Dried Coriandrum sativum seeds were purchased from TongRenTang Chinese Medicine (Beijing, China). A total of 100 g of coarsely ground seeds was extracted twice with boiling water. The first extraction was performed with 1000 mL of water for 1 hour, followed by filtration. The residue was then re-extracted with 800 mL of water for 45 minutes. Both filtrates were combined and concentrated under reduced pressure to a final volume of 100 mL, and stored at 4 °C.

For sample preparation, 1 mL of the concentrated extract was mixed with 1 mL of ethyl acetate and vortexed for 60 seconds using an XW-80A vortex mixer (QiLinBeiEr, China). After centrifugation at 12,000 rpm for 5 minutes at 4 °C (H1750R centrifuge, XiangYi, China), the upper organic layer was filtered through a 0.22 μm membrane and evaporated to dryness under nitrogen using an MD200–1 nitrogen evaporator (AoSheng, China). The residue was reconstituted in 200 μL ethyl acetate and stored at −20 °C for GC-MS analysis.

GC-MS analysis was carried out using a GCMS-QP2010SE instrument (Shimadzu, Japan) equipped with an EI ion source and a DB-5MS capillary column (30 m × 0.25 mm, 0.25 μm). The injection volume was 1 μL in splitless mode. The injector temperature was set at 250 °C. The column oven temperature was programmed as follows: 70 °C (held for 2 min), increased to 180 °C at 2 °C/min (held for 5 min), then to 230 °C at 10 °C/min (held for 5 min). The total run time was 72 minutes. Helium was used as the carrier gas at a constant flow of 1.0 mL/min.

Mass spectrometric detection was performed in full-scan mode (m/z 40–350) with the following settings: EI energy at 70 eV, ion source temperature 200 °C, interface temperature 230 °C, quadrupole temperature 150 °C, scan speed 1111 amu/s, and a solvent delay of 6 minutes.

Compounds were identified by comparing their mass spectra with entries in the NIST 17 library, with only those showing a match factor ≥90% being retained. When multiple matches were available, the compound with the highest similarity was selected. Relative abundances were calculated using peak area normalization.

A total of thirty-two male KM mice (5 weeks old, 32–34 g) were procured from SPF (Beijing) Biotechnology Co., Ltd. (Beijing, China). Animals were housed under specific pathogen-free (SPF) conditions at Tianjin University of Traditional Chinese Medicine, maintained at a controlled temperature (24 ± 1 °C), relative humidity of 50%–70%, and a 12 h light/dark cycle. All experimental procedures were conducted in accordance with the ARRIVE guidelines and approved by the Animal Experiment Ethics Commit-tee of Tianjin University of Traditional Chinese Medicine (Approval No. TCM-LAEC2024162z1030).

Following acclimatization, mice were randomly assigned to four groups (n = 7 per group): Normal Diet (ND), High-Sugar High-Fat Diet (HSFD), low-dose aqueous Coriandrum sativum seed extract (YS-L, 1.0 g/kg), and high-dose aqueous Coriandrum sativum seed extract (YS-H, 2.0 g/kg). Mice in the ND group were provided with standard chow and distilled water, while those in the HSFD, YS-L, and YS-H groups received a high-sugar high-fat diet comprising 67% standard chow, 10% lard, 20% sucrose, 2.5% cholesterol, and 0.5% sodium cholate, along with 20% fructose solution ad libitum (SPF Bio-technology Co., Ltd., Beijing, China). In addition, mice in the YS-L and YS-H groups were administered low- and high-dose YS aqueous extract via oral gavage daily for 56 consecutive days; control groups (ND and HSFD) received an equivalent volume of saline.

Upon completion of the intervention, all mice were anesthetized with an intraperitoneal injection of pentobarbital sodium at a dosage of 50 mg/kg. Blood samples were collected via retro-orbital bleeding, after which the animals were euthanized by cervical dislocation. Following cardiac perfusion with phosphate-buffered saline (PBS, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China), liver, pancreas, and colon tissues were harvested. Portions of each tissue were fixed in 4% paraformaldehyde for histological analysis, while remaining samples were snap-frozen in liquid nitrogen and stored at −80 °C for subsequent biochemical assays. Additionally, cecal contents were collected and stored at −80 °C for subsequent microbial genomic analysis via full-length 16S rRNA sequencing.

