Effects of limonin and nomilin on lipid metabolic homeostasis in hyperlipidemic mice

Abstract

This study investigated the effects of different doses (5 mg mL⁻¹ and 10 mg mL⁻¹) of limonin and nomilin on the lipid metabolic homeostasis of high-fat diet-fed mice by oral gavage. The results showed that both limonin and nomilin interventions could significantly reduce the serum levels of TC, TG and LDL-C in high-fat diet-fed mice (p < 0.05), while greatly increased the level of HDL-C (p < 0.05), particularly with an increase of 71.95% after nomilin intervention at a high dose. In addition, the pathological damage of colonic adipose tissue improved after either limonin or nomilin intervention, and goblet cells increased by 17.19%–31.21% when compared with the HFD group. Limonin and nomilin interventions could improve the structure of the intestinal flora of hyperlipidemic mice, along with probiotics, e.g., Lachnospiraceae increased and harmful bacteria, e.g., Helicobacter and Desulfovibrio, abundance decreased significantly (p < 0.05). Metabolomic analysis revealed distinct metabolic remodeling effects, including the modulation of key metabolites involved in lipid metabolism and steroid hormone biosynthesis; notably, nomilin exhibited superior lipid-regulating efficacy compared to limonin, with specific upregulation of steroid hormone biosynthesis-related metabolites (e.g., estrone glucuronide and estrone 3-sulfate). This study could advance our understanding of the therapeutic potential of citrus-derived limonoids in restoring lipid metabolic homeostasis and provides a mechanistic basis for their application in functional foods or nutraceuticals.

---

1 Introduction

In recent years, the global obesity rate has continued to rise, and the number of obese people is increasing. Studies have shown that obesity may cause chronic diseases such as hyperlipidemia, cardiovascular disease¹ and diabetes.² Hyperlipidemia is a kind of disease caused by abnormal metabolism or transport of lipid components in the blood, which may cause a variety of serious complications, such as atherosclerosis. However, the present known lipid-lowering drugs have side effects and are expensive, which make them unsuitable for long-term prevention. The development of biological functional components as health products is one of the potential methods for the prevention and treatment of hyperlipidemia. For example, natural active components in plants such as polyphenols,³ polysaccharides,⁴ terpenoids, etc.,⁵ have show good lipid-lowering and hypoglycemic effects in vitro.

Limonoids belong to tetracyclic triterpenoids and have the skeleton structure of 4,4,8-trimethyl-17-furyl steroids. Limonoids mainly include nomilin, limonin, xanthone, etc. Nomilin is the main precursor for the synthesis of limonin.⁶ Although limonin and nomilin share a similar core structure, nomilin possesses a unique acetoxy group. Studies have shown that nomilin can act as a biological activator and directly influence metabolic processes by activating receptors and regulating gene expression through multiple pathways.⁷ Besides, relevant studies have found that limonin has a lipid-lowering effect and can significantly improve blood lipid levels and inhibit the accumulation of fat in the liver of nutritional obesity rats.⁸ Limonin can also reduce lipid accumulation by inhibiting the Akt-FOXO1-PPARγ axis in 3T3-L1 adipocytes, thereby inhibiting fat formation, clarifying the inhibitory effect of limonin on fat formation.⁹ Similarly, nomilin also has a variety of biological activities, such as anti-inflammatory, detoxification, anti-obesity and so on.^10–12 Studies have shown that treatment of mice with metabolic disorders with 2.0% nomilin for 77 days can reduce body weight, blood glucose, and serum insulin, and increase glucose tolerance.¹³ Moreover, nomilin can alleviate nonalcoholic steatohepatitis and liver fibrosis in mice by inhibiting the expression of the inflammatory factor IL-6 and improving anti-inflammatory ability.¹⁴ Fan et al. have found that nomilin can extend the lifespan by activating PXR-mediated detoxification function.¹¹ Overall, both limonin and nomilin belong to limonoid triterpenoids and show good activity in lowering blood lipids in vitro. However, the differences in lipid-lowering activity between limonin and nomilin remain unknown.

Therefore, in this study, a hyperlipidemia model was constructed in high-fat diet-induced mice, and the interventions of simvastatin, limonin, and nomilin on hyperlipidemic mice for six weeks were compared. High-throughput sequencing of the 16SrRNA gene and non-targeted metabolomics technology were used to explore the effects of different doses of limonin and nomilin on intestinal microbiota and fecal metabolite changes in HFD mice. This study compared the lipid-lowering activities of limonin and nomilin to provide a theoretical basis for the development of lipid-lowering products based on limonin and nomilin.

2 Materials and methods

2.1 Materials

Citrus-derived limonin and nomilin with purity ≥98% were supplied by Shanghai Yuanye Biotechnology Co., Ltd. Simvastatin Tablets (ST) were provided by Shanghai Diran Dancheng Pharmaceutical Co., Ltd (Shanghai, China). The total cholesterol (TC), triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) kits were purchased from the Nanjing Jiancheng Institute of Bioengineering (Nanjing, China).

2.2 Animals and treatments

SPF-grade C57BL/6 mice (male, 6–8 weeks of age, and weighing 20 ± 2 g) were provided by Beijing Huafu Kang Biotechnology Co., Ltd. The animals were raised in the animal room of the 900th Hospital of the Joint Support Force of the People's Liberation Army of China (Fuzhou, China, license no.: SYXK (Military) 2025-0016). The ethical review was passed by the Experimental Animal Welfare Ethics Committee of the 900th Hospital of the Joint Support Force of the People's Liberation Army of China, with the review number 2025-04. The temperature of the raising environment was 22 ± 2 °C, and the relative humidity was 50% to 80%.

After 56 C57BL/6 mice were adaptively fed for one week, they were randomly divided into 7 groups, with 8 mice in each group. One group was fed a basic diet, which was recorded as the blank control group. The other six groups were fed the same high-fat diet for modeling. After successful modeling, the high-fat diet-fed mice were treated with gavage of limonin or nomilin solutions at concentrations of 5 and 10 mg mL⁻¹ (prepared with 0.5% sodium carboxymethyl cellulose; the intervention concentration was referred to in Wang⁸et al. and Ono¹³et al.). Meanwhile, one group treated with ST solution at a concentration of 0.4 mg mL⁻¹ was used as the positive control group. Overall, the blank control group (NC), the model group (HFD), the high-dose limonin group (H-HL, 10 mg mL⁻¹), the low-dose limonin group (H-LL, 5 mg mL⁻¹), the high-dose nomilin group (H-HN, 10 mg mL⁻¹), the low-dose nomilin group (H-LN, 5 mg mL⁻¹), and the simvastatin group (ST, 0.4 mg mL⁻¹) were prepared, respectively. The mice were administered treatments by gavage at a fixed time every morning for six weeks, and their feces were collected every two weeks.

After the intervention, the mice were fasted and deprived of water for twelve hours. The mice were weighed and their body weights were recorded, and then blood was collected from the eyeballs. The mice were then sacrificed, and their body lengths were measured before dissection. Subsequently, the contents of the colon, the liver, and the testicular fat were collected, and the liver and epididymal adipose tissue were weighed and recorded. The eyeball blood was kept static for 2 hours and centrifuged at 3000 rpm at 4 °C for 30 min, and the supernatant was frozen at −80 °C. Colonic contents were flash-frozen in liquid nitrogen and stored at −80 °C. Colonic and epididymal adipose tissues were stored in 4% paraformaldehyde fixative for pathological observation. Lee's index and organ index were calculated using the following formulas:

Organ index = organ weight (g)/body weight (g).

