Autophagy and ethanol neurotoxicity

Abstract

Excessive ethanol exposure is detrimental to the brain. The developing brain is particularly vulnerable to ethanol such that prenatal ethanol exposure causes fetal alcohol spectrum disorders (FASD). Neuronal loss in the brain is the most devastating consequence and is associated with mental retardation and other behavioral deficits observed in FASD. Since alcohol consumption during pregnancy has not declined, it is imperative to elucidate the underlying mechanisms and develop effective therapeutic strategies. One cellular mechanism that acts as a protective response for the central nervous system (CNS) is autophagy. Autophagy regulates lysosomal turnover of organelles and proteins within cells, and is involved in cell differentiation, survival, metabolism, and immunity. We have recently shown that ethanol activates autophagy in the developing brain. The autophagic preconditioning alleviates ethanol-induced neuron apoptosis, whereas inhibition of autophagy potentiates ethanol-stimulated reactive oxygen species (ROS) and exacerbates ethanol-induced neuroapoptosis. The expression of genes encoding proteins required for autophagy in the CNS is developmentally regulated; their levels are much lower during an ethanol-sensitive period than during an ethanol-resistant period. Ethanol may stimulate autophagy through multiple mechanisms; these include induction of oxidative stress and endoplasmic reticulum stress, modulation of MTOR and AMPK signaling, alterations in BCL2 family proteins, and disruption of intracellular calcium (Ca2+) homeostasis. This review discusses the most recent evidence regarding the involvement of autophagy in ethanol-mediated neurotoxicity as well as the potential therapeutic approach of targeting autophagic pathways.

Keywords:

• alcohol
• alcoholism
• development
• fetal alcohol spectrum disorders
• neurodegeneration
• oxidative stress
• protein degradation

Abbreviations:

• AD, Alzheimer disease
• ALS, autophagy-lysosome system
• AMPK, adenosine 5′-monophosphate-activated protein kinase;
• ATG, autophagy-related
• CNS, central nervous system
• ER, endoplasmic reticulum
• FASD, fetal alcohol spectrum disorders
• FOXO3, forkhead box O3
• GSK3B, glycogen synthase kinase 3 β
• HD, Huntington disease, HNSCs, hippocampal neural stem cells
• LC3, microtubule-associated protein 1 light chain 3
• MTOR, mechanistic target of rapamycin (serine/threonine kinase)
• MTORC1, MTOR complex 1
• NOX, NADPH oxidase
• NFE2L2, nuclear factor, erythroid 2-like 2
• PD, Parkinson disease
• PI3K, class I phosphoinositide 3-kinase
• ROS, reactive oxygen species
• SQSTM1/p62, sequestosome 1
• TSC1/2, tuberous sclerosis 1/ 2
• UPR, unfolded protein response

Introduction

Heavy ethanol consumption produces profound damage to the brain, heart, liver, pancreas, lungs, bone, skeletal muscles, endocrine, and immune systems. The developing nervous system is particularly vulnerable to ethanol exposure. Prenatal exposure to ethanol causes fetal alcohol spectrum disorders. Fetal alcohol syndrome is the most severe form of FASD, characterized by mental retardation, malformation of the nervous system, deficits in craniofacial development, and growth deficiency.^1-5 The incidence of FASD in the United States is between 2% and 5%.³ The lifetime cost of FASD is estimated to be approximately $2 million per case, and the total annual cost of FASD was estimated at $4 billion in 2004.⁶ Neurodegeneration and the depletion of neurons in the brain, which contribute to some behavioral abnormalities observed in FASD, is the most devastating consequence of developmental exposure to ethanol.^5,7 Therefore, it is critical to elucidate the mechanisms of ethanol-induced neuroapoptosis in order to develop effective therapeutic approaches to ameliorate ethanol neurotoxicity.

