Laurolitsine ameliorates insulin resistance, ovarian dysfunction and gut microbiota dysbiosis in DHEA plus HFD-induced PCOS mice

Laurolitsine (PubChem CID: 22179, Catalog Number: BP0922, Purity ≥ 98%) was supplied by Desite (Chengdu, China). Dehydroepiandrosterone (DHEA) (PubChem CID: 5881, Catalog Number: GC11070, Purity ≥ 99%) was supplied by Glpbio (USA). Metformin (PubChem CID:4091, Catalog Number: S30880, Purity ≥ 98%) was supplied by Shyuanye (Shanghai, China). A high-fat diet (HFD, Catalog Number: D12492, 60 Kcal% Fat) was supplied by Future Biotech (Beijing, China). The triglyceride (TG, Cat Number: A110-1-1), cholesterol (CHO, Cat Number: A111-1-1), low density lipoprotein cholesterol (LDL-c, Cat Number: A113-1-1), high density lipoprotein cholesterol (HDL-c, Cat Number: A112-1-1), aspartate aminotransferase (AST, Cat Number: C010-2-1), and alanine aminotransferase (ALT, Cat Number: C009-2-1) assay kits were purchased from Njjcbio (Nanjing, China). The serum testosterone (T, Cat Number: E-OSEL-M0003), luteinizing hormone (LH, Cat Number: E-EL-M3053), estradiol (E2, Cat Number: E- E-OSEL-M0008), and follicle-stimulating hormone (FSH, Cat Number: E-EL-M0511) enzyme-linked immunosorbent assay (ELISA) kits were purchased from Elabscience (Wuhan, China).

A total of 40 three-week-old female C57BL/6J mice were purchased from Gempharmatech Co., Ltd. (Jiangsu, China). The animal study was conducted according to the regulations of the National Institutes of Health (NIH), USA. The experimental protocol was approved by the Ethics Committee of Hainan Medical University (Approval No. SYXK-2017 0013). All mice were housed in standard cages under controlled environmental conditions (25°C, appropriate humidity, and a 12-h light/12-h dark cycle), with free access to food and water. Daily health monitoring was performed throughout the experimental period. After three days of acclimation, the mice were randomly divided into two groups: the normal group (n=8) and the PCOS model group (n=32). The normal group was fed a basic diet, whereas the model group was fed an HFD. From day 8 to day 28 (weeks 2 to 4), the model group received daily injections of DHEA (0.6 mg/kg/d) in combination with the HFD to induce PCOS. Meanwhile, the normal group continued on the basic diet and received vegetable oil injections as a control.

In the week 5, body weight, estrous cycle and oral glucose tolerance test (OGTT) were assessed to confirm successful establishment of the PCOS model. Subsequently, the PCOS mice were randomly divided into four groups (n=8): PCOS model group, low-dose laurolitsine group (LL-L group, 50 mg/kg/d), high-dose laurolitsine group (LL-H group, 100 mg/kg/d), and metformin group (Met group, 200 mg/kg/d) (19). The normal group continued to receive a basic diet, while the PCOS model group, LL-L group, LL-H group, and Met group continued with HFD. After the week 9, all animals were fasted for 8 h, blood samples were collected. Serum was centrifuged for analysis, and tissue samples were obtained for further experiments.

Vaginal smears were used to assess the estrous cycle of the mice (20). The vagina was flushed two to three times with 25 mL of sterile saline, and the vaginal fluid was then collected and placed on a slide. After air-drying at room temperature, slides were stained with methylene blue (MB). Vaginal cytology was used to assess the estrous cycle’s stage.

After 8 h of fasting, the blood glucose levels of the mice were measured, and the oral glucose dose was set at 2 g/kg body weight. Blood glucose levels were measured by drawing blood from the tail vein at 15, 60, 90, and 120 min after glucose administration (21).

After 8 h of fasting, the body weight of the mice was measured, and blood glucose levels were tested before the injection. The mice received an intraperitoneal injection of insulin at a dose of 0.5 U/kg body weight. Blood glucose levels were collected from the tail at 30, 60, 90, and 120 min after the injection.

Blood samples were collected from the mouse’s eyeball and allowed to stand at room temperature for 4 h. Subsequently, the samples were centrifuged at 3500 rpm for 15 min at 4°C, and the serum was transferred to new centrifuge tubes for further analysis. Serum TG, CHO, LDL-c, HDL-c, ALT, and AST levels were measured according to the instructions of commercial assay kits. Additionally, serum levels of T, E2, LH, and FSH were determined using ELISA kits.

