁴ However, miR-30a-5p restored BDNF levels and decreased alcohol consumption.¹⁰⁵ In humans, differential expression of miR-92, miR-122, and miR-146 also has been observed, with miR-92 and miR-122 generally downregulated in people with AUD^106-108 and miR-146 upregulated,^106,109,110 particularly in persons classified as heavy drinkers (i.e., who reported two or more instances of binge drinking or consumed the equivalent of ≥1.0 oz. of pure ethanol per day⁶⁸). Another study showed that miR-124-3p was increased in the nucleus accumbens and decreased in the limbic forebrain after ethanol withdrawal in animal models of alcohol dependence.¹¹¹ Only one study found sex differences in differential expression of miRNAs after ethanol exposure; for example, miR-125a-3p and let-7a-5p were upregulated in males, but not females, whereas miR-881-3p and miR-504 were downregulated in females, but not in males.¹¹² A number of mechanistic studies also have assessed behavioral endpoints. For instance, inhibition of miR-137¹¹³and overexpression of let-7d¹¹⁴ have been shown to reduce alcohol consumption, indicating their potential as therapeutic targets.

Several miRNAs associated with different stages of AUD (i.e., alcohol consumption, dependence, and withdrawal) are known to regulate mRNA networks involved in neurotransmission, neuroadaptation, and synaptic plasticity.¹¹⁵ For example, miR-130a targets mRNAs for the inositol 1,4,5-triphosphate receptor type 2 (ITPR2) and the Na⁺/K⁺ transporter ATP1A2, which are involved in ion channel regulation.¹¹⁶ Genetic variations in miRNA genes may also contribute to alcohol misuse. For instance, a G>A polymorphism (rs2910164) in miR-146a was significantly more prevalent in people with AUD compared to sex-matched people without AUD, indicating a genetic predisposition linked to miRNA regulation.¹¹⁷ In another study, the level of gamma-aminobutyric acid receptor delta (GABAAR-delta) protein in the dorsal hippocampus was inversely related with miR-365-3p in high-alcohol preferring mice, suggesting genetic background can contribute to miRNA-mediated post-transcriptional mechanisms.¹¹⁸

These studies highlight the multifaceted role of miRNAs in AUD as they influence various neurobiological processes and offer potential avenues for diagnostic and therapeutic interventions. Some of these studies have used saliva samples to assess the association between miRNAs and AUD,^106,119 which suggests the possibility of developing point-of-care devices for conveniently assessing AUD risk and severity. However, further research is necessary to explore and validate these current findings.

Neurological inflammation

Appendix 3B summarizes studies assessing the regulatory roles of miRNAs in alcohol-related neuroinflammation, neurotoxicity, and cellular stress responses. These studies used samples ranging from blood to various brain regions in species ranging from mice to macaques and humans, and employed a variety of analytic techniques. miR-155 was frequently upregulated in response to chronic ethanol feeding in rats and mice, contributing to neuroinflammation and inflammatory cytokine production via toll-like receptor 4 (TLR4) and TLR7 signaling pathways.^120,121 Other studies using mouse models such as TLR4 knockout mice also highlighted alterations in miRNAs that modulate TLR4 and NF-kappa-B pathways following ethanol exposure.^122-125 miR-155 is a pro-inflammatory mediator of the central nervous system and is upregulated in the brain of people affected by many neurodegenerative diseases.^126,127 It may also modulate numerous mechanisms contributing to the etiology of Alzheimer’s disease.¹²⁸ The induction of neuroinflammatory miRNAs following ethanol exposure during development and in adults offers potential targets for therapeutic interventions in alcohol-related neurobiology.

FASD/neurodevelopmental disorders

Research has evaluated the impact of ethanol exposure on miRNA expression across various species, tissues, and developmental stages (see Appendix 3C). Using diverse systems, from zebrafish and mice to sheep and humans, and sample types ranging from blood to embryonic tissues, and diverse analytic techniques, studies have shown that ethanol exposure significantly alters miRNA expression, impacting a variety of biological processes.

Several miRNAs, including miR-9, miR-15b, miR-19b, and miR-20a, were dysregulated in ovine blood upon ethanol exposure,¹²⁹ indicating they could be candidate biomarkers for maternal ethanol exposure, as well as biomarkers in newborns to test for fetal exposure. In human embryonic stem cells, miR-145 mediated alcohol toxicity by targeting the transcription factor Sox-2 and extracellular signal-regulated kinases, resulting in neural progenitor depletion.¹³⁰ In another study, the signaling cascade involving the transcription factor SP1, protein kinase R (PKR), and PKR-associated protein X (RAX) was linked to ethanol-induced cerebellar neuron apoptosis via miR-29b in mice.¹³¹ 

miR-9, the miRNA most often identified in studies of neurological and developmental effects (see Appendix 3), is important for neural stem cell maturation and brain development.^132,133 Multiple studies across different species and tissues, including zebrafish embryos,¹³⁴ murine blood,¹³⁵ ovine blood,¹²⁹ monkey brain,¹³⁶ and human brain and fetal central nervous system-derived extracellular vesicles¹³⁷ have shown dysregulation of miR-9 upon ethanol exposure, highlighting its significance in these processes. These findings highlight the complex interplay between ethanol exposure and miRNA regulation, underscoring the potential of miRNAs such as miR-9 as biomarkers for ethanol-induced developmental disorders and as therapeutic targets for mitigating the adverse effects of prenatal alcohol exposure.

Knowledge gaps: miRNA regulation and miRNA effects

A major focus of studies of miRNAs in the context of AUD and associated conditions has been on screening tissues to identify and enumerate the miRNAs that are responsive to alcohol exposure in brain tissues and in organs that are secondary AUD targets (see Table 3). Relatively few studies have focused on mechanisms that link ethanol to changes in miRNAs. Some linking mechanisms that have been identified include methylation changes in miRNA gene loci^43,138 and promoter occupancy by chromatin remodeling factors such as the BRG1/BRM-Associated Factor (Baf) complex.¹³⁹ In general, however, this area has been poorly investigated.

