Available online at www.sciencedirect.com

ScienceDirect Journal of Nutritional Biochemistry xx (2015) xxx – xxx

Diets rich in fructose, fat or fructose and fat alter intestinal barrier function and lead to the development of nonalcoholic fatty liver disease over time☆,☆☆ Cathrin Sellmann a , Josephine Priebs a , Marianne Landmann a , Christian Degen a , Anna Janina Engstler a , Cheng Jun Jin a , Stefanie Gärttner a , Astrid Spruss b , Otmar Huber c, d , Ina Bergheim a,⁎ a

Institute of Nutritional Sciences, Model Systems of Molecular Nutrition, Friedrich-Schiller-University Jena, Jena, Germany b Department of Nutritional Medicine, University of Hohenheim, Stuttgart, Germany c Department of Biochemistry II, University Hospital Jena, Friedrich-Schiller-University Jena, Jena, Germany d Center of Sepsis Control and Care, Jena University Hospital, Jena, Germany

Received 22 December 2014; received in revised form 8 May 2015; accepted 13 May 2015

Abstract General overnutrition but also a diet rich in certain macronutrients, age, insulin resistance and an impaired intestinal barrier function may be critical factors in the development of nonalcoholic fatty liver disease (NAFLD). Here the effect of chronic intake of diets rich in different macronutrients, i.e. fructose and/or fat on liver status in mice, was studied over time. C57BL/6J mice were fed plain water, 30% fructose solution, a high-fat diet or a combination of both for 8 and 16 weeks. Indices of liver damage, toll-like receptor 4 (TLR-4) signaling cascade, macrophage polarization and insulin resistance in the liver and intestinal barrier function were analyzed. Chronic exposure to a diet rich in fructose and/or fat was associated with the development of hepatic steatosis that progressed with time to steatohepatitis in mice fed a combination of macronutrients. The development of NAFLD was also associated with a marked reduction of the mRNA expression of insulin receptor, whereas hepatic expressions of TLR-4, myeloid differentiation primary response gene 88 and markers of M1 polarization of macrophages were induced in comparison to controls. Bacterial endotoxin levels in portal plasma were found to be increased while levels of the tight junction protein occludin and zonula occludens 1 were found to be significantly lower in the duodenum of all treated groups after 8 and 16 weeks. Our data suggest that chronic intake of fructose and/or fat may lead to the development of NAFLD over time and that this is associated with an increased translocation of bacterial endotoxin. © 2015 Elsevier Inc. All rights reserved. Keywords: Endotoxin; Macronutrients; NAFLD; Tight junction; Aging

1. Introduction Nonalcoholic fatty liver disease (NAFLD) is a disease comprising a continuum of liver damage ranging from simple steatosis to nonalcoholic steatohepatitis (NASH) and cirrhosis. A recent survey

Abbreviations: 4-HNE, 4-hydroxynonenal protein adducts; ALT, alanineaminotransferase; Arg-1, arginase 1; iNOS, inducible nitric oxide synthase; IR, insulin receptor; IRS-1, insulin receptor substrate 1; MCP-1, monocyte chemotactic protein 1; MyD88, myeloid differentiation primary response gene 88; NAFLD, nonalcoholic fatty liver disease; NAS, NAFLD activity score; NASH, nonalcoholic steatohepatitis; SOPF, specified and opportunistic pathogen-free; TLR-4, toll-like receptor 4; TNFα, tumor necrosis factor α; ZO-1, zonula occludens 1. ☆ Funded by grants from Bundesministerium für Bildung und Forschung, FKZ: 01EA1305 and FKZ: 01KU1214A (both IB). ☆☆ Conflicts of interest: None. ⁎ Corresponding author. Institute of Nutritional Sciences, SD Model Systems of Molecular Nutrition, Friedrich-Schiller-University Jena, Dornburger Strasse 25, 07743 Jena, Germany. Tel.: +49-3641-949730; fax: +49-3641-949672. E-mail address: [email protected] (I. Bergheim). http://dx.doi.org/10.1016/j.jnutbio.2015.05.011 0955-2863/© 2015 Elsevier Inc. All rights reserved.

reviewing 260 epidemiological studies published in Europe in the last 5 years reported that NAFLD has to be accounted to the most frequent liver diseases in Europe [1]. Furthermore, results of Vernon et al. even suggest that NAFLD is the most common cause of elevated liver enzymes worldwide [2]. Also, it has been suggested that fatty livers, long thought to be a benign state of liver disease, are more vulnerable to injury from various causes increasing the probability to progress to later stages of the disease and further liver-related morbidity and mortality [3,4]. However, despite intense research efforts during the last decades, molecular and pathological changes involved in the onset and even more so the progression of NAFLD are not yet fully understood and therapeutic options are still rather limited. Therefore, a better understanding of the alterations causally involved in the early stages of NAFLD is desirable to improve prevention and intervention strategies. General overnutrition, particularly when being rich in fat and/or sugars like fructose, is being discussed to be key factors in the development of NAFLD [5]. However, results of not only animal-based studies but also some human studies suggest that not only the development of NAFLD in settings of overnutrition may result from the extra energy adding to an enhanced de novo synthesis and storage of

