A High-Fat Diet Enriched with Fish Oil Affects Metabolic Pathways in The Hypothalamus of Rats: A Proteomic Approach


 Background: Fatty acids of the n-3 series are recognized as a healthy type of fat but the effects of excess n-3 consumption have not been fully explored. The hypothalamus is a key site in the control of energy homeostasis and a better understanding of the effects of dietary challenges on this brain site is necessary.Objective: To evaluate whether the long-term intake of a high-fat diet enriched fish-oil, rich in n-3 fat, modify the proteins content of the hypothalamus.Methods: Male rats received for 8 weeks either a Control normolipidic diet or a hyperlipidic Fish oil diet. The hypothalami were dissected and processed for proteomic analysis, performed by LC-MS/MS. False Discovery Rate (FDR) correction for multiple testing was performed. Enriched pathways were determined using the Panther online platform.Results: Twenty-nine proteins were significantly altered between the Control and the Fish groups, of which 15 were down-regulated and 14 were up-regulated in the Fish group. Pathway-enrichment analysis showed that 10 of those proteins were related to “Metabolism of Carbohydrate” and “TCA Cycle and Respiratory Electron Transport” pathways. Of these, 7 proteins were upregulated (Aldo-keto reductase family 1 member B1; Aspartate aminotransferase, mitochondrial; Glucose-6-phosphate isomerase; Malate dehydrogenase mitochondrial; L-lactate dehydrogenase A chain; Dihydrolipoyllysine-residue acetyltransferase component of pyruvate dehydrogenase complex, mitochondrial; Alpha-enolase) and 3 were downregulated (Phosphoglycerate mutase 1; NADH dehydrogenase [ubiquinone] flavoprotein 2, mitochondrial; NADH dehydrogenase [ubiquinone] 1 alpha subcomplex subunit 9, mitochondrial). Conclusions: The chronic intake of a hyperlipidic diet rich in fish oil led to modifications on the expression of several hypothalamic proteins related to metabolic pathways. There are indications of possible pro-inflammatory activation and impairment of respiratory chain Complex 1 function, probably due to the excess fat content of the diet. Also, a protective role played by the type of fat was suggested by the findings indicating possible decrease of reactive oxygen species production, maintenance of NADH delivery to the respiratory chain, and production of anorexigenic signals. The results suggest that, although n-3 fat has recognized healthy properties, its consumption in excess affects the hypothalamus with deleterious effects. Further studies are necessary to verify these effects.


Background
Obesity prevalence has been pointed out as a pandemic health condition worldwide. The rise of body fat is known to increase the risk of development of comorbidities as cardiometabolic diseases, osteoarthritis, dementia, depression, and some types of cancer. The imbalance between food intake and energy expenditure is one of the causes of obesity and the excess ingestion of fatty acids favors this condition [1][2]. Polyunsaturated fatty acids (PUFA) induce more fat oxidation and energy expenditure than saturated fatty acids, indicating that the type of fat is relevant to the development of obesity [3]. PUFAs n-3 have been implicated as healthy fatty acids [4]. Our laboratory has previously evaluated the effects of a high-fat sh-oil diet in rats. We have shown that it prevented the activation of hypothalamic orexigenic nuclei and in ammatory mediators induced by a high-fat soy-oil diet [5][6]. It has also induced lower serum triglycerides levels and prevented the development of depressive-like behaviors, in comparison to a high-fat lard diet [7]. However, we have also found that the high-fat sh-oil diet increased serum glucose levels and impaired serotonin-induced hypophagia [8][9]. The above results indicate the need of further studies to better understand the role played by n-3 fatty acid in the development of the metabolic derangements of obesity. Since the hypothalamus is a key site in the control of energy homeostasis [10], a better understanding of the effects of dietary challenges on this brain site is necessary.
The non-target proteomic technique has been widely used to perform exploratory analysis of several conditions. It has been shown that the intake of a high-fat lard diet by mice modulated hypothalamic proteins involved in metabolic pathways [11]. We have previously demonstrated that the obesity induced by intrauterine growth restriction affected hypothalamic pathways related to glucose metabolism and mitochondrial function in adult male rats while glutamate metabolism and GABAergic synapse were affected in adult female rats [12][13]. We have not been able to nd studies utilizing this approach to evaluate the effects of PUFA n-3-rich diets on the hypothalamus.
The present study thus used the proteomic technique to access the changes in this brain area after a chronic intake of a high-fat sh oil diet.

