Identification and validation of key target genes of penning formula for treating rat chronic endometritis model.

The average error rate of reads across different samples, as well as the distribution of error rates and base content in the transcriptomic sequencing data, confirmed the reliability of the data for subsequent analyses (Supplementary Fig. S1 ). The volcano plot displays a total of 442 differentially expressed genes (DEGs1), with 201 up-regulated and 241 down-regulated. Additionally, 409 DEGs2 were identified, comprising 209 up-regulated and 200 down-regulated genes. The top 10 genes with the most significant changes were annotated (Fig.  1 a,b). The heatmap further illustrated the expression trends of the 20 genes with the most significant changes (Fig.  1 c,d). Furthermore, a total of 47 PNF-related genes (PRGs) were identified, including 19 intersection genes between the up-regulated genes of DEGs1 and the down-regulated genes of DEGs2, and 28 intersection genes between the down-regulated genes of DEGs1 and the up-regulated genes of DEGs2 (Fig.  1 e,f). Fig. 1 Identification of 370 DEGs1, 464 DEGs2, and 47 PRGs. ( a ) Volcano plot of differentially expressed genes (DEGs1). Green dots represent down-regulated genes, red dots represent up-regulated genes, and gray dots represent genes with no significant differential expression. ( b ) Volcano plot of differentially expressed genes (DEGs2). Green dots represent down-regulated genes, red dots represent up-regulated genes, and gray dots represent genes with no significant differential expression. ( c ) Heatmap of differential expression analysis for DEGs1. Yellow indicates high expression, and purple indicates low expression. ( d ) Heatmap of differential expression analysis for DEGs2. Yellow indicates high expression, and purple indicates low expression. ( e ) Venn diagram of the intersection of DEG1 up-regulated genes and DEG2 down-regulated genes. ( f ) Venn diagram of the intersection of DEG1 down-regulated genes and DEG2 up-regulated genes. Identification of 370 DEGs1, 464 DEGs2, and 47 PRGs. ( a ) Volcano plot of differentially expressed genes (DEGs1). Green dots represent down-regulated genes, red dots represent up-regulated genes, and gray dots represent genes with no significant differential expression. ( b ) Volcano plot of differentially expressed genes (DEGs2). Green dots represent down-regulated genes, red dots represent up-regulated genes, and gray dots represent genes with no significant differential expression. ( c ) Heatmap of differential expression analysis for DEGs1. Yellow indicates high expression, and purple indicates low expression. ( d ) Heatmap of differential expression analysis for DEGs2. Yellow indicates high expression, and purple indicates low expression. ( e ) Venn diagram of the intersection of DEG1 up-regulated genes and DEG2 down-regulated genes. ( f ) Venn diagram of the intersection of DEG1 down-regulated genes and DEG2 up-regulated genes. Functional enrichment analyses were performed on DEGs1, DEGs2, and PRGs separately. Among the 256 Gene Ontology (GO) categories enriched by DEGs1, 212 were biological processes (BP), such as regulation of small molecule metabolic processes, 8 were cellular components (CC), such as L-type voltage-gated calcium channel complex, and 36 were molecular functions (MF), including DNA-binding transcription activator activity (Fig.  2 a). DEGs2 were enriched in 351 GO categories, including 304 BP, such as nucleoside metabolic processes, 20 CC, such as the mitotic spindle, and 27 MF, including titin binding (Fig.  2 b). PRGs were enriched in 179 GO categories, with 149 BP, such as multicellular organismal response to stress, 13 CC, such as voltage-gated sodium channel complex, and 17 MF, such as estrogen response element binding. The top 7 most significant categories for each set of genes were visualized (Fig.  2 c). Additionally, 10 KEGG pathways, including vascular smooth muscle contraction, were enriched by DEGs1 (Fig.  2 d), while 30 pathways, including the cAMP signaling pathway, were enriched by DEGs2, with the 20 most significant pathways visualized (Fig.  2 e). Similarly, 8 pathways, including the prolactin (PRL) signaling pathway, were enriched by PRGs (Fig.  2 f). These results suggest that the genes are primarily linked to their respective functions and pathways, providing insight into the potential roles of PNF treatment-related genes in the biological and pathological processes associated with CE. Fig. 2 Related functional pathways of DEGs1, DEGs2, and PRGs. ( a ) Gene Ontology (GO) enrichment analysis of DEGs1. ( b ) GO enrichment analysis of DEGs2. ( c ) GO enrichment analysis of PRGs. ( d ) Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of DEGs1. ( e ) KEGG enrichment analysis of DEGs2. ( f ) KEGG enrichment analysis of PRGs. Related functional pathways of DEGs1, DEGs2, and PRGs. ( a ) Gene Ontology (GO) enrichment analysis of DEGs1. ( b ) GO enrichment analysis of DEGs2. ( c ) GO enrichment analysis of PRGs. ( d ) Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of DEGs1. ( e ) KEGG enrichment analysis of DEGs2. ( f ) KEGG enrichment analysis of PRGs. A total of 25 active ingredients, including salidroside and baicalin, were identified as potential key substances through which PNF exerts its therapeutic effects in CE (Supplementary Table S1 ). Subsequently, 909 target genes (TGs), including TARS, were