Distinct physiological, plasma amino acid, and ... - Wiley Online Library

6 downloads 0 Views 666KB Size Report
3 Key Laboratory of Human Functional Genomics Jiangsu Province, Nanjing Medical University, Nanjing,. P. R. China. 4 Norwich Medical School, University of ...
Mol. Nutr. Food Res. 2016, 60, 1199–1205

1199

DOI 10.1002/mnfr.201500789

FOOD & FUNCTION

Distinct physiological, plasma amino acid, and liver transcriptome responses to purified dietary beef, chicken, fish, and pork proteins in young rats Shangxin Song1 , Guido J. E. J. Hooiveld2 , Mengjie Li1 , Fan Zhao1 , Wei Zhang3 , Xinglian Xu1 , 4 ¨ , Chunbao Li1∗ and Guanghong Zhou1 Michael Muller 1

Key Laboratory of Meat Processing and Quality Control, MOE, Key Laboratory of Animal Products Processing, MOA, Jiangsu Synergetic Innovation Center of Meat Processing and Quality Control, Nanjing Agricultural University, Nanjing, P. R. China 2 Nutrition, Metabolism and Genomics Group, Division of Human Nutrition, Wageningen University, Wageningen, the Netherlands 3 Key Laboratory of Human Functional Genomics Jiangsu Province, Nanjing Medical University, Nanjing, P. R. China 4 Norwich Medical School, University of East Anglia Norwich, Norwich Medical School, University of East Anglia, Norwich, UK Scope: We report on the impact of purified dietary meat proteins from four species on plasma insulin, lipid and amino acid (AA) concentrations, and hepatic transcriptome (RNA-sequencing). Methods and results: Young rats received semi-synthetic diets for 1 wk that differed only regarding protein source; casein (reference) was replaced by beef, chicken, fish, or pork proteins. Compared to casein, all proteins, except pork, increased total plasma AA concentrations. Pork protein reduced adipose tissue mass and liver triacylglycerol, which was accompanied by increased plasma triacylglycerol concentrations. Plasma cholesterol was reduced by fish protein. The number of differentially expressed genes ranged between 609 (pork) and 1258 (chicken); on average one-third of the changes were specific for each meat protein. Pathway responses were most similar for beef and chicken, followed by pork and fish. Although the extent varied, all meat proteins induced mRNA translation, antigen processing/presentation, intracellular vesicular trafficking, and oxidoreductive-transformation pathways, and suppressed signal-transduction (Notch, TGFB/SMAD, insulin) and mitochondrial biogenesis pathways. Lipid- and AA-metabolic pathways were repressed, except by pork. AA-transport pathways were induced by beef and fish only, and complement/coagulation-pathways were suppressed by chicken and beef. Fish suppressed nuclear-transport and cofactor metabolism. Conclusion: To conclude, short-term feeding of different meat proteins resulted in distinct physiological and transcriptome changes in young rats.

Received: October 7, 2015 Revised: January 15, 2016 Accepted: January 18, 2016

Keywords: Dietary protein / Meat protein / Metabolism / Molecular nutrition / Nutrigenomics



Additional supporting information may be found in the online version of this article at the publisher’s web-site

1

Introduction

From a nutritional perspective, meat and meat products are rich sources of high biological value protein and important Correspondence: Dr. Guanghong Zhou E-mail: [email protected] Abbreviations: AA, amino acid; GSEA, gene set enrichment analysis; TC, total cholesterol; TG, triacylglyceride  C 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

∗ Additional corresponding author: Dr. Chunbao Li, E-mail: [email protected] Colour online: See the article online to view Fig. 2 in colour

www.mnf-journal.com

1200

S. Song et al.

micronutrients that are needed for good health throughout life [1,2]. However, increased consumption of meat, especially of red meat, has been associated with increased mortality, in particular due to cardiovascular diseases [3], but also to cancer [4]. Red meat intake has also been linked to an increased risk of type II diabetes [5]. As many of these detrimental effects have been connected to the content in meat of fats, especially of saturated fats, and heme, dietary guidelines recommend the consumption of lean, white meat over that of red meat [6]. Still, the role of meat as an important source of protein is unequivocal [1, 2]. The higher protein digestibility corrected amino acid (AA) scores (PDCAAS) of meat proteins compared to plant proteins (e.g. soy protein) indicate higher protein value in human nutrition for meat than plant protein [2, 7, 8]. However, except for the protein digestibility scores, detailed knowledge on physiological effects of dietary meat proteins is still limited. Recently, it has been found that fish protein may have beneficial effect on insulin sensitivity [9, 10], but comparative data on the effect of different meat proteins on physiological and gene expression responses are scarce. The aim of this study was to investigate and compare in young rats the effects of a short-term (1 wk) dietary intervention with purified meat proteins from beef, pork, chicken, or fish on growth performance, body fat, plasma insulin, and metabolite profiles, and the hepatic transcriptome. Beef and pork can be classified as red meats, whereas chicken and fish are considered white meat.

