Liver Fatty Acid Composition and Inflammation in Mice Fed with High-Carbohydrate Diet or High-Fat Diet

Both high-carbohydrate diet (HCD) and high-fat diet (HFD) modulate liver fat accumulation and inflammation, however, there is a lack of data on the potential contribution of carbohydrates and lipids separately. For this reason, the changes in liver fatty acid (FA) composition in male Swiss mice fed with HCD or HFD were compared, at the time points 0 (before starting the diets), and after 7, 14, 28 or 56 days. Activities of stearoyl-CoA desaturase-1 (SCD-1), ∆-6 desaturase (D6D), elongases and de novo lipogenesis (DNL) were estimated. Liver mRNA expression of acetyl-CoA carboxylase 1 (ACC1) was evaluated as an additional indicator of the de novo lipogenesis. Myeloperoxidase activity, nitric oxide (NO) production, and mRNA expressions of F4/80, type I collagen, interleukin (IL)-6, IL-1β, IL-10, and tumor necrosis factor-α (TNF-α) were measured as indication of the liver inflammatory state. The HCD group had more intense lipid deposition, particularly of saturated fatty acids (SFAs) and monounsaturated fatty acids (MUFAs). This group also showed higher DNL, SCD-1, and D6D activities associated with increased NO concentration, as well as myeloperoxidase activity. Livers from the HFD group showed higher elongase activity, stored more polyunsaturated fatty acids (PUFAs) and had a lower omega-6/omega-3 fatty acid (n-6/n-3) ratio. In conclusion, liver lipid accumulation, fatty acids (FA) composition and inflammation were modulated by the dietary composition of lipids and carbohydrates. The HCD group had more potent lipogenic and inflammatory effects in comparison with HFD.


Introduction
Non-alcoholic fatty liver disease (NAFLD) is a condition whereby there is high hepatic lipid accumulation even under low alcohol intake [1]. NAFLD is not necessarily a disease, since it may be reversed by physical exercise [2], food restriction and body weight reduction [3]. However, some patients with NAFLD progress to nonalcoholic steatohepatitis (NASH) that can lead to fibrosis, cirrhosis, and eventually, liver failure [4].
NAFLD is the most common cause of liver diseases since this organ has a limited capacity for lipid storage. It is associated with obesity, insulin resistance, type 2 diabetes, hypertriglyceridemia, Table 1. Fatty acid composition (mg 100/g) of the high-carbohydrate (HCD) and the high-fat (HFD) diets.

Measurements of Diet and Liver Fatty Acid Composition
Total lipid contents of diets and livers were extracted using the method of Bligh and Dyer [14], in reduced scale.
The homogenized sample was weighed (1.000 ± 0.001 g) in 10 mL glass tubes and extraction performed by adding 2 mL methanol, 2 mL chloroform and 1 mL distilled water in different steps. The tubes were vortexed for 9 min and, after extraction, centrifuged at 3000 rpm for 5 min. The residual chloroform phase, containing total lipid extracted, was separated and the solvent removed using nitrogen gas flow. Fatty acid methyl esters (FAME) of diet and liver homogenates were prepared by ultrasound assisted total lipid methylation as described by Santos et al. [15]. FAME separation was performed by gas chromatography in a Thermo Scientific™ TRACE™ Ultra Gas Chromatographer (Thermo Scientific™, Waltham, MA, USA), fitted with a flame ionization detector (FID) and a fused-silica capillary column (100 m × 0.25 mm i.d., 0.25 µm cyanopropyl CP-7420 select FAME). The ultra-pure gas flows were 1.2 mL·min −1 carrier gas (hydrogen), 30 mL·min −1 make-up gas (nitrogen), 350 mL·min −1 synthetic air and 35 mL·min −1 hydrogen flame gas. The injected sample volume was 2.0 µL with split injection ratio 1:80. The injector and detector temperatures were 200 • C and 240 • C, respectively. The column temperature was maintained at 165 • C for 7 min, followed by a heating rate of 4 • C per min until reaching 185 • C, which was maintained for 4.67 min. After that, a new heating rate of 6 • C per min was applied until reaching 235 • C, which was maintained for 5 min, totalling 30 min of analysis.
Retention times and peak areas were determined using the Chrom-Quest™ software (Thermo Scientific™). For identification of FA, retention times were compared to those of standard methyl esters. FA contents in the diets and livers were expressed as mg/100 mg sample.

Determination of Nitric Oxide (NO) Production in the Liver
NO gas released from cells into the media reacts with water to produce nitrite and nitrate [16]. Nitrite and nitrate were then measured in the cell culture supernatant (200 µL) in triplicate. Griess reagent, containing sulfanilamide (1 g) in phosphoric acid (2.5 mL) and dihydrochloride of N-(1-naphtyl) ethylenediamine in milli-Q water (0.1 g), was added at room temperature. Absorbance was measured at 550 nm using an ELISA plate reader. NO production was calculated using a standard curve of sodium nitrite and results expressed as µM.

