Characterization of the Cafeteria Diet as Simulation of the Human Western Diet and Its Impact on the Lipidomic Profile and Gut Microbiota in Obese Rats

The obesity pandemic has been strongly associated with the Western diet, characterized by the consumption of ultra-processed foods. The Western lifestyle causes gut dysbiosis leading to impaired fatty acid metabolism. Therefore, this study aimed to evaluate shifts in gut microbiota and correlate these with serum fatty acid profiles in male Wistar rats fed a cafeteria diet. Ten male rats were fed with standard diet (CTL, n = 5) and cafeteria diet (CAF, n = 5) for fifteen weeks. Body weight and food intake were recorded once and three times per week, respectively. At the end of the study, fresh fecal samples were collected, tissues were removed, and serum samples were obtained for further analyses. Gut microbiota was analyzed by sequencing the V3-V4 region of 16S rRNA gene. Serum fatty acid profiles were fractioned and quantified via gas chromatography. The CAF diet induced an obese phenotype accompanied by impaired serum fatty acids, finding significantly higher proportions of total saturated fatty acids (SFAs) and C20:3 n-6, and lower C18:1 n-7 and C18:3 n-3 in the phospholipid (PL) fraction. Furthermore, circulating C10:0, total n-3 and n-7 decreased and total monounsaturated fatty acids (MUFAs), including oleic acid C18:1 n-9, increased in the cholesterol ester (CE) fraction. The obesity metabotype may be mediated by gut dysbiosis caused by a cafeteria diet rich in C16:0, C18:0, C18:1 n-9 and C18:2 n-6 fatty acids resulting in a 34:1 omega-6/omega-3 ratio. Therefore, circulating C10:0 was associated with several genera bacteria such as Prevotella (positive) and Anaerotruncus (negative). Two classes of Firmicutes, Bacilli and Erysipelotrichi, were positively correlated with PL- C20:3 n-6 and CE- 18:1 n-9, respectively. TM7 and Bacteroidetes were inversely correlated with PL-SFAs and CE- 18:2 n-6, respectively.


Introduction
Nowadays, globalization has fostered a transformation towards a more industrialized food environment [1]. Lifestyle and environment can influence ultra-processed food choices that consist of industrial formulations rich in sodium, added sugar and fats, among other substances such as preservatives, artificial colors and artificial flavors [2]. This has led to an obesity epidemic resulting from a prolonged imbalance between energy intake and expenditure caused by a complex multifactorial interaction [3]. To explore the mechanisms involved in the development of obesity, animal models have been used to simulate the common cause of human obesity, employing hypercaloric diets such as high-fat and highsugar diets [4]. However, these diets do not adequately replicate the Western dietary

Fatty acid Composition of Cafeteria Diet
Lipid extracts of dry food samples (3 g) were obtained via the n-hexane extraction method [16]. Then, the lipid extracts were dried under vacuum in a rotary evaporator system (Büchi Rotavapor R-215, Büchi, Flawil, Switzerland) and the total lipid content was determined gravimetrically. The recovered lipids were transferred to vials and stored at −80 • C for fatty acid (FA) analysis.
The composition of the FAMEs was analyzed on an Agilent 8860 gas chromatography system (GC) equipped with a flame ionization detector (Agilent Technologies. Inc., Santa Clara, CA, USA) and a 100 m × 0.25 mm i.d., 0.2 mm film thickness fused-silica capillary column SP-2380 (Supelco Inc., Bellefonte, PA, USA). The injection samples (2 µL) were performed in split mode 1:15 ratio at 260 • C. The GC parameters and settings were set up according to the method of Castillo et al. [17]. The identification and quantification of FAs was carried out by comparing the retention times and peak areas of a standard mixture of FAMEs (GLC 566, Nu Chek Prep Inc., Elysian, MN, USA) to an internal standard, as explained by Castillo et al. [17]. The FA content was expressed as the percentage of dry matter and lipid content.