obtained. From the intersection of PRGs and TGs, three key target genes—Bhmt, Scn10a, and Esr2—were identified (Fig.  3 a). Analysis based on the gene’s localization on the chromosome showed that these genes are located on chromosomes 2, 8, and 6, respectively (Fig.  3 b). Notably, all three key target genes were highly expressed in the model group and underexpressed in the test group (Fig.  3 c). The chromosomal locations of these key target genes provided insights into their distinct roles in CE and offered a novel perspective for exploring their functional associations at the chromosomal level. Fig. 3 Chromosome localization and expression of three key target genes. ( a ) Screening of key target genes. ( b ) Distribution map of key genes on chromosomes. ( c ) Analysis of expression levels of key target genes. Chromosome localization and expression of three key target genes. ( a ) Screening of key target genes. ( b ) Distribution map of key genes on chromosomes. ( c ) Analysis of expression levels of key target genes. Further analysis revealed that the key target genes were enriched in 179 Gene GO categories, including 149 BPs, such as multicellular organismal response to stress (GO: 0033555), rhythmic process (GO: 0048511), and response to hydroxyisoflavone (GO: 0033594); 13 CCs, including voltage-gated sodium channel complex (GO: 0001518), sodium channel complex (GO: 0034706), and clathrin coat (GO: 0030118); 17 MFs, such as estrogen response element binding (GO: 0034056), voltage-gated monoatomic ion channel activity involved in regulation of presynaptic membrane potential (GO: 0099508), and nuclear steroid receptor activity (GO: 0003707). The top 7 most significant categories are shown in Fig.  4 a. Additionally, 8 KEGG pathways, including the PRL signaling pathway (KEGG: rno04917), GnRH secretion (KEGG: rno04929), and glycine, serine, and threonine metabolism (KEGG: rno00260), were enriched by these key target genes (Fig.  4 b). Fig. 4 Crucial functional pathways of key target genes. ( a ) GO enrichment analysis of key target genes. ( b ) KEGG enrichment analysis of key target genes. Crucial functional pathways of key target genes. ( a ) GO enrichment analysis of key target genes. ( b ) KEGG enrichment analysis of key target genes. A PNF herbal medicine-active ingredient-key target gene-pathway network was constructed, comprising 11 PNF herbal medicines, 48 active ingredients, and 8 pathways. A totalof 92 interactionswereobserved, which revealed the potential regulatory relationships between PNF components and key target genes, as well as their associated functional pathways (Fig.  5 ). Fig. 5 Pathway network diagram of herbs, ingredients, and key target genes. Yellow represents the Chinese medicine names, orange represents the active ingredients of the medicine, cyan represents the key target genes, and blue represents the KEGG pathways. Pathway network diagram of herbs, ingredients, and key target genes. Yellow represents the Chinese medicine names, orange represents the active ingredients of the medicine, cyan represents the key target genes, and blue represents the KEGG pathways. An interaction network was established to illustrate the multiple interactions between the three key target genes and 20 associated genes, highlighting their roles in biological functions such as alpha-amino acid biosynthetic processes, cellular amino acid biosynthesis, and alpha-amino acid metabolic processes (Fig.  6 a). Notably, the physical interaction between Bhmt and Pdzd2 was particularly strong, and Bhmt shared similar functions with Bhmt2. Further analysis of the functional similarities among the key target genes revealed that Esr2 exhibited the greatest functional overlap with the other key target genes, indicating its central role in the gene expression regulatory network (Fig.  6 b). This network provides a comprehensive view of the complex interactions among key target genes, suggesting that PNF may influence the progression of CE by modulating these interactions. Fig. 6 Complex interaction network of key target genes. ( a ) Network diagram of key target genes and their co-expressed genes. The size of the circles (nodes) represents the “degree,” which is the number of interactions the node has with other nodes. The higher the degree, the larger the circle, indicating the importance of the node in the network or the richness of its interactions. The colors in the circles represent the functions of the related target proteins, and different line colors represent different network relationships. ( b ) Functional similarity analysis of key target genes. Different colors represent different genes, and the x-axis represents gene expression levels. Complex interaction network of key target genes. ( a ) Network diagram of key target genes and their co-expressed genes. The size of the circles (nodes) represents the “degree,” which is the number of interactions the node has with other nodes. The higher the degree, the larger the circle, indicating the importance of the node in the network or the richness of its interactions. The colors in the circles represent the functions of the related target proteins, and different line colors represent different network relationships. ( b ) Functional similarity analysis of key target genes. Different colors represent different genes, and the x-axis represents gene expression levels. Upon intersecting the miRNAs predicted