Inhibition of Farnesoid X Receptor Rescues Fat Depression Induced by Dietary Berberine in Grass Carp, Ctenopharyngodon Idella

Berberine (BBR) depresses lipid accumulation in sh, but the mechanism remains unknown. In this study, we hypothesize FXR signaling participates in this physiological process of grass carp. Three diets, namely the control, BBR (1.0 g/kg), and BBR + Gly-β-MCA (an FXR inhibitor) were formulated to feed juvenile grass carp (9.90 ± 0.07) for 8 weeks. Fish fed BBR presented signicantly lower IPF index, hepatic TG and TC contents, as well as whole body lipid levels, whereas these were rescued by Gly-β-MCA. Serum TG and HDL-c contents were signicantly decreased in sh fed BBR compared to those in the control. The serum ALT activity, combined with the TG, TC, HDL-c, and LDL-c concentrations were all signicantly increased in sh fed BBR + Gly-β-MCA than those fed BBR. Dietary BBR signicantly increased the mRNA and protein expression of FXR, decreased the mRNA level of FGF19 in the intestine, whereas these were reversed by Gly-β-MCA. In the hepatopancreas, the inhibitor recovered the suppression of the CYP7A1, CYP8B1, and CYP27A1 expression induced by dietary BBR. Fish fed BBR showed signicantly lower mRNA expression of SREBP-1c and FAS, whereas these two genes were all up-regulated in response to inhibitor. Dietary BBR increased the gene expression of PPARα, ATGL, CPT-1, which were all abolished by dietary Gly-β-MCA treatment. Fish fed BBR and BBR + Gly-β-MCA showed signicantly lower total OTUs, ACE index, chao 1 index, and simpson index of the gut microbiota. Overall, our results demonstrate that inhibition of FXR leads to the rescue of lipid suppression induced by dietary BBR in grass carp.


Introduction
Aquaculture is the most promising strategy in solving increasingly demand for high quality protein in society (Reverter et al., 2020). In China, grass carp (Ctenopharyngodon idella) is the most cultured freshwater sh with nearly 20% in all the freshwater sh production (Fisheries Bureau of Ministry of Agriculture, 2019). However, there are some issues in front of the farmers and scientists that should be addressed, one of them is the excessively accumulated fat in the visceral, such as in the mesentery and liver, which is a big threaten to the health and market value of these cultured sh and sustainable development of aquaculture (Sun et al., 2021a;Tian et al., 2020).
Berberine (BBR, C 20 H 18 NO 4 ) is a natural plant alkaloid extracted from Berberis aristate and Coptis chinensis (Huanglian), an ancient anti-diarrhoeal medication. In mammals, BBR has antimicrobial activity against a variety of microorganisms (Zhou et al.), inhibits in ammatory process (Lin et al., 2019), and remits anti-oxidative stress (Moghaddam et al., 2014). Remarkably, BBR has been proved to reduce lipid accumulation by inhibiting lipogenesis and triggering fatty acid oxidative in mice (Kim et   . However, though marked molecular in related to the lipogenesis and lipid catabolism has been tested, the primary mechanism that induced these changes is still not understood. BBR is poorly absorbed into the systemic circulation but is signi cantly accumulated in the intestine (Pan et al., 2019), implying an interestingly potential mechanism that BBR may in uence the body's physiological changes through the intestine.
Actually, BBR has been found to impact the abundance and composition of gut microbiota both in mammals and sh (Pan et al., 2019;Zhang et al., 2020). Hence, it is supposed intestine is a target of BBR in depressing lipid accumulation in sh.
Farnesoid X receptor (FXR, NR1H4), a member of the nuclear receptor superfamily, is an intestinal molecular that has a vital role in regulating lipid metabolism of the whole body (Sun et al., 2021b). The natural ligands of FXR are bile acids (BAs) (Forman et al., 1995), which are reabsorbed by enterocytes and activate FXR that induce the expression of broblast growth factor (FGF) 19/15 hormone. After exiting the portal circulation, FGF19/15 binding to its liver FGFR4/β-Klotho co-receptor complex ultimately leads to the inhibition of the rate-limiting enzyme of BAs synthesis: the cholesterol 7 alpha-hydroxylase (CYP7A1) (De Magalhaes Filho et al., 2017). In addition to in uence the homeostasis of BAs, FXR also acts as an important target in modulating lipid accumulation by regulating lipid anabolism and lipid catabolism molecular (Jiao et al., 2015). Recently, we demonstrate that activating FXR decreased triglyceride (TG) and total cholesterol (TC) content in the liver, whereas inhibiting FXR increased the lipid accumulation in the liver and adipose tissue. Mechanically, inhibition of FXR would increase the expression of SREBP-1c and induce the TGs/cholesterol synthesis and lipid droplets formation (Tian et al., 2021). Interestingly, gut microbiota promote diet-induced obesity and associated phenotypes through FXR in mice (Parséus et al., 2017). Similarly, by modulating gut microbiota via antibiotic mixture, the FXR signaling of grass carp was attenuated and the lipid content in liver was increased (Tian et al., 2021).
These studies indicate FXR is an effective molecular in controlling fat deposition in the intestine and may be in uenced by the gut microbiota composition.
Collectively, considering the role BBR in depressing lipid content in sh and its characteristic of deposition in the intestinal and impact on the gut microbiota composition in grass carp. We hypothesize that FXR is a key molecular that mediates the function of BBR in controlling fat in the viscera. Therefore, in this study, we designed three semi-puri ed diets utilizing a FXR inhibitor glycine-β-muricholic acid (Gly-β-MCA), which was used to feed grass carp for 8 weeks, aiming to investigate the role of FXR in the inhibition of lipid accumulation in grass carp responding to dietary BBR.