Autophagy and Neuronal Survival

There are 2 major protein catabolism systems, namely, the ubiquitin-proteasome system (UPS) and the autophagy-lysosome system (ALS). Autophagy is a cellular catabolic process for amino acid/energy recycling and mitigation of metabolic stress.⁸ Although some of these functions overlap with those of the UPS, ALS mainly regulates the turnover of long-lived proteins and the maintenance of amino acid pools in the condition of cellular stresses. The ALS is able to degrade entire organelles such as mitochondria, endoplasmic reticulum (ER), and peroxisomes.⁸ This process is considered to be adaptive and plays an essential role in various cellular activities, including development, differentiation, survival, metabolism, and immunity.^9,10

Autophagy is activated by nutrient deprivation and other stress signals, such as exposure to toxic agents and metabolic stress.^11-14 Studies in the past decade have shown not only that autophagy provides cells with oxidizable substrates, but it can also offer protection against numerous pathologies such as cancer, neurodegeneration, cardiac disease, diabetes, and infections.^15,16 For the CNS, constitutive elimination of damaged proteins by basal autophagy is important, particularly to neurons.¹⁷ The constitutive activation of autophagy is essential for neuronal cell survival.¹² Disruption in the normal flow of autophagic pathways has been shown in the aging process of the brain and in some neurodegenerative diseases, including Alzheimer (AD), Parkinson (PD), and Huntington (HD) diseases.¹⁷ In some neuropathies, including AD, PD, HD, amyotrophic lateral sclerosis, and frontotemporal dementia, autophagy is beneficial due to its role in clearing toxic factors and enhancing cell survival particularly in neurons.^12,18-20 For example, impaired autophagy has been observed in several neurodegenerative diseases, such as AD models and PD, among others. Genetic activation of autophagy by overexpressing the autophagy-related (ATG) protein BECN1 promotes degradation of amyloid-β and reduces AD pathology in animal models.²¹ Of potential clinical interest is that chemical stimulation of autophagy by rapamycin, or MTOR-independent autophagy inducers, such as lithium, valproic acid, carbamazepine, rilmenidine, or trehalose, can rescue the cellular toxicity and neurodegenerative phenotype in cellular and animal models of certain neurodegenerative disorders including AD, PD, and HD.²²

The role of autophagy in regulating cell survival, however, is still under debate. Since the activation of autophagy and cell death are frequently observed occurring at the same time, some scientists suggest that autophagy may be cytotoxic.^23-25 The terminology “autophagic cell death” or “type II cell death” has been used to describe cell death caused by autophagy.^23,24 For example, insulin withdrawal causes autophagic cell death of adult hippocampal neural stem cells (HNSCs) in culture. Insulin-starved HNSCs accumulate markers of autophagy, including increases in lipidated LC3 and BECN1, and knockdown of ATG7 suppresses autophagy, reducing cell death. Conversely, activation of autophagy by rapamycin increases death of insulin-deficient cells. Furthermore, even though the HNSCs have an intact apoptotic pathway, there is no evidence of caspase activation. This suggests a causative role of autophagy in HNSC death after insulin withdrawal.²⁴ It is possible that there is a mutual regulation of autophagy and apoptosis. For example, there is some evidence showing that autophagy is required for apoptosis.²⁴ Many pro-apoptotic signals induce autophagy and apoptosis simultaneously, suggesting that cell survival is determined by the balance of protective autophagy versus the cellular accumulation of toxins, such as damaged mitochondria.^12,26 The activation of apoptosis might inhibit autophagy by impairing some essential autophagy proteins such as BECN1.^12,26 However, activation of the autophagic pathway beyond a certain threshold may be toxic due to a collapse of cellular functions and cellular atrophy. The activation of autophagy may thus have dual functions. On the one hand, autophagy is a beneficial process to cope with metabolic stresses and immunological challenges. On the other hand, sustained activation of autophagy may induce pathogenic conditions by impairing protein recycling processes and depleting critical cellular constituents, resulting in cell death.²⁷

Autophagic Protection Against Ethanol-Mediated Neuronal Damage

Since immature neurons are particularly sensitive to ethanol, a number of mechanisms have been suggested to explain ethanol-induced neurodegeneration in the developing brain. These potential mechanisms include interference with neurotrophic factor- and retinoid acid-mediated signaling, induction of oxidative stress, DNA damage and ER stress, increased proinflammatory proteins, disruption of the neurotransmitter system, and inhibition of translational regulation.^28-33 As mentioned above, autophagy can be activated by various types of cellular stress including deprivation of nutrients, hypoxia, oxidative stress, ER stress, DNA damage, and protein aggregates, as well as by the presence of intracellular pathogens.^16,34