2

Materials and methods

All animals were handled in accordance with the guidelines of the Ethical Committee of Experimental Animal Center of Nanjing Agricultural University. The samples used in this study are part of a larger research project in which the effects of a whole set of dietary proteins from both plant and animal origin are investigated in young and old rats. Here we report on the effects of short-term (1 wk) exposure to meat proteins in young rats. After 1-wk adaption period, 4-wk-old male Sprague Dawley rats were fed for 7 days the nutritionally balanced semisynthetic AIN-93G diet [11], or the same diet in which the protein source (casein; milk protein) was fully replaced by purified meat proteins from beef, pork, chicken, or fish (n = 10 rats in each group). These meat proteins were chosen because previous work from our group demonstrated that the protein composition of these protein sources was very different [12], and different responses were thus anticipated. Detailed methods for meat protein and diet preparation and animal feeding can be found in Supporting Information– Detailed Methods. All diets were balanced for energy (4056.0 kcal/kg) and macronutrient content (protein 177g/kg, fat 70 g/kg, and carbohydrate 679.5 g/kg). See Supporting Information Table 1 for detailed diet information. The AA and mineral composition of the meat protein powders and diets are available in Supporting Information Tables  C 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Mol. Nutr. Food Res. 2016, 60, 1199–1205

2–4. Body weight and feed intake were determined daily. Liver and epididymal adipose tissue weight were measured at the dissection of the animals. Liver triacylglyceride (TG) and total cholesterol (TC) contents, and plasma insulin, glucose, TG, TC, and free AA concentrations were determined using commercially available kits. Liver gene expression profiles were determined by RNA sequencing of three biological replicates (i.e. three rats) from each protein group. Detailed methods for all analyses can be found in Supporting Information Detailed Methods. Differentially expressed genes were identified by using linear models and a moderated t-statistic [13,14]. To simplify the interpretation of the rather complicated RNA-seq dataset, each meat protein was only compared to casein. Genes that satisfied the criterion of p < 0.05 were considered to be significantly regulated. Changes in gene expression were related to biologically meaningful changes using gene set enrichment analysis (GSEA) [15]. The Enrichment Map plugin (Release 2.10) for Cytoscape 3.2.0 was used for visualization and interpretation of the GSEA results [16]. Enrichment maps were generated with gene sets that passed conservative significance thresholds (p < 0.05, False Discovery Rate (FDR) < 0.25). For visual comparison of differential regulation, a merged (union) Enrichment Map was generated that consisted of gene sets that were regulated by at least one of the meat proteins. This merged network served as canvas on which the significantly regulated gene sets per meat protein were overlaid. To further enhance biological interpretation, mutually overlapping nodes of high similarity were semi-automatically arranged close together, circled, and labeled to highlight the prevalent biological functions among related gene sets. For each meat protein, regulation of these ‘super-gene sets’ was subsequently visualized in a clustered heatmap according to the regulated fraction of gene sets. Except for the RNA-seq data, all statistical analyses were performed using the SPSS version 16.0 (Chicago, IL, USA). The effect of each meat protein on the measured variables was only compared to casein and was analyzed by ANOVA, followed by the two-tailed Dunnett’s posthoc test. Statistical significance was set at p < 0.05. Values are shown as mean ± SD.

3

Results

Rats in each protein group had similar initial body weight (Supporting Information Table 6). Although the rats fed chicken and fish protein had higher daily feed intake than casein (p < 0.05), daily body weight gain and final body weight did not differ (Supporting Information Table 6). To evaluate the effect of meat protein ingestion on body adiposity, epididymal adipose tissue weight and liver lipid content were measured (Fig. 1A–C). Compared to casein, only pork protein reduced epididymal adipose tissue weight (p < 0.05). Pork, but also beef protein significantly reduced liver weight and TG content (p < 0.05). Fish and chicken proteins reduced www.mnf-journal.com