Similarly, gaps remain in understanding the effects of miRNAs, whether acting alone or in groups of co-regulated miRNAs. Most studies on the roles of miRNAs in AUD have focused on single miRNAs. For instance, an early study showed that striatal explant cultures obtained from juvenile male rats and treated with ethanol exhibited an increase in miR-9 and a loss of splice variant transcripts of the BK potassium channel whose 3’UTR contained miR-9 binding sites.¹⁰² The authors of that study interpreted this as a mechanism for ethanol tolerance in striatal neurons, an important phenotype for the emergence of AUD. A subsequent study implicated another important brain-expressed miRNA, miR-124a, and its regulated target, BDNF, in alcohol preference.¹⁴⁰ Elevation of miR-124a levels by microinjection into the dorsal striatum of adult male rats resulted in decreased BDNF levels and increased alcohol preference in a conditioned place preference and two-bottle choice paradigm; conversely, miR-124a inhibition resulted in decreased voluntary alcohol consumption.¹⁴⁰ Importantly, these studies only included male rats; therefore, it is unknown if similar miRNA mechanisms facilitate alcohol tolerance, preference, and consumption in females. The lack of studies powered to detect differences between males and females represents a significant limitation to early studies that is only recently being addressed.

These early studies also did not specify whether for the miRNA in question the guide or the passenger strand of the miRNA strand was retained following DICER cleavage. It may be reasonable to assume that the studies implicated the dominant (guide) strand of the miRNA (i.e., miR-9-5p and miR-124a-5p). However, caution is warranted, because passenger strand miRNAs (e.g., miR-9-3p) also can be retained,¹⁴¹ be differentially regulated by ethanol,¹⁴² and exhibit a unique functional specificity.¹⁴³

According to the dominant conceptual model, miRNAs exert their influence by binding to the 3’UTRs of target mRNAs to destabilize the mRNAs and thereby inhibit gene translation. A number of studies utilized the approach of creating reporter constructs in which the luciferase gene (whose activity can easily be measured) was linked to the 3’UTR of the presumptive target mRNA to explore the mechanistic link between ethanol-sensitive miRNA and mRNA translation.^{97,102,144-146}

Early studies already noted the potential for miRNAs to cooperatively target mRNA translation.³⁵ Most miRNA-regulated mRNAs have multiple miRNA target sequences within their 3’UTRs, and most miRNAs target multiple mRNA 3’UTRs. Therefore, the potential for cross-talk between miRNAs is an important but poorly studied component of AUD biology. An early study did present evidence that miRNAs inhibited by ethanol in neural progenitor cells cooperatively controlled cell survival and target gene expression.⁴¹ A more recent study identified a cluster of 11 miRNAs that are elevated in the second trimester of pregnancy, in the plasma of some alcohol-exposed pregnant women. These miRNAs collectively explained 24%–30% of the variance in growth metrics in newborn infants, collectively inhibited placental growth in mice, and inhibited invasiveness of human trophoblast cells.¹⁴⁷ However, the contribution of differentially regulated miRNA clusters to the pathogenesis of AUD has rarely been investigated to date.

Other sources of miRNA variation

RNA sequencing studies have shown that individual miRNAs within a single organism can exhibit substantial sequence heterogeneity due to processing variations and post-transcriptional modifications. DICER processing of a given pre-miRNA can be imprecise and result in a cell-type–dependent heterogeneity of mature miRNA products,¹⁴⁸ for example, in the 5’-seed sequence for targeting mRNAs. Nucleotide polymorphisms within miRNA genes can also contribute to variation in the size and sequence of mature miRNAs.

Another cause of miRNA variation are adenosine deaminases of RNAs (ADARs), a class of enzymes that can change adenosine to inosine (A-to-I editing) in RNAs, including miRNAs. A-to-I editing of pri-miRNAs can block cleavage by DICER.¹⁴⁹ However, ADARs can also form a complex with DICER to facilitate miRNA processing; conversely, loss of ADARs can globally inhibit miRNA expression.¹⁵⁰ A-to-I editing within a miRNA seed sequence can alter targeting specificity and function.¹⁵¹ Moreover, A-to-I editing within mRNA transcripts can also reveal new miRNA regulatory sites,¹⁵² suggesting that mechanisms that control RNA editing can also result in dynamic changes in miRNA:mRNA interactions. RNA editing of protein-coding genes has been shown to regulate alcohol consumption,¹⁵³ but the effects of alcohol on editing miRNAs are unknown.

Like lncRNAs (discussed below) and mRNAs, miRNAs can also be post-transcriptionally modified by methylation. For instance, some miRNAs that are associated with AUD targets (e.g., miR-17-5p, miR-21-5p, miR-200c-3p, and let-7a-5p) exhibit increased methylation in cancer.¹⁵⁴ A recent paper provided evidence that members of the let-7 miRNA family were subject to 7-methylguanosine (m7G) methylation by the enzyme METTL1 and that m7G methylation of pri-Let-7e promoted more efficient processing of pri-miRNA to pre-miRNA by Drosha.¹⁵⁵ 6-methyladenosine (m6A) methylation has also been described in miRNAs and shown to be important for promoting pri-miRNA processing.¹⁵⁶ Numerous studies have shown that ethanol both increases and decreases a large number of miRNAs and that at least some of the regulation may occur at the level of methylation. However, no studies to date seem to have identified ethanol effects on the METTL RNA methyltransferase family, although one study noted that operant ethanol self-administration resulted in significant elevation of tRNA methyltransferase 1 mRNA transcripts in prefrontal cortex of young adult male rats.¹⁵⁷ Although the study included no females, representing a common gap in our knowledge, these data support a more in-depth analysis of miRNA methylation machinery in AUD pathology.