Animal Procedures
Ten 6-week-old male Wistar rats supplied from "Círculo ADN" S.A. de C.V. company (Mexico City, Mexico) underwent a microbiological study upon arrival at the Center for Research and Development in Health Sciences, as established by the Center's biosafety committee. The rats were kept without performing procedures in a two-week acclimatization period and were housed individually in a temperature-controlled room (21-23 • C) under a 12:12 h light:dark cycle with ad libitum water and food access (standard chow diet). Animals were randomly distributed into two groups (n = 5 per group) and were assigned a standard chow diet (CTL group) or cafeteria diet (CAF group) for fifteen weeks.
Food intake was recorded three times per week and body weight once a week. The daily calorie intake and food efficiency were calculated as previously described by Etxeberria et al. [18].
Fresh fecal samples were gathered in 15 mL Falcon tubes one day before the endpoint of the experiment, early in the morning and after the overnight fasting period, via abdominal massage. Fecal samples were frozen at −80 • C for further analyses.
At the end of the study, rats were euthanized after an overnight fasting, and blood was collected from the trunk. Serum samples were obtained via centrifugation (2000× g for 15 min) and were stored at −80 • C until further analyses. White adipose tissue depots including mesenteric (mWAT), epididymal (eWAT), retroperitoneal (rWAT) and subcutaneous (sWAT), as well as brown adipose tissue (BAT), soleus and gastrocnemius muscles, the liver, and the spleen were collected, weighed and immediately frozen in liquid nitrogen.
The research was conducted in agreement with the national and institutional guidelines of the Animal Research Bioethics Committee and was approved by the Ethics Committee of the Faculty of Public Health and Nutrition (CE 1/2022-01).

Fatty Acid Profiling of Serum Cholesterol and Phospholipids Fractions
The lipids of serum samples (150 µL) were extracted via a modification of the Folch method [17]. Serum lipids were fractioned in five sorts of lipids with dissimilar polarities; namely, cholesterol ester (CE), triglyceride (TG), mono and di-glyceride (MG+DG), free fatty acid (FFA) and phospholipid (PL) fractions, employing solid-phase extraction aminopropyl cartridges (500 mg, 3 mL, Bond Elut NH2, Agilent Technologies Inc., Santa Clara, CA, USA), as described by Castillo et al. (2020). Cholesteryl nonadecanoate (CE-C19:0, 450 µg/mL), from Nu Chek Prep Inc. (Elysian, MN, USA), and 1,2-Diheptadecanoyl-sn-glycerol-3phos-phoryl-choline (PC-C17:0/C17:0, 300 µg/mL, Abcam, Cambridge, UK) were added as internal extraction standards. The CE and PL lipid fractions were selected for their characterization and were transmethylated to produce the FAMEs, as previously reported by Castillo et al. [17]. TG-C11:0 (110 µg/mL) was added to each fraction as an internal quantification standard. The FAMEs of CE and PL fractions were analyzed using an Agilent 8860 GC-FID. The identification and quantification of FAs were performed, as previously mentioned. The absolute concentration (µg/mL) of individual FAs in each fraction was normalized to the total amount of FAs and expressed as the relative percent concentration.

Gut Microbiota Analysis from Male Wistar Rats by Illumina Mi-Seq Sequencing
DNA from the fecal samples was extracted using the QIAamp DNA Stool Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. DNA quality and quantity were evaluated using spectrophotometry (NanoDrop 8000, Thermo Scientific, Wilmington, DE, USA) and a Picogreen fluorometer, respectively. The V3 and V4 regions of the 16S rRNA gene were amplified using specific forward (5 TCGTCGGCAGCGTCAGATG TGTATAAGAGACAGCCTACGGGNGGCWGCAG 3 ) and reverse primers (5 GTCTCGTG GGCTCGGAGATGTGTATAAGAGACAGGACTA CHVGGGTATCTAATCC 3 ) containing the Illumina adapter overhang nucleotide sequences. PCRs were carried out using the following parameters: 3 min 95 • C pre-denaturation; followed by 25 amplification cycles: 95 • C for 30 s, 63 • C for 30 s, 72 • C for 30 s and 72 • C for 5 min. Amplicon size was analyzed via 3% agarose gel electrophoresis (PowerPac, Bio-Rad, Hercules, CA, USA) and amplicons were purified using AMPure XP. The indices of the Nextera XT kit were added via new PCR, following Illumina's protocol [19].
DNA sequencing was performed using the Illumina MiSeq platform according to the protocol suggested by Illumina (16S metagenomic sequencing library preparation). At the end of the process, a file in FASTQ format was obtained for each sample that included the readings. The data obtained were subjected to quality control and analyzed using QIIME v.1.9. Sequences with 97% identity were grouped into operational taxonomic units (OTUs) and taxonomically assigned by comparison against the Greengenes database. The readings were assigned to the phylum, class, order, family and genus levels. In addition, the α-diversity (Shannon index) was determined.

