Novel insights and mechanisms of diet-induced obesity: Mid-term versus long-term effects on hepatic transcriptome and antioxidant capacity in Sprague-Dawley rats

Life Sciences 324 (2023) 121746 Available online 29 April 2023 0024-3205/© 2023 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Novel insights and mechanisms of diet-induced obesity: Mid-term versus long-term effects on hepatic transcriptome and antioxidant capacity in Sprague-Dawley rats Alejandro García-Beltr´ an a , 1 , Rosario Martínez a , 1 , Jesus M. Porres a , * , Francisco Arrebola b , Inmaculada Ruiz Artero b , Milagros Galisteo c , Pilar Aranda a , Garyfallia Kapravelou a , 1 , María L´ opez-Jurado a , 1 a Department of Physiology, Institute of Nutrition and Food Technology (INyTA), Centre for Biomedical Research (CIBM), Institute for Research in Sport and Health (IMUDS), Universidad de Granada, 18016 Granada, Spain b Department of Histology, Institute of Neurosciences, Centre for Biomedical Research (CIBM), Universidad de Granada, 18016 Granada, Spain c Department of Pharmacology, School of Pharmacy, Centre for Biomedical Research (CIBM), Universidad de Granada, Campus Universitario de Cartuja s/n, 18071 Granada, Spain ARTICLE INFO Keywords: Obesity Aging Liver Glucose & lipid metabolism Antioxidant capacity Rats ABSTRACT Aims: The study of molecular mechanisms related to obesity and associated pathologies like type 2-diabetes and non-alcoholic fatty liver disease requires animal experimental models in which the type of obesogenic diet and length of the experimental period to induce obesity deeply affect the metabolic alterations. Therefore, this study aimed to test the influence of aging along a rat model of diet-induced obesity in gene expression of the hepatic transcriptome. Main methods: A high-fat/high-fructose diet to induce obesity was used. Mid- (13 weeks) and long-term (21 weeks) periods were established. Caloric intake, bodyweight, hepatic fat, fatty acid profile, histological changes, antioxidant activity, and complete transcriptome were analyzed. Key findings: Excess bodyweight, hepatic steatosis and altered lipid histology, modifications in liver antioxidant activity, and dysregulated expression of transcripts related to cell structure, glucose & lipid metabolism, antioxidant & detoxifying capacity were found. Modifications in obese and control rats were accounted for by the different lengths of the experimental period studied. Significance: Main mechanisms of hepatic fat accumulation were de novo lipogenesis or altered fatty acid catabolism for midor long-term study, respectively. Therefore, the choice of obesity-induction length is a key factor in the model of obesity used as a control for each specific experimental design. 1. Introduction Nowadays, obesity and associated pathologies including nonalcoholic fatty liver disease (NAFLD), metabolic syndrome (MetS), and alterations in bone functionality, constitute an important burden for national health systems. The prevalence of these diseases is very high and the fact that life expectancy, as a result of the increase in obesity, could decrease for the first time in recent times has generated great alarm [1]. Due to their high prevalence in the world population, obesity and MetS are considered pandemics. This condition affects approximately 20–40 % of the population in industrialized nations, and its prevalence is expected to rise further in the next decades [2]. Their incidence increases alarmingly every year mainly due to environmental factors, although genetic factors are also involved [3]. Concerning environmental factors, energy imbalance is influenced by a diet high in saturated fat and refined sugars as well as a sedentary lifestyle [4]. The regular intake of high fat and high fructose diets is directly related to the development of obesity, which is an important risk factor for other * Corresponding author at: Institute of Nutrition and Food Technology, Centre for Biomedical Research, Universidad de Granada, Avda. del Conocimiento S/N., Armilla 18016, Granada, Spain. E-mail address: [email protected] (J.M. Porres). 1 These authors contributed equally to this work. Contents lists available at ScienceDirect Life Sciences journal homepage: www.elsevier.com/locate/lifescie https://doi.org/10.1016/j.lfs.2023.121746 Received 21 February 2023; Received in revised form 8 April 2023; Accepted 25 April 2023 Life Sciences 324 (2023) 121746 2 associated chronic pathologies [3]. Experimental animal models are indispensable tools for the study of the alterations in morphology and metabolic pathways involved in the development of overweight and obesity. Although numerous models have been developed, they are usually classified into two main categories such as genetic or diet-induced obesity (DIO). Genetic models exhibit strong symptoms and can be associated with MetS or its hepatic manifestation, NAFLD. However, genetic obesity has a lower distribution compared to other forms of obesity that are more related to environmental factors. In this regard, DIO models may be closer to the actual situation that is more prevalent in the human population. DIO models are usually generated using a high-fat semi-synthetic diet (HFD) with either 60 or 45 % of the total caloric content present as fat [5,6], although cafeteria diets prepared with a mixture of ingredients representative of a western type diets (fried potatoes, biscuits, bacon, standard chow diet, pork pate base and liquid chocolate) are also used [7]. In recent years, and to resemble more accurately the consume trends by humans, the use of diets with 45 % of kcal present as fat in combination with a high mono or di-saccharide content (fructose or sucrose) has been recommended for DIO models [5,6]. Furthermore, other diet-related considerations present in the former diets should be considered like the presence of casein as the main protein source, which per se has been reported to induce significant alterations in rat's glucose and lipid metabolism [8]. Nevertheless, a general trend of outcomes may be described for most of the diet-induced obesity interventions that incorporate overweight, metabolic dysregulation or alterations in liver histology. Likewise, another factor that should be considered when designing DIO models is the extensive length of time that is usually needed to implement such models and develop the adequate bodyweight and metabolic alterations sought. Such periods may lead to an aging process of the animals that will in turn significantly affect the extent to which different genes are expressed [9], and thus metabolic pathways are affected. The longer the experimental period implemented, the more extensive the damage in different organ functionality caused by the dietary conditions implemented to produce the development of obesity. All these changes are matched by a progressive functional decline in various organs over time, changes in the biotransformation of xenobiotics, and impairment of normal cellular functions by free radicals. Aged animals have altered activity of drug-metabolizing enzymes such as glutathione S-transferases [10], and aging is associated with downregulated expression of genes related to antioxidant function, thus impairing the antioxidant capacity of the liver and increasing oxidative damage in old animals [11]. Moreover, the adaptation of key transcription factors involved in lipid metabolism in response to nutritional status changes is impaired in old rats, and this might contribute to the development of hepatic steatosis with aging [12]. In the liver of Wistar rats, aging caused an increase in the mRNA abundance of lipogenic transcription factors and enzymes, and a decrease in mRNA levels of enzymes associated with mitochondrial fatty acid oxidation such as carnitine-palmitoyl transferase-1 (CPT-1a) [11]. In this experiment, we have developed an animal model of dietinduced obesity (DIO) using a HFD (45 % of total dietary Kcal) and refined sugars (20 % fructose solution) to resemble the events that take place in the human population. 10-week-old male Sprague-Dawley rats were used and fed ad libitum during 13 or 21 weeks with an unbalanced high-fat high-fructose diet that produces a pathological state similar to human obesity. Specifically, we aimed to (i) develop a rat experimental model of obesity and non-alcoholic fatty liver disease along two different maturity stages of the experimental animals, (ii) perform an in vivo study of hepatic transcriptome profile and detection of the main markers associated with NAFLD, and (iii) test the influence of aging along DIO in potential metabolic and histological damage, as well as in gene expression of hepatic transcripts related to cell structure components, glucose and lipid metabolism, or antioxidant and detoxifying defense system. 2. Material and methods 2.1. Chemical compounds Cumene hydroperoxide, KCN, 2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide (XTT), dichlorophenolindophe nol (DCIP), flavin adenine dinucleotide (FAD), 1-chloro-2,4-dinitrobenzene (CDNB), NADPH, xanthine, xanthine oxidase, reduced glutathione, and sodium azide were from Sigma-Aldrich (Madrid, Spain). Diethylene triamine pentaacetic acid (DETAPC) was from Panreac Applichem (Barcelona, Spain). 2.2. Animals and experimental design A total of 32 male Sprague-Dawley rats (Charles River, Barcelona) were housed in group cages with solid-bottom and nest shavings (4 rats in each cage), and located in a well-ventilated, thermostatically controlled room (21 ±2 ◦C) under a 12-h light/dark cycle to ensure animal welfare (Unidad de Experimentaci´ on Animal, CIC, University of Granada). After a week of acclimatization, animals aged 10 weeks with a starting average body weight of 357 ±2 g, were randomly divided into four experimental groups of eight animals per group. All the cages in each specific experimental group were labeled accordingly and placed in order to avoid possible identification errors. All the researchers and animal facility personnel were trained to ensure the correct identification of cages in each experimental group. The handling of the animals was refined to the minimum necessary to ensure their comfort during the experiments and to avoid causing unnecessary stress. Two groups of animals consumed a standard diet (20 % kcal protein, 10 % kcal fat) (TD.08806, ENVIGO, Madison, WI) for 13 or 21 weeks (SD1 and SD2, respectively) while the remaining two experimental groups consumed a HFD (19 % kcal protein, 45 % kcal fat) (TD.06415, ENVIGO, Madison, WI) and were fed 20 % fructose in the