Novel high‐docosahexaenoic‐acid tuna oil supplementation modulates gut microbiota and alleviates obesity in high‐fat diet mice

Abstract Studies have documented the benefits of fish oil in different diseases because of its high n‐3 polyunsaturated fatty acid content. However, these studies mostly used commercially available fish oil supplements with a ratio of 18/12 for eicosapentaenoic acid and docosahexaenoic acid (DHA). However, increasing DHA content for this commonly used ratio might bring out a varied metabolic effect, which have remained unclear. Thus, in this study, a novel tuna oil (TO) was applied to investigate the effect of high‐DHA content on the alteration of the gut microbiota and obesity in high‐fat diet mice. The results suggest that high‐DHA TO (HDTO) supplementation notably ameliorates obesity and related lipid parameters and restores the expression of lipid metabolism‐related genes. The benefits of TOs were derived from their modification of the gut microbiota composition and structure in mice. A high‐fat diet triggered an increased Firmicutes/Bacteroidetes ratio that was remarkably restored by TOs. The number of obesity‐promoting bacteria—Desulfovibrio, Paraeggerthella, Terrisporobacter, Millionella, Lachnoclostridium, Anaerobacterium, and Ruminiclostridium—was dramatically reduced. Desulfovibrio desulfuricans, Alistipes putredinis, and Millionella massiliensis, three dysbiosis‐related species, were especially regulated by HDTO. Regarding the prevention of obesity, HDTO outperforms the normal TO. Intriguingly, HDTO feeding to HFD‐fed mice might alter the arginine and proline metabolism of intestinal microbiota.

chronic inflammation, or impaired satiety signaling, is linked to various chronic diseases, such as cardiovascular diseases, type 2 diabetes, hypertension, and several forms of cancer (Fabbrini et al., 2010;Kahn, 2008;Lavanya & Rana, 2019;Lee et al., 2013). Then, what are the factors resulting in obesity? Host-microbial interactions have been documented in studies of obesity-related diseases (Cani et al., 2012). In addition to genetic susceptibility, environmental impact, lack of physical activity and other factors, the greatest risk is the expansion of high-fat/high-sugar diets (Crinò et al., 2018;Stoner et al., 2016;Zhang & Yang, 2016). However, the gut microbiota is a key interface for energy acquisition of the host (Ikuo et al., 2013;Mayu et al., 2015). Accumulating evidence has indicated that highfat-diet-triggered obesity leads to alterations in the gut microbial composition, as well as reductions in microbial diversity and changes in specific bacterial taxa (Daniel et al., 2014;Turnbaugh et al., 2006Turnbaugh et al., , 2009Zhu et al., 2018). These variations in the gut microbiota might result in gut microbial dysbiosis and play an important role in the pathogenesis of obesity. Thus, developing a new strategy to treat obesity surrounding manipulations of the gut microbiota and its metabolites has become a primary public health goal.
However, most of the currently available studies investigating the effect of dietary n-3 PUFAs employed commercially available fish oil supplements, which are characterized by higher amounts of EPA than DHA, in a distinctive ratio of 18/12 (Fard et al., 2019;Turchini et al., 2009). Nonetheless, the molecular structures of DHA and EPA are different. DHA contains a longer carbon chain (22 vs. 20) and an additional double bond (6 vs. 5) per molecule compared with EPA, which might result in the different metabolic effects between the two molecules. Moreover, increasing evidence has shown that EPA and DHA exert heterogeneous effects on human health; dietary DHA or EPA have different metabolic fates in animal models; for example, DHA is preferentially retained over EPA, and EPA is β-oxidized more than DHA (Ghasemifard et al., 2015;Serhan, 2005).
Mammalian brains were also shown to be invariably rich in DHA (Crawford et al., 2009). Thus, it is necessary to obtain a better understanding of the potential metabolic and health effects of DHA, uncoupled by a higher EPA content. Tuna, one of the most important sources of EPA and DHA, is characterized by higher DHA and less EPA in its oil (e.g., EPA:DHA = 4.2:19.8 [Castellano et al., 2011], 3:13 [Ninio et al., 2005]). In this study, not only general tuna oil (60 mg of EPA and 260 mg of DHA per gram of oil) but also its fractionated and concentrated product with a higher DHA content (60 mg of EPA and 340 mg of DHA per gram of oil) was employed to investigate the effect of high-DHA tuna oil on the alteration of the gut microbiota and obesity in high-fat diet mice. We aimed to provide evidence for the clinical therapeutic potential of the dietary administration of high-DHA fish oil preparations.