Experimental diets
Three isonitrogenous and isoenergetic semi-puri ed diets with 36.0% crude protein and 6.0% crude lipid were designed and formulated based on the method previous described (Lovell and Tom, 1998). Table 1 shows the formation and composition of the experimental diets. Casein and gelatine were used as the protein sources, the sh oil and soybean oil were used as the oil sources. The rst diet that was not added extra chemicals was designed as the control. The second diet was added BBR (10g/kg; MedchemExpress LLC, Monmouth Junction, NJ, USA) at the expense of cellulose. The third diet was extra added the FXR inhibitor Gly-β-MCA (2.5mg/kg; MedchemExpress LLC) on the base of the second diet. The concentration of BBR and Gly-β-MCA were referred the study previously (Pan et al., 2019; Tian et al., 2021). We added 0.1% butylated hydroxytoluene (BHT; Sigma-Aldrich, USA) to the diets as the antioxidant ( Table 1). The ingredients were manually mixed at the sequence from less to more by the magnitude of the quantity, which were blended with 70% sterile pure water to make a dough, which was pressed into noodle-like pellets (2-mm diameter). The pellets were dried under forced air at 25°C for 24h, packaged and stored at -20°C until utilization. The proximate composition of diets was determined according to the Association of O cial Analytical Chemists (AOAC) procedures.

Experimental procedure
Experimental grass carp were purchased from Tongwei Aquaculture Co., Ltd. (Foshan, Guangdong, China), which were originated from the same parental stock. Fish were reared in aquaria (0.73×0.46×1.0m) and fed the commercial diet for 1 week, then the control diet for 2 week to acclimatize them to the experimental environment. The sh were not starved for 24h until the feeding experiment. A total of 180 healthy sh with uniform size (9.90 ± 0.07) were randomly distributed into nine aquaria (20 sh/aquarium). Each diet was randomly assigned into three aquaria. Fish were hand-fed to apparent satiation twice daily (at 8:30 and 16:30) for 8 weeks and the feed intake was recorded daily. During the feeding experiment, the water-dissolved O 2 was 5.6-6.4 mg/L, pH was 6.8-7.1, the temperature was 28.0-33.0°C. The photoperiod was 12h light-12h dark (from 8:00 to 20:00).
Sampling procedure After 8 weeks of feeding experiment, all of the sh were fasted for 24h. Afterwards, the sh were weighed and anaesthetized with MS-222 at the concentration of 0.06g/L to maximally reduce the pain of sh during sampling. Six sh per aquarium were randomly selected for blood collection from the caudal vein by using 1ml injection. The obtained blood was placed at 4°C (at least 6h) for clotting; the serum was collected after centrifugation (825 g, 4°C, 10 min) of the blood. The sampled serum was frozen in liquid N 2 and stored at -80°C for biochemistry analysis. The remaining nine sh were dissected, the viscera, hepatopancreas, intraperitoneal fat (IPF) were stripped and weighed. Samples of the hepatopancreas from three sh per aquarium were xed in 4% paraformaldehyde solution for histology analysis. Samples of hepatopancreas and mid-intestine from another six sh per aquarium were frozen in liquid N 2 then stored at -80°C for TG, TC, gene expression, and protein analysis. The intestine from three sh in each aquarium were aseptically removed, and the faecal matter were obtained by squeezing and scrapping the intestinal mucosa using a sterile spatula. The last ve sh per aquarium were stored at -80°C for whole body fat analysis. Speci c growth rate (SGR), feed conversion ratio (FCR), visceral index (VI), hepatopancreas index (HI), intraperitoneal fat index (IPFI), and survival rate (SR) were calculated using the following formulae: SGR = (Ln nal weight-Ln initial weight) × 100/days, FCR = amount of feed given/weight gain (g), VI (%) = visceral weight (g) × 100/body weight (g), HI (%) = hepatopancreas weight (g) × 100/ body weight (g), IPFI (%) = IPF weight × 100/body weight (g), SR (%) = nal number of sh×100/initial number of sh.