Statistical Analysis
Statistical analyses were conducted using SPSS 25.0 (SPSS Inc., Chicago, IL, USA). Results are expressed as mean ± SEM. Normality of data distribution was obtained using the Shapiro-Wilk test and Levene's test with a p-value cutoff of 0.05. For normally distributed data, a t-Student test for independent samples was performed to analyze the statistical differences between group means and its non-parametric equivalent, the Mann-Whitney U test was used for non-normally distributed data. The correlation between different variables was performed using the Pearson test or its non-parametric Spearman equivalent. A level of probability of p < 0.05 was set as statistically significant. Sample sizes can be found in the figure legends, where n represents the number of animals used in each analysis.

Effect of Cafeteria Diet on Body Weight and Biometric Parameters in Male Wistar Rats
In the present study, the CAF group had higher daily food intake (32.36 ± 1.23 g) as well as higher food efficiency (3.22 ± 0.12 g/100 kcal) compared with the CTL group (27.20 ± 0.42 g, 2.82 ± 0.03 g/100 kcal, respectively). Consistent with those findings, after five weeks eating the cafeteria diet, greater body weight gain was obtained in CAF group (475.76 ± 23.93 g) compared with the CTL group (410.48 ± 10.78 g) (p < 0.05). This difference was maintained throughout the study ( Figure 1). However, although the CAF group indicated higher percentage of rWAT (3.29 ± 0.21 g/100 g BW) than the CTL group (2.87 ± 0.75 g/100 g BW), the difference was not statistically significant. Likewise, changes in the gastrocnemius muscle were observed between the CTL group (0.45 ± 0.03 g/100 g BW) and the CAF group (0.37 ± 0.02 g/100 g BW) without statistically significant differences (Table 2). dicated higher percentage of rWAT (3.29 ± 0.21 g/100 g BW) than the CTL group (2.87 ± 0.75 g/100 g BW), the difference was not statistically significant. Likewise, changes in the gastrocnemius muscle were observed between the CTL group (0.45 ± 0.03 g/100 g BW) and the CAF group (0.37 ± 0.02 g/100 g BW) without statistically significant differences ( Table 2).  0.14 ± 0.03 a 0.17 ± 0.02 a Gastrocnemius (%) 0.45 ± 0.03 a 0.37 ± 0.02 a Soleus (%) 0.10 ± 0.01 a 0.08 ± 0.01 a Results are expressed as mean ± standard error (n = 5 per group). Statistical analysis was performed using Student's t test for two independent samples. Mean values with different letters (a,b) show Figure 1. Effect of cafeteria diet on body weight during fifteen weeks in male Wistar rats. All the results are expressed as mean ± standard error (n = 5 per group). Statistical analysis was performed using Student's t test for two independent samples. Mean values with different letters (a,b) show significant differences (p < 0.05). CTL, control group; CAF, cafeteria group.

Shifts in Gut Microbiota after Eating Cafeteria Diet for 15 Weeks in Male Wistar Rats
The gut microbiota α-diversity values were evaluated using Shannon's diversity index. As shown in Figure 2, significant differences between the CTL group (6.51 ± 0.11) and the CAF group (5.77 ± 0.30) were found (p < 0.001). In fact, Figure 3 shows the relative abundance of 10 phyla found in both groups (>0.05%). The results indicate significant differences in the relative abundance of Bacteroidetes and TM7 phyla between groups (p < 0.05 and p < 0.001, respectively). Moreover, the relationship between Firmicutes and Bacteroidetes relative abundance, expressed as Firmicutes/Bacteroidetes ratio (F/B ratio), was statistically significant between the CTL group (0.80 ± 0.11%) and the CAF group (1.41 ± 0.22%) (p < 0.05).

