PROTECTIVE EFFECTS OF HYPOTHALAMIC BETA-ENDORPHIN NEURONS AGAINST ALCOHOL-INDUCED LIVER INJURIES AND LIVER CANCERS IN RAT ANIMAL MODELS

Abstract

Background

Recently, retrograde tracing has provided evidence for an influence of hypothalamic β-endorphin (BEP) neurons on the liver, but functions of these neurons are not known. We evaluated the effect of BEP neuronal activation on alcohol-induced liver injury and hepatocellular cancer.

Methods

Male rats received either BEP neuron transplants or control transplants in the hypothalamus and randomly assigned to feeding alcohol-containing liquid diet or control liquid diet for 8 weeks or to treatment of a carcinogen diethylnitrosamine (DEN). Liver tissues of these animals were analyzed histochemically and biochemically for tissue injuries or cancer.

Results

Alcohol-feeding increased liver weight and induced several histopathological changes such as prominent microvesicular steatosis and hepatic fibrosis. Alcohol feeding also increased protein levels of triglyceride, hepatic stellate cell activation factors and catecholamines in the liver and endotoxin levels in the plasma. However, these effects of alcohol on the liver were reduced in animals with BEP neuron transplants. BEP neuron transplants also suppressed carcinogen-induced liver histopathologies such as extensive fibrosis, large focus of inflammatory infiltration, hepatocelluar carcinoma, collagen deposition, numbers of preneoplastic foci, levels of hepatic stellate cell activation factors and catecholamines, as well as inflammatory milieu and the levels of NK cell cytotoxic factors in the liver.

Conclusion

These findings are the first evidence for a role of hypothalamic BEP neurons in influencing liver functions. Additionally, the data identify that BEP neuron transplantation prevents hepatocellular injury and hepatocellular carcinoma formation possibly via influencing the immune function.

INTRODUCTION

Alcoholic liver disease (ALD) is a serious health problem among alcoholics, characterized by the development of fibrosis and cirrhosis coupled with a state of chronic, low-grade inflammation. Chronic alcohol consumption is considered as one of the well-defined risk factors for hepatocellular carcinoma (HCC), the most common type and lethal malignancy of the liver (Warren and Murray, 2013). Fibrosis and its end-stage consequences are commonly accepted as the precursors to cirrhosis and nodule formation in the liver, which may lead to altered hepatic function (De Minicis et al., 2013).

Many central nervous system (CNS) regions including hypothalamus integrate peripheral signals delivered by the blood (e.g., hormones, cytokines) and by neuronal input from peripheral organs (Uyama et al., 2004). Studies have shown that liver receives sympathetic and parasympathetic innervations from the hypothalamus and neurons descents from paraventricular nucleus (PVN) of the hypothalamus play a key role in regulating cellular dynamics of the liver and modulate many physiological processes, such as regulation of hepatic glucose production (Kalsbeek et al., 2010; Swanson and Sawchenko,1980). However, little is known about the functional role of specific hypothalamic neural populations in controlling liver functions in response to pathologies.

One neuronal population, which produces Proopiomelanocortin (POMC) gene and gives rise to various peptides, including BEP, is originated from the arcuate nucleus of the hypothalamus and distributed throughout the PVN. Recent evidence suggests that these BEP neuronal populations project indirectly through multisynaptic pathways to the liver (Stanley et al., 2010). Also, neurons of PVN send projections to different brain regions including the brain stem, anterior pituitary, and the median eminence to mediate physiological, autonomic and endocrine functions (Tasker and Dudek, 1991). Many indirect evidences supports the possibility that sympathetic nervous system (SNS) hyperactivity, in response to various insult of the liver may be involved in the aetiopathogenesis and progression of fibrosis, cirrhosis and liver cancer (Oben and Diehl, 2004). It has also been shown that SNS affects the function of both hepatic stellate cells (HSCs) and hepatic oval cells (OCs) and therefore directly regulates liver repair, and overactivity of SNS neurotransmitters initiates and perpetuates hepatic fibrosis, whereas inhibitors of SNS reduce experimentally induced liver fibrosis (Dubuisson et al., 2012).