Determination of Myeloperoxidase Activity in the Liver
The livers were homogenized in phosphate buffered saline (PBS) and the homogenate was stirred in a vortex and centrifuged (2500 rpm) for 5 min.
The IMI was calculated by the sum of expressions of pro inflammatory factors divided by the sum of expressions of anti-inflammatory factors as follows: interleukin (IL)-6 + IL-1β + tumor necrosis factor-α (TNF-α) + F4/80 + type 1 collagen/IL-10.

Statistical Analysis
Results are reported as means ± standard deviation of the means and were analyzed by one-way ANOVA. Tukey and Student t-test using Graph-Pad Prism Version 5.0 software (GraphPad Software, San Diego, CA, USA) were used to assess differences between means. p-values < 0.05 indicate statistical significance.

Fatty Acid Composition of the Livers
In agreement with the diet composition (Table 1), livers from HCD and HFD mice had higher content of palmitic acid (SFA), oleic acid (MUFA) and linoleic acid (PUFA) as compared with other FAs (Table 2).   Lipid accumulation, calculated by the sum of all FA, was intensified (p < 0.05) during the experimental period for both HCD and HFD. The HCD group had high (p < 0.05) FA accumulation, particularly with respect to SFA and MUFA ( Table 3). The HCD group exhibited low (p < 0.05) levels of PUFAs (n-3 and n-6), due to more intense reduction of linoleic acid, eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) levels during the 56-day period (Tables 2 and 3). Table 3. Fatty acid family composition (mg 100/g of sample), and n-6/n-3 fatty acid, PUFA/SFA and MUFA/SFA ratios in the liver from mice fed with high carbohydrate diet (HCD) or high fat diet (HFD) at 0 (before starting the diets) or after 7, 14, 28 or 56 days. Arachidonic acid (AA) content was increased during the 56-day period in the livers of both groups but the changes were more pronounced in HFD mice ( Table 2). Decrease of n-6/n-3 fatty acid ratios were observed in the HFD group throughout the 56-day period ( Table 3). The PUFA/SFA ratio was reduced by both HCD and HFD, however, it was more pronounced in the HCD group (Table 3).
The MUFA/SFA ratio was increased in livers from the HCD group from day 7 ( Table 3).

Estimated Activities of SCD-1, D6D, Elongase and DNL in the Liver
Elevations (p < 0.05) of SCD-1 and D6D activities during the 56-day period were observed in the livers of HCD mice (Table 4). Liver ACC1 expression was increased in the HCD mice; the values expressed as mean ± standard deviation of 5-6 mice per group were: 1.02 ± 0.22 for HCD and 0.634 ± 0.18 for HFD. These values were significantly different as indicated by the Student t-test for p < 0.05.
The elongase activities were increased (p < 0.05) and decreased (p < 0.05) in the liver of the HFD and the HCD groups, respectively, during the experimental period. DNL was increased throughout the study either by HCD or HFD but the changes were more pronounced in HCD mice. Table 4. Estimations of enzyme activities (SCD-1, D6D and elongase) and of de novo lipogenesis (DNL) in the liver from mice fed with high carbohydrate diet (HCD) or high fat diet (HFD) at 0 (before starting the diets) or after 7, 14, 28 or 56 days.

Liver Myeloperoxidase Activity and Nitric Oxide Levels
Eleven mice that received the diets for 56 days were used in this analysis. Myeloperoxidase activity and NO levels were increased (p < 0.05) after 56 days of receiving HCD as compared to the HFD group (Table 5). Eleven mice that received the diets for 56 days were used in this analysis. F4/80, type I collagen, IL-6, IL-1β, TNF-α and IL-10 ( Table 6) mRNA expressions were not significantly different between HCD and HFD. However, the IMI was increased in the liver of HCD animals (p < 0.05) indicating a more intense inflammatory state in this group.

