Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Animal Models

Isocaloric high-fat feeding directs hepatic metabolism to handling of nutrient imbalance promoting liver fat deposition



Consumption of fat-rich foods is associated with obesity and related alterations. However, there is a group of individuals, the metabolically obese normal-weight (MONW) subjects, who present normal body weight but have metabolic features characteristic of the obese status, including fat deposition in critical tissues such as liver, recognized as a major cause for the promotion of metabolic diseases. Our aim was to better understand metabolic alterations present in liver of MONW rats applying whole genome transcriptome analysis.


Wistar rats were chronically fed a high-fat diet isocaloric relative to Control animals to avoid the hyperphagia and overweight and to mimic MONW features. Liver transcriptome analysis of both groups was performed.


Sustained intake of an isocaloric high-fat diet had a deep impact on the liver transcriptome, mainly affecting lipid metabolism. Although serum cholesterol levels were not affected, circulating triacylglycerols were lower, and metabolic adaptations at gene expression level indicated adaptation toward handling the increased fat content of the diet, an increased triacylglycerol and cholesterol deposition in liver of MONW rats was observed. Moreover, gene expression pointed to increased risk of liver injury. One of the top upregulated genes in this tissue was Krt23, a marker of hepatic disease in humans that was also increased at the protein level.


Long-term intake of a high-fat diet, even in the absence of overweight/obesity or increase in classical blood risk biomarkers, promotes a molecular environment leading to hepatic lipid accumulation and increasing the risk of suffering from hepatic diseases.


Epidemiological studies have identified a significant positive correlation between dietary fat intake, the development of obesity and the incidence of certain related disorders such as insulin resistance, cardiovascular diseases and/or cancer.1, 2, 3, 4 Although body mass index is widely used as a simple and first measure to define the development of certain risk factors associated with the intake of diets rich in fat, numerous studies have shown that a diet-induced disturbed metabolism does not necessarily correlate with obesity.5, 6 In fact, there is a special group of subjects who are not obese on the basis of weight and height but have metabolic features that are usually related to obesity: they are hyperinsulinemic, insulin resistant and vulnerable to type 2 diabetes mellitus and cardiovascular diseases.7, 8 These individuals have been denominated ‘metabolically obese normal-weight’ (MONW).9 The presence of individuals with MONW features constitutes a serious public health problem, as an increasing percentage of the population could be at higher metabolic risk, even presenting alterations in the classical blood biomarkers, but not being diagnosed because of the absence of overweight or obesity.10, 11, 12 The common characteristic of MONW subjects is the presence of higher visceral adiposity and fat deposition in tissues such as a liver, and muscle.13, 14 This has been related to insulin resistance, an alteration with a central role in metabolic syndrome development.15 The mechanisms involved in the harmful effects of ectopic fat accumulation are not completely understood, but may involve physical compression of the tissue/organ, changes in tissue secretory profiles as well as lipotoxicity.16 Fat deposition in the liver is recognized as a major cause for the promotion of metabolic diseases, including non-alcoholic fatty liver diseases, that are not only linked to other metabolic alterations such as insulin resistance or cardiovascular diseases, but can also promote more severe liver diseases including non-alcoholic steatohepatitis, hepatic cirrhosis and hepatocellular carcinoma (HCC).17, 18 Based on this, our objective was to study the molecular changes occurring in liver of MONW animals in depth by whole genome gene expression analysis.

The intake of unbalanced diets (rich in fats or carbohydrates) is one of the main causes involved in the appearance of MONW subjects.3, 4, 19 Although diets high in carbohydrates (especially those rich in sucrose and fructose) may also predispose to the development of MONW features,20 we focused our study on the effects of high-fat diets because of epidemiological evidence showing a dramatic increase in the intake of fat-rich foods in Western society.21 To perform our study, we designed an experiment administrating a high-fat diet isocaloric to a control diet in order to avoid the hyperphagia and increased body weight characteristic of these diets when administered in ad libitum conditions.22

Materials and methods

Ethics statement

Animal procedures were according the directive of the European Communities Council (2010/63/EU) and approved by the Ethical Committee of the University of the Balearic Islands.

Animals and diets

We used 2-month-old male Wistar rats (purchased from Charles River Laboratories España, SA, Barcelona, Spain) that were fed for 4 months with a control diet (Control group; n=7) or with a high-fat diet (HF group; n=7) administered in isocaloric conditions to the control one as previously described in detail, as these animals are part of a larger study.23 Sample size has been previously used in similar experiments, and has been shown appropriated to obtain statistical differences. Control diet (D12450B, Research Diets, Inc., New Brunswick, NJ, USA) was normolipidic (10% of Kcal from fat), whereas the HF diet (D12492, Research Diets, Inc.) contained 60% Kcal from fat (mainly lard). Both diets were obtained from Brogaarden (Gentofte, Denmark). Detailed diet composition (macronutrient proportion and ingredients) and fatty acid profile are included in Supplementary Table 1. Food intake was daily registered to calculate cumulative caloric intake. Moreover, body weight was recorded and fat percentage measured using an EchoMRI-700TM (from Echo Medical Systems, LLC, Houston, TX, USA). At the end of the experimental period, when animals were 6 months old, they were killed by decapitation at the beginning of the light cycle in the fed state and blood was collected from the neck to obtain serum. Liver was harvested, weighed and stored at −70 °C until analysis, and different adipose tissue depots (epididymal, inguinal, mesenteric and retroperitoneal) were weighed to determine the adiposity index. Although animals were killed in the fed state, we also collected blood in fasted conditions 1 week before killing in order to calculate the HOMA-IR (Homeostatic Model Assessment of Insulin Resistance) score using the formula of Matthews et al.24

Quantification of circulating total cholesterol levels and other serum parameters

Total circulating cholesterol was measured in serum using an enzymatic kit (BioSystems, Barcelona, Spain). Other circulating parameters were measured also using enzymatic kits: free fatty acids (Wako Chemicals, Neuss, Germany), triacylglycerols and β-hydroxybutyrate (both using kits from Sigma Diagnostics, Madrid, Spain). ELISA kits were used to measure serum insulin (DRG Instruments, Marburg, Germany) and tumor necrosis factor-α levels (R&D Systems Europe, Minneapolis, MN, USA). Circulating glucose was measured using an Accu-Chek Glucometer (Roche Diagnostics, Barcelona, Spain).