Alterations in the Host-Microbiome Lipid Co-Metabolism
Although no correlations were found between the relative abundance of Firmicutes and fatty acid levels, the class Bacilli, order Lactobacillales, family Lactobacillaceae and genus Lactobacillus were positively correlated with PL-C20:3 n-6. In addition, the genus Eubacterium, from the phylum Firmicutes, also demonstrated a positive correlation with homo-γ-linoleic acid. Erysipelotrichi, Erysipelotrichales, Erysipelotrichaceae, a class, order, and family of the phylum Firmicutes, respectively, were positively correlated with CE-C18:1 n-9.
Anaerotruncus, a bacterial genus from the phylum Firmicutes, was positively correlated with CE-total MUFAs, CE-C18:1 n-9, CE-C18:2 n-6 and CE-total n-9, but inversely with CE-and PL-total n-7 and C18:1 n-7. Additionally, as expected, these MUFAs n-7 were negatively correlated with relative abundance of Bacilli, Lactobacillales, Lactobacillaceae and Lactobacillus. Figure 5 shows a heatmap of the Pearson's correlation coefficients between PL a CE fatty acids with several gut microbiota at different taxonomic levels. The Shannon dex, an indicator of bacterial diversity, was positively correlated with both PL-and C C18:1 n-7 and C18:3 n-3 fatty acids, as well as PL-C20:2 n-6, PL-total n-7, CE-C20:4 n and CE-total n-3. However, PL-total SFAs, CE-total MUFAs, CE-C18:1 n-9, CE-total n and CE-n-6/n-3 were inversely correlated with the Shannon index. Although no correlations were found between the relative abundance of Firmicu and fatty acid levels, the class Bacilli, order Lactobacillales, family Lactobacillaceae and gen Lactobacillus were positively correlated with PL-C20:3 n-6. In addition, the genus Eubac rium, from the phylum Firmicutes, also demonstrated a positive correlation with homo linoleic acid. Erysipelotrichi, Erysipelotrichales, Erysipelotrichaceae, a class, order, and fam of the phylum Firmicutes, respectively, were positively correlated with CE-C18:1 n-9.
Circulating CE C10:0 was associated with several genera of bacteria but had especially high correlations with Prevotella (positive) and Anaerotruncus (negative).