Liver FA Accumulation
In agreement with previous studies in mice [12] and humans [18], livers from the HFD and HCD groups had higher content of palmitic acid, stearic acid, oleic acid, linoleic acid, and arachidonic acid in comparison with other FA ( Table 2).
The more intense (p < 0.05) deposition of FA in the HCD group, which was inferred from the sum of all FA, was due to MUFAs being the main contributors (Table 3). These results may be explained as a consequence of increased DNL [19] and SCD-1 activity (Table 4). In fact, increased carbohydrate supply has been reported to stimulate DNL and SCD-1 activity [20].
The mechanisms by which DNL increases due to high carbohydrate diet involve SREBP-1c and ChREBP, which influence the expression of key genes involved in DNL such as acetyl-CoA carboxylase. Acetyl-CoA, generated from glucose, activates the transcription factors SREBP1c and ChREBP in the liver, which stimulate DNL [21]. In contrast, dietary FA are directly incorporated into triglycerides by diacylglycerol acyltransferase, and are not able to activate DNL [22]. n-3 PUFAs, found in high concentrations in the liver of HFD mice, prevent liver steatosis by inhibiting DNL via down regulation of SREBP-1c gene expression and lipogenic gene (FAS, ACC, and SCD-1) expressions, and stimulation of FA oxidation [23].
PUFA levels are modulated by the ingestion of linoleic acid and α-linolenic acid [24]. In accordance with this statement, high (p < 0.05) hepatic levels of linoleic acid and α-linolenic acid and their products of elongation and desaturation were found in HFD mice ( Table 2) as consequence of the high amount of essential PUFA in their diet (Table 1).
Low PUFA/SFA ratio has been associated with increased risk of atherosclerosis, cardiovascular diseases and diabetes [25]. Low liver PUFA/SFA ratio may indicate a predisposition for HCD mice to develop these latter diseases (Table 3).
In accordance with the fact that n-3 PUFAs reduce lipid content in NAFLD [26], the HFD group accumulated less FA, as expected by the high (p < 0.05) levels of EPA and DHA (Table 2). In addition, the serum levels of glucose, total cholesterol and triglyceride were not different between the two groups.

Activities of Elongase, Desaturase and SCD-1
Elongase is regulated by diet composition and changes in its activity can lead to alterations in cell lipid composition [27,28]. Moon et al. [29] have reported accumulation of palmitic acid, palmitoleic acid and reduced contents of stearic acid and oleic acid in knockout mice for this enzyme.
In rodents, about 90% of the stearic acid is synthesized from palmitic acid by elongase activity [29]. In accordance, mice fed with HFD, presenting high (p < 0.05) elongase activity, also had higher stearic acid levels ( Table 2). The contents of vaccenic acid, an elongation product of palmitoleic acid [30], was higher (p < 0.05) in the HCD group, in spite of the fact that the HCD group had lower elongase activity (Table 4). These results may be explained by the higher availability of palmitoleic acid (a precursor of vaccenic acid) in the livers from the HCD group (Table 2).
SCD-1 introduces a double bond on the ∆9 position in palmitic acid and stearic acid whereas D6D causes an unsaturation in linoleic acid and α-linolenic acid. These enzymes are expressed in several tissues, including the liver [30], and have an important role not only in the maintenance of plasma membrane lipid composition but also in the production of lipid signaling molecules such as eicosanoids [31].
An increase in SCD-1 and D6D activity was observed in the livers of the HCD group (Table 4). The changes in SCD-1 may be attributed to the high carbohydrate content of the diet [32,33]. The higher SFA level in the liver from the HCD group (Table 3) also increases SCD-1 activity and its RNAm expression [34]. Knockout mice for hepatic SCD-1 were protected against liver steatosis induced by HCD and this was associated with low rates of DNL [35].
In accordance with the fact that the HFD group had lower (p < 0.05) D6D and SCD-1 activity (Table 4), Vessby et al. [40] reported that this enzyme activity is inhibited by AA and PUFA which are increased in the HFD group (Tables 2 and 3).