Other sources of RISC complex variation

As mentioned above, the activities of RISCs can be modulated by ADARS. A large number of RISC binding partners have been identified, suggesting that the association between argonautes and their binding partners can contextualize miRNA control over cellular translation. The RNA-binding protein fragile X messenger ribonucleoprotein 1 (Fmr1) is an example of the potential intersection between RISC function and AUD pathology. Fmr1 by itself can promote Ago activity; however, when it binds to the RISC-associated helicase, Mov10, it inhibits the translational-repression activity of Ago proteins.¹⁵⁸ Fmr1 phosphorylation promotes the translation-inhibitory activity of Ago, while Fmr1 dephosphorylation associated with cell-surface G-protein-coupled-receptor (GPCR) signaling results in increased protein translation.¹⁵⁹ This implies that GPCR signaling at a synapse can switch on local protein synthesis by controlling the phosphorylation state of an Ago-associated protein. The contextual environment of a RISC complex member like Fmr1—that is, phosphorylation state and binding partner availability—is important because a number of studies have linked ethanol consumption with elevations in brain Fmr1.^160-162 These studies have also linked elevations in Fmr1 with shifts in glutamate and GABA plasticity, advancing a potential miRNA-mediated mechanism in AUDs. It will be important for future studies to directly link RISC proteins and their activation states with synaptic plasticity responses to ethanol among others.

miRNA compartmentalization

miRNAs can be found in different parts of the cell or can be secreted from cells. The dominant conceptual model is that miRNAs are processed and act in the cytoplasm of cells and tissues to inhibit the translation of their target mRNAs. Indeed, most of the reviewed studies used this conceptual framework. However, once miRNAs have been processed in the cytoplasm, they can be translocated back to the nucleus. Evidence in support of nuclear functions of miRNAs comes from studies showing that Ago proteins (which act as chaperones of RISC complex and thus miRNAs) can bind the protein trinucleotide repeat containing adaptor 6A (TNRC6A), which has both nuclear localization and export signals and can shuttle between nucleus and cytoplasm.¹⁶³ Mature miRNAs also have been shown to translocate to the nucleus, where they can regulate transcription. For instance, Ago1 protein was found in the promotor region of the mouse cyclin B1 (Ccnb1) gene, and the miRNA miR-744, which targets a site in the promoter of the Ccnb1 gene, increased RNA polymerase-2 occupancy at the transcription start site, Ccnb1 transcription, and cell proliferation.¹⁶⁴ Several publications subsequently documented both transcription activation and transcription silencing due to nuclear miRNA and RISC complex activity (for a review, see Bhattacharjee et al. 2019¹⁶⁵). One recent study linked alcohol exposure to nuclear actions of miRNAs.¹⁶⁶ The authors showed that in a hepatocellular carcinoma cell line, ethanol exposure resulted in decreased levels of miR-29c, a miRNA that was present in the nucleus and exhibited Ago2-dependent binding to the enhancer region of the alcohol dehydrogenase 6 (ADH6) gene, but not to the 3’UTR of ADH6 mRNA. In their model, miR-29c activated the enhancer and recruited RNA polymerase II to ADH6 and other ADH genes to promote gene transcription. These data provided evidence that a miRNA acting within the nucleus could mediate epigenetic effects of ethanol. Given the link between alcohol and cancer, such transcription control may represent an important mechanism for tumor progression that warrants further investigation.

Other studies showed that miRNAs could also be secreted in biofluids and serve as biomarkers for disease.^167-169 Subsequently, evidence emerged that secreted miRNAs were chaperoned by Ago proteins,¹⁷⁰ lipoproteins such as high-density lipoprotein, ¹⁷¹ or in small extracellular vesicles.¹⁷² AUD-related studies showed that miRNAs whose levels were elevated in plasma extracellular microvesicles could serve as biomarkers for ALD in mice.⁷³ Moreover, in an ovine model of pregnancy, maternal circulating miRNAs were biomarkers for fetal alcohol exposure.¹²⁹

Circulating miRNA research accounts for about 18% of the publications on miRNAs in AUDs identified in this scoping review. To date, most published papers have reported on screening circulating miRNAs as biomarkers for prenatal alcohol exposure and associated pathology. Only a few studies have used circulating miRNAs in predictive models for future disease burden due to alcohol. For example, two papers showed that a cluster of microRNAs that was elevated in blood plasma of pregnant women during the second trimester predicted growth deficits of newborn infants.^129,147 A transition in the field from enumerating the effects of alcohol and AUD pathology on circulating miRNA expression to a focus on predicting future health outcomes associated with alcohol exposure, including future disease burden and mortality from miRNA profiles, is likely to yield substantial gains in health management for AUDs.

Circulating miRNAs also have the potential to serve as endocrine molecules—that is, when they are secreted into circulation by cells and tissues, they can affect the biology of recipient cells and tissues. Some studies have tried to assess the endocrine potential of secreted miRNAs. This has been useful and informative when the focus has been on individual miRNAs. For example, one study showed that alcohol exposure resulted in increased release of miR-122 from hepatocytes in extracellular microvesicles, and that recipient monocytes exhibited inhibition of the heme oxygenase 1 pathway and increased sensitization to lipopolysaccharide stimulation.¹⁷³ A second study showed that miR-27a release from alcohol-treated monocytes into extracellular vesicles resulted in increased expression of the mannose receptor CD206 and the hemoglobin-haptoglobin scavenger receptor CD163, as well as transformation of naïve monocytes into M2-type macrophages.¹⁷⁴ Such macrophages may be mediators of fibrosis and and/or tumorigenesis in tissues such as liver.¹⁷⁵ A third paper showed that ethanol exposure resulted in increased miR-140-3p levels in extracellular microvesicles secreted by cultured mouse neural progenitor cells and that a molecule mimicking miR-140-30 (miR-140-3p mimetic) increased cell proliferation of neural progenitors.¹⁷⁶ Finally, one study presented a novel, atypical mechanism for secreted miRNA activation of an inflammatory response, resulting in a transfer of Let-7b from Ago to the proinflammatory high mobility group box 1 protein in microvessels, for presentation to TLR7.¹⁷⁷ This mechanism was associated with subsequent neurotoxicity. This last study emphasizes the diversity of chaperone–miRNA interactions that may be activated by ethanol.