Our previous studies shows that transplantation of neural stem cells derived BEP neurons into the PVN prevented prostate and mammary cancer growths and metastases by activating innate immune system and reducing inflammatory milieu through the stimulation of parasympathetic and inhibition of sympathetic neuronal functions (Sarkar et al., 2008; 2011). Therefore, the purpose of the present study was to assess the influence of transplanted BEP neurons in the PVN on alcohol-induced liver injury and on diethylnitrosamine (DEN)-induced rat hepatocellular carcinoma.

MATERIALS AND METHODS

Implantation of BEP neurons into the PVN of the hypothalamus

Pregnant female Sprague-Dawley rats and adult male Fisher-344 rats were obtained from Charles River, (Wilmington, MA) and were maintained in a controlled environment with a 12 h light/dark cycle. Animal care was performed in accordance with institutional guidelines and complied with National Institutes of Health policy. Neural stem cells from 17 days old fetal rat brains of Sprague-Dawley rats were obtained and then differentiated to BEP neurons in cultures to use in this study. To control for transplantation, cortical cells prepared from 17-day-old fetal rat brains were used. These cell preparations and justifications for use are previously described (Sarkar et al., 2008; 2011). On Postnatal day 60, male Fisher-344 rats were anesthetized and injected with cortical neurons (control) or BEP neurons (20,000 cells/μl/each side) in both sides of PVN of the hypothalamus using stereotactic procedures as previously described (Sarkar et al., 2008). The animals were allowed to recover for 2 weeks from the surgery and then assigned to alcohol induced liver injury or liver cancer studies.

Chronic alcohol induced liver injury model

Male rats with control or BEP neuronal transplants were randomly assigned to a liquid diet containing ethanol (AF) or pair-fed an isocaloric liquid diet (PF) (Bio-Serv, Frenchtown, NJ). To habituate the rat with an alcohol diet, they were fed a liquid diet containing 1.7% (v/v) and 5.0% (v/v) ethanol for 4-days. Subsequently, the animals were fed with 6.7% (v/v) of ethanol providing about 35% of the total dietary calories for 8 weeks. The amount of isocaloric liquid diet presented to rats in the PF group was determined from the average amount of alcohol diet consumed by AF animals on the previous day. Freshly prepared liquid diets were presented to experimental animals daily between 1600 and 1700 h, and the intake amount by each animal was recorded daily. Rats of AF and PF groups consumed 105.6 ± 3.3 ml/day and 100.2 ± 3.1 ml/day liquid diet daily. Both control neuron transplanted and BEP neuron transplanted animals consumed similar volume of alcohol diet (~106 ml/day). The blood alcohol concentration 2h after introduction of freshly prepared diet at the dark phase was similar in both control neuron transplanted and BEP neuron transplanted animals and were between 120 and 150 mg/dl.

Diethylnitrosamine (DEN) induced hepatocellular carcinoma (HCC) model

Adult male rats with neuronal transplants were given a single intraperitoneal (ip) injection of DEN (200 mg/kg in saline; Sigma Chemical Company, St Louis, MO) followed by two weeks after intragastric (ig) administrations of 0.02% 2-acetylaminofluorene (2-AAF in 0.5% carboxymethyl cellulose; Sigma) three days per week for 13 weeks. Control animals were injected with vehicle (0.9% saline ip once and 0.5% CMC ig for 13 weeks). Animals were left with out any treatment for additional 3 weeks and then euthanized for tissue and blood samples collections for biochemical and/or histochemical analysis.

Blood, tissue collection and analysis of histopathological and fibrotic changes of the liver

At the end of treatments (8 weeks for alcohol toxicity and 18 weeks for liver cancer), animals were sacrificed and trunk blood samples were collected into sterile endotoxin-free EDTA-containing tubes for separating plasma. Liver samples were removed and snap-frozen in liquid nitrogen and stored at −80°C or formalin-fixed (10% neutral buffered formalin) for histological analyses. The fixed tissues were dehydrated in graded ethanol, embedded in paraffin, sectioned at 5 mm and collected on slides for H&E, Sirius red and Massson's trichrome staining. H&E stained slides were used for histopathological evaluations by two investigators blinded to treatment. Fibrosis was semi quantitatively scored on Sirius red and Massson's trichrome stained sections as absent (0), mild (1), moderate (2) or severe (3).