Triacylglycerol and total cholesterol liver determination

Lipid extracts were obtained from liver as previously described.25 Triacylglycerol (TG) content was measured in those extracts using the Serum Triglyceride Determination Kit (Sigma Diagnostics), and total cholesterol content using an enzymatic kit based on the reaction of cholesterol esterase and cholesterol oxidase/peroxidase (BioSystems).

Quantification of other liver parameters

Hepatic glycogen content was measured using an anthrone method,26 hepatic lipid content using the Folch method27 and protein content using the Bradford method.28

Liver histological analysis

Liver tissue samples were embedded in paraffin blocks for light microscopy and 5 μm thick sections were mounted on slides, all as previously described.29 For each liver sample, 2–3 photos from representative sections of the slides were digitalized. To determine the presence of hepatic steatosis we analyzed the entire area (2 cm2) of histological stained liver sections of 3–5 animals for each of the conditions, and searched for the presence of fatty vesicles. Image analyses from all samples and groups were examined in a blind manner.

Total RNA isolation

Total RNA from liver samples was extracted using Tripure Reagent (Roche Diagnostics). We further purified RNA using Qiagen RNeasy Mini Kit spin columns (Izasa SA, Barcelona, Spain). After quantification, we analyzed RNA integrity using an Agilent 2100 Bioanalyzer with RNA 6000 Nano chips (Agilent Technologies, South Queensferry, UK).

Microarray processing

We used a total of 14 individual liver RNA samples from the Control (n=7) and HF (n=7) groups for the microarray analysis; all arrays were hybridized simultaneously. Labeling and microarray processing (hybridization, washing and scanning) were performed as previously described.30 Briefly, 1 μg RNA of each sample was reverse transcribed, and half was labeled with Cy5 and the other half with Cy3.31 After purification and quantification, Cy3-labeled cRNA samples were pooled on an equimolar basis. Each individual Cy5-labeled sample was hybridized with Cy3 pool on 4x44K G4131F rat whole genome Agilent microarrays (from Agilent Technologies).

Microarray data analysis

Quantification and quality control of the array data was performed as previously described.32 Data were deposited in NCBI (National Center for Biotechnology Information) Gene Expression Omnibus (accessible through GEO Series accession number GSE57858). Statistical differences between the HF group vs the Control group was assessed by Student’s t-test in GeneMaths XT using a multiple testing correction according to Benjamini–Hochberg.33 Corrected P-values of <0.05 were considered as statistically significant. Fold change equals HF/Control ratio (for increase) or equals −1/ratio (for decrease). For pathways analysis using MetaCore (Thomson Reuters, New York, NY, USA), a threshold of P<0.05 (false discovery rate) was selected. The list of genes significantly affected was manually analyzed using different databases (PubMed, Rat Genome Database, Genecards, KEGG, NCBI, Reactome, UniProt, USCN and WikiPathways) and distributed in accordance to their functions and the biological processes in which they are involved. In addition, we supplemented the significantly enriched biological processes with non-annotated genes from the selected gene set using biological databases (Biocarta, SOURCE, GenMAPP and KEGG) and scientific literature. In some cases we annotated probes manually in order to have the best annotated list of probes.

Real-time reverse transcription-PCR analysis

In order to validate the data analysis, real-time reverse transcription-PCR was used to measure mRNA expression levels of selected genes in liver samples of animals of the Control and HF groups. The genes analyzed were: Abcg5, Cpt1a, Cyp2e1, Elovl6, Fasn, Me1, Scd1 and Scd2. Conditions used for gene expression analysis have been previously described.29 Primers were obtained from Sigma Genosys (Sigma Aldrich Química S.A., Madrid, Spain); sequences are provided in Supplementary Table 2. Data were normalized against the well-known reference gene β-actin.

Western blot analysis of KRT23 in liver

KRT23 protein levels were determined by western blot using 40 μg of total protein per lane in a 4–15% Criterion TGX Precast Gel (Bio-Rad, Madrid, Spain). Membranes were incubated overnight with a rabbit polyclonal anti-KRT23 antibody (Catalog number: STJ24357, St John’s Laboratory, London, UK) diluted 1:500 in Tris-buffered saline and Tween-20. β-Actin antibody (Catalog number: 37005, from Cell Signaling Technology, Inc., Danvers, MA, USA) was used as transfer and loading control. Infrared-dyed secondary anti-IgG antibodies (LI-COR Biosciences, Lincoln, NE, USA) were used, membranes were scanned in Odyssey Imager (LI-COR Biosciences) and bands were quantified using the software Odyssey V3.0 (LI-COR Biosciences).

Immunohistochemistry analysis of KRT23 in liver

Sections (5 μm) of liver samples from the Control and HF groups were immunostained by means of the avidin-biotin technique34 as previously described,29 using a primary rabbit polyclonal anti-KRT23 antibody (St John’s Laboratory) diluted 1:50 in phosphate-buffered saline.

Statistical analysis

Data analysis was performed without blinding. Student’s t-test was used to analyze differences between groups. All groups had a similar variance (homogeneity of variance). The test used for each comparison is detailed in the footnote of tables and figures. SPSS for windows (version 20, SPSS, Chicago, IL, USA) was used to perform statistical analysis. P<0.05 was defined as threshold of significance. For microarray data, the statistical analysis has been previously indicated in a specific section.


Body weight, adiposity and serum parameters

Data of body weight, adiposity and serum parameters of the experimental model used in this study have been previously published,23 and are summarized in Supplementary Table 3. In brief, compared with Control rats, 6-month-old animals fed an isocaloric HF diet for 4 months did not have increased body weight, but displayed increased adiposity as well as signs of insulin resistance, such as increased serum glucose levels and increased HOMA-IR index. Circulating free fatty acid (FFA) levels were not affected, whereas triacylglycerol levels were decreased in the HF animals. Serum tumor necrosis factor-α levels, measured as a marker of inflammation, were increased in the HF group. In addition, we detected β-hydroxybutyrate in serum of animals fed the HF diet.