Fresh tissues were fixed in 4% formaldehyde at room temperature for 24 h. After fixation, the tissues were dehydrated through a graded ethanol series. Subsequently, the tissues were infiltrated with paraffin at 65°C, embedded, and sectioned into 4 μm-thick slices. The sections were floated in a 40°C water bath, placed on glass slides, and baked in a 60°C oven for adherence. Subsequently, the sections were deparaffinized in xylene three times and rehydrated through 100% and 95% ethanol for 5 min each. The sections were stained with hematoxylin for 5 min, rinsed for 10 min, differentiated in 0.7% hydrochloric acid ethanol for 10 s, and returned to blue with tap water for 10 min. The sections were then stained with eosin for 5 min, dehydrated with 95% and absolute ethanol for 5 min each, cleared twice in xylene, and finally mounted with neutral resin (22).

Ovaries were collected from three mice in each of Normal, PCOS and LL (100 mg/kg) groups for analysis. Total RNA was extracted and its purity and integrity were verified using agarose gel electrophoresis, a Qubit 4.0 fluorometer/MD, and a Q sep 400 bioanalyzer. A cDNA library was constructed, and its quality was evaluated using an Agilent Bioanalyzer 2100 system. The library was sequenced on the Illumina NovaSeq platform, generating 150 bp paired-end reads. Differentially expressed genes (DEGs) were identified and analyzed using SangerBOX (http://vip.sangerbox.com↗), along with Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses.

Fresh mouse fecal samples were collected, and total DNA was extracted using a commercial kit from Wetware (Wuhan, China). The V3-V4 region of 16S rDNA as then amplified using specific primers. Upstream primers with barcode tags were employed to distinguish different samples within the same library. The TruSeq Nano DNA LT sample preparation kit was used for library construction, and sequencing was performed on the Illumina MiSeq PE300 platform. Data quality control and analysis were conducted using the QIIME2 platform.

Statistical analyses were conducted using GraphPad Prism 8 (GraphPad Software Inc., La Jolla, CA). The normality of data distribution was assessed using the Shapiro-Wilk test and the Kolmogorov-Smirnov (K-S) test. All data included in the statistical analyses were confirmed to follow a normal distribution. For data that followed a normal distribution, one-way analysis of variance (ANOVA) was utilized to compare differences between groups. Results were presented as mean ± standard deviation (± SD). Differences were considered statistically significant at P< 0.05.

To establish a PCOS mouse model, we used a combination of DHEA plus HFD (Figure 1A) (23). Compared with the normal group, the model mice exhibited significantly increased body weight and impaired glucose tolerance, as demonstrated by the OGTT results (Figures 1B, C), indicating severe glucose intolerance. Additionally, the estrous cycle of the DHEA plus HFD-induced mice were disrupted (Figures 1D, E). These results collectively confirm the successful establishment of the PCOS mouse model.

To evaluate the effects of LL on PCOS, we administered LL intervention in a PCOS mouse model induced by DHEA combined with HFD. Compared with the normal group, the body weight, liver weight, and gonadal fat weight of PCOS mice were significantly increased (Figures 2A–C). However, after oral administration of LL (100 mg/kg/d), the body weight, liver weight, and gonadal fat weight of PCOS mice were significantly reduced (Figures 2A–C). Furthermore, the serum TG, CHO, and LDL-c levels were significantly elevated in PCOS mice (Figures 2D–G), but LL intervention significantly reduced these lipid parameters, indicating a notable improvement in lipid metabolism. Serum ALT and AST levels showed no significant changes (Figures 2H, I), and both LL-L and LL-H did not exhibit hepatotoxicity (Figure 2J), demonstrating a favorable safety profile. Adipose tissue is considered a key factor in the etiology of PCOS, particularly in metabolic abnormalities associated with overweight and obesity (24). H&E staining revealed marked accumulation of gonadal fat in PCOS mice, accompanied by more numerous and larger lipid droplets (Figure 2J). LL intervention significantly reduced the size of lipid droplets. Furthermore, PCOS mice exhibited significantly elevated fasting blood glucose (FBG), impaired glucose tolerance, and increased insulin resistance (Figures 2K–O). After LL treatment, the FBG levels, glucose tolerance, and insulin resistance were significantly improved in PCOS mice. Collectively, these findings indicate that LL significantly alleviates lipid metabolism disorders, impaired glucose tolerance, and insulin resistance induced by DHEA plus HFD in PCOS mice.