On day 56 of the intervention period, fasting blood glucose (FBG) was measured using tail tip blood samples collected with a calibrated glucometer (ACCU-CHEK Instant S, Roche Diagnostics, Germany). For serum preparation, whole blood was obtained via retro-orbital venous plexus puncture. Samples were allowed to clot at room temperature for 2 hours, then centrifuged at 4500 rpm for 10 minutes at 4 °C. The resulting serum was collected for lipid profiling. Serum levels of total cholesterol (TC, Cat# A111-1-1), triglycerides (TG, Cat# A110-1-1), low-density lipoprotein cholesterol (LDL-C, Cat# A113-1-1), and high-density lipoprotein cholesterol (HDL-C, Cat# A112-1-1) were determined using commercial enzymatic colorimetric assay kits (Nan-jing Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer’s instructions.

Liver tissues were homogenized in ice-cold saline (1:9, w/v), and the homogenates were centrifuged at 2000 rpm for 10 minutes at 4°C. The resulting supernatants were used for quantification of hepatic biochemical indices, including TC, TG, alkaline phosphatase (AKP, Cat# A059-2-1), alanine aminotransferase (ALT, Cat# C009-2-1), aspartate aminotransferase (AST, Cat# C010-2-1), superoxide dismutase (SOD, Cat# A001-1-1), malondialdehyde (MDA, Cat# A003-1-1), and glycogen (Cat# A043-1-1), using corresponding assay kits (Nanjing Jiancheng Bioengineering Institute).

Serum insulin concentrations were determined using a mouse insulin ELISA kit (Jiangsu Meimian Industrial Co., Ltd., Jiangsu, China) following the manufacturer’s instructions. Briefly, serum samples were added to 96-well microplates pre-coated with purified mouse insulin antibodies and incubated with horseradish peroxidase (HRP)-conjugated detection antibodies. After thorough washing to remove unbound components, chromogenic substrate solution was applied, and the colorimetric reaction was carried out at 37°C. The reaction was stopped with stop solution, and absorbance was measured at 450 nm using a microplate reader. Insulin concentrations were calculated based on a standard curve generated from known standards. All as-says were conducted in duplicate to ensure accuracy and reproducibility.

Homeostasis model assessment of insulin resistance (HOMA-IR) was calculated using the following formula:

HOMA-IR = (FBG (mmol/L) × FINS (mIU/L))/22.5.

where FBG represents fasting blood glucose and FINS denotes fasting serum insulin levels. This index was used to evaluate systemic insulin resistance in each experimental group.

Total RNA was extracted from approximately 50 mg of mouse liver tissue using the TransZol Up Plus RNA extraction kit (TransGen Biotech, Beijing, China), strictly adhering to the manufacturer’s guidelines. Tissue homogenization was performed thoroughly in pre-cooled reagent, and the resultant homogenates underwent standard chloroform phase separation, isopropanol precipitation, and ethanol washes. Purified RNA pellets were dissolved in RNase-free water and subsequently assessed for con-centration using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA).

For quantitative real-time PCR (RT-qPCR), RNA samples with optical density ratios (260/280 nm and 260/230 nm) indicative of high purity (≥1.9) were selected. Re-verse transcription reactions were performed using the Quantscript RT-PCR kit (Tiangen Biotech, Beijing, China), employing random primers and oligo (dT) primers according to standard protocols provided by the manufacturer. The resulting cDNA samples were diluted appropriately for subsequent amplification.

Gene expression analysis was carried out using specific primer pairs synthesized by Sangon Biotech Co., Ltd. (Shanghai, China), detailed sequences of which are summarized in Supplementary Table A1. Real-time PCR amplifications were conducted in triplicate on an Applied Biosystems quantitative PCR system (Bio-Gener Technology Co., Ltd, Hangzhou, China). Reaction mixtures comprised SYBR Green fluorescent dye, template cDNA, forward and reverse primers, and RNase-free water to a final volume as per the manufacturer’s instructions. PCR cycling parameters were as follows: initial denaturation at 95 °C for 5 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 30 s, with melting curves generated to confirm the specificity of amplified products.

Relative gene expression levels were quantified using the comparative Ct method (2−ΔΔCt), and data were normalized against β-actin expression, selected as the stable in-ternal control gene.

In this experiment, total cellular proteins were extracted using a RIPA lysis buffer (supplemented with protease inhibitors) and quantified by the BCA method to ensure equal loading (approximately 20-30 µg per lane). After denaturation in a 5× SDS loading buffer, the samples were separated by SDS-PAGE on a 12% gel, followed by transfer onto a PVDF membrane pre-activated in 95% methanol. The membrane was blocked with 5% non-fat milk in TBST for 1 hour, then incubated overnight at 4 °C with primary antibodies specific to PPARα, IL-6, NF-κB, and LDL-R. Subsequently, HRP-conjugated secondary antibodies were applied for 1 hour at room temperature, and protein bands were visualized using an ECL chemiluminescent substrate, followed by image acquisition with a gel imaging system for quantitative analysis.