These types of studies on individual miRNAs are valuable but have limited impact for understanding AUD pathology, because effects of ethanol are often associated with changes in the extracellular release profile of a large number of miRNAs. The collective action of this miRNA cohort on recipient tissues may not reflect the behavior of their individual component miRNAs. Some studies have attempted to make predictions about endocrine biology of secreted miRNA cohorts using pathway-enrichment analysis (e.g., see Jing et al. 2015¹⁷⁸). However, such an approach provides, at best, a theoretical framework that needs to be tested experimentally. Two studies did attempt a comprehensive analysis of the cohort behavior of secreted miRNAs.^147,179 Based on data showing that 11 maternal circulating miRNAs that were elevated in the second trimester of pregnancy predicted infant birth outcomes following prenatal alcohol exposure,¹⁸⁰ these studies assessed the behavior of individual and grouped miRNAs on placental trophoblast gene expression, trophoblast behavior, and fetal growth. One study found that as a group, the 11 miRNAs inhibited epithelial-to-mesenchymal transition (EMT) genes in human trophoblast cells, inhibited trophoblast invasion, and explained the effects of prenatal ethanol on EMT in rodent and primate models.¹⁴⁷ Moreover, injecting eight out of these 11 maternally elevated miRNAs into a naïve pregnant mouse at mid gestation resulted in fetal growth restriction compared to controls, a hallmark phenotype of prenatal alcohol exposure. Importantly, the behavior of no single secreted miRNA predicted the behavior of the miRNA cohort. A subsequent transcriptomic study showed that this cohort of miRNAs collectively upregulated Notch signaling pathway genes, dysregulated angiogenic genes in placenta, and decreased umbilical cord blood flow.¹⁷⁹ Collectively, these data emphasize the functional, endocrine nature of miRNAs that are secreted into biological fluids in conjunction with AUD pathology. The translational potential of these findings is strong, suggesting that miRNA mimetics or antagonists may be delivered into circulation to mitigate effects of alcohol exposure.

Sex as a biological variable in miRNA studies

Only a minority of miRNA studies (approximately 39%) reported using female subjects, and it is not clear if the studies were adequately powered to detect sex differences in miRNA regulation or miRNA-dependent biology. Newer statistical methodologies, such as bootstrap-resampling approaches—which involve repeatedly sampling from the data with replacement to estimate the accuracy and variability of statistical measures—can help overcome false discovery rate barriers. These methods can be applied in secondary data analyses to identify the contributions of important biological variables, such as genetic sex, age, and others. For instance, a secondary analysis of maternal miRNA profiles in pregnancy using such a bootstrap resampling approach uncovered distinct maternal circulating miRNA profiles in pregnancies with male and female fetuses.¹⁸¹

Long Noncoding RNAs (lncRNAs)

lncRNAs have been arbitrarily defined as RNAs that exceed 200 nucleotides in length, but many exceed several kilo bases (kb) in length. For instance, lncRNA kcnq1ot1/NONCODE transcript ID# NONHSAT017523.2, which is encoded on the antisense strand for the protein-coding gene kcnq1, is estimated to exceed 91kb in length.¹⁸² The human genome is annotated to encode more than 172,000 lncRNA transcripts, while the mouse genome encodes more than 131,000 lncRNA transcripts.¹⁸² However, most lncRNAs are not well-characterized, and the functions of many are simply unknown. While a large number of lncRNAs are conserved, many others are specific to humans, with functions that are not present in other species. This means that lncRNA biology relevant to human alcohol use and to AUD may be both evolutionarily conserved and human-specific. Compared with miRNA research, little work has been done on lncRNAs and alcohol. Approximately 10% of the assessed papers in this review dealt with lncRNAs. Especially given the large number of annotated lncRNAs, the published literature is sparse and presents a limited window on their contribution to the biology of AUD and their effects. Appendix 4 summarizes the methods and key findings of various studies examining lncRNAs in the context of alcohol-related diseases.

Modes of lncRNA action: Nuclear functions

lncRNAs are generally thought to play a critical role in the packaging and modification of the nuclear chromatin structure for epigenetic regulation of gene expression, but may have a range of additional and mostly unexplored functions (for a consensus statement, see Mattick et al. 2023¹⁰). In humans, lncRNA UBE3A-ATS (SNHG14), which is encoded on the antisense strand for the gene encoding the ubiquitin ligase UBE3A,¹⁸³ is an important example of functional epigenetic control. This lncRNA, encoded from the paternal allele in the nervous system, epigenetically silences the paternal UBE3A allele; therefore, loss of the maternal UBE3A allele (e.g., due to mutation or deletion) can result in developmental disability (Angelman Syndrome).¹⁸⁴ Studies with other gene-silencing lncRNAs, such as XIST, have shown that gene silencing is mediated by the presence of repeating stem-loop structures, whereas chromatin binding is mediated by other portions of this lncRNA.¹⁸⁵ Moreover, the organization of targeted chromatin structure can also influence the efficacy of lncRNAs. For instance, regions of chromatin that have a low density of long interspersed nuclear element-1 repeats are relatively resistant to XIST-mediated silencing.¹⁸⁶ These data suggest that sequence and structure of lncRNAs as well as the structure of target chromatin may be important determinants of epigenetic regulation by lncRNAs. Finally, epigenetic control mechanisms mediated by a single lncRNA may be dependent on cell-types and physiological context. For instance, cell-types such as B-lymphocytes, which are subject to substantial genetic recombination to generate antibody diversity, utilize XIST in a unique complex of proteins to stall translation of specific X chromosome-linked immune genes such as TLR7 that do not contain sequences (i.e., CpG islands) susceptible to DNA methylation.¹⁸⁷ Such cell-contextual partnering between lncRNAs and epigenetic machinery has not been explored in the context of AUD and related effects.

Like UBE3A-ATS, numerous lncRNAs are transcribed from the DNA strand opposite to that which encodes a protein; they are therefore referred to as long antisense RNAs, with a nomenclature that acknowledges the antisense position relative to the protein-coding gene. BDNF-AS, the lncRNA encoded on the opposite strand to the gene coding for BDNF. Decreased BDNF in the dorsolateral striatum itself resulted in increased alcohol self-administration in a rat model.¹⁸⁸ However, BDNF-AS, which acts as an inhibitor of the BDNF,¹⁸⁹ was found to be elevated post-mortem in the amygdala of persons with a diagnosis of early-onset AUD; additionally, BDNF-AS recruited the methyltransferase protein enhancer of zeste homolog 2 (EZH2) to the BDNF promoter, resulting in promoter silencing.¹⁹⁰ Antisense lncRNA transcripts are associated with many important growth and plasticity genes, and their actions, as with the BDNF/BDNF-AS gene may mediate many pathogenic effects of ethanol, at the genomic level, but little is known about how such pairs of genes function.