Immunohistochemical and Immunoblotting of proteins

For immunohistochemistry 5 μm sections of liver were stained using the ABC Elite Vectastain kit (Vector Labs, Burlingame, CA) according to manufacturer's instructions using various primary antibodies: tumor necrosis factors (TNF-α; 1: 200), cell proliferation-associated antigen (Ki-67; 1: 200), nuclear factor κB(NF-κB; 1: 250) and alpha smooth muscle actin (α-SMA; 1:300) were all from Abcam, Cambridge, MA, and glutathione S-transferase pi (GST-pi; 1:300) from MBL International, Watertown, MA. Five microscopic fields per animal were photographed using Nikon-TE 2000 inverted microscope (Nikon Corp., Tokyo, Japan) and used for evaluation. Intensity of staining was categorized as negative (−), weak (+), moderate (++) and strongly positive (+++). The percentage of NFκB– nuclear positive cells was calculated from the number of cells showing brown staining in the nucleus divided by the total cell number of cells × 100. Data are means ± SEM obtained from three serial sections of 5-6 animals in each group.

Liver samples were homogenized and used for isolation of cytosolic and nuclear fraction extracts as described previously (Lee et al., 2004). For Western Blotting, tissue extracts (50 μg) and Western blot chemiluminescence reagent (Thermo Fisher Scientific Inc., Rockford, IL) were used. Primary antibodies used were: anti-perforin rabbit polyclonal (1:250), anti-granzyme B mouse monoclonal (1:200), and anti–IFN-γ mouse monoclonal (1:250) all from Santa Cruz Biotechnology, Inc., (Dallas, TX); NF κB (1: 500), anti-lamin B1 antibody (1:2000) were from Abcam (Cambridge, MA). and anti-β-actin (1:5000) from EMD Millipore (Billeries, MA). The densitometry was performed using Image J analysis software (National Institutes of Health, Bethesda, MD). Each protein value in the total or cellular extract was normalized with the β-actin value and in the nuclear extracts was normalized with lamin B1 value.

Plasma endotoxin and liver triglycerides, epinephrine and norepinephrine levels

Plasma endotoxin levels were determined by an end point Limulus Amebocyte Lysate (LAL) assay kit with a chromogenic substrate (HyCult biotechnology, Uden, The Netherlands). Liver triglycerides (TG) were measured with commercially available TG colorimetric assay kit (Cayman chemical company, Ann Arbor, MI). The liver supernatants were extracted as described (Mizelle et al., 1988) and used for epinephrine/norepinephrine assay by using BICAT® Adrenaline & Noradrenaline ELISA Kit (Eagle Biosciences, Inc., Nashua, NH). Results were normalized to liver weight.

Statistical analysis

The data represented in figures are mean ± SEM, Differences between the groups were analyzed using one-way ANOVA with the Newman-Kuel's post-test. The chi-squared test was used to analyze the difference in tumor incidence at the termination of experiment. A value of P<0.05 was considered to be significant.

RESULTS

BEP neuron transplants in the hypothalamus suppress alcohol effects on liver weight and triglyceride content

Alcohol feeding markedly increased the liver weight in animals with control neuron transplants (Table 1). This growth promoting effect of alcohol was suppressed by BEP neuronal transplantation. Alcohol feeding also increased triglycerides contents in liver of control neuron transplanted groups at a higher level than in BEP neuron transplanted groups. These data suggest that BEP neurons exhibited an ability to counteract the ethanol-induced changes in the liver weight and triglyceride content.

Table 1.

Liver weight and triglycerides content in animal pair-fed liquid diet (PF) or alcohol containing liquid diet (AF) and transplanted with BEP neurons or control cell transplants.

Treatment	Liver weight (g wet tissue)	Liver triglyceride (mg/g wet tissue)
PF + control	6.18 ± 0.08a	2.08 ± 0.14a
PF+ BEP	6.21 ± 0.08a	1.8 ± 0.14a
AF + control	10.24 ± 0.15b,*	5.82 ± 0.37b,*
AF + BEP	8.13 ± 0.11c,*	3.28 ± 0.11c,*

Data are means ± SEM obtained from 5 animals in each group.

^aP < 0.05 significantly different from those groups not sharing a common superscript letter.

^bP < 0.05 significantly different from those groups not sharing a common superscript letter.

^cP < 0.05 significantly different from those groups not sharing a common superscript letter.

^*P < 0.01 significantly different from PF treated groups.