Liver, pancreas, and colon tissues were harvested from experimental mice and immediately fixed in 4% paraformaldehyde solution (Servicebio, Wuhan, China). For paraffin embedding, tissues underwent routine dehydration in graded ethanol, clearing with environmentally-friendly dewaxing solution, embedding, and subsequent sectioning at 4 µm thickness using a Leica RM2016 microtome (Leica, China). Liver samples designated for lipid analysis were embedded in optimal cutting temperature (OCT) compound, snap-frozen, and sectioned at 8 µm thickness on a Cryostar NX50 cryostat (Thermo Fisher, USA).

For hematoxylin-eosin (HE) staining, paraffin sections were dewaxed, rehydrated, stained with hematoxylin for 5 min, differentiated, blued in ammonia solution, and counterstained briefly with eosin for 15 s. Sections were dehydrated, cleared, and permanently mounted with neutral resin. Images were captured using a Nikon Eclipse E100 microscope equipped with a DS-U3 imaging system (Nikon, Japan). Nuclei were visualized as blue and cytoplasm as varying shades of pink.

For Oil Red O staining, liver cryosections were fixed with 4% paraformaldehyde for 15 min, followed by staining in freshly prepared Oil Red O working solution for 10 min at 4 °C. Differentiation was briefly conducted in 60% isopropanol, and nuclei were counterstained with hematoxylin. Slides were mounted with glycerol gelatin (Service-bio). Lipid droplets were visualized as bright orange-red against blue nuclei. Pancreatic and colonic tissues underwent identical HE staining procedures. All histopathological evaluations were performed independently by blinded observers.

Fecal microbial DNA was isolated using established protocols as previously outlined (48). The entire 16S rRNA gene spanning regions V1–V9 was amplified by PCR with primers 27F (5’-AGRGTTYGATYMTGGCTCAG-3’) and 1492R (5’-RGYTACCTTGTTACGACTT-3’). Following amplification, PCR products were evaluated on 2% agarose gels, purified according to the Pacific Biosciences guidelines, and subsequently utilized to generate SMRTbell libraries through blunt-end ligation. Amplicon sequencing was performed commercially by Shanghai Biozeron Biotechnology Co., Ltd. (Shanghai, China).

Perform quality control using FastQC. The raw sequences were processed using QIIME2 (version 2023.05, https://docs.qiime2.org/). The further analysis, including noise reduction, correcting marginal sequence errors, removing chimeric sequences, removing singletons, joining paired-end reads, and dereplication, was achieved using DADA2 plugin. ASV (amplicon sequence variants) were defined based on 100% sequence similarity between the sequences. The species annotation results were based on the SILVA (version 138.1) (http://www.arb-silva.de/). The normalized ASV abundance data was subjected to alpha-diversity analysis, beta-diversity analysis, difference analysis and ecological network analysis using R (version 4.1.2). The Vegan package (version 2.6–6.1) was used to calculate the alpha-diversity and beta diversity. The ggplot2 package (version 3.5.0) was used to visualization.

Statistical analyses were conducted using GraphPad Prism software (version 9.0; GraphPad Software Inc., San Diego, CA, USA). Data are expressed as mean ± standard error of the mean (SEM). Group comparisons were performed using one-way analysis of variance (ANOVA), followed by Tukey’s multiple comparison test to assess inter-group differences. A two-tailed p-value < 0.05 was considered statistically significant. All statistical procedures adhered to established guidelines to ensure analytical robustness and reproducibility.

The chemical composition of the aqueous extract of Coriandrum sativum seeds was characterized by gas chromatography–mass spectrometry (GC–MS). A total of nine compounds were identified based on retention time, retention index (RI), molecular formula, and mass spectral fragmentation patterns (Figure 1; Table 1). Compound identities were further supported by similarity index (SI) values and confirmed through comparison with standard mass spectral libraries (Table 2). Representative electron ionization (EI) mass spectra of these nine compounds, including their characteristic fragment ions and chemical structures, are presented in (Supplementary Figure A1). Each spectrum displays the mass-to-charge ratio (m/z) and relative intensity of key fragment ions, providing additional structural confirmation for compound identification.

Among the identified constituents, monoterpenoid alcohols included 2,6-dimethylocta-3,7-diene-2,6-diol (Compound 1, RI 1197) and its structural isomer (2E)-2,6-dimethylocta-2,7-diene-1,6-diol (Compound 4, RI 1325), with relative abundances of 1.35% and 2.58%, respectively. These compounds shared major fragment ions at m/z 43 and 67, attributed to cleavage of the terminal hydroxylated isoprenoid chains. Their spectral profiles were distinguishable by the presence of additional peaks such as m/z 93 and 152 in Compound 4.