Aside from epigenetic mechanisms, lncRNAs can simultaneously exhibit multiple alternate nuclear functions. The abundant nuclear lncRNA, metastasis associated lung adenocarcinoma transcript 1 (MALAT1), also known as nuclear enriched abundant transcript 2 (NEAT2), was upregulated in postmortem brain samples from people with a diagnosis of AUD and increased in rat brain following withdrawal from ethanol.¹⁹¹ It was associated with “nuclear speckles”¹⁹²—sites of active mRNA transcription.¹⁹³ MALAT1/NEAT2’s partner in the nucleus, NEAT1, which is elevated in ALD,¹⁹⁴ is a pertinent example alternate isoforms of individual lncRNAs having different functions. For instance, NEAT1 exists in two isoforms of 23kb and 3.7kb length. The 23kb isoform normally localizes within the nucleus where it serves as an architectural RNA for the assembly of paraspeckle bodies¹⁹⁵ (i.e., interchromatin structures adjacent to MALAT1/NEAT2-enriched nuclear speckles) that may be important for retaining double-stranded RNA that has been subject to A-to-I editing.¹⁹⁶ However, NEAT1 can be subject to alternate 3’-polyadenylation, and the 3.7kb polyadenylated isoform does not retain the capacity to assemble paraspeckle bodies;¹⁹⁷ instead, it may inhibit paraspeckle body formation and have alternate, perhaps antagonistic functions in cells.^198,199 Alternate polyadenylation and splicing variations in lncRNAs have not been investigated in AUD pathology.

Post-mortem analysis of brain tissues documented that a large number of lncRNAs, including long antisense RNAs and pseudogene transcripts (RNAs transcribed from presumably non-functional gene copies of protein-coding genes) were either elevated or suppressed coincident with a diagnosis of AUDs.²⁰⁰ The authors of that study used gene ontology—an analytic approach for systematic and consistent classification of gene functions, regardless of the organism—to advance a hypothesis that the dysregulation of these lncRNAs was associated with genome-wide mis-splicing of RNA transcripts. This hypothesis is an innovative and testable link between lncRNAs and alcohol’s pathogenic effects, but it is unlikely to be the only or even major mechanism of alcohol-related pathogenesis. Increasing evidence suggests that most, if not all, lncRNAs have multiple, context-dependent functions, and each lncRNA needs substantial further investigation.

Modes of lncRNA action: Cytoplasmic functions

Changes in location of ncRNAs within cells (i.e., translocation between nucleus and cytoplasm) can dramatically alter the function of an ncRNA. For instance, under stress or inflammatory conditions, NEAT1 in macrophages can also localize to the cytoplasm and, for example, stabilize caspase-1 enzyme and promote the production of interleukin-1b.²⁰¹ It will be important to analyze how cytoplasmic lncRNAs interact with their protein partners in AUDs—for example, using RNA-immunoprecipitation methodologies.²⁰²

Within the cytoplasm, NEAT1 also functions as a “competing endogenous RNA” (ceRNA), serving as a “sponge” to sequester miRNAs.^203,204 In a study of ASH, Ye and colleagues showed that levels of NEAT1 and miR-129-5p were inversely related to each other and that the lncRNA/miRNA pair behaved as functional antagonists, which suggests a ceRNA function for NEAT1.¹⁹⁴ Its nuclear partner, MALAT1/NEAT2, can also serve as a miRNA sponge.²⁰⁵ Similarly, BDNF-AS lncRNA, which was linked to epigenetic mechanisms, also exhibits miRNA sponge activity by sequestering the miR-9-5p,²⁰⁶ which has been implicated in animal models in the pathogenesis of both fetal effects of ethanol^{41,139,207,208} and of AUD.¹⁰² BDNF-AS therefore classifies as a ceRNA, although its potential ceRNA activity in AUD and related consequences has not been explored. Another antisense RNA, HOX transcript antisense RNA (HOTAIR), has been linked to pathogenic effects of ethanol in bone and liver^209-211 and has been shown to exhibit miRNA sponge activity in its contribution to ethanol-related pathogenesis.^209,210 Other antisense strand-encoded lncRNAs may exhibit similar ceRNA activity in the context of ethanol effects and need to be further investigated.

Potential ceRNA activity for lncRNAs can include not only miRNA sponge activity, but also the direct control of gene translation via regulation at the 3’UTRs of mRNAs. An example is a lncRNA encoded from a pseudogene locus (Oct4pg9), which was created by duplication of the Oct4/Pou5f1 gene (a core transcription factor for stem cell identity). In developing cortical neural progenitor cells, Oct4pg9 was elevated following ethanol exposure.²¹² This study showed that Oct4pg9 lncRNA not only binds miRNAs and their Ago chaperones (i.e., has a sponge activity) but also directly and independently inhibited translation of a reporter gene linked to the Oct4/Pou5f1 3’UTR. This suggested a novel ceRNA mechanism whereby Oct4pg9 lncRNA directly repressed translation of Oct4/Pou5f1,²¹³ providing evidence that the paired protein-coding gene and pseudogene-encoded lncRNA could mediate loss of stem cell identity following ethanol exposure. In general, the ceRNA function of lncRNAs warrants substantial further investigation.

Modes of lncRNA action: Endocrine/paracrine lncRNAs

lncRNAs, like miRNAs, also are secreted by cells into extracellular spaces and fluids, where they may serve as endocrine molecules. For instance, studies documented that NEAT1 is secreted in extracellular vesicles and can regulate pathogenic processes, including cardiac fibrosis²¹⁴ and cancer²¹⁵ by exploiting its various modes of function in recipient target cells. The role of endocrine lncRNAs in alcohol pathology is largely unknown. Moreover, the presence of lncRNAs in body fluids such as blood presents an exciting opportunity to assess these molecules for diagnostic purposes. Some published studies have indeed assessed lncRNAs in biological fluids as markers for ALD^194,216,217 and esophageal cancer,²¹⁸ and these molecules could be further investigated as biomarkers for other AUD-associated diseases.