2.3 Determination of serum lipid levels

The levels of serum TC, TG, HDL-C and LDL-C in mice were determined according to the instructions of the biochemical kit.

2.4 Pathological observation of colon and epididymal adipose tissue

The colon and epididymal adipose tissue were fixed in 4% paraformaldehyde for 24 h and embedded in paraffin. The colon was stained with Alcian blue (AB-PAS), and the epididymal adipose tissue was stained with hematoxylin–eosin (H&E).

2.5 Determination of SCFA contents in mouse feces

After successful modeling, the feces of mice were collected every two weeks and recorded as weeks 0, 2, 4, and 6. The contents of SCFAs were determined by gas chromatography with a flame ionization detector (Thermo Fisher Scientific, USA) with an Agilent HP-INNOWAX capillary column (30 m × 0.25 mm ID × 0.25 μm). GC analysis conditions: FID detector, carrier gas N₂, flow rate 20.0 mL min⁻¹, no split. The air flow rate was 300 mL min⁻¹ and the H₂ flow rate was 30 mL min⁻¹. The detector temperature was 260 °C, and the injection port temperature was 240 °C. Heating procedure: the initial column temperature was maintained at 100 °C for 0.5 minutes, then increased to 150 °C at a rate of 4 °C min⁻¹, and further increased to 185 °C at a rate of 5 °C min⁻¹. The sample injection volume was 1 μL, and each measurement time was 20 min. Data analysis was conducted using the HP Chem Workstation software (Agilent). Each sample was independently analyzed three times.

2.6 Analysis of the intestinal flora of mice

The colon contents of mice were analyzed through 16S rRNA high-throughput sequencing. The total DNA was extracted according to the operation requirements of the manual DNA extraction kit (Thermo Fisher Scientific, USA), and PCR amplification was performed on the V3–V4 variable region of the 16S rRNA gene. The upstream primer was 338F (5′-ACTCCTACGGGAGGCAGCAG-3′), while the downstream primer was 806R (5′-GGACTACHVGGGTWTCT-AAT-3′), and the amplification length was 468 bp. The PCR amplification products were purified, and libraries were constructed and quality-inspected. The sequencing results were submitted to the cloud platform of Shanghai Majorbio for comprehensive statistical analysis and visualization (https://www.majorbio.com).

2.7 Non-targeted metabolomics detection of colonic contents

LC-MS/MS analysis was performed on the colonic contents of mice using Thermo Fisher's ultra-performance liquid chromatography–tandem Fourier transform mass spectrometry UHPLC-Q Exactive HF-X system (Shanghai Meiji Biomedical Technology Co., Ltd.). Chromatographic conditions: 2 µL of the sample was separated using an Accucore C30 column (100 mm × 2.1 mm i.d., 2.6 µm; Thermo Fisher Scientific) and then introduced into the mass spectrometer for detection. Mobile phase A was a 10 mM ammonium acetate 50% acetonitrile aqueous solution (containing 0.1% formic acid), and mobile phase B was 2 mM ammonium acetate acetonitrile/isopropanol/water (10/88/2) (containing 0.02% formic acid). The flow rate was 0.40 mL min⁻¹ and the column temperature was 40 °C. Mass spectrometry conditions: the sample mass spectrometry signal was collected using positive and negative ion scanning modes, with a mass scanning range of m/z: 200–2000. The positive mode ion spray voltage was 3000 V, the negative mode ion spray voltage was −3000 V, the sheath gas was 60 psi, the auxiliary heating gas was 20 psi, the ion source heating temperature was 370 °C, the cyclic collision energy was 20–40–60 V, and the DDA mode was used to collect data. Non-target metabolomic analysis was conducted by Shanghai Meiji Biomedical Technology Co., Ltd. The data analysis and processing were performed using the Meiji Biological cloud platform (https://www.majorbio.com).

2.8 Statistical analysis

One-way analysis of variance (ANOVA) was used to compare the differences between groups. GraphPad Prism 9.0 and IBM SPSS Statistics 27 were used to complete data analysis and visualization. Data are presented as mean ± standard deviation. P < 0.05 was considered statistically significant.

3 Results and discussion

3.1 Effects of limonin and nomilin on lipid accumulation in hyperlipidemic mice

As shown in Fig. 1A–D, with high-fat feeding for two weeks, the body weight of mice was much higher than that of the NC group fed a non-high-fat diet. With intervention treatment for six weeks, the body weight, Lee's index and epididymal fat index of mice fed with a high-fat diet were significantly decreased by both limonin and nomilin (p < 0.05), and the effects were similar to those of positive drug ST treatment. In particular, high-dose nomilin could significantly reduce the body weight (decreased by 17.00%) and epididymal fat index (decreased by 49.00%) of mice compared with the HFD group (p < 0.05). For Lee's index, each intervention group decreased significantly compared with the HFD group (p < 0.05), where the H-HL and H-HN groups decreased to 0.938 and 0.934 of the HFD group, respectively, and showed no significant difference compared with the NC group. However, for the liver index, the reduction effect of high-dose limonin was higher than that of high-dose nomilin. In conclusion, it was suggested that both limonin and nomilin could reduce the weight gain and visceral fat index of high-fat diet-fed mice, and the effect of nomilin was better than that of limonin.

Fig. 1 Effects of limonin and nomilin on body weight, organ indices and blood lipid levels in mice: (A) changes in the body weight of mice; (B) Lee's index; (C) epididymal fat index; (D) liver index; (E) serum TC; (F) serum TG; (G) serum HDL-C; (H) serum LDL-C. Note: H-HL: high-dose limonin; H-LL: low-dose limonin; H-HN: high-dose nomilin; H-LN: low-dose nomilin; NC: normal control group; ST: positive control group; HFD: high-fat model group. All results are expressed as the mean ± SD, n = 6. Different letters indicate significant differences (p < 0.05).

As shown in Fig. 1E–H, compared with the NC group, the contents of serum TC, TG and LDL-C in the HFD group increased significantly (p < 0.05), which were 176.40%, 121.70% and 185.60% of those in the NC group, respectively. Similar to ST treatment, both limonin and nomilin interventions could significantly reduce the levels of TC, TG and LDL-C in the serum of high-fat diet-fed mice (p < 0.05). In particular, the TC level among the limonin and nomilin groups was reduced to around 70% of that in the HFD group. Moreover, the HDL-C level was significantly increased with both limonin and nomilin interventions (p < 0.05), and the highest effect was found in the high-dose nomilin group, with an increase of 71.95%. The above results indicated that limonin and nomilin could effectively improve the abnormal serum lipid contents in mice on a high-fat diet. Previous studies have also found that limonin has the ability to improve lipid metabolic homeostasis,¹⁵ and 0.2% w/w nomilin can reduce blood glucose and improve glucose tolerance in HFD mice.¹³ Overall, these results showed that both limonin and nomilin interventions had the effect of improving blood lipids. Notably, for the epididymal fat index and HDL-C, the improvement effect of high-dose nomilin was significantly better than that of limonin intervention groups (p < 0.05), suggesting potentially different lipid regulation mechanisms of limonin and nomilin.