by three databases, five common miRNAs for the key target genes were identified, including rno-miR-29a-3p, rno-miR-29b-3p, and rno-miR-29c-3p for Bhmt, as well as rno-miR-381-3p and rno-miR-219a-5p for Esr2 (Fig.  7 a). A miRNA-key target gene regulatory network was visualized (Fig.  7 b), suggestingthat the efficacy of PNF in CE may be linked to the regulatory roles of specific miRNAs acting upstream of the key target genes. Fig. 7 Specific miRNA-key target genes regulatory network. ( a ) Venn plot of miRNA intersection predicted by EIMMo, mirDB, and Microcosm databases. ( b ) miRNA-key target gene regulatory network diagram. Orange represents key target genes, and blue represents miRNAs. Specific miRNA-key target genes regulatory network. ( a ) Venn plot of miRNA intersection predicted by EIMMo, mirDB, and Microcosm databases. ( b ) miRNA-key target gene regulatory network diagram. Orange represents key target genes, and blue represents miRNAs. Binding assays revealed that the key target genes had strong binding activity with several corresponding active ingredients. Esr2 exhibited robust binding with Baicalin, Erycibenin_D, Garbanzol, Apigenin, and Quercetin (|Total Score| > 8.0). The conformations of the Esr2-active ingredient interactions with the highest |Total Score| were visualized. Additionally, Scn10a showed strong binding activity with Ergosterol (|Total Score| > 8.0), and the corresponding conformations were also visualized (Supplementary Fig. S2 ). These results suggest that PNF may exert its therapeutic effects on CE through the specific binding of its active ingredients to the key target genes. Gene expression analysis revealed that, compared to the control group, Scn10a and Esr2 were significantly upregulated in the CE model group ( P  < 0.05), consistent with earlier findings. Furthermore, compared to the CE group, Bhmt, Scn10a, and Esr2 were significantly downregulated in the treatment group ( P  < 0.01), showing a reversal trend. This suggests that PNF may have a corrective or normalizing effect on the pathological processes associated with CE (Fig.  8 a–c). Fig. 8 Expression patterns of key target genes in CE, control, and treatment groups ( a : Bhmt, b : Scn10a, c : Esr2). Expression patterns of key target genes in CE, control, and treatment groups ( a : Bhmt, b : Scn10a, c : Esr2).

Transcriptome sequencing data were obtained from rat samples, which were divided into three groups: five samples from the control (normal) group labeled A1-5, five samples from the model (CE) group labeled B1-5, and five samples from the test (PNF-treated) group labeled C1-5. The PNFwas composed of the following herbal ingredients: Da Xue Teng ( Caulis Sargentodoxae ) 30 g, Bai Jiang Cao ( Herba Patriniae ) 30 g, Bai Hua She She Cao ( Herba Hedyotis )30 g, San Leng ( Rhizoma Sparganii ) 15 g, E Zhu ( Rhizoma Curcumae ) 15 g, Zao Jiao Ci ( Spina Gleditsiae ) 9 g, Lu Lu Tong ( Fructus Liquidambaris ) 15 g, Yan Hu Suo ( Rhizoma Corydalis ) 15 g, Tu Bie Chong ( Eupolyphaga Seu Steleophaga )9 g, Jiu Xiang Chong ( Aspongopus ) 3 g, Dan Shen ( Radix Salviae Miltiorrhizae )15 g, Dang Gui ( Radix Angelicae Sinensis ) 15 g, Chuan Xiong ( Rhizoma Ligustici Chuanxiong ) 6 g, Shan Yao ( Rhizoma Dioscoreae )15 g, Chao Bai Zhu ( Rhizoma Atractylodis Macrocephalae ) (stir-fried) 6 g, Huang Qi ( Radix Astragali ) 12 g. The herbs were soaked in 500 mL of water for 30 min, then decocted by boiling for 30 min. An additional 500 mL of water was added, followed by continued boiling for another 30 min.The decoction was filtered through medical gauze to obtain the final herbal extract, which was used for the construction of the subsequent rat model. Schematic diagram ofthe animal experiment design is shown in Supplementary Figure S3 . Seventeen SPF-grade healthy, nulliparous female SD rats (aged 6–8 weeks), with an initial body weight of 160–200 g, were purchased from Spaf-Bio (Beijing) Biotechnology Co., Ltd.After one week of adaptive feeding, 19 SD rats were randomly divided into three groups: sham control surgery group (Control group, n  = 5), CE group ( n  = 7), and PNF group ( n  = 7). In the CE and PNF groups, after anesthesia, the 1 cm long mid-abdominal incision was made. The abdominal cavity was exposed, and both sides of the uterus were identified. The 4-gauge needle was inserted at the uterine bifurcation, and 3.5 ml/kg of LPS (20ul) was slowly injected towards the ovaries 26 – 28 . The Control group received an injection of an equal amount of physiological saline. After suturing the incision, the rats were kept in an inverted position for 1–2 min, and appropriate back-and-abdomen massage was applied to aid the even distribution of the infusion.One week after modeling, two rats from each group (Control, CE, and PNF) were randomly dissected. Gross observation revealed that the left uterine wall of the rats in the Control, CE, and PNF groups appeared thin, with a large amount of yellow or clear fluid visible in the uterine cavity. The uterine lumen was severely dilated, and the serosal surface of the uterus showed congestion, swelling, and a dark red color, indicating successful modeling.On the day of successful modeling, the PNF group rats were given 25 g/kg/d of PNF via gavage at a fixed time every day (based on the ratio of body surface area between humans and animals, assuming an adult weight of 60 kg, and converting the dose for rats using a factor of 6.25). The