Modes of lncRNA action: Posttranscriptional processing

Following transcription, all RNAs may be modified by the addition of methyl groups, mainly to adenosines (m6A²¹⁹), although methylation at cytosines and guanosines (m5C, m7G, etc.) can also occur. Such epi-transcriptomic modifications can control a variety of processes, including RNA splicing, export, stability, and translation.²²⁰ RNA can also be more directly modified by direct editing of base sequences, such as A-to-I editing by ADARS,²²¹ resulting in alterations in the information content of an RNA sequence. Both types of post-transcriptional modifications have been described for lncRNAs as well,^222,223 but remain a poorly investigated area of study in general, and in AUDs specifically.

In cancer cell-culture models, ethanol exposure has been shown to reduce m6A methylation around the stop codons for opioid receptor mRNAs.²²⁴ In human populations, mutation patterns and expression of genes regulating m6A mRNA methylation predicted disease-free intervals in patients with alcohol-associated hepatic carcinoma,²²⁵ supporting the hypothesis that RNA methylation mechanisms may mediate pathogenic effects of ethanol. However, the role of such epi-transcriptomic modifications of lncRNAs that mediate ethanol effects remains to be determined.

Other studies in animal models found that editing of protein-coding mRNAs, such as the brain 5HT2C serotonin receptor, increased as a consequence of ethanol exposure.^153,226 Moreover, inhibiting A-to-I editing of serotonin receptor in the nucleus accumbens decreased alcohol intake after chronic ethanol exposure,¹⁵³ suggesting that A-to-I editing has functional consequences for both molecular and behavioral manifestations of how alcohol impacts the individual (i.e., AUD phenotypes). However, the effects of ethanol on lncRNA editing or the consequences of such editing remain to be studied.

lncRNAs can also be post-transcriptionally processed to generate small RNAs with divergent functions. The 3’-end of the ethanol-induced nuclear MALAT1/NEAT2 lncRNA can be cleaved by RNAse-P to generate a small tRNA-like fragment, MALAT1-associated small cytoplasmic RNA (mascRNA), that, unlike its parent lncRNA, localizes to the cytoplasm. MascRNA was found to globally promote gene translation by stabilizing a component of the multi-tRNA synthase complex.²²⁷ However, mascRNA can also promote protein ubiquitination and proteasomal degradation,²²⁸ suggesting a more nuanced role for this small RNA as a mechanism to control protein turnover in cells. It remains to be determined whether elevated levels of MALAT1/NEAT in AUD imply a commensurate increase in mascRNA. However, a role for mascRNA in AUD pathology is suggested by findings that mascRNA modulates effects of TLRs, inhibiting downstream activation of TLR4 but facilitating TLR3.²²⁸ Both TLR3 and TLR4 have been shown to be increased in rat brain following ethanol exposure and mediate inflammatory responses in these models.^229,230 Thus, ncRNAs such as mascRNA may add nuance to AUD-associated inflammation by shaping the balance between inflammatory pathways.

Modes of lncRNA action: Micro-peptide translation

lncRNAs have been labelled as non-protein–coding RNAs because they do not appear to possess conventional protein-encoding open reading frames. However, uncertainty persists around whether lncRNAs truly are non-protein-coding or could be translated under some conditions. For example, early evidence suggested that lncRNAs could associate with ribosomes, suggesting that they may in fact be translated.²³¹ A 2014 study examining the structure of lncRNAs suggested that these RNAs may indeed contain short open reading frames that may encode small peptides (micropeptides).²³² Finally, in 2015, researchers reported that a micropeptide termed myoregulin was indeed encoded by a lncRNA (human annotation: LINC00948, mouse: 2310015B20Rik), was expressed in skeletal muscle, and regulated Ca²⁺ uptake into the sarcoplasmic reticulum.²³³ Since that report, several publications have documented that, aside from their non-protein-coding functions, lncRNAs do contain short open reading frames and can simultaneously encode micropeptides.^234-236 Moreover, pri-miRNAs, a class of lncRNAs that have hitherto been solely regarded as precursors for miRNAs, also may simultaneously encode micropeptides with independent functions,²³⁷ suggesting a new level of biological complexity associated with lncRNAs that parallels the complexity of miRNA biology. The emerging data suggest that the activity of lncRNAs may be contextual and result, at times, in these RNAs serving as templates for translating small peptides. This contextuality of lncRNA biology presents an exciting area of research that has not been explored in the field of AUD biology.

lncRNAs mediate sex-specific effects of ethanol

Genetic sex is increasingly identified as a key contributor to individual differences in disease susceptibility; however, until recently, sex differences in AUD have been poorly documented (for a review, see Becker and Koob 2016²³⁸). This deficiency also exists for the emerging knowledge base of lncRNA contributions to biology. For instance, the first two lncRNAs to be identified—H19, a paternally imprinted lncRNA^14,239 associated with developmental growth control,²⁴⁰ and XIST, the female-specific lncRNA^13,241 responsible for inactivating the supernumerary X-chromosome in female cells as a dose compensation mechanism²⁴²—are also targets of ethanol, both during development and in the adult, but with sex-specific differences. For example, prenatal alcohol exposure in a mouse model increased brain H19 lncRNA, but only in male offspring.²⁴³ It will be important to identify a functionally equivalent lncRNA in female offspring. Indeed, when lncRNAs are regulated or act in a sex-specific manner, it will be important to identify equivalent biological mechanisms that contribute to common outcomes. Increased H19 levels were shown to promote processing of another lncRNA, Homer, into a stable, covalently closed loop (i.e., circularization) leading to higher levels of circHomer (for more on circRNAs, see below). The potential circHomer-specific effects were observed only in male offspring.²⁴³ It remains unknown whether equivalent regulatory mechanisms contribute to alcohol’s effects in females, highlighting an important area for future research.