We have recently shown that ethanol activates autophagy in neurons in vitro and in the developing brain.³⁵ Autophagy involves the nucleation of the phagophore (the initial sequestering compartment), expansion and cargo sequestration, and subsequent maturation and degradation of autophagosomes. This dynamic flow is called “autophagic flux.”³⁶ LC3 lipidation or visualization of autophagosomes by electron microscopy has been frequently used to measure autophagic activity. However, an increase in LC3-II may be caused by either a stimulation of autophagosomal formation or inhibition of autophagosomal degradation, or may result from autophagy-independent mechanisms.³⁶ Therefore, both generation and degradation of autophagosomes need to be carefully evaluated when LC3-II is used to measure autophagy activity. Ethanol increases LC3 lipidation (Fig. 1), and reduces SQSTM1/p62 levels (Fig. 2) in neuronal cells, which was accompanied by a significant increase of LC3 puncta (Fig. 3). At both 0.4% and 0.8%, ethanol is effective, but at 0.8% it produces a greater effect.³⁵ We therefore present the data using 0.8% to better illustrate the effect of ethanol, although we note that for in vitro studies, a concentration of 0.4% ethanol is considered physiologically relevant to heavy alcohol exposure in humans.³⁷ SQSTM1 levels are controlled by autophagy-dependent degradation. Bafilomycin A₁, an inhibitor of the fusion of autophagosomes with the lysosome, prevents SQSTM1 degradation, further increasing the level of this protein (Fig. 2). Rapamycin, an activator of autophagy, enhances ethanol-mediated upregulation of LC3-lI Levels and the formation of LC3 puncta (Figs. 1 and 3). In contrast, wortmannin, an inhibitor of initiation/maturation of autophagosomes, and shRNA for BECN1 block ethanol-increased LC3-II levels and LC3 puncta (Figs. 1 and 3). Bafilomycin A₁ alone also drastically increases LC3 puncta (Fig. 3). Taken together, these results indicate that ethanol enhances the development of autophagosomes rather than inhibiting autophagosome-lysosome fusion. Therefore, ethanol promotes autophagic flux in neurons.

Figure 1. Ethanol increases LC3 lipidation. SH-SY5Y neuroblastoma cells were treated with or without ethanol (0.8%) in the presence/absence of wortmannin (Wort: 10 μM), bafilomycin A₁ (Baf: 10 nM), rapamycin (Rap: 10 nM) or BECN1 shRNA (BECN1). The protein samples were collected 8 h after the treatment. The levels of LC3 were examined with immunoblotting (top panel). Relative levels of LC3-II were quantified by densitometry and normalized to actin levels (bottom panel). * P < 0.05, **P < 0.01. (from ref. 35 with permission of the publisher).

Figure 2. Ethanol decreases SQSTM1 levels. Notations are the same as in Fig. 1 (from ref. 35 with permission of the publisher).

Figure 3. Ethanol increases LC3 puncta. (A) SH-SY5Y cells were transfected with a GFP-LC3 plasmid and exposed to ethanol (0.8%) in the presence/absence of bafilomycin A₁ (Baf: 10 nM) or wortmannin (Wort: 10 μM) for 6 h. In some experimental groups, cells were cotransfected with BECN1 shRNA (BEC). The formation of GFP-LC3 puncta was examined under a fluorescence microscope 6 h after ethanol exposure. (B) GFP-LC3 puncta/cell was quantified. * P < 0.05. (from ref. 35 with permission of the publisher).

Ethanol causes damage to mitochondria resulting in the production of intracellular reactive oxygen species.^38-40 Because mitochondria are essential in regulating apoptosis, elimination of defective mitochondria by selective autophagy (mitophagy) is one of the mechanisms by which autophagy offers protection,^11,41 and ethanol increases mitophagy.³⁵ Ethanol-induced mitophagy is inhibited by BECN1 shRNA while it is increased by rapamycin. Rapamycin alleviated ethanol-mediated ROS accumulation and attenuates ethanol-induced neuroapoptosis in cultured neurons and in the developing brain, whereas wortmannin or knocking down BECN1 promotes ethanol-stimulated ROS accumulation and potentiated ethanol-induce neuroapoptosis. These results agree with a previous finding showing that in the context of liver cells, removal of defective mitochondria by autophagy is protective against mitochondrial release of pro-apoptotic factors and accumulation of toxic ROS.^42,43 Removing damaged mitochondria by autophagy not only promotes cell survival, but also may increase the threshold for the stress signals that are pro-apoptotic. Taken together, these findings indicate that neurons increase autophagic flux in response to ethanol neurotoxicity; enhanced autophagy is a protective mechanism that alleviates ethanol-mediated oxidative stress and neuronal death.

