Evo M-MLV reverse transcription, SYBR^® Green Pro Taq HS premix qPCR kit from Accurate Biotechnology(Hunan, China) Co.,Ltd. Propylthiouracil from Beijing Solarbo Technology Co., Ltd (Beijing, China). Triiodothyronine (T₃), thyroid stimulating hormone (TSH), Tetraiodothyronine (T₄), cyclic adenosine monophosphate (cAMP), thyrotropin releasing hormone (TRH), free fatty acid (FFA), lipoprotein lipase (LPL), Na⁺-K⁺-ATPase, nitric oxide (NO), cyclic guanosine monophosphate (cGMP), Ca²⁺-ATPase, hepatic glycogen (Glycogen), succinate dehydrogenase (SDH) kits were obtained from Shanghai Kexing Biotechnology (Shanghai, China) Co., Ltd. GAPDH, TRH, THRB1, THRA 1 + 2 antibody and HRP-IgG (H + L) from Beijing Bioss Biotechnology (Beijing, China) Co., Ltd. PBS from Dailian Meilun (Dailian, China) Co., Ltd. Acetonitrile and methanol of HPLC quality were acquired from Merck (China). Aladdin (China) supplied formic acid of HPLC quality. The Milli-Q system (Millipore, Bedford, MA, USA) was used to prepare ultrapure water. The remaining substances were all analytical grade.

Cistanche deserticola Y.C.Ma were collected from Alashan (Inner Mongolia, China) and authenticated by Professor Yanjun Zhai (Liaoning University of Traditional Chinese Medicine). Specimens of botanical drugs (CPU20190508-1, CPU20190508-2) were deposited in Liaoning Traditional Chinese Medicine Processing Technology Industry Innovation Center. Refer to the basis of our previous research (Liu et al., 2020), the optimal wine-steaming process parameters were used to prepare wine-processed CD.

500 g of CD/wine-processed CD was extracted with water three times, each for 1 h. The filtrates were combined, concentrated, and 95% ethanol was added to the appropriate concentration. After stirring, the solution was kept at 4°C overnight, then centrifuged. The supernatant was collected and concentrated until no ethanol remained. The solution was passed through a D101 macroporous resin column, and the eluate with 40% ethanol was collected and concentrated to obtain the total glycosides fraction. The total glycosides of raw CD [yield 7.96% (w/w)] and the total glycosides of wine-processed CD [yield 4.83% (w/w)] were stored in a refrigerator at 4°C. Utilizing HPLC-PDA(Waters, Milford, United States), chemical compound studies of the content of RG and WG extracts were carried out.

Changsheng Biotechnology (Liaoning, China) Co., Ltd. was the commercial supplier of 40 male Sprague-Dawley (SD) rats weighing 200 ± 20 g and licensed under license number SCXK-(Liao) 2020-0001. Each rat was kept under SPF-standard lab conditions, which included full access to food and water, a temperature of 25°C ± 2°C, a humidity of 40% ± 10%, and a 12-hour cycle of light and dark. The rats were acclimated for 1 week before being randomly assigned to four groups (n = 8), which included the control group (CN), model group (MD), the 82 mg/kg/day total glycosides of raw CD treated group (RG), the 50 mg/kg/day total glycosides of wine-processed CD treated group (WG), and positive drug group (PS). The dosage was converted from the human dosage in the Chinese pharmacopoeia to the rat dosage according to the proportion of total glycosides of raw/wine-processed CD and the conversion of human and rat body surface area.

The animal studies were carried out in strict conformity with the standards of Chinese Experimental Animals Administration Legislation, and the Animal Care and Use Committee of Liaoning University of Traditional Chinese Medicine authorized the experimental protocol (NO. SYXK [Liao] 2019-0004). The rats were intragastrically administered propylthiouracil (10 mL/kg) for 15 days to induce the KYD rat model. The treatment group was given RG/WG extracts for 15 consecutive days, while the control (CN) and model group (MD) were given the same amount of saline. The positive group (PS) was given 0.936 g/kg Guifu Dihuang Pills.

