Mitochondrial depolarization after acute ethanol treatment drives mitophagy in living mice

ABSTRACT

Ethanol increases hepatic mitophagy driven by unknown mechanisms. Type 1 mitophagy sequesters polarized mitochondria for nutrient recovery and cytoplasmic remodeling. In Type 2, mitochondrial depolarization (mtDepo) initiates mitophagy to remove the damaged organelles. Previously, we showed that acute ethanol administration produces reversible hepatic mtDepo. Here, we tested the hypothesis that ethanol-induced mtDepo initiates Type 2 mitophagy. GFP-LC3 transgenic mice were gavaged with ethanol (2–6 g/kg) with and without pre-treatment with agents that decrease or increase mtDepo–Alda-1, tacrolimus, or disulfiram. Without ethanol, virtually all hepatocytes contained polarized mitochondria with infrequent autophagic GFP-LC3 puncta visualized by intravital microscopy. At ~4 h after ethanol treatment, mtDepo occurred in an all-or-none fashion within individual hepatocytes, which increased dose dependently. GFP-LC3 puncta increased in parallel, predominantly in hepatocytes with mtDepo. Mitochondrial PINK1 and PRKN/parkin also increased. After covalent labeling of mitochondria with MitoTracker Red (MTR), GFP-LC3 puncta encircled MTR-labeled mitochondria after ethanol treatment, directly demonstrating mitophagy. GFP-LC3 puncta did not associate with fat droplets visualized with BODIPY558/568, indicating that increased autophagy was not due to lipophagy. Before ethanol administration, rhodamine-dextran (RhDex)-labeled lysosomes showed little association with GFP-LC3. After ethanol treatment, TFEB (transcription factor EB) translocated to nuclei, and lysosomal mass increased. Many GFP-LC3 puncta merged with RhDex-labeled lysosomes, showing autophagosomal processing into lysosomes. After ethanol treatment, disulfiram increased, whereas Alda-1 and tacrolimus decreased mtDepo, and mitophagy changed proportionately. In conclusion, mtDepo after acute ethanol treatment induces mitophagic sequestration and subsequent lysosomal processing.

Abbreviations : AcAld, acetaldehyde; ADH, alcohol dehydrogenase; ALDH, aldehyde dehydrogenase; ALD, alcoholic liver disease; Alda-1, N-(1,3-benzodioxol-5-ylmethyl)-2,6-dichlorobenzamide; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GFP, green fluorescent protein; LAMP1, lysosomal-associated membrane protein 1; LMNB1, lamin B1; MAA, malondialdehyde-acetaldehyde adducts; MAP1LC3/LC3, microtubule-associated protein 1 light chain 3; MPT, mitochondrial permeability transition; mtDAMPS, mitochondrial damage-associated molecular patterns; mtDepo, mitochondrial depolarization; mtDNA, mitochondrial DNA; MTR, MitoTracker Red; PI, propidium iodide; PINK1, PTEN induced putative kinase 1; PRKN, parkin; RhDex, rhodamine dextran; TFEB, transcription factor EB; Tg, transgenic; TMRM, tetramethylrhodamine methylester; TOMM20, translocase of outer mitochondrial membrane 20; VDAC, voltage-dependent anion channel

INTRODUCTION

Alcoholic liver disease (ALD) is one of the most prevalent liver diseases worldwide [1,2]. Alcohol injures many organs; however, as the primary site of alcohol metabolism, the liver experiences the most severe tissue injury caused by heavy drinking. Alcohol-induced pathological changes in the liver include steatosis, hepatitis, and fibrosis/cirrhosis, which typically co-exist and eventually progress to end-stage liver disease and, in many instances, liver cancer [2–4]. Fibrosis/cirrhosis develops after years and even decades of heavy alcohol drinking, but steatosis occurs quickly after binge drinking [5]. Overall, about 35% of heavy drinkers develop advanced liver disease [5]. ALD is responsible for almost half of global mortality due to cirrhosis [1,6]. Effective therapies for ALD are lacking except cessation of drinking, which is often difficult to achieve in alcohol-addicted individuals.

Although the pathogenesis of alcohol-induced liver injury remains incompletely understood, it is widely recognized that ALD is mediated by multiple-hit mechanisms [7]. Potential mechanisms of pathogenesis include oxidative/nitrative stresses, methionine metabolism perturbation, mitochondrial alterations, endoplasmic reticulum stress, gut microbiome-liver interactions, adipose tissue dysfunction, inflammatory cytokines, genetic polyphorphisms, and epigenetic changes [5,7–11].

Mitochondrial dysfunction appears to play an important role in development of ALD [4,12,13]. Mitochondrial changes such as megamitochondria and mitochondrial DNA deletion are common findings in ALD patients, as well as in ethanol-treated animals [4,14,15]. Exposure to ethanol leads to an adaptive increase of ethanol metabolism accompanied by a mitochondrial respiratory burst in both rodent and human livers [16]. Although increased respiration should in theory lead to increased ATP generation, ethanol treatment actually decreases hepatic ATP, suggesting uncoupling [17–20].

Many studies also observe increased mitochondrial autophagy (mitophagy) after ethanol treatment [21–24]. Type 1 mitophagy removes functional mitochondria to provide metabolic precursors during nutrient deprivation or to remove mitochondria in excess of metabolic needs [25–27]. By contrast in Type 2 mitophagy, mitochondrial depolarization (mtDepo) initiates autophagic sequestration [25]. After mtDepo, PINK1 (PTEN induced putative kinase 1) accumulates on mitochondria to promote PRKN (parkin) binding, ubiquitination of outer membrane proteins, and subsequent autophagosome formation [25]. In this way, Type 2 mitophagy removes potentially harm-inducing dysfunctional/damaged mitochondria. Although previous studies show increased mitophagy after ethanol treatment, the mechanism driving mitophagy after ethanol exposure and the type of mitophagy that occurs remain unclear. Previously, we showed widespread mtDepo occurring after acute and chronic ethanol treatment due to acetaldehyde (AcAld) production during ethanol metabolism [13,19]. Therefore, in this study, we tested the hypothesis that mtDepo drives hepatocellular mitophagy after acute ethanol treatment.