Mol. Nutr. Food Res. 2016, 60, 1199–1205

1201

Figure 1. Effect of dietary animal proteins on epididymal adipose tissue weight (A1 & A2), liver weight (B1 & B2), liver TG and TC levels (C1 & C2), plasma TG and TC levels (D1 & D2), plasma glucose and insulin levels (E1 & E2), and HOMA-IR (E3). Values are mean ± SD (n = 10 per group); * : indicate significant difference (p < 0.05) compared to casein (reference) according to ANOVA with two-way Dunnett’s posthoc test. EATW: absolute weight of epididymal adipose tissue; EATW/BW: relative weight of epididymal adipose tissue to body weight; LW: absolute weight of liver; LW/BW: relative weight of liver to body weight; TG, triacylglyceride; TC: total cholesterol. HOMAIR was calculated using the formula: HOMA-IR = [glucose (mmol/L) × insulin (mIU/L)/22.5], using fasting values.

liver weight (p < 0.05), but did not change liver TG content. Liver TC content was not affected by any meat protein. Plasma TG concentrations were significantly increased by pork protein only (p < 0.05), whereas plasma TC concentrations were significantly reduced by fish protein only (p < 0.05, Fig. 1D). Only beef protein significantly reduced plasma glucose concentrations (p < 0.05, Fig. 1E1). Plasma insulin levels and HOMA-IR were reduced by all meat proteins (p > 0.05), but compared to casein this did not reach statistical significance (Fig. 1E). Compared to casein, total plasma AA concentrations were significantly increased by beef, chicken, and fish (p < 0.05), but not pork proteins (Table 1). For these three animal protein groups, this was due to increased concentrations of approximately half of the measured AA, including nutritionally essential and nonessential AA. In contrast, concentrations of the essential AA leucine, valine, and methionine were reduced by pork protein. RNA sequencing was performed to identify differentially expressed genes and corresponding biological pathways in rat liver. Compared to casein, the number of differentially expressed genes were for the beef (1076), chicken (1258), and fish (1172) protein groups about twice the size of the pork protein group (609) (see Supporting Information Fig. 1A), and on average one-third of the changes were specific for each meat protein. See Supporting Information Table 7 for all genes. The number of significantly enriched gene sets identified by GSEA reflected the number of differentially regulated genes  C 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

by each protein (Supporting Information Fig. 1B). Biological interpretation of enriched gene sets showed that clusters of functionally related gene sets describing mRNA translation, antigen processing/presentation, intracellular vesicular trafficking, and oxidoreductive transformation reactions were significantly increased by all four meat proteins, although the extent varied (Fig. 2, a high-resolution map that includes names of all gene sets is available in Supporting Information Fig. 2). In contrast, clusters relating to several signal transduction (Notch, TGFB/SMAD, insulin) and mitochondrial biogenesis pathways were suppressed by all meat proteins. The cluster of lipid metabolic pathways was repressed mostly by beef, chicken, and fish proteins but much less by pork protein. In more detail, gene sets describing fatty acid beta-oxidation, triglyceride and bile acids biosynthesis, and phospholipid metabolism were suppressed by beef, chicken, and fish proteins, whereas fish protein also inhibited sets of genes linked to the (regulation) of cholesterol biosynthesis. Pork protein had only a minor effect on gene sets related to bile acids biosynthesis. Regarding AA metabolism, beef, chicken, and fish proteins showed similar inducing effects on gene sets related to AA transport, but suppressed gene sets that described metabolism of tryptophan, histidine, and branched chain AA. Pork protein had no significant impact on AA metabolism. In addition, gene sets related to oxidative phosphorylation/electron transport chain (TCA/OXPHOS) were increased by chicken and pork protein, but repressed www.mnf-journal.com

1202

S. Song et al.

Mol. Nutr. Food Res. 2016, 60, 1199–1205

Table 1. Concentrations of free amino acids in plasma in rats fed casein and purified meat proteins

␮mol/L Total Leu Ile Val Phe Tyr Met Cys Arg Pro Lys Thr His Ser Asp Glu Gly Ala

Casein 2729.35 118.96 83.2 160.22 52.46 73.58 71.66 6.73 83.28 223.11 549.48 318.5 75.46 207.34 11.28 80.54 220.1 396.14

Pork ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

831.56 37.11 27.48 45.4 12.74 29.09 25.95 4.04 28.6 51.35 243.33 92.63 32.55 62.72 5.18 29.97 78.27 123.58

2707.12 87.98 67.56 121.28 45.36 57.76 51.32 8.72 105.42 289.1 475.24 488.52 60.98 229.04 8.2 63.4 250.12 297.12