3.2 Effects of limonin and nomilin on pathological changes of colon and adipose tissue

The results of colonic histopathological staining are shown in Fig. 2A. In the NC group, the colonic tissue structure and crypt structure of mice were intact and clear, and the goblet cells were neatly arranged without any pathological changes observed. In the HFD group, the mucus layer was partially destroyed, and the crypts were damaged in colonic tissue. Previous studies have also shown that mice fed a high-fat diet have a thinner mucus layer, which leads to constipation, triggers colonic inflammation and disrupts intestinal health.¹⁶ Compared with the HFD group, both limonin and nomilin interventions significantly alleviated structural damage, significantly improved mucus layer damage, and increased the number of goblet cells in the colonic tissue of hyperlipidemic mice (p < 0.05). As shown in Fig. 2C, the number of goblet cells in the HFD group was much lower than that in the NC group (p < 0.05), which was 69.88% of that in the NC group. With either limonin or nomilin intervention, the number of goblet cells increased by 17.19%–31.21% compared to that in the HFD group; in particular, high-dose nomilin showed the highest effects. These results indicated that both limonin and nomilin could effectively improve the colonic barrier injury in hyperlipidemic mice and maintain the integrity of the intestinal barrier. Studies have shown that limonin can improve ulcerative colitis by downregulating the levels of p-STAT3 and miR-214,¹⁷ while nomilin exhibits anti-inflammatory activity by binding to the key protein MD-2 and inhibiting the LPS-TLR4/MD-2-NF-κB signaling pathway.¹⁰ This might be the reason for the different effects of these two substances on intestinal tissues.

Fig. 2 Effects of limonin and nomilin on colonic and epididymal adipose tissue morphology in HFD mice: (A) colonic AB-PAS staining images (magnification ×100); (B) epididymal fat H&E staining images (magnification ×200); (C) number of colonic goblet cells; (D) mean area of epididymal adipocytes. Note: H-HL: high-dose limonin; H-LL: low-dose limonin; H-HN: high-dose nomilin; H-LN: low-dose nomilin; NC: normal control group; ST: positive control group; HFD: high-fat model group. All results are expressed as the mean ± SD, n = 3. Different letters indicate significant differences (p < 0.05).

The arrangement and size of adipocytes can directly reflect the lipid level of high-fat diet-fed mice.¹⁸ The epididymal histological staining results of mice are shown in Fig. 2B and D. In the NC group, adipocytes of mice were closely arranged and small in size, while in the HFD group, adipocytes increased to varying degrees. The average area of adipocytes in hyperlipidemic mice increased significantly by 180.63% when compared with the NC group (p < 0.05). Both limonin and nomilin interventions could significantly reduce the average size of adipocytes compared to the HFD group (p < 0.05), especially low-dose nomilin was more effective than low-dose limonin; however, there was no significant difference in the reduction effect between high-dose limonin and nomilin interventions. In conclusion, both limonin and nomilin could inhibit the expansion of adipose tissue in mice and improve the inflammatory state of adipose tissue, and the intervention effect of nomilin was better than that of limonin.

3.3 Effects of limonin and nomilin on fecal SCFAs in hyperlipidemic mice

SCFAs mainly include acetic acid, propionic acid, butyric acid, etc., which have multiple functions such as regulating energy metabolism, improving blood glucose and lipid levels, maintaining the integrity of the intestinal barrier, and regulating intestinal health.^19,20Fig. 3 shows the variation in the contents of SCFAs. With high-fat feeding, the acetic acid, propionic acid and valeric acid contents in hyperlipidemic mice were significantly lower compared with the NC group (p < 0.05), but the butyric acid content did not change significantly (p > 0.05). With the interventions of limonin and nomilin, the contents of acetic acid, propionic acid and valeric acid in the limonin and nomilin groups were much higher than those in the HFD group (p < 0.05). Compared with the HFD group, acetic acid, propionic acid and valeric acid increased by 30.00%–34.20%, 40.10%–42.60% and 55.80%–59.25%, respectively, after limonin intervention. With the nomilin intervention, acetic acid increased by 52.5%–53.8%, propionic acid increased by 35.77%–58.21%, and valeric acid increased by 32.41%–41.98% compared with the HFD group. Moreover, the high-dose effect of limonin and nomilin interventions on SCFAs was better than that of the low-dose ones. In the sixth week, the acetic acid concentration in the nomilin intervention group was significantly higher than that in the limonin group (p < 0.05), and the acetic acid concentration in the H-HN group was 113.60% of that in the H-HL group. For propionic acid, there was no significant difference in the effect of limonin and nomilin interventions. However, in terms of valeric acid levels, the nomilin group was significantly lower than the limonin group (p < 0.05), where the valeric acid concentration in the H-HN group was 91.10% of that in the H-HL group. These findings validate the potential of these citrus-derived limonoids as natural modulators of intestinal SCFA metabolism, offering a mechanistic basis for their previously observed lipid-lowering and gut-protective effects via enhancing SCFA production. Besides, nomilin exhibited a more potent ability to increase acetic acid, while limonin was more effective at increasing valeric acid. These nuances highlight the structural specificity of limonoids in targeting microbial metabolic pathways, providing valuable insights for optimizing natural product-based interventions tailored to specific SCFA-related health outcomes.

Fig. 3 Effects of limonin and nomilin on SCFA content in hyperlipidemic mice: (A) acetic acid concentration; (B) propionic acid concentration; (C) butyric acid concentration; (D) valeric acid concentration. Note: H-HL: high-dose limonin; H-LL: low-dose limonin; H-HN: high-dose nomilin; H-LN: low-dose nomilin; NC: normal control group; ST: positive control group; HFD: high-fat model group. All results are expressed as the mean ± SD, n = 5. Different letters indicate significant differences (p < 0.05).

3.4 Effects of limonin/nomilin on intestinal flora in hyperlipidemic mice

3.4.1 Diversity analysis. For α-diversity analysis (Fig. 4A–C), there was no significant difference in the diversity, evenness, and richness of intestinal microbiota in mice of different groups (p > 0.05). For β-diversity analysis (Fig. 4D–F), there was a significant difference between the NC group and the HFD group (p < 0.05), indicating that the high-fat diet changed the overall structure of the intestinal flora in mice. Compared with the HFD group, the intestinal flora of hyperlipidemic mice treated with different doses of limonin and nomilin was significantly different (p < 0.05). This finding indicated that both limonin and nomilin could change the intestinal flora of hyperlipidemic mice, and the stress of NMDS was less than 0.2, indicating that the data analysis was reasonable.

Fig. 4 Effects of limonin and nomilin on intestinal flora diversity: (A–C) α-diversity; (D–F) β-diversity. Note: H-HL: high-dose limonin; H-LL: low-dose limonin; H-HN: high-dose nomilin; H-LN: low-dose nomilin; NC: normal control group; ST: positive control group; HFD: high-fat model group. All results are expressed as the mean ± SD, n = 5. Different letters indicate significant differences (p < 0.05).