The involvement of XIST as a mediator of alcohol effects, in contrast, suggests the existence of female-specific mechanisms. XIST is a female-specific lncRNA that randomly and epigenetically silences one of two X chromosomes (for review, see Brockdorff et al. 2020²⁴⁴), resulting in chromatin condensation and Barr body²⁴⁵ formation. Three papers have connected ethanol exposure to XIST. A study in human liver stellate cells showed that ethanol increased levels of XIST,²⁴⁶ while another study in mouse neuroblastoma cells found that ethanol exposure transiently increased and then decreased XIST levels.²⁴⁷ These two papers present a partly overlapping outcome resulting from ethanol exposure. Finally, a third study, using a single-cell RNA-seq approach, showed that a single episode of ethanol exposure in a pregnant mouse at the onset of fetal neurogenesis resulted in widespread decreases in expression of XIST across multiple cell types in the developing female fetal cerebral cortex, with evidence of loss of X chromosome inactivation (i.e., increased transcript abundance from X chromosome-encoded genes).²⁴⁸ The partial discrepancy between studies highlights the fact that more research is needed to define the contributory role of intervening factors such as variations in the sensitivity of various cell/tissue types to ethanol, and the effects of alcohol exposure pattern (e.g., episodic/acute vs. chronic) and dose on outcomes. Moreover, since XIST is a female-specific transcript, it will be important to determine whether equivalent, but XIST-independent, mechanisms exist in males. At a more global level, a greater understanding of the sex-specific roles of lncRNAs will pave the way for precision medicine approaches to treating AUD.

Circular RNAs

circRNAs are a type of ncRNA with a looped structure without 5' or 3' ends. This structure makes them more stable than linear RNAs due to their resistance to exonuclease. circRNAs are generated from precursor mRNAs (pre-mRNAs) through back-splicing of exons (exonic circRNAs), introns (circular intron RNA), or both (exon-intron circRNAs).²⁴⁹ Among these, exonic circRNAs are the most abundant type. There are three main models for the biogenesis of circRNA: lariat-driven, intron-pairing-driven, and RNA-binding proteins (RBPs)-driven circularization. In the lariat-driven circularization, the pre-mRNA forms a loop (lariat) intermediate that contains introns and/or exons. The pre-mRNA undergoes splicing, including removal of the exons (exon skipping), followed by reverse splicing that leads to the formation of the circRNA.²⁵⁰ In contrast, intron-pairing–driven and RBP-driven circularization events are both direct back-splicing mechanisms, where two flanking introns are brought into close proximity by repeated, short, complementary DNA sequences (ALU elements), other complementary sequences without repetition, or RBPs respectively.²⁵¹

When first identified more than 30 years ago, circRNAs were thought to be non-functional. However, today, they have been acknowledged to have several important functions, such as acting as miRNA sponges, regulating host gene expression, interacting with proteins, and protein transportation.²⁵¹ They can also serve as diagnostic biomarkers and therapeutics because of their higher nuclease stability and longer half-lives in cells.²⁵² In their function as miRNA sponges, circRNA act as ceRNAs, modulating miRNA function and regulating transcriptional gene expression.²⁵² Some circRNA can also directly regulate the transcription of their parent gene by interacting with the RNA polymerase II complex.²⁵³ Finally, circRNAs function as protein sponges by providing the binding sites for proteins like RBPs and inhibiting the biological activity of proteins.²⁵⁴ Like lncRNAs, circRNAs can also be translated into functional peptides, although this is relatively rare and not fully understood.²⁵²

Overall, circRNAs are widely studied in the fields of cancer; cardiovascular, neurological, immune-related, metabolic, and infectious diseases; and aging.²⁵⁴ However, the association between circRNAs and alcohol use and misuse is an emerging area of research that has only gained attention recently. Most of the studies to date were conducted to identify the circRNAs and their functional outcomes in AUD, including differential expression of circRNAs in prenatal alcohol exposure,^243,255,256 ALD,^257,258 and alcohol dependence²⁵⁹ (Appendix 5). circRNAs have also been reported to influence neurotransmitter and signal-transduction pathways after chronic intermittent alcohol exposure,²⁶⁰ and to regulate cell survival and apoptosis.²⁶¹ Various studies have investigated the differential expression of circRNAs in animal models, such as mice and rats, and in humans. A study on human serum and exosomes identified hsa_circ_0004771 as a potential biomarker for alcohol dependence (as defined in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition).²⁵⁹ A number of circRNAs have been found to be regulated by prenatal alcohol exposure. Thus, downregulation of circHomer1 occurred in the brain of prenatally exposed adult male mice,²⁴³ and sex-specific upregulation of circSatb2 and downregulation of circPtchd2 was found in the brain of prenatally exposed male mouse pups.²⁵⁶ Moreover, upregulation of circVopp1 in both blood leukocytes and spinal cord tissues but downregulation of spinal immune-regulatory circRNAs, circItch and circRps6ka3, was found in prenatally alcohol exposed adult rats.²⁵⁵ The interactions of circRNA with miRNAs such as circRNA-406742 and miR-1200²⁶² or circRNA-1639 and miR-122,²⁵⁷ indicates a potential role of circRNAs as ceRNAs by acting as miRNA sponges, thereby modulating miRNA activity and thus influencing translation of gene networks associated with AUD.

These scoping analyses of studies assessing circRNAs in AUD highlight several key gaps in the current understanding of the circRNAs’ roles. Thus, little is known about the link between alcohol and the biology of circRNAs. Another important area that deserves greater focus is diagnosis—the extent to which circRNAs predict AUD onset or progression and are biomarkers for treatment effectiveness. Research also is needed to understand how alcohol influences stability, regulation, localization, degradation, and modification of circRNAs. Although general models for circRNA biogenesis exist, the specific impact of alcohol on these processes remains largely unexplored. Although studies have shown associations of circRNA with AUD along with their functional outcomes, more research of individual circRNAs, but more importantly, coordinately regulated groups of circRNAs is necessary to understand their contributions to molecular and cellular processes affected by alcohol. Further, the existing research on circRNA interaction with other ncRNAs, such as miRNAs and lncRNAs, is quite limited, as is data on their interactions with proteins. Each of these areas warrants further investigation.