Materials And Methods
This study was approved by the Research Ethics Committee of the Universidade Federal de São Paulo -UNIFESP (CEUA N 1673/07) and animal protocols were conducted following the ethics procedures established by the Conselho Nacional de Experimentação Animal (CONCEA).
Two-month-old male Wistar rats were maintained (4-5 per cage) under controlled conditions of lighting (12:12h light-dark cycle, lights on at 6 am) and temperature (24 ± 1 o C) and randomly assigned to receive either standard chow diet (Control group: 2.7 Kcal/g, 15% energy from fat) (Nuvilab CR1, Nuvital, PR, Brazil) or a high-fat diet enriched with sh oil (Fish group: 4.4Kcal/g, 45% energy from fat) ad libitum for 8 weeks. The sh diet contained (w/w) 50% of standard chow diet, 20% of sh oil (ROPUFA® '75′ω-3, Roche, DSM Nutritional Products, Brazil), 10% sucrose and 20% casein, to obtain a protein/energy ratio similar to that of the Control diet. Once a week, each rat was weighed, and a known amount of diet was introduced in the cage and weighed after 24 hours. Food intake was estimated by rat by dividing the cage value by the number of rats in the cage. Feed e ciency was calculated as (body weight gain/energy intake) x 100 [14].
After the 8 weeks of treatment, the animals were fasted overnight and euthanized. The hypothalamic were rapidly dissected and frozen in liquid nitrogen and kept at -80 o C until further analysis.

Protein sample preparation
Proteomic analysis was performed as previously described, with minor modi cations [12]. Because it is a complex structure, the entire hypothalamus was used. Each hypothalamus was homogenized in 1 mL of lysis buffer (8 M urea, 75 mM NaCl, 1 M Tris, and protease inhibitor cocktail (Roche Diagnostics, Indianapolis, IN, USA)). After centrifugation (19,000 g for 30 minutes at 4 o C) the protein concentration in the supernatants was determined (2-D Quant Kit GE Healthcare, Waukesha, WI, USA). Samples were pooled (3 samples/pool, 2 pools/group) and aliquots of 900 µg of protein were subjected to dialysis (Ultra-4 Centrifugal 3,000 NMWL lter devices, Merck Millipore) against 50 mM ammonium bicarbonate.

LC-MS/MS, data processing and database search
Proteomic analysis was performed by data-independent acquisition experiments in a nanoAcquity UPLC system coupled to a Synapt G2 HDMS Q-TOF mass spectrometer (Waters). Digested samples (2 pools per group, 3 technical replicates) were injected onto a trap column (nanoAquity C18 trap column Symmetry 180 µm x 20 mm, Waters), and then transferred by an elution gradient (phase B gradient from 7-35% for 92 minutes at a 275 nL/min) to an analytical column (nanoAcquity C18 BEH 75 µm x 150 mm, 1.7 mm, Waters). The mobile phase A was 0.1% formic acid in water and the mobile phase B was 0.1% formic acid in acetonitrile. Data were acquired in HDMS E mode, switching from low (4 eV) to high (ramped from 19 to 45 eV) collision energy. For external calibration, Glu-brinopeptide B solution (500 fmol/mL, in 0. % formic acid in 50% acetonitrile; Waters) was infused using a nanoLockSpray apparatus and scanned every 30 seconds [12].
Mass spectrometry raw data was processed with ProteinLynx Global Server software version 3.0.1 (Waters). Database search was performed using Rattus norvegicus sequences obtained in the UniProtKB/Swiss-Prot database (www.uniprot.org, including 9485 entries). The search parameters included: carbamidomethylation of cysteine as xed modi cation, oxidation of methionine, N-terminal acetylation, glutamine and asparagine deamidation as variable modi cations, up to 2 missed cleavage sites allowed for trypsin digestion, and automatic fragment and peptide mass tolerance. The protein identi cation criteria comprised: a minimum of 1 fragment ion per peptide, 5 fragment ions per protein and 2 peptides per protein, and the false discovery identi cation rate was set at 5%. Label-free quantitative assessments based on peptide intensities were performed for relative quantitation [15].
Results were exported into Excel les. Proteins identi ed in at least 2 technical replicates of each pool were included in the analysis. Normalization was performed in each sample according to the sum of protein intensities.