Discussion
In recent decades, the world has been involved in a more industrialized food environment, increasing the prevalence of obesity. Although obesity is a multifactorial disease, a factor associated with this pandemic is the high consumption of a Western diet [20]. The typical Western diet consists of a high-energy dietary pattern with large portions and a high content of fat, sugars, and sodium. Regarding fat content, it is noteworthy that the Western diet contains a high percentage of saturated fat and is characterized by a 20:1 n-6/n-3 ratio [10].
For the study of obesity, different animal models have been used, as well as different dietary patterns or hypercaloric diets such as high-fat, high-sugar and cafeteria diets [21,22]. However, a greater impact on the development of diet-induced obesity has been seen in models fed a cafeteria diet [5,21,23]. This may be because body weight is determined by a balance between energy intake and expenditure. In this sense, the cafeteria diet has been shown to be highly palatable, mimicking the Western diet, which contributes to weight gain and metabolic alterations associated with obesity [21]. In our study, similar to other authors [18,23], we found an increase in food efficiency in the CAF group compared to the CTL group. The foods used to make the cafeteria diet are ultra-processed and include hydrogenated fats and other ingredients such as artificial colors and flavors, making foods highly palatable, translating into a higher energy intake [2,24]. To understand the effect of the cafeteria diet on the development of obesity, we first characterized the fatty acid content of the food. As expected, the cafeteria diet is composed of a higher proportion of palmitic acid C16:0, stearic acid C18:0, oleic acid C18:1 n-9 and linoleic acid (LA) C18:2 n-6. Furthermore, it resulted in a 34:1 n-6/n-3 ratio. Palmitic acid is a long-chain saturated fatty acid (LCSFA) and the most abundant in animal products and other vegetable fats such as palm oil used in ultra-processed foods [25]. Palmitic acid has been associated with a pro-inflammatory response by activating Toll-like receptors-(TLRs) 2 and 4. Unlike lauric acid, a medium-chain saturated fatty acid (MCSFA), palmitic acid increased inflammation in male C57BL/6 mice fed a high-fat diet supplemented with 3% palmitic acid for 12 weeks [26]. However, although in our study circulating levels of palmitic acid were not found to be elevated in rats fed a cafeteria diet, other longer-chain or unsaturated fatty acids can be formed from it. In fact, palmitic acid can be converted to palmitoleic acid C16:1 n-7 by ∆-9 desaturation.
In our study, monounsaturated fatty acid C16:1 n-7 was found in both fractions of phospholipids (PL) and cholesterol esters (CE), without significant differences between groups. On the other hand, circulating PL-C20:3 n-6 was higher in the CAF group compared with the CTL group.
Furthermore, oleic acid is a monounsaturated fatty acid of the omega-9 series and widely present in vegetable oils such as safflower oil. The saturated form of this acid is stearic acid C18:0, which, in addition to being present in the Western diet, can also be synthesized from palmitic acid [29]. However, it has been seen that stearic acid is not absorbed in the intestine, so in our study no circulating levels in serum of stearic acid C18:0 were observed in Wistar rats fed a cafeteria diet. However, increased levels in serum of CE-C18:1 n-9 were found in the CAF group compared with the CTL group. On the other hand, although γ-linolenic acid (GLA) C18:3 n-6 increased in the CAF group, this was not significant. In a longitudinal study, biomarkers associated with metabolic disorders such as the HOMA index and insulin sensitivity were associated with the fatty acid profile. The authors reported that the higher circulating levels of C18:3 n-6 were correlated with lower insulin sensitivity. In contrast, circulating PL-C18:1 n-7 was positively associated with insulin sensitivity and pancreatic β-cell function [30].
Although in the cafeteria diet the content of capric acid C10:0 was not highlighted, the result of circulating C10:0 in rats fed cafeteria diet was striking, because although it is a MCSFA, it decreased in the CAF group compared with the CTL group. Capric acid is digested and absorbed in the stomach and is catalyzed by gastric lipase. In addition, this fatty acid increases energy expenditure, decreased adiposity and increases insulin sensitivity [12,31].
In fact, a study conducted by Li et al. demonstrated that C10:0 decreased weight gain and increased bile acid excretion in hypercholesterolemic C57BL/6J mice fed a diet high in medium-chain triglycerides [32].
Circulating fatty acids can come from the diet, as substrates for chemical reactions of metabolism, but they can also be derived from the product of bacteria that inhabit the host's intestine, thus contributing to the pathogenesis of obesity. Thus, alterations in the lipidome may be related to diet-induced changes in bacterial composition and diversity [33].