RESULTS

Acute ethanol treatment causes hepatic mitochondrial depolarization and increases GFP-LC3 puncta formation in a dose-dependent fashion

Autophagy increases after ethanol treatment by an unknown mechanism [21,22,28]. Whether ethanol-induced autophagy is linked to mtDepo remains unclear. Since ethanol-induced mtDepo is an in vivo phenomenon and does not occur in vitro, we examined mitochondrial polarization status after acute ethanol treatment in living green fluorescent protein (GFP)-MAP1LC3/LC3 (microtubule-associated protein 1 light chain 3) transgenic (Tg) mice by intravital multiphoton microscopy. Polarization status was monitored with the red cationic fluorophore tetramethylrhodamine methylester (TMRM), which labels negatively polarized mitochondria in proportion to the mitochondrial membrane potential [29,30]. In control mice, TMRM fluorescence was punctate in virtually all hepatocytes, indicating mitochondrial polarization (Figure 1A, left). At about 4 h after acute ethanol treatment, mitochondria in many hepatocytes became depolarized, as shown by loss of TMRM fluorescence (Figure 1B, left, group of cells in middle and upper left), whereas mitochondria in other hepatocytes remained polarized, as shown by retention of TMRM puncta (Figure 1B, ring of TMRM-labeled hepatocytes surrounding poorly TMRM-labeled hepatocytes), which is consistent with our previous observation [19]. For individual hepatocytes, mtDepo after ethanol treatment almost always occurred in a virtually all-or-none manner. In mice treated with 2 g/kg of ethanol, mtDepo occurred in 41% of hepatocytes (p < 0.01 vs. vehicle, Fig. 1 and 2). As the ethanol dose increased, mtDepo progressively increased to a maximum of 98% of hepatocytes after 6 g/kg (Figure 2A). These results showed that ethanol treatment caused mtDepo in a dose-dependent fashion, as expected.

Figure 1.

Increased hepatic mitochondrial depolarization and GFP-LC3 puncta formation in vivo after acute ethanol treatment. GFP-LC3 Tg mice were gavaged with saline (A) or ethanol (2 g/kg, i.g., B). TMRM (1.1 µmol/mouse) was infused arterially at ~4 h after ethanol treatment, as described in Materials and Methods. Intravital multiphoton microscopy of TMRM (red fluorescence, left panels) and GFP-LC3 (green fluorescence, middle panels) was performed using a 60x silicone oil objective lens ~10 min later. Right panels are overlays of TMRM and GFP-LC3 images. Bar: 10 µm. Shown are representative images from 3–4 mice/group. The inset in B shows part of the field at higher magnification. The dotted line is the border between a hepatocyte with mtDepo (above) and a hepatocyte without mtDepo (below). Arrows point to presumptive phagophore assembly sites. An asterisk (*) overlies an area where all mitochondria are depolarized.

Figure 2.

GFP-LC3 puncta formation increases with hepatic mitochondrial depolarization and occurs principally in hepatocytes with depolarized mitochondria. GFP-LC3 Tg mice were gavaged with one dose of ethanol (2–6 g/kg) or equal volume of saline, and mitochondrial polarization status was detected by intravital multiphoton microscopy of TMRM at ~4 h after treatment, as described in Figure 1. (A), the numbers of hepatocytes with mtDepo were counted in ~10 random fields per mouse, and the percentage of hepatocytes with mtDepo was plotted. (B), GFP-LC3 puncta in all hepatocytes (both with and without mtDepo) were counted in 10 random fields per mouse, and average numbers of GFP-LC3 puncta per hepatocyte were calculated and plotted. (C), numbers of GFP-LC3 puncta were counted separately for hepatocytes with polarized and depolarized mitochondria in 10 random fields per mouse. Values are means ± SEM (n = 3–4 per group). **, p < 0.01 vs no ethanol. a, p < 0.05 vs no ethanol; b, p < 0.05 vs hepatocytes with polarized mitochondria; c, p < 0.05 vs the pooled ethanol-treated polarized group.

In livers of vehicle-treated mice, small numbers of green GFP-LC3 puncta were observed in hepatocytes (1.4 puncta/cell, Figure 1A, middle panel, and Figure 2B). After ethanol treatment, GFP-LC3 puncta markedly increased in hepatocytes, as averaged for all hepatocytes both with and without mtDepo (Figure 1B, middle panel). In mice treated with 2 g/kg ethanol, average GFP-LC3 puncta increased to 7.3 per cell (Figure 2B; p < 0.01 vs vehicle). As the ethanol dose increased, GFP-LC3 puncta progressively increased to 12.6/cell after 6 g/kg (Figure 2B). Nuclei characteristically showed diffuse GFP-LC3 fluorescence, which was not different between vehicle- and ethanol-treated mice. These results showed that ethanol treatment increased GFP-LC3 puncta in a dose-dependent fashion in parallel to the extent of mtDepo.

In hepatocytes with mtDepo after ethanol treatment, puncta sometimes appeared adjacent to mitochondria that were still polarized to some extent as shown by retention of TMRM (Figure 1). Likewise in hepatocytes without mtDepo after ethanol treatment, GFP-LC3 puncta often were immediately adjacent to mitochondria. Since the TMRM-labeled mitochondria were not inside these puncta, the puncta may represent phagophore assembly sites rather than mitophagosomes [31].

LC3 puncta form primarily in hepatocytes with depolarized mitochondria

To explore further the relationship between mtDepo and LC3 expression, we calculated the average number of GFP-LC3 puncta in hepatocytes with mtDepo in comparison to hepatocytes without mtDepo (Figure 2C). In livers of vehicle-treated mice, hepatocytes with mtDepo were rare (<2%, Figure 2A). GFP-LC3 puncta in hepatocytes with polarized and depolarized mitochondria were ~1.3/cell and 2.6/cell, respectively (n.s.). In livers of mice receiving 2, 4 and 6 g/kg of ethanol, GFP-LC3 puncta in cells with polarized mitochondria were 3.8 to 4.5/cell, which was not statistically different from cells with polarized mitochondria in vehicle-treated mice (Figure 2C). However, when all ethanol-treated groups (2–6 g/kg) were pooled together, the difference for cells with polarized mitochondria between the pooled ethanol-treated groups vs vehicle-treated mice became statistically significant (Figure 2C, P < 0.05), indicating a small increase of autophagy after ethanol treatment in hepatocytes with polarized mitochondria.

By contrast, GFP-LC3 puncta after ethanol treatment in hepatocytes with mtDepo increased to 12.6 to 12.8/cell over a dose range of 2 to 6 g/kg. In cells without mtDepo from the same ethanol-treated animals, GFP-LC3 puncta averaged 3.8 to 4.5/cell, as described above, which was substantially less than hepatocytes with mtDepo (p < 0.001). These data demonstrated that GFP-LC3 puncta developed principally in hepatocytes with mtDepo. The results also indicated that the increase of total GFP-LC3 puncta in the liver with increasing ethanol dosage was due to recruitment of more and more hepatocytes to mtDepo as ethanol increased.