Beef ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

190.6 10.25* 6.11 11.72* 8.63 6.41 6.36* 3.44 14.85 46.91* 60.49 102.19* 7.43 30.73 3.05 10.1 48.08 40

Chicken

3807.14 123.26 82.02 166.64 59.42 74.24 55.3 4.9 142.72 353.34 687.84 709.84 69.92 338.34 17.14 120.26 363.8 438.16

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

322.98* 22.99 15.81 24.91 7.4 16 12.29 6.18 20.53* 58.78* 92.06 106.02* 8.37 48.53* 4.37* 15.99* 55.98* 96.7

3831.5 130.24 90.5 173.08 84.94 90.46 69.18 11.9 149.29 416.62 711.6 623.16 78.52 313.24 17.94 109.56 369.24 420.3

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

Fish 335.61* 14.15 17.14 21.05 14.62* 13.99 15.87 6.39 27.71* 52.83* 85.24 105.23* 10.07 29.81* 4.03* 28.4* 46.48* 71.2

4047.46 122.68 81.04 170.14 61.26 77.3 60.34 5.42 150.1 393.74 776.76 692.98 80.08 352.98 17.7 109.2 413.36 482.38

± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±

674.99* 32.25 23.69 32.25 11.3 13.77 15.05 3.87 27.25* 58.22* 208.52* 194.02* 15.37 57.61* 5.37* 20.27* 60.14* 92.31

Values are mean ± SD; n = 10 per group; * : indicate significant difference (p < 0.05) compared to casein (reference) according to ANOVA with two-way Dunnett’s posthoc test.

by fish protein. In line with this, the gene set describing glutathione metabolism in the gene set cluster of oxidoreductive transformation was also increased by pork and chicken proteins. A large cluster of overlapping gene sets, which contained descriptors of many aspects of (control of) the cell cycle and cell proliferation was induced by all meat proteins, except fish protein. Especially fish protein suppressed the expression of genes involved in the generation of vitamin-derived coenzymes for metabolism. A cluster of gene sets involved in the insulin signaling pathway was inhibited most by chicken protein and to a lesser extent by the other meat proteins. The clustered heatmap (Fig. 2B) revealed that, compared to casein, the pathway responses were most similar for beef and chicken, followed by pork protein. The impact of fish protein was most dissimilar.

4

Discussion

We are the first to report on a comprehensive comparison of the effects of four meat proteins given at the recommended level on plasma insulin, lipid and AA concentrations, and the hepatic transcriptome in young male rats. The dietary meat proteins differentially affected the plasma AA pool, which was similar between chicken, fish, and beef on the one hand, and pork and casein on the other hand. The AA compositions of the diets (Supporting Information Table 3) was comparable between the chicken and fish protein diets, and between beef and pork protein diets. This coincided with higher feed intake of the chicken and fish protein, whereas the intake of the beef and pork protein was comparable to casein. Although the blood AA profiles did not fully reflect the AA intake, we believe that this may  C 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

partially explain the observed similar plasma AA patterns. Alternatively, differences in protein digestion and absorption in the gastrointestinal tract or AA metabolism in other organs may contribute to this as well. The four meat proteins showed distinct effects on liver lipid metabolism. Compared to casein, pork protein drastically prevented hepatic lipid accumulation, which coincided with increased plasma TG concentrations. Liver TG content and plasma TG concentrations showed similar trends in response to the other meat proteins, although, except for the reduction in liver TG for beef protein, this did not reach statistical significance. Even so, these observations are in accordance with data reported before [17], and indicate that TG may be shuttled from the liver into the blood in the form of TGrich VLDL particles [18]. Out of all four meat proteins, only fish protein reduced plasma TC levels compared to casein. This cholesterol-lowering property of fish protein has also been reported before [19, 20]. Our RNA-seq analyses suggest that reduced expression of genes of cholesterol biosynthesis by fish protein may contribute to this hypocholesterolemic effect. Notably, genes expression data showed that the lipid metabolism affected by the four meat proteins coincided with that of AA metabolism, indicating correlation between these two processes. We observed that especially chicken protein inhibited the expression of genes of the insulin signaling pathway. Even though the difference was not statistically significant (p > 0.05), plasma insulin concentrations and HOMA-IR were reduced by chicken protein by over 45% compared to casein. Nevertheless, this may point to chicken protein being an effective insulin-sensitizing protein source, although beef, chicken, and fish proteins did not change postprandial insulin levels in healthy human subjects [21]. However, other studies www.mnf-journal.com