Food and macronutrient intake

Animals of the Control and HF groups did not present any difference in accumulated caloric intake (10 227±182 Kcal in the Control and 9789±203 Kcal in HF group) or in daily food intake (73.2±3.4 Kcal in the Control and 69.8±3.1 Kcal in the HF group). However, as shown in Figure 1a, there were important differences in the proportion of macronutrients ingested: HF pair-fed animals ingested 5.7 times more fat than Controls during the experimental period (4 months) and, consequently, the intake of carbohydrates in this group was less, 3.7 times lower than in Control animals.

Figure 1

Cumulative macronutrient (a) and cholesterol (b) intake; and serum (c) and liver (d) cholesterol levels of male Wistar rats isocalorically fed with either a control or a HF diet from the age of 2 until the age of 6 months. Cumulative macronutrient intake was calculated by multiplying the amount (in g) of carbohydrates, fats and proteins ingested by the animals of the different groups during the 4 months of the experimental design by the Kcal g−1 of each macronutrient (4 Kcal g−1 for carbohydrates and proteins and 9 Kcal g−1 for fats). Cumulative cholesterol intake was calculated by multiplying for each group cumulative ingested diet (in g) by the content of cholesterol per g of diet (Supplementary Table 1). Total cholesterol levels in liver were extracted as described in the Materials and methods section and measured, as well as serum cholesterol, also as described in the Materials and methods. Results represent mean±s.e.m. (n=7). *Versus Control group (Student’s t-test, P<0.05).

Serum and hepatic cholesterol levels

Cholesterol content in the HF diet was 5.4 times higher than in the control diet (Supplementary Table 1). This was translated into an important difference in cumulative cholesterol intake between both groups, and this was 3.9 times higher in the HF group (Figure 1b). In spite of this, serum total cholesterol levels were similar in the HF and Control animals but increased liver cholesterol content was found in the HF group (Figures 1c and d).

Liver composition

Data revealed liver fat deposition in animals of the HF group, evidenced by higher total lipid and triacylglycerol content in liver (Figure 2a). In agreement, histological analysis revealed hepatic steatosis in the HF animals (Figure 2b). A lower weight of liver in the HF group was seen that could be explained, in part, by the lower glycogen content in these animals and the consequent loss in glycogen-bound water.

Figure 2

Liver composition and morphological analysis in the same animals indicated in Figure 1 (Control group and HF group) from the age of 2 until the age of 6 months. (a) Hepatic glycogen, lipid, TG and protein content measured as described in the Materials and methods section. Results represent mean±s.e.m. (n=7). *Versus Control group (Student’s t-test, P<0.05). (b) Representative sections of liver with hematoxylin/eosin staining are shown, evidencing hepatic steatosis in the HF animals.

Liver transcriptomics analysis

Our whole genome microarray transcriptome analysis showed that hepatic gene expression was affected by sustained intake of an isocaloric HF diet. Approximately 54% of all probes tested in the array were considered to be expressed (22 765 out of 41 012 probes). Of these, 3280 changed as a result of the intake of the HF diet, in comparison with the Control group (Student’s t-test, P<0.05). Application of false discovery rate adjustment gave a total of 233 probes that were differentially expressed in liver of isocaloric HF-fed animals in comparison with Controls. Duplicates with the highest P-values were removed, resulting in 187 unique genes being differentially expressed in the HF group (147 known and 40 unknown). Of the known genes, 34% were downregulated and 66% upregulated. Pathway analysis revealed that lipid metabolism was the most affected pathway, with 8 of the top 10 affected pathways being related to this process; the other affected pathways were carbohydrate and vitamin metabolism (Table 1). Of note, the single pathway related to carbohydrate metabolism, the pentose phosphate pathway, in fact directly related to lipid metabolism as it produces the necessary NADPH needed for fatty acid synthesis.

Table 1 Pathway analysis of liver genes affected by isocaloric HF diet feeding

Subsequently, we classified the 147 unique known genes. Genes affected by the HF diet were distributed into different processes (Supplementary Figure 1) that were in order of relevance as follows: energy homeostasis (mainly lipid metabolism), gene expression, signal transduction, cell cycle, sterol metabolism and cell communication. A subset of genes (n=47) were not included in the classification as they were involved in other biological processes. As could be expected, and in line with the results of the pathway analysis, the most significant and prominent process affected by the intake of the HF diet was lipid metabolism. Isocaloric HF diet feeding affected the expression of 34 genes related to energy homeostasis: 16 of these genes related with lipid synthesis and glucose metabolism were downregulated, whereas 18 genes related to fatty acid oxidation were upregulated. Another relevant affected process was sterol metabolism, with genes involved in the transport and biotransformation of dietetic cholesterol and in de novo cholesterol synthesis being differentially expressed. Metabolic interconnection of the most relevant HF-regulated genes involved in energy homeostasis and sterol metabolism are shown schematically in Figure 3.

Figure 3

Scheme of the effects on mRNA expression of energy metabolism-related genes in liver of rats with a chronic intake (4 months) of a HF diet isocaloric to a control diet based on microarray (and real-time reverse transcription-PCR (RT-PCR)) data analysis. The high percentage of fat and the lower carbohydrate content present in the diet of the HF group produced severe changes in liver metabolism aimed to compensate dietary macronutrient unbalance: decreased expression of genes involved in fatty acid/cholesterol synthesis as well as in glycolysis, in contrast to increased expression of key genes involved in β- and ω-oxidation, cholesterol biotransformation, ketone body synthesis and gluconeogenesis. However, the observed changes were not enough to avoid liver TG and cholesterol deposition. In addition, lower expression levels of Apoa4 that codes for a major component of triacylglycerol-rich very-low-density lipoproteins (VLDL) could be related to lower serum TG levels in the HF group, suggesting a decreased flow of TG into the blood stream, thus promoting the storage of TG in the liver. Increased liver cholesterol content may also be in relation with a lower cholesterol secretion as VLDL. Vertical arrows show up- or down-regulation of gene expression under conditions of HF diet.