Elevated androgen levels, decreased ovulation, and polycystic ovarian morphologic alterations are major characteristics of PCOS in patients, which are similarly observed in the PCOS mice model. We evaluated serum FSH, LH, E2, and T levels to assess the effects of LL on hormonal regulation. Compared with the normal group, serum T and LH levels were significantly increased, while FSH levels were significantly decreased in PCOS mice. No significant change was observed in E2 levels (Figures 3A–D). Following LL intervention, serum T and LH levels in PCOS mice were significantly decreased, similar to the effects observed Met group (Figure 3A, C). In addition, both LL and Met dramatically reduced the LH/FSH ratio in PCOS mice (Figure 3E). These results indicate that LL effectively alleviates hormonal imbalance, thereby improving PCOS-like features in mice.

Ovarian dysfunction is a key characteristic of PCOS. To confirm the ameliorative effects of LL on folliculogenesis and ovulation, we assessed ovarian morphology in PCOS mice. Compared to normal mice, PCOS mice showed a reduced number of corpora lutea and a significantly increased number of follicles, accompanied by vacuolization and structural disorganization (Figures 4A–C). After LL intervention, the formation of corpora lutea increased, and the number of follicles with vacuolated cystic dilatation decreased in PCOS mice (Figures 4A–C). However, LL did not significantly affect ovary weight or uterus weight in PCOS mice (Figures 4D, E). These results suggest that LL may improve ovarian function in PCOS mice.

To further explore the potential mechanisms of LL’s effects in PCOS mice, we conducted RNA-seq analysis on ovary tissues from PCOS mice with or without LL treatment. Principal component analysis (PCA) showed a clear separation between the PCOS group and the LL (100mg/Kg/d) group (Figure 5A). The volcano plot revealed 53 upregulated and 25 downregulated genes in the LL group compared with the PCOS group (Figure 5B). A heatmap illustrated differential expression of these genes (Figure 5C). Among them, we observed significant modulation of Lhcgr (luteinizing hormone/choriogonadotropin receptor) and Tnc (tenascin-c), both of which are involved in critical ovarian function. Lhcgr is essential for LH-mediated follicular maturation and ovulation; its downregulation in PCOS may lead to impaired follicle maturation and anovulation, while LL treatment restored Lhcgr expression, suggesting a potential mechanism for LL in improving ovarian function (25, 26). In contrast, Tnc known to mediate tissue remodeling and inflammation, was upregulated in PCOS mice, reflecting excessive extracellular matrix deposition and granulosa cell apoptosis (27). LL treatment downregulated Tnc, suggesting that LL may alleviate ovarian inflammation and fibrosis, promoting healthier follicular development. GO analysis revealed that LL treatment primarily impacted biological processes such as response to organic substances, hormones and endogenous stimuli (Figures 5D–F). These findings are consistent with the potential therapeutic effects of LL on ovarian function and metabolic health in PCOS. Additionally, KEGG pathway enrichment analysis indicated that LL primarily affected pathways associated with ovarian steroidogenesis, apoptosis signaling, ether lipid metabolism, and arginine and proline metabolism (Figure 5G). These pathways are critical for maintaining ovarian function and metabolic homeostasis, further supporting LL’s role in improving ovarian function and metabolic regulation. Further analysis through Gene Set Enrichment Analysis (GSEA) demonstrated that oxidative phosphorylation, fatty acid metabolism, mitochondrial translation, and gene groups related to translation were significantly downregulated in the ovary tissues of PCOS mice, while LL reversed these genomic changes (Figures 5H–K). Notably, oxidative stress has been reported to play a critical role in hyperandrogenemia-induced organ dysfunction, including kidney injury, further supporting the involvement of oxidative stress pathways in PCOS pathogenesis (28). These results suggest that LL may improve PCOS by modulating key metabolic processes, including hormone response, fatty acid metabolism, and lipid metabolism in the ovary. The gene expression patterns after LL treatment are closely associated with the phenotypic improvements in ovarian function and metabolic regulation. Specifically, the upregulation of Lhcgr and downregulation of Tnc indicate that LL restores key signaling pathways involved in folliculogenesis and reduces ovarian inflammation. GO, KEGG, and GSEA analyses further support that LL exerts therapeutic effects by modulating pathways such as steroidogenesis, oxidative stress, and mitochondrial function, thereby improving ovarian function and alleviating metabolic dysregulation in PCOS mice.

The gut microbiota has been recognized as an essential mediator of metabolic diseases including PCOS, and modulation gut microbiota is considered an effective strategy for the prevention and treatment of these related diseases (29). α-diversity reflects within-sample richness and evenness, while β-diversity describes between-sample compositional differences. Through 16S rDNA sequencing analysis, principal co-ordinates analysis (PCoA) and unweighted pair-group method with arithmetic mean (UPGMA) showed that both LL-L and LL-H interventions distinctly shifted the gut microbiota composition away from that of the PCOS group (Figures 6A, B). However, α-diversity analyses (Shannon, Simpson, Chao1, and ACE) did not reveal significant differences in species richness and diversity among the groups (Figures 6C–F), indicating that LL intervention primarily affects microbial composition rather than overall diversity.