3.4.2 Species composition and differential analysis. To further analyze the effects of limonin and nomilin on the structure of the gut microbiota, the top ten taxa at the phylum level were determined. As shown in Fig. 5, it was observed that Firmicutes, Bacteroidetes, Verrucomicrobiota and Campylobacterota were the main taxa at the phylum classification level. Among them, Firmicutes and Bacteroidetes are the main phyla. Studies have shown that an increase in the ratio of Firmicutes and Bacteroidetes (F/B) may cause the occurrence of obesity,²¹ which can effectively reflect the advocated microecological balance and is especially closely related to the intestinal health of hyperlipidemic mice.²² As shown in Fig. 5B–F, compared with the NC group, the relative abundance of Firmicutes in the HFD group increased significantly (p < 0.05), while the relative abundance of Bacteroidetes decreased significantly (p < 0.05). The ratio of Firmicutes to Bacteroidetes (F/B) in the HFD group was significantly increased (p < 0.05), which was 3.262 compared with the NC group. Meanwhile, the abundance of Campylobacterota increased, and the abundance of Verrucomicrobiota in the HFD group was significantly lower than that in the NC group and the H-LL group (p < 0.05). However, the intervention of nomilin and ST had no significant effect on the abundance of Verrucomicrobiota. Relevant studies have proved that a decrease in the abundance of Verrucomicrobiota may promote the occurrence of intestinal metabolic disorders and inflammation.²³ The abundance of Campylobacter is positively correlated with obesity and intestinal inflammation.²⁴ After the interventions of limonin and nomilin, the F/B ratios of H-HL, H-HN and H-LN groups decreased significantly (p < 0.05), and the effect of nomilin was significantly better than that of limonin (p < 0.05). Campylobacter abundance in the limonin and nomilin intervention groups was significantly lower (p < 0.05). In conclusion, the effect of nomilin on Firmicutes and Bacteroidetes was better than that of limonin, and the effect of limonin on Verrucomicrobiota was better than that of nomilin. At the F/B level, the effects of the H-HN and H-LN groups were significantly better than those of the ST group (p < 0.05), and there was no significant difference compared with the NC group.

Fig. 5 Differences in intestinal species at the phylum level: (A) relative abundance at the phylum level; (B) relative abundance of Firmicutes; (C) relative abundance of Bacteroidetes; (D) Firmicutes/Bacteroidetes; (E) relative abundance of Campylobacterota; (F) relative abundance of Verrucomicrobiota. Note: H-HL: high-dose limonin; H-LL: low-dose limonin; H-HN: high-dose nomilin; H-LN: low-dose nomilin; NC: normal control group; ST: positive control group; HFD: high-fat model group. All results are expressed as the mean ± SD, n = 5. Different letters in the figure indicate significant differences (p < 0.05).

At the genus level, the top ten bacterial genera are shown in Fig. 6A, and there were significant changes in Lachnospiraceae, Helicobacter, Bacteroides, and Desulfovibrio in hyperlipidemic mice treated with limonin and nomilin (p < 0.05). As shown in Fig. 6B–E, high-dose nomilin significantly improved the abundance of Lachnospiraceae and reduced the abundance of Helicobacter, Bacteroides and Desulfovibrio. Moreover, for Lachnospiraceae, the improvement effects of limonin and nomilin were significantly better than those of the NC group and the ST group (p < 0.05). As previous studies have demonstrated, Lachnospiraceae has the potential to regulate the host immune system, reduce the level of inflammation in the intestine, and maintain the balance of intestinal flora,²⁵ and an increase in the abundance of Lachnospiraceae may enhance the production of SCFAs.²⁶ It has been reported that Helicobacter may cause dyslipidemia by activating inflammatory factors in the gastrointestinal tract,²⁷ and the reduction of Helicobacter abundance can effectively relieve colitis.²⁸ Studies have shown that the abundance of Bacteroides increases in mice with alcoholic fatty liver disease, which affects lipid metabolism and intestinal health.²⁹ The abundance of Helicobacter and Bacteroides in the intervention group was significantly lower than that of the HFD group (p < 0.05). However, there was no significant difference in the improvement effects of limonin and nomilin on the abundance levels of Lachnospiraceae, Helicobacter and Bacteroides. For the harmful bacterium Desulfovibrio, both limonin and nomilin interventions could significantly reduce the abundance of Desulfovibrio (p < 0.05). Regarding the abundance of Desulfovibrio, the nomilin group was significantly better than the limonin group. The limonin intervention could reduce the abundance of Desulfovibrio to 65.51%–84.37% of that in the HFD group. The abundance of Desulfovibrio was reduced to 37.69% of that in the HFD group after the low-dose nomilin intervention, and that in the high-dose nomilin group was not significantly different from that in the NC group (p > 0.05). Studies have shown that the abundance of Desulfovibrio can affect lipid metabolism and is positively correlated with the occurrence of obesity.³⁰

Fig. 6 Difference analysis at the genus level: (A) relative abundance at the genus level; (B) relative abundance of Lachnospiraceae; (C) relative abundance of Helicobacter; (D) relative abundance of Bacteroides; (E) relative abundance of Desulfovibrio. Note: H-HL: high-dose limonin; H-LL: low-dose limonin; H-HN: high-dose nomilin; H-LN: low-dose nomilin; NC: normal control group; ST: positive control group; HFD: high-fat model group. All results are expressed as the mean ± SD, n = 5. Different letters in the figure indicate significant differences (p < 0.05).

These findings identified genus-specific gut microbial dysbiosis associated with HFD-induced dyslipidemia, confirming that HFD disrupted the abundance of functionally critical bacterial genera—including the beneficial Lachnospiraceae and harmful taxa such as Helicobacter, Bacteroides, and Desulfovibrio. Both limonin and nomilin exerted targeted regulatory effects on these functionally relevant genera in a dose-dependent manner. This finding validated their potential as natural gut microbiota modulators, providing a critical mechanistic link between their previously observed lipid-lowering/intestinal-protective effects and the regulation of key microbial taxa. Besides, nomilin exhibited superior efficacy in reducing Desulfovibrio abundance. This distinction emphasized the structural specificity of limonoids in interacting with gut microbial communities, suggesting nomilin as a more promising candidate for targeting Desulfovibrio-related metabolic disorders (e.g., obesity and dyslipidemia). Overall, the above research provides new insights into how limonin and nomilin participate in regulating the intestinal microecology during metabolic disease states. The findings provide some valuable preliminary data for further exploration of functional foods and dietary supplements that focus on regulating the intestinal microbiota, as well as for developing new clinical intervention strategies targeting dyslipidemia and related metabolic complications.

3.5 Effects of limonin and nomilin on intestinal metabolites in hyperlipidemic mice

3.5.1 Effects of limonin and nomilin on differential metabolites in hyperlipidemic mice. According to the above research results, the high-dose effects of limonin/nomilin interventions were generally better than those of the low-dose one; thus, the high-dose groups were selected for subsequent metabolite analysis. The characteristic metabolites of intestinal contents can effectively reflect the lipid metabolism level, and the related metabolic pathways are able to reflect the metabolic differences in mice in the high-fat feeding environment.³¹ The volcano map of metabolite differences is shown in Fig. 7A–C, which contains related metabolic pathways mainly including the biosynthesis of amino acids, secondary bile acid biosynthesis, arachidonic acid metabolism, tryptophan metabolism, primary bile acid biosynthesis, cholesterol metabolism, steroid hormone biosynthesis, etc. As shown in Fig. 7D, arachidonic acid metabolism and steroid hormone biosynthesis can effectively adjust the dyslipidemia caused by a high-fat diet;³² tryptophan metabolism is negatively correlated with adiponectin and is significantly elevated in prediabetic and metabolically unhealthy obesity, which may contribute to obesity-related metabolic diseases.³³ Besides, studies have shown that primary bile acid biosynthesis is closely related to HFD-induced hypercholesterolemia and can promote the excretion of cholesterol in the feces of high-fat diet-fed mice.³⁴ Based on the above metabolic pathways, this study performed descriptive statistics on related metabolites to study the effect of limonin/nomilin on fecal metabolites in high-fat diet-fed mice and to provide a theoretical basis for the subsequent analysis of related metabolites.