Statistical analysis
Statistical analysis was performed in IBM SPSS Statistics 21 Software (IBM, Armonl, NY, USA). The Shapiro-Wilk normality test was applied to all variables, with exception of proteome data. Comparisons between Control and Fish groups were performed by unpaired Student's "t" test for parametric variables or Wilcoxon test for nonparametric variables.
Proteome data were analyzed by Student's "t" test. False Discovery Rate (FDR) correction for multiple testing was performed on R software (https://www.r-project.org/). Statistical signi cance was set at FDR < 0.05.

Pathway enrichment analysis
Enriched pathways were determined using the online platform Panther (version 15.0, released 2020-07-28; http://www.pantherdb.org/) [16]. The overrepresentation test was carried out with proteins differentially expressed as the input list, whole Rattus norvegicus proteome was used as reference and Reactome Pathway (version 65, released 2020-11-17) was chosen as annotation data set. Enriched pathways were based on Fisher test followed by False Discovery Rate correction and signi cance was set to FDR < 0.05.

Body weight and food intake
The cumulative caloric intake during the 8 weeks of diet treatment was signi cantly higher in the sh group and this led to a nal body weight 4.5% higher than that of the control group (Table 1). Figure 1 shows that the caloric intake of the Fish group was higher at the 1st rst and 5th weeks. The feed e ciency was higher in the Fish group during the 5th week of diet treatment. Together, these results indicate that, throughout diet treatment, the Fish group had light increases of cumulative caloric intake and feed e ciency, what was probably important to limit body weigh gain in response to the hyper caloric diet.

Proteomics
The mean number of proteins detected in each hypothalamic replicate was 553. After aplication of inclusion criteria for relative quantitation (presence in at least two technical replicates of each pool), 339 proteins were compared between the groups (additional le 1). Twenty-nine proteins were signi cantly altered between sh-diet and control rats, of which 15 were down-regulated and 14 were up-regulated in the sh-diet group ( Table 2). The pathway-enrichment analysis showed that 4 pathways were signi cantly affected by the intake of the sh-oil diet. In "Metabolism of carbohydrate" pathway, 6 proteins were affected, of which 5 were included in the "Glucose Metabolism" pathway. The "TCA cycle and respiratory electron transport" pathway had 5 proteins affected, of wich 3 were include in the "Pyruvate metabolism and TCA cycle" pathway. The "Innate immune system" pathway had 10 proteins affected, of which 7 were included in the "Neutrophil degranulation" pathway. Five affected proteins were related to "Cellular response to stress" (Table 3).