At the phylum level, Proteobacteria (or Pseudomonadota) was significantly elevated in PCOS mice. This phylum is closely associated with microbial dysbiosis and low-grade inflammation driven by endotoxins (30), a chronic inflammatory state that has been identified as a key contributor to insulin resistance and hyperandrogenism in PCOS. Notably, LL-L and LL-H significantly reduced the abundance of Proteobacteria, which may help alleviate gut-derived inflammation and systemic metabolic stress (Figures 6G, H).

At the species level, 29 species were found to be significantly altered between the PCOS group and LL-treated groups (Figure 6I). Specifically, Faecalibaculum rodentium, Lactobacillus johnsonii, Bacteroides vulgatus, and Sphingomonas echinoides were markedly enriched in PCOS mice. These species have been linked to low-grade intestinal inflammation, gut microbiota dysbiosis, and disrupted estrogen metabolism, all of which are central to PCOS pathogenesis (31, 32). Notably, Lactobacillus johnsonii is known for its high β-glucuronidase activity, which may affect estrogen recycling and exacerbate hormonal imbalance (33). LL treatment markedly decreased the abundance of these species, thereby helping to restore gut microbial balance. Meanwhile, LL intervention, particularly LL-H promoted the enrichment of beneficial species such as Akkermansia muciniphila, which has been shown to be a beneficial bacterium associated with intestinal mucosal barrier integrity and enhanced anti-inflammatory responses. It plays a positive role in improving obesity-related metabolic syndromes and regulating glucose and lipid metabolism (34). These findings suggest that LL modulates the gut microbiota in a dose-dependent manner, targeting key species involved in metabolic regulation and hormonal balance. This remodeling of the gut microbial community may underlie the potential therapeutic effects of LL in alleviating PCOS-related phenotypes.

Through T-test and LEfSe analysis, we further investigated the changesn gut microbiota following LL intervention. In the PCOS group, Lactobacillus johnsonii was dominant species. In contrast, Lachnospiraceae bacterium 10_1 dominated in the LL-L group (Figures 7A, B, E). In the LL-H group, Akkermansia muciniphila and Burkholderiales bacterium YL45 became the dominant species, while Lactobacillus johnsonii remained the dominant species in the PCOS group (Figures 7C, D, F). Notably, Lactobacillus johnsonii, enriched in the PCOS group, has been linked to intestinal inflammation and hormonal disruption, potentially contributing to the pathogenesis of PCOS. After LL intervention, the abundance of Lactobacillus johnsonii decreased, while the beneficial bacterium Akkermansia muciniphila increased, which are associated with anti-inflammatory effects and improved metabolic health. These findings further support the gut microbiota remodeling effect of LL intervention.

OTU (Operational Taxonomic Unit) is a term commonly used in microbial ecology to classify groups of closely related individuals based on DNA sequence similarity, typically in 16S rRNA gene sequencing analysis. Further analysis of OTU changes using DESeq2 revealed that, compared to the normal group, 13 bacterial OTUs were upregulated and 10 were downregulated in the PCOS group (Figures 8A, D). In the LL-L group, 18 bacterial OTUs were upregulated and 32 were downregulated compared to the PCOS group (Figures 8B, E), while in the LL-H group, 29 OTUs were upregulated and 23 were downregulated (Figures 8C, F). Venn diagram analysis further demonstrated that LL significantly regulated OTUs associated with PCOS. Specifically, LL-L treatment significantly downregulated OTUs such as OTU_25, OTU_111, OTU_140, and OTU_686, while LL-H treatment reduced OTUs including OTU_25, OTU_111, OTU_140, OTU_87, and OTU_69, while upregulating OTU_51 (Figures 8G–J). These findings indicated that LL significantly altered gut microbiota at the OTUs levels, educing PCOS-related pathogenic taxa and enriching beneficial microbial populations, thereby contributing to the improvement of gut dysbiosis in PCOS mice. LL treatment may specifically modulate the gut microbiota composition in the PCOS model, particularly by increasing the abundance of Akkermansia muciniphila, while reducing the relative abundance of Proteobacteria and Lactobacillus johnsonii. These changes may help improve the host’s metabolic function, inflammatory responses, and endocrine balance, thereby offering a potential microbiota-targeted intervention strategy for the treatment of PCOS.