Fig. 7 Metabolite volcano diagrams and related metabolic pathways: (A) NC vs. HFD; (B) H-HL vs. HFD; (C) H-HN vs. HFD; (D) KEGG pathway. Note: H-HL: high-dose limonin; H-LL: low-dose limonin; H-HN: high-dose nomilin; H-LN: low-dose nomilin; NC: normal control group; ST: positive control group; HFD: high-fat model group.

Compared with the NC group, the HFD group showed 537 significantly upregulated metabolites and 206 significantly downregulated metabolites (p < 0.05). As shown in Tables 1–3, the significantly upregulated metabolites included L-histidinol, isocitric acid, L-saccharopine, 2-aminobenzoic acid, etc., and the significantly downregulated metabolites included pretyrosine, (R)-2,3-dihydroxy-3-methylvalerate, etc. (p < 0.05). Compared with the HFD group, in the high-dose limonin and nomilin intervention groups, 151 and 252 significantly upregulated metabolites were detected, while 58 and 55 significantly downregulated metabolites were found, respectively (p < 0.05). Among them, the upregulated metabolites included (S)-10,16-dihydroxyhexadecanoic acid, 2′E,4′Z,7′Z,8E-colnelenic acid, 15-deoxy-D-12,14-Pgj2, etc. The downregulated metabolites included dehydroepiandrosterone, 4-hydroxyphenylpyruvic acid, 2-hydroxyethanesulfonate, cholesterol sulfate, etc. As previous studies have reported, a certain amount of colnelenic acid can improve lipid metabolism disorders caused by a high-fat diet.³⁵ In addition, the above metabolic pathways including cutin, suberin and wax biosynthesis, α-linolenic acid metabolism, arachidonic acid metabolism, steroid hormone biosynthesis, etc., are closely related to lipid metabolism. The findings suggested the ability of limonin and nomilin to remodel the metabolome toward a lipid-friendly state, and both of them target functionally relevant metabolites by upregulating beneficial metabolites such as (S)-10,16-dihydroxyhexadecanoic acid and 2′E,4′Z,7′Z,8E-colnelenic acid, while downregulating detrimental metabolites including dehydroepiandrosterone, 4-hydroxyphenylpyruvic acid, and cholesterol sulfate. Notably, nomilin exhibited a more potent regulatory capacity (252 upregulated and 55 downregulated metabolites) compared to limonin (151 upregulated and 58 downregulated metabolites).

Table 1 Statistics of metabolites with significant differences in metabolic pathways between the NC and HFD groups

Metabolic pathway	Total	Metabolites
		Up	Down
The difference between the NC and HFD groups is expressed as ***p < 0.001, **p < 0.01, and *p < 0.05; N.D. not detected.
Biosynthesis of amino acids	6	L-Histidinol*/isocitric acid**/L-saccharopine*/2-aminobenzoic acid**	Pretyrosine**/(R)-2,3-dihydroxy-3-methylvalerate**
Secondary bile acid biosynthesis	5	N.D.	Taurocholic acid**/taurochenodeoxycholic acid***/deoxycholic acid***/isolithocholic acid***/glycochenodeoxycholic acid*
Arachidonic acid metabolism	7	Prostaglandin I2**/15-deoxy-D-12,14-Pgj2***/dinoprost*/thromboxane*/arachidonic acid**	Prostaglandin A2*/20-Hete***
Tryptophan metabolism	6	Tryptamine***/3-(3-indolyl)-2-oxopropanoic acid*/tryptophol***/2-aminobenzoic acid**/N′-formylkynurenine**/indole-3-acetic acid***	N.D.
Primary bile acid biosynthesis	6	7Alpha-hydroxy-3-oxo-4-cholestenoate*	Taurocholic acid**/taurochenodeoxycholic acid***/7alpha,12alpha-dihydroxycholest-4-en-3-one*/5B-cyprinol sulfate***/glycochenodeoxycholic acid*
Cholesterol metabolism	3	N.D.	Taurochenodeoxycholic acid***/glycochenodeoxycholic acid*/taurocholic acid**
Steroid hormone biosynthesis	7	Estriol***	Corticosterone***/21-deoxycortisol***/17A,20A-dihydroxycholesterol***/cortisol**/desoxycortone*/dehydroepiandrosterone***

---

Table 2 Statistics of metabolites with significant differences in metabolic pathways between the H-HL and HFD groups

Metabolic pathway	Total	Metabolites
		Up	Down
The difference between the H-HL and HFD groups is expressed as ***p < 0.001, **p < 0.01, and *p < 0.05; N.D. not detected.
Steroid hormone biosynthesis	4	Estriol*/2-methoxyestrone 3-glucuronide***	Dehydroepiandrosterone*/cholesterol sulfate*
Arachidonic acid metabolism	2	Prostaglandin I2**/15-deoxy-D-12,14-Pgj2*	N.D.
Biosynthesis of amino acids	4	Ketoleucine**/L-histidinol*/L-saccharopine*/citric acid*	N.D.
Biosynthesis of alkaloids derived from histidine and purine	2	Caffeine*/citric acid*	N.D.
Steroid degradation	1	N.D.	Dehydroepiandrosterone*
Cutin, suberine and wax biosynthesis	1	(S)-10,16-Dihydroxyhexadecanoic acid*	N.D.
α-Linolenic acid metabolism	1	(2′E,4′Z,7′Z,8E)-Colnelenic acid*	N.D.