Serum biochemical parameters
Serum samples from of two sh were pooled and three two pooled samples for each aquarium were tested with the following methods. The serum biochemical parameters were assayed by enzymic procedures using an automatic analyzer (Mindray, Shenzhen, China). All of the serum biochemical parameters, including the alanine aminotransferase (ALT), aspartate aminotransferase (AST), TG, TC, high-density lipoprotein cholesterol (HDL-c), and low-density lipoprotein cholesterol (LDL-c) were tested using assay kits from Mindray.
For oil red O staining, xed hepatopancreases were incubated with 30% sucrose at 4°C for three days.
They were then embedded in an optimal cutting temperature compound (Leica), cut into 6-10 µm sections, and rinsed with distilled water. Slides were permeated with 60% isopropanol for 20-30 s and stained with oil red O (Sigma-Aldrich) for 10 min. Slides were immediately destained in 60% isopropanol for 3 min and washed with distilled water to clean the background. The sections were counterstained with Mayer's haematoxylin for 1 min, and washed with distilled water for 10 min. The slides were seal-capped with glycerogelatin and were photographed using a light microscope (Olympus BX41, Olympus Corporation).

Hepatic TG and TC content
The hepatopancreases from three sh per aquarium were used for the TG and TC content assays, which were determined using TG and TC assay kits, respectively (Jian Cheng Bioengineering Institute, Nanjing, China). The manufacturer's instructions were followed, and each sample was analysed in triplicate.

Whole body lipid content
The whole body fat content (3 individuals/aquarium) was determined according the Association of O cial Analytical Chemists (AOAC) procedures (1995) methods. Brie y, the whole sh was homogenized and its crude lipid was measured by ether extraction using the Soxhlet method after freeze drying to a constant weight at -40°C.

Quantitative real-time PCR
The total RNA of the sampled hepatopancreases or intestines was extracted based on the method described above. After removing the DNA in the total RNA, cDNA was synthesised using a PrimeScript ® RT reagent kit (TaKaRa) according to the manufacturer's protocol. The primers are listed in Table S1.
Based on a series of 6-step 10-fold dilutions of the target template, the ampli cation e ciency of these primers ranged from 92.46-103.68% after qRT-PCR assay. The determination coe cient (R 2 ) values were > 0.99. The β-actin was used as the reference gene based on preliminary tests using geNorm (version 3.5) and NormFinder algorithms. Real-time quantitative reverse transcription PCR (qRT-PCR) was performed three times (LightCycler 96, Roche Diagnostics, Basel, Switzerland) in a nal volume of 20 µL containing 0.6 µL of each primer (0.5 µM), 1 µL diluted rst-strand cDNA product, 10 µL 2 × SYBR Premix Ex Taq II (TaKaRa), and 7.8 µL sterilised double-distilled water. The cycling parameters were: 95°C for 5 min, followed by 40 cycles of 95°C for 15 s, and 60°C for 1 min. After the PCR reaction, the melting curve was analysed over a range of 60-95°C (in 5 s steps) to con rm a single product. Negative controls, including a no-cDNA control and a DNasetreated non-reverse transcribed tissue RNA sample, were used to ensure that only the cDNA was quanti ed in each sample. A relative quanti cation method was used to calculate the gene expression values using the comparative CT method (2 −ΔΔCt ) previously described in the literature (Livak and Schmittgen, 2001;Pfa , 2001).