| Preparation of high-DHA tuna oil
High-DHA tuna oil was prepared according to the method proposed in our previous report . The composition of fatty acids was detected by gas chromatography-mass spectrometry (GC-MS; Agilent 7890/M7-80EI system with a VOCOL column [60 m × 0.32 mm]). The initial temperature of the oven program was set at 60°C, increased to 260°C at a rate of 5°C/min, and then was maintained at 260°C for 40 min. The gas flow rate was 50 ml/ min. The temperature of the injector was maintained at 260°C.
The detected mass ranged from 30 to 425 m/z (Lu et al., 2017;Satil et al., 2003). The content ratios of EPA and DHA in the general tuna oils and its fractionated and concentrated product were 60 mg of EPA and 260 mg of DHA per gram of oil, 60 mg of EPA and 340 mg of DHA per gram of oil, respectively.

| Animals and experimental design
All experimental procedures and animal care were performed ac- Co., Ltd) and were randomly divided into five groups (six mice per group). Next, each group was housed in two separate cages (three mice per cage) under controlled conditions (23 ± 1°C; 12:12 hr light/ dark cycle; 60 ± 5% relative humidity) with free access to food and water for 6 weeks. The five groups were as follows: (a) control group (Control), normal chow feeding (protein with 20% kcal, carbohydrate with 70% kcal, fat with 10% kcal; purchased from Laboratory Animal Center of Ningbo University, Ningbo, China) and received 200 ml of saline per day by gavage; (b) high-fat diet group (HFD), high-fat diet feeding (protein with 20% kcal, carbohydrate with 35% kcal, fat with 45% kcal purchased from Laboratory Animal Center of Ningbo University); (c) positive-control drug group (Zocord), high-fat diet feeding and received 3 mg kg −1 day −1 of Zocord by gavage; (d) high-DHA tuna oil group (HDTO), high-fat diet feeding and received 260 mg kg −1 day −1 of high-DHA tuna oil by gavage; (e) conventional tuna oil group (TO), high-fat diet feeding and received 260 mg kg −1 day −1 of conventional tuna oil by gavage.
The body weight of the mice was monitored every 5 days. At the end of 6 weeks, stool samples were collected, immediately immersed in liquid nitrogen and stored at −80°C for later microbiota analysis. After 12 hr of food deprivation, the fasting body weight was examined and all the mice were anesthetized with ether. Blood was collected from the orbital plexus, and the serum was further isolated by centrifugation at 1500 g at 4°C for 15 min and then stored at −80°C for subsequent biochemical testing. All the animals were sacrificed. Visceral tissues, including adipose (epididymal, subcutaneous, visceral, and interscapular) and liver tissues were dissected, weighed, instantly immersed in liquid nitrogen, and then stored at −80°C for further analysis.

| Real-time qPCR for the expression of genes related to obesity
The livers were homogenized in liquid nitrogen. RNA extraction was performed according to the RNA extraction kit instructions. The RNA was reverse transcribed into cDNA using a high-capacity cDNA reverse transcription kit (Applied Biosystems, Life Technologies).

qRT-PCR was performed using SYBR Green Master Mix and Quant
Studio 6 Flex (Thermo Fisher Scientific Inc.) according to the manufacturer's instructions. The quantification of RNA samples was carried out using the Nanodrop 2000C system (Thermo Fisher Scientific Inc.). All the samples were run in duplicate on a single plate, and relative quantification was detected using the 2 −ΔΔCt method.
The qRT-PCR primers used in this study are presented in Table 1.
β-Actin was used as an internal control.