Discussion
The polyunsaturated n-3 fatty acid has been implicated as a health fatty acid [4]. The present study showed that 8 weeks of intake of the high-fat sh oil diet, rich in EPA (eicosapentaenoic acid) and DHA (docosahexaenoic acid), caused only a small non-signi cant increase of body weight in comparison to the control diet, along with a light increase of caloric intake and feed e ciency. These results agree with previous ndings of our laboratory [5][6][7][8] and indicate that the sh oil diet was protective against the potentially deleterious effects of high-fat intake. Accordingly, sh oil intake has been found to reduce the adverse effect of overfeeding in early life in rats [17].
The hypothalamus receives inputs from gastrointestinal-borne molecules, including the nutrients themselves and factors induced by nutrients, and this sensing affects the hypothalamic control of energy homeostasis [18]. Hypothalamic neuroin ammation induced by high-fat intake of saturated or n-6 PUFAs has been shown to impair this homeostatic role of the hypothalamus [19]. Contrarily, the chronic high-fat intake of sh oil diet decreased hypothalamic levels of IL-6 and TNF-α [5].
To better understand the effects of high-fat sh oil diet intake, we performed a proteomic analysis of the hypothalamus. We found that the 29 proteins differentially expressed in the Fish group belonged to 4 different main pathways: Metabolism of Carbohydrate, TCA cycle and Respiratory Electron Transport, Innate Immune System, and Cellular Response to Stress.
Glucose-6-phosphate isomerase, Phosphoglycerate mutase 1, Alpha-enolase participate in the glycolytic pathway, but they do not play a regulatory role in the pathway. Glucose-6-phosphate isomerase catalyzes the isomerization of glucose-6-phosphate to fructose-6-phosphate. A 20% decrease of this protein in the cerebral cortex of diabetic rats after carotid occlusion has been shown to exacerbate the ischemiainduced damage [20]. This enzyme is also known as neuroleukin and is involved in neuronal growth, differentiation, and motility, besides its function in glucose metabolism. Inhibition of this protein in PC12 cells increased the susceptibility to apoptosis [21]. In the present work, its increased expression in the Fish group could represent a protective mechanism.
Phosphoglycerate mutase 1 acts on the conversion of 3-phosphoglycerate to 2-phosphoglycerate in glycolysis. Its levels have been found to increase in human glioma tissue, proportionally to tumor grade, as well as in C6 glioma cells, and its knockdown inhibited cell proliferation and promoted apoptosis in U87 glioma cells [22][23]. The increased expression of this protein in cancer cells is consistent with the increase of glycolysis as main source of energy supply to the tumor [24]. On the other hand, its superexpression in mouse hippocampus facilitated phosphorylation of cAMP response element-binding protein (CREB) and increased cellular proliferation and neuroblast differentiation, responses indicative of a possible role of this protein in promoting tissue regeneration [25]. In the HT22 hippocampal cell line, phosphoglycerate mutase 1 activation attenuated ischemia-induced cell damage and decreased reactive oxygen species levels. In rodents, super-expression of this protein decreased the damage of the CA1 hippocampal area, after 4 days of brain ischemia [26]. The above data indicated that, although deleterious when expressed in cancer cells, phosphoglycerate mutase 1 had bene cial effects in the response of hippocampal cells to injury. We could not nd papers on the hipothalamic status of this protein in response to challenges potentialy affecting the hypothalamus. Importantly, in rat liver, six weeks of a high-fat saturated diet induced steatosis and increased phosphoglycerate mutase 1 by 57%, while the combination of Fuzhuan brick tea restored enzyme levels to those of control animals and ameliorated the hepatic fat accumulation [27].
Alpha-enolase is one of the sub-units of the enzyme enolase, responsible for the interconversion of 2phosphoglycerate and phosphoenolpyruvate in glycolysis. In human brain, alpha-enolase has been located mainly in the glia [28]. A proteomic study has found increase of alpha-enolase in the hippocampus and cingulate gyrus of Alzheimer`s disease patients [29]. On the other hand, analyzing oxidized proteins in the cortex Alzheimer`s disease patients another proteomic study found increased levels. The authors suggested that oxidation could lead to loss of protein function and impair glycolysis [30]. This is in agreement with data showing glucose hypometabolism in this pathology [31]. Interestingly, a prospective study found that the intake of n-3 fatty acids-rich seafoods by elderly subjects protected from the development of Alzheimer`s Disease, a nding that was associated with DHA intake [32].