drinking water for 13 or 21 weeks (HFHF1 and HFHF2, respectively) since the combination of high fat and high fructose generates a higher weight gain, hyperinsulinemia, hepatic steatosis and oxidative stress than a high fat, low-carbohydrate diet [5,6] and promotes de novo lipogenesis and the aggravation of glucose and fat metabolism disorders [6]. During the experimental period, the animals had free access to fluid (water or 20 % fructose solution), and animals of the HFD group consumed the experimental diet ad libitum, while the animals of the standard-diet group followed a pair-fed design of food consumption based on previous experiments [9] to maintain adequate caloric intake in a normocaloric group. Food intake was recorded daily whereas bodyweight was measured weekly. All experiments were undertaken according to Directional Guides Related to Animal Housing and Care [13], and all procedures were approved by the Animal Experimentation Ethics Committee of the University of Granada, Spain (Project Reference 16/ 07/2019/132). To select the number of rats assigned to each experimental group (n =8), we implemented the 3Rs principle [14]. General health monitoring of all animals was performed every day. Criteria for the health monitoring include wound, bleeding, hair brilliance, nasal discharge, eye discharge, convulsions, alterations in heart rate, anal and genital discharge, and general motor activity. An end point criterion was established if the animals suffered any type of physical damage, showed symptoms of anorexia with a decrease in intake of 30 % or more, or weight loss of 25 % or more. If, based on the manifestations presented by the animal, it is decided that it is in an irreversible state of suffering, the experimental death of the specific animal is declared, and its sacrifice is considered, in case it should be necessary. No adverse effect derived from the obesity induction was observed in the experimental animals at either length of the experimental period. In addition, no animal became severely ill or died before the experimental endpoints. At the end of the 13or 21-week experimental period, the animals were anesthetized with ketamine (75 mg⋅kg −1 body weight) and xylazine (10 mg⋅kg −1 body weight), and blood was collected by A. García-Beltr´ an et al. Life Sciences 324 (2023) 121746 3 abdominal aorta puncture using heparin as an anticoagulant. The liver was extracted, weighed, divided into various portions and immediately frozen in liquid nitrogen and stored at −80 ◦C except for 100 mg that were immersed in RNA preserving solution (RNAlater, Ambion). 2.3. Total hepatic lipid content A liver portion was lyophilized to assess the moisture content. Hepatic lipids were extracted using hexane from an aliquot of the freezedried liver portion using the method described by Folch et al. [15] with the modifications made by Kapravelou et al. [16]. Total liver lipids were measured gravimetrically after solvent extraction under N 2 stream. 2.4. Microscopic liver study A portion of the liver was fixed in 10 % phosphate-buffered formalin, dehydrated in ethanol, embedded in paraffin, and sectioned for histological examination using hematoxylin-eosin (HE) staining for general microscopy morphology (Servicio de Microscopía, CIC, University of Granada). Eight animals were evaluated per experimental group (n =32 samples per group) and four different preparations of each staining were analyzed for each animal. Histological alterations were evaluated in zones 1, 2, and 3 of the acinus. For semi-quantitative evaluation of liver damage, an initial scoring was done using the Brunt evaluation based on the following parameters: macrovesicular steatosis, microvesicular steatosis, ballooning, periportal inflammation, centrilobular inflammation, and fibrosis. For the semi-quantitative evaluation, the following cross-scale was used: (−) non-existent, (+) mild, (++) mild-moderate, (+++) moderate, (++++) severe [17]. In case of differences among treatments being found in the Brunt evaluation, the Non-Alcoholic Steatohepatitis (NASH) semi-quantitative scoring system of Kleiner et al. [18] as adapted by Chen et al. [19] was used to evaluate the degree of NASH development following the recommendations of Martinez et al. [9]. The scoring system comprised 14 histological features, 4 of which were evaluated semi-quantitatively: steatosis (0–3), lobular inflammation (0–2), hepatocellular ballooning (0–2), and fibrosis (0–4). Another nine features were recorded as present or absent. NAS score was calculated by the sum of steatosis grade, lobular inflammation, and ballooning. NAS of >5 correlated with a diagnosis of “NASH”, and scores of <3 were diagnosed as “not NASH” [18]. 