Top 10 up- and downregulated genes

The top 10 genes upregulated and downregulated by HF diet in liver (Table 2) show changes in the same processes and direction than the pathway analysis. Of the 10 downregulated genes, 8 were related to lipid synthesis, 1 to the monosaccharide uptake and the last one to protein turnover. On the other hand, the top 10 upregulated genes were involved in more diverse processes: cell communication, cholesterol metabolism, signal transduction and ω-oxidation, among others.

Table 2 Top ten genes differentially expressed in liver of the high-fat diet (HF) group

Confirmation of microarray results by real-time reverse transcription-PCR

Real-time reverse transcription-PCR analysis was performed on RNA of liver samples of the two experimental groups to substantiate the microarray data. A total of eight genes were verified (Supplementary Table 4): six genes related to lipid metabolism (Cpt1a, Elovl6, Scd1, Scd2, Fasn and Me1), one gene involved in cholesterol metabolism (Abcg5) and one in detoxification/ω-oxidation (Cyp2e1). As in the microarray analysis, the eight genes were significantly affected by the isocaloric HF diet (Student’s t-test, P<0.05) and in the same direction.

Western blot and immunohistochemical analysis of KRT23 in liver

We then focused our attention on one of the top 10 upregulated genes, Keratin 23 (histone deacetylase inducible) (fold change +3.62, P=0.008, Student’s t-test, false discovery rate, P<0.05), the one encoding KRT23. We found this protein of interest as it has been recently reported, in humans, as a molecular marker of steatohepatitis and progression to HCC.35 Western blot analysis of liver tissue indicated that the animals fed with isocaloric HF diet had higher KRT23 levels compared with Control animals (Figure 4a). The supporting inmunohistochemical assay in the hepatic tissue revealed that KRT23 is localized mainly around the central vein in the group fed with HF diet, whereas it has a more disperse location in Controls (Figure 4b).

Figure 4

KRT23 protein expression analysed in liver of the same animals indicated in Figure 1. (a) KRT23 protein levels measured by western blot. Results represent mean±s.e.m. (n=7) of ratios of specific protein levels to β-actin (loading control). Data of Control group was set to 100%. *Versus Control group (Student’s t-test, P<0.05). Representative bands obtained in the western blot are shown; 40 μg of protein was loaded per lane. The represented bands correspond to samples that have been resolved together in the same gel. (b) KRT23 protein expression analysed by immunohistochemistry in liver of the Control and HF groups. Positive staining in HF animals was observed around the central vein, whereas positivity in Control animals was more dispersed.


The consumption of animal products and processed foods from animals is rising dramatically in Western societies,36 and this could be related to increased prevalence of MONW individuals. This is supported by this study and a previous study where rats were fed a isocaloric diet rich in lard,23 resulting in increased adiposity (mainly visceral), liver fat deposition and a metabolic profile related to obesity and metabolic syndrome without the development of overweight and obesity. Liver is a key tissue in the handling of the different macronutrients. It receives, transforms and redirects nutrients to extrahepatic tissues according to their requirements and functions.37 Increased liver fat deposition is a starting point of several metabolic diseases18, 38 and, for that reason, we performed a global gene expression analysis using microarray technology of the liver of MONW animals.

Our study revealed a strong alteration in rat liver transcriptome profile as a result of the chronic intake of an isocaloric HF diet during 4 months. In our animals, the substantial decrease in dietary carbohydrates in comparison with the control diet produced an increase in the expression of genes involved in gluconeogenesis (Fbp2 and G6pd), and decreased expression of genes involved in glycolysis (Gpi, Gapdh, Pkm2 and Pkdl). In addition, as expected, and in accordance with other microarray studies performed in liver of mice under long-term HF diet, but without isocaloric intervention, the largest number of genes expressed differentially were those involved in lipid metabolism.39 Particularly, the most drastic changes were observed in genes related to lipid synthesis and 7 of the top 10 downregulated genes were linked to this process (in order of fold change: Scd1 (with a fold change of −38.0), Elovl6, Fasn, Fabp5, Pnpla3, Pnpla5 and Me1). Another of the top 10 downregulated genes was Slc2a5 that codes for a glucose/fructose transmembrane transporter; this decreased expression may be associated with the lower carbohydrate content of the diet. When the relative consumption of carbohydrates is reduced, the liver switches from a glucose-based to a fat-based metabolism in which fatty acid β-oxidation serves as the primary source of energy.40 Indeed, and in accordance with other studies,39 our results show an upregulation of fatty acid oxidative capacity through the activation of genes related to fatty acid β-oxidation (Cpt1a, Echs1, Hadha and Hadhb). Interestingly, another process strongly affected by our HF diet was fatty acid ω-oxidation, with four genes being upregulated: Car/Nr1i3, Cyp3a3, Cyp2a6 and Cyp2e1. Car encodes the central nuclear receptor necessary for the expression of Cyp3a3, Cyp2a6 and Cyp2e1 that code for key microsomal cytochrome P450 enzymes involved in ω-oxidation. In normal conditions, dietary fatty acids are metabolized first in peroxisomes (very-long-chain FFA) and further in mitochondria (long, medium and short FFA) by β-oxidation.41 However, when the load of fatty acids is increased, the mitochondrial oxidative system becomes insufficient, and ω-oxidation is additionally activated to metabolize FFA.42, 43 Increased ω-oxidation has been related to liver steatosis and, although it is related to reactive oxygen species production, it has been proposed as an adaptation to help to prevent and decrease cellular lipotoxicity.44 Both ω- and β-oxidation are characterized by a high production of acetyl-CoA, the unique precursor of ketone bodies.45 Accordingly, we found increased expression of key genes involved in the production of ketone bodies (Bdh2 and Acat1), in concordance with the presence of β-hydroxybutyrate in the serum of the HF-fed animals. Although the amount of carbohydrates in our HF diet (20% of total Kcal) was not low enough to induce a strong ketogenic state,46 our animals presented ketogenic features, also reflected in decreased hepatic glycogen levels. Glycerol originated as result of the hydrolysis of the dietary TG could serve as substrate to produce glucose because of the previously commented increased gluconeogenic capacity in liver of our HF-fed animals. Intriguingly, we observed increased expression of Sord, a gene involved in the conversion of glucose to fructose through the polyol pathway, that is implicated in the development of diabetic complications.47 Notwithstanding, despite protective molecular adaptations that include inhibition of lipogenic genes, liver composition and morphologic analysis showed hepatic TG deposition in the HF animals. This increased lipid content is unlikely to be because of increased FFA uptake, as one of the most downregulated genes was Fabp5, involved in FFA uptake, transport and metabolism. Decreased Fabp5 expression in response to HF diets has been also reported by other authors,48 and can be interpreted as an adaptation to avoid liver lipotoxicity. Another important mechanism related to lipid deposition in liver is the control of secretion of TG and cholesterol into the circulation via the formation of very-low-density lipoproteins.18 VerHague et al.49 described that expression in mouse liver of apolipoprotein A-IV, which is a major component of triacylglycerol-rich very-low-density lipoproteins, enhanced TG secretion and reduced hepatic lipid content by promoting very-low-density lipoprotein expansion. Interestingly, our data reveal substantial lower mRNA levels of Apoa4 (fold change of −2.09, P=0.0002, Student’s t-test, false discovery rate, P<0.05) that could be one of the points to explain increased TG deposition observed in liver. In fact, this is in accordance with lower TG levels in serum of our MONW animals.