PINK1 and PRKN translocate to mitochondria after acute ethanol treatment

In vitro, mtDepo causes accumulation of PINK1, a protein kinase that mediates mitophagy, on the surface of mitochondria [32]. Accordingly, we examined if PINK1 accumulates in vivo after acute ethanol treatment by immunoblotting. In liver extracts from vehicle-treated mice, a weak band of PINK1 was detected (Figure 3A and D), consistent with previous reports [33–35]. After ethanol treatment (2, 4, and 6 g/kg), total liver PINK1 increased in a dose-dependent fashion (Figure 3A and D). After ethanol treatment at 4 g/kg, PINK1 increased ~170% but was not significantly altered in the cytosolic fraction (Figure 3B, C, E, and F), indicating that PINK1 accumulated in mitochondria.

Figure 3.

Increased mitochondrial PINK1 and PRKN after acute ethanol. PINK1, PRKN, and proteins of housekeeping genes Tomm20 and Gapdh were detected by immunoblotting. (A, D, and G), C57Bl/6 mice were treated with saline or ethanol (2, 4, and 6 g/kg, i.g.). (B, C, E, F, H, and I), mice were treated with saline or ethanol (4 g/kg, i.g.). (A), representative image of immunoblots of PINK1, PRKN, and GAPDH in liver extracts; (B), representative image of immunoblots of PINK1, PRKN, and TOMM20 in the mitochondrial fractions; (C), representative image of immunoblots of PINK1, PRKN, and GAPDH in the cytosolic fractions; (D), PINK1:GAPDH ratios from quantification of immunoblot images in liver extracts by Image-J; (E), PINK1:TOMM20 ratios in the mitochondrial fractions; (F), PINK1:GAPDH ratios in the cytosolic fractions; (G), PRKN:GAPDH ratios in liver extracts. (H), PRKN:TOMM20 ratios in the mitochondrial fractions; (I), PRKN:GAPDH ratios in the cytosolic fractions. Values are means ± SEM (n = 4/group). *, p < 0.05, **, p < 0.01 vs no ethanol.

PRKN, an E3 ubiquitin ligase that promotes mitophagy [32], also increased in a dose-dependent fashion in total liver extracts after ethanol treatment (2, 4, and 6 g/kg, Figure 3A and G). After ethanol treatment at 4 g/kg, mitochondrial PRKN increased 111%, whereas cytosolic PRKN decreased ~30% after acute ethanol, consistent with translocation of PRKN from cytosol to mitochondria (Figure 3B, C, H, and I).

Mitophagy occurs after acute ethanol treatment

Formation of LC3 puncta is a basic feature of macroautophagic processes. To document directly autophagic sequestration of mitochondria, MitoTracker Red (MTR) was injected 1 h before ethanol treatment. Like TMRM, cationic MTR enters polarized mitochondria. Unlike TMRM, MTR binds covalently to mitochondrial proteins after entry. Therefore, after mitochondrial uptake of MTR and covalent attachment of MTR to mitochondrial proteins, mitochondria retain MTR even after subsequent mtDepo [36,37]. In livers of vehicle-treated mice, the few GFP-LC3 puncta observed were rarely in association with MTR-labeled mitochondria (Figure 4A, left). After acute ethanol treatment (4 g/kg), GFP-LC3 green puncta markedly increased and became larger in size. Many puncta were adjacent to or encircling MTR-labeled mitochondria (Figure 4A, right, inset). These findings document unequivocally autophagic sequestration of mitochondria in vivo after ethanol treatment.

Figure 4.

Autophagic sequestration of hepatic mitochondria and lipophagy in vivo after acute ethanol. (A), GFP-LC3 Tg mice were injected with MTR, as described in Materials and Methods, and gavaged with vehicle (left) or 4 g/kg ethanol (right) after 1 h. Intravital multiphoton microscopy of MTR (red) and GFP-LC3 (green) fluorescence was performed ~4 h after ethanol treatment. Inset on right is an enlarged view of a portion of the same image. (B), mice were gavaged with saline (left) or 4 g/kg ethanol (right). BODIPY 558/568 was arterially infused ~4 h after ethanol treatment. Intravital confocal microscopy of BODIPY 558/568 (red) and GFP-LC3 (green) fluorescence was performed at ~10 min after BODIPY 558/568 injection. Bars are 5 µm. Shown are representative overlay images from 3 mice/group.

Steatosis but not lipophagy occurs at an early stage after acute ethanol treatment

Formation of GFP-LC3 puncta can also occur due to other macroautophagic processes, such as lipophagy. We showed previously in our model that fat droplets accumulate quickly after acute ethanol treatment [19]. This steatosis may stimulate lipophagy. To examine the role of lipophagy early after acute ethanol treatment, we infused BODIPY 558/568 into mice 4 h after ethanol gavage (4 g/kg) when active mitophagy was already occurring (see Figure 1B). BODIPY 558/568, a red fluorophore, selectively labels fat droplets. In livers of vehicle-treated mice, few GFP-LC3 puncta were observed, and red fat droplets were rare (Figure 4B, left). In mice treated with 4 g/kg ethanol, both fat droplets and GFP-LC3 puncta increased (Figure 4B, right). However, virtually no GFP-LC3 puncta were adjacent to or merged with fat droplets. Thus, lipophagy was not driving GFP-LC3 puncta formation at this early time point at 4 h after ethanol treatment.

Disulfiram increases acetaldehyde adduct formation, mitochondrial depolarization, and LC3 puncta formation

If mtDepo is stimulating mitophagy, then enhancement of mtDepo should increase whereas suppression of mtDepo should decrease formation of GFP-LC3 puncta. We showed previously that AcAld plays an essential role in inducing mtDepo [19]. Since AcAld is highly reactive, volatile, and rapidly reacts with other molecules (e.g., other aldehydes and proteins), direct measurement of AcAld formation is difficult. Therefore, malondialdehyde-acetaldehyde (MAA) adducts, a surrogate of AcAld production, was measured in liver tissue. After treatment with ethanol (2, 4, and 6 g/kg), MAA increased in a dose-dependent manner (Figure 5A and B) in parallel to increases of mtDepo and LC3 puncta (Figure 2). To determine if increased AcAld enhances mtDepo, mice were treated with disulfiram (DSF, an inhibitor of ALDH2 [aldehyde dehydrogenase 2, mitochondrial]) and a low dose of ethanol (2 g/kg). In mice treated with DSF only, mtDepo was 2.9%, which was not significantly different from vehicle-treated mice without ethanol treatment (2.6%), although GFP-LC3 puncta after DSF alone increased slightly from 1/cell to 1.8/cell (p < 0.05) (Figure 5C, D and E). The basis for this very small increase of GFP-LC3 puncta by DSF remains unclear. In mice treated with a low dose of ethanol (2 g/kg), DSF markedly increased mtDepo from ~40% to ~90% in parallel with an increase of GFP-LC3 puncta from 5.7/cell to 9.4/cell (Figure 5C, D and E).