| DNA extraction, 16S rRNA sequencing, and bioinformatic analysis
DNA from different samples was extracted using the E.Z.N.A.
®Stool DNA Kit (D4015; Omega, Inc.) according to the manufacturer's instructions, and the quantification of genomic DNA was executed using the Thermo NanoDrop 2000C system. The V3-V4 region of the 16S rRNA gene was amplified using primers 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5′-GGACTACHVG GGTWTCTAAT-3′). The 5′ ends of the primers were tagged with specific barcodes per sample and universal sequencing primers. The cycling and reaction conditions were 98°C for 30 s followed by 35 cycles of denaturation at 98°C for 10 s, annealing at 54°C/52°C for 30 s, and extension at 72°C for 45 s and then a final extension at 72°C for 10 min. The PCR products were confirmed by 2% agarose gel electrophoresis, and they were purified using AMPure XT beads (Beckman Coulter Genomics) and quantified by Qubit (Invitrogen).
Samples were sequenced using the Illumina MiSeq platform according to the manufacturer's recommendations provided by LC-Bio. Quality filtering on the raw tags was performed under specific filtering conditions to obtain high-quality clean tags according to Representative sequences were chosen for each OTU, and taxonomic data were then assigned to each representative sequence using the RDP (Ribosomal Database Project) classifier. Alpha diversity and beta diversity analysis were performed using QIIME (Version 1.8.0). Linear discriminant analysis (LDA) scores derived from the LDA effect size (LEfSe, https://hutte nhower.sph.harva rd.edu/galax y/root?tool_id=lefse_upload) was executed to identify the specific bacteria (p < .05 and LDA score of >6.0; Segata et al., 2011). The correlations between the relative abundance of the key species and related metabolic indices and genes were conducted by Spearman's correlation in R software (version 3.6.2). The prediction of functional pathway variations in the gut microbiome at the OTU level was performed using the Tax4Fun R package (Aßhauer et al., 2015).

TA B L E 1 Primer sequences used in real-time PCR analysis
Different analyses were conducted in STAMP (Parks et al., 2014), and Welch's t test was used for the comparison of two groups.

| Statistical analysis
The data are expressed as the means ± SEM (standard error of the mean) and were analyzed using SPSS 23.0 statistics and OriginPro software. Differences between two groups were assessed using unpaired two-tailed Student's t tests. Repeated measures oneway analysis of variance (ANOVA) and Tukey's post hoc test (SPSS) were used. For the data whose distribution did not conform to the Gaussian model of heterogeneity, nonparametric Kruskal-Wallis analysis was conducted. Differences were considered statistically significant at p < .05.

| HDTO and TO improve the features of obesity in high-fat diet-fed mice
Four groups of mice were, respectively, provided with HFD and were administered 260 mg kg −1 day −1 of HDTO or TO, as well as Zocord, to identify the impact of HDTO and TO on the development of obesity. The positive-control drug Zocord served as a treatment reference. As illustrated in Figure 1a, the mice that consumed a high-fat diet gained more weight than the other groups from the 10th day, reaching a twofold increase at the final feeding trial compared with the control. In the Zocord group, the body weight was significantly decreased, but the body fat was higher than that in the HFD group ( Figure 1b). Notably, two types of tuna oil could attenuate the body gain and reduce fat build-up, and the suppressing effect was significantly better than that in Zocord treatment. However, the suppressing effect between HDTO and TO on the body weight and body fat rate showed no significant difference.
Compared with the control, 6-week high-fat diet feeding led to boosted levels of TG and LDL-C and decreased levels of HDL-C in the serum (Figure 2a-c). Similarly, the levels of TC and TG in the liver of the HFD groups were higher than those in the control group.
Furthermore, HDTO and TO supplementation significantly decreased the serum TG and increased the serum HDL-C levels in the mice fed a high-fat diet.
Additionally, the level of serum LDL-C was decreased due to HDTO and TO supplementation, and p < .05 compared with the HFD groups. The effect of Zocord supplementation was similar to that of the tuna oil. Notably, HDTO and TO also reduced the liver weights of the HFD-fed mice to restore to the control level ( Figure 2f).
Regarding the concentrations of liver TG and TC, HDTO and TO supplementation led to decreased liver TG and TC levels (Figure 2d/e).
Additionally, exceptions occurred for Zocord supplementation; the concentrations of liver TC were higher than those of mice fed a high-fat diet. These results indicate that HDTO and TO can improve blood and liver metabolic parameters in obese mice, and the effect of HDTO treatment is more obvious.

| HDTO and TO ameliorate the expression of lipid metabolism-related genes and IL-6 expression in high-fat diet-fed mice
The In this study, interleukin-6 (IL-6) expression levels were notably elevated due to HFD induction compared with those in chow-fed mice.
Next, the expression pattern of this cytokine was reduced by HDTO and TO supplementation. Additionally, the reversal effects of Zocord administration were also observed in all genes. Overall, HDTO more significantly reversed the change triggered by HFD in mice.

