Liver-Specific Deletion of Phosphatase and Tensin Homolog Deleted on Chromosome 10 Significantly Ameliorates Chronic EtOH-Induced Increases in Hepatocellular Damage

Alcoholic liver disease is a significant contributor to global liver failure. In murine models, chronic ethanol consumption dysregulates PTEN/Akt signaling. Hepatospecific deletion of phosphatase and tensin homolog deleted on chromosome 10 (PTENLKO) mice possess constitutive activation of Akt(s) and increased de novo lipogenesis resulting in increased hepatocellular steatosis. This makes PTENLKO a viable model to examine the effects of ethanol in an environment of preexisting steatosis. The aim of this study was to determine the impact of chronic ethanol consumption and the absence of PTEN (PTENLKO) compared to Alb-Cre control mice (PTENf/f) on hepatocellular damage as evidenced by changes in lipid accumulation, protein carbonylation and alanine amino transferase (ALT). In the control PTENf/f animals, ethanol significantly increased ALT, liver triglycerides and steatosis. In contrast, chronic ethanol consumption in PTENLKO mice decreased hepatocellular damage when compared to PTENLKO pair-fed controls. Consumption of ethanol elevated protein carbonylation in PTENf/f animals but had no effect in PTENLKO animals. In PTENLKO mice, overall hepatic mRNA expression of genes that contribute to GSH homeostasis as well as reduced glutathione (GSH) and oxidized glutathione (GSSG) concentrations were significantly elevated compared to respective PTENf/f counterparts. These data indicate that during conditions of constitutive Akt activation and steatosis, increased GSH homeostasis assists in mitigation of ethanol-dependent induction of oxidative stress and hepatocellular damage. Furthermore, data herein suggest a divergence in EtOH-induced hepatocellular damage and increases in steatosis due to polyunsaturated fatty acids downstream of PTEN.


Introduction
Alcoholic liver disease (ALD) and non-alcoholic fatty liver disease (NAFLD) are two of the leading causes of liver disease in the United States today. Both ALD and NAFLD are characterized by progressive hepatocellular damage manifested in increased steatosis, steatohepatitis, fibrosis and ultimately progression to cirrhosis [1]. In the western world, the prevalence of diet-induced nonalcoholic steatohepatitis (NASH) has dramatically increased in the last decade. According to the latest data from the Centers of Disease Control, current estimates indicate 30-35% of all Americans are obese and 69% are overweight (http://www.cdc.gov/ nchs/fastats/obesity-overweight.htm). Moderate alcohol consumption by human subjects with increased body mass index is strongly correlated with increased hepatic damage as determined by plasma alanine amino transferase (ALT) and Gamma-Glutamyl Transferase levels [2]. Given these statistics, obesity and its concomitant steatosis are predictable cofactors for ALD.
In the liver, insulin is a major regulator of lipid metabolism and alteration of insulin signaling can induce hepatocellular lipid accumulation evident in both ALD and NAFLD [3][4][5][6][7]. Downstream of the insulin receptor, the PTEN/Akt pathway regulates insulin signaling [8]. Recruitment and full activation of Akt requires interaction of N-terminal pleckstrin homology domains with phosphatidylinositol (3,4,5) trisphosphate (PIP 3 ) on intracellular membranes [9,10]. Through its ability to catalyze the hydrolysis of the 3' position phosphate on PIP 3 , PTEN negatively regulates Akt activation and insulin signaling [11]. This is highlighted by bypassing the Akt arm of insulin receptor signaling using mice containing a hepatospecific deletion of PTEN (PTEN LKO ). PTEN LKO mice possess constitutive Akt activation, hepatic insulin hypersensitivity and increased steatohepatitis [12][13][14]. In our previous study using the PTEN LKO model, we determined that increased consumption of a diet high in polyunsaturated fatty acids potentiates hepatocellular damage and decreases hepatocellular redox capacity when compared to chow fed controls [15].
During chronic alcohol consumption, there is pronounced lipid accumulation and enhanced oxidative stress [16]. In murine models of ALD, increased de novo lipogenesis (DNL) has been proposed to be a contributing factor in lipid accumulation [17,18]. Chronic consumption of EtOH combined with dietary polyunsaturated fatty acids decreases PTEN expression/activity increasing activation of Akt2 [19]. In other models, chronic EtOH consumption has also been demonstrated to enhance the Akt activated transcription factor SREBP1, increasing fatty acid synthesis [17]. Furthermore, using a short term model of EtOH toxicity, preadministration of insulin reduced oxidative stress and hepatocellular damage but did not diminish steatosis further demonstrating the contribution of insulin signaling in EtOH toxicity [20]. Most recently, an environment of increased insulin hypersensitivity created by PTEN LKO was demonstrated to provide protection against murine endotoxemia [21]. Thus, it would be reasonable to predict that if insulin signaling was completely bypassed by using mice possessing a constitutively activated form of Akt, then EtOH administration would not increase hepatocellular damage but still increase steatosis. In the present study, mice possessing constitutively activated Akt (PTEN LKO ) were used in a well characterized 6-week model of chronic EtOH consumption to further elucidate the contribution of PTEN/Akt signaling in EtOH-induced steatosis and hepatotoxicity. In PTEN LKO mice, chronic EtOH consumption did not increase hepatocellular damage and corresponded with elevated glutathione metabolism. We hypothesize that the aforementioned elevation in glutathione metabolism assists in mitigation of hepatocellular damage induced by EtOH.