Western-blotting
The intestine and hepatopancreas were homogenized with glass tenbroeck tissue grinders on ice (3 individuals per aquarium). The cell lysis buffer supplemented with protease and phosphatase inhibitor cocktails (Beyotime, Nanjing, China) was added before homogenization. Afterwards, the crude lysates were centrifuged for 10 min (4°C; 13,000g), the upper mixture were collected for further analysis. The total protein concentration was determined by a bicinchoninic acid protein assay kit (Thermo Fisher Scienti c Inc., United States). The protein samples were separated by SDS-PAGE and transferred to a polyvinylidene di uoride membrane (Beyotime, Nanjing, China) by electroblotting. The membranes were incubated with the primary antibodies overnight at 4°C. After washing, the secondary antibody was added and incubated for 2h at room temperature. The protein bands were visualized by ECL Plus (Beyotime, Nanjing, China). The membranes were then stripped and reprobed with anti-β-actin antibody. Densitometric quantitation was performed using a Sagecreation imaging system with Sagecreation Quantity One software (Sagecreation Co., Ltd.). The following antibodies were used: anti-bodies against anti-FXR (56kDa; Rabbit; Abcam, Cambridge, MA); anti-bodies against anti-CYP7A1 (57kDa; Rabbit; Invitrogen) Gut microbiota diversity analysis The total gut microbial DNA was extracted using the NucleoSpin Soil kit (Macherey-Nagel, Dueren, Germany) according to the manufacturer's instructions. DNA integrity and quality were monitored on 1% agarose gels. The quantity of the DNA was measured using an Implen NanoPhotometer (Implen, Inc.).
The V3 + V4 regions of the 16S rDNA were then ampli ed by the broad fusion primers of 338F (5'-ACTCCTACGGGAGGCAGCA-3') and 806R (5'-GGACTACHVGGGTWTCTAAT-3'). PCRs were performed using 10 µL of reaction volume: 5 µL of KOD FX Neo Buffer, 2µL of dNTP, 50 ng of DNA, 0.3 µL of each primer (10 µM), 0.2 µL of KOD FX Neo, and ddH 2 O up to 10 µL. The ampli cation program had an initial denaturation phase at 95°C for 3 min followed by 25 cycles of denaturation at 95°C for 30 s, an annealing stage at 50°C for 30 s, and an extension stage at 72°C for 40 s, followed by an extension at 72°C for 7 min. The ampli cations were extracted from 1.8% agarose gels and puri ed using Agencourt® AMPure XP Beads (A63881, Beckman Coulter Life Sciences, Indianapolis, IN, USA). Then, the gene quality was assessed using a Qubit® 2.0 Fluorometer (Q32866, Invitrogen, Carlsbad, CA, USA) and an Agilent 2100 Bioanalyzer (G2939AA, Agilent Technologies, Santa Clara, CA, USA). Finally, the quali ed ampli cations were run on an Illumina HiSeq 2500 platform from Biomarker Technologies Co., Ltd.
The obtained raw data were then jointed using FLASH v1.2.7 software to produce raw tags with an overlap length > 10 bp and a maximum mismatch rate < 0.2. High-quality tags were obtained by ltering raw tags using the Trimmomatic v 0.33 software with average Phred scores < 20. Chimeric sequences were discarded with the use of UCHIME v 4.2 software. The tags were clustered into one operational taxonomic unit (OTU) at a 97% identify threshold. The OTUs were classi ed and annotated using the QIIME software based on the Silva database (http://www.arb-silva.de). The alpha diversity indexes, including the number of OTUs, abundance-based coverage estimator (ACE), and Chao1, and the Shannon and Simpson diversity indexes were calculated.

Statistical analyses
The data are expressed as the mean ± standard deviation (S.D.). One way analysis of variance was used to compare differences, followed by Tukey's post hoc test. Percentage data were arcsine-transformed prior to analysis. All analyses were conducted using PASW Statistics 18 (SPSS, Chicago, IL, USA). A signi cance level of P < 0.05 was used for all tests.