---

Table 3 Statistics of metabolites with significant differences in metabolic pathways between the H-HN and HFD groups

Metabolic pathway	Total	Metabolites
		Up	Down
The difference between the H-HN and HFD groups is expressed as ***p < 0.001, **p < 0.01, and *p < 0.05; N.D. not detected.
Steroid hormone biosynthesis	4	Estrone 3-sulfate***/estrone glucuronide***/17A,21-dihydroxy-5B-pregnane-3,11,20-trione***	Dehydroepiandrosterone*
Biosynthesis of amino acids	3	Ketoleucine**/L-saccharopine**	4-Hydroxyphenylpyruvic acid*
Isoquinoline alkaloid biosynthesis	3	Morph***/coclaurine*	4-Hydroxyphenylpyruvic acid*
Bile secretion	2	Estrone 3-sulfate***/etoposide***	N.D.
2-Oxocarboxylic acid metabolism	3	Ketoleucine**/7-methylthioheptyl glucosinolate*	4-Hydroxyphenylpyruvic acid*
Taurine and hypotaurine metabolism	1	N.D.	2-Hydroxyethanesulfonate**

---

Compared with nomilin, limonin specifically upregulated prostaglandin I2 in the arachidonic acid metabolic pathway, (2′E,4′Z,7′Z,8E)-colnelenic acid and indole alkaloids in the α-linolenic acid metabolic pathway, and 17-O-deacetylvindoline in the biosynthetic pathway. Compared with limonin, nomilin intervention could specifically upregulate estrone glucuronide and estrone in the steroid hormone biosynthesis metabolic pathway, and N′-formylkynurenine in the 3-sulfate and 17A,21-dihydroxy-5B-pregnane-3,11,20-trione and tryptophan metabolism pathways. The results revealed distinct metabolic pathway-specific targets of limonin and nomilin, suggesting the structural specificity of these citrus-derived limonoids in reshaping host metabolism. In addition, Li et al. have found that steroid hormone biosynthesis can effectively regulate lipid metabolism in high-fat diet-fed mice and is closely related to intestinal inflammation.³⁶ Similarly, our study showed that nomilin's targeted upregulation of key metabolites in this pathway aligns with its previously observed superior lipid-regulating effects. These results can preliminarily establish a potential mechanistic link between the metabolic reprogramming of the specific pathways affected by limonin and nomilin and their beneficial functions. By comparing the differential effects of limonin and nomilin on the microbiota and metabolism, this study may offer a theoretical reference for the future development of related lipid-lowering products.

3.5.2 Correlation analysis between characteristic fecal metabolites and blood lipid levels. To further explore the correlation between metabolites and blood lipid levels in hyperlipidemic mice with limonin/nomilin interventions, Spearman correlation analysis was conducted on fecal characteristic metabolites and serum biochemical indicators (TC, TG, HDL-C, LDL-C), as shown in Fig. 8. The results showed that these metabolites had significant correlations with serum biochemical indicators. Among them, seven characteristic metabolites—sphingosine, 5B-cyprinol sulfate, 12-oxo leukotriene B4, 7-dehydroxychol-8(14)-enoic acid, 6,9,12,15,18, 21-tetracosahexaenoic acid, laurocapram and deoxycholic acid—were significantly positively correlated with TC and LDL-C, and significantly negatively correlated with HDL-C (p < 0.05). In contrast, 2′,3-dihydroxy-4,4′,6′-trimethoxychalcone and N-acetylneuraminic acid were significantly positively correlated with HDL-C and significantly negatively correlated with LDL-C (p < 0.05). Similarly, previous studies have shown that diabetes can cause an increase in the metabolite sphingosine, affecting the overall health of the body.³⁷ The level of deoxycholic acid in the intestine of HFD mice increases, which may cause colitis.³⁸ Glucose tolerance can be effectively improved, and insulin levels are reduced after intervention with N-acetylneuraminic acid.³⁹ These findings provide evidence that fecal metabolite profiles are tightly coupled to systemic lipid homeostasis, and elucidating how limonin- and nomilin-mediated metabolite remodeling may simultaneously regulate lipid levels and mitigate diet-induced pathological risks.

Fig. 8 Correlation analysis of metabolites with blood lipid levels.

The overall findings of this study demonstrated that nomilin exhibited superior lipid-lowering activity relative to limonin, and this difference may be closely associated with the structural differences between the two limonoid derivatives. Both limonin and nomilin belong to the limonoid family and share a similar core structural framework; however, nomilin possesses distinct structural features that contribute to its enhanced biological activity in lipid metabolism regulation. Specifically, nomilin has a more stable structural skeleton and contains a unique acetoxy group, as discussed above, which may increase its polarity and chemical reactivity compared to limonin. These structural disparities tended to govern the interaction efficiency of the two compounds with biological targets (e.g., receptors, enzymes, or signaling molecules), thereby leading to differences in their biological functions and activity strengths. Existing research has further supported the structural specificity of limonoids in mediating different biological effects. For example, it has been found that limonin shows better anti-inflammatory activity in the aspect of inflammation,⁴⁰ while nomilin has good effects in detoxification⁴¹ and regulating metabolism.⁴² Notably, nomilin can act as a TGR5 agonist, directly interfering with downstream signaling pathways to modulate systemic metabolism, thereby exerting more robust anti-obesity and hypoglycemic effects¹³—biological activities that are closely linked to lipid homeostasis regulation. This aligns with the results of our study, which revealed distinct metabolic remodeling patterns induced by limonin and nomilin. Moreover, our study showed that limonin could specifically upregulate prostaglandin I2 in the arachidonic acid metabolic pathway, (2′E,4′Z,7′Z,8E)-colnelenic acid and indole alkaloids in the α-linolenic acid metabolic pathway, as well as 17-O-deacetylvindoline in the biosynthetic pathway, while nomilin intervention specifically up-regulated estrone glucuronide and estrone in the steroid hormone biosynthesis metabolic pathway, and N′-formylkynurenine in the 3-sulfate and 17A,21-dihydroxy-5B-pregnane-3,11,20-trione and tryptophan metabolism pathways. These divergent metabolic regulatory effects further highlight the structural specificity of limonoids in targeting key pathways associated with lipid metabolism. In conclusion, the unique acetoxy group of nomilin may enhance its solubility in the intestinal microenvironment, promote its absorption and bioavailability, and strengthen its binding affinity for target proteins involved in lipid metabolism. Additionally, the more stable structural framework of nomilin may prolong its half-life in vivo, allowing for sustained regulation of metabolic pathways and gut microbiota, thereby contributing to its superior lipid-lowering efficacy. The pathway-specific regulation observed in this study (e.g., nomilin targeting steroid hormone biosynthesis and tryptophan metabolism) further explains why nomilin outperforms limonin in the regulation of lipid metabolic homeostasis. These observations can provide a theoretical basis for structural modification and optimization of limonoids to enhance their therapeutic potential in lipid disorders.

4 Conclusion

This study investigated the regulatory effects of limonin and nomilin on lipid metabolic homeostasis in hyperlipidemic mice via intestinal microbiota and metabolomics. Both compounds significantly inhibited HFD-induced body weight gain, reduced epididymal white fat and liver weights, lowered serum TC, TG and LDL-C levels, alleviated colonic injury and attenuated adipocyte hypertrophy, with high-dose nomilin showing stronger effects than limonin. Besides, the intervention of limonin and nomilin could also increase the abundance of probiotics such as Lachnospiraceae in the intestines of high-fat diet-fed mice, and significantly reduce the abundance of Helicobacter, Bacteroides and Desulfovibrio (p < 0.05). Non-targeted metabolomics showed that limonin mainly modulated arachidonic acid and α-linolenic acid metabolism, while nomilin targeted steroid hormone biosynthesis and tryptophan metabolism. These distinct metabolic and microbial signatures highlight the structure-specific actions of limonoids in lipid regulation. Moreover, this work provides a scientific foundation for elucidating the lipid-lowering mechanisms of limonin derivatives and facilitates their future development as promising nutritional or therapeutic agents against hyperlipidemia and related metabolic syndrome.

Author contributions

Nan Zhang & Yujia Ou: investigation and writing – original draft. Tao Hong: formal analysis and writing – review & editing. Fan He & Zhipeng Li: conceptualization and formal analysis. Yanbing Zhu: formal analysis, validation, and writing – review & editing. Hui Ni & Zedong Jiang: conceptualization and writing – review & editing. Yang Hu & Mingjing Zheng: writing – review & editing, supervision, and funding acquisition.