Two mitochondrial enzymes of the malate-aspartate shuttle (MAS), aspartate aminotransferase and malate dehydrogenase, showed increased levels in the Fish group. MAS transfers to mitochondria the electrons of the NADH generated in the citosol during glycolysis. Decreased levels of mitochondrial MAS proteins have been found in a proteomic analysis of postmortem brain tissue (putamen, thalamus and parietal lobe) of schemic stroke victims [33]. In microglial Bv2 cells, the inhibition of aspartate aminotransferase decreased intracellular ATP levels and caused activation of apoptosis [34].
In the present study, aldo-keto reductase family 1 member B1, an enzyme important in detoxi cation processes, showed increased levels in the Fish group. In rats, increased levels of this enzyme were found in astrocytes after spinal cord injury, and associated with increased proliferation of astrocytes through stimulation of Akt pathways and glucose utilization [35]. On the other hand, it has also been attributted a pro-in ammatory role through its involvement in prostaglandins metabolism [36]. Increased levels have been reported in diabetes, asthma and sepsis and its inhibition decreased oxdative stress in these pathologies [37].
Additionaly, malate dehydrogenase is the last enzyme of the TCA cycle, convering malate to oxaloacetate with the production of 1 NADH. In synaptossomes, high levels of oxidative stress impaired NADH production by blocking aconitase and alpha-ketoglutarate dehydrogenase enzymes [38]. In HT22 hippocampal cells, the induction of oxidative stress up regulated mitochondrial malate dehydrogenase mRNA levels by 22% [39]. Since high-fat intake may induce oxidative stress [40], the increase of malate dehydrogenase seen in the present study could represent an attempt to maintain NADH levels in TCA.
The TCA cycle and respiratory electron transport pathway had other 4 affected proteins. Two proteins were upregulated: L-lactate dehydrogenase A chain (22%) and dihydrolipoyllysine-residue L-lactate dehydrogenase A chain is part of the lactate dehydrogenase enzyme, responsible for the conversion of glucose-derived pyruvate into lactate. In the brain, this enzyme is present mainly in astrocytes [41]. Astrocytes form lactate and release it to neurons, where it is transformed in pyruvate [42].
Lactate has been shown to have anorexigenic properties and L-lactate dehydrogenase is important to this function. The intracerebroventricular (i.c.v.) injection of pyruvate or lactate decreased food intake and body weight gain, effects abolished by the suppression of L-lactate dehydrogenase [43]. Blocking Llactate dehydrogenase centrally has been shown to also suppress the hypophagia induced by intraperitoneal lactate [44]. The increased of L-lactate dehydrogenase found in the present study could explain, at least in part, the absence of excess caloric intake of the Fish group.
The protein dihydrolipoyllysine-residue acetyltransferase is a component of the E2 enzyme of the pyruvate dehydrogenase complex, responsible for formation of acetyl-CoA from glycolysis-derived pyruvate. This enzyme has been attributed a protective role in situations of metabolic impairment. After expose of rats to chronic mild stress, the up-regulation of its hippocampal levels protected against the development of depressive-like symptoms [45]. Mild brain injury led to increased levels of the 3 enzymes of pyruvate dehydrogenase complex while severe injury decreased their levels. The authors indicated that the increased levels in mild brain injury was to protect the metabolism during a transiently malfunction of mitochondria assembly and the cells were unable to utilize galactose as a carbon source, indicating that oxidative phosphorylation was disrupted [49]. Also, decressed levels of this enzyme have been found in the brain of Alzheimer's Disease patients [50].

Conclusions
In conclusion, the chronic intake of a hiperlipidic diet rich in sh oil led to modi cations on the expression of several hypothalamic proteins related to metabolic pathaways. As shown in Fig. 2, there are indications of possible pro-in ammatory activation and impairment of respiratory chain Complex 1 function, probably due to the excess fat content of the diet. Also, a protective role played by the type of fat was suggested by the ndings indicating possible decrease of ROS production, maintenance of NADH delivery to the respiratory chain, and production of anorexigenic signals. Further studies are necessary to verify these effects. Availability of data and materials: "Not applicable".
Competing interests: "The authors declare that they have no competing interests". Caloric intake and feed e ciency during intake of control or sh-oil diet Caloric intake was an estimate of ingestion per rat in cage; N = 4 cages for Control Group; N = 3 cages for Fish Group. Feed e ciency N = 18 rats for Control Group; N = 13 rats for Fish Group. * p < 0.05 vs. control.