| HDTO and TO supplementation alters the structure of the gut microbiota in high-fat dietfed mice
To elucidate the effects of dietary high-DHA tuna oil on the gut microbiome in high-fat diet-induced obese mice, high-throughput sequencing of 16S rRNA based on V3-V4 hypervariable regions was used to analyze the variations in the gut microbial structure. The gut microbiota of the mice fed HFD + HDTO and HFD + TO was profiled and compared with that of the control group, HFD group and Zocord group. After double-end splicing, quality control, and chimera filtering, 243,397 high-quality sequencing reads were obtained from 15 fecal samples, and then the sequencing reads were clustered into OTUs at a 97% similar level.
Alpha diversity was applied in analyzing the complexity and diversity of species for samples using four parameters: two richness estimators (Chao1 and ACE) and two diversity indices (Shannon and Simpson). As suggested in Figure 4, HDTO and TO supplementation could prevent HFD-induced reduction in microbial richness and diversity to some extent, especially HDTO treatment, but not to a statistically significant level (p > .05, Figure 4a Absiella was also observed in the Zocord group.

F I G U R E 3 Effects of HDTO and
TO supplementation on the relative expression of ACC, FAS, CPT-1, HMGCR, SREBP-2, and IL-6 in the liver. The values are expressed as means ± SEM, and the different letters represent significant differences between different groups (p < .05)

| Pivotal phylotypes of gut microbiota corresponding to HDTO and TO supplementation and their correlation with obesityrelated metabolism parameters
The LDA effect size (LEfSe) method was employed to analyze the 16s rRNA sequencing data to discern the specific altered bacterial phenotypes and biomarkers of HFD and the dietary intervention groups ( Figure 6)  However, Mucispirillum schaedleri was negatively correlated with liver HDL-C, and Escherichia fergusonii ATCC 35469 was concurrently negatively correlated with serum MDA and TG, HDL-C and CPT-1 genes in the liver. The above correlation was significant at p < .05.

| HDTO and TO supplementation alters the metabolic pathways
Because dietary HDTO resulted in the variation of the gut microbiota structure in high-fat diet consumption mice, the functional profiles related to HDTO supplementation should be further predicted. This prediction was performed by Tax4Fun