2

C. Sellmann et al. / Journal of Nutritional Biochemistry xx (2015) xxx–xxx

lipids but also other factors may be involved. Indeed, we showed that the development of NAFLD in mice chronically fed a fructose-rich diet is associated with increased endotoxin levels in portal blood and an activation of the toll-like receptor TLR-dependent signaling cascades [6,7]. Similar findings were also reported form NAFLD models exposing animals chronically to a high-fat diet [8,9]. It was further shown that chronic intake of fat and fructose, respectively, is associated with a decrease in tight junction proteins in the small intestine [10,11]. However, if the combination of a chronic exposure to a fat- and fructose-rich diet has an additive or even synergistic effect on the loss of tight junctions, the increased permeation of bacterial endotoxin and subsequently the development of liver damage and if these effects progress with time have not yet been systematically studied. Starting from this background, the aim of our study was to assess the effects of a chronic intake of a fructose, fat or fructose- and fat-enriched diet on intestinal tight junction proteins, portal endotoxin levels and the development of NAFLD over time in mice. 2. Materials and Methods 2.1. Animals and treatment Six- to eight-week-old female C57BL/6J mice (Janvier Labs, Le Genest-Saint-Isle, France), shown before to be more susceptible to fructose-induced NAFLD than male mice [12], were housed in a specified and opportunistic pathogen-free (SOPF) barrier facility accredited by the Association for Assessment and Accreditation of Laboratory Animal Care. All procedures were approved by the local Institutional Animal Care and Use Committee. Mice (n=6 per group) had free access to plain tap water and standard chow (ssniff, Soest, Germany), to water enriched with 30% (w/v) fructose and proteinand micronutrient-enriched chow, to a high-fat diet (30 kJ% fat, 50 kJ% carbohydrates) or water enriched with 30% (w/v) fructose and a high-fat diet (60 kJ% fat, 24 kJ% carbohydrates) for 8 and 16 weeks, respectively. In groups fed the fructose-enriched water, protein- and micronutrient-enriched chow was used to avoid malnutrition as it was shown before that mice fed to a 30% (w/v) fructose solution decrease their chow intake [13]. Body weight as well as consumption of chow and drinking solution was assessed twice weekly throughout the 8 and 16 weeks of feeding, respectively. Mice were anesthetized with 80 mg ketamine and 6 mg xylazine per kilogram body weight by ip injection and blood was collected from the portal vein prior to sacrifice. Portions of liver and small intestine were snap-frozen immediately or fixed in neutralbuffered formalin.

2.2. Clinical chemistry and histological evaluation of liver sections Alanine-aminotransferase (ALT) activity (n=4–6 per group for lack of plasma in some groups) was determined by an Olympus AT200 Autoanalyzer (Olympus Europa Holding GmbH, Hamburg, Germany) using commercially available kits (Beckman Coulter Biomedical GmbH, Krefeld, Germany). Furthermore, paraffin-embedded sections of liver (5 μm) were stained with hematoxylin and eosin to evaluate liver histology by scoring photomicrographs captured at original magnifications ×200 and ×400 (Leica DM4000 B LED, Wetzlar, Germany) using the semiquantitative ‘Nonalcoholic Steatohepatitis Clinical Research Network System for Scoring Activity and Fibrosis in Nonalcoholic Fatty Liver Disease’ (modified from Kleiner et al. and Brunt) [14,15]. According to this system, scores were as follows: steatosis grade, 0: b5%, 1: 5– 33%, 2: 34–66%, 3: N66%; lobular inflammation, 0: none, 1: b2, 2: 2–4, 3: 4; hepatocellular ballooning, 0: none, 1: few ballooned cells, 2: many ballooned cells. Neutrophils were stained with naphthol AS-D chloroacetate esterase (Sigma Aldrich Chemie GmbH, Steinheim, Germany) and hematoxylin. To determine means, counting from 8 fields (original magnification ×200) of each tissue section was used.

2.3. Immunohistochemical staining for inducible nitric oxide synthase (iNOS) and 4-hydroxynonenal protein adducts (4-HNE) in liver and tight junction proteins occludin and zonula occludens 1 (ZO-1) in duodenum and ileum Paraffin-embedded liver and duodenal or ileal tissue sections were cut (4 μm) and stained for iNOS, 4-HNE and the tight junction proteins occludin and ZO-1, respectively, using polyclonal antibodies (4-HNE: AG Scientific, San Diego, CA, USA; iNOS: Thermo Fisher Scientific, Waltham, MA, USA; occludin: rabbit anti-occludin, Invitrogen, Camarillo, CA, USA; ZO-1: rabbit anti-ZO-1, Invitrogen, Camarillo, CA, USA) as described previously [13,16]. Using an image acquisition and analysis system incorporated in the microscope (Leica DM4000 B LED, Wetzlar, Germany), the extent of staining in liver sections was defined as percent of the field area within the default colour range determined by the software. To determine means, data from 8 fields (original magnification ×200) of each tissue section were used.

2.4. ELISA Concentration of tumor necrosis factor α (TNFα) was determined in plasma of mice using a commercially available mouse-TNFα ELISA kit according to the instructions of the manufacturer (Assaypro, St. Charles, USA). 2.5. Endotoxin assay Plasma endotoxin levels were measured as detailed before [16]. In brief, samples were heated at 70°C for 20 min and endotoxin levels were determined using a commercially available limulus amebocyte lysate assay with a concentration range of 0.015–1.2 EU/ml (Charles River, L’Arbaesle, France). 2.6. RNA isolation and real-time RT-PCR RNA isolation and real-time RT-PCR have been previously detailed by Kanuri et al. [17]. SYBR Green Supermix (Agilent Technologies, Waldbronn, Germany) was used to prepare the PCR mix. Primers were added to a final concentration of 3 pmol. Primer sequences are shown in Supplementary Table 1. The amplification reactions were carried out in a thermocycler (Agilent Technologies Stratagene Mx3005P, Waldbronn, Germany) with an initial hold step (95°C for 10 min) and 40 cycles of a two-step PCR (95°C for 30 s and 60°C for 60 s). The comparative CT method was used to determine the amount of target, normalized to an endogenous reference (the eukaryotic elongation factor Eef2) and relative to a calibrator (2−ΔΔCt). The purity of PCR products was verified by melting curves and gel electrophoresis. 2.7. Statistical analyses All results are reported as means±S.E.M. One-way analysis of variance with Tukey’s post-hoc test was used for the determination of statistical significance among treatment groups (GraphPad Prism Software, USA). A P≤.05 was selected as the level of significance before the study. Grubb’s test was used to identify outliers before statistical analysis (GraphPad Prism Software, USA).