Discussion

Summary of Evidence

This scoping review identified 338 primary research studies addressing the contribution of ncRNAs to alcohol use and misuse and related pathologies across the lifespan, spanning tissues and organs from brain to heart, liver, gastrointestinal system, musculoskeletal system, and placenta. In contrast, a PubMed search using the terms “(Protein) AND (Alcohol Use Disorder)” recovered more than 14,900 results, suggesting that the dominant focus to date has been on protein-based mechanisms. The human genome is estimated to contain 19,411 protein-coding genes; in contrast, ncRNAs are encoded from approximately 43,675 gene loci,⁶ and protein-coding gene loci also encode ncRNAs such as circRNAs, miRNAs, and antisense RNAs. Thus, it is evident that research on ncRNA biology in AUD is in its infancy. Moreover, most research has focused on very few classes of ncRNAs. For instance, research on miRNAs accounted for approximately 86% of all published ncRNA research in AUD in this scoping review. Additionally, a majority of studies at this stage have focused on enumerating ncRNAs that are differentially regulated, and a few ncRNAs, primarily miRNAs such as members of the Let7 family, miR-9, and miR-155, have appeared repeatedly in analyses. This means that the majority of ncRNAs remain to be studied, and entire ncRNA classes such as tRNAs and tRNA fragments, snRNAs, and snoRNAs, remain largely unstudied. Significant gaps in the literature remain, particularly regarding the mechanisms by which ethanol regulates ncRNA expression at both the genomic and post-transcriptional levels. Also, increasing evidence suggests that lncRNAs often express short open reading frames and have the capacity to generate micropeptides under defined biological conditions, an outcome that is completely unstudied. Similarly, it remains unknown what other mechanisms ncRNAs mediate. It is increasingly clear that the functions of individual ncRNAs may be determined by their location within cellular compartments. When localized to the nucleus, an ncRNA may control transcription or splicing, whereas when localized to the cytoplasm, the same ncRNA may control mRNA translation. Another question is whether there are unique, nonparadigmatic mechanisms, for instance, secretion of ncRNAs into biofluids that lend themselves to not only better diagnosis, but also to intervention.

The contribution of key biological variables, such as genetic sex; race and ethnicity; and co-occurring AUD risk factors such as psychological states, environmental stress, and polysubstance use, all remain to be explored. It should be noted that about 19% of the assessed studies did not report the subjects’ genetic sex, and about 41% of studies only used males.

A majority of studies used methodologies such as PCR and microarrays that, while highly quantitative, are insensitive to transcript variations in individual ncRNAs. Only about 5% of studies used RNA sequencing-based approaches, which might yield information about individual transcript variation but were not analyzed for this purpose. Identifying sequence variations within individual ncRNAs is important because they are known to result in functional variation.

Overall, the extent to which ethanol regulates ncRNA expression and function remains incompletely understood. While some studies have documented that specific ncRNAs are modulated by ethanol exposure, the broader regulatory mechanisms and functional consequences of these changes are largely unknown. This is true particularly in the context of cellular metabolism and transcription, which require further investigation. Given the recency of this field—60% of the papers reviewed here were published from 2018 to 2023—it is clear that the immediate next steps are within the domain of the biomedical research community. However, the emerging data on ncRNAs that are secreted into biofluids suggest that their use in diagnostics and particularly, predictive modeling of disease progression is likely to have a significant impact on treatments at all stages of AUD pathology. Further in time lies the promise of ncRNAs as therapeutics. As the field of ncRNA research matures, researchers will increasingly understand the principles of ncRNA action that result from both their primary sequence and their capacity to fold and interact with other cellular components, enabling them to design novel RNA molecules. Moreover, as the roles of chaperones such as lipoproteins and microvesicles are better understood, investigators will be positioned to merge RNA design and delivery strategies to intervene in disease processes.

Limitations of the Review

The initial search strategy, including selection of search terms, may have resulted in the elimination of informative studies on alcohol effects on ncRNAs. Other equally informative papers may have also been inadvertently omitted. A unique limitation to this review is that because the field of ncRNA biology is new, terminology and gene annotations have evolved over time. For instance, early studies on miRNAs did not discriminate between the 5’ and 3’ strands of pre-miRNAs, which can generate unique miRNAs, and passenger strand miRNAs were annotated with an asterisk. While the 5’ strand is often the retained guide strand, and the 3’ strand is often the eliminated passenger strand, this is not always the case. Moreover, gene annotations have evolved over time, and with that, so have classifications of RNA transcripts as ncRNAs. The continued evolution of this field lends some degree of ambiguity to the published literature, precision of search terms, and data interpretation. Finally, the search may be subject to other publication biases, as it only includes literature published in English, potentially overlooking significant studies in other languages; additionally, biases may arise from selective publication of positive findings or limited access to unpublished studies.

Conclusions

This scoping review identified 338 research publications that linked ncRNAs to alcohol use. While some studies established associations, others established causal links. Analysis using citation impact tools such as iCite suggests that these studies have a significant and growing impact on research into alcohol use, and the translation potential of these studies is strong. Moreover, these studies link ncRNAs to every stage of alcohol use and misuse, from the induction of acute tolerance to alcohol-associated pathologies, cancer, and FASD. Nevertheless, substantial gaps remain in the knowledge base. ncRNA-coding genes outnumber protein-coding genes, and protein-coding genes can also transcribe ncRNAs. Moreover, ncRNAs can be edited, processed, transferred between sub-cellular compartments, or even secreted into extracellular spaces, where they may act as endocrine molecules. The majority of ncRNAs and their associated mechanisms remain unexplored. Finally, ncRNAs can exhibit dramatic changes with evolution and speciation, and important ncRNAs exhibit sex-specific or sex-biased patterns of expression and function. Consequently, in this field, significant attention needs to be paid to the impact of key biological variables, genetics, and species differences when investigating the linkage between ncRNAs and alcohol use and its consequences.