Different studies have evaluated the effect of a diet rich in long-chain fatty acids (LCFA) in gut microbiota and the metabolic effects associated with gut dysbiosis [31]. For example, in a study with C57BL/6J mice fed a diet rich in palmitic acid, the authors observed increased body weight and increased relative abundance of Firmicutes and decreased Bacteroidetes [34]. The upregulation of Firmicutes abundance has been associated with impaired intestinal permeability, obesity and associated metabolic disturbances [35]. The phylum Firmicutes includes bacterial species with genes encoding enzymes involved in lipid and carbohydrate metabolism, so an increase in Firmicutes/Bacteroidetes (F/B) ratio has been associated with body weight gain and the development of obesity [13]. In our study, the F/B ratio was inversely correlated with CE C10:0. Gual-Grau et al. evaluated the effect of different hypercaloric diets on bacterial abundance and diversity, finding that the cafeteria diet decreased α-diversity more than the other obesogenic diets, compared to the standard diet, although the results were not statistically significant [21]. The decrease in bacterial diversity has been associated with obesity; in fact, in our study, we found a lower α-diversity in the CAF group compared with the CTL group. In the association between α-diversity and fatty acids, we found a positive correlation with seven fatty acids and a negative correlation with six of the total number of fatty acids that presented a significant difference between groups. The strongest association was with PL-C18:1 n-7, a MUFA that decreased in the CAF group. Furthermore, this fatty acid was positively associated with the relative abundance of TM7, TM7_3, CW040 and F16. To date, little is known about the phylum TM7; however, a recent study found an association between TM7 and adiposity markers [36]. In our study we found the opposite, as the relative abundance of TM7 decreased significantly in the CAF group. Similar with our results, Hu et al. found a negative correlation between TM7 and body weight [37]. Therefore, more studies on the phylum TM7 are needed to understand its role in the obesity phenotype. What we can highlight in this study is that a relationship of TM7 with total n-7 was found in the two fractions PL and CE and was involved in certain metabolic processes such as less fat storage in the liver, as well as improved insulin sensitivity [38]. In addition, the total n-7 was negatively correlated with the genus Anaerotruncus, from the phylum Firmicutes.
Although there are few studies reporting the abundance of Anaerotruncus, Bortolin et al. observed an increase in this bacterial genus in male Wistar rats fed a cafeteria diet for 18 weeks. Interestingly, this genus has been associated with weight gain, metabolic diseases and intestinal permeability [39].
Other bacterial genera of the phylum Firmicutes that increased significantly in the CAF group were Eubacterium, Lactobacillus and Allobaculum. A study in 3-week-old BALB/c mice found an increase in Eubacterium abundance after two weeks of being fed a high-fat diet, but with 1:2 n-6/n-3 PUFAS, however, the authors did not report metabolic changes [40]. On the other hand, in a review on gut microbiota and obesity, the authors reported an increase in Eubacterium abundance in Japanese people with obesity [41]. Thus, the increased abundance of genus Eubacterium from the phylum Firmicutes could be associated with obesity, and in this study it was positively associated with PL-C20:3 n-6 and negatively with total n-7, as well as with C10:0. On the other hand, the genus Prevotella, from the phylum Bacteroidetes, has been associated with a diet rich in fiber [42]. Indeed, volunteers with risk of metabolic syndrome decreased Prevotella abundance after 4 weeks of consuming a highfat n-6 PUFA diet, whereas the MUFA diet induced an increase in the relative abundance of Prevotella, lowering BMI [43]. In this sense, it has been seen that overweight adults with an upregulation of Prevotella abundance lose more weight than subjects with a downstream Prevotella abundance after consuming a high-fiber diet. In our study, a positive correlation of Prevotella abundance was found with C10:0 and PL-C18:1 n-7, and negative association with CE C18:2 n-6.
In summary, the increased Firmicutes/Bacteroidetes ratio, including classes Bacilli and Erysipelotrichi, orders Lactobacillales and Erysipelotrichales, genera Anaerotruncus, Eubacterium and Lactobacillus (Phylum: Firmicutes), was positively associated with PL-C20:3 n-6 that was higher in the CAF group compared with the CTL group. On the other hand, relative abundance of the phylum Bacteroidetes, including class Bacteroidia, order Bacteroidales, family Prevotellaceae, and genus Prevotella significantly decreased in rats with diet-induced obesity. In addition, changes in bacterial diversity were found in male Wistar rats fed a cafeteria diet for 15 weeks, which in turn were positively associated with total circulating levels of n-3 and n-7 and negatively with n-6/n-3 ratio in the two PL and CE fractions.

Conclusions
In the present study, we demonstrated that CAF diet induced obesity in male Wistar rats. This obesity phenotype was accompanied by increased energy intake and food efficiency. Characterization of the cafeteria diet indicated that it is rich in palmitic acid C16:0, stearic acid C18:0, oleic acid C18:1 n-9 and linoleic acid C18:2 n-6. The n-6/n-3 ratio of the cafeteria diet was 34:1. This indicates that a diet rich in omega-6 caused gut dysbiosis by decreasing bacterial diversity and relative abundance of Bacteroidetes, Bacteroidia, Bacteroidales, and Prevotella, as well as TM7, TM7_3, CW040, and F16. In contrast, an increase in the relative abundance of Firmicutes, Bacilli, Lactobacillales, Lactobacillaceae and Lactobacillus was found. These changes were associated with the obesity metabotype, finding an increase in the circulating levels of PL-SFAs, PL-C20:3 n-6, and CE C18:1 n-9 and a decrease in PL-C18:1 n-7 and PL-C18:3 n-3.