The HF diet used in our experiment was rich in cholesterol and, according to a higher cumulative cholesterol intake in the isocaloric HF group, we observed transcriptome changes in genes related to cholesterol homeostasis. It is well known that increased cholesterol intake is related to a compensatory downregulation of de novo synthesis;50 in agreement, we found decreased expression of Dnaja4, a gene involved in cholesterol biosynthesis.50 In addition, we found a higher mRNA expression of Akr1c21, Cesc1c, Cyp3a62, Cyp7b1 (one of the top 10 upregulated genes) and Sult2a6, genes related to cholesterol metabolism and excretion.51, 52, 53, 54 Inhibition of cholesterol synthesis and activation of pathways involved in cholesterol clearance could explain the fact that even though cholesterol intake in HF animals was around 4 times higher than in Controls, serum cholesterol levels were not affected. de Meijer et al.55 observed higher serum cholesterol levels in rodents fed with 60% HF under isocaloric conditions, but dietary lipid profile was rich in saturated and trans fat that are known to increase serum cholesterol levels.56, 57 The high amount of unsaturated fatty acids in our diet (>60% of total fat in comparison with a 31% in the study of de Meijer et al.55) could also explain maintenance of serum cholesterol, as monounsaturated and polyunsaturated fatty acids have been related to decreased serum cholesterol levels,58 in part resulting from an overall decrease of hepatic transcripts involved in de novo cholesterol biosynthesis.59 In addition, we found increased mRNA expression of Abcg5 (one of the top 10 upregulated genes) that plays an important role in the transport of dietary cholesterol in and out of the enterocyte and in the selective sterol excretion by the liver into bile.60 Nevertheless, and although no change was observed at serum level, the compensatory mechanisms were not enough to compensate increased cholesterol intake, and liver cholesterol content was increased. This fact could also be related to the previously commented decreased Apoa4 mRNA expression that would be related to a lower cholesterol secretion from liver into the bloodstream.

Fat and cholesterol deposition is one of the first stages in the development of steatosis, steatohepatitis, hepatic cirrhosis and HCC.18, 61, 62 Our gene expression analysis shows that three pathways highly affected by chronic isocaloric HF feeding were gene expression, signal transduction and cell cycle. In fact, one of the top 10 upregulated genes was Usp2; high liver protein levels of USP-2 have been related to the development of human HCC.63 We furthermore observed high gene expression levels of the sulfotransferases Sult1e1 and Sult2a6; these key enzymes in cholesterol metabolism have been shown to contribute to proliferation of hepatocellular carcinoma cells, and may promote the progression of HCC.64 Finally, Krt23, which codes for keratin 23, was the sixth upregulated gene. KRT23 overexpression in liver has been proposed as a biomarker for steatohepatitis as well as a marker for progression to HCC.35 Apart from increased mRNA expression, KRT23 protein levels were also increased in liver. Moreover, immunohistochemical staining revealed that in HF animals this protein was located around the central vein. All in all, long-term maintenance of the metabolic status generated by the intake of HF diet, even if this is administered in isocaloric conditions to a control diet and is not related to overweight/obesity (as in MONW individuals), could lead to a molecular microenvironment related to increased risk of hepatic diseases, including cancer. These data are in the same direction as what is described by Kim et al.,65 who suggested that the MONW phenotype is associated with increased risk of developing advanced colorectal adenoma.

In conclusion, in our animals the consumption of a HF diet controlled in calories during 4 months promoted metabolic features typical of MONW subjects. Liver microarray analysis revealed compensatory molecular adaptations to higher fat intake, but nevertheless hepatic TG and cholesterol levels were increased (see Figure 3 for a schematic interaction of the key affected genes involved in energy homeostasis and cholesterol metabolism). Changes at gene expression level indicated increased risk for hepatic pathologies related to increased liver fat deposition. Particularly, we observed increased KTR23 levels (mRNA and protein), a molecular marker of steatohepatitis/hepatocellular carcinoma, that constitutes the end stage of chronic liver diseases. MONW is not related to overweight/obesity, and we even did not observe changes in serum cholesterol levels, considered as one of the main markers of cardiovascular diseases; however, important adverse changes were occurring in the liver. These results suggest the need for early identification of MONW individuals in order to minimize health risks by implementation of prevention strategies.

Accession codes


Gene Expression Omnibus


  1. 1

    Moreno LA, Gottrand F, Huybrechts I, Ruiz JR, Gonzalez-Gross M, DeHenauw S et al. Nutrition and lifestyle in european adolescents: the HELENA (Healthy Lifestyle in Europe by Nutrition in Adolescence) study. Adv Nutr 2014; 5: 615S–623S.

    CAS  Article  Google Scholar 

  2. 2

    Diethelm K, Huybrechts I, Moreno L, De Henauw S, Manios Y, Beghin L et al. Nutrient intake of European adolescents: results of the HELENA (Healthy Lifestyle in Europe by Nutrition in Adolescence) Study. Public Health Nutr 2014; 17: 486–497.