BEP neuron transplants in the hypothalamus reduces alcohol-induced histopathologies in the liver

Histological examination of liver by H&E staining showed normal structures in the liver of PF animals with control neurons (Fig. 1A) or BEP neurons transplants (Fig. 1B). Alcohol feeding produced micro and macro vesicular steatosis (accumulation of fats) in the hepatocytes around the central vein (Fig. 1C) and dense inflammatory cell infiltration cells near the portal area (Fig. 1D) in the liver of rats with control transplants. Interestingly, alcohol feeding in BEP neurons transplanted rats produced no steatosis (Fig. 1E) but produced mild infiltration of inflammatory cells (Fig. 1F). Examination of Sirius red and Masson's trichrome stained sections of liver revealed a minimal collagen fibril deposition in PF animals that had control transplants (Fig. 1G, K) or BEP cells transplants (Fig. 1H, L). AF animal treated with control transplants had abundant labeling of collagen not only in the portal area (Fig 1M) but also in the parenchyma to form a mesh-like pattern of fibrosis with collagen fibrils completely surrounding single hepatocytes (Fig 1I). In BEP cells transplanted AF animals, the liver tissue showed normal centrolobular vein (Fig 1N) and only mild fibrotic changes around the hepatocytes of the parenchyma (Fig 1J).

Figure. 1.

Effects of BEP neuron transplants in the hypothalamus on alcohol-induced histopathologies the liver. (A-F) Representative photomicrograph of H&E-stained liver sections from pair-fed (PF) and alcohol-fed (AF) rats transplanted with cortical neurons (cont) or BEP neurons (BEP). Liver sections showing normal morphology of centrolobular veins (A, B, E), huge micro and macro vesicular steatosis (accumulation of fat, arrows) around the central vein (C), inflammatory cell infiltration in the portal area (arrow) (D) and in the parenchyma (F). (G-N) Representative photomicrographs of Sirus red and Masson's trichrome stained liver sections of similarly treated animals showing normal collagen staining (G, H, K, L), severe fibrosis around the centrolobular vein (M) and form a mesh-like pattern of fibrosis with collagen fibrils completely surrounding single hepatocytes (peri-hepatocyte; I). Whereas, mild fibrotic changes around the centrolobular vein (N) and in peri-hepatocyte (arrow) (J). All images are at 20x magnification except those inside boxes that are at 40x.

BEP neuron transplants in the hypothalamus suppress alcohol effects on hepatic stellate cell activation factors in the liver

Because, previous studies indicated that ethanol-induced steatosis is associated with HSC activation (Nanji et al., 1997), we examined changes in a HSC activation marker protein a-SMA in livers. PF animals irrespective of cell treatment had very low or absence of α-SMA immunoreactivity in their liver (Fig. 2A,B). In AF animals with control transplants, α-SMA staining is localized in portal fibroblasts (Fig 2 C, solid arrows) and in hepatic stellate cells of the sinusoids (Fig 2C, broken arrows). Whereas, minimal expression of alpha-SMA can been seen in BEP transplanted animals (Fig 2D) even in large portal areas or not seen in sinusoidal hepatic stellate cells. The liver content of α-SMA protein was also markedly increased in AF rats with control transplants but very moderately increased in AF rats with BEP transplants (Fig 2I).

Figure. 2.

Effects of BEP neuron transplants in the hypothalamus on alcohol-induced changes in hepatic stellate cell activation factors in the liver. Animals were treated as described in Figure 1. (A-H) Representative immunohistochemical photographs of liver sections stained for α-SMA (A-D) and TNF-α (E-H). Brown-stained spots are immunopositive cells. All images are at 20x magnification except those inside boxes that are at 40x. (I, J) Western blot analysis of α-SMA (I) and TNF-α (J) levels in livers. (K) ELISA analysis of plasma endotoxin levels. N=6. * P < 0.05, *** P < 0.001, compared to PF groups. ^a, P < 0.001 compared to BEP + AF.

Studies have shown that HSC expresses key enzymes for catecholamine synthesis and HSC-derived catecholamines contribute to the liver injury (Dubuisson et al., 2012). As shown in Table 2, alcohol feeding in animals with control transplants significantly increased liver epinephrine and norepinephrine levels, as compared to PF treated and control cell transplanted animals. BEP neuron transplants suppressed alcohol effects on liver contents of both catecholamines. BEP neuronal transplants moderately reduced liver contents of epinephrine in PF animals.