Figure 5.

Malondialdehyde-acetaldehyde adducts after ethanol treatment and increased hepatic mitochondrial depolarization and GFP-LC3 puncta formation after aldehyde dehydrogenase inhibition with disulfiram. (A and B), C57Bl/6 mice were treated with saline or ethanol (2–6 g/kg, i.g.). Livers were harvested for detection by immunoblotting and quantification by ImageJ of MAA adducts and GAPDH. #, p = 0.06, *, p < 0.05 and **, p < 0.01 vs no ethanol (0 g/kg). (C-E), GFP-LC3 Tg mice were injected with DSF (200 mg/kg, ip) or vehicle and then gavaged with one dose of ethanol (2 g/kg) 1 h later. TMRM and GFP-LC3 fluorescence were detected by intravital multiphoton microscopy after ~4 h (representative images shown in C). Bar is 10 µm. Numbers of hepatocytes with polarized and depolarized mitochondria, as well as numbers of GFP-LC3 puncta in all hepatocytes (both with and without mtDepo), were counted in 10 random fields per mouse. The percentage of hepatocytes with mtDepo (D) and the number of GFP-LC3 puncta per cell (E) are plotted. Values are means ± SEM (n = 3–4/per group). a, p < 0.05 vs no ethanol (0 g/kg); b, p < 0.05 vs 2 g/kg of ethanol treatment.

Alda-1 and tacrolimus decrease mitochondrial depolarization and LC3 puncta formation

We further examined the effects of Alda-1, an ALDH2 activator that accelerates acetaldehyde degradation [38,39], on GFP-LC3 puncta formation in relation to mtDepo. In livers of GFP-LC3 Tg mice treated with 4 g/kg of ethanol, mitochondria in ~75% of hepatocytes became depolarized (Figure 6A and B). Simultaneously, GFP-LC3 puncta increased to 10.2/cell overall (Figure 6A and C). Alda-1 pretreatment produced commensurate decreases of both mtDepo to ~49% of hepatocytes and of GFP-LC3 puncta to 5.6/cell (Figure 6).

Figure 6.

Alda-1 and tacrolimus decrease hepatic mitochondrial depolarization and GFP-LC3 puncta formation after ethanol. GFP-LC3 Tg mice were injected with Alda-1 (50 mg/kg, ip), tacrolimus (2 mg/kg, ig), or vehicle and then gavaged 1 h later with one dose of ethanol (4 g/kg). TMRM and GFP-LC3 fluorescence was detected by intravital multiphoton microscopy after ~4 h, as described for Figure 1. Representative images are shown in A. Bar is 10 µm. Numbers of hepatocytes with polarized and depolarized mitochondria, as well as numbers of GFP-LC3 puncta in all hepatocytes (both with and without mtDepo), were counted in 10 random fields per mouse. The percentage of hepatocytes with mtDepo (B) and the number of GFP-LC3 puncta per cell (C) are plotted. Values are means ± SEM (n = 3–5/per group). *p < 0.05, **, p < 0.01 vs ethanol.

Tacrolimus, an immunosuppressant and PPP3/calcineurin inhibitor, protects mitochondria after ischemia/reperfusion injury and blocks calcium ionophore-induced mtDepo in astrocytes and neurons, possibly by inhibiting the rise in intracellular free Ca²⁺ [40,41]. Tacrolimus also decreases hepatic mtDepo after acute ethanol treatment [42]. In GFP-LC3 mice after treatment with 4 g/kg ethanol, we found that tacrolimus pretreatment produced commensurate decreases of mtDepo from ~75% to ~34% of hepatocytes and of GFP-LC3 puncta from 10.2 to 3.8/cell overall (Figure 6A, B and C), showing that inhibition of mtDepo by a mechanism other than decreasing AcAld also decreases GFP-LC3 puncta formation.

Increased mitophagosome processing into lysosomes after acute ethanol treatment

Once formed, mitophagosomes fuse with lysosomes for hydrolytic digestion of their contents. We examined whether acute ethanol treatment alters the levels of TFEB, the master regulator of lysosomal gene expression [43,44]. Compared to livers of vehicle-treated mice, TFEB expression in livers of ethanol (4 g/kg)-treated mice was not significantly different (Figure 7A and E). Nonetheless, after ethanol treatment, TFEB increased by 101% in the nuclear fraction (Figure 7B and F) and decreased 34% in the cytosolic fraction (Figure 7C and G), indicating nuclear translocation of TFEB. Additionally, LAMP1 (lysosomal-associated membrane protein 1), a lysosomal marker, increased ~50% after ethanol treatment (Figure 7D and H).

Figure 7.

TFEB expression and nuclear translocation and LAMP1 expression after acute ethanol. C57Bl/6 mice were treated with saline or ethanol (4 g/kg, i.g.). TFEB and LAMP1 in liver tissue, TFEB in nuclear and cytosol fractions, and housekeeping proteins GAPDH (cytosol) and LMNB1 (nucleus) were detected by immunoblotting. (A, B, and C), representative immunoblots of TFEB in whole liver, nucleus and cytosol, respectively; (D), representative immunoblots of LAMP1 in whole liver extracts; (E), TFEB:GAPDH ratios in liver extracts from quantification of immunoblots by ImageJ; (F), TFEB:LMNB1 ratios in nuclear fractions; (G), TFEB:GAPDH ratios in cytosol; (H), LAMP1:GAPDH ratios in liver extracts. Values are means ± SEM (n = 3-4/group). **, p < 0.01 vs no ethanol.

We also performed intravital multiphoton microscopy after labeling lysosomes with rhodamine dextran (RhDex). Lysosomes selectively accumulate RhDex, which enters cells by macroendocytosis but is resistant to degradation and consequently becomes a fluorescent marker of lysosomes [45]. In livers of vehicle-treated mice after RhDex labeling, many red-fluorescing lysosomes were detected, and RhDex-positive areas were ~6% (Figure 8A and C). At ~4 h after acute ethanol treatment (4 g/kg), RhDex-positive areas increased to ~11%, documenting increased lysosomal mass. In livers of vehicle-treated mice, red-fluorescing RhDex-labeled lysosomes were rarely in association with the small number of green GFP-LC3 puncta present (Figure 8A). After ethanol treatment, GFP-LC3 puncta markedly increased, and many GFP-LC3 puncta were adjacent to or completely merged with lysosomes (Figure 8B, and inset in the right panel), demonstrating processing of autophagosomes into lysosomes. Colocalization of lysosomes and GFP-LC3 puncta was assessed by two different methods. Spearman’s rank correlation value between RhDex and GFP-LC3 increased from 0.11 in vehicle-treated mice to 0.51 in ethanol-treated mice, and Pearson’s R value increased from 0.09 to 0.4, both consistently indicating increased colocalization of lysosomes and GFP-LC3 puncta after ethanol treatment (Figure 8D and E).