Phenolic and unsaturated alcohol derivatives were represented by 4-ethenyl-2-methoxyphenol (Compound 2, RI 1293) and 3-methylhepta-1,6-dien-3-ol (Compound 3, RI 888). Compound 2 was the most abundant component (11.62%), showing diagnostic ions at m/z 150, 107, and 77, consistent with fragmentation of a vinyl-substituted guaiacol core. Compound 3 presented a simpler fragmentation pat-tern dominated by m/z 71 and 43.

Compounds 5 to 7 included dodecanal (RI 1402), hexadecane (RI 1612), and 2,6,10,14-tetramethylhexadecane (RI 1753), with relative abundances ranging from 0.51% to 1.47%. Dodecanal produced characteristic aldehyde fragments at m/z 57 and 82, while the two saturated hydrocarbons showed typical alkyl ion series, including m/z 57, 71, and 99.

Long-chain fatty acids were identified as hexadecanoic acid (Compound 8, RI 1968; 6.13%) and octadecanoic acid (Compound 9, RI 2167; 4.28%). Both compounds exhibited typical McLafferty rearrangement ions at m/z 60 and 73, in addition to sequential β-cleavage fragments such as m/z 129 and 143. Their spectral data were in agreement with reference spectra of saturated fatty acids.

The complete fragmentation ion distribution and intensity values for all compounds are provided in Supplementary Table A2. The identification of structurally diverse compounds, including monoterpenoids, phenolics, aldehydes, alkanes, and fatty acids, establishes the chemical basis for further pharmacological evaluation of the extract.

Following phytochemical identification, we evaluated the physiological impact of Coriandrum sativum seed extract in a murine model of metabolic dysfunction. As illustrated in Figure 2A, mice receiving HSFD exhibited a continuous elevation in body weight across the 8-week period, markedly exceeding that of the ND group. In contrast, mice treated with YS, particularly at the higher dose, showed a gradual attenuation of weight gain beginning at week 3, with a clearer divergence observed from week 5 onward.

Final body weights recorded at week 8 (Figure 2B) reinforced the longitudinal findings. HSFD-fed mice displayed significantly higher body weights compared to ND animals (p < 0.01), while both YS-L and YS-H groups demonstrated weight reductions. The effect was more pronounced in the YS-H group (p < 0.01 vs. HSFD), suggesting a potential dose-dependent response.

Analysis of organ indices (Figure 2C) revealed that HSFD led to liver enlargement, evidenced by a significant increase in liver index compared to ND (p < 0.05). This elevation was substantially diminished by YS administration, with YS-H reducing liver index values to near-baseline levels (p < 0.001 vs. HSFD). While spleen indices remained largely unaltered, the renal index was moderately elevated in the HSFD group and significantly reduced in YS-H-treated mice (p < 0.05), indicating a potential normalization of renal mass.

As shown in Figure 2D, HSFD significantly elevated epididymal fat mass to ND controls (p < 0.01), consistent with visceral adipose accumulation. Both YS-treated groups exhibited reductions in fat mass, with YS-H achieving a statistically significant decrease compared to HSFD alone (p < 0.05). These reductions paralleled the observed suppression of body weight gain.

Together, these findings establish a physiological basis for the metabolic effects of the extract, highlighting its capacity to mitigate HSFD-induced weight gain, hepatomegaly, renal index elevation, and visceral fat expansion in a dose-responsive manner.

Given the extract’s effectiveness in reducing HSFD-induced weight gain and adiposity, we next evaluated its role in glycemic regulation and pancreatic integrity—two closely linked endpoints of metabolic dysfunction (Figures 3A–E).

Consistent with the obesity-related metabolic burden observed in the previous section, fasting blood glucose levels were markedly elevated in HSFD-fed mice relative to the normal diet group (p < 0.001, Figure 3A). Both low and high doses of Coriandrum sativum seed extract reduced glucose levels, with the high-dose group restoring normoglycemia to a greater extent (p < 0.001 vs. HSFD). Parallel changes were observed in circulating insulin concentrations (Figure 3B). HSFD mice exhibited marked hyperinsulinemia (p < 0.001), while extract treatment attenuated this elevation in a dose-responsive manner (p < 0.01 and p < 0.001 for low and high doses, respectively), suggesting improved insulin resistance.