Results
Growth performance, feed utilization, and biological parameters During the feeding experiment, all sh accepted the diets well and no dead sh was found ( Table 2). As shown in table 2, grass carp fed with BBR+Gly-β-MCA showed signi cantly higher nal weight and SGR than sh fed the control and BBR diets (P<0.05). The FCR of the BBR+ Gly-β-MCA group was signi cantly lower than that of the BBR group (P<0.05). Fish fed with BBR showed lower VI, HI, and IPFI than the other two groups, of which VI was signi cant difference than sh fed Gly-β-MCA, and IPFI was signi cant difference than the other two groups (P<0.05). Table 3 shows the serum biochemical parameters in sh fed different diets. The serum from sh fed BBR showed signi cantly lower TG, and HDL-c contents than those from sh fed the control diet (P<0.05). However, when the grass carp was fed the diet supplemented with Gly-β-MCA, the ALT activity, as well as the TG, TC, HDL-c, and LDL-c concentrations were all signi cantly increased in comparison to the sh fed the BBR diet (P<0.05).

Hepatic lipid content
The hepatopancreas was stained with H&E and oil red O to directly observe the lipid accumulation of the sh (Fig. 1A). Histological section showed that the hepatopancreas of sh fed BBR contained distinctly less lipid droplets than those of the control sh, whereas there were obviously larger lipid droplets in the hepatopancreas of sh fed the BBR+Gly-β-MCA than in those of sh fed the control and BBR feed. Quantitatively, hepatic TG and TC contents were signi cantly lower in the BBR group than those in the control and BBR+Gly-β-MCA groups (P<0.05, Fig. 1B).

Whole body lipid content
The whole body of sh fed different diets were analyzed the crude fat contents (Fig. 2). It is found that sh fed BBR had signi cantly lower lipid content than the sh fed the control diet, whereas the whole body lipid content was signi cantly recovered after administration Gly-β-MCA in the BBR diet (P<0.05).

FXR signaling gene and protein expression
The mRNA and protein expression of the molecular related to FXR signaling are shown in Fig. 3. Fish fed BBR showed signi cantly higher expression level of FXR than the control in the intestine (P<0.05, Fig. 3A), whereas the relative expression of the mRNA for FXR was signi cantly decreased in the sh fed BBR+Glyβ-MCA than that in the BBR group (P<0.05, Fig. 3A). The target gene of FXR, FGF19, was signi cantly down-regulated in the intestine of the BBR fed sh, whereas this was recovered to a signi cant level by Gly-β-MCA administration (P<0.05, Fig. 3A). Similarly, the protein expression of FXR was signi cantly increased in the intestine of sh fed BBR, whereas was suppressed by Gly-β-MCA (P<0.05, Fig. 3B).

Lipid metabolism related gene expression
The mRNA expression of genes related to lipid anabolism in the hepatopancreas are shown in Fig. 4. There was no signi cant difference of the mRNA expression of peroxisome proliferator activated receptor γ (PPARγ) between the control and BBR groups (P>0.05). However, sh fed with BBR diet showed signi cantly decreased mRNA expression of sterol-regulatory element binding proteins-1c (SREBP-1c) and fatty acid synthase (FAS) (P<0.05). When the diet was supplemented with BBR+Gly-β-MCA, the mRNA expression of PPARγ, SREBP-1c and FAS were all signi cantly increased compared to the BBR group (P<0.05).
Several genes related to lipid catabolism in the hepatopancreas are shown in Fig. 5. Signi cant difference of the relative mRNA expression of peroxisome proliferator activated receptor α (PPARα), adipose triglyceride lipase (ATGL), and carnitine palmitoyltransferase 1(CPT-1) was found between the control and BBR groups (P<0.05), they were all up-regulated in the sh consuming BBR. Moreover, these three genes were down-regulated to a signi cant level in the sh fed inhibitor Gly-β-MCA diet than sh fed the BBR diet (P<0.05).

Gut microbiota composition
The alpha-diversity of the gut microbiota in sh feeding different diets are shown in table 4. Fish fed with BBR and BBR+Gly-β-MCA showed signi cantly lower total OTUs, ACE index, chao1 index, and simpson index than sh fed the control diet (P<0.05, Table 4). However, no signi cant difference of the shannon index was found amongst the treatments.
At the phylum level, the most abundant phyla were Proteobacteria, Fusobacteria, Bacteroidetes, and Firmicutes (Fig. 6A, Fig. S1). No signi cant difference of the relative abundance of these phyla was found among groups. However, numerically, the abundance of Proteobacteria was higher in BBR group (49.67%) than in control (35.75%), whereas this was suppressed by dietary Gly-β-MCA to 22.57%. The abundance of Firmicutes was decreased from 5.58% to 1.24% after feeding BBR, but was increased to 10.87% after feeding inhibitor. At the genus level, Cetobacterium, Rhodobacter, Phreatobacter were the most abundance genus (Fig. 6B, Fig. S1). Phreatobacter was the signi cant abundance genus, which was increased from 7.25% to 14.90% after feeding with BBR, whereas signi cantly decreased to 2.38% in response to BBR+Gly-β-MCA diet (P<0.05). No signi cant difference of the other genus of the gut microbiota was found among the groups (P>0.05).