Mol. Nutr. Food Res. 2016, 60, 1199–1205

1203

Figure 2. Enrichment maps and clustered heatmap of gene sets significantly regulated by meat proteins versus casein. (A) The networks display all significant gene sets changed by meat proteins compared to casein. Each nodes represent a gene set, and its size is proportional to the total number of genes within each gene set (varying from 15 to 500), while edges represent the fraction of overlapping genes between gene sets. Gene sets that did not pass the enrichment significance threshold (p < 0.05 and false discovery rate (FDR) < 0.25) for all of the four meat proteins are not shown. Red node color represents enrichment (induction) in the meat protein-fed rats, blue represents enrichment in the casein-fed rats (or suppression by the meat protein), whereas gray indicates no change. Mutually overlapping nodes of high similarity were semi-automatically arranged close together, circled, and labeled to highlight the prevalent biological functions among related gene sets. A high-resolution map that includes names of all gene sets is available in Supporting Information Fig. 2. (B) For each meat protein, regulation of the overlapping ‘super-gene sets’ was subsequently visualized in a clustered heatmap according to the regulated fraction of gene sets. The color scale represents the meat-specific fraction of regulated gene sets per super-gene set; the more intense the color the higher the fraction. Red means enriched, whereas blue means suppressed in the meat protein fed rats when compared to casein.  C 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.mnf-journal.com

1204

S. Song et al.

investigating diabetic human subject suggest a key role of amino acid metabolism in the early pathogenesis of diabetes [22, 23], but the molecular mechanism on the interaction between amino acid metabolism and insulin sensitivity remain to be elucidated. Of note is that a hypothesized mechanism links insulin signaling to the leucine-mediated activation of the mammalian target of rapamycin complex 1 [24]. Our RNA-seq data further showed that glutathione metabolism was increased by pork and chicken proteins. This induction was expected since the mitochondrial electron transport chain, that was also increased by pork and chicken proteins, is one of the major cellular generators of reactive oxygen species [25]. It is well known that glutathione plays a crucial role in antioxidant defense [26]. Glutathione is synthesized mainly in the liver from glutamate, cysteine, and glycine [26]. The availability of cysteine is thought to be the rate-limiting factor for synthesis of the glutathione [26]. In present study, pork and chicken protein groups had much higher level of plasma cysteine (8.74 and 11.9 ␮mol/L, respectively) than the beef and fish protein groups (4.90 and 5.42 ␮mol/L, respectively). Therefore, we believe the antioxidative capacity of meat proteins is likely to be related to the plasma cysteine level. In summary, we showed that purified beef, chicken, fish, and pork proteins provided at recommended levels caused distinct physiological, metabolic, and gene expression responses in young rats, and these changes were mainly related to differences in plasma AA profiles. Since young and adult rats have a very different physiological status, it remains to be demonstrated whether the observed responses will be alike in other age groups. We would like to thank Professors Ron Tume, Feng Gao, and Weihua Chen from Nanjing Agricultural University for their helpful suggestions in designing the experiments. This work was funded by grants 31471600 (National Natural Science Foundation of China), NCET-11-0668 (Ministry of Education of the P. R. China), and CXZZ13-0286 (Jiangsu Provincial Department of Education, China). G.Z., C.L., M.M., X.X., and S.S. designed research; S.S., M.L., F.Z., and W.Z. conducted research; S.S., C.L., and G.H. analyzed data; S.S., C.L., and G.H. wrote the paper; G.Z. and C.L. take primary responsibility for final content. All authors read and approved the final manuscript. The authors have declared no conflict of interest.

5

References

[1] McNeill, S., Van Elswyk, M. E., Red meat in global nutrition. Meat Sci. 2012, 92, 166–173. [2] Pereira, P. M., Vicente, A. F., Meat nutritional composition and nutritive role in the human diet. Meat Sci. 2013, 93, 586–592. [3] Schmid, A., The role of meat fat in the human diet. Crit. Rev. Food Sci. Nutr. 2011, 51, 50–66.