The homeostasis model assessment of insulin resistance (HOMA-IR) further sup-ported these findings (Figure 3C). HSFD feeding led to a significant increase in HOMA-IR values (p < 0.001), reflecting systemic insulin resistance. Treatment with Coriandrum sativum seed extract reduced HOMA-IR scores, with the high-dose group demonstrating a significant improvement (p < 0.05 vs. HSFD), indicating partial restoration of insulin responsiveness.

In line with these improvements, hepatic glycogen stores—often depleted under insulin-resistant conditions—were significantly reduced in HSFD-fed mice (p < 0.001 vs. ND, Figure 3D). Both extract-treated groups exhibited a restoration of glycogen content, with the high-dose group achieving greater recovery (p < 0.05 vs. HSFD; p < 0.001 vs. HSFD), suggesting enhanced hepatic glucose utilization and storage.

Histological assessment of pancreatic tissue provided morphological evidence supporting the observed metabolic improvements (Figure 3E). In ND-fed mice, hematoxylin and eosin staining revealed well-defined islet architecture with intact cellular boundaries and uniform acinar distribution. In contrast, HSFD-fed mice exhibited marked islet distortion, including irregular borders, vacuolar degeneration, and disorganized exocrine lobules, indicative of diet-induced pancreatic injury. Notably, both extract-treated groups demonstrated attenuation of these pathological features. In the high-dose group, the islets retained their structural integrity with reduced cellular swelling and more regular contours, while the surrounding acinar cells showed improved alignment and nuclear morphology, approaching the histological presentation of the ND group. Collectively, these data reveal that Coriandrum sativum seed extract not only reduces hyperglycemia and insulin resistance but also mitigates the downstream hepatic and pancreatic consequences of prolonged HSFD exposure. This functional transition from systemic indicators to organ-level protection supports its broader therapeutic potential in metabolic regulation.

Following improvements in glucose homeostasis and pancreatic morphology, we next examined the effects of Coriandrum sativum seed extract on hepatic lipid metabolism, liver enzyme activities, oxidative stress markers, and histopathological features in HSFD-fed mice (Figures 4A–K).

Mice subjected to HSFD exhibited significant dyslipidemia, with elevated serum total cholesterol (TC), triglycerides (TG), and low-density lipoprotein cholesterol (LDL-C) levels compared to the normal diet group (p < 0.001, Figures 4A–C). Both extract-treated groups showed dose-dependent reductions in these lipid markers, with the high-dose group achieving the most pronounced effects (TC: p < 0.001; TG: p < 0.001; LDL-C: p < 0.001 vs. HSFD), clearly indicating effective systemic lipid-lowering activity.

Consistent with serum profiles, hepatic lipid accumulation was also aggravated in HSFD-fed mice, as evidenced by increased liver TC and TG levels (p < 0.001 vs. ND, Figures 4D, E). Treatment with Coriandrum sativum seed extract significantly attenuated hepatic lipid content in a dose-dependent manner. The high-dose group demonstrated significantly greater reductions in hepatic TC and TG compared to the low-dose group (p < 0.01), approaching near-baseline values.

Liver enzyme activities were further assessed to evaluate hepatic injury. Serum levels of alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase (AKP) were all significantly elevated in HSFD-fed mice (Figures 4F–H), indicating hepatocellular damage. Both low- and high-dose extract treatments ameliorated these elevations, with the high-dose group showing significantly greater reductions in ALT (p < 0.05), AST (p < 0.01), and AKP (p < 0.001 vs. HSFD), demonstrating a clear dose-dependent protection against diet-induced hepatic injury.

To further investigate hepatic oxidative status, malondialdehyde (MDA) and superoxide dismutase (SOD) were measured (Figures 4I, J). HSFD feeding significantly increased hepatic MDA levels (p < 0.05) while decreasing SOD activity (p < 0.05), indicating enhanced lipid peroxidation and weakened antioxidant capacity. Extract administration partially reversed these effects, with the high-dose group demonstrating significantly greater reductions in MDA and improvements in SOD levels compared to the low-dose group (p < 0.05), supporting a dose-dependent antioxidative role of the extract.

Histological analysis of liver tissues further substantiated the biochemical findings. In HE-stained sections (Figure 4K, top row), livers from ND-fed mice displayed normal hepatic architecture with clear hepatic cords, intact nuclei, and regular sinusoidal structure. HSFD-fed mice, however, exhibited notable pathological alterations, including hepatocellular ballooning, cytoplasmic rarefaction, and sinusoidal congestion, indicative of steatosis and early inflammatory response. Treatment with Coriandrum sativum seed extract ameliorated these histopathological changes in a clear dose-dependent manner. Livers from the high-dose group demonstrated significantly greater restoration of hepatic cord alignment, reduction in cytoplasmic vacuolation, and normalization of sinusoidal spacing compared to the low-dose group.