Discussion
BBR has been used in traditional Chinese medicine and Ayurvedic medicine and current research evidences support its use for various therapeutic areas (Singh and Mahajan, 2013). One of its remarkable function is reduction of lipid accumulation, this is conservative from sh to mammals (Hao et , 2017). Thus, the mechanism of dietary BBR on the lipid metabolism still remain elusive. In the present study, we explored the role of FXR, a modulator of lipid metabolism molecular in the intestine, in this process of grass carp through pharmacological methods. We show that dietary BBR activated the FXR signaling pathway, and inhibition of FXR abolished the fat suppression function, as well as some lipogenic genes induced by dietary BBR in grass carp.
During the feeding experiment, all sh accepted the diets well, and no dead sh was recorded, suggesting that dietary BBR and FXR inhibitor had no negative impact on the experimental sh. Dietary BBR had no impact on the growth of grass carp, which is consistent with previous studies in grass carp (Pan et al., 2019) and black sea bream (Wang et al., 2021). The decreased VI, HI, and IPFI in sh fed BBR are in line with the lipid content of the whole sh, re ecting BBR decreased lipid accumulation in a macroscopic view. On the contrary, dietary BBR + Gly-β-MCA increased the growth, VI, HI, and IPFI in sh compared to the BBR group, which can be explained by the fat accumulation in the vicera. However, the compensatory growth of the sh (such as intestine, hepatopancreas, IPF etc.) in respones to Gly-β-MCA to meet the FXR signalling requirements in tissues cannot be excluded.
The activities of serum ALT and AST are generally important indicators of the liver function, they also re ect the health of other tissues, such as spleen, kidney, etc. (Tian et al., 2014). Our study showed that the activity of serum ALT was decreased in the BBR feeding sh, whereas increased in the BBR + Gly-β-MCA feeding sh. The serum AST also showed similar trends. The changes of these two enyzmes are consistent with the VI, HI, and IPFI, suggesting that dietary BBR may be bene cial for the function of tissues in the viscera of grass carp. However, administration with FXR inhibitor may do harm to the health of the cells in the sh. This is possibly due to the lipid droplets accumualtion in cells that disordered the cellular homeostasis (Hyun et al., 2005). . We further proved that administration with Gly-β-MCA rescued the suppresion of these parameters, suggesting that inhibition of FXR could recover the serum fat homeostasis induced by dietary BBR.
By staining with H&E and oil red O, the lipid droplets in the hepatocytes were decreased the accumulation by the dietary BBR (marked in vacuole and red particles, respectively). Because lipid droplets are composed by TG and sterol esters (Lundquist and Susanto, 2020), our quantitative data also showed signi cantly decreased TG and TC content in BBR feeding sh, suggesting the component of the lipid droplets were also reduced. Our results are in line with previous studies in grass carp (Yang et al., 2019a), black sea bream , and blunt snout bream (Zhou et al., 2019). As a constrast, BBR + Glyβ-MCA increased the lipid content in the heaptopancreas, suggesting that FXR signalling play a negative role in fat deposition. Our previous study showed that solely treatment with Gly-β-MCA also increased the lipid content in grass carp (Tian et al., 2021), combined with this study, we could conclude that inhibition of FXR could eliminate the fat depression function of dietary BBR. However, it is still not con rmed BBR decreased the lipid accumulation via FXR signalling. We then explored the molecular events in the FXR signalling. In the intestine, activated FXR stimulates the expression of FGF19 hormone, which binds to the liver FGFR4/β-Klotho co-receptor complex to inhibit the rate-limiting enzyme of the bile acid synthesis CYP7A1 after transportation in the portal circulation (Zheng et al., 2017). Intriguingly, dietary BBR increased the mRNA (and protein) expression of FXR and decreased its downstream gene FGF19 in the intestine, as well as decreased the gene (or protein) expression of CYP7A1, CYP8B1, and CYP27A1 in the hepatopancreas, suggesting that BBR did activate the signalling of FXR. Expectedly, the decreased molecular exrpression of FXR combined with the up-regualtion of the downstream of the FXR genes (or protein) by Gly-β-MCA treatment implied that pharmacological treatment succeeded in abolishment in the FXR signaling induced by BBR. These data also indirectly indicate that FXR signalling participated in the fat suppression of dietary BBR in grass carp.