    Article  Google Scholar 

  3. 3

    Bel-Serrat S, Mouratidou T, Huybrechts I, Labayen I, Cuenca-Garcia M, Palacios G et al. Associations between macronutrient intake and serum lipid profile depend on body fat in European adolescents: the Healthy Lifestyle in Europe by Nutrition in Adolescence (HELENA) study. Br J Nutr 2014; 112: 2049–2059.

    CAS  Article  Google Scholar 

  4. 4

    Labayen I, Ruiz JR, Ortega FB, Huybrechts I, Rodriguez G, Jimenez-Pavon D et al. High fat diets are associated with higher abdominal adiposity regardless of physical activity in adolescents; the HELENA study. Clin Nutr 2013; 33: 859–866.

    Article  Google Scholar 

  5. 5

    Caterson ID, Gill TP . Obesity: epidemiology and possible prevention. Best Pract Res Clin Endocrinol Metab 2002; 16: 595–610.

    Article  Google Scholar 

  6. 6

    Yajnik CS, Yudkin JS . The Y-Y paradox. Lancet 2004; 363: 163.

    Article  Google Scholar 

  7. 7

    Lee SH, Ha HS, Park YJ, Lee JH, Yim HW, Yoon KH et al. Identifying metabolically obese but normal-weight (MONW) individuals in a nondiabetic Korean population: the Chungju Metabolic disease Cohort (CMC) study. Clin Endocrinol (Oxf) 2011; 75: 475–481.

    Article  Google Scholar 

  8. 8

    Lopez-Miranda J, Perez-Martinez P . It is time to define metabolically obese but normal-weight (MONW) individuals. Clin Endocrinol 2013; 79: 314–315.

    Article  Google Scholar 

  9. 9

    Ruderman N, Chisholm D, Pi-Sunyer X, Schneider S . The metabolically obese, normal-weight individual revisited. Diabetes 1998; 47: 699–713.

    CAS  Article  Google Scholar 

  10. 10

    Molero-Conejo E, Morales LM, Fernandez V, Raleigh X, Gomez ME, Semprun-Fereira M et al. Lean adolescents with increased risk for metabolic syndrome. Arch Latinoam Nutr 2003; 53: 39–46.

    PubMed  Google Scholar 

  11. 11

    Dvorak RV, DeNino WF, Ades PA, Poehlman ET . Phenotypic characteristics associated with insulin resistance in metabolically obese but normal-weight young women. Diabetes 1999; 48: 2210–2214.

    CAS  Article  Google Scholar 

  12. 12

    Park YW, Zhu S, Palaniappan L, Heshka S, Carnethon MR, Heymsfield SB . The metabolic syndrome: prevalence and associated risk factor findings in the US population from the Third National Health and Nutrition Examination Survey, 1988-1994. Arch Intern Med 2003; 163: 427–436.

    Article  Google Scholar 

  13. 13

    Conus F, Rabasa-Lhoret R, Peronnet F . Characteristics of metabolically obese normal-weight (MONW) subjects. Appl Physiol Nutr Metab 2007; 32: 4–12.

    Article  Google Scholar 

  14. 14

    Thomas EL, Parkinson JR, Frost GS, Goldstone AP, Dore CJ, McCarthy JP et al. The missing risk: MRI and MRS phenotyping of abdominal adiposity and ectopic fat. Obesity 2012; 20: 76–87.

    CAS  Article  Google Scholar 

  15. 15

    Arsenault BJ, Beaumont EP, Despres JP, Larose E . Mapping body fat distribution: a key step towards the identification of the vulnerable patient? Ann Med 2012; 44: 758–772.

    CAS  Article  Google Scholar 

  16. 16

    Despres JP, Lemieux I . Abdominal obesity and metabolic syndrome. Nature 2006; 444: 881–887.

    CAS  Article  Google Scholar 

  17. 17

    Browning JD, Szczepaniak LS, Dobbins R, Nuremberg P, Horton JD, Cohen JC et al. Prevalence of hepatic steatosis in an urban population in the United States: impact of ethnicity. Hepatology 2004; 40: 1387–1395.

    Article  Google Scholar 

  18. 18

    Koo SH . Nonalcoholic fatty liver disease: molecular mechanisms for the hepatic steatosis. Clin Mol Hepatol 2013; 19: 210–215.

    Article  Google Scholar 

  19. 19

    Choi J, Se-Young O, Lee D, Tak S, Hong M, Park SM et al. Characteristics of diet patterns in metabolically obese, normal weight adults (Korean National Health and Nutrition Examination Survey III, 2005). Nutr Metab Cardiovasc Dis 2012; 22: 567–574.

    CAS  Article  Google Scholar 

  20. 20

    Cao L, Liu X, Cao H, Lv Q, Tong N . Modified high-sucrose diet-induced abdominally obese and normal-weight rats developed high plasma free fatty acid and insulin resistance. Oxid Med Cell Longev 2012; 2012: 374346.

    Article  Google Scholar 

  21. 21

    Rahati S, Shahraki M, Arjomand G, Shahraki T . Food pattern, lifestyle and diabetes mellitus. Int J High Risk Behav Addict 2014; 3: e8725.

    Article  Google Scholar 

  22. 22

    Palou M, Priego T, Sanchez J, Torrens JM, Palou A, Pico C . Moderate caloric restriction in lactating rats protects offspring against obesity and insulin resistance in later life. Endocrinology 2010; 151: 1030–1041.

    CAS  Article  Google Scholar 

  23. 23

    Diaz-Rua R, Garcia-Ruiz E, Caimari A, Palou A, Oliver P . Sustained exposure to diets with an unbalanced macronutrient proportion alters key genes involved in energy homeostasis and obesity-related metabolic parameters in rats. Food Funct 2014; 5: 3117–3131.

    CAS  Article  Google Scholar 

  24. 24

    Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC . Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 1985; 28: 412–419.

    CAS  Article  Google Scholar 

  25. 25

    Caimari A, Oliver P, Palou A . Adipose triglyceride lipase expression and fasting regulation are differently affected by cold exposure in adipose tissues of lean and obese Zucker rats. J Nutr Biochem 2012; 23: 1041–1050.