3. Results 3.1. Effect of a diet rich in fructose, fat or fat and fructose on caloric intake and body weight In Table 1 and Fig. 1A and B, intake of the different diets as well as body weight gain throughout the 8 and 16 weeks long feeding period, respectively, is summarized. Caloric intake and absolute body weight gain of mice fed standard chow and plain water did not differ from that of mice fed the high-fat diet. In line with earlier reports of our group [13], mice fed a 30% fructose solution reduced their caloric intake from chow significantly by ~45% in comparison to mice fed a standard chow. Still, absolute caloric intake was significantly higher in mice fed the 30% fructose solution (8 weeks: 41%, 16 weeks: 36% in comparison to controls), and in comparison to controls, absolute weight gain was ~0.6 g higher after 8 weeks and was ~1.1 g higher after 16 weeks of feeding; however, as weight gain varied considerable between mice, differences did not reach the level of significance. Mice fed a combination of 30% fructose solution and high-fat diet also reduced their chow intake when compared to mice only fed a high-fat diet; however, despite a significantly lower liquid intake in comparison to mice only fed a 30% fructose solution, total caloric intake per mouse and week was significantly higher than that in the control group. Indeed, body weight gain was significantly higher in this group than in all other groups along with a massive accumulation of visceral adipose tissue, both after 8 and 16 weeks of feeding (8 weeks: +~46% and 16 weeks: +~294% in comparison to weight gain of controls) (see Fig. 1A, last column). 3.2. Effect of a diet rich in fructose, fat or fructose and fat on liver status Data summarizing liver status are shown in Fig. 1C and D and Table 1. Chronic intake of a diet rich in fat was associated with a slight accumulation of fat in the liver after 8 weeks of feeding, which was also found after 16 weeks [8 weeks: NAFLD activity score (NAS) + ~3-fold and 16 weeks: NAS + ~3-fold when compared to the respective chow controls, both Pb.05]. In line with these findings, neither liver weight

C. Sellmann et al. / Journal of Nutritional Biochemistry xx (2015) xxx–xxx

3

Table 1 Effect of fructose and/or fat feeding on food intake, weight gain, liver weight, ALT levels and neutrophils. 8 weeks

Food intake (kcal/mouse/week) Chow Liquid

Total Weight gain (g) Liver weight (g) Liver-to-body weight ratio (%) ALT (U/L) Neutrophils (number per microscope field)

16 weeks

C

F

Fat

F+Fat

C

F

Fat

68.4±1.7 Intake of plain tap water was not assessed. 68.4±1.7 3.6±0.3 1.2±0.1 5.7±0.2 22.7±2.2 0.9±0.1

36.8±1.1a,c,d 59.5±4.6

73.3±1.7 Intake of plain tap water was not assessed. 73.3±1.7 3.0±0.2 b 1.1±0.0 b 5.2±0.1 36.3±8.6 2.3±0.2 a

72.9±1.5 23.5±1.3 b

72.4±1.2 Intake of plain tap water was not assessed. 72.4±1.2 4.6±0.3 1.1±0.0 5.0±0.1 15.1±4.7 1.3±0.1

33.9±0.9a,c,d 64.7±2.7

71.9±1.8 Intake of plain tap water was not assessed. 71.9±1.8 3.9±0.3 1.1±0.0 b 4.8±0.2 54.8±18.8 2.7±0.3 a

96.7±5.2a,c 4.2±0.1 1.3±0.1 6.4±0.2a,c,d 20.0±4.9 2.6±0.5 a

96.4±1.8a,c 5.3±0.4a,c 1.2±0.1 5.5±0.1 32.3±11.1 2.7±0.4 a

98.2±2.9a,c 5.7±0.4 1.5±0.0 6.2±0.1a,c,d 18.0±0.8 3.3±0.2 a

F+Fat 79.2±1.9a,c 27.9±1.7 b

106.2±3.1a,c 18.0±3.0a,b,c 1.7±0.2a,c 4.8±0.1 28.2±6.9 2.4±0.2a,b

Values represent means±S.E.M. (n=4–6). Significance refers to groups within one time point. C, control; F, fructose. a Pb.05 compared with respective water-fed control mice. b Pb.05 compared with mice fed with 30% fructose solution. c Pb.05 compared with mice fed with high-fat diet. d Pb.05 compared with mice with a combination of 30% fructose solution and high-fat diet.

nor liver-to-body weight ratio differed significantly from that of controls at the two time points. However, number of neutrophils was significantly higher in livers of mice fed a high-fat diet when compared to controls. Mice chronically fed a 30% fructose solution showed a marked accumulation of fat in the liver after 8 weeks of feeding when compared to chow controls. After 16 weeks of feeding, NAS of mice fed the 30% fructose solution was significantly higher not only than that of chow control but also than that of mice fed the high-fat diet. While liver weight did not differ from that of chow-fed animals, liver-tobody weight ratio and number of neutrophils in liver were significantly higher in livers of mice fed the fructose solution than in chow-fed mice, both after 8 and 16 weeks of feeding. In mice fed a combination of a high-fat diet and a 30% fructose solution, NAS was significantly higher than in chow-fed controls and mice only fed a high-fat diet after 8 and 16 weeks of feeding. Indeed, while number of fatty hepatocytes was not different from that found in mice only fed fructose solution, mice fed a combination of fat and fructose were found to display larger fat droplets and signs of hepatic inflammation after 16 weeks of feeding. Similar changes were not found in any of the other groups. Liver weight was also found to be significantly higher in these mice than in controls and mice fed the high-fat diet after 16 weeks of feeding; however, liver-to-body weight ratio was only significantly increased in mice fed fructose in comparison to all other groups. Number of neutrophils in the liver was significantly higher in livers of all feeding groups at both time points studied when compared to controls but did not, with the exception of fat and fat and fructose after 16 weeks of feeding, differ between feeding groups. As even after 16 weeks of feeding, only a beginning of steatohepatitis was found in liver of mice fed fructose and fat and ALT plasma levels did not differ between groups. 3.3. Effect of a diet rich in fructose, fat or fructose and fat on markers of insulin resistance in liver tissue and plasma After 8 weeks of feeding, expression of insulin receptor (IR) and insulin receptor substrate 1 (IRS-1) were significantly lower by ~40% in livers of mice fed a 30% fructose solution, a high-fat diet or 30% fructose solution and a high-fat diet, respectively, when compared to mice fed a standard chow and tap water (see Table 2). However, after 16 weeks of feeding, a similar effect of the different diets on IR mRNA expression was only found in livers of mice fed a 30% fructose solution. Expression of IRS-1 mRNA was still decreased in all groups when compared to standard chow and tap water controls at this time point. TNFα protein