Conflicts of interest

The authors confirm that they have no conflicts of interest with respect to the work described in this manuscript.

Ethics statements

All animal procedures were performed in strict accordance with the Guidelines for the Care and Use of Laboratory Animals of the People's Republic of China and the institutional guidelines established by the 900th Hospital of the Joint Support Force of the People's Liberation Army of China (Fuzhou, China). Animals were housed in the animal facility of the 900th Hospital of the Joint Support Force of the People's Liberation Army of China (Fuzhou, China; license no.: SYXK (Military) 2025-0016). Ethical approval for all animal experiments was obtained from the 900th Hospital of the Joint Logistic Support Force of the People's Liberation Army of China (approval number: 2025-04).

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgements

This work was supported by the Fujian Provincial Department of Science and Technology, China (grant number 2023J01773), and the Natural Science Foundation of Xiamen, China (grant number 3502Z202373037).

References

1. R. Lahey and S. S. Khan, Trends in Obesity and Risk of Cardiovascular Disease, Curr. Epidemiol. Rep., 2018, 5, 243–251,  DOI:10.1007/s40471-018-0160-1.
2. S. Klein, A. Gastaldelli, H. Yki-Järvinen and P. E. Scherer, Why does obesity cause diabetes?, Cell Metab., 2022, 34, 11–20.
3. J. Xie, L. Zhang, Y.-Y. Bai, W.-J. Wang, X. Hu, S. Li and Y. Tian, Isolation and purification of polyphenols, hypoglycemic and hypolipidemic and active constituent analysis of walnut septum polyphenols, Food Chem., 2025, 467, 142287.
4. Y. Chen, J. Hong, Z. Jiang, L. Wu, X. Wang, Y. Zhu, Z. Jiang, H. Ni and M. Zheng, Effect of red algae powder on gel properties and in vitro hypolipidemic activity of fish balls, Algal Res., 2023, 74, 103223.
5. J. D. Bermudes-Contreras, M. V. Gutiérrez-Velázquez, E. A. Delgado-Alvarado, R. Torres-Ricario and J. Cornejo-Garrido, Hypoglycemic and hypolipidemic effects of triterpenoid standardized extract of agave durangensis gentry, Plants, 2025, 14(6), 894.
6. S. Hasegawa and Z. Herman, Biosynthesis of limonoids: Conversion of deacetylnomilinate to nomilin in Citrus limon, Phytochemistry, 1986, 25, 2523–2524.
7. F. T. Zohra, Y. Tominaga, Y. Matsumoto, N. Taguchi, A. Okubo, S. Mitsutake and N. Kotoda, Relationship between the limonoid content in different parts of the sour orange (Citrus aurantium L.) and the ligand activity of a bile acid receptor, TGR5, Hortic. J., 2020, 89(4), 384–393.
8. S. Wang, T. Lan, H. Chen, H. Sheng, C. Xu, L. Xu, F. Zheng and F. Zhang, Limonin, an AMPK Activator, Inhibits Hepatic Lipid Accumulation in High Fat Diet Fed Mice, Front. Pharmacol., 2022, 13, 833705.
9. S. Baba, Y. Ueno, T. Kikuchi, R. Tanaka and K. Fujimori, A limonoid kihadanin B from immature citrus unshiu peels suppresses adipogenesis through repression of the akt-FOXO1-PPARγ axis in adipocytes, J. Agric. Food Chem., 2016, 64, 9607–9615.
10. Y. Chen, S. Guo, G. Chen, C. Liu, M. Zhang and X. Wang, Nomilin Attenuates Lipopolysaccharide-Induced Inflammatory Response by Binding with Myeloid Differentiation Protein-2Comb, Chem. High Throughput Screening, 2023, 26, 2469–2475.
11. S. J. Fan, Y. X. Yan, Y. Xia, Z. Y. Zhou, L. L. Luo, M. N. Zhu, Y. L. Han, D. Q. Yao, L. J. Zhang, M. L. Fang, L. N. Peng, J. Yu, Y. Liu, X. Y. Gao, H. D. Guan, H. L. Li, C. H. Wang, X. J. Wu, H. H. Zhu, Y. Cao and C. Huang, Pregnane X receptor agonist nomilin extends lifespan and healthspan in preclinical models through detoxification functions, Nat. Commun., 2023, 14, 3368.
12. Y. Ding, H. Guan, Y. Yan, Y. Chen and C. Huang, Pharmacokinetic and pharmacodynamic drug-drug interaction of Nomilin with atorvastatin in hyperlipidemic mice, Heliyon, 2023, 9, e22016.
13. E. Ono, J. Inoue, T. Hashidume, M. Shimizu and R. Sato, Anti-obesity and anti-hyperglycemic effects of the dietary citrus limonoid nomilin in mice fed a high-fat diet, Biochem. Biophys. Res. Commun., 2011, 410, 677–681.
14. Y. L. Han, L. L. Luo, H. L. Li, L. J. Zhang, Y. X. Yan, M. L. Fang, J. Yu, X. Y. Gao, Y. Liu, C. Huang and S. J. Fan, Nomilin and its analogue obacunone alleviate NASH and hepatic fibrosis in mice via enhancing antioxidant and anti-inflammation capacity, BioFactors, 2023, 49, 1189–1204.
15. W. Wu, Y. Pan, T. Zheng, H. Sun, X. Li, H. Zhu, Z. Wang and X. Zhou, Limonin alleviates high-fat diet-induced dyslipidemia by regulating the intestinal barrier via the microbiota-related ILC3–IL22–IL22R pathway, Food Funct., 2024, 15, 2679–2692.
16. R. Mukai, O. Handa, Y. Naito, S. Takayama, Y. Suyama, C. Ushiroda, A. Majima, Y. Hirai, K. Mizushima, T. Okayama, K. Katada, K. Kamada, K. Uchiyama, T. Ishikawa, T. Takagi and Y. Itoh, High-Fat Diet Causes Constipation in Mice via Decreasing Colonic Mucus, Dig. Dis. Sci., 2019, 65, 2246–2253.
17. S. Liu, S. Zhang, X. Lv, J. Lu, C. Ren, Z. Zeng, L. Zheng, X. Zhou, H. Fu, D. Zhou and Y. Chen, Limonin ameliorates ulcerative colitis by regulating STAT3/miR-214 signaling pathway, Int. Immunopharmacol., 2019, 75, 105768.
18. L. Zheng, J. Wu, J. Mo, L. Guo, X. Wu and Y. Bao, Polydatin Inhibits Adipose Tissue Inflammation and Ameliorates Lipid Metabolism in High-Fat-Fed Mice, BioMed Res. Int., 2019, 2019, 1–10.
19. E. E. Blaak, E. E. Canfora, S. Theis, G. Frost, A. K. Groen, G. Mithieux, A. Nauta, K. Scott, B. Stahl, J. van Harsselaar, R. van Tol, E. E. Vaughan and K. Verbeke, Short chain fatty acids in human gut and metabolic health, Benefic. Microbes, 2020, 11, 411–455.