Table 2.

Liver epinephrine and norepinephrine contents in animal pair-fed liquid diet (PF) or alcohol containing liquid diet (AF) and transplanted with BEP neurons or control cell transplants.

Treatment	Liver epinephrine (ng/g wet tissue)	Liver norepinephrine (ng/g wet tissue)
PF + control	39.11 ± 0.86a	42.61 ± 0.89a
PF+ BEP	32.74 ± 0.3b	45.14 ± 0.11a
AF + control	65.24 ± 0.52c,*	92.12 ± 0.91c,*
AF + BEP	42.01 ± 0.29d,*	75.12 ± 0.89c,*

Data are means ± SEM obtained from 8 animals in each group.

^aP < 0.05 significantly different from those groups not sharing a common superscript letter.

^bP < 0.05 significantly different from those groups not sharing a common superscript letter.

^cP < 0.05 significantly different from those groups not sharing a common superscript letter.

^*P < 0.01 significantly different from PF treated groups.

Simultaneous increases in the hepatic TNF-α and plasma endotoxin levels were reported in alcohol-induced chronic liver diseases and fibrosis (Mizelle et al., 1988; Hoek, 1999). In this study we show that PF animals, irrespective of cell transplants, had minimal number or absence of TNF-α positive cells in the liver (Fig. 2E, F). However, AF treated animals with control cell transplants had many TNF-α positive cells in the liver (Fig. 2G) while AF treated animals with BEP neuron transplants had sparsely populated TNF-α positive cells in the liver (Fig. 2H). Biochemical data verified the immunocytochemical findings and indicated that BEP transplants prevent alcohol-induced increase in TNF-α levels (Fig, 2J). In agreement with the result on TNF-α level, a higher level of plasma endotoxin was observed in AF rats with control transplants (Fig. 2K). Alcohol-induced increase in endotoxin level was attenuated in the animals that had BEP neuron transplants.

BEP neuronal transplantation in the hypothalamus prevents carcinogen-induced histopathologies of the liver

We investigated the influence of BEP neurons on DEN-induced hepatocellular carcinoma, a multistage hepatocarcinogenesis animal model for studying human HCC (Ellinger-Ziegelbauer et al., 2008). Gross inspection of liver after carcinogen treatment indicated that animals with control transplants developed hepatic nodules (60% tumor incidence; N=10, Fig. 3C), but animals with BEP transplants showed no appearance of hepatic nodule (0% tumor incidence; N=16; Fig. 3D). As illustrated by H&E staining, tumors in control transplants-treated group were well-differentiated hepatocellular carcinomas with compressed hepatic vein (Fig. 3G). Carcinogen-treated animals with BEP transplants had almost normal liver and vein architecture (Fig. 3H). The liver of cortical or BEP transplanted rats that are given vehicle also had no tumor (Fig. 3A, B) and the histology was normal (Fig 3E, F). Immunostaining of Ki-67 indicated that the number of proliferating cells in the liver was markedly increased after carcinogen treatment in rats with control transplants and only moderately increased in rats with BEP neuron transplants (Fig. 3K & L), as compared to vehicle-treated controls (Fig 3I, J).

Figure. 3.

Effects of BEP neuron transplants in the hypothalamus on the DEN-induced histopathologies in the liver. (A-D) Representative livers showing normal appearance (A, B,D) or tumors (arrow; C) of animals treated with vehicle and control or BEP cell transplants or DEN and control or BEP cell transplants. (E-H) Representative photomicrographs of H&E-stained liver sections showing well differentiated hepatocellular carcinomas with compressed hepatic vein (G, arrow) or normal liver and vein architecture (H, arrow). (I-L) Immunohistochemical analysis of the cell proliferation marker (Ki-67; brown stained cells; IL), liver fibrosis markers by Sirius red (M-P) and Masson's trichrome staining (Q-T). (U-X) Immunohistochemical analyses of hepatic malignance by glutathione S-transferase pi-class (GST-pi; brown staining). All images were captured at 20x magnification except those inside boxes that are at 40x.