At the end of the 31st day, a minimum of 6 pills of feces were taken from every rat, separated into two sections in a sterilized EP tube containing 10 μL of 1% sodium azide (NaN₃), and kept at −80°C for the 16S rRNA gene sequencing and feces metabonomics analyses that followed. After being put to sleep with a 20% urethane solution, the rats were killed. Rats from each group were slaughtered simultaneously. The plasma for the ELISA analysis was obtained by centrifuging certain blood specimens in a cylindrical centrifuge adding anticoagulant sodium for 15 min at 4°C, while other blood samples were kept at −80°C. The tissues of the liver, spleen, thymus, kidney and pathogenic axis (pituitary, thyroid, and hypothalamus) were quickly isolated on an ice surface, cleaned with saline, wiped with filter paper, frozen with liquid nitrogen, and stored at −80°. The thyroid was preserved for histological examination in 4% paraformaldehyde. Rats’ kidney, thymus, and spleen were measured in order to determine the organ index (Lian et al., 2024). Organ index % = weight   of   the   organ   g weight   of   the   body   g × 100 %

ELISA kits were used to measure the amounts of thyroid hormones, such as T₃, TRH, TSH, FFA, T₄, LPL, and NO in serum samples. A microplate reader was used to measure the amounts of Ca²⁺-ATPase, SDH, glycogen, and Na⁺-K⁺-ATPase of rat liver homogenate. Using ELISA kits, the amounts of key cell metabolic regulators, such as cAMP and cGMP in plasma samples were determined.

Following the directions provided by the manufacturer, TRIzol reagent (Invitrogen, Carlsbad, CA, USA) was employed to extract total RNA of rapidly frozen hypothalamus and hypophysis tissues. Both the reverse transcription and genetic DNA removal reactions were carried out as directed. Amplification using SYBR^® Green Pro Taq HS qPCR was performed via an Applied Biosystems (ABI) 7500. Accurate Biotechnology (Hunan, China) Co., LTD generated the primer sequences for TRH and TSH using the Primer Premier 5.0 program, using the housekeeper gene β-actin as an internal reference. The following were the conditions for amplification: predenaturation at 95°C, denaturation at 95°C, and annealing at 60°C (30 s), with a total of 40 cycles. At the end of the reaction, the specificity of PCR amplification was determined by dissolution curve analysis, the Ct value was read, and the relative expression of target gene mRNA was calculated by the 2^−ΔΔCt method. Primer sequences are shown in Table 1.

For the general histological test, 5 μm slices of the specimens were stained with hematoxylin and eosin (H&E) after they had been dried in an ordered sequence of alcohol and covered with paraffin. Lastly, a light microscope was used to view the thyroid gland’s abnormal appearance. The thyroid tissues were dehydrated and waxed, and then thin sections (4 μm) were blocked, deparaffinized, and placed for 12 h at 4°C with anti-Thyroid Stimulating Hormone Receptor (TSHR), anti-Thyroid peroxidase (TPO), and anti-Thyroglobulin (TG) (1:220; Abcam Technology, United Kingdom). As directed by the manufacturer, sections were incubated for 1 hour with secondary antibodies diluted 1:510 with horseradish peroxidase (HRP) labeled. To view immunological complexes, slides were cleaned and incubated for 10 min in PBS including 0.02% 3,3-diaminobenzidine and 0.01% H₂O₂. Finally, cells that had been positively labeled were examined under a light microscopy.

After weighing 30 mg of the material into an EP tube, 500 μL of the resulting solution (methanol:acetonitrile:water = 2:2:1, a standard inner mixture that was isotopically labeled) was applied. These specimens were subsequently sterilized for 5 minutes in an ice water bath after undergoing homogenization for 4 minutes at 35 Hz. Three iterations of the homogenized and ultrasound cycle were conducted. Following a one-hour incubation period at −40°C, the specimens were spun up for 15 min at 4°C at 12,000 rpm. Quality control (QC) sample was prepared by mixing an equal aliquot of the supernatants from all of the samples. LC-MS/MS analyses were performed using an UPLC system (Vanquish, Thermo Fisher Scientific) with a UPLC BEH Amide column (2.1 mm × 100 mm, 1.7 μm) coupled to Q Exactive Orbitrap MS spectrometer (Thermo). Acetonitrile (A) and fluid at a pH of 9.75 containing 25 mmol/L ammonia hydroxide and ammonium acetate (B) made up the mobile phase. The auto-sampler temperature was 4°C. Since the QE HFX mass spectrometer can get MS/MS spectra in information dependent acquisition (IDA) mode under the supervision of the gathering system (Xcalibur, Thermo), it was utilized and the injection volume was 2 μL. During this mode, the data collection program constantly assesses the entire scan MS spectrum. The ESI source situations were established as follows: 3.7 kV positive and −3.0 kV negative spray voltage, 31 Arb sheath gas flow, 8000 MS/MS resolution, 24 Arb auxiliary gas flow, 350 °Ccapillary temperature, 65000 full MS resolution.