20. J. Zheng, Y. An, Y. Du, Y. Song, Q. Zhao and Y. Lu, Effects of short-chain fatty acids on blood glucose and lipid levels in mouse models of diabetes mellitus: A systematic review and network meta-analysis, Pharmacol. Res., 2024, 199, 107041.
21. G. H. Baek, K. M. Yoo, S. Kim, D. H. Lee, H. Chung, S. Jung, S. Park and J. Kim, Collagen peptide exerts an anti-obesity effect by influencing the Firmicutes/Bacteroidetes ratio in the gut, Nutrients, 2023, 15, 2610.
22. F. Lin, X. Li, X. Guo, X. Lu, X. Han, G. Xu, P. Du and L. An, Study on the hypolipidemic effect of Inonotus obliquus polysaccharide in hyperlipidemia rats based on the regulation of intestinal flora, Food Sci. Nutr., 2022, 11, 191–203.
23. X. Miao, P. Luo, J. Liu, J. Wang and Y. Chen, Dihydromyricetin ameliorated nonalcoholic steatohepatitis in mice by regulating the composition of serous lipids, bile acids and ileal microflora, Lipids Health Dis., 2023, 22, 112.
24. M. Bailén, C. Bressa, S. Martínez-López, R. González-Soltero, M. G. Montalvo Lominchar, C. San Juan and M. Larrosa, Microbiota Features Associated With a High-Fat/Low-Fiber Diet in Healthy Adults, Front. Nutr., 2020, 7, 583608.
25. C. Fontes-Candia, P. Jiménez-Barrios, B. Miralles, I. Recio, A. López-Rubio and M. Martínez-Sanz, Development of polysaccharide-casein gel-like structures resistant to in vitro gastric digestion, Food Hydrocolloids, 2022, 127, 107505.
26. L. Gan, J. Han, C. Li, J. Tang, X. Wang, Y. Ma, Y. Chen, D. Xiao and X. Guo, Tibetan highland barley fiber improves obesity and regulates gut microbiota in high-fat diet-fed mice, Food Biosci., 2023, 53, 102620.
27. Y. Furuto, M. Kawamura, J. Yamashita, T. Yoshikawa, A. Namikawa, R. Isshiki, H. Takahashi and Y. Shibuya, Relationship between Helicobacter pylori infection and arteriosclerosis, Int. J. Gen. Med., 2021, 14, 1533–1540.
28. P. Li, N. Xiao, L. Zeng, J. Xiao, J. Huang, Y. Xu, Y. Chen, Y. Ren and B. Du, Structural characteristics of a mannoglucan isolated from Chinese yam and its treatment effects against gut microbiota dysbiosis and DSS-induced colitis in mice, Carbohydr. Polym., 2020, 250, 116958.
29. M. H. Do, H. H. L. Lee, M. Park, M.-J. Oh, E. Lee, M. Kweon and H.-Y. Park, Morinda citrifolia extract prevents alcoholic fatty liver disease by improving gut health, J. Med. Food, 2022, 25, 1102–1111.
30. J. Zhang, C. Yi, J. Han, T. Ming, J. Zhou, C. Lu, Y. Li and X. Su, Novel high–docosahexaenoic–acid tuna oil supplementation modulates gut microbiota and alleviates obesity in high–fat diet mice, Food Sci. Nutr., 2020, 8, 6513–6527.
31. Z. Zhang, Y. Cui, X. Zhang, X. Hu, S. Li and T. Li, Gut microbiota combined with serum metabolites to reveal the effect of Morchella esculenta polysaccharides on lipid metabolism disordered in high-fat diet mice, Int. J. Biol. Macromol., 2024, 281, 136380.
32. J. Gao, L. Ma, J. Ma, S. Xia, S. Gong, Y. Yin and Y. Chen, Camellia (Camellia oleifera Abel.) Seed Oil Regulating of Metabolic Phenotype and Alleviates Dyslipidemia in High Fat-Fed Mice through Serum Branch-Chain Amino Acids, Nutrients, 2022, 14, 2424.
33. J. Lischka, A. Schanzer, M. Baumgartner, C. de Gier, S. Greber-Platzer and M. Zeyda, Tryptophan Metabolism Is Associated with BMI and Adipose Tissue Mass and Linked to Metabolic Disease in Pediatric Obesity, Nutrients, 2022, 14, 286.
34. L. Jin, H. Dang, J. Wu, L. Yuan, X. Chen and J. Yao, Supplementation of Weizmannia coagulans BC2000 and ellagic acid inhibits high-fat-induced hypercholesterolemia by promoting liver primary bile acid biosynthesis and intestinal cholesterol excretion in mice, Microorganisms, 2023, 11, 264.
35. J. Zhang, O. Wang, Y. Guo, T. Wang, S. Wang, G. Li, B. Ji and Q. Deng, Effect of increasing doses of linoleic and α-linolenic acids on high-fructose and high-fat diet induced metabolic syndrome in rats, J. Agric. Food Chem., 2016, 64, 762–772.
36. X. Li, K. S. W. Tan and L. Wang, Black rice bran improved lipid metabolism in hyperlipidemia mice via steroid hormone biosynthesis and PPAR signaling pathway, J. Pharm. Biomed. Anal., 2025, 262, 116830.
37. Y. Sun, P. Sun, Y. Hu, L. Shan, Q. Geng, Y. Gong, H. Fan, T. Zhang and Y. Zhou, Elevated testicular apoptosis is associated with elevated sphingosine driven by gut microbiota in prediabetic sheep, BMC Biol., 2022, 20, 121.
38. C. Wang, Q. Chu, W. Dong, X. Wang, W. Zhao, X. Dai, W. Liu, B. Wang, T. Liu, W. Zhong, C. Jiang and H. Cao, Microbial metabolite deoxycholic acid-mediated ferroptosis exacerbates high-fat diet-induced colonic inflammation, Mol. Metab., 2024, 84, 101944.
39. W. Zhang, L. Zhou, X. Huang, X. Zhao, H. Zheng, D. Guo, X. Chen, L. Pan, Y. Li and H. Li, Preventive and controlling effects of N-acetylneuraminic acid in regulating glucose and lipid metabolism disorders via the gut-liver axis in high-fat diet mice, Food Funct., 2025, 16, 5151–5165.
40. M. Song, H. Luo, Z. Li, L. Qiu, Y. Zhao, C. He, X. Zhang, W. Ye, L. Lin and Q. Zhang, Limonoids from the roots of Melia azedarach and their anti-inflammatory activity, Phytochemistry, 2023, 216, 113869.
41. S. Fan, Y. Yan, Y. Xia, Z. Zhou, L. Luo, M. Zhu, Y. Han, D. Yao, L. Zhang, M. Fang, L. Peng, J. Yu, Y. Liu, X. Gao, H. Guan, H. Li, C. Wang, X. Wu, H. Zhu, Y. Cao and C. Huang, Pregnane X receptor agonist nomilin extends lifespan and healthspan in preclinical models through detoxification functions, Nat. Commun., 2023, 14, 3368.
42. W. Li, N. Wang, J. Liu, Y. Ma, Q. Wang, W. Li, Y. Mu, Q. Wu and C. Wang, Isolation and bioactivities of nomilin from citrus family: a review, J. Future Foods, 2025, 6, 730–739.

---

Footnote

† These authors contributed equally to this work.

---