Animal Model
To generate liver-specific PTEN deletion mice (PTEN LKO ), mice carrying PTEN conditional knockout alleles (PTEN loxP/loxP ; Alb-Cre − (PTEN f/f )) were bred with an Albumin (Alb)-Crerecombinase transgenic mouse as previously described [12,15]. At 5 weeks of age, tail snips were taken and mice were genotyped using PTEN f/f and Cre specific primers (Transnetyx, Memphis TN). At 12 weeks of age, mice were isocalorically pair fed (PF) in groups of 6 either a modified Lieber-DeCarli diet (45% polyunsaturated fat derived calories mostly from corn oil) (Bio-Serv, Frenchtown, NJ) or EtOH-fed (EtOH-derived caloric content with an initial EtOH concentration of 10.8%, which was subsequently increased weekly to 16.2, 21.5, 26.9, 29.2, 31.8% ethanol derived calories for a total of 6-weeks. [19,22]. Upon completion of the study, animals were anesthetized via intraperitoneal injection with sodium pentobarbital and euthanized by exsanguination. Blood was collected from the inferior vena cava and plasma was separated through centrifugation at 4°C and assayed for ALT activity (Sekisui Diagnostics, P.E.I., Canada). Excised livers were weighed, homogenized and subjected to differential centrifugation and subcellular fractionation (cytosolic, mitochondrial, microsomal and nuclear fractions) as previously described [23]. All procedures involving animals were approved by the Institutional Animal Care and Use Committee of the University of Colorado and were performed in accordance with published National Institutes of Health guidelines.

Biochemical Analysis
Liver triglycerides were measured using a 2:1 chloroform:MEOH extract of liver homogenate using a kit from Diagnostic Research Inc. Protein concentrations were determined using a modified Lowry Protein Assay from Bio-Rad (Hercules, CA). Blood ethanol concentrations were determined from freshly isolated serum as previously described. [25,26] Microarray analysis: For microarrays, total RNA was extracted from fresh frozen pooled tissue isolated from triplicates of PF/EtOH-fed PTEN f/f /PTEN LKO mice. Following transcription, cDNA was processed and analyzed as previously described [22,27].

Statistical Analysis
Relative densitometry of Western blots was quantified using ImageJ (http://rsb.info.nih.gov/ij/). All data and statistical analysis was performed by two-way Analysis of Variance or a students t-test using Prism 5 for Windows (GraphPad Software, San Diego, CA). All data are expressed as mean +/-S.E.M. and p values <0.05 were considered significant.

Results
Chronic EtOH consumption increases Akt phosphorylation in WT mice but has no further effect in PTEN LKO mice To verify deletion of PTEN in our model, hepatic cytosolic extracts were prepared from PF and EtOH-fed PTEN f/f and PTEN LKO mice. As shown in S1A and S1B Fig, EtOH consumption decreased overall PTEN expression and increased Akt phosphorylation in PTEN f/f mice. In PF PTEN LKO mice, PTEN expression was decreased by greater than 95% and cytoplasmic pSer 473 Akt levels significantly increased by 7-fold above normal. Addition of EtOH did not significantly affect total cytosolic levels of or phosphorylation of the cytosolic Akt in the PTEN LKO mice. Combined these data support constitutive activation of Akt's in PTEN LKO mice irrespective of EtOH addition.