2.5. Fatty acid profile of the liver A freeze-dried liver portion was extracted and methylated according to Lepage and Roy [20] for gas chromatography analysis of fatty acid profile using an Agilent 7890A chromatograph equipped with CTC Pal combi-xt model sampler and a Waters Quattro micro GC mass spectrometer detector. Individual fatty acid methyl esters (FAMEs) were separated with a 30 ×0.25 mm ZB Fame capillary column (0.2 μ m thickness) (Phenomenex, Torrance, CA, USA). The gas chromatography conditions were as follows: injector temperature 250 ◦C, injection volume 2 μ L Split (proportion 10:1), temperature gradient from 100 ◦C to 210 ◦C with a rate of 4 ◦C/min, hold time 5 min. The flow rate of the carrier gas (Helium) was 1 mL/min. The analysis time was 40 min and the measurement range 45–450 uma (scan mode). Chromatographic data were recorded and integrated using Masslynx, version 4.1 software. FAMEs were identified using analytical standards and mass spectral library. Peak areas were measured and used to calculate the percentage of each fatty acid related to the total sum of all the fatty acid areas in the sample. Furthermore, some products-to-precursor fatty acid ratios were used as indices of desaturase or desaturase-elongase enzyme activities in the liver as described by Gonzalez-Torres et al. [21] using the following formulas. Delta-6-elongase-desaturase activity: (i) docosahexaenoic acid/linolenic acid (ii) arachidonic acid/linoleic acid Stearoyl-CoA activity (SCD): (i) palmitoleic acid/palmitic acid (ii) oleic acid/stearic acid Delta-5 desaturase activity: (i) arachidonic acid/eicosatrienoic acid 2.6. RNA extraction Total RNA was isolated from the liver of all rats in each experimental group (n =8/group). One hundred milligrams of tissue were homogenized in 1 mL of Tri-Reagent® (Sigma-Aldrich). The RNA was solubilized in RNAse-free H 2 O and treated with DNase (Applied Biosystems) to remove any DNA present in the sample. 2.7. Transcriptomics For hepatic transcriptomics analysis, six total-RNA samples per experimental group were randomly selected. Library preparation and Illumina sequencing were carried out at the IPBLN Genomics Facility (CSIC, Granada, Spain). Total RNA quality was verified by Bioanalyzer RNA 6000 Nano chip electrophoresis (Agilent Technologies). Every RNA sample showed a RIN value above 8.4. RNA-seq libraries were prepared using Truseq stranded mRNA kit (Illumina®) from 200 ng of input total RNA. Quality and size distribution of PCR-enriched libraries were validated through Bioanalyzer High Sensitivity DNA assay and concentration was measured on the Qubit® fluorometer (Thermo). Final libraries were pooled in an equimolecular manner and then diluted and denatured as recommended by Illumina NextSeq 500 library preparation guide. The 75 ×2 nt paired-end sequencing was conducted on a NextSeq 500 sequencer with a final output of 70 Gbp and a quality score (Q30) of 97 %. The reads from Illumina paired-end sequencing were qualitychecked using FastQC v0.11.9 [22] and MultiQC v1.9 [23]. It was verified that throughout the sequence of the reads, their average quality presented a Phred nitrogen base quality score >30, so it was not necessary to filter the data. Then, the mapping of the reads was carried out. For this, HISAT2 v2.2.1 was used [24]. The Rattus norvegicus genome obtained from Ensembl (Rattus_norvegicus.Rnor_6.0.dna.toplevel.fa file) was chosen as the reference genome for alignment and the reads. The SAM files resulting from these mappings were ordered, transformed into BAM files, and indexed, all using SAMtools v1.10 [25]. In the next step, the count of the reads that mapped against each of the genes was carried out, for which the BAM files obtained in the previous step, the annotation file in GTF format of the reference genome (accession GCA_000001895. 4), and the feature counts v2.0.1 [26] were used. Finally, differential expression analysis was performed using the count tables obtained in the previous step and the programs DESeq [27], DESeq2 [28], and edgeR [29]. Transcripts were defined as differentially expressed when the fold change (FC) between the groups (SD1, HFHF1, SD2, HFHF2) was >1.3 (percentage of change +30 %; upregulated) or <−1.3 (percentage of change −30 %; downregulated) and the P-value of the Student's t-test was <0.05. Identical or similar filter criteria were used in several recent studies [30,31]. 2.8. RNA extraction and quantitative RT-PCR To validate the transcriptomic analysis, targeted gene expression was conducted using RT-PCR. A total of 100–250 ng of RNA was reversetranscribed according to standard protocols using a Lifepro Thermal Cycler (Bioer Serves Life, P. R. China). Quantitative RT-PCR was performed with QuantStudio 12 K Flex Real-Time PCR System (Applied Biosystems), using primers for genes involved in cell structure, glucose A. García-Beltr´ an et al. Life Sciences 324 (2023) 121746 4 & lipid metabolism, and antioxidant & detoxifying capacity (Table 1). The PCR master mix reaction included the first strand cDNA template, primers, and 2×TaqMan® Fast Universal PCR Master Mix, No AmpErase® UNG (Applied Biosystems). Expression of the test gene was related to that of Actb reference measured in parallel in the same sample using the ΔCt method. The 2 −ΔΔCt method was used to analyze the data in reference to the control group. 