Oil Red O staining (Figure 4K, bottom row) revealed minimal lipid deposition in ND livers, whereas HSFD induced prominent macrovesicular steatosis, characterized by widespread lipid droplet accumulation within hepatocyte cytoplasm. Extract treatment markedly reduced lipid deposition, with the high-dose group showing significantly greater decreases in intracellular lipid accumulation and improved lobular architecture compared to the low-dose group, clearly demonstrating a dose-dependent capacity to mitigate hepatic steatosis. These results demonstrate that Coriandrum sativum seed extract effectively mitigates hepatic lipid accumulation, reduces biochemical markers of liver injury, enhances antioxidant defense, and preserves hepatic histomorphology under HSFD challenge, further supporting its multi-targeted hepatoprotective potential.

Given the observed improvements in hepatic lipid accumulation, liver enzyme activities, and histopathological features, hepatic gene expression profiling was conducted to further characterize the molecular responses induced by Coriandrum sativum seed extract. Quantitative RT-PCR was performed on liver tissues from the high-dose group (2.0 g/kg), which consistently showed superior biochemical and morphological improvements in previous analyses.

High-sugar, high-fat feeding markedly disrupted the expression of lipid metabolic regulators. Hepatic mRNA levels of FAS, a central enzyme in de novo lipogenesis, were significantly upregulated in HSFD-fed mice compared to the ND group (p < 0.01; Figure 5A). Similarly, transcript levels of PPARα, which governs lipid oxidation, and LDLR, involved in cholesterol clearance, were both elevated under HSFD conditions (p < 0.01; Figures 5C, D), reflecting a shift in hepatic lipid handling toward a pathophysiological state. At the same time, markers of hepatic inflammation were markedly induced. The expression of NF-κB, a transcription factor activated in response to metabolic stress, was significantly increased (p < 0.01; Figure 5B), alongside elevated expression of IL-6, a downstream pro-inflammatory cytokine (p < 0.01; Figure 5E).

Intriguingly, Western blot analysis revealed a distinct regulatory pattern at the protein level: the protein expressions of both PPARα and LDLR were upregulated in the extract-treated group, a trend opposite to their mRNA changes. This divergence suggests that the extract may enhance the translational efficiency or stabilize the proteins of PPARα and LDLR, representing a sophisticated, multi-layered regulatory mechanism. For inflammatory markers, a consistent downregulatory trend was observed: both NF-κB and IL-6 were significantly downregulated at mRNA and protein levels after extract treatment, aligning with the reduced hepatic inflammatory burden (Figure 5F).

The coordinated expression profile observed in the high-dose group corresponds closely with the extract-induced improvements in hepatic lipid levels, oxidative stress, and tissue morphology. These multi-level changes reflect a complex and harmonized modulation of hepatocellular metabolic and immune pathways under dietary challenge, where the upregulation of key functional proteins may be crucial for restoring metabolic homeostasis.

Building upon the hepatic transcriptomic evidence of reduced inflammation and metabolic stress, we further evaluated the effects of Coriandrum sativum seed extract on intestinal tissue architecture, given the well-established gut–liver axis and its relevance in metabolic disease progression. Morphological assessment of hematoxylin and eosin-stained colonic sections was performed to determine whether dietary intervention could preserve mucosal structure under long-term HSFD exposure (Supplementary Figure A2).

Colon tissue from ND-fed mice exhibited normal histological features, including intact mucosal layers, tightly packed crypts, uniform epithelial cell arrangement, and well-distributed goblet cells. In contrast, HSFD-fed mice showed pronounced mucosal injury, characterized by crypt distortion, epithelial attenuation, increased submucosal spacing, and infiltration of inflammatory cells within the lamina propria—indicative of impaired intestinal barrier integrity.

Treatment with Coriandrum sativum seed extract attenuated these pathological changes in a dose-dependent manner. The low-dose group demonstrated partial restoration, with moderate improvements in crypt organization and epithelial density. Notably, the high-dose group displayed substantial morphological preservation, with nearly continuous epithelial lining, restored crypt depth, and normalization of goblet cell architecture, closely resembling the histological presentation observed in the ND group.

These structural findings indicate that Coriandrum sativum seed extract confers protective effects on colonic mucosa under metabolic challenge and provide morphological context for the subsequent investigation of gut microbiota composition.