Figure 8.

Increased hepatic mitophagosome processing into lysosomes in vivo after acute ethanol. GFP-LC3 Tg mice were injected with RhDex and gavaged with saline (A) or ethanol (4 g/kg, B) ~18 h later, as described in Materials and Methods. Intravital multiphoton microscopy of GFP-LC3 (left panels) and RhDex (middle panels) was performed using a 60x silicone oil objective lens at ~4 h after ethanol treatment. Right panels are overlays of the RhDex and GFP-LC3 images. Scale bar: 2 µm. Shown are representative images from 3 mice/group. (C), quantification of RhDex-positive areas by image analysis using ImageJ. Colocalization of RhDex and GFP-LC3 puncta was analyzed using ImageJ by 2 different methods and expressed as Spearman’s rank correlation value (D) and Pearson’s R value (E), respectively. Values are means ± SEM (n = 3/group). *, p < 0.05, **, p < 0.01 vs no ethanol.

DISCUSSION

Mitophagy occurs in livers of living mice after acute ethanol treatment

Previous studies show an increase of hepatic autophagy or mitophagy after ethanol treatment [21,22]. After acute ethanol treatment, mitophagy seems protective, since suppression of mitophagy by chloroquine, Atg7 knockdown, and PRKN deficiency increase liver injury [21,22,46]. After chronic ethanol treatment, some studies show enhanced autophagic flux; whereas other studies show disrupted lysosomal function and autophagosomal processing, leading to autophagosome accumulation [47–50]. Mitophagy decreases mitochondrial content if not coordinated with compensating mitochondrial biogenesis. Indeed, mtDNA depletion occurs after acute ethanol exposure [51,52].

To show early onset of mitophagy after ethanol treatment in living animals, we used MTR to label mitochondria before ethanol treatment in GFP-LC3 Tg mice. MTR enters polarized mitochondria and then covalently binds to mitochondrial proteins, such that MTR is retained when mtDepo subsequently occurs (e.g., after ethanol treatment). After ethanol treatment, GFP-LC3 associated with or encircled many MTR-labeled mitochondria, demonstrating unequivocally mitophagic sequestration in vivo as early as 4 h after ethanol exposure (Figure 4A). Mitochondria release TMRM upon mtDepo (Figure 1), whereas covalently linked MTR is retained (Figure 3) [36,37] Thus, MTR labeling inside puncta does not signify the presence of polarized mitochondria inside mitophagosomes. Importantly, TMRM unlike MTR did not appear inside GFP-LC3 puncta except as an infrequent superimposition artifact, since image slices have a finite thickness (~1 µm) and one structure under another can produce the appearance of colocalization. The appearance of MTR but not TMRM labeling inside GFP-LC3-decorated mitophagosomes supports the conclusion that mitophagosomes contain only depolarized mitochondria.

Mitochondrial depolarization initiates ethanol-induced type 2 mitophagy

Although many studies show autophagy or mitophagy after ethanol exposure, what triggers mitophagy and which type of mitophagy occurs are not understood. Mitophagy has distinct variants. Type 1 mitophagy occurs in conditions such as nutrient deprivation and cytoplasmic remodeling to provide metabolic precursors and to adjust for changing metabolic needs [25–27]. PIK3C3/VPS34, BECN1/beclin 1, and other proteins trigger the formation of cup-shaped sequestration membranes that then recruit LC3, enlarge, and wrap around individual mitochondria to form mitophagosomes, often with coordinate mitochondrial fission. In Type 1 mitophagy, mitochondria remain polarized until after their sequestration into mitophagosomes [31]. Acidification of the outer compartment of mitophagosomes then occurs, followed by mtDepo. Thus, Type 1 mitophagy removes functional mitochondria.

In contrast, mitochondrial depolarization/damage precedes and initiates autophagic sequestration in Type 2 mitophagy [25,26,53]. Ordinarily, functional mitochondria import full-length PINK1 using the mitochondrial protein import machinery and driven by membrane potential (ΔΨ). PINK1 degradation in the mitochondrial matrix keeps endogenous PINK1 at low levels [32]. When mitochondria depolarize, however, PINK1 import is arrested, and PINK1 accumulates on the surface of mitochondria, leading to binding of PRKN, an E3 ligase, and ubiquitination of outer membrane proteins. The ubiquitinated proteins are recognized by receptor proteins like SQSTM1/p62, leading to association of LC3-containing membranes that form autophagosomes enveloping the depolarized mitochondria [26,32,54]. Unlike Type 1 mitophagy, Type 2 mitophagy specifically targets damaged/dysfunctional mitochondria. Type 1 and 2 mitophagy are forms of macroautophagy and require LC3 participation [55]. In Type 3 mitophagy or micromitophagy, vesicles containing oxidized proteins bud off from mitochondria in a PINK1 and PRKN-dependent but LC3-independent fashion. These mitochondria-derived vesicles internalize into multivesicular bodies, a type of lysosome, by vesicle scission into the lumen [56,57].

In cultured hepatocytes, we showed previously that photodamage to individual mitochondria leads to mtDepo, association of LC3, mitophagic sequestration of the damaged organelles, and processing to lysosomes [58]. Similarly, others show that uncoupler treatment in vitro to cause depolarization of all mitochondria also leads to mitophagy [59,60]. Since a reversible mtDepo occurs in vivo after acute ethanol treatment is related to increased AcAld, the purpose of this study was to understand the cellular mechanisms underlying how alcohol consumption and metabolism leads to mitophagy in vivo. Specifically, we tested the hypothesis that mtDepo driven by ethanol metabolism to AcAld triggers Type 2 mitophagy. Onset of mtDepo occurs as early as 1 h after a single bolus of ethanol, peaks at 6–12 h, and recovers substantially by 24 h [19]. AcAld from ethanol metabolism triggers mtDepo, since deficiency of ADH (alcohol dehydrogenase) and acceleration of AcAld degradation by ALDH2 activation markedly decrease mtDepo after ethanol treatment. Although a report shows that mitochondria isolated from mice treated chronically with ethanol exhibit greater sensitivity to a calcium-induced mitochondrial permeability transition (MPT) [61], hepatic mtDepo in vivo is not the MPT, since mtDepo is reversible and not decreased by MPT blockers like cyclosporin A [19].