2.9. Antioxidant and detoxifying enzymes activity assays A fresh liver aliquot was homogenized (1:10 w/v) in 50 mM phosphate buffer (pH 7.8) containing 0.1 % Triton X-100 and 1.34 mM diethylene triamine pentaacetic acid (DETAPAC) using a Micra D-1 homogenizer (ART moderne labortechnik) at 18,000 rpm for 30 s followed by treatment with a Sonoplus HD 2070 ultrasonic homogenizer (Bandelin) at 50 % power three times for 10s. Liver homogenates were centrifuged at 13,000 ×g, 45 min, 4 ◦C, and the supernatant was used to determine the activity of antioxidant enzymes. Catalase (CAT) activity was measured by the method of Cohen et al. [32] and expressed as enzyme units calculated by the following formula: ln(A1/A2)/t, where ln is the natural log, A1 and A2 are the observed absorbances at the two selected time points, and t is the reaction time between the two points. Total Glutathione Peroxidase (GPx) activity was determined by the coupled assay of NADPH oxidation [33] using cumene hydroperoxide as a substrate. The enzyme unit was defined as nmol of NADPH oxidized per min. Total superoxide dismutase (SOD) activity was measured as described by Ukeda et al. [34]. Mn-SOD activity was determined by the same method after treating the samples with 4 mM KCN for 30 min. CuZn-SOD activity resulted from subtracting the Mn-SOD activity from the total SOD activity. One unit of SOD activity was defined as the enzyme needed to inhibit 50 % 2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide (XTT) reduction. NADH: Quinone reductase (QR) activity was determined according to the method of Ernster [35] using dichlorophenolindophenol (DCIP) and flavin adenine dinucleotide (FAD) as electron acceptor and inhibitor, respectively. The activity was expressed as enzyme units per min. The glutathione S-transferase (GST) was assayed by the method of Habig et al. [36], employing 1-chloro-2,4-dinitrobenzene (CDNB) as substrate, and expressed as enzyme units per min. The protein concentration was assayed by the method of Bradford [37]. 2.10. Statistical analysis The effects of dietary intervention with HFHF vs SD, and length of the experimental period as a measure of aging (13 vs 21 weeks) on hepatic weight, total fat content, fatty acid profile and indices, and antioxidant or detoxifying enzyme activity were analyzed by 2 ×2 factorial ANOVA, with dietary intervention and length of the experimental period as main treatments. The use of 2 ×2 factorial ANOVA is based on the potential interactions among the two interventions assayed (high-fat highfructose diet, length of experimental period) being significant in our statistical model in addition to single effects. To reinforce the potential integrative strength of the statistical model implemented, the R 2 statistic has been included in the tables as a measure of the goodness of fit of the model, given that the coefficient of determination indicates the proportion of variability in a data set that can be accounted by the statistical model. Results are given as mean values and pooled SEM. Duncan's test was used to detect differences between treatment means. Differences were considered significant at P <0.05. Model assumptions were checked using the Shapiro-Wilk normality test and Levene's test for homogeneity of variance, and by visual inspection of frequency histogram, quantile-quantile plot, and residual and fitted value plots. Pearson's test was carried out on the different antioxidant & detoxifying enzyme data to test the correlation between the transcript expression and enzymatic activity; when Pearson's test showed r >0.4 and P < 0.05, results were considered statistically significant. SPSS v.25 was used for the statistical treatment. Student’s t-test was used to detect differences in fold change of transcript expression between the control and experimental groups (SD1, HFHF1, SD2, HFHF2). Table 1 Gene distribution in categories for expression study. Function Gen Protein Assay ID (Applied biosystems) Cell structure Anatomic cell structure Mmp15 Matrix metalloproteinase-15 Rn01536925_m1 Psmb9 Proteasome subunit beta type-9 Rn00680664_g1 Cell adhesion Col26a1 Collagen alpha-1(XXVI) chain Rn01499402_m1 Cell damage Cdkn1a Cyclin-dependent kinase inhibitor 1 Rn00589996_m1 Gadd45a Growth arrest and DNA damage-inducible protein alpha Rn01425130_g1 Cell proliferation Gdf15 Growth/differentiation factor 15 Rn00570083_m1 Myc Myc proto-oncogene protein Rn00561507_m1 Solute transport Abcg5 ATP-binding cassette sub-family G member 5 Rn01499073_m1 Slc2a2 Solute carrier family 2, facilitated glucose transporter member 2 Rn00563565_m1 Slc34a2 Sodium-dependent phosphate transport protein 2B Rn00584515_m1 Glucose & lipid metabolism Lipogenic action Agpat3 1-acyl-sn-glycerol-3-phosphate acyltransferase gamma Rn01428234_m1 Fabp5 Fatty acid-binding protein 5 Rn01461858_g1 Fasn fatty acid synthase Rn00569117_m1 Pnpla3 Patatin-like phospholipase domain-containing protein 3 Rn01502361_m1 Scd1 stearoyl-CoAdesaturase-1 Rn00594894_g1 Srebf1 Sterol regulatory element binding transcription factor 1 Rn01495769_m1 Pparg Peroxisome proliferator-activated receptor gamma Rn00440945_m1 Lipolytic action Cpt1a Carnitine palmitoyl transferase 1A Rn00580702_m1 Crot Peroxisomal carnitine O-octanoyltransferase Rn01526940_m1 Ppara Peroxisome proliferator activated receptor alpha Rn00566193_m1 Cholesterol metabolism Cyp1a1 Cytochrome P450 1A1 Rn01418021_g1 Cyp1a2 Cholesterol 25-hydroxylase Rn00561082_m1 Cyp7a1 Cholesterol 7 alpha-hydroxylase Rn00564065_m1 Hmgcr 3-hydroxy-3-methylglutaryl-coenzyme A reductase Rn00565598_m1 Glucose metabolism Gck Hexokinase-4 Rn00688285_m1 Antioxidant & detoxifying capacity Akr7a3 Aflatoxin B1 aldehyde reductase member 3 Rn00680664_m1 Gpx1 Glutathione peroxidase 1 Rn00577994_g1 Nfe2l2 nuclear factor erythroid 2 like 2 