Considering the observable dose-related metabolic benefits, gut microbiota profiling was conducted in the high-dose Coriandrum sativum seed extract group (YS-H) to evaluate its potential role in reshaping intestinal microbial ecology under high-fat high-sugar diet (HSFD) conditions. Compared to the normal diet (ND) group, the HSFD group showed a downward trend in alpha diversity indices, which was partially restored by YS-H intervention, though none of the differences reached statistical significance. (Figures 6A–D). To better understand the broader community structure, β diversity was explored using principal coordinates analysis (PCoA), non-metric multidimensional scaling (NMDS), and partial least squares discriminant analysis (PLS-DA), each based on Bray–Curtis dissimilarity matrices (Figures 6E–G). A clear separation was observed between ND and HSFD groups, highlighting the extent of dysbiosis induced by dietary overload.

Phylum-level microbial composition was assessed to determine dietary and intervention-induced shifts in dominant taxonomic groups (Figure 7A). Across all three groups, Firmicutes and Bacteroidota remained the predominant phyla. HSFD feeding resulted in a visible reduction in the relative abundance of Bacteroidota, accompanied by increased proportions of Desulfobacterota and Proteobacteria, as indicated by both column height and heatmap intensity. Verrucomicrobiota, which was undetectable in the HSFD group, was present in the ND group and reappeared in extract-treated mice. Actinobacteriota and Campylobacterota exhibited elevated levels in the HSFD group and lower abundance in YS-H mice. In the extract-treated group, the relative abundance of Desulfobacterota and Proteobacteria appeared lower than in the HSFD group, suggesting a directional shift in phylum-level structure following intervention, although Firmicutes remained dominant across all groups. At the genus level (Figure 7B), further analysis of the microbial community structure revealed significant shifts in the composition of dominant genera due to dietary intervention. Compared to the ND group, the HSFD group exhibited a marked reduction in the relative abundance of Lactobacillus and the Lachnospiraceae_NK4A136_group. Conversely, HSFD feeding resulted in a significant increase in the proportions of Faecalibaculum and Desulfovibrio, indicating specific diet-induced dysbiosis. Notably, these trends were partially reversed following intervention with Coriandrum sativum seed extract. Compared to the HSFD group, the abundance of Lactobacillus was restored, while the relative abundance of Faecalibaculum decreased. This suggests that the extract intervention helps regulate the genus-level composition of the gut microbiota, shifting it back toward a normal state.

To characterize group-specific microbial differences at higher taxonomic resolution, linear discriminant analysis effect size (LEfSe) was applied to identify discriminant taxa among the ND, HSFD, and Coriandrum sativum seed extract high-dose (2.0 g/kg) groups. The analysis was conducted at both the genus and species levels and further supported by heatmap visualization (Figures 8A–D).

At the genus level, ND-fed mice exhibited consistent enrichment of Bacteroidota-affiliated taxa, including Bacteroides, Prevotellaceae_UCG-001, Mucispirillum, Alloprevotella, and Monoglobus. These genera displayed LDA scores ranging from 3.3 to 3.9, and their relative abundance was elevated in ND samples on the genus-level heatmap. Cladogram mapping showed a coherent phylogenetic distribution within Bacteroidota branches. In contrast, the HSFD group was dominated by Muribaculaceae, Dubosiella, Helicobacter, and Ileibacterium, each with LDA scores exceeding and p-values < 0.01. These genera were distributed across diverse phyla, including Proteobacteria and Actinobacteriota, and corresponded to elevated signal intensity in HSFD samples. The YS-H group demonstrated a distinct genus-level profile enriched in Firmicutes-related taxa, such as Lactobacillus (LDA = 4.51), Desulfovibrio (LDA = 4.44), Blautia, Lachnospiraceae_UCG-006, and Enterorhabdus. These genera were largely absent in the other two groups and were located primarily within Firmicutes branches in the phylogenetic tree.

Species-level analysis further refined these group-specific differences. ND-fed mice were enriched in Bacteroides acidifaciens (LDA = 3.94), Mucispirillum sp. (LDA = 3.64), and Lactobacillus hominis (LDA = 3.21), each displaying localized signal in ND samples and minimal representation elsewhere. HSFD mice showed increased abundance of Helicobacter ganmani (LDA = 4.09), Corynebacterium stationis (LDA = 4.42), and Ileibacterium valens (LDA = 3.98), reflecting the taxonomic expansion observed under high-fat dietary conditions. In the YS-H group, Lactobacillus murinus emerged as the most dominant discriminative species, with the highest LDA score observed (4.51, p = 0.004). Additional extract-enriched taxa included Adlercreutzia caecicola and Streptococcus danieliae, both detected exclusively in the extract-treated group with strong signal intensity and Firmicutes-based phylogenetic placement.