levels in plasma did not markedly differ between groups at any time point as data varied considerable; however, in some groups, we only had plasma samples from 4 mice per group (see Table 2). 3.4. Effect of a diet rich in fructose, fat or fructose and fat on tight junction protein levels in the small intestine and endotoxin levels in portal plasma as well as toll-like receptor 4 (TLR-4) and myeloid differentiation primary response gene 88 (MyD88) mRNA expression in the liver To determine if the damaging effects of the different diets on the liver were related to alterations of intestinal barrier function or an increased translocation of bacterial endotoxins, protein levels of the tight junction proteins occludin and ZO-1 in the small intestine and endotoxin levels in portal plasma were determined. Data are summarized in Fig. 2, Fig. 3 and Supplementary Fig. 1. Protein levels of both occludin and ZO-1, as determined by immunohistochemical staining, were significantly lower in the duodenum of mice fed a highfat diet, a 30% fructose solution or a combination of fat and fructose in comparison to control chow after 8 and 16 weeks of feeding. Similar alterations were not found in ileum of these mice (representative staining; see Supplementary Fig. 1). Endotoxin levels in portal plasma was markedly higher in all dietary intervention groups when compared to chow controls after 8 and 16 weeks of feeding (see Fig. 3A); however, as data varied considerable within groups and only 4 plasma samples were available in some groups, differences only reached the level of significance for the comparison of the highfat group and the chow control group after 8 weeks of feeding and for the high-fat and high-fat and fructose group and the chow control group after 16 weeks of feeding. In line with these findings, expression of TLR-4 mRNA and to a lesser extent of MyD88 mRNA was also markedly higher in livers of mice fed the different diets in comparison to chow controls at both time points studies (TLR-4 mRNA: Pb.05 for all groups in comparison to chow controls at both time points; MyD88 mRNA: 8 weeks: Pb.05 for fructose and the combination of fructose and fat, 16 weeks: Pb.05 for fat and the combination of fructose and fat in comparison to chow controls) (see Fig. 3). 3.5. Effect of a diet rich in fructose, fat or fructose and fat on 4-HNE and iNOS protein levels as well as markers of macrophage polarization in the liver Fig. 4A and B depicts representative pictures of staining of 4-HNE and iNOS protein levels in livers of mice fed the different diets and Fig. 4C and D shows a summary of the densitometric analysis of the

4

C. Sellmann et al. / Journal of Nutritional Biochemistry xx (2015) xxx–xxx

Fig. 1. Effect of the different diets on indices of liver damage and weight gain in mice after 8 and 16 weeks. (A) Representative pictures of abdomen and (B) summary of body weight gain. (C) Evaluation of liver histology using the semiquantitative “Nonalcoholic Steatohepatitis Clinical Research Network System for Scoring Activity and Fibrosis in Nonalcoholic Fatty Liver Disease” (modified from Kleiner et al. and Brunt) and (D) representative photomicrographs of hematoxylin and eosin staining of liver sections (original magnifications ×200 and ×400). Data are shown as means±S.E.M. (n=6). aPb.05 compared with respective water-fed control mice, cPb.05 compared with mice fed with high-fat diet. ⁎Pb.05, ⁎⁎Pb.01, ⁎⁎⁎Pb.001. C, control; F, fructose.

stainings. Staining of 4-HNE in the liver was slightly but not significantly higher in mice fed a high-fat diet when compared to chow controls. In livers of mice fed a 30% fructose solution or a combination of fructose solution and a high-fat diet, levels of 4-HNE

staining were significantly higher in comparison to chow-fed mice after 8 weeks of feeding. After 16 weeks of feeding, levels of 4-HNE staining in the liver were only found to be significantly different between mice fed a 30% fructose solution in combination with a high-fat

C. Sellmann et al. / Journal of Nutritional Biochemistry xx (2015) xxx–xxx

5

Table 2 Effect of fructose and/or fat feeding on markers of insulin resistance. 8 weeks C IR mRNA expression (−fold change) IRS-1 mRNA expression (−fold change) TNFα (pg/ml)

16 weeks F

Fat a

1.6±0.1

F+Fat a

2.3±0.2

1.3±0.1

3.2±0.5

1.4±0.1 a

1.8±0.3 a

12.1±2.0

12.7±0.7#

6.5±2.2#

C

F

Fat

2.2±0.1

1.4±0.1

2.7±0.2

2.1±0.1 b

1.5±0.2 a

3.4±0.5

1.7±0.1 a

2.3±0.1 a

1.7±0.2 a

13.5±1.6#

8.1±1.0#

1.5±0.1

11.9±1.7

7.3±1.9#

a,b

F+Fat

a

14.3±4.5

#

Values represent means±S.E.M. (n=4 –6). Significance refers to groups within one time point. C, control; F, fructose; IRS, insulin receptor substrate. a Pb.05 compared with respective water-fed control mice. b Pb.05 compared with mice fed with high-fat diet.

diet and chow controls as data varied considerable in some groups. Furthermore, after 8 and 16 weeks of feeding, protein levels of iNOS in liver tissue were significantly higher in all dietary intervention groups in comparison to chow controls, with levels being highest in livers of mice fed a combination of fat and fructose. Indeed, iNOS protein levels in livers of these animals were even higher than in those of mice fed a high-fat diet or the fructose solution after 8 weeks. After 16 weeks of feeding, similar differences were only found between mice fed a highfat diet and those fed a combination of fat and fructose. In line with these observations and also the findings for endotoxin plasma levels, TLR-4 and MyD88 mRNA expression, levels of arginase 1 (Arg-1) mRNA, being a marker for a more anti-inflammatory polarization of macrophages [18], was also significantly lower in livers of mice fed a diet rich in fat, fructose or a combination of fructose and fat in comparison to chow controls at both time point studies. Expression of monocyte chemotactic protein 1 (MCP-1), being a marker of a proinflammatory polarization of macrophages, was markedly higher; however, as expression of MCP-1 varied considerable in some groups, level of significance was not reached for the comparison of fat-fed mice and chow controls. 4. Discussion Worldwide, the number of individuals affected by NAFLD is still increasing. Results of several human studies suggest that lifestyle and herein particular physical activity and overnutrition are critical [5,19]; however, in more recent years, it has been suggested that not only a general overnutrition but also dietary composition may be critical [20]. Indeed, epidemiological studies suggest that a diet rich in fat and cholesterol may increase the odds to develop NAFLD [21]. Similar data have also been shown for a high dietary intake of sugar and herein particularly fructose [22]. However, only limited data are available comparing the different macronutrients and systematically determining their additive or even synergistic effects. In the present study, we used a mouse model to determine the effects of a diet enriched with fructose, fat or a combination of both on food intake, weight gain and the development of NAFLD. Applying these different diets, we were able to show that both diets rich in fat and fructose add to the development of liver steatosis over time; however, a combined feeding of these two macronutrients was found to exacerbate the damaging effect of each macronutrient by itself. Somewhat contrary to the studies of others feeding high-fat, high-fructose or so-called Western-style diets [19,23–26], in the present study, none of the feeding groups reached the full stage of NASH or even developed early fibrosis. Reasons for this might be related to the fact that mice were kept under SOPF conditions in the present study, whereas in most other studies, feeding experiments were conducted in standard pathogen-free facilities implying that intestinal microbiota was markedly different. Indeed, it was shown by us before that development of NAFLD, despite using the same feeding protocol, was markedly delayed when mice were kept under SOPF conditions [12]. Furthermore, while number of fat infiltrated hepatocytes was almost similar between livers of mice fed only fructose and those fed fructose and fat, fat droplets were markedly bigger