Control and CE groups received the same volume of physiological saline gavage at the same time. During the drug administration, vital signs were monitored weekly, and the medication continued for 14 days after successful modeling. Two hours after the last dose, the rats were anesthetized, euthanized by cervical dislocation, and the abdominal cavity was opened to collect entire uterine tissue from the Control, CE, and PNF groups.Additionally, uterine tissue was stored at -80 °C for sequencing. And the weight changes of rats are shown in Supplementary Table S2 . Total RNA was isolated and purified using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s procedure. The RNA amount and purity of each sample was quantified using NanoDrop ND-1000 (NanoDrop, Wilmington, DE, USA). The RNA integrity was assessed by Bioanalyzer 2100 (Agilent, CA, USA) with RIN number > 7.0, and confirmed by electrophoresis with denaturing agarose gel. Poly (A) RNA is purified from 1 µg total RNA using Dynabeads Oligo (dT)25-61005 (Thermo Fisher, CA, USA) using two rounds of purification. Then the poly(A) RNA was fragmented into small pieces using Magnesium RNA Fragmentation Module (NEB, cat.e6150, USA) under 94 ℃ 5–7 min. Then the cleaved RNA fragments were reverse-transcribed to create the cDNA by SuperScript™ II Reverse Transcriptase (Invitrogen, cat. 1896649, USA), which were next used to synthesise U-labeled second-stranded DNAs with E. coli DNA polymerase I (NEB, cat.m0209, USA), RNase H (NEB, cat.m0297, USA) and dUTP Solution (Thermo Fisher, cat.R0133, USA). An A-base is then added to the blunt ends of each strand, preparing them for ligation to the indexed adapters. Each adapter contains a T-base overhang for ligating the adapter to the A-tailed fragmented DNA. Single- or dual-index adapters are ligated to the fragments, and size selection was performed with AMPureXP beads. After the heat-labile UDG enzyme (NEB, cat.m0280, USA) treatment of the U-labeled second-stranded DNAs, the ligated products are amplified with PCR by the following conditions: initial denaturation at 95℃ for 3 min; 8 cycles of denaturation at 98℃ for 15 s, annealing at 60℃ for 15 s, and extension at 72 ℃ for 30 s; and then final extension at 72 ℃ for 5 min. The average insert size for the final cDNA library was 300 ± 50 bp. At last, we performed the 2 × 150 bp paired-end sequencing (PE150) on an illumina Novaseq™ 6000 (LC-Bio Technology CO., Ltd., Hangzhou, China) following the vendor’s recommended protocol. Fastpsoftware ( https://github.com/OpenGene/fastp ) were used to remove the reads that contained adaptor contamination, low quality bases and undetermined bases with default parameter. Then sequence quality was also verified using fastp. We used HISAT2 ( https://ccb.jhu.edu/software/hisat2 ) to map reads to the reference genome of Homo sapiens GRCh38. The mapped reads of each sample were assembled using StringTie ( https://ccb.jhu.edu/software/stringtie ) with default parameters. Then, all transcriptomes from all samples were merged to reconstruct a comprehensive transcriptome using gffcompare ( https://github.com/gpertea/gffcompare/ ). After the final transcriptome was generated, StringTie and was used to estimate the expression levels of all transcripts. StringTie was used to perform expression level for mRNAs by calculating FPKM (FPKM = [total_exon_fragments/mapped_reads(millions) × exon_length(kB)]). This study implemented a systematic data processing workflow to enhance the reliability of gene expression analysis: First, missing values in raw data were replaced with zeros to ensure data integrity. Subsequently, non-finite values (e.g., Inf, -Inf, or NaN) generated during log2 transformation due to zero or extremely low expression values were uniformly replaced with 0 to maintain computational stability. Total read counts per sample were then calculated to assess data quality, followed by log2 transformation to compress the dynamic range of expression data and improve distribution symmetry. The FPKM matrix was integrated with clinical grouping information, and samples or genes containing residual missing values were pruned to eliminate potential confounding factors. Finally, row-wise standardization (Z-score) was applied to normalize gene expression values, mitigating the impact of expression magnitude differences across genes and enabling clearer visualization of cross-sample transcriptional variation patterns in heatmaps. This multi-step collaborative optimization balanced data fidelity with analytical robustness. And the Phred base quality score formula was applied to calculate the average error rate of sequencing reads across different samples, using a Phred score threshold of 30 for quality filtering. Additionally, the error rate distribution and nucleotide base content composition were systematically assessed. The flowchart of this study is shown in Supplementary Figure S4 . DEGs1 between the model and control groups, and DEGs2 between the model and test groups, were identified using the DESeq2 package (v 1.40.2) 29 , with a |log 2 fold change (FC)| > 0.5 andadj.p (FDR) < 0.05 30 – 32 . Subsequently, PRGs were derived by identifying the intersection between up-regulated genes of DEGs1 and down-regulated genes of DEGs2, as well as between down-regulated genes of DEGs1 and up-regulated genes of DEGs2, using the VennDiagram package (v 1.7.3) 33 . To explore the biological functions