ProteoWizard was utilized to transform the raw data to the mzXML style, and an internal program built with R centered on XCMS was utilized for cleaning the data for maximum identification, mining, placement, and merging. Then, for metabolites annotation, an internal MS2 library (Biotree DB) was selected. Markup limit was determined at 0.3. For multidimensional statistical examination, all standard deviations were entered as unmodified information into the SIMCA 16.0.0 application (Sartorius Stedim Data Analytics, Umea, Sweden).

All rats’ feces were taken on days 0, 15, and 22 of the medication therapy. Following gathering, each feces specimen was quantified at 0.2 g (the mean across 5 specimens of feces), frozen right away using liquid nitrogen, followed by frozen at −80°C for further DNA extraction. As stated earlier, ERIC-PCR procedures, division, and whole genome DNA collection and evaluation have been carried out (Chi et al., 2019). The Illumina MiSeq platform was used to treat cleaned DNA specimens in preparation for 16S amplicon reading (Illumina, San Diego). Using barcoded 515F along with 806R broad primers designed for the V3-V4 zone as previously indicated, the 16S rRNA gene was selected for deletion (Abellan-Schneyder et al., 2021). In short, a restricted PCR cycle was utilized to combine the reads utilizing dual-index tags and Illumina DNA sequencing connections. The Nextera XT method was adopted for library normalization after purification with Agencourt AMPure beads (Agencourt, USA) (Rinke, 2018). Afterward, samples were placed into a single flow cell for sequencing on the MiSeq sequencing platform (Illumina, San Diego).

In the initial run, clumps were auto-generated using double tag reads having an average size of 2 × 300 bp and paired-end processed. PANDAseq was used to gather paired-end indicates and Raw FASTQ data were acquired in accordance with the preceding procedure (Langille et al., 2013). Barcodes and primers were removed from the sequences. “N” and any data sequences shorter than 250 bp were eliminated (Martijn et al., 2019). Solo readings and replicas were then eliminated after the processed scans grouped at k = 10 (97% similarity) (Zhang M. et al., 2023). Finally, BLASTn was used to OTUs with a selected library that was sourced via NCBI, RDPII, and GreenGenes (Li et al., 2022).

In short, isolate DNA using CTAB as directed by the manufacturer. PCR was utilized to amplify the 16S rRNA gene’s V3-V4 domain. An Agilent 2100 bioanalyzer (Agilent, USA) and an Illumina Novaseq platform were used for the sequencing process. Software called VSearch (v2.3.4) can be used to filter hybrid genomes. Following DADA₂ removal, we were left with a characteristic list and a metabolite series. The Silva (version 138) classifier states that the characteristic levels are normalized using each specimen’s relative abundance. After dereplication using DADA₂, we obtained feature table and feature sequence. According to SILVA (release 138) classifier, feature abundance was normalized using relative abundance of each sample. Five indices—Chao1, observed species, goods coverage, Shannon, and Simpson—are used to analyze the degree of species diversity. QIIME2 was used to calculate each index for our collections. QIIME2 was utilized for assessing beta diversity, and the R program was used to create the visualizations. Aligned sequence alignment was done using Blast, and every instance sequence’s characteristic regions were added to using the SILVA library. The R package (version 3.5.2) was used to create additional graphics. (Xie et al., 2024).