BEP neuronal transplantation prevents carcinogens-induced liver fibrosis, decreases preneoplastic foci, and hepatic stellate cell activation factors levels in the liver

As evaluated by Sirius red and Masson's trichrome staining, livers of rats with control transplants and DEN treatment exhibited severe fibrosis around the centrolobular vein (Fig. 3O, S) and throughout the parenchyma to form a “chickenwire fibrosis” characterized by a mesh-like pattern of fibrosis with collagen fibrils surrounding single hepatocytes (Fig. 3O). Whereas, very low hepatic fibrosis was observed in the liver of BEP neurons transplanted rats treated with carcinogen (Fig. 3P, T). The liver of cortical or BEP transplanted rats that are given vehicle had normal collagen staining (Fig. 3M, Q, N, R). These findings suggest that BEP neurons decreased DEN-induced fibrogenesis and HCC.

Immunostaining of GST-pi, a biomarker of hepatic preneoplastic lesions (Tsuda et al., 2010), revealed that DEN treated animals with control transplants had large foci consisting of GST-pi positive hepatocytes (Fig. 3W). These foci appeared to be evenly distributed throughout the entire section of the liver. Whereas, only few single GST-pi positive hepatocytes were observed in the livers of BEP neurons transplanted rats (Fig. 3X). No GST-pi positive foci or hepatocytes were detected in the livers of vehicle treated animals (Fig 3U,V). These data indicate that the suppression of GST-pi foci formation might be a mechanism by which BEP neurons prevent liver tumor formation.

Immunohistochemical localization of α-SMA, a HSC activation marker protein, revealed that these protein-positive cells were abundant in portal fibroblasts (Fig 4C, solid arrows) and in hepatic stellate cells of the sinusoids throughout the hepatic parenchyma (Fig 4C, broken arrows) after carcinogen treatment in control cells transplanted animals (Fig. 4C). However, α-SMA-labeled cells were found in low abundance only at vascular structures in carcinogen treated and BEP neurons transplanted rats (Fig. 4D). Very low level or absence of α-SMA was detected in the liver of rats that were given vehicle (Fig 4A, B). Western blot measurements of α-SMA contents in livers (Fig. 4M) verified the immunohistochemical data and further provided the evidence that BEP neuronal transplant reduces the ability of DEN to activate HSC in the liver.

Figure 4.

Effects of BEP neuron transplants in the hypothalamus on DEN-induced hepatic stellate cell activation factors and immune markers in the liver. Animals were treated as described in Figure 3. (A-L) Representative photographs of liver sections immunostained for α-SMA (A-D), TNF-α (E-H) and NF-kB (I-L). Brown-stained spots are immuno-positive cells. All images were captured at 20x magnification except those inside boxes that are at 40x. (M, N) Western blot analysis of α-SMA (M) and TNF-α (N) levels in livers. (O) Percentage of nuclear NF-κB–positive cells in liver tissues. (P, Q) Western blot analysis of NF-kB levels in cellular and nuclear fractionates of liver tissues. (R) ELISA analysis of endotoxin level in plasma. (S-U) Western blot analysis of Perforin (S), Granzyme B (T), and IFN-γ levels (U) in livers. N=6. *** P < 0.001, compared to control + carcinogen treated groups. ^(a-c) P < 0.05, values not sharing a common superscript letter are significantly different.

Next, to determine if altered SNS input to the liver might be a part of the protective effect of BEP, we measured liver contents of E and NE (Table 3). The NE content of liver was not different in rats treated with vehicle. The E content was moderately suppressed in vehicle treated and BEP transplanted rats. Both E and NE contents in livers were found to be higher in carcinogen treated animals with control transplants but not with BEP neuron transplants, suggesting that BEP neuron influences the SNS input in the liver during carcinogenesis.

Table 3.

Liver epinephrine and norepinephrine levels in animal treated with with carcinogen or vehicle and transplanted with BEP neurons or control cell transplants.

Treatment	Liver epinephrine (ng/g wet tissue)	Liver norepinephrine (ng/g wet tissue)
Vehicle + control	31.01 ± 0.21a	54.91 ± 2.20a
Vehicle + BEP	27.28 ± 0.71b	49.63 ± 2.61a
Carcinogen + control	72.28 ± 0.78c,*	107.29 ± 3.82b,*
Carcinogen + BEP	55.92 ± 0.91d,*	82.461 ± 2.45c,*

Data are means ± SEM obtained from 8 animals in each group.