Deletion of PTEN confers significant protective effects against EtOH induced liver injury. Effects of PTEN LKO and EtOH on hepatocellular function
We have previously demonstrated the consumption of a diet rich in polyunsaturated fatty acids exerts an additive effect with respect to hepatocellular damage in PTEN LKO mice. The data presented in Table 1 presents the effects of either a PF diet or PTEN LKO on EtOH induced hepatotoxicity. In the PTEN LKO PF animals, serum ALT increased 10.7-fold when compared to PTEN f/f animals. In the PTEN f/f animals, chronic EtOH resulted in a 2.18-fold increase in serum ALT, a result comparable to values obtained using the C57BL6/J strain [22]. Interestingly, EtOH ingestion in the PTEN LKO background resulted in a significant (p<0.05) 42% decrease in ALT relative to the PTEN LKO pair fed group. Comparing liver to body weight ratios, EtOH consumption significantly increased liver/body weight in the PTEN f/f group. Compared to PF PTEN f/f mice, liver to body weight was 3.2-fold higher in the PTEN LKO PF animals. The addition of EtOH did not have a further effect on liver to body weight ratio in PTEN LKO animals. As expected, under conditions of sustained lipid synthesis, hepatic triglycerides were elevated by 6-fold in the PF PTEN LKO animals compared PF PTEN f/f groups [12,29]. In the EtOH-fed groups, a significant increase in hepatic triglycerides in occurred in PTEN f/f animals but triglycerides significantly decreased in EtOH-fed PTEN LKO animals when compared to PF PTEN LKO animals (p = 0.05).
Overall, 2-way ANOVA indicated a significant interaction between PTEN LKO and EtOH with respect to ALT, liver triglycerides and liver to body weight. In summary, when compared to the PTEN LKO PF group, chronic EtOH consumption results in decreased hepatocellular damage as evidenced by decreased ALT and hepatic triglycerides in PTEN LKO mice.

Chronic EtOH consumption decreases periportal steatosis in PTEN LKO mice
Given the finding that the addition of EtOH resulted in a decrease in ALT, the effects of EtOH in the PTEN LKO background was further explored with respect to hepatocellular pathology using histology. Using hematoxylin and eosin staining, at low magnification (100X), mild steatosis occurred in EtOH-fed PTEN f/f mice that was barely visible (Fig 1 panels A-D). Steatosis was much more significant in liver sections from PTEN LKO PF mice but steatosis decreased with EtOH (Arrows Panel D). To better assess specific changes in hepatic pathology a higher magnification was employed. At higher magnification (400X) (Fig 1 panels E-H), formation of mild steatosis is much more visible following consumption of EtOH in PTEN f/f mice. In the PTEN LKO mice, consumption of the high fat diet in the PF group induced severe panlobular steatosis. In agreement with hepatic triglyceride accumulation, EtOH consumption by the PTEN LKO mice decreased steatosis primarily in the periportal region which parallels the observed decrease in hepatic triglycerides presented in Table 1. As an additional method to support the change in periportal lipid accumulation, tissues sections were stained with adipophilin. As shown in Fig 1 Panels I-L, compared to the PTEN LKO PF group, adipophilin staining decreased in the periportal region of PTEN LKO EtOH mice. These data support decreased periportal lipid accumulation in EtOH-fed PTEN LKO mice.
To further explore differences in overall pathology in PTEN LKO PF/EtOH fed mice, a modified Kleiner Scoring was performed on hematoxylin and eosin stained hepatic tissue sections [30]. As shown in Table 2, with the exception of a mild increase in steatosis, PTEN f/f PF/EtOH groups did not exhibit significant differences in overall hepatic pathology. Not surprisingly, both PF and EtOH-fed PTEN LKO groups exhibited increased biliary hyperplasia [12,13]. Although there was an increasing trend, at least by pathology, steatosis was not significantly increased by EtOH in the PTEN f/f control groups. Steatosis in the PTEN LKO PF group was extensive and panlobular. Addition of EtOH resulted in a significant decrease in steatosis primarily in zone 1. Both PTEN LKO groups exhibited macro and microsteatosis. Overall, the combined modified Kleiner score was significantly decreased by 6-weeks consumption of EtOH in PTEN LKO mice. An examination by Two-Way analysis of variance revealed a significant interaction between the PTEN LKO genotype and EtOH with respect to overall hepatic steatosis and overall modified Kleiner score.
In a previous report, PTEN LKO mice develop fibrosis by 40 weeks [12,13]. Although the mice used in the present study were only 18 weeks old at study completion, we hypothesized that EtOH may also affect fibrosis. Therefore, we examined the effects of 6 weeks of EtOH consumption on fibrosis using Picrosirius red staining [31]. To eliminate showing "selected" fields of the liver, lower magnification are presented in Panels M-P (white light) and lower panels ((Panels Q-T) polarized light). Chronic EtOH consumption did not result in an increase in picrosirius red staining in the PTEN f/f animals. In PF PTEN LKO , only mild fibrosis was present Liver Specific PTEN Deletion Is Protective against EtOH-Induced Liver Damage compared to PF PTEN f/f mice. Following 6-weeks consumption of EtOH, no significant differences were evident in PTEN LKO mice (quantification not shown), indicating that fibrosis as evidenced by collagen deposition was not affected by EtOH consumption in PTEN LKO mice.