KEGG, HMDB, METLIN and MetaboAnalyst were used to import the different metabolites that were picked from feces metabonomics for pathway enriched research. Through the application of the PICRUSt 2 program to compare the variety makeup parameters derived through the 16S sequenced data, the practical structure of microorganism specimens was calculated, allowing for the analysis of the practical distinctions between various specimens or groups (Langille et al., 2013). Lastly, applying the G-TEST and Fisher test procedures in the STAMP apps, a major distinction test among specimens was conducted at the degree of genus for biodiversity between various specimens (t-test, P < 0.05). Generally speaking, comparative examinations of KEGG metabolic pathways can predict variations and shifts in the biochemical pathways associated with function genes of microbial communities across samples within other groups.

A Pearson correlational study was conducted between the genus of disturbed gut microorganisms assessed from 16S rRNA gene sequencing analysis and the modified metabolites filtered from fecal metabonomics (Zhao et al., 2022). Those correlation coefficient values (|r| ≥ 0.70, P < 0.05) were considered to be significantly correlated. Eventually, significant related metabolites and gut microbe genus were obtained and displayed as heat maps (*P < 0.05, **P < 0.01). Red indicates a positive correlation while blue indicates a negative correlation.

The findings were displayed as mean ± SEM, with the number of repetitions for each group indicated in the caption. Software named SPSS 20.0 (IBM, USA) was used to conduct the statistical analysis. The degrees of statistical validity were ascertained using the LSD test and one-way ANOVA. The threshold for the level of significance was chosen P < 0.05 or P < 0.01. For intestinal bacterial study analysis, GraphPad Prism (GraphPad Prism version 5.1) was utilized for statistical examination and graphical representation of the data. To identify substantial variations, a Kruskal-Wallis test with Bonferroni correction was employed. For alpha and beta diversity analysis, the R packages phyloseq, ggplot₂ and scales were utilized. For multidimensional statistical evaluation, including PCA, PLS-DA and OPLS-DA, all normalized parameters were entered as raw data into the SIMCA-P 13.0 software (Umetrics, Sweden). The PLS-DA model has been confirmed using the method of 200 permutation testing, OPLS-DA was validated using both an a whopping seven a cross-valid and an ANOVA of the CV-ANOVA technique.

Both RG and WG have phenylethanoid glycoside compositions that differ significantly, the contents of echinacoside, acetoside, and 2′-acetylacteoside decrease, while the content of isoacetoside increases significantly (Figure 1). Our prior research commented on the transition process (Li et al., 2021).

For the purpose of metabolism evaluation, all of the aforementioned putative indicators were imported into MetaboAnalyst (Xu et al., 2023). These pertinent paths were found using parameters (effect value >0.1) as the ranking factor. As seen in Figures 5B,C, the outcome was displayed as an interactive visual system. Pyrimidine, one carbon pool by folate, amino acid, vitamin B6, arachidonic acid, glycine, serine, threonine, GPI-anchor biosynthesis metabolism were the metabolic pathways that were strongly linked to the pharmaceutical regimen of propylthiouracil-induced KYD rats, according to the outcomes of metabolic pathway enhancement. Figure 6 displayed the main routes of metabolism linked to biomarkers. We examined the reaction of metabolites disrupted by modeling formation following CDG treatment in order to assess the curative effect and explain the mechanisms of CDG on KYD rats. Among the possible biomarkers linked to KYD were Guanidoacetic acid, Uridine, Lecithin, Phosphatidylethanolamine, N-Acetyl-D-glucosaminyldiphosphodolichol, Pyridoxamine, Folic acid. The variation trends of typical metabolites were shown respectively in Supplementary Figure S4. Compared with the MD group, the levels of Guanidoacetic acid, Lecithin, N-Acetyl-D-glucosaminyldiphosphodolichol, Pyridoxamine and Folic acid were increased, after RG and WG treatment, which was close to CN group, especially for WG treatment.

Figure 7 displays findings of the 16S rRNA technology assessment of the gut bacteria. The Shannon and Simpson curves remained relatively flat as the sequencing amount increased. This trend suggests that the sequencing data are robust and reliable. The species within the sample are not only abundant but also hardly distributed, strongely capturing underlying units structures (Figures 7A,B). Alpha diversity primarily reflects species richness and evenness through the Chao1 index (Figure 7C). In this study, Thirty samples were sequenced using Illumina MiSeq sequencing to produce original gene tags, and the paired-end information was spliced using the overlaid approach for chimeric sequence identification and quality control. In the Venn diagram of rat intestinal microbial diversity (Figure 7D), 108 shared ASVs were identified across the five groups. The overlap between the WG and CN groups was larger than that between the RG and CN groups, indicating that the WG group is more similar to the CN group. These findings suggest that CDG has the potential to ameliorate the disorder of the microbiota structure in KYD rats, with the WG treatment demonstrating a more pronounced effect (Figures 7E,F). Principal Coordinate Analysis (PCoA) results are presented in Figure 7J. The first two principal coordinates (PCoA1 and PCoA2) account for 7.69% and 6.09% of the total variance, respectively.