The protective role of autophagy against ethanol neurotoxicity is further supported by a study of hypoxic preconditioning. A modest hypoxic preconditioning (1% oxygen) can induce autophagy without affecting the viability of neuronal cells. This hypoxic preconditioning protects neuronal cells against ethanol-induced cell death.⁴⁴ Inhibition of autophagy blocks hypoxic preconditioning-mediated protection, whereas activation of autophagy enhances neuroprotection caused by hypoxic preconditioning.⁴¹

During development, there is a temporal susceptibility to ethanol neurotoxicity. Using a mouse model of postnatal ethanol exposure, which is equivalent to human third trimester, it was found that ethanol treatment induces widespread neuroapoptosis in the brain of C57BL/6 mice at postnatal d 4 (PD4), but had little effect on PD12 mice.⁴⁵ We studied the expression pattern of genes involved in apoptosis, the pathways of the unfolded protein response (UPR) and autophagy in PD4 and PD12 mice. The expression of CASP3/caspase 3, a pro-apoptotic protein, is much higher in the ethanol-sensitive period, PD4, than in ethanol-resistant periods, corresponding to PD12; in contrast, the expression of genes related to UPR and autophagy are much higher on PD12 than PD4; these genes include Atg4, Atf6, Atg9, Atg10, Becn1, Bnip3, Cebpb, Ctsb, Ctsd, Ctss, Eif2ak3, Hspa5/Grp78, Ern1/Ire1α, Lamp1, Pik3c3, and Sqstm1.⁴² The data suggest that the susceptibility of the developing nervous system to ethanol may be due to high expression of pro-apoptotic proteins and an immaturity of the stress mitigation systems such as UPR and autophagy.

Short-term ethanol exposure induces a protective response of upregulation of autophagy in the developing brain. However, chronic ethanol exposure in adult mice activates MTOR and causes impairment of the autophagy-lysosome pathway in the brain.⁴⁶ This implies that long-term exposure to ethanol may damage this protective system and further jeopardize CNS neurons to subsequent ethanol exposure and other insults. In turn, this finding suggests that upregulation of autophagy may be beneficial not only for the damage caused by acute ethanol exposure in the developing brain but also for deficits in the adult brain induced by chronic exposure.

Indeed, the activation of autophagy protects against ethanol-induced liver damage. Autophagy is essential for liver health, and induction of autophagy is viewed as a promising approach for treating liver injury.^47-49 Ethanol induces autophagy in liver cells; suppression of autophagy exacerbates alcohol-induced apoptosis of liver cells.^50,51 Conversely, pharmacological promotion of autophagy alleviates steatosis and liver injury caused by ethanol exposure.^52,53 Therefore, in both the liver and developing brain, autophagy is a protective mechanism against alcohol-induced cellular damage. However, in some model systems, the activation of autophagy seems detrimental. For example, autophagy seems to have a negative effect in ethanol-induced cardiac defects and apoptosis because the activation of autophagy by rapamycin and AICAR exacerbates ethanol-induced cell damage in cardiomyocytes.⁵⁴ Patients or animals with alcoholic cirrhosis and hepatitis have severe muscle loss due to the degeneration of skeletal muscle protein. Furthermore, the degeneration of skeletal muscle protein and loss of skeletal muscle mass is mediated by ethanol-induced autophagy.⁵⁵ These examples suggest that the activation of autophagy may also contribute to the adverse effect of ethanol exposure in some situations.