in mice fed a fructose- and fat-rich diet. Also, early signs of steatohepatitis were found in livers of these mice after 16 weeks of feeding whereas those fed fructose or fat only showed simple steatosis. The markedly more pronounced fat accumulation and damage in livers of mice fed a fructoserich diet or fructose- and fat-rich diet might at least in part have resulted from overnutrition found in these two groups. In contrast, in mice only fed fat, caloric intake was similar to controls. Indeed, while all groups had ad libitum excess to chow and drinking water, only fructose-fed mice and mice fed a combination of fructose and fat had an ~36% and ~47% higher total caloric intake, respectively. It has been suggest that fructose through its insulin-independent metabolism may be converted into fatty acids at a greater rate than other sugars (i.e. glucose) [27] thereby increasing de novo lipogenesis and adding to the development of NAFLD. In the present study, the markedly higher overall caloric intake and unregulated metabolism of fructose might have added to the more pronounced NAFLD in livers of fructose-fed mice and even more so in mice fed fructose and fat. Indeed, results of earlier studies of our group suggested that, besides others factors discussed below (i.e. increased endotoxin levels in portal plasma and activation of TLR-4-dependent signaling cascades), the insulin-independent metabolism seems to also contributed to the markedly higher fat accumulation found in livers of mice fed fructose when compared to other sugars [7,13]. Interestingly, while only taking in slightly more calories than mice fed a fructose-rich diet, mice fed the combination of fructose and fed gained significantly more weight and accumulated massive visceral adipose tissue. These data suggest that mice fed the diet enriched in fat and fructose changed their energy expenditure, decreased their physical activity or more efficiently metabolized consumed energy. Others feeding a so-called Western-style diet being rich in fat, sucrose and cholesterol or diets rich in high-fructose corn syrup in combination with a high-fat diet also report that mice fed these diets develop obesity after 6–8 weeks of feeding [20,24,26]; however, mechanisms involved in the massive weight gain remain to be determined. Taken together, our data suggest that, in mice, chronic intake of a diet rich in fat or fructose leads to the development of hepatic steatosis, which progresses with time and is most pronounced when both macronutrients are combined. However, our data also suggest that, contrary to feeding only one macronutrient, a combined feeding of a fatand fructose-enriched diet results in the development of overweight and a massive increase in visceral adipose tissue. These data by no means preclude that feeding a fat-rich diet or a fructose-rich diet with time also results in the development of overweight in mice but rather suggests that a combination of fat and fructose might enhance and accelerate the development of overweight and subsequently also the development of metabolic diseases like NAFLD. 4.1. Chronic elevated intake of fructose, fat or fructose and fat is associated with alterations of genes associated with insulin signaling in the liver To further characterize the different feeding models and to determine mechanisms underlying the differences in damage found, markers of insulin resistance were determined. Indeed, alterations of

6

C. Sellmann et al. / Journal of Nutritional Biochemistry xx (2015) xxx–xxx

Fig. 2. Effect of the different diets on the intestinal barrier function after 8 and 16 weeks. (A) Representative photomicrographs of ZO-1 and (B) of occludin staining of duodenum sections (original magnification ×400). (C) Densitometric analysis of ZO-1 and (D) occludin staining. Data are shown as means±S.E.M. (n=6). aPb.05 compared with respective water-fed control mice. C, control; F, fructose.

IR as well as IRS-1 and IRS-2 mRNA expression but also increases in TNFα protein in plasma and liver tissue have repeatedly been shown to be associated with impairments of insulin signaling and insulin resistance [28–32]. Recently, we reported that, despite not yet showing altered glucose tolerance tests and only slightly but not pathologically elevated insulin levels, expression of IRS-1 was significantly lower in livers of patients with steatosis and steatohepatitis than in controls [32]. Furthermore, it has been shown in rodents that decreased insulin sensitivity in liver is associated with a lower IR mRNA expression [33]. In the present study, no differences were found between groups in regard to plasma protein levels of TNFα as plasma samples were limited. However, the development of NAFLD was associated with a lower

expression of IRS-1 in the liver of all dietary treatment groups at both time points studied. In contrast, the decreased expression of IR mRNA found after 8 weeks of feeding in all dietary treatment groups persisted only in mice fed fructose over time. These data suggest that mice fed fat or fat in combination with fructose might have adapted to the different diets at least in part over time; however, molecular mechanisms involved remain to be determined. Taken together, these data suggest that the chronic intake of a diet rich in fat or fructose or even a combination of both is associated with alterations in insulin signaling in the liver; however, future studies will have to address if mice develop some kind of adaptation to these alteration and what the impact of these alterations on the development and even more so the progression of the disease is.

C. Sellmann et al. / Journal of Nutritional Biochemistry xx (2015) xxx–xxx

Fig. 3. Effect of the different diets on portal endotoxin levels and mRNA expression levels of TLR-4 and MyD88 in livers of mice after 8 and 16 weeks. (A) Portal endotoxin levels and expression of (B) TLR-4 and (C) MyD88 mRNA in livers of mice. Expression of mRNA was determined by real-time RT-PCR with Eef2 as a reference gene in the liver. Data are shown as means±S.E.M. (n=6). aPb.05 compared with respective water-fed control mice, bPb.05 compared with mice fed with 30% fructose solution, dPb.05 compared with mice fed with a combination of 30% fructose solution and a high-fat diet. C, control; F, fructose.