Figures 7H,I displayed the composition of the phylumlevel flora in every group. Firmicutes and Bacteroidetes made up almost all of the gut bacteria among all groups. While the total amount of Bacteroides in all groups receiving treatment was substantially less than that of the MD group (P < 0.05), the richness of Bacteroides in the MD group was 87.2%, which was far greater than that of CN group (80.1%). The percentage of Firmicutes to Bacteroidetes, which usually reflects the composition of the main intestinal flora (Li J. et al., 2024), was 7.52,13.22,10.51,11.31 and 8.31 in each group, respectively (Figure 7M). Based on LEfSe analysis, it was found that the abundance of Actinobacteria and Chloroflexi in MD group was significantly increased compared with CN group, while the abundance of Actinobacteria and Chloroflexi in other treatment groups was decreased compared with MD group. In addition, propylthiouracil intervention resulted in a significant reduction in the abundance of Bacteroidetes and Fusobacteria in the intestine of rats, which could be reversed by CDG intervention (Figures 7K,L). In conclusion, CDG improves the disturbance and structural changes of kidney-yang deficiency microflora induced by propylthiouracil at the phylum level.

The top 10 species in each group were identified by additional bacteria analysis at the general level in terms of sample organization and range (Figure 7G), comprising, Pediococcus, Lactobacillus, Firmicutes, Bacteroidetes, Porphyromonadaceae, Bacteroidales, Clostridiales, Lachnospiraceae, Muribaculaceae, Clostridium and Bifidobacterium, which were dominant among all groupings. The predominant bacteria found in the CN group were Firmicutes, Lactobacillus, Pediococcus. Crucially, the MD group had a considerably larger amount of Lactobacillus than the rest of the groups (P < 0.05). Figure 7M depicts that in MD, the number of Bifidobacterium, Lactobacillus, Porphyromonadaceae and Pediococcus increased significantly, while the number of Bacteroidales, Clostridiales, Lachnospiraceae, Muribaculaceae, Clostridium, Staphylococcus and Ruminococcaceae decreased significantly. On the contrary, in the RG/WG treatment group, the abundance trends of 11 genera such as Bifidobacterium, Bacteroidales, Clostridiales, Lachnospiraceae, Muribaculaceae were reversed (P < 0.05). The relative abundance of the microbial community had a greater impact by the WG group than by the RG group.

Fecal metabolites mediate the interaction between gut microbiota and HPT axis ELISA indicators in rats. Metabolites play an important role in host-microbiome interactions. First, we used pearson correlation analysis to evaluate HPT axis ELISA indicators and key fecal metabolites. As shown in Figures 8A, T₃, T₄ and TRH were positively correlated with Pyridoxamine, Tyrosine, Stearic acid and Serine, while Glycogen was strongly negatively correlated with Uridine, Folate and Phosphatidylcholine. These data suggested that RG/WG may play a role in anti-KYD through regulating specific fecal metabolites. Secondly, in order to better understand the functional association between fecal metabolites that interact with intestinal flora, correlation analysis was performed on bacterial genera and metabolites with significant differences. As shown in Figure 8B, Serine was positively correlated with Prevotella, Folate was positively correlated with Bifidobacterium, Pyridoxamine was positively correlated with Lactobacillus, similar to Uridine and Firmicutes-to-Bacteroidetes, Pyridoxamine and Pseudomonas. RG/WG may significantly improve KYD rats through the connection between intestinal flora and arachidonic acid metabolism. For example, the observed Phosphatidylcholine between intestinal flora and fecal metabolites was negatively correlated with lleibacterium, and Phosphatidylethanolamine was significantly negatively correlated with Streptococcus.