Mechanisms Underlying Ethanol Activation of Autophagy in Neurons

Many cellular mechanisms potentially contribute to ethanol-induced autophagy. Ethanol directly and indirectly induces many varied cellular consequences on factors such as ROS generation, MTOR and AMPK signaling, anti-apoptotic protein expression, and intracellular Ca²⁺ homeostasis (Fig. 4). Autophagy is a complex process that can be better understood by determining the altered roles of these various cellular factors. More specifically, through understanding and monitoring representative biological pathways the function of autophagy in the context of neurotoxicity can be better understood. These potential mechanisms are further addressed in the following section.

Figure 4. Autophagy as a potential protective response to ethanol neurotoxicity. Ethanol induces oxidative stress and ER stress in neurons of the developing brain, which results in neurodegeneration. Autophagy is activated in response to ethanol neurotoxicity. Several potential mechanisms contribute to ethanol-induced autophagy; these include the modulation of MTOR, PI3K-AKT, or AMPK signaling, alterations in BCL2 family proteins and disruption of intracellular calcium homeostasis. Ethanol-induced oxidative stress and ER stress may also trigger autophagy. There is considerable crosstalk among these signaling pathways during the activation of autophagy.

Oxidative stress

It is well established that the generation of reactive oxygen species, redox signaling and oxidative stress all stimulate autophagy.^56,57 ROS may mediate the balance between autophagy and apoptosis.⁵⁶ Free radicals regulate the activities of transcriptional factors, such as NFKB (nuclear factor of kappa light polypeptide gene enhancer in B-cells), TP53, and NFE2L2 (nuclear factor, erythroid 2-like 2), and affect various protein kinase cascades that are involved in the interaction between autophagy and apoptosis. Ethanol oxidation by ADH (alcohol dehydrogenase) and CYP2E1 (cytochrome P450, family 2, subfamily E, polypeptide1) produces ROS and causes oxidative stress in cells. In HepG2 liver cells, both ADH-generated acetaldehyde and CYP2E1-generated oxidants activates autophagy.⁵⁸

In neurons, ethanol-induced ROS mainly come from damaged mitochondria and the activation of NOX (NADPH oxidase).^38-40,59 NOX-derived production of ROS activates autophagy.⁵⁷ Antioxidants partially block ethanol-induced autophagic flux and mitophagy, suggesting that ROS is involved in ethanol activation of autophagy.³⁵

ER stress

The ER lumen is loaded with polypeptides at different stages of assembly. ER quality control systems monitor incorrect protein folding, stochastic errors, and off-pathway intermediates. As the accumulation of unfolded or misfolded proteins in the ER lumen overflows the capacity of the ER chaperone system, ER stress or UPR occurs. Induction of the UPR attempts to eliminate unfolded or misfolded proteins so that ER stress can be alleviated. The UPR is orchestrated by 3 ER transmembrane receptors, namely, EIF2AK3 (eukaryotic translation initiation factor 2-α kinase 3), ATF6 (activating transcription factor 6), and ERN1 (endoplasmic reticulum to nucleus signaling 1). In resting cells, all 3 UPR receptors are kept inactive through their association with an ER chaperone called HSPA5/Grp78 (heat shock 70 kDa protein 5 [glucose-regulated protein, 78 kDa]).

It is now known that ER stress can activate autophagy and there is considerable crosstalk between the UPR and autophagy.^52,61,62 There are multiple mechanisms underlying ER stress-induced autophagy. ER stress can activate autophagy by disrupting intracellular Ca²⁺ homeostasis or activating kinases and transcription factors that regulate autophagy; these include ERN1, MAPK/JNK, DDIT4/REDD1, AKT, DDIT3/CHOP, XBP1, DAPK1, and NFE2L1/NFE2L2.^52,61,62 Since ethanol induces ER stress in cultured neurons and in the brain,^31,33 it may trigger autophagy by inducing ER stress in neurons.