| D ISCUSS I ON
Obesity triggers many diseases, such as heart disease, dyslipidemia, cancer, and 2 type diabetes. In recent years, accumulating studies in animals or humans showed that dietary supplementation with n-3 PUFAs is a potentially feasible nutritional strategy to prevent obesity.
Thus, many studies to date have exhibited positive effects of supplementation with fish oil containing EPA and DHA on obesity. However, the fish oils used in these studies were commercially available fish oil with a higher EPA content (Molinar-Toribio et al., 2015;Parker et al., 2019). Additional evidence has indicated the role of the ratios of DHA and EPA in the prevention and treatment of chronic disease in rat models Molinar-Toribio et al., 2015). Furthermore, a study (Cottin et al., 2011) explored the known differential effects of EPA and DHA in human subjects and concluded that there is an F I G U R E 5 Changes in the gut microbiota related to HDTO and TO supplementation. (a) Relative abundance of the bacterial profile at the phylum level. (b) Firmicutes/Bacteroidetes (F/B) ratio. (c) Heatmap of the relative abundance of the bacterial profile at the genus level.
(d) Mean proportions of the significantly different genera in mice between different groups (Welch's t test was performed to assess the difference between two groups). The data are shown as the means ± SEM, *p < .05 evident potency of DHA to improve several cardiovascular risk factors.
Thus, fish oils with a higher content of DHA than EPA might have different health benefits compared with the high-EPA fish oil traditionally used. In this study, to fill this knowledge gap and extend the work of the currently available literature, tuna oil with an EPA/DHA ratio of 6:26 and its fractionated and concentrated oil (EPA:DHA = 6:34) were employed to gain a better understanding of the potential effects on obesity mitigation and gut microbiota in HFD mice.
As presented in this study, HDTO and TO supplementation effectively attenuated the features of obesity in high-fat diet-fed mice.
The weight gain, liver weight, and body fat rate of the mice were significantly reduced accompanied by HDTO and TO consumption.
Deterioration induced by HFD in most of the levels of lipid metabolism parameters in the serum and liver-TC, TG, LDL-C, and HDL-Ccould be suppressed and even restored to the control level. Hence, regarding the variation in these parameters, HDTO and TO could improve the obesity feature in obese mice, and the overall treatment effect of HDTO was better than that of TO.
These improvements could be further demonstrated by the amelioration of lipid metabolism-related gene expression in high-fat diet-fed mice. Often, obesity is associated with lipid metabolism disorders. The liver is the center of fat metabolism, and, together with the gallbladder, it can achieve fat digestion. However, in the case of metabolic disorders, the above process cannot be carried out smoothly. Transfats F I G U R E 6 Key phylotypes of gut microbiota corresponding to HDTO and TO supplementation. LDA effect size (LEfSe) was performed to identify the different abundant taxa (LDA score was 6.0). Gut dysbiosis is a critical factor in the development of obesity and metabolic syndrome. The modulation of gut microbiota has become a promising pharmacological approach in the prevention of various chronic diseases. According to previous reports, an increased richness in the gut microbial diversity is negatively correlated with obesity and various disease states (Ji et al., 2019;Sánchez et al., 2017), and obesity triggers an increase in the relative abundance of Bacteroidetes and a decrease in the relative abundance of Firmicutes. Obesity can be marked by this increased proportion of F/B. However, growing evidence has indicated that the gut microbiota composition and structure can be reshaped by the interaction between dietary components and intestinal microorganisms . HDTO and TO supplementation increase the alpha diversity of gut microbiota in HFD mice; PCoA analyses revealed a significant separation of microbiota communities between the HFD group and the two groups treated with HDTO and TO. Additionally, HDTO supplementation resulted in a farther reduction in the F/B ratio. Bacteroidetes and Firmicutes are two main communities that affect energy metabolism homeostasis (Xu et al., 2017). A lower F/B ratio often reflects less energy extraction from the diet, and the interventions that prevented obesity in animals and humans are potent (Sasaki et al., 2013;Turnbaugh et al., 2006). Intestinal bacteria influence mammalian physiology and contribute to nutrient acquisition, inflammatory reactions, energy harvest and lipid metabolism, and they are closely related to obesity and metabolic diseases (Frazier et al., 2011;Kasubuchi et al., 2015). At the species level, as a potential diagnostic biomarker of dysbiosis, the increased relative abundance of Desulfovibrio desulfuricans was associated with metabolic disorders and related inflammation (Sun et al., 2015;Weglarz et al., 2003). The Desulfovibrio genus is known to include a sulfate-reducing bacterium that can metabolize sulfate to produce hydrogen sulfide.
The latter in the intestinal tract inhibits the metabolic pathway of intestinal epithelial cells using butyric acid, damages the intestinal epithelial mucosa, and induces chronic inflammation (Pitcher & Cummings, 1996).
According to our study, the relative abundance of this species was also positively correlated with FAS gene expression. Adipose FAS mRNA expression is significantly associated with obesity, predominantly visceral fat accumulation, impaired insulin sensitivity, and circulating adipokines (Berndt et al., 2007). However, the increased prevalence of this species by a high-fat diet was significantly attenuated by HDTO and TO administration, indicating the potential effects of HDTO and TO in the prevention of obesity and related metabolic disease. The other biomarker, Alistipes putredinis, belongs to the genus Alistipes and is positively related to the gene expression of FAS, HMGCR, SREBP-2, and IL-6 in the liver and serum LDL-C. As reported previously, Alistipes is a harmful microorganism, and its abundance is positively correlated with some obesity-related parameters, such as weight and serum TG and IL-6 gene expression (Kang et al., 2019;Xu et al., 2015). These results indicate that the anti-obesity regulation of HDTO and TO is not only associated with lipid absorption but also with inflammatory immunity. Compared with Alistipes putredinis and Desulfovibrio desulfuricans, knowledge about Millionella massiliensis is very sparse.

| CON CLUS ION
In summary, high-DHA tuna oil significantly ameliorated obesity and metabolic dysfunctions in mice fed a high-fat diet, and these anti-obesity effects might be mediated by gut microbiota. HDTO and TO markedly decreased the number of obesity-promoting bacteria, Desulfovibrio, Paraeggerthella, Terrisporobacter, Millionella, Lachnoclostridium, Anaerobacterium and Ruminiclostridium, and restored the increase in the F/B ratio and specific regulation of community structure. Particularly, Desulfovibrio desulfuricans, Alistipes putredinis, and Millionella massiliensis, as potential diagnostic biomarkers of dysbiosis, were dramatically increased in relative abundances by HDTO treatment. Furthermore, according to the prediction, the functional pathway of HDTO supplementation-regulated obesity might be arginine and proline metabolism in intestinal microbiota. Overall, the regulatory effect of HDTO occurred before that of TO. Tuna oil with a higher DHA content might be a promising therapeutic option in mitigating obesity. However, its quantitative usage and possible anti-obesity mechanisms need to be further clarified. Further understanding of the mechanisms that underlie microbial resilience toward external perturbations will be a crucial requirement for microbiome-directed precision treatment.