and signaling pathways involved in DEGs1 and DEGs2, GO and KEGG 34 35 analyses were performed on DEGs1, DEGs2, and PRGs using the clusterProfiler package (v 4.7.1.003) 36 , with a significance threshold of P  < 0.05. The active ingredients of PNF, which include Dang Gui, E Zhu, Huang Qi, Lu Lu Tong, Bai Hua She She Cao, Yan Hu Suo, Chuan Xiong, Shan Yao, Dan Shen, Da Xue Teng, Tu Bie Chong, San Leng, Bai Zhu, Zao Jiao Ci, Bai Jiang Cao, and Aspongopus chinensis, were sourced from the Encyclopedia of Traditional Chinese Medicine (ETCM) database ( http://www.tcmip.cn/ETCM/ ) and the Traditional Chinese Medicine Systems Pharmacology (TCMSP) database ( https://old.tcmsp-e.com/tcmsp.php ), with oral bioavailability (OB) ≥ 30%, drug-likeness (DL) ≥ 0.18, and FDR (Benjamini and Hochberg (BH)) < 0.05. The potential targets of these active ingredients were retrieved and mapped to target gene names using the UniProt database ( https://www.uniprot.org/ ), with species set to rat, thereby identifying the target genes of PNF (TGs). Finally, key target genes of PNF for treating CE were determined by identifying the intersection between PRGs and TGs, which were then utilized for subsequent analyses. The RCircos package (v1.2.2) 37 was utilized to map the locations of key target genes on chromosomes. To assess the expression trends of these genes in CE, their expression levels in the control, model, and test groups were visualized. To explore the biological functions and signaling pathways associated with key target genes in the onset and treatment of the disease, GO and KEGG analyses were performed on the key target genes using the clusterProfiler package (v 4.7.1.003), with a significance threshold of P  < 0.05. To further explore the interactions between key target genes, PNF, and relevant pathways, a PNF herbal medicine-active ingredient-key target gene-pathway network was constructed based on the enriched pathways and key target genes. This network was visualized using Cytoscape software (v 3.1.1) 38 . Additionally, to investigate interactions between key target genes and other genes with related functions, a co-expression network was established using the GeneMANIA algorithm ( http://www.genemania.org/ ). To uncover the functional relationships and relevance among key target genes, Functional Relationships among Integrated Data for Eukaryotic Systems (Friends) analysis was conducted using the GOSemSim package (v 2.24.0) 39 . Key target gene-associated microRNAs (miRNAs) were predicted using the EIMMo database 40 , mirDB database ( https://mirdb.org/ ), and the microcosm database ( https://mycocosm.jgi.doe.gov/mycocosm/home ). The intersecting miRNAs predicted across the three databases were then identified as common miRNAs.It should be noted that the miRNA network analysis is based on predicted results from databases and has not yet been combined with actual miRNA expression data. Subsequently, a miRNA-key target gene regulatory network was constructed and visualized using Cytoscape software (v 3.1.1). Molecular docking of the key target genes with their corresponding active ingredients in PNF was conducted. The 3D protein structures of molecular receptors (key target genes) and molecular ligands (active ingredients) were retrieved from the UniProt database ( https://www.uniprot.org/ ) and the PubChem database ( https://pubchem.ncbi.nlm.nih.gov/ ), respectively. After charge calculations using Mgtools ( https://github.com/mgtools/ ), the receptors and ligands were uploaded to AutoDock Vina software (v 1.2.5) 41 for docking, with binding free energy calculations performed. The detailed information regarding protein and ligand was provided in the Supplementary Table S3 . The resulting complexes were visualized using PyMol software (v 3.1.1) 42 . RNAs from tissue samples of five rats in the control, model, and test groups were isolated using TRIzol reagent (Ambion, USA) at Zhejiang Provincial People’s Hospital. RNA integrity was confirmed via electrophoresis, and purity was assessed using a NanoPhotometer N50. cDNA synthesis was performed with the SweScript First Strand cDNA Synthesis Kit (Servicebio, China). Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) was conducted using the Universal Blue SYBR Green qPCR Master Mix (Servicebio, China) on a CFX Connect RT-qPCR detection system (BIO-RAD, USA). Primers for key target genes (Bhmt, Esr2, and Scn10a) and the internal reference gene GAPDH are listed in Supplementary Table S4 . Detailed reverse transcription reaction conditions are provided in Supplementary Table S5 and S6, while RT-qPCR reaction conditions and program settings are shown in Supplementary Table S7 and S8.The RNA concentration is shown in Supplementary Table S9 . The dissociation amplification curve is shown in Supplementary Fig. S5 . The relative gene expression was quantified using the 2 −ΔΔCт method. Statistical analysis of group differences was performed using GraphPad Prism 5, with P-values calculated and presented. This study was approved by the Ethics Committee of Zhejiang Provincial People’s Hospital, following the ARRIVE guidelines (Approval date: May 4, 2023, No. KT2023016). And all methods were performed in accordance with the relevant quidelines and regulations. Bioinformatics analyses were conducted using R language (v 4.2.2). The Wilcoxon rank sum test was employed to evaluate group differences, with significance set at P  < 0.05.