^aP < 0.05 significantly different from those groups not sharing a common superscript letter.

^bP < 0.05 significantly different from those groups not sharing a common superscript letter.

^cP < 0.05 significantly different from those groups not sharing a common superscript letter.

^*P < 0.01 significantly different from vehicle-treated groups.

BEP neuron transplants decrease inflammatory milieu during chemically induced hepatocarcinogenesis in rats

Accumulating evidence indicates that TNF-α production and activation of NF-kB-mediated signal transduction play a major role in DEN-induced carcinogenesis (Karin and Greten, 2005). Determination of TNF-α levels in livers revealed that this cytokine-positive hepatocytes are expressed abundantly following DEN treatment in control cell transplanted rats (Fig. 4G) but expressed sparsely in BEP neurons transplanted rats (Fig. 4H). TNF-α positive cells were rarely found in livers of vehicle treated animals (Fig. 4E, F). Western blot measurement of TNF-α contents in livers (Fig. 4N) verified the immunohistochemical data. It has been suggested that in non-malignant conditions, NF-kB is present in the cytoplasm bound to an inhibitory protein, IkB, whereas in response to various stimuli, NF-kB translocates into the nucleus, where it induces transcription of a large variety of target genes that induce inflammation (Karin and Greten, 2005). In the present study, we observed that livers of animals with control transplants had mostly nuclear staining of NF-kB after DEN treatment (Fig. 4I and K; Fig. 40). BEP treatment decreased the amount of nuclear NF-kB staining in DEN treated animals (Fig. 4K and O) but not in vehicle-treated control (Fig. 4I and L; Fig. 4O). Western blot analysis of liver tissues confirms that levels of NF-kB protein were decreased in cytosolic extracts (Fig 4P) with a concomitant increase in nuclear extracts (Fig 4Q) after carcinogen treatment in control cell transplanted rats. BEP transplantation was able to reduce the effect of carcinogen on nuclear translocation (Fig. 4P and Q). These findings indicate that the protective effect of BEP neurons against liver tumor formation is likely by a reduced inflammatory signaling.

Literature suggests that endotoxin promotes chemically induced HCC (Yu et al., 2010). We show here that the plasma level of endotoxin was markedly elevated following DEN treatment in control cells transplanted animals as compared to vehicle treated controls (Fig. 4R). BEP neurons transplants significantly reduced DEN-induced increase in plasma levels of endotoxin. These data suggest that plasma endotoxin may be a critical cofactor in chemically induced hepatocarcinogenesis.

BEP neuron transplants enhances the level of NK cell cytotoxic factors

Experimental evidence suggests that NK cells are important surveillance mechanism for in vivo anti-tumor activity during chemical carcinogenesis (Gillgrass and Ashkar, 2011). Measurements of levels of NK cell cytotoxic proteins (perforin, granzyme B and IFN-γ) in the liver revealed that carcinogen treatment decreased levels of liver perforin (Fig. 4S), granzyme B (Fig. 4T) and IFN-γ (Fig. 4R) in rats with control cells transplants but not with BEP neurons transplants. Thus, these results suggest that NK cells derived cytotoxic factors are modulated by BEP neuronal activity during the hepatocarcinogenesis.

DISCUSSION

It is well accepted that alcohol-induced liver injury is mediated through one or more factors such as accumulation of fat, oxidative damage, proinflammatory cytokines, increased collagen deposition and activation of various non-parenchymal cells (Sarkar and Zhang, 2013). In the present study, we demonstrated that transplanted BEP neurons in the PVN alleviated the detrimental effects of alcohol and DEN-induced lesions. In alcohol-induced liver injury model, BEP neuron transplants reduced liver weight and accumulation of triglycerides and less pathological changes such as infiltration of inflammatory cells and steatosis in the hepatocytes. In the carcinogenesis study, DEN induced liver malignancies and cell proliferations were prevented in rats with BEP neuron transplants, supporting the concept that BEP neuron has an anti-tumor effect (Sarkar et al., 2008; 2011).