    CAS  Article  Google Scholar 

  26. 26

    Seifter S, Dayton S, Novic B, Muntwyler E . The estimation of glycogen with the anthrone reagent. Arch Biochem 1950; 25: 191–200.

    CAS  PubMed  Google Scholar 

  27. 27

    Folch J, Lees M, Sloane Stanley GH . A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem 1957; 226: 497–509.

    CAS  Google Scholar 

  28. 28

    Bradford MM . A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976; 72: 248–254.

    CAS  Article  Google Scholar 

  29. 29

    Garcia-Ruiz E, Reynes B, Diaz-Rua R, Ceresi E, Oliver P, Palou A . The intake of high-fat diets induces the acquisition of brown adipocyte gene expression features in white adipose tissue. Int J Obes (Lond) 2015; 39: 1619–1629.

    CAS  Article  Google Scholar 

  30. 30

    van Schothorst EM, Pagmantidis V, de Boer VC, Hesketh J, Keijer J . Assessment of reducing RNA input for Agilent oligo microarrays. Anal Biochem 2007; 363: 315–317.

    CAS  Article  Google Scholar 

  31. 31

    Duivenvoorde LP, van Schothorst EM, Bunschoten A, Keijer J . Dietary restriction of mice on a high-fat diet induces substrate efficiency and improves metabolic health. J Mol Endocrinol 2011; 47: 81–97.

    CAS  Article  Google Scholar 

  32. 32

    Diaz-Rua R, Keijer J, Caimari A, van Schothorst EM, Palou A, Oliver P . Peripheral blood mononuclear cells as a source to detect markers of homeostatic alterations caused by the intake of diets with an unbalanced macronutrient composition. J Nutr Biochem 2015; 26: 398–407.

    CAS  Article  Google Scholar 

  33. 33

    Hochberg Y, Benjamini Y . More powerful procedures for multiple significance testing. Stat Med 1990; 9: 811–818.

    CAS  Article  Google Scholar 

  34. 34

    Hsu SM, Raine L, Fanger H . Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 1981; 29: 577–580.

    CAS  Article  Google Scholar 

  35. 35

    Starmann J, Falth M, Spindelbock W, Lanz KL, Lackner C, Zatloukal K et al. Gene expression profiling unravels cancer-related hepatic molecular signatures in steatohepatitis but not in steatosis. PloS One 2012; 7: e46584.

    CAS  Article  Google Scholar 

  36. 36

    Cordain L, Eaton SB, Sebastian A, Mann N, Lindeberg S, Watkins BA et al. Origins and evolution of the Western diet: health implications for the 21st century. Am J Clin Nutr 2005; 81: 341–354.

    CAS  Article  Google Scholar 

  37. 37

    Otero YF, Stafford JM, McGuinness OP . Pathway-selective insulin resistance and metabolic disease: the importance of nutrient flux. J Biol Chem 2014; 289: 20462–20469.

    CAS  Article  Google Scholar 

  38. 38

    Yoon HJ, Cha BS . Pathogenesis and therapeutic approaches for non-alcoholic fatty liver disease. World J Hepatol 2014; 6: 800–811.

    Article  Google Scholar 

  39. 39

    Kim S, Sohn I, Ahn JI, Lee KH, Lee YS, Lee YS . Hepatic gene expression profiles in a long-term high-fat diet-induced obesity mouse model. Gene 2004; 340: 99–109.

    CAS  Article  Google Scholar 

  40. 40

    Kennedy AR, Pissios P, Otu H, Roberson R, Xue B, Asakura K et al. A high-fat, ketogenic diet induces a unique metabolic state in mice. Am J Physiol Endocrinol Metab 2007; 292: E1724–E1739.

    CAS  Article  Google Scholar 

  41. 41

    Rao MS, Reddy JK . Peroxisomal beta-oxidation and steatohepatitis. Semin Liver Dis 2001; 21: 43–55.

    CAS  Article  Google Scholar 

  42. 42

    Hsu MH, Savas U, Griffin KJ, Johnson EF . Human cytochrome p450 family 4 enzymes: function, genetic variation and regulation. Drug Metab Rev 2007; 39: 515–538.

    CAS  Article  Google Scholar 

  43. 43

    Reddy JK, Rao MS . Lipid metabolism and liver inflammation. II. Fatty liver disease and fatty acid oxidation. Am J Physiol Gastrointest Liver Physiol 2006; 290: G852–G858.

    CAS  Article  Google Scholar 

  44. 44

    Hardwick JP, Osei-Hyiaman D, Wiland H, Abdelmegeed MA, Song BJ . PPAR/RXR regulation of fatty acid metabolism and fatty acid omega-hydroxylase (CYP4) isozymes: implications for prevention of lipotoxicity in fatty liver disease. PPAR Res 2009; 2009: 952734.

    PubMed  Google Scholar 

  45. 45

    Cotter DG, Schugar RC, Crawford PA . Ketone body metabolism and cardiovascular disease. Am J Physiol Heart Circ Physiol 2013; 304: H1060–H1076.

    CAS  Article  Google Scholar 

  46. 46

    Selfridge JE, Wilkins HM, E L, Carl SM, Koppel S, Funk E et al. Effect of one month duration ketogenic and non-ketogenic high fat diets on mouse brain bioenergetic infrastructure. J Bioenerg Biomembr 2014; 47: 1–11.

    Article  Google Scholar 

  47. 47

    Abdel-Hamid NM, Nazmy MH, Abdel-Bakey AI . Polyol profile as an early diagnostic and prognostic marker in natural product chemoprevention of hepatocellular carcinoma in diabetic rats. Diabetes Res Clin Pract 2011; 92: 228–237.

    CAS  Article  Google Scholar 

  48. 48

    Renaud HJ, Cui JY, Lu H, Klaassen CD . Effect of diet on expression of genes involved in lipid metabolism, oxidative stress, and inflammation in mouse liver-insights into mechanisms of hepatic steatosis. PLoS One 2014; 9: e88584.