Effects of PTEN LKO and EtOH on ADH and ALDH2 expression
When ingested, EtOH is first metabolized by alcohol dehydrogenase 1 (ADH1) forming acetaldehyde which is then further metabolized by aldehyde dehydrogenase 2 (ALDH2) to produce acetate [32]. To determine the effects of PTEN LKO on EtOH metabolism, expression of ADH1, and ALDH2 was examined. As shown in Fig 2A and 2B, EtOH ingestion significantly increased ALDH2 expression in PTEN f/f and PTEN LKO mice. Chronic EtOH consumption had no effect on ADH expression in the PTEN f/f mice but in PTEN LKO , a significant increase was present in both PF and EtOH mice. This suggested that metabolism of EtOH might be increased in PTEN LKO mice. Therefore, overall blood ethanol concentrations (BEC) were examined using serum isolated from each group. In EtOH-fed PTEN f/f and PTEN LKO mice, BEC was increased (PTEN f/f 137.38±66.69, PTEN LKO 120.25±26.48) but no significant differences were evident between the two genotypes indicating that metabolism of EtOH is not significantly affected by hepatospecific deletion of PTEN.

Consumption of EtOH does not increase protein carbonylation in PTEN LKO mice
Chronic EtOH consumption results in increased accumulation of lipid aldehyde modified hepatic proteins (protein carbonylation) [19,33]. To examine the effects of constitutive Akt activation on protein carbonylation, liver sections prepared from PTEN f/f and PTEN LKO pairs were examined for expression of Cyp2E1, acrolein, 4-HNE and MDA via immunohistochemistry (Fig 3). Expression of Cyp2E1 was clearly elevated in the centrilobular region of both genotypes following EtOH consumption. Examining protein carbonylation, in PTEN f/f animals, EtOH-induced a periportal increase in post-translational modification of proteins by oxidative stress induced lipid peroxidation products acrolein, MDA and 4-HNE. Surprisingly, in the