This study focuses on the therapeutic effect of PNF on chronic endometritis (CE) and explored its underlying mechanism. Then, based on bioinformatics methods such as transcriptomics and network pharmacology, we further screened the key target genes (Bhmt, Scn10a and Esr2) of PNF based on the active components and target genes of PNF. These three key targets were significantly down-regulated in rats samples compared with the disease group as verified by RT-qPCR. It provides a new theoretical basis and scientific reference for the potential mechanism of PNF in the treatment of CE. Key target genes of biological functions and pathways and potential regulatory mechanisms were explored.RT-qPCR validation demonstrated a significant downregulation of these three targets in rats samples compared to the disease group. These findings offer a novel theoretical framework and scientific reference for understanding PNF’s potential mechanism in CE treatment. The study further explores the biological functions, pathways, and regulatory mechanisms of these key target genes. Bhmt (betaine-homocysteine methyltransferase), a member of the methyltransferase family, plays a pivotal role in the one-carbon metabolism cycle 6 .The relationship between CE and Bhmt has not been extensively studied, but there are some underlying mechanisms that can be discussed. Gene single-nucleotide polymorphisms and DNA methylation levels of BHMT are associated with folate therapy efficacy for Hyperhomocysteinaemia(HHcy). DNA methylation of BHMT and BHMT_1 mediated 34.84% and 33.06% of the genotype rs3733890’s effect on the efficacy of folate therapy. During pregnancy, the BHMT pathway is affected by folate status and by the variant BHMT c.716 A allele 6 .This maybe improves the intrauterine metabolic pathway in patients with chronic endometritis.Previous studies have shown that Bhmt induces fat expansion by stimulating the p38 mitogen-activated protein kinase (MAPK) /Smad pathway 7 . Also in the study of endometrial tissue, the molecular mechanisms regulating angiogenesis and VEGF expression have been relevant to the p38 MAPK signaling pathway 8 .In our study, PNF treatment resulted in a significant decrease in BHMT expression levels in CE samples, suggesting that PNF may target the BHMT-mediated methylation pathway. This in turn affects the occurrence and development of CE. Therefore, BHMT may serve as a potential biomarker and therapeutic target for this disease. The Scn10a gene encodes the Nav1.8 sodium channel 9 .Expression of Nav1.8, can affect pain transmission and thus mediate the human pain phenotype 10 . This may elucidate a mechanism by which PNF is used in the treatment of CE with persistent pelvic pain. The sodium (Na+) leak channel, nonselective (NALCN) is upregulated by P4 and downregulated by E2 in human myometrial cell lines during the quiescent period of pregnancy.NALCN may play an important role in maintaining resting membrane potential 11 . Based on this, it can be speculated that the gene Scn10a by modulating sodium ion channels affects the pathological development of CE.In our study, the key target genes were enriched in the voltage-gated sodium channel complex, and PNF treatment resulted in a significant decrease in the expression level of Scn10a in CE samples, suggesting that PNF treatment may affect the pathological progression of CE by regulating the expression level of Scn10a, affecting the voltage-gated sodium channel complex pathway. Therefore, SCN10A may have important clinical significance in the pathological mechanism of CE, providing potential targets and strategies for clinical treatment. The gene Esr2 encodes estrogen receptor 2, a member of the nuclear receptor family that mediates transcription factor activity, through which estrogen influences target cells 12 .Stromal ESR1-regulated genes in the mouse uterus included several growth factors and cytokines, which are potential factors that regulate epithelial and stromal tissue interaction, and also genes involved in lipid homeostasis 13 . The effects of estrogen are mediated primarily via nuclear estrogen receptors, including ERα and ERβ, which are encoded by ESR1 and ESR2 14 . Previous studies have suggested that estrogen receptor 2 (ESR2)-selective agonists might be therapeutic in a rodent endometriosis model 15 . And ESR2 interacts with the inflammasome complex and cytoplasmic apoptotic machinery to enhance the proliferative and adhesive activities of endometriotic tissues 16 . Studies have shown that the estrogens can inhibit the activated inflammation and immune response biomarker genes; and the ESR1 and ESR2, the signaling receptors of estrogens, were identified in the viral-host signaling network and intensively interact with other signaling targets on the network 12 . Based on this, it can be speculated that PNF may change the activity of estrogen signaling pathway by affecting the expression of Esr2, thereby having a therapeutic effect on CE. Therefore, PNF may achieve an effective treatment for CE by comprehensively regulating the expression and function of three key target genes. In the future, if the relevant therapeutic strategies target the three key target genes at the same time, it may have a synergistic effect and enhance the efficacy. This multi-target approach may be more effective than a single-target therapeutic strategy. ESR2-driven effects impair glycemic homeostasis. Estrogens can suppress inflammatory and