Class I PI3K and MTORC1

The pathways associated with the regulation of autophagy are very complex, involving more than 34 known proteins that assemble the autophagic machinery.⁶³ The best characterized regulator for autophagy is the mechanistic target of rapamycin complex 1 (MTORC1). MTORC1 is a protein complex that consists of a serine/threonine kinase, MTOR, a regulatory associated partner of MTOR (RPTOR) and MLST8 (MTOR associated protein, LST8 homolog [S. cerevisiae]). MTORC1 phosphorylates its substrates, EIF4EBP1 (eukaryotic translation initiation factor 4E binding protein 1) and the serine/threonine kinase RPS6KB/p70S6K (ribosomal protein S6 kinase, 70 kDa), and regulates their activity.⁶⁴ MTORC1 is regulated by the class I phosphoinositide 3-kinase (PI3K). PI3K phosphorylates AKT1 (v-akt murine thymoma viral oncogene homolog 1), which activates TSC1 and TSC2 (tuberous sclerosis 1/2). Other kinases, such as MAPK1/ERK2 (mitogen-activated protein kinase 1), MAPK3/ERK1, RSK1 (ribosomal protein S6 kinase, 90kDa, polypeptide 1), AMPK (adenosine 5’-monophosphate-activated protein kinase) and GSK3B (glycogen synthase kinase 3 β) also regulate the TSC1-TSC2 complex.⁶⁴ TSC1 and TSC2 form a functional complex with MTORC1 and negatively regulate its activity.⁶⁴ MTORC1 regulates autophagy through the modulation of specific proteins controlling the autophagic machinery; MTORC1 phosphorylates the mammalian homolog of yeast Atg13 (ATG13) and the mammalian Atg1 homologs ULK1/2 (unc-51 like autophagy activating kinase 1/2), inhibiting the activity of the ATG13-ULK1/2-RB1CC1/FIP200-ATG101 complex and preventing the activation of autophagy.^22,65 However, some evidence suggests that MTORC2 may also participate in the regulation of autophagy.⁶⁶

Ethanol inhibits the phosphorylation of EIF4EBP1 and RPS6KA, 2 main substrates of MTORC1, indicating that ethanol suppresses MTORC1 activity.³⁵ Rapamycin, a well-known inhibitor of MTORC1, and ethanol together produce greater dephosphorylation of RPS6KA and EIF4EBP1 than when cells are treated with either rapamycin or ethanol alone; this result is consistent with the finding that rapamycin potentiates ethanol-induced autophagy. Both rapamycin and the knockdown of MTOR by siRNA protect neurons against ethanol-induced apoptosis, further verifying the importance of the MTORC1 pathway in ethanol activation of autophagy and neurotoxicity.³⁵ Ethanol may regulate autophagy by MTORC1-dependent or -independent pathways. Further study is necessary to clearly establish whether ethanol-induced autophagy is mediated through the MTORC1 pathway; this can be achieved by determining whether a constitutive MTORC1 activation blocks ethanol-induced autophagy.

It has been previously shown that ethanol modulates MTORC1 activity in the brain.⁶⁷ Other evidence shows that ethanol suppresses MTOC1 activity in hepatocytes, skeletal muscles, and myocytes.^52,68-70 Ethanol-induced inhibition of MTORC1 activity in liver cells may be mediated by ROS.⁵² However, chronic exposure to ethanol upregulates MTOR phosphorylation, leading to an inhibition of autophagic pathways in the brain.⁴⁶ The mechanisms of ethanol inhibition of MTORC1 are unclear. Ethanol induces ROS production and affects several upstream kinases of MTORC1, such as PI3K-AKT and GSK3B in neurons.^30,71-73 These potential mechanisms, particularly how ethanol-induced oxidative stress regulates MTORC1 signaling and autophagy, are worth further investigation.

AMPK

AMPK is a heterotrimeric protein kinase and is activated by alterations in the cellular AMP-to-ATP ratio. Upon activation, AMPK inhibits energy-requiring cellular activities and meanwhile stimulates catabolic pathways. AMPK plays a key role in cell survival during metabolic stress by maintaining metabolic homeostasis.⁷⁴ Recent studies indicate that AMPK plays a critical role in regulating autophagy in neurons.^75,76 AMPK is upstream of MTORC1 and the activation of AMPK inhibits MTORC1 activity through the phosphorylation of the MTORC1-associated proteins RPTOR and TSC2, thereby activating autophagy.²² AMPK is activated by oxidative stress and also controls the redox-state and mitochondrial function.^74,77 AMPK is required for ROS-stimulated autophagy, which enhances cell survival in response to nutrient starvation, hypoxia, or ischemia.^56,77 Although it is commonly thought that activation of AMPK induces autophagy, some studies show that inhibition of AMPK stimulates autophagy.⁷⁸ Ethanol has been shown to modulate AMPK in neuronal cells as well as other cell types in vitro and in vivo.^74,79-83 Therefore, ethanol may activate autophagy by regulating AMPK activity in neurons.