GSH homeostasis is increased in PTEN LKO mice
In mice, 6-weeks consumption of EtOH results in a 30% decrease in total hepatic GSH concentrations [34]. By its ability to conjugate GSH with reactive aldehydes, glutathione S-transferase A4 (GSTA4) represents a primary mechanism for regulation of hepatic protein carbonylation [35][36][37][38]. Overall these data support increased GSH homeostasis in PTEN LKO mice. KEGG pathways analysis was used to gain addition understanding of the ramifications of changes in the genes presented in Fig 4. From the data presented in S1 Table, hepatospecific deletion of PTEN significantly affected glutathione metabolic and xenobiotic detoxification pathways. Effects of PTEN LKO and EtOH on overall glutathione oxidation and reduction (redox) capacity In hepatocytes, reactive aldehydes are removed via conjugation to GSH [39]. Oxidation of glutathione (GSSG) occurs under conditions of oxidative stress and a significant decrease in the ratio of GSH:GSSG is an accepted marker of increased oxidative stress [40,41]. In wild type murine models, chronic addition of EtOH results in a decrease in total reduced GSH but no significant change in GSSG [34]. We hypothesized that in PTEN LKO mice, an increase in GSH contributes to decreased carbonylation following chronic EtOH consumption. From Fig 5A, EtOH decreased GSH by 30% in the PTEN f/f animals, a result similar to data previously obtained in WT C57BL6/J mice [16]. Comparing both genotypes, deletion of PTEN significantly increased GSH and GSSG. In the PTEN LKO model, EtOH slightly increased hepatic GSH but did not affect GSSG. Examining redox status, the ratio of GSH:GSSG significantly decreased in PTEN f/f mice following EtOH. This effect was reversed in PTEN LKO mice. Compared to either group of PTEN f/f mice, cellular redox ratios significantly decreased in the PTEN LKO mice. Using 2-way ANOVA, genotype effects were evident for all three parameters and a significant interaction occurred with respect to GSH and redox ratio. Combined with data presented in Figs 1, 2 and 3, these data indicate that in PTEN LKO mice there is increased GSH, enhanced mRNA expression of glutathione S-transferases correlating with mitigation of protein carbonylation, decreased ALT and decreased hepatic triglycerides.

Protein glutathionylation is increased in PTEN LKO mice
In the PTEN LKO mice, there is a significant increase in GSH and GSSG. We hypothesized that increased cellular GSH would also result in an increased in protein glutathionylation. As Liver Specific PTEN Deletion Is Protective against EtOH-Induced Liver Damage shown in Fig 5B, in the PTEN f/f group, protein-SSG is increased around the central vein following consumption of EtOH indicating an increase in oxidative stress. Examining PF/EtOH fed PTEN LKO mice, staining of protein SSG was increased when compared to respective PTEN f/f controls but no effect was evident with respect to EtOH.