Genus- and species-level abundance heatmaps confirmed these patterns, showing group-specific clustering and minimal overlap among differential taxa. Several Firmicutes-affiliated taxa, including Lactobacillus murinus, Lachnospiraceae_UCG-006, and Streptococcus danieliae, displayed elevated relative abundance exclusively in the YS-H group, with minimal detection in the ND and HSFD samples. These taxa were consistently identified across LDA rankings, heatmap clusters, and Firmicutes-associated branches in the phylogenetic tree.

To clarify the associations between extract-modulated gut microbiota and host metabolic outcomes, Spearman correlation analysis was performed, focusing on representative genera and species altered by high-dose Coriandrum sativum seed extract (Figure 9A). The correlation matrix incorporated indicators of glucose metabolism, lipid metabolism, liver injury, oxidative stress, and inflammation, and statistical significance was denoted using p-value thresholds (p < 0.05, p < 0.01, p < 0.001).

Several extract-enriched taxa demonstrated significant inverse correlations with pathological parameters induced by a high-sugar, high-fat diet. Lachnospiraceae_UCG-006 and Lactobacillus murinus showed negative correlations with fasting blood glucose, serum LDL-C, IL-6, and NF-κB levels, while positively correlating with hepatic glycogen and SOD activity. These patterns aligned with the metabolic improvements observed in extract-treated animals. Streptococcus danieliae and Christensenellaceae_R-7_group, also enriched in the extract group, exhibited inverse associations with liver injury markers (ALT, AST), lipogenesis-related genes (FAS), and systemic inflammatory indicators.

In contrast, Desulfovibrio, which was elevated in HSFD-fed mice, showed significant positive correlations with ALT, liver TG, and pro-inflammatory markers, while displaying negative associations with SOD and hepatic glycogen content, reflecting its alignment with a metabolically impaired state.

The relative abundance profiles of the key genera and species (Figures 9B–G) confirmed their differential expression across groups. Lachnospiraceae_UCG-006 and Lactobacillus murinus were markedly increased in the high-dose extract group (p < 0.05 or p < 0.01), while Desulfovibrio was selectively elevated in the HSFD group. Streptococcus danieliae was minimally detected in ND and HSFD mice but was enriched in the extract-treated group.

These data provide a taxonomic framework for understanding how specific microbial populations may contribute to the metabolic benefits observed following Coriandrum sativum seed extract intervention, serving as a bridge to further mechanistic investigations.

To further investigate the functional implications of Coriandrum sativum seed extract-mediated microbial shifts, PICRUSt2-based predictive metagenomic profiling was performed using 16S rRNA sequencing data from the high-dose group. KEGG pathway annotations at three hierarchical levels (L1–L3) were used to assess the impact of microbial changes on host-relevant metabolic capacities (Figures 10A–C).

At KEGG level 1 (L1), the dominant functional category across all groups was “Metabolism”, which accounted for the largest proportion of predicted gene content. HSFD feeding modestly suppressed overall metabolic functions, while Coriandrum sativum extract treatment restored and further enhanced this functional category (Figure 10A). Other KEGG L1 pathways, such as “Genetic Information Processing” and “Environmental Information Processing”, also showed mild increases in the extract group compared to the HSFD group.

More detailed categorization at KEGG level 2 (L2) revealed extract-induced enhancements in microbial pathways related to “Carbohydrate metabolism”, “Amino acid metabolism”, and “Energy metabolism” (Figure 10B). Notably, the abundance of genes associated with carbohydrate metabolism was substantially increased in the Coriandrum sativum-treated group relative to both ND and HSFD groups, indicating a restoration of fermentative capacity that may contribute to improved glycemic control. Similarly, pathways associated with amino acid turnover and cofactor/vitamin metabolism were elevated, suggesting enhanced microbial biosynthetic potential under extract intervention.

At KEGG level 3 (L3), further resolution identified enrichment in specific metabolic modules including “Metabolic pathways”, “Biosynthesis of secondary metabolites”, and “Microbial metabolism in diverse environments” (Figure 10C). Among them, “Metabolic pathways” represented the most prominent shift, showing a marked increase in mean gene abundance in the Coriandrum sativum extract group. Additionally, elevated prediction scores for “Biosynthesis of amino acids”, “Carbon metabolism”, and “ABC transporters” indicated a broader reprogramming of microbial functional capacity toward nutrient utilization and environmental adaptability.

Collectively, these predictions indicate that Coriandrum sativum seed extract potentially regulates key functions in energy production, biosynthesis, and host-microbiome interactions, as a potential basis for the improved metabolism seen in mice on a high-sucrose high-fat diet (HSFD).