Previous studies have shown that BBR suppresses lipid accumulaiton via inhibiting lipogenesis and promoting lipid oxidative in sh (He et al., 2021;Wang et al., 2021). In the present study, we show that dietary BBR decreased the relative mRNA expreesion of lipogenic genes, SREBP-1c and FAS. Moreover, we also did nd the difference of lipid catabolism genes, PPARα, ATGL and CPT-1. These results are consistent with early studies. Importantly, our results provide evidences that inhibition of FXR abolished the down-regulation of lipid anabolism genes, as well as the up-regualtion of lipid catabolism genes induced by dietary BBR, suggesting FXR signalling plays an important role in the modulation of lipid metabolism related transcripts induced by dietary BBR, similar to our early study (Tian et al., 2021).
The results that dietary BBR decreased the total OTUs, ACE index, and Chao 1 index suggest that BBR decreased the richness of gut microbiota in grass carp. This might be due to the antibacterial property of BBR . Interestingly, dietary BBR also decreased the simpson index, which implied increased diversity of the gut microbiota in BBR treated sh, in line with early studies (Pan et al., 2019). The total decreased richness but increasd diversity may be due to increased community evenness for the compostion of gut microbita. Furthermore, we also showed that dietary Gly-β-MCA had no obvious differnce of the alpha-diversity of the gut microbiota, indicating that the drug used in this study did not in uence the gut microbiota composition. In mammals, the richness and diveristy of gut microbiota are regarded to be a key factor in impacting the body fat accumulation and metabolic disease (Clemente et al., 2012;Turnbaugh et al., 2006). Obesity people have more diversity of the gut microbiota than the thin ones (Turnbaugh et al., 2009). Similarly, the changes of the gut microbiota in sh are also linked to the changes in fat content (Sheng et al., 2018;Tian et al., 2021). It is generally accepted that gut microbiota decogugate and further metabolise primary BAs into secondary BAs through bile salt hydrolase (BSH) (Wahlström et al., 2017). BAs differ widely in their ability to activate FXR (de Boer et al., 2020), the changes in the BAs composition altered by the may altered the signalling of FXR. From this view, BBR may in uence the FXR via modulating gut microbiota, which altered the composition of BAs and activated FXR signalling, further decreased the lipid accumulation in grass carp.
It is suggested that obesity mice or hunman have higher content of Firmicutes and lower Bacteroidetes (Kallus and Brandt, 2012). This is possibly due to the differences of these two bacteria in the contents for the enzymes related to lipid/carbonhydrate metabolism (Stephens et al., 2018). In our study, though have no obvious difference of the gut microbiota composition in the phyla level. The concentration of Firmicutes had a decreasd trend in the BBR treatment, whereas had an increased trend in the BBR + Gly-β-MCA treatment, which is in line with the fat content of grass carp. In addition, several other bacteria have similar trends with lipid content, such as acidobacteria and gemmatimonadetes, as well as the genus Phreatobacter, but the relationship between these bacteria and lipid metabolism are scarcely studied.
In conclusion, in this study, we explored the mechanism of dietary BBR on the lipid metabolism by using the FXR inhibitor Gly-β-MCA in grass carp. Dietary BBR modulated gut microbiota composition, activated FXR signalling, and suppressed lipid accumulation. Meanwhile, inhibiton FXR could recover the suppression the fat deposition induced by BBR. It is speculated that FXR signalling play important role in the function of BBR in modulating lipid accumulaiton in grass carp. More studies in related to the gut microbita and BAs composition that altered by BBR are needed to be further addressed. Con icts of interest: The authors declared that they have no con icts of interest to this work.
Ethics approval: The protocol was approved by the Institutional Animal Care and Use Ethics Committee of the Chinese Academy of Fishery Sciences. 44. Zhou, X., Yang, C., Li, Y., Liu, X., Wang, Y., Potential of berberine to enhance antimicrobial activity of commonly used antibiotics for dairy cow mastitis caused by multiple drug-resistant Staphylococcus epidermidis infection.