Discussion
Steatosis is an early consequence of ALD as well as NASH. Given that in the Western world, obesity and NASH are rapidly increasing it is critical to understand that combinatorial effects of alcohol and NASH. Recent research has clearly demonstrated that a major regulator of hepatocellular lipid accumulation is the PTEN/Akt pathway [12]. We previously demonstrated that in C57BL/6J WT mice, chronic consumption of EtOH increases PTEN phosphorylation, carbonylation and decreases PTEN expression [19]. This translated to an increase in Akt activation contributing to the formation of steatosis [19,42]. To further elucidate the contribution of PTEN/Akt signaling in EtOH-induced hepatocellular toxicity, we utilized PTEN LKO mice as a model of increased Akt activation and preexisting steatosis [11,15,19]. Our initial hypothesis was that hepatocellular damage would not increase following EtOH consumption and that concurrent steatosis in PTEN LKO mice in conjunction with EtOH would exert an additive effect with respect to hepatic triglyceride accumulation. From the results obtained in this study, chronic consumption of EtOH reversed high fat diet induced increases in hepatocellular damage as evidenced by decreased ALT and also decreased hepatic triglycerides. Compared to the pair-fed PTEN f/f genotype, pair-fed PTEN LKO mice displayed a dramatic 3-fold increase in overall liver weight, an increased liver:body weight ratios. Hepatic triglycerides and ALT both increased by 10-fold accompanied by a dramatic reduction of GSH/GSSG ratio. These data demonstrate that the high polyunsaturated fat PF diet produces significant liver injury by itself in the PTEN LKO mice and is associated with increased oxidative stress and inflammation. In the PTEN f/f animals, chronic EtOH consumption resulted in a mild but significant increase in hepatic triglycerides, steatosis and hepatocellular damage as shown by increased ALT as well as by increased protein carbonylation. In the PTEN LKO model, surprisingly, chronic EtOH addition resulted in a significant decrease in ALT, periportal steatosis and hepatic triglycerides. Furthermore, compared to PF PTEN LKO , an increase in protein modification by reactive aldehydes did not occur. This is in agreement with previous data that demonstrates that PTEN LKO mice are resistant to additional hydrogen peroxide-induced oxidative stress [43]. Furthermore, in chow-fed PTEN LKO mice, basal oxidative stress is increased when compared to WT controls [44].
In the PTEN LKO PF group, expression of GSH metabolizing enzymes (GST's, GCLC, GSS) are increased when compared to PTEN f/f controls. This corresponds to increased GSH and GSSG and indicates a plausible mechanism for resistance to EtOH-induced oxidative stress. In a previous report, expression of GSTm6 was decreased in PTEN LKO mice whereas we find no significant change [14]. In that report, array analysis was performed at 10 weeks of age and the authors hypothesize that downregulation of GSTm6 may contribute to an increase in inflammation that occurs after 10 weeks of age. Our data originated from 18 week old PTEN LKO mice and demonstrate no significant change in GSTm6 expression in PTEN LKO PF groups when compared to PTEN f/f PF groups. In the previous study, analysis did not report differences in other GST isoforms. In this study, we find that GSTa1/2/3/4/5 and GSTm3 are all upregulated in PF PTEN LKO mice when compared to PF PTEN f/f mice indicating an enhanced response to diet induced inflammation. Chronic EtOH challenge however, decreases GSTm6 expression and decreased GSTA isoforms suggesting that inflammation may be decreased. In a recent publication, PTEN LKO mice were protected against endotoxemia [21]. In that study the authors determined that heme oxygenase was upregulated by PPARγ. We do not see HO-1 upregulation in this study but we do see increased PPARγ in the PTEN LKO model [15]. A difference between our study, is that Guenzl et al, used mice younger than 12-weeks for their studies and there are some reports that suggest that 12 weeks of age is necessary for the full Cre-recombination and PTEN deletion to take effect [12][13][14]45].
Interestingly, using the same model, 6-weeks consumption of alcohol and WT SV and C57BL/6J mice, protein adducts are increased in the periportal region [19,34,46]. This also occurs in the PTEN f/f controls. In this study, protein-SSG is only increased around the central venous region in PTEN f/f mice but is increased panlobularly with EtOH in the PTEN LKO model. This suggests that glutathionylation may also prevent carbonylation of proteins. Recent evidence suggests that GSTμ by its protein-SSG regulatory function, exerts a positive influence against ER stress [47]. In support of this mechanism, both GSTμ and protein-SSG are elevated in PF PTEN LKO mice when compared to PF PTEN f/f controls. Increased GST activity exerts a protective effect in NASH [48]. Post-translational modification of cysteine residues occurs in both glutathionylation and carbonylation [48][49][50][51]. Increased glutathionylation would protect critical cysteine residues by preventing carbonylation in a reversible mechanism. Future studies will be necessary to fully elucidate the impact of increased glutathionylation and inflammation in PTEN LKO mice chronically fed EtOH and to determine clinical relevance.
In conclusion, this study examined the effects of chronic EtOH consumption in the background of constitutively activated de novo lipogenesis, increased steatosis and increased cellular respiration. The data obtained provide new insight into the hepatocellular outcomes of preexisting steatosis due to increased Akt activation during alcohol consumption [19]. In PTEN LKO mice, EtOH consumption did not exacerbate hepatocellular damage suggesting that an environment of increased hepatocellular glutathione concentration may be protective. We hypothesize that reduced injury evident in this study is in part from enhancement of protective oxidative stress responses due to increased cellular respiration occurring during constitutive Akt activation (PTEN LKO ) (S1 Fig). This preexisting condition in PTEN LKO mice results in an alteration of glutathione homeostasis creating a compensatory hepatocellular environment primed to mitigate increased oxidative stress and hepatocellular damage by EtOH. Alternatively, in the background of increased de novo lipogenesis, the addition of EtOH mitigates the effects of a high fat diet. Additional studies will be necessary to determine the sustainability these compensatory mechanisms under conditions of long-term ethanol ingestion and to determine proteins downstream of PTEN that contribute to the protective effect.  S1A Fig (actin normalized). Data are means± SEM as analyzed by students ttest (PF/EtOH) and two-way ANOVA with a Bonferroni post hoc analysis (PTEN f/f group compared to PTEN LKO group) (N = 3 mice/group ( Ã p<0.05, ÃÃÃ p<0.001)). (DOCX) S1 Table. Pathway analysis of up/downregulated oxidative stress proteins in PF/EtOH fed PTEN f/f /PTEN LKO mice. Proteins identified in Fig 4 were examined using KEGG pathway analysis as previously described [49]. (XLSX)