4.2. Chronic intake of fat, fructose or fat and fructose is associated with alterations of tight junction proteins in the upper part of the small intestine, increased translocation of bacterial endotoxin and induction of TLR-4-dependent proinflammatory signaling pathways Besides insulin resistance, changes in intestinal barrier function and an increased translocation of bacterial endotoxin have been suggested to be associated with the development of NAFLD not only in animal models of NAFLD but also in humans [32]. Indeed, chronic intake of a diet rich in fructose has repeatedly been shown to be

7

associated with a loss of tight junction proteins in the upper parts of the small intestine, elevated translocation of endotoxin and an induction of the expression of TLRs in the liver [6,16]. Results of rodent studies suggest that chronic intake of a high-fat diet is associated with similar alterations [34–36]. In the present study, we found that the intake of a diet rich in fat, fructose or a combination of both is associated with a marked loss of the tight junction proteins occludin and ZO-1 in duodenum but not in ileum of mice. Interestingly, in comparison to controls, the extent of the reduction of protein concentrations was similar between macronutrients and did not markedly progress with time. In line with these findings, bacterial endotoxin levels in portal plasma but also hepatic TLR-4 and to a lesser extent also MyD88 expression were markedly elevated in mice fed fat. In mice fed fructose or fructose and fat, portal endotoxin levels were only slightly different from controls when all four groups were compared (paired comparison with controls Pb.05 for both groups). However, TLR-4 levels shown before to be induced by endotoxin [32,37] were markedly higher in livers of mice fed fructose or fructose and fat. In line with these findings, MyD88 mRNA expression was also induced in livers of mice fed fructose and fructose and fat at the early time points studied, while after 16 weeks of feeding, this induction was only found in livers of mice fed fat and fructose. The lack of a significant elevation of portal endotoxin levels in mice fed fructose or fructose and fat might have resulted from the lack of availability of plasma samples in some groups and the variability of endotoxin levels in some groups, leading to a reduction of statistical power in the analysis. Indeed, TLR-4 expression in livers of these mice suggests that, despite being lower than in mice fed only fat, endotoxin levels were sufficient to induce TLR-4 and TLR-4-dependent signaling cascades. However, it could also be that fructose through so far unknown mechanism might have either induce hepatic TLR-4 expression or sensitize the liver to endotoxin. Indeed, it has been shown by us before that, in livers of mice concomitantly treated with antibiotics targeting Gram-negative bacteria while being fed a fructose solution, TLR-4 levels are still slightly higher than in controls (+~50%, n.s., P=.09) [6]. Most other TLRs were at the level of controls or below [6]. In addition, others have reported before that nutrient composition of the diet, probably through altering intestinal microbiota but also other factors (see below), might also impact plasma endotoxin levels [38,39]. In the present study, we tried to adjust fat and fructose intake between groups according to previous findings of our group [13] where we found that mice fed a fructose-enriched drinking water reduced their chow intake. Accordingly, it was necessary that the two fat groups were fed different high-fat diets. It cannot be ruled out that intestinal microbiota differed between mice only fed fat and those fed fat and fructose, which in turn might have affected portal endotoxin levels. Reasons for the “lack” of response of MyD88 mRNA expression after 16 weeks of feeding might have resulted from an adaptive response, a desensitization of the liver or a shift toward other TLR-4 signaling cascades (i.e. IRF3/7); however, mechanisms involved need to be clarified in future studies. The markedly higher levels of endotoxin in portal plasma of mice fed fat might have resulted from an enhanced uptake associated with an increased formation of chylomicrons found in rodents exposed to high-fat diets. Indeed, it has been shown that high amounts of fat in the diet promote the formation of chylomicrons and that this assists endotoxin absorption [40]. However, in line with our findings, results of Vreugdenhil et al. suggested that endotoxin bond to a complex with lipopolysaccharidebinding protein and chylomicrons leads to a decreased secretion of cytokines like TNFα [41]. The lipid A domain of lipopolysaccharides, which has been shown to be critical in the hepatocellular response to endotoxins, is blocked by the chylomicron binding subsequently leading to less hepatic inflammation [42,43]. In line with the findings for endotoxin and TLR-4, levels of iNOS protein and 4-HNE were also found to be induced in livers of all dietary treatment groups.

8

C. Sellmann et al. / Journal of Nutritional Biochemistry xx (2015) xxx–xxx

Fig. 4. Effect of the different diets on markers of oxidative stress and macrophage polarization in livers of mice after 8 and 16 weeks. (A) Representative photomicrographs of hepatic 4-HNE protein adduct staining and (B) iNOS staining of liver sections (original magnification ×200), (C) densitometric analysis of 4-HNE protein adduct and (D) iNOS staining. (E) Expression of Arg-1 and (F) MCP-1 mRNA levels in livers of mice. Expression of mRNA was determined by real-time RT-PCR with Eef2 as a reference gene in the liver. Data are shown as means±S.E.M. (n=6). aPb.05 compared with respective water-fed control mice, bPb.05 compared with mice fed with 30% fructose solution, cPb.05 compared with mice fed with high-fat diet. C, control; F, fructose.