Rn00477784_m1 Nqo1 NAD(P)H:quinone oxidoreductase 1 Rn00566528_m1 Sod1 Superoxide dismutase [Cu-Zn] Rn00566938_m1 Ucp2 Mitochondrial uncoupling protein 2 Rn01754856_m1 A. García-Beltr´ an et al. Life Sciences 324 (2023) 121746 5 3. Results 3.1. Food, fluid, caloric intake, and bodyweight gain The influence of dietary treatment (SD or HFHF) and length of the experimental period (13 or 21 weeks) on food, fluid, caloric intake, and bodyweight gain of rats are shown in Fig. 1A–D. There was a marked dietary treatment effect on food intake derived from the higher amount of food consumed throughout the experimental period by rats on the standard diet (SD) compared to animals on the high-fat diet (HFHF) groups. In contrast, caloric and fluid intake was higher in rats fed HFHF along the entire experimental period vs SD-fed animals. Such differences led to greater bodyweight gain in rats that consumed the HFHF vs SD diet. 3.2. Liver weight and hepatic fat content Dietary treatment with a high-fat high-fructose diet (HFHF1, HFHF2) had a significant effect on both liver weight and total hepatic fat content (Fig. 2), which exhibited higher values in the former groups compared to those treated with the SD diet. The length of the experimental period significantly affected the amount of total liver fat in HFHF groups, resulting in greater values for that parameter at 21 vs 13 weeks. 3.3. Liver histological study Liver histological changes resulting from DIO along 13 or 21 weeks of experimental period are described in Figs. 3A–C and Supplemental Tables S1 and S2. Animals on the SD or HFHF dietary treatments for 13 weeks exhibited mild to moderate hepatic steatosis, mainly microvesicular, with a low degree of macrovesicular steatosis or cell ballooning (Fig. 3A–B). In the SD experimental group, between 33 and 66 % of rats exhibited steatosis, whereas that percentage raised to >66 % in the animals of the HFHF group (Supplementary Table S1). Conversely, rats fed for 21 weeks with the SD diet exhibited a similar degree of steatosis (33–66 %), but a lower extent of microvesicular and slightly higher macrovesicular changes. Between 33 and 66 % of the rats in the 21-week HFHF experimental group showed steatosis. However, macrovesicular changes experienced a marked increase at the expense of microvesicular steatosis, whereas a significantly higher number of hepatocytes showed clear signs of ballooning. Interestingly, macrovesicular steatosis took place mainly on the convex area of the selected 10 12 14 16 18 20 22 24 26 28 1 2 3 4 5 6 7 8 9 101112131415161718192021 yadreptarrepsmarg Week Food intake SD1 HFHF1 SD2 HFHF2 10 20 30 40 50 60 70 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 yadreptarrepsmarg Week Fluid intake SD1 HFHF1 SD2 HFHF2 A B Fig. 1. Effect of DIO and length of experimental period on food (grams per rat per day, Fig. 1A), fluid (tap water or 20 % fructose solution, grams per rat per day, Fig. 1B) or caloric intake (kcal per rat per day, Fig. 1C), and bodyweight gain (grams, Fig. 1D) of Sprague-Dawley rats. To differentiate the results more clearly, only the most representative weeks of the experimental period are included in the graphs. SD1, standard diet for 13 weeks; HFHF1, high-fat high-fructose diet for 13 weeks; SD2 standard diet for 21 weeks; HFHF2, high fat-diet for 21 weeks. Points and bars represent the mean and standard error of the mean, respectively (n =8). A. García-Beltr´ an et al. Life Sciences 324 (2023) 121746 6 hepatic lobule in the form of columns that extended radially (Fig. 3C). NAFLD scoring index (NAS), points out to hepatic histological alterations that may be potentially reverted such as steatosis, lobular inflammation, and ballooning. Scores for rats fed the SD diet along the 13 or 21-week experimental period corresponded to a value of 2 (no NASH), whereas DIO during 13 or 21 weeks led to values of 3 and 4 (probable NASH), respectively (Supplementary Table S2). 3.4. Fatty acid profile The effect of dietary treatment and length of the experimental period on hepatic fatty acid profile and ratios is presented in Table 2. The highest proportion of hepatic fatty acids corresponded to saturated (palmitic and stearic acids), followed by mono- (palmitoleic, oleic, and octadecenoic acids) and poly-unsaturated fatty acids (linoleic, dihomoγ-linolenic (DGLA), arachidonic and docosahexaenoic). Diet-induced obesity (DIO) had a significant effect on hepatic fatty acid profile, resulting in higher oleic and linoleic acid percentages, together with lower percentages of saturated acids (palmitic and stearic), palmitoleic, stearic, DGLA, and docosahexaenoic acids. Elongase–desaturase ratios (docosahexaenoic/linolenic and arachidonic/linoleic) were significantly modified by diet administration, with lower ratios in the animals that consumed SD vs HFHF. Furthermore, stearoyl CoA desaturase activity was differentially affected by DIO, showing higher values for the oleic/stearic ratio and lower for the palmitoleic/palmitic ratio. The palmitoleic/palmitic ratio was also altered by the length of the experimental period, declining at week 21. Δ5Desaturase activity (arachidonic/DGLA) was characterized by the opposite effects of DIO related to the length of experimental period (decrease at 13 weeks or increase at 21 weeks), thus resulting in a significant diet ×time interaction. 