Experimental evidence suggests that ethanol-induced and carcinogen-induced liver injuries are mediated through a secondary compensation for the circulatory disturbances that accompany fibrosis and cirrhosis (Lands, 1995; Szabo et al., 2012). Among the effector molecules, simultaneous increase in the plasma endotoxin level and proinflammatory cytokines such as TNF-α play a critical role in the initiation and development of liver injury (Enomoto et al., 2000; 2001). In our study, we found that plasma endotoxin levels, the expression of TNF-α and activated NF-kB (in cancer study) in the liver were significantly lower in BEP neurons transplanted rats. Studies have suggested that enhanced Kuffer cells activity by endotoxin in the liver is the main source of TNF-α production after liver injuries (Hansen et al., 1994; Nath and Szabo, 2009; An et al., 2012). Our results are encouraging and warrant further investigation, including the depletion of KC cells that directly assess the mechanistic role of BEP on liver KC and its involvement in the onset and progression of ALD and HCC. Modulating effects of BEP neurons on liver pathologies in alcoholic liver disease could also be due to their actions on the gut-brain axis. In particular, it may alter the gut permeability to endotoxin and the impact of these changes on immune cell activation in the liver and the interaction of these effects with the “direct” effects of alcohol within the liver (e.g. alcohol metabolism and oxidative stress/ROS/acetaldehyde production). Also, BEP transplantation might modulate GI-function and may also impact the hepatic response to chronic alcohol feeding and/or hepatocarcinogens, which will be expanded in future studies.

In burn patients, blood endotoxin appears to be involved in the suppression of NK cells activity, a major mediator of host defense against infections and tumor cells (Roberts et al., 2007). Our previous study indicated that BEP transplants elevates NK and macrophage cell numbers in PBMCs, suggesting that the transplants also promote migration of these immune cells to the blood and/or to the target tissue to promote host defense (10). This evidence provides another possible explanation of why BEP neurons transplanted animals had lower plasma endotoxin, TNF-α and activated NF-kB expressions in the liver. In liver cancer study, we found that, after carcinogen treatment, BEP neurons transplanted animals had higher levels of NK cells cytotoxic proteins in the liver tissue than cortical cells transplanted animals, providing additional support to the concept that BEP neuron enhances innate immune functions.

Placental form of GST-pi, a putative preneoplastic marker was considered to play a key role in the process of hepatocarcinogensis in rats (Tatematsu et al., 1983). A study shows that NK cell activity of DEN-treated rats progressively decreased with the duration of treatment, and those changes inversely correlated with the induction of preneoplastic hepatic foci (Lee et al., 1999). Interestingly, we found that livers of BEP neurons transplanted animals that did not developed tumors after DEN treatment had higher levels of NK cells cytotoxic proteins and had very few GST-pi positive cells. Whereas, the livers of control transplants treated rats that developed tumors after DEN treatment had lower levels of NK cells cytotoxic proteins and had increased GST-pi positive cells throughout the tissue. These results strongly support the concept that BEP neurons enhances the NK cell cytolytic function and suppresses the progression of hepatocarcinogenesis.

It has been shown that the activation of SNS increases the liver and plasma levels of catecholamines, especially E and NE, there by activating the non-parenchymal cells such as HSC's (Oben et al, 2004). Higher levels of E and NE were found in the liver of control cell transplanted animals compared to BEP neurons transplanted animals in both alcohol injury and liver cancer studies. Since, the levels of E and NE are considered as a marker for SNS tone, our data support the possibility that BEP neurons suppressed the SNS overactivity that causing aetiopathogenesis of the liver. Moreover, the lower levels of a-SMA in the liver in BEP transplanted animals support the concept that SNS overactivity in the liver is involved in the activation of HSC's.

Together, these findings suggest that BEP neurons in the hypothalamus have the ability to suppress hepatocellular injuries and hepatocellular carcinoma formation possibly by influencing the ANS controls of innate immune functions in the liver. Our studies shows that preventive effects of BEP transplantation on liver cancer is very persuasive while its therapeutic effects need to be further investigated in liver cancer models which could have a potential therapeutic use in HCC patients.

List of abbreviation

AD

ad libitum-fed

AF

alcohol-fed

ALD

alcoholic liver disease

a-SMA

alpha smooth muscle actin

BEP

β-endorphin

DEN

diethylnitrosamine

GST-pi

glutathione S-transferase pi-class

HCC

hepatocellular carcinoma

HSC

hepatic stellate cell

NF-kB

nuclear factor kB

PF

pair-fed

PVN

paraventricular nucleus

TNF-α

tumor necrosis factors

SNS

sympathetic nervous system