In addition to inhibiting MTORC1 activity, AMPK regulates the activity of several transcription factors, such as CREB (cAMP response element binding protein), PPARG (peroxisome proliferator-activated receptor gamma), PPARA, and the FOXO (forkhead box O) family of transcription factors, as well as and NFE2L2.⁸⁴ Among these transcription factors, FOXO3 and NFE2L2 induce expression of genes regulating autophagy.^62,84 FOXO3 mediates ethanol-induced autophagy in primary hepatocytes and in the liver.⁸⁵ Overexpression of a dominant negative form of FOXO3 inhibits ethanol-induced autophagy-associated gene expression and enhances ethanol-induced cell death of hepatocytes. Similarly, the induction of FOXO3 nuclear localization activates autophagy and offers protection against ethanol-induced death of hepatocytes.⁸³ There is considerable interaction between ethanol and NFE2L2 in neuronal cells and other cell types.^86-89 Both FOXO3 and NFE2L2 are also regulated by oxidative stress.^62,90 Therefore, ethanol may also activate autophagy by modulating the activity of FOXO3 and NFE2L2.

BCL2 family proteins

It is well known that proteins of the BCL2 family play a critical role in the regulation of cell death/survival. Recent studies indicate that BCL2 family proteins have both a positive and negative regulatory role in autophagy.^91,92 Anti-apoptotic BCL2 family proteins, such as BCL2, BCL2L1/BCL^-xL, and MCL1, inhibit autophagy by a direct association with BECN1. It seems that only the endoplasmic reticulum-BCL2 protein inhibits autophagy.^93,94 In addition to binding with BECN1, BCL2, and BCL2L1 also inhibit autophagy by modulating ER Ca²⁺ homeostasis, which plays an important role in the initiation of autophagy.^95,96 In contrast to anti-apoptotic BCL2 family proteins, pro-apoptotic BH3-only proteins, such as BNIP3, BNIP3L/NIX, BAD, BIK, PMAIP1/NOXA, BBC3/PUMA, and BCL2L11/BIM, generally induce autophagy through their binding to BCL2 and subsequent release of the inhibitory BCL2 from BECN1. However, the role of pro-apoptotic proteins BAX and BAK1 in autophagy is controversial. Both positive and negative regulation of autophagy by BAX and BAK1 have been reported.⁵²

Ethanol modulates BCL2 family proteins in neurons in vitro and in vivo.^73,97-99 Generally, ethanol inhibits the expression of anti-apoptotic BCL2 family proteins, but increases the expression of pro-apoptotic proteins, which may result in the activation of autophagy.

Calcium signaling

Intracellular Ca²⁺ is a ubiquitous second messenger regulating diverse cellular and physiological processes such as differentiation, survival, development, proliferation, muscle contraction, and secretion. Alterations in intracellular Ca²⁺ homeostasis could result in neuroapoptosis in the CNS. Homeostasis of intracellular Ca²⁺ is achieved in part by the buffering role of mitochondria and the ER. Recent studies indicate that intracellular Ca²⁺ signaling is involved in autophagy.^100,101 Although it has been suggested that autophagy may be activated by an increase in cytosolic Ca²⁺ concentration or a perturbed release from intra- or extracellular stores,¹⁰⁰ more compelling evidence has indicated that an increase in intracellular Ca²⁺ inhibits autophagic flux.^102-104 Both mitochondria and the ER play a critical role in Ca²⁺ homeostasis and regulation of autophagy.^96,100 Ethanol can damage mitochondria and cause ER stress, which may disrupt Ca²⁺ homeostasis.^31,38,39,105 Ethanol also disturbs intracellular Ca²⁺ homeostasis in neurons in vitro and in vivo.^106,107 Therefore, it is possible that ethanol may regulate autophagy by disrupting intracellular Ca²⁺ homeostasis in neurons.