3.5. Liver transcriptomic analysis A complete hepatic transcriptomic profile (>32,000 genes, Fig. 4) was carried out in SD or HFHF-fed rats to study the influence of dietary treatment and length of experimental period on cell structure, glucose, 0 20 40 60 80 100 120 140 1 2 3 4 5 6 7 8 9 101112131415161718192021 yadreptarreplack Week Caloric intake SD1 HFHF1 SD2 HFHF2 200 300 400 500 600 700 800 900 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 grams Week Bodyweight gain SD1 HFHF1 SD2 HFHF2 C D Fig. 1. (continued). 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 5 10 15 20 25 SD1 HFHF1 SD2 HFHF2 Total fat (g) Liver weight (g) Liver Weight Total hepatic fat Aa Cb Aa Bb Fig. 2. Effect of dietary treatment and length of experimental period on liver and total hepatic fat weight (g). Columns or points and bars represent the mean and standard error of the mean, respectively (n =8). Values within each treatment followed by different letters are significantly different (A–C, liver weight; a–b, total liver fat) P <0.05. SD1, standard diet for 13 weeks; HFHF1, high-fat high-fructose diet for 13 weeks; SD2 standard diet for 21 weeks; HFHF2, high-fat high-fructose diet for 21 weeks. Liver weight, Diet effect: P < 0.001; Diet ×Time: P =0.02. Total fat, Diet effect: P <0.001. A. García-Beltr´ an et al. Life Sciences 324 (2023) 121746 7 Ballooning AB Portal infiltration Macrovesicular Microvesicular A B C Fig. 3. Light microscopic images of liver preparations from rats fed SD or HFHF diets along 13 or 21 weeks of experimental period. Fig. 3A, effect of dietary induction of obesity and length of experimental period on liver histology (hematoxylineosin stain) of Sprague Dawley rats (50×). (A) SD1, (B) SD2, (C) HFHF1, (D) HFHF2. Photographs are representative of livers of 8 different rats for each experimental group. Fig. 3B, major alterations in hepatic histology caused by the dietary treatment and length of experimental period. A, steatosis, B, inflammation (hematoxylineosin stain, 200×). Photographs are representative of livers of 8 different rats for each experimental group. Fig. 3C, micrographs depicting the convex area of hepatic preparations from different experimental groups of Sprague-Dawley rats (hematoxylin-eosin stain, 20×). A, standard rodent chow (13 weeks), B, HFHF2. Photographs are representative of livers of 8 different rats for each experimental group. SD1, standard diet for 13 weeks; HFHF1, high-fat high-fructose diet for 13 weeks; SD2 standard diet for 21 weeks; HFHF2, high-fat high-fructose diet for 21 weeks. A. García-Beltr´ an et al. Life Sciences 324 (2023) 121746 8 and lipid metabolism as well as antioxidant and detoxifying enzyme activity. As expected, profound changes in hepatic gene expression were induced by these two factors. When comparing the DIO and lean rats on week 13 of the experimental period, a total of 4506 hepatic transcripts were differentially regulated according to the two-filter criteria (FC > 1.3 or <−1.3, P <0.05). Of those, 645 transcripts were upregulated and 3861 downregulated in the HFHF1 vs SD1 group. Similar comparisons were made among the former groups and those in which the experimental period lasted for 21 weeks: SD2 vs SD1 (948 upregulated, 1551 downregulated), HFHF2 vs HFHF1 (3505 upregulated, 1364 downregulated), and HFHF2 vs SD2 (1222 upregulated, 1781 downregulated). Among those hepatic transcripts differentially regulated depending on the consumption of HFHF diet or length of experimental period, 32 were selected based on their physiological action and filtered into three categories: cell structure, glucose/lipid metabolism, and antioxidant & detoxifying capacity (Table 1). 3.6. Hepatic mRNA expression To confirm the results of the hepatic transcriptome analysis, RTqPCR expression analysis was carried out in the genes selected. The effects of dietary treatment and length of experimental period on the hepatic expression of transcripts involved in cell structure, lipid metabolism, and antioxidant & detoxifying activity are presented in Tables 3A–3D. The expression of genes grouped in the category of cell structure was differentially affected by the dietary treatment with an HFHF diet that resulted in a significant induction of Col26a1, Psmb9, Gadd45a, and Slc2a2 compared to the SD diet on week 13 (Table 3A), whereas Cdkn1a, Myc, Abcg5, and Slc34a2 were significantly downregulated. On the other hand, HFHF treatment caused a significant increase in expression of Cdkn1a, Gdf15, Myc, and Slc34a2 vs SD on week 21, whereas it significantly decreased that of Slc2a2. Length of experimental period also exhibited a significant effect on gene expression, causing a general decrease in expression of most of the transcripts studied with the exception of Mmp15 in both SD and HFHF groups or CdKn1a and Myc only in HFHF treatment. Regarding the expression of lipogenic genes (Table 3B), mid-term DIO caused a marked up-regulation of transcription factor Srebf1 and transcripts Hmgcr, Pnpla3, Fasn, Scd1, and Fabp5 associated to cholesterol synthesis, hepatic fat accumulation, lipid and fatty acid synthesis or lipid transport on week 13. In contrast, most of the former transcripts and those related to cholesterol metabolism were