    Article  Google Scholar 

  49. 49

    VerHague MA, Cheng D, Weinberg RB, Shelness GS . Apolipoprotein A-IV expression in mouse liver enhances triglyceride secretion and reduces hepatic lipid content by promoting very low density lipoprotein particle expansion. Arterioscler Thromb Vasc Biol 2013; 33: 2501–2508.

    CAS  Article  Google Scholar 

  50. 50

    Robichon C, Varret M, Le Liepvre X, Lasnier F, Hajduch E, Ferre P et al. DnaJA4 is a SREBP-regulated chaperone involved in the cholesterol biosynthesis pathway. Biochim Biophys Acta 2006; 1761: 1107–1113.

    CAS  Article  Google Scholar 

  51. 51

    Pandak WM, Hylemon PB, Ren S, Marques D, Gil G, Redford K et al. Regulation of oxysterol 7alpha-hydroxylase (CYP7B1) in primary cultures of rat hepatocytes. Hepatology 2002; 35: 1400–1408.

    CAS  Article  Google Scholar 

  52. 52

    Quiroga AD, Li L, Trotzmuller M, Nelson R, Proctor SD, Kofeler H et al. Deficiency of carboxylesterase 1/esterase-x results in obesity, hepatic steatosis, and hyperlipidemia. Hepatology 2012; 56: 2188–2198.

    CAS  Article  Google Scholar 

  53. 53

    Matsubara T, Kim HJ, Miyata M, Shimada M, Nagata K, Yamazoe Y . Isolation and characterization of a new major intestinal CYP3A form, CYP3A62, in the rat. J Pharmacol Exp Ther 2004; 309: 1282–1290.

    CAS  Article  Google Scholar 

  54. 54

    Rizner TL, Penning TM . Role of aldo-keto reductase family 1 (AKR1) enzymes in human steroid metabolism. Steroids 2014; 79: 49–63.

    CAS  Article  Google Scholar 

  55. 55

    de Meijer VE, Le HD, Meisel JA, Akhavan Sharif MR, Pan A, Nose V et al. Dietary fat intake promotes the development of hepatic steatosis independently from excess caloric consumption in a murine model. Metabolism 2010; 59: 1092–1105.

    CAS  Article  Google Scholar 

  56. 56

    Rees K, Dyakova M, Wilson N, Ward K, Thorogood M, Brunner E . Dietary advice for reducing cardiovascular risk. Cochrane Database Syst Rev 2013; 12: CD002128.

    Google Scholar 

  57. 57

    Garshick M, Mochari-Greenberger H, Mosca L . Reduction in dietary trans fat intake is associated with decreased LDL particle number in a primary prevention population. Nutr Metab Cardiovasc Dis 2014; 24: 100–106.

    CAS  Article  Google Scholar 

  58. 58

    Adamsson V, Cederholm T, Vessby B, Riserus U . Influence of a healthy Nordic diet on serum fatty acid composition and associations with blood lipoproteins - results from the NORDIET study. Food Nutr Res 2014; 58: 24114.

    Article  Google Scholar 

  59. 59

    Rossmeisl M, Medrikova D, van Schothorst EM, Pavlisova J, Kuda O, Hensler M et al. Omega-3 phospholipids from fish suppress hepatic steatosis by integrated inhibition of biosynthetic pathways in dietary obese mice. Biochim Biophys Acta 2014; 1841: 267–278.

    CAS  Article  Google Scholar 

  60. 60

    Wang J, Mitsche MA, Luetjohann D, Cohen JC, Xie XS, Hobbs HH . Relative roles of ABCG5/ABCG8 in liver and intestine. J Lipid Res 2014; 56: 319–330.

    Article  Google Scholar 

  61. 61

    Van Rooyen DM, Larter CZ, Haigh WG, Yeh MM, Ioannou G, Kuver R et al. Hepatic free cholesterol accumulates in obese, diabetic mice and causes nonalcoholic steatohepatitis. Gastroenterology 2011; 141: 1403 e1–1403 e5.

    Article  Google Scholar 

  62. 62

    Jiang CM, Pu CW, Hou YH, Chen Z, Alanazy M, Hebbard L . Non alcoholic steatohepatitis a precursor for hepatocellular carcinoma development. World J Gastroenterol 2014; 20: 16464–16473.

    CAS  Article  Google Scholar 

  63. 63

    Calvisi DF, Wang C, Ho C, Ladu S, Lee SA, Mattu S et al. Increased lipogenesis, induced by AKT-mTORC1-RPS6 signaling, promotes development of human hepatocellular carcinoma. Gastroenterology 2011; 140: 1071–1083.

    CAS  Article  Google Scholar 

  64. 64

    Yang X, Xu Y, Guo F, Ning Y, Zhi X, Yin L et al. Hydroxysteroid sulfotransferase SULT2B1b promotes hepatocellular carcinoma cells proliferation in vitro and in vivo. PLoS One 2013; 8: e60853.

    CAS  Article  Google Scholar 

  65. 65

    Kim MC, Kim CS, Chung TH, Jeong J, Lee SH, Kim SR et al. MONW phenotype is associated with advanced colorectal adenoma in Korean men. Obesity (Silver Spring) 2012; 20: 1876–1881.

    Article  Google Scholar 

Download references


We thank Enzo Ceresi for technical assistance in the morphological and immunohistochemical analysis. CIBER de Fisiopatología de la Obesidad y Nutrición is an initiative of the ISCIII. This work was supported by the Spanish Government (Ministerio de Educación y Ciencia, EPIMILK (AGL2012-33692)) and by the EU FP7 project (BIOCLAIMS (FP7-244995)). Laboratory of Molecular Biology, Nutrition and Biotechnology and Human and Animal Physiology are members of the European Research Network of Excellence NuGO (The European Nutrigenomics Organization, EU Contract: FOOD-CT-2004-506360 NUGO). RD-R was recipient of a fellowship from the Spanish Government.

Author information



Corresponding author

Correspondence to A Palou.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on International Journal of Obesity website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Díaz-Rúa, R., van Schothorst, E., Keijer, J. et al. Isocaloric high-fat feeding directs hepatic metabolism to handling of nutrient imbalance promoting liver fat deposition. Int J Obes 40, 1250–1259 (2016).

Download citation

Further reading


Quick links