C. Sellmann et al. / Journal of Nutritional Biochemistry xx (2015) xxx–xxx

It has been shown before that not only chronic dietary fructose intake but also intake of a high-fat diet is associated with an increase on hepatic iNOS expression and lipid peroxidation [44,45]. We further showed that, in mice chronically fed a fructose-enriched diet, iNOS was induced through endotoxin-dependent mechanisms and that a loss of iNOS markedly protected mice from lipid peroxidation and the development of fructose-induced NAFLD [44]. These findings support the hypothesis that an increased formation of reactive oxygen species — be it through endotoxin-dependent signaling cascades or other mechanisms — is critical in the development of NAFLD. It was further suggested that an endotoxin-dependent activation of Kupffer cells in the liver leads to an M1 polarization of these cells, which in turn further contributes to the onset and progression of inflammatory processes in the liver [18,46,47]. In the present study, elevated endotoxin and TLR-4 levels found in mice fed fructose, fat or a combination of both were associated with a reduced expression of Arg-1 and induction of MCP-1 in liver, suggesting that the development of NAFLD in these mice was associated with a change in polarization of Kupffer cells and infiltrating macrophages in liver. Taken together, these findings lend further support to the hypothesis that an impaired intestinal barrier function is associated with chronic intake of fat but also fructose, subsequently leading to an “activation” of Kupffer cells in the liver. However, our data also suggest that the effects of the different macronutrients on these parameters do not increase with time and are already present in the early phases of the disease development. Molecular mechanisms involved in regard to the differences in portal endotoxin levels, the response of the liver and this apparent “adaption” and their implications in disease progression remain to be determined. 5. Conclusion Taken together, our data suggest that chronic intake of fat, fructose and even more so a combined intake of fat and fructose leads to the development of NAFLD in mice, which progresses with time of intake. Our data further suggest that the development of NAFLD in feeding models using fat, fructose or a combination of both is associated with impairments of intestinal barrier function and an activation of TLR-4dependent signaling cascades in the liver. However, further studies are needed to determine mechanisms involved in the enhanced progression of the disease when fat and fructose are combined and especially the markedly more pronounced effect on body weight gain and to determine if similar effects of different macronutrients are also found in humans. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jnutbio.2015.05.011. Conflict of Interest The authors declare that they have no competing interests or other interests that might be perceived to influence the results and discussion reported in this paper. Acknowledgements This work is funded by grants from Bundesministerium für Bildung und Forschung, FKZ: 01EA1305 and FKZ: 01KU1214A (both IB). References [1] Blachier M, Leleu H, Peck-Radosavljevic M, Valla DC, Roudot-Thoraval F. The burden of liver disease in Europe: a review of available epidemiological data. J Hepatol 2013;58:593–608. [2] Vernon G, Baranova A, Younossi ZM. Systematic review: the epidemiology and natural history of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis in adults. Aliment Pharmacol Ther 2011;34:274–85.

9

[3] Bedogni G, Miglioli L, Masutti F, Tiribelli C, Marchesini G, Bellentani S. Prevalence of and risk factors for nonalcoholic fatty liver disease: the Dionysos nutrition and liver study. Hepatology 2005;42:44–52. [4] Bhala N, Jouness RI, Bugianesi E. Epidemiology and natural history of patients with NAFLD. Curr Pharm Des 2013;19:5169–76. [5] Fan JG, Cao HX. Role of diet and nutritional management in non-alcoholic fatty liver disease. J Gastroenterol Hepatol 2013;28:81–7. [6] Wagnerberger S, Spruss A, Kanuri G, Volynets V, Stahl C, Bischoff SC, et al. Toll-like receptors 1–9 are elevated in livers with fructose-induced hepatic steatosis. Br J Nutr 2012;107:1727–38. [7] Spruss A, Kanuri G, Wagnerberger S, Haub S, Bischoff SC, Bergheim I. Toll-like receptor 4 is involved in the development of fructose-induced hepatic steatosis in mice. Hepatology 2009;50:1094–104. [8] Gabele E, Dostert K, Hofmann C, Wiest R, Scholmerich J, Hellerbrand C, et al. DSS induced colitis increases portal LPS levels and enhances hepatic inflammation and fibrogenesis in experimental NASH. J Hepatol 2011;55:1391–9. [9] Ye D, Li FY, Lam KS, Li H, Jia W, Wang Y, et al. Toll-like receptor-4 mediates obesity-induced non-alcoholic steatohepatitis through activation of X-box binding protein-1 in mice. Gut 2012;61:1058–67. [10] Suzuki T, Hara H. Dietary fat and bile juice, but not obesity, are responsible for the increase in small intestinal permeability induced through the suppression of tight junction protein expression in LETO and OLETF rats. Nutr Metab (Lond) 2010;7:19. [11] Volynets V, Spruss A, Kanuri G, Wagnerberger S, Bischoff SC, Bergheim I. Protective effect of bile acids on the onset of fructose-induced hepatic steatosis in mice. J Lipid Res 2010;51:3414–24. [12] Spruss A, Henkel J, Kanuri G, Blank D, Puschel GP, Bischoff SC, et al. Female mice are more susceptible to nonalcoholic fatty liver disease: sex-specific regulation of the hepatic AMP-activated protein kinase-plasminogen activator inhibitor 1 cascade, but not the hepatic endotoxin response. Mol Med 2012;18:1346–55. [13] Bergheim I, Weber S, Vos M, Kramer S, Volynets V, Kaserouni S, et al. Antibiotics protect against fructose-induced hepatic lipid accumulation in mice: role of endotoxin. J Hepatol 2008;48:983–92. [14] Kleiner DE, Brunt EM, Van NM, Behling C, Contos MJ, Cummings OW, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease. Hepatology 2005;41:1313–21. [15] Brunt EM. Histopathology of non-alcoholic fatty liver disease. Clin Liver Dis 2009; 13:533–44. [16] Spruss A, Kanuri G, Stahl C, Bischoff SC, Bergheim I. Metformin protects against the development of fructose-induced steatosis in mice: role of the intestinal barrier function. Lab Invest 2012;92:1020–32. [17] Kanuri G, Weber S, Volynets V, Spruss A, Bischoff SC, Bergheim I. Cinnamon extract protects against acute alcohol-induced liver steatosis in mice. J Nutr 2009; 139:482–7. [18] Wan J, Benkdane M, Teixeira-Clerc F, Bonnafous S, Louvet A, Lafdil F, et al. M2 Kupffer cells promote M1 Kupffer cell apoptosis: a protective mechanism against alcoholic and nonalcoholic fatty liver disease. Hepatology 2014;59:130–42. [19] Schultz A, Neil D, Aguila MB, Mandarim-de-Lacerda CA. Hepatic adverse effects of fructose consumption independent of overweight/obesity. Int J Mol Sci 2013;14: 21873–86. [20] Desmarchelier C, Ludwig T, Scheundel R, Rink N, Bader BL, Klingenspor M, et al. Diet-induced obesity in ad libitum-fed mice: food texture overrides the effect of macronutrient composition. Br J Nutr 2013;109:1518–27. [21] Ferramosca A, Zara V. Modulation of hepatic steatosis by dietary fatty acids. World J Gastroenterol 2014;20:1746–55. [22] Ouyang X, Cirillo P, Sautin Y, McCall S, Bruchette JL, Diehl AM, et al. Fructose consumption as a risk factor for non-alcoholic fatty liver disease. J Hepatol 2008; 48:993–9. [23] Ganz M, Csak T, Szabo G. High fat diet feeding results in gender specific steatohepatitis and inflammasome activation. World J Gastroenterol 2014;20:8525–34. [24] Ishimoto T, Lanaspa MA, Rivard CJ, Roncal-Jimenez CA, Orlicky DJ, Cicerchi C, et al. High-fat and high-sucrose (western) diet induces steatohepatitis that is dependent on fructokinase. Hepatology 2013;58:1632–43. [25] Kohli R, Kirby M, Xanthakos SA, Softic S, Feldstein AE, Saxena V, et al. Highfructose, medium chain trans fat diet induces liver fibrosis and elevates plasma coenzyme Q9 in a novel murine model of obesity and nonalcoholic steatohepatitis. Hepatology 2010;52:934–44. [26] Reichold A, Brenner SA, Spruss A, Forster-Fromme K, Bergheim I, Bischoff SC. Bifidobacterium adolescentis protects from the development of nonalcoholic steatohepatitis in a mouse model. J Nutr Biochem 2014;25:118–25. [27] Aoyama Y, Yoshida A, Ashida K. Effect of dietary fats and fatty acids on the liver lipid accumulation induced by feeding a protein-repletion diet containing fructose to protein-depleted rats. J Nutr 1974;104:741–6. [28] Araki E, Lipes MA, Patti ME, Bruning JC, Haag B, Johnson RS, et al. Alternative pathway of insulin signalling in mice with targeted disruption of the IRS-1 gene. Nature 1994;372:186–90. [29] Gonzalez-Rodriguez A, Mas Gutierrez JA, Sanz-Gonzalez S, Ros M, Burks DJ, Valverde AM. Inhibition of PTP1B restores IRS1-mediated hepatic insulin signaling in IRS2-deficient mice. Diabetes 2010;59:588–99. [30] Kubota N, Tobe K, Terauchi Y, Eto K, Yamauchi T, Suzuki R, et al. Disruption of insulin receptor substrate 2 causes type 2 diabetes because of liver insulin resistance and lack of compensatory beta-cell hyperplasia. Diabetes 2000;49: 1880–9. [31] Withers DJ, Gutierrez JS, Towery H, Burks DJ, Ren JM, Previs S, et al. Disruption of IRS-2 causes type 2 diabetes in mice. Nature 1998;391:900–4.