In this study, we used intravital multiphoton microscopy to characterize the relationship of hepatic mtDepo to LC3 puncta formation in living GFP-LC3 Tg mice. The following evidence supports the conclusion that mtDepo drives mitophagy after ethanol treatment. First, GFP-LC3 puncta increased dose dependently after ethanol treatment in parallel to increased mtDepo (Fig. 1 and 2). Second, GFP-LC3 puncta increased principally in hepatocytes with mtDepo (Fig. 1 and 2). Importantly, the increase of total hepatic GFP-LC3 puncta with increasing ethanol was primarily due to recruitment of more hepatocytes with mtDepo (Figure 2). Third, interventions that increased and decreased mtDepo also increased and decreased mitophagy. Specifically, DSF increased both mtDepo and GFP-LC3 puncta formation (Figure 5), whereas Alda-1 and tacrolimus, unrelated drugs that inhibit mtDepo through different mechanisms, decreased mtDepo and GFP-LC3 puncta to similar extents (Figure 6). Last, mitochondrial localization of PINK1 and PRKN, important mediators of mtDepo-driven mitophagy, increased after ethanol treatment (Figure 3). Previously, PRKN deficiency was shown to decrease mitophagy after ethanol treatment [22], consistent with mtDepo-driven mitophagy and a role of PINK1 and PRKN in ethanol-induced mitophagy (see also [62,63]). Together, these findings support the novel conclusion that mtDepo is the primary initiator of mitophagic sequestration and GFP-LC3 puncta formation after ethanol treatment (Figure 9).

Figure 9.

Mitochondrial depolarization initiates mitophagy during adaptive ethanol metabolism. AcAld increases rapidly after acute ethanol due to ethanol oxidation by ADH and CYP2E1(cytochrome P450, family 2, subfamily e, polypeptide 1) in the non-mitochondrial cytoplasm. AcAld then diffuses into mitochondria to be further oxidized and detoxified to acetate by ALDH2, an enzyme activated by Alda-1 and inhibited by DSF. AcAld accumulation then initiates mtDepo, which is inhibited by tacrolimus. Subsequently, PINK1 and PRKN accumulate on depolarized mitochondria to initiate Type 2 mitophagy, which removes damaged mitochondria. AcAld also induces VDAC closure, which together with mtDepo suppresses mitochondrial beta -oxidation of fatty acids and leads to steatosis. AcAld and possibly the competing demands for mitophagy by widespread mtDepo also inhibit lipophagy at this stage, which further promotes fat accumulation. Thus, steatosis develops rapidly after acute ethanol treatment.

Hepatic PINK1 and PRKN protein expression increased dose dependently after acute ethanol treatment (2–6 g/kg) (Figure 3). Although the increment of PINK1 and PRKN seems modest compared with the formation of GFP-LC3 puncta, formation of LC3 puncta represents the redistribution of LC3 from the cytosol to autophagosomes rather than a change of total protein expression. By contrast, PINK1 and PRKN increase due to their stabilization after binding to depolarized mitochondria. Thus, GFP-LC3 puncta formation and PINK1/PRKN protein expression need not be commensurate.

A small increase of autophagy after ethanol treatment was also observed in hepatocytes with still polarized mitochondria (Figure 2C). The basis for this small increase may be that hepatocytes switch from polarized to depolarized and back stochastically after ethanol. Autophagic puncta in hepatocytes with polarized mitochondria would then represent residua from a prior mtDepo event. Another possibility is that ethanol also induces some Type 1 mitophagy related to bioenergetic stresses caused by mtDepo in nearby hepatocytes.

Overall, mitophagy occurring in vivo after acute ethanol treatment had the following characteristics: 1) dependence on mtDepo, 2) association with increased mitochondrial PINK1 and PRKN, and 3) involvement of LC3. Type 1 mitophagy removes functional mitochondria and does not require mtDepo for mitophagosome formation. PINK1 and PRKN are involved in both Type 2 and 3 mitophagy but not in Type-1 mitophagy, whereas LC3 encircling of mitochondria does not occur in Type 3 mitophagy. Therefore, we can conclude that mitophagy after acute ethanol treatment is mainly Type 2 mitophagy (Figure 9).

Lipophagy is minimal early after acute ethanol treatment

LC3 mediates other macroautophagy processes, such as lipophagy [32]. Steatosis is a characteristic of ALD and occurs within just a few hours after acute ethanol treatment [4,19]. Lipophagy, the macroautophagic degradation of fat droplets, is active in hepatocytes to resorb lipid droplets, thereby maintaining lipostasis [64,65]. To assess whether lipophagy drives LC3 puncta formation after acute ethanol treatment, we labeled fat droplets with BODIPY 558/568 at 4 h after vehicle or ethanol treatment, a timepoint when autophagy becomes active in GFP-LC3 Tg mice. In vehicle-treated mice, few hepatic fat droplets were present (Figure 4B, left). By contrast, fat droplets increased after acute ethanol treatment but were rarely adjacent to or merged with GFP-LC3 puncta (Figure 4B, right). Therefore, increased macroautophagy was not due to lipophagy at this early stage after acute ethanol treatment. The reason for lack of lipophagy at this timepoint remains unclear. In vitro, treatments with ethanol (100 mM) and AcAld (100–500 µM) suppress lipophagy in ethanol-metabolizing VA-13^{ADH+/CYP2E1−} cells, an effect alleviated by ALDH2 activation with Alda-1 [5,65]. Thus, increased intrahepatic AcAld after acute ethanol treatment may be inhibiting lipophagy (Figure 9).

Multiple alterations after ethanol treatment, including mtDepo, closure of mitochondrial voltage-dependent anion channels (VDAC), and suppression of lipophagy, appear linked to AcAld formation and may work synergistically to cause fatty liver [5,13,19,26,65]. Previously, we showed that steatosis after acute ethanol treatment occurs principally in hepatocytes with mtDepo, strongly implying that steatosis is a response to mtDepo [19]. mtDepo inhibits oxidative phosphorylation and decreases ATP production after acute ethanol administration [17–20]. Since ATP is required for acylCoA formation required for β-oxidation, ATP depletion promotes hepatic triglyceride accumulation and steatosis. Other compounds and conditions causing acute mitochondrial inhibition or damage also produce hepatic steatosis, such as Reye syndrome [66–69].