Inflammation Associated with Liver FA Accumulation
FA modulate eicosanoid metabolism, act on plasma membrane and cytosolic signaling processes and influence activities of transcription factors involved in inflammation like NFκB and PPAR-γ [41].
Liver content of AA, a precursor of pro-inflammatory prostaglandins, thromboxans, leukotrienes and lipoxins [42] was increased in the HFD group (Table 2).
MUFAs have anti-inflammatory effects [43], and were found at high levels in the liver from the HCD group. Thus, a more prominent liver pro-inflammatory state in the HFD group could be predicted. However, the HFD liver had higher (p < 0.05) levels of n-3 PUFAs ( Table 2) that have anti-inflammatory effects [44,45]. This latter group had lower (p < 0.05) n-6/n-3 fatty acid ratio ( Table 3) in spite of the fact that similar n-6:n-3 fatty acid proportions were described for both high carbohydrate diet and high fat diet. The HFD group also had lower SFA levels (Table 3), especially palmitic acid (Table 2), which is able to raise nuclear factor kappa B (NF-κB) activity by binding to Toll-like receptor-2 (TLR-2) and 4, increasing expressions of pro-inflammatory cytokines such IL-6 and TNF-α [18,46].
Considering the pro-inflammatory properties of AA and SFA, and the anti-inflammatory effects of n-3 PUFA, the inflammatory states of the livers from the HCD and HFD groups were compared.
The liver inflammatory state was herein evaluated by the measurements of myeloperoxidase activity, NO content and inflammatory gene expressions (F4/80, type I collagen, IL-6, IL-1β, TNF-α and IL-10).
Increased myeloperoxidase activity and NO contents in the liver indicate a more prominent (p < 0.05) inflammatory state in HCD mice (Table 5). Myeloperoxidase activity is a marker of tissue neutrophil infiltration [47]. The higher (p < 0.05) activity of myeloperoxidase (Table 5) in the livers of HCD mice suggests a more intense hepatic neutrophil infiltration [47].
The F4/80 is one of the most specific cell-surface markers for macrophages. F4/80 is highly and constitutively expressed in resident tissue macrophages, including Kupffer's cells in the liver [51]. There were no differences in F4/80 mRNA expression, indicating similar macrophage infiltration in livers from both HCD and HFD groups ( Table 6).
There were no differences in liver cytokine mRNA expression between the HCD and HFD groups. However, the IMI was increased in livers from the HCD group mainly because IL-10 was poorly expressed in this group (Table 6). This information, together with the results of myeloperoxidase, NO and high n-6/n-3 fatty acid ratio, indicates an increased inflammatory state in livers from mice fed with HCD for 56 days, as compared with the HFD group. In accordance with these results, high postprandial blood glucose levels induced by high carbohydrate food intake increase NO generation, that in turn can combine with superoxide to produce peroxynitrite, a potent long-lived pro-oxidant molecule, contributing either to acute or chronic low-grade inflammation [52].
The results in livers from mice fed with HCD or HFD are summarized (Figure 1). The HCD (a) generates glucose in excess that reaches the liver at high quantities and stimulates DNL and SCD-1 activity. This stimulation leads to production of high quantities of SFAs and MUFAs. SFA, especially 16:0, activates NF-κB that increases NO production and tissue neutrophil infiltration indirectly. In addition, low quantities of n-3 PUFA and increased n-6/n-3 PUFA ratio, prompt inflammation; HFD mice (b) had lower FA accumulation and lower n-6/n-3 PUFA ratio in the liver that is well known as an anti-inflammatory condition. Low n-6/n-3 PUFA ratio inhibits DNL, reduces SCD-1 and NF-κB activities and results in low lipid accumulation. Abbreviations: DNL, de novo lipogenesis; SCD-1, stearoyl CoA desaturase-1; SFA, saturated fatty acids; MUFAs, monounsaturated fatty acids; PUFAs, polyunsaturated fatty acids; NF-κB, nuclear factor kappa B; NO, nitric oxide; IMI: inflammatory marker index.

Conclusions
The proportion of fat and carbohydrates in the diet modulated the deposition of lipids and composition of liver fatty acids. The liver from the HFD group had higher elongase activity and stored more n-3 and n-6 PUFAs. Increased lipid deposition particularly of SFAs and MUFAs, higher SCD-1 and D6D activities and ACC1 expression and DNL were reported in the liver from the HCD group. These changes were associated with a more intense liver inflammation state ( Figure 1). Figure 1. Summary of the results in livers from mice fed with either high carbohydrate diet (HCD) or high fat diet (HFD). Bold words indicate higher availability in comparison with the other group; black arrows indicate strong stimulation; green arrows indicate weak stimulation; red arrows represent inhibition. The HCD (a) generates glucose in excess that reaches the liver at high quantities and stimulates DNL and SCD-1 activity. This stimulation leads to production of high quantities of SFAs and MUFAs. SFA, especially 16:0, activates NF-κB that increases NO production and tissue neutrophil infiltration indirectly. In addition, low quantities of n-3 PUFA and increased n-6/n-3 PUFA ratio, prompt inflammation; HFD mice (b) had lower FA accumulation and lower n-6/n-3 PUFA ratio in the liver that is well known as an anti-inflammatory condition. Low n-6/n-3 PUFA ratio inhibits DNL, reduces SCD-1 and NF-κB activities and results in low lipid accumulation. Abbreviations: DNL, de novo lipogenesis; SCD-1, stearoyl CoA desaturase-1; SFA, saturated fatty acids; MUFAs, monounsaturated fatty acids; PUFAs, polyunsaturated fatty acids; NF-κB, nuclear factor kappa B; NO, nitric oxide; IMI: inflammatory marker index.

Conclusions
The proportion of fat and carbohydrates in the diet modulated the deposition of lipids and composition of liver fatty acids. The liver from the HFD group had higher elongase activity and stored more n-3 and n-6 PUFAs. Increased lipid deposition particularly of SFAs and MUFAs, higher SCD-1 and D6D activities and ACC1 expression and DNL were reported in the liver from the HCD group. These changes were associated with a more intense liver inflammation state ( Figure 1).