Autophagy as a Potential Therapeutic Target for Treating Ethanol Neurotoxicity

Since the activation of autophagy appears protective against ethanol neurotoxicity, pharmacological manipulation of autophagy should be considered a method of potential therapeutic approaches. The most important and well-studied drug identified as an autophagy inducer is rapamycin, an MTORC1 inhibitor. Rapamycin is a FDA-approved drug and a widely used immune suppressor in clinical transplant patients. Rapamycin and its analogs (CCI-779, RAD001 and AP23573) have been used to activate autophagy and offer protection in a number of neurodegenerative models.^22,108 We have recently shown that rapamycin protects neurons against thiamine deficiency-induced neurodegeneration¹⁰⁹; in vitro studies indicate that rapamycin alleviates ethanol-induced death of neuronal cells by activating autophagy.^35,44 Similarly, rapamycin mitigates ethanol-induced liver pathogenesis.⁵² In an animal model of ethanol-induced cardiac defect (in which autophagy plays a negative role), however, rapamycin exacerbates ethanol-induced cell injury in cardiomyocytes.⁹⁷ In addition to directly targeting MTOR, activation of autophagy by manipulating other pathways upstream of MTORC1, such as AMPK, PI3K-AKT, GSK3B, or TSC1/2 are alternative approaches.

Although rapamycin has been used clinically for immune suppression, it is unclear how it will affect the human CNS. In addition to the induction of autophagy, rapamycin may have potential side effects because MTORC1 signaling affects multiple pathways, such as protein translation, cell growth, and metabolism.^110,111 Chronic stimulation of autophagy may impose additional risks, because sustained autophagy may result in depletion of healthy mitochondria, causing neuronal death. The discovery of MTOR-independent autophagy enhancers may thus provide an alternative method for stimulating this process. A number of studies of chemical screening have identified chemicals and small molecule enhancers that regulate autophagy independent of MTOR, including some FDA-approved drugs.²² It is worthwhile to test these chemicals for their induction of autophagy in the CNS and potential neuroprotection against ethanol neurotoxicity. For example, various mood-stabilizing drugs such as lithium, carbamazepine, L-690,330, and sodium valproate, could activate autophagy through the suppression of IMPA (inositol monophosphatase).^63,109 Lithium-induced autophagy seems to mediate its protective function in the neurodegeneration see in the PD model.^112,113 In fact, lithium alleviates ethanol-induced neuroapoptosis in cultured neurons and developing brain.^7,32,73 Although these drugs may be beneficial by activating autophagy in the brain, the potential teratogenic effects need to be carefully evaluated. GSK3B is an important regulator of brain development and these drugs alter the activity of this kinase.¹¹⁴

A potentially more effective approach to induce autophagy is the application of a cocktail of autophagy activators. Combination therapy for enhancing autophagy with lower concentrations of each chemical inducer may be safer for chronic drug administration and may lessen side effects. Additive effects in enhancing autophagy can be achieved through simultaneous stimulation by MTOR-dependent and MTOR-independent routes. For example, combined treatment with lithium and rapamycin causes greater induction of autophagy compared to the effects of each individual compound.²² Similar additive effects for activating autophagy have been shown by using rapamycin with other MTOR-independent autophagy enhancers, such as trehalose, calpastatin, or small molecule enhancers.²²

Future Directions

Several critical questions remain regarding the role of autophagy in ethanol neurotoxicity. These include: 1) During brain development, there is a spatiotemporal susceptibility to ethanol neurotoxicity. Is there a correlation between this selective vulnerability and the status of autophagy activation? What are the factors that mediate the region and cell type specificity of autophagic activation in the brain? 2) What are the primary molecular targets of ethanol that initiate autophagy? Are these targets region-specific and developmentally regulated in the brain? 3) Recent attempts targeting autophagy in the CNS bring tremendous excitement for the development of novel therapeutic approaches. What is the most effective paradigm of manipulation of autophagy in the brain? How can the potential for detrimental side effects be limited? What combination of autophagy inducers will be most effective? Taken together, the studies carried out to date suggest that autophagy will undoubtedly be an attractive target for future therapeutic efforts. A better understanding of neuronal autophagy shows promise in providing avenues for future therapy against ethanol neurotoxicity.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgment

I would like to thank Ashley Comer and Jacqueline Frank for reading this manuscript.

Funding

This research is supported by grants from National Institutes of Health (NIH) (AA019693 and AA015407).