Accumulating evidence indicates that VDAC is a dynamic regulator or “governator” of mitochondrial metabolite exchange [70]. In adaptive alcohol metabolism, VDAC closure occurs coordinately with mtDepo to promote selective mitochondrial AcAld oxidation by ALDH2, since AcAld, but not other respiratory substrates, easily crosses the outer membrane independently of VDAC (Figure 9). VDAC closure also prevents futile mitochondrial hydrolysis of cytosolic ATP after mtDepo to allow glycolysis, which is activated after acute ethanol treatment, to partially maintain non-mitochondrial ATP levels [13,26,71,72]. Because VDAC closure blocks entry of fatty acyl-CoA to access the carnitine shuttle, beta-oxidation becomes further inhibited. Simultaneously, AcAld may inhibit lipophagy. As a result, lipostasis in hepatocytes is disrupted, leading to rapid development of fatty liver (Figure 9) [19,65,72]. However, direct activation of mitophagy by AcAld or AcAld adducts seems unlikely, since tacrolimus, an agent not targeting AcAld metabolism, decreased mtDepo and consequently mitophagy.

Acute ethanol treatment stimulates lysosomal processing of mitophagosomes

Mitophagosomes are delivered to lysosomes for degradation [26]. Chronically after ethanol treatment, suppressed lysosomal processing leads to accumulation of dysfunctional mitochondria and release of mitochondrial damage-associated molecular patterns (mtDAMPS), which may lead to liver injury, inflammation, and fibrosis [13,26,50,73]. Mitochondrial DNA (mtDNA) is a major mtDAMP recently shown to be a stellate cell activator and driver of liver fibrosis [74]. Accordingly, we investigated mitophagosome processing into lysosomes after acute ethanol treatment.

TFEB is the master regulator of lysosomal biogenesis and function, which controls the coordinated expression, import, and activity of lysosomal enzymes [75,76]. TFEB also modulates gene expression involved in related processes, including lysosomal acidification, exocytosis, and endocytosis [44,75]. Previous studies show that nuclear translocation of TFEB increases after acute ethanol treatment but decreases after chronic ethanol treatment in mice and in patients with alcoholic hepatitis, implying suppression of lysosomal biogenesis and autophagy after chronic ethanol exposure [50,77]. In this study, although autophagy increased markedly, TFEB protein expression in whole cell extracts remained unchanged, whereas TFEB in nuclear extracts increased (Figure 7). This nuclear translocation of TFEB was accompanied by increased expression of LAMP1, a marker of lysosomes (Figure 7). Moreover, RhDex-labeled lysosomes increased markedly (Figure 8). Together, these demonstrate that TFEB nuclear translocation stimulated expansion of lysosomal mass and processing of autophagosomes after acute ethanol treatment.

We observed a ~ 9-fold increase of LC3 puncta after acute ethanol treatment (Figure 2). This observation can only be reasonably explained by an increase in autophagic sequestration of mitochondria rather than a decrease of autophagic processing into lysosomes, since agents like bafilomycin that block lysosomal processing typically produce only an approximate doubling of autophagosome content in in vitro models [78]. Moreover, the increase of lysosomal cross-sectional area, enhancement of LAMP1 expression, and encirclement of MTR-labeled mitochondria by GFP-LC3 after ethanol treatment also signify active processing of mitophagosomes into lysosomes (Fig. 4, 7 and 8). These would all decrease or remain unchanged if processing to lysosomes were suppressed. Last, others show that ethanol treatment inhibits autophagic processing only in chronic models [50]. Our findings confirm and extend previous conclusions that mitophagy and mitochondrial turnover increase after ethanol treatment and further demonstrate that this phenomenon is principally mtDepo-driven Type 2 mitophagy.

Conclusions and future directions

In conclusion, these novel findings document a chain of causality whereby AcAld formation from ethanol metabolism triggers mtDepo, which in turn leads to Type 2 mitophagic sequestration and processing of mitophagosomes into lysosomes (Figure 9). Steatosis develops also, but mitophagy rather than lipophagy predominates in the autophagic process at this early stage after acute ethanol exposure.

Several questions remain unanswered. For example, polarized mitochondria reappear ~24 h after ethanol treatment [19]. After all-or-nothing mtDepo in individual hepatocytes, LC3 puncta seem disproportionately less abundant than depolarized mitochondria, which suggests that not all depolarized mitochondria undergo mitophagy. Accordingly, further damage (a second hit) may be needed for mitophagy to progress or that factors like PINK1, PRKN, ubiquitination, and others are insufficient to support mitophagic removal of all mitochondria. Alternatively, repolarization may be mediated by biogenesis of new mitochondria. Why and the mechanisms how mtDepo occurs as an all-or-nothing event in individual hepatocytes are also unknown. Possibly, all-or-nothing mtDepo avoids futile cycling inside individual hepatocytes of ATP generation by polarized mitochondria and ATP hydrolysis by depolarized mitochondria. Last, it remains to be determined whether hepatocytes cycle between mtDepo and a normal polarized state to avoid more major disruption of cellular homeostasis from sustained mtDepo. These questions will be addressed in future studies.

MATERIALS AND METHODS

Synthesis of Alda-1

Alda-1 was synthesized as described previously [79]. The purity of the product was >99% as detected by HPLC chromatography, and the structure was verified by NMR spectroscopy.

Animal treatments

Male C57BL/6 mice (Jackson Loboratory) and GFP-LC3 transgenic mice (8–9 weeks) were used in this study. GFP-LC3 Tg mice were bred at the Medical University of South Carolina from original breeding pairs kindly provided by Dr. Noboru Mizushima (Tokyo Medical and Dental University) [80]. Fed mice were gavaged with a single dose of ethanol (2–6 g/kg in normal saline, 20 µl/g body weight) or equal volume of vehicle. Food was removed immediately after gavaging and was not returned before imaging or tissue collection. At 1 h before ethanol treatment, some mice were pretreated with Alda-1 (50 mg/kg, i.p., 30 µl/mouse), an ALDH2 (aldehyde dehydrogenase 2, mitochondrial) activator [38], DSF (200 mg/kg, i.p., 1 µl/g body weight; Millipore Sigma Company, 86,720), an ALDH2 inhibitor [81], or tacrolimus (2 mg/kg, i.g., 2 µl/g body weight; Abcam Biotechnology, 120,223), a drug that decreases mtDepo after ethanol treatment [40,41]. Control mice were treated with equal volumes of vehicles for Alda-1 and DSF (DMSO) or tacrolimus (normal saline). All animals were given humane care using protocols approved by the Animal Care and Use Committee of the Medical University of South Carolina.