immune response biomarker gene activation.Additionally, ESR1 and ESR2, estrogen signaling receptors, are implicated in the viral-host signaling network, interacting extensively with other network targets 17 .Based on these findings, it is plausible that PNF modulates estrogen signaling activity by regulating Esr2 expression, thus providing a therapeutic effect for CE. In summary, PNF may provide an effective CE treatment by regulating the expression and function of three key target genes. Although these genes are distributed on different chromosomes, their functions suggest the possibility of multiplexing synergies. Future therapeutic strategies targeting all three genes simultaneously could yield synergistic effects, enhancing treatment efficacy. This multi-target approach might surpass the efficacy of single-target strategies. Network pharmacology and molecular docking analyses reveal that Esr2 serves as a drug target for Baicalin, Erycibenin_D, Garbanzol, Apigenin, and Quercetin, while Scn10a is a potential drug target for Ergosterol. Baicalin demonstrates significant anti-inflammatory activity in treating inflammatory bowel disease, rheumatoid arthritis, autoimmune hepatitis, and chronic liver disease. It has been shown to alter macrophage polarization (M1/M2) and modulate the JAK/STAT pathway, reducing myocardial ischemia-reperfusion injury and inflammation 18 . Apigenin protects against LPS-induced endometritis through Nrf2 pathway activation and NF-kB pathway inhibition 19 . These findings underscore the potential of PNF for CE-targeted therapy. KEGG pathway analysis showed that the key target genes were significantly enriched in PRL signaling pathway, GnRH secretion, and glycine, serine, and threonine metabolism pathways. PRL plays critical roles in pregnancy, mammary gland development, immune regulation, reproduction, and islet cell differentiation. PRL exerts its effects by binding to its receptor, PRLR, a member of the class I cytokine receptor superfamily lacking intrinsic kinase activity. Upon ligand binding, PRLR activates downstream signaling cascades, including JAK-STAT, AKT, and MAPK pathways, thereby promoting cell proliferation, stemness, migration, apoptosis resistance, and chemoresistance. PRL signaling is frequently upregulated in various hormone-dependent malignancies, such as breast, prostate, ovarian, and endometrial cancers 20 .GnRH and its analogues have the potential to treat CE-related diseases, especially in the context of the need to reduce gonadal steroid production, and to be effective in alleviating disease symptoms 21 . Protein kinases are categorized based on their target amino acid residues, including serine/threonine, tyrosine, aspartic/glutamic acid, histidine, and mixed-specificity kinases. Among these, serine/threonine kinases are central to numerous essential cellular processes. Notably, TGF-β signaling, mediated by the serine/threonine activity of its receptors, along with Rho kinase (ROCK), contributes significantly to the pathogenesis and progression of fibrosis across multiple human diseases, including systemic sclerosis (SSc) 22 . Go pathway analysis showed that the key target genes were significantly enriched in voltage-gated sodium channel complex, estrogen response element binding and nuclear steroid receptor activity. Associated with many human diseases, the voltage-gated sodium channel complex regulates the flow of sodium ions through specific molecular mechanisms in response to changes in membrane potential, which is essential for physiological processes such as nerve signaling and muscle contraction 23 .estrogen response element binding regulates the binding of steroid receptors to their DNA response elements, influencing the effects of estrogen and vitamin D, which in turn induce a resistance response to these hormones in the body 24 .Nuclear steroid receptor activity may play an important role in the occurrence and progression of cancer and other diseases by activating gene expression and interacting with coactivators to regulate cellular physiological functions 25 . These insights suggest that PNF may exert therapeutic effects on CE by modulating key gene expression, thereby influencing pathway-specific functions and downstream signaling events. This study employs a series of bioinformatics analyses to elucidate the potential mechanisms underlying PNF’s action in CE, offering valuable perspectives for clinical application and therapeutic development. Nonetheless, there are some limitations to this study. First, although clarifying the role of a single component in a PNF formulation is critical to elucidating synergies and key active compounds, due to experimental limitations, we have not yet systematically evaluated the independent contribution of each component to transcriptomic changes, which may affect the full elucidation of therapeutic mechanisms. Second, the stability of the internal reference gene has not been verified, which may introduce quantitative bias. Therefore, we plan to determine the active ingredient in PNF in future studies, compare the effect and toxicity characteristics of a single component of PNF with the complete formulation through in vitro and in vivo models, and use a multi-reference gene validation strategy to improve data reliability. Furthermore, robust clinical validation supported by extensive evidence-based research is essential to establish the practical applicability of these targets.