down-regulated by HFHF consumption on week 21, which showed instead a significant upregulation of transcription factor Pparg and transcript Gck associated to glucose metabolism. Aging showed a differential action on either SD or HFHF groups. In the former, it resulted in down-regulation of transcription factors and marked up-regulation of genes associated to cholesterol metabolism, hepatic fat accumulation, lipid and fatty acid synthesis. In contrast, aging down-regulated most transcripts related to lipogenesis in the HFHF intervention with the exception of Gck that was up-regulated. Gene expression of lipolytic transcripts Cpt1a, and Cyp7a1 was significantly down-regulated by the dietary treatment at both midand long-term obesity induction (Table 3C), while Crot was not affected at 13 weeks and exhibited a significant up-regulation on week 21. No major effect was found for the transcription factor Ppara. Length of experimental period also affected gene expression in both control and obesogenic groups; in the former, a significant up-regulation of the transcription factor Ppara was observed whereas the opposite was true for Crot, Cpt1a, and Cyp7a1 expression. Similar effects were observed in the obesogenic groups. Results of antioxidant and detoxifying enzyme gene expression are marked by two different trends. First, a significant interaction between DIO and the length of experimental period characterized the expression of Akr7a3 or Sod1 that were significantly up-regulated by DIO on week 13, but significantly down-regulated by this same treatment on week 21 (Table 3D). Second, a constant effect of the dietary treatment was observed along the experimental period for Gpx (up-regulated) or Ucp2 (down-regulated) expression. A clear aging effect in both SD and HFHF groups was only observed for the expression of the transcription factor Nfe212 and for Sod1 which were lower on week 21 vs 13. The expression of Akr7a3 and Nqo1 transcripts was not affected by length of experimental period on the control SD groups, but exhibited a significant down-regulation in the obese animals. 3.7. Hepatic antioxidant and detoxifying enzyme activities The effects of dietary treatment and length of experimental period on hepatic antioxidant and detoxifying enzyme activities are shown in Table 4. DIO had a significant effect during the entire experimental period in CAT, Mn-SOD, GST, and QR activities. CAT and GST were increased by the HFHF treatment whereas the opposite action was Table 2 Influence of diet-induced obesity and length of experimental period on hepatic fatty acid profile (%) and fatty acid indices of rats. SD 13-weeks HFHF 13-weeks SD 21-weeks HFHF 21-weeks R 2 SEM Diet effect Time effect Diet ×Time effect Palmitic (C16:0) 29.9 A 27.0 A 33.8 B 26.4 A 0.58 0.84 P <0.001 P =0.096 P =0.026 Palmitoleic (C16:1) 5.92 B 4.37 AB 5.39 B 2.46 A 0.44 0.50 P <0.001 P =0.038 P =0.222 Stearic (C18:0) 11.4 B 7.56 A 10.1 AB 9.41 AB 0.27 0.68 P =0.006 P =0.71 P =0.048 Oleic (C18:1n9) 25.6 A 31.2 B 26.4 AB 29.3 AB 0.21 1.26 P =0.006 P =0.713 P =0.339 Linoleic (C18:2n6) 9.56 AB 14.1 BC 8.61 A 17.4 C 0.55 1.10 P <0.001 P =0.356 P =0.098 Dihomo-γ-linolenic (C20:3n6) 0.36 B 0.32 AB 0.36 B 0.21 A 0.36 0.03 P =0.004 P =0.063 P =0.086 Arachidonic (C20:4n6) 10.5 A 7.84 A 9.55 A 9.73 A 0.13 0.91 P =0.242 P =0.631 P =0.18 Docosahexaenoic (C22:6n3) 2.28 B 1.23 A 2.04 B 1.61 AB 0.32 0.18 P =0.001 P =0.741 P =0.137 Octadecenoic (C18:1n7) 4.08 B 3.91 AB 3.16 AB 2.83 A 0.27 0.26 P =0.391 P =0.002 P =0.789 Others 1.28 A 1.61 B 1.19 A 1.34 A 0.44 0.06 P =0.001 P =0.01 P =0.161 SFA 41.4 B 35.6 A 44.8 C 36.0 A 0.75 0.76 P <0.001 P =0.034 P =0.101 MUFA 35.6 A 39.6 A 35.0 A 34.7 A 0.07 1.62 P =0.309 P =0.138 P =0.239 PUFA 23.0 AB 24.8 AB 20.2 A 29.3 B 0.29 1.63 P =0.006 P =0.645 P =0.058 DC/LN 11.2 AB 5.65 A 16.1 B 5.20 A 0.53 1.42 P <0.001 P =0.169 P =0.103 ARA/LE 1.13 B 0.62 B 1.22 B 0.61 A 0.53 0.09 P <0.001 P =0.709 P =0.599 PE/PI 0.20 B 0.16 B 0.16 AB 0.09 A 0.39 0.02 P =0.006 P =0.004 P =0.403 OLE/STE 2.61 A 4.65 A 2.75 A 3.75 A 0.16 0.52 P =0.014 P =0.516 P =0.375 ARA/EI 34.9 AB 27.0 A 26.8 A 49.8 B 0.44 3.53 P =0.068 P =0.077 P =0.001 SFA, saturated fatty acids; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; DC/LN, docosahexaenoic/linolenic acid (desaturase–elongase); ARA/LE, arachidonic/linoleic acid (desaturase–elongase); PE/PI, palmitoleic/palmitic acid (Scd-1 activity); OLE/STE, oleic/stearic acid (Scd-1 activity); ARA/EI, arachidonic/DGLA acid (5-desaturase). Results are means of 8 rats. R 2 , coefficient of determination, SEM, pooled standard error of the mean. A-C Means within the same row with different letters differ significantly (P <0.05). SD, standard diet, HFHF, high-fat high-fructose diet. A. García-Beltr´ an et al. Life Sciences 324 (2023) 121746 9 Fig. 4. Volcano plot showing the differentially regulated hepatic transcripts between the experimental groups (n =6): (a) SD1 vs HFHF1, (b) SD1 vs SD2, (c) SD2 vs HFHF2, and (d) HFHF1 vs HFHF2. The double filtering criteria are indicated by horizontal (P-value <0.05) and vertical (fold change (FC): >log2 (1.3) or <log2 (−1.3)) dashed lines. Transcripts in the upper left and the upper right corner represent the downregulated and the upregulated transcripts, respectively. A. García-Beltr´ an et al.