Detection of GFP-LC3 expression and mitochondrial depolarization in mouse liver by intravital multiphoton microscopy

At ~4 h after ethanol treatment, TMRM, a red-fluorescing fluorophore that enters negatively polarized mitochondria, was used to monitor mitochondrial polarization status, and GFP-LC3 fluorescence was monitored to evaluate autophagy. Under ketamine and xylazine (90 mg/kg and 10 mg/kg, i.p.) anesthesia, the trachea was intubated with a 20 G catheter, which was connected to a small animal respirator. TMRM (1.1 µmol/mouse; ThermoFisher Scientific, T668) was infused slowly into the carotid artery over about 10 min. After laparotomy, mice were placed on the stage of microscope in a prone position, and the liver was gently positioned over a No. 1.5 glass coverslip above a silicone oil objective lens (60X, 1.3 N.A.). Body temperature was maintained at 37°C using an environmental chamber. Intravital imaging was performed with an Olympus FluoView 1200 MPE multiphoton microscope (Olympus, Center Valley, PA) equipped with a Spectra Physics Mai Tai Deep Sea tunable multiphoton laser (Newport, Irvine, CA). Using 920-nm multiphoton excitation, fluorescence of TMRM and GFP-LC3 was imaged simultaneously through 575–630 nm and 495–540 nm band pass emission filters, respectively. The respirator was turned off for 5–10 sec during imaging to eliminate breathing artifacts. Ten or more random images were collected for each mouse.

In 10 images per mouse, parenchymal cells were scored in a blinded fashion for punctate TMRM fluorescence representing cells with polarized mitochondria or a dimmer diffuse cytosolic fluorescence indicating mtDepo. Hepatocytes with mtDepo were counted and the percentage of total cells with mtDepo calculated. Green GFP-LC3 fluorescent puncta in hepatocytes with polarized and depolarized mitochondria were also calculated separately in 10 images from each mouse.

Detection of mitophagy by mitotracker red and GFP-LC3

MTR (0.5 µmol/mouse, i.v.; ThermoFisher, M7512) was injected into GFP-LC3 Tg mice 1 h before acute ethanol treatment (4 g/kg). MTR enters polarized mitochondria electrophoretically and forms covalent protein adducts that are not released after mtDepo. Encirclement of MTR-labeled mitochondria by GFP-LC3 signified directly the occurrence of mitophagic sequestration. Red and green fluorescence of MTR and GFP-LC3 was imaged by intravital multiphoton microscopy using 920-nm multiphoton excitation, as described above.

Detection of lipid droplets

We showed previously that steatosis after acute ethanol treatment occurs primarily in hepatocytes with depolarized mitochondria [19]. To determine whether GFP-LC3 puncta associate with lipid droplets to signify lipophagy, GFP-LC3 Tg mice were treated with ethanol (4 g/kg) or equal volume of saline. About 4 h later, BODIPY 558/568 (0.08 mg/mouse; ThermoFisher Scientific, D3922), a red fluorophore that labels fat droplets, was infused slowly into the carotid artery. Intravital confocal microscopy was then performed using a Zeiss LSM 880 confocal microscope (Zeiss, Thornwood, NY) and a 63× oil-immersion planapochromat objective lens (1.4 N.A.). BODIPY 558/568 and GFP-LC3 fluorescence was imaged simultaneously using 561-nm and 488-nm excitation, respectively, and emission was collected at 490–553 nm and 570–735 nm.

Detection of autophagosome processing into lysosomes

Rhodamine-dextran (RhDex, 70 kDa; ThermoFisher Scientific, D1824), which labels lysosomes, was injected (100 mg/kg, i.p.) into GFP-LC3 Tg mice, and ethanol (4 g/kg) was gavaged 18 h later. Intravital multiphoton microscopy was performed 4 h after ethanol treatment using an Olympus FluoView 1200 MPE multiphoton microscope. Imaging conditions for RhDex and GFP-LC3 were as described above for TMRM and GFP-LC3. Using ImageJ, lysosomal mass was quantified from RhDex positive cross-sectional areas of 5 images/mouse in the red channel, and autophagosome processing into lysosomes was accessed by co-localization of GFP-LC3 puncta with RhDex-labeled lysosomes. Spearman’s Rank correlation value and Pearson’s R value were analyzed in 3 images per mouse from 3 random cytoplasmic regions per image.

Immunoblotting

Liver tissue was collected under ketamine-xylazine anesthesia 4 h after ethanol treatment, snap-frozen in liquid nitrogen, and stored at −80°C until use. Nuclear, mitochondrial, and cytosolic fractions were isolated using a Cytosol and Nuclear Isolation Kit (Invent Biotech, NT-032) and Mitochondrial Isolation Kit (Sciencell Research, 8268) according to the manufacturers’ instructions. Proteins of interest in liver extracts and nuclear, mitochondrial, and cytosolic fractions were detected by immunoblotting, as described previously [82]. Primary antibody against glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was obtained from Cell Signaling Technology (2118). Primary antibodies against PINK1, PRKN, TOMM20 and LMNB1 (lamin B1) were from Santa Cruz Biotechnology (sc-517,353, sc-32,282, sc-11,415, and sc-6216, respectively). Primary antibody against TFEB was from Bethyl laboratories (A303-673A). Primary antibody against LAMP1 was from Developmental Studies Hybridoma Bank (1D4B). Primary antibody for malondialdehyde-acetaldehyde (MAA) adducts was a gift from Dr. Todd Wyatt (University of Nebraska Medical Center). Immunoblots were exposed to primary antibodies at a dilution of 1:1,000 overnight at 4°C and then to the secondary antibodies (Jackson Immuno Research Laboratory, 111,035,003; 115,035,003) at a dilution of 1:10,000 at room temperature for 1 h. For visualization, membranes were incubated with the SuperSignal Reagent (ThermoFisher, 34,076) for about 1 min, and images of blots were captured using the Chemidoc Touch Imaging System (Bio-Rad). Band intensities were quantified using NIH ImageJ software.

Statistical analysis

All groups were compared using the Student’s t-test or ANOVA plus Student-Newman-Keuls’ or Bonferroni post-hoc test as appropriate. Values are means ± SEM. Differences were considered significant at p < 0.05.

Funding Statement

This work was supported by the National Institute on Alcohol Abuse and Alcoholism [AA025379]; National Institute on Alcohol Abuse and Alcoholism [AA017756]; National Institute on Alcohol Abuse and Alcoholism [AA021191]; National Institutes of Health [1 P30 CA138313]; National Institutes of Health [1 S10 OD018113]; National institute of Diabetes and Digestive and Kidney Diseases [1 P30 DK123704]; National Institutes of General Medicines [1 P20 GM130457]; National Institute of General Medicines [1 P30 GM140964].

Disclosure statement

No potential conflict of interest was reported by the author(s).