Chronic endometritis (CE) is a latent inflammatory condition of the endometrium, primarily caused by Escherichia coli ( E. coli ), Mycobacterium pyogenes , and Staphylococcus aureus 1 .Patients with CE are usually asymptomatic or present with nonspecific clinical symptoms, such as chronic lower abdominal pain, menstrual irregularities, dyspareunia, abnormal uterine bleeding, vaginitis, recurrent cystitis, and mild gastrointestinal discomfort. In severe cases, sepsis, adnexitis, etc., and infertility can be induced 2 . If the disease is not treated in time, it will have a great impact on the health of the patient. At present, CE is often treated with antibiotics such as metronidazole tablets in clinical practice 3 . However, the widespread use of antibiotics can lead to increased drug resistance, which may affect the effectiveness of treatment. Therefore, it is necessary to further study the underlying molecular mechanisms in the development of CE to discover new biomarkers that may reflect the entire endometrial microenvironment and provide a new reference for the clinical treatment of CE. Traditional Chinese Medicine (TCM) is a unique medical system with a holistic approach and combination therapy at its core, and its formulas work synergistically through multi-ingredient and multi-target synergies. In recent years, the therapeutic potential of traditional Chinese medicine in CE treatment has gained increasing recognition. Baogong decoction (BGD) is based on Dang Gui ( Radix Angelicae Sinensis ), Huang Qi ( Radix Astragali ), Bai Zhu ( Rhizoma Atractylodis Macrocephalae ) and other medicinal materials, and has shown clear efficacy in the treatment of chronic endometritis (CE) 4 . Penning Formula (PNF), derived from Song’s gynecology, consists of Da Xue Teng ( Caulis Sargentodoxae ), Bai Jiang Cao ( Herba Patriniae ), Bai Hua She She Cao ( Herba Hedyotis ), San Leng ( Rhizoma Sparganii ), E Zhu ( Rhizoma Curcumae ), Zao Jiao Ci ( Spina Gleditsiae ), Lu Lu Tong ( Fructus Liquidambaris ), Yan Hu Suo ( Rhizoma Corydalis ), Tu Bie Chong ( Eupolyphaga Seu Steleophaga ), Jiu Xiang Chong ( Aspongopus ), Dan Shen ( Radix Salviae Miltiorrhizae ), Dang Gui ( Radix Angelicae Sinensis ), Chuan Xiong ( Rhizoma Ligustici Chuanxiong ), Shan Yao ( Rhizoma Dioscoreae ), Chao Bai Zhu ( Rhizoma Atractylodis Macrocephalae ) ( stir-fried ), Huang Qi ( Radix Astragali ), which is ptimized and expanded on the basis of BGD. The main ingredients are Da Xue Teng, Bai Jiang Cao and Bai Hua She She Cao, which have the effects of reducing swelling and dampness, clearing heat and detoxifying; Others are auxiliary drugs, Dang Gui and Huang Qi can replenish qi and blood, strengthen the right and dispel evil; Tu Bie Chong, Jiu Xiang Chong, Yan Hu Suo, Dan Shen, and Chuan Xiong can relieve pain, invigorate blood and chase away stasis, and San Leng and E Zhu can break blood and qi, eliminate accumulation and relieve pain; Lu Lu Tong and Zao Jiao Ci can improve blood circulation, relieve rheumatic pain, promote water metabolism and reduce edema; Chao Bai Zhu and Shan Yao can take care of the spleen and stomach. The combination of various medicines can clear away heat and detoxify, invigorate blood and relieve pain. Previous research demonstrated that PNF is highly effective in treating CE, outperforming antibiotic therapy. Its therapeutic mechanismin uterine tissue likely involves inhibition of TLR4 signaling pathway activation and downregulation of downstream factors such as MyD88 and NF-κB 5 . However, the exact molecular mechanism and key target genes of PNF in alleviating CE have not yet been elucidated. Therefore, this study aimed to explore the potential mechanisms of PNF in CE treatment by integrating transcriptomics and network pharmacology to identify PNF’s active components and key target genes in a CE rat model. Based on these target genes, the study further elucidated the potential mechanisms of PNF’s action and proposed novel targets and strategies for the clinical management of CE.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4 Supplementary Material 4 Supplementary Material 5 Supplementary Material 5 Supplementary Material 6 Supplementary Material 6 Supplementary Material 7 Supplementary Material 7 Supplementary Material 8 Supplementary Material 8 Supplementary Material 9 Supplementary Material 9 Supplementary Material 10 Supplementary Material 10 Supplementary Material 11 Supplementary Material 11 Supplementary Material 12 Supplementary Material 12 Supplementary Material 13 Supplementary Material 13 Supplementary Material 14 Supplementary Material 14 Supplementary Material 15 Supplementary Material 15