 C 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Mol. Nutr. Food Res. 2016, 60, 1199–1205

[4] Bouvard, V., Loomis, D., Guyton, K. Z., Grosse, Y. et al., Carcinogenicity of consumption of red and processed meat. Lancet Oncol 2015, 16, 1599–1600. [5] Pan, A., Sun, Q., Bernstein, A. M., Schulze, M. B. et al., Red meat consumption and risk of type 2 diabetes: 3 cohorts of US adults and an updated meta-analysis. Am. J. Clin. Nutr. 2011, 94, 1088–1096. [6] Willett, W. C., Stampfer, M. J., Current evidence on healthy eating. Annu. Rev. Public Health 2013, 34, 77–95. [7] Dietary protein quality evaluation in human nutrition. Report of an FAQ Expert Consultation. FAO Food Nutr. Pap. 2013, 92, 1–66. [8] Hoffman, J. R., Falvo, M. J., Protein - Which is Best? J. Sports Sci. Med. 2004, 3, 118–130. [9] Madani, Z., Louchami, K., Sener, A., Malaisse, W. J. et al., D., Dietary sardine protein lowers insulin resistance, leptin and TNF-alpha and beneficially affects adipose tissue oxidative stress in rats with fructose-induced metabolic syndrome. Int. J. Mol. Med. 2012, 29, 311–318. [10] Ouellet, V., Marois, J., Weisnagel, S. J., Jacques, H., Dietary cod protein improves insulin sensitivity in insulin-resistant men and women: a randomized controlled trial. Diabetes Care 2007, 30, 2816–2821. [11] Reeves, P. G., Nielsen, F. H., Fahey, G. C., Jr., AIN-93 purified diets for laboratory rodents: final report of the American Institute of Nutrition ad hoc writing committee on the reformulation of the AIN-76A rodent diet. J. Nutr. 1993, 123, 1939–1951. [12] Wen, S., Zhou, G., Song, S., Xu, X. et al., Discrimination of in vitro and in vivo digestion products of meat proteins from pork, beef, chicken, and fish. Proteomics 2015, 15, 3688–3698. [13] Ritchie, M. E., Phipson, B., Wu, D., Hu, Y. et al., limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015, 43, e47. [14] Law, C. W., Chen, Y., Shi, W., Smyth, G. K., voom: precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 2014, 15, R29. [15] Subramanian, A., Tamayo, P., Mootha, V. K., Mukherjee, S. et al., Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. USA 2005, 102, 15545–15550. [16] Merico, D., Isserlin, R., Stueker, O., Emili, A. et al., Enrichment map: a network-based method for gene-set enrichment visualization and interpretation. PLoS One 2010, 5, e13984. [17] Brandsch, C., Shukla, A., Hirche, F., Stangl, G. I. et al., Effect of proteins from beef, pork, and turkey meat on plasma and liver lipids of rats compared with casein and soy protein. Nutrition 2006, 22, 1162–1170. [18] Schwarz, J., Tome, D., Baars, A., Hooiveld, G. J. et al., Dietary protein affects gene expression and prevents lipid accumulation in the liver in mice. PLoS One 2012, 7, e47303. [19] Hosomi, R., Fukunaga, K., Arai, H., Kanda, S. et al., Fish protein decreases serum cholesterol in rats by inhibition of cholesterol and bile acid absorption. J. Food Sci. 2011, 76, H116-H121.

www.mnf-journal.com

Mol. Nutr. Food Res. 2016, 60, 1199–1205

[20] Shukla, A., Bettzieche, A., Hirche, F., Brandsch, C. et al., Dietary fish protein alters blood lipid concentrations and hepatic genes involved in cholesterol homeostasis in the rat model. Br. J. Nutr. 2006, 96, 674–682. [21] Uhe, A. M., Collier, G. R., O’Dea, K., A comparison of the effects of beef, chicken and fish protein on satiety and amino acid profiles in lean male subjects. J. Nutr. 1992, 122, 467– 472. [22] Michaliszyn, S. F., Sjaarda, L. A., Mihalik, S. J., Lee, S. et al., Metabolomic profiling of amino acids and beta-cell function relative to insulin sensitivity in youth. J. Clin. Endocrinol. Metab. 2012, 97, E2119–E2124.

 C 2016 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

1205 [23] Wang, T. J., Larson, M. G., Vasan, R. S., Cheng, S. et al., Metabolite profiles and the risk of developing diabetes. Nat. Med. 2011, 17, 448–453. [24] Lynch, C. J., Adams, S. H., Branched-chain amino acids in metabolic signalling and insulin resistance. Nat. Rev. Endocrinol. 2014, 10, 723–736. [25] Liu, Y., Fiskum, G., Schubert, D., Generation of reactive oxygen species by the mitochondrial electron transport chain. J. Neurochem. 2002, 80, 780–787. [26] Wu, G., Fang, Y. Z., Yang, S., Lupton, J. R. et al., Glutathione metabolism and its implications for health. J. Nutr. 2004, 134, 489–492.

www.mnf-journal.com