10

C. Sellmann et al. / Journal of Nutritional Biochemistry xx (2015) xxx–xxx

[32] Kanuri G, Ladurner R, Skibovskaya J, Spruss A, Konigsrainer A, Bischoff SC, et al. Expression of toll-like receptors 1–5 but not TLR 6–10 is elevated in livers of patients with non-alcoholic fatty liver disease. Liver Int 2013;18. [33] Kong WJ, Zhang H, Song DQ, Xue R, Zhao W, Wei J, et al. Berberine reduces insulin resistance through protein kinase C-dependent up-regulation of insulin receptor expression. Metabolism 2009;58:109–19. [34] Serino M, Luche E, Gres S, Baylac A, Berge M, Cenac C, et al. Metabolic adaptation to a high-fat diet is associated with a change in the gut microbiota. Gut 2012;61:543–53. [35] Cani PD, Bibiloni R, Knauf C, Waget A, Neyrinck AM, Delzenne NM, et al. Changes in gut microbiota control metabolic endotoxemia-induced inflammation in highfat diet-induced obesity and diabetes in mice. Diabetes 2008;57:1470–81. [36] Lam YY, Ha CW, Campbell CR, Mitchell AJ, Dinudom A, Oscarsson J, et al. Increased gut permeability and microbiota change associate with mesenteric fat inflammation and metabolic dysfunction in diet-induced obese mice. PLoS One 2012;7:e34233. [37] Cani PD, Osto M, Geurts L, Everard A. Involvement of gut microbiota in the development of low-grade inflammation and type 2 diabetes associated with obesity. Gut Microbes 2012;3:279–88. [38] Cani PD, Amar J, Iglesias MA, Poggi M, Knauf C, Bastelica D, et al. Metabolic endotoxemia initiates obesity and insulin resistance. Diabetes 2007;56:1761–72. [39] Cani PD, Neyrinck AM, Fava F, Knauf C, Burcelin RG, Tuohy KM, et al. Selective increases of bifidobacteria in gut microflora improve high-fat-diet-induced diabetes in mice through a mechanism associated with endotoxaemia. Diabetologia 2007;50:2374–83.

[40] Ghoshal S, Witta J, Zhong J, de VW, Eckhardt E. Chylomicrons promote intestinal absorption of lipopolysaccharides. J Lipid Res 2009;50:90–7. [41] Vreugdenhil AC, Rousseau CH, Hartung T, Greve JW, van’t Veer C, Buurman WA. Lipopolysaccharide (LPS)-binding protein mediates LPS detoxification by chylomicrons. J Immunol 2003;170:1399–405. [42] Kasravi FB, Lee DH, Weisgraber K, Harris HW. Lipoprotein-bound endotoxin exerts an immunomodulatory effect on hepatocytes through the lipid A domain of LPS. J Endotoxin Res 2005;11:19–24. [43] Kasravi B, Lee DH, Lee JW, Dada S, Harris HW. Chylomicron-bound LPS selectively inhibits the hepatocellular response to proinflammatory cytokines. J Surg Res 2008;146:96–103. [44] Spruss A, Kanuri G, Uebel K, Bischoff SC, Bergheim I. Role of the inducible nitric oxide synthase in the onset of fructose-induced steatosis in mice. Antioxid Redox Signal 2011;14:2121–35. [45] Hassanin A, Malek HA, Saleh D. Heparin modulation on hepatic nitric oxide synthase in experimental steatohepatitis. Exp Ther Med 2014;8:1551–8. [46] Seth RK, Das S, Pourhosseini S, Dattaroy D, Igwe S, Basuray J, et al. M1 Polarization bias and subsequent NASH progression is attenuated by nitric oxide donor DETA NONOate via inhibition of CYP2E1 induced oxidative stress in obese mice. J Pharmacol Exp Ther 2014;27. http://dx.doi.org/10.1124/jpet.114.218131. [47] Wan J, Benkdane M, Alons E, Lotersztajn S, Pavoine C. M2 kupffer cells promote hepatocyte senescence: an IL-6-dependent protective mechanism against alcoholic liver disease. Am J Pathol 2014;184:1763–72.