Identification of quality markers and mechanisms of Anzi Tiaochong Fang in the treatment of antiphospholipid syndrome-related recurrent pregnancy loss: chemical analysis, network pharmacology, and in vitro approaches

Abstract

Introduction

Anzi Tiaochong Fang (ATF) is a traditional Chinese medicine (TCM) Fangji widely used to treat antiphospholipid syndrome-related recurrent pregnancy loss (APS-RPL). This study aimed to identify the quality markers and elucidate the mechanisms of ATF in treating APS-RPL.

Methods

Chemical, network pharmacology, and in vitro verification were employed to identify quality markers and mechanisms of ATF. HPLC-MS/MS was used to identify and quantify ATF compounds. APS-RPL targets were identified using databases such as HERB, similarity ensemble approach, PharmMapper, Swiss Target Prediction, Gene Expression Omnibus, Genecards, and DisGeNET. GO and Reactome analyses were conducted using KOBAS-i. In vitro verification was performed using CCK-8, FDA staining, and ELISA.

Results

This study identified 23 compounds and 942 targets, including 132 APS-RPL targets and 42 targets between ATF and APS-RPL. GO analysis demonstrated significant enrichment in cytokine-mediated signaling pathway, positive regulation of angiogenesis, response to hypoxia, inflammatory response, and platelet degranulation. Reactome analysis indicated significant enrichment in Immune System, Cytokine Signaling in the Immune system, Signaling by Interleukins, Platelet activation, signaling and aggregation, and Signaling by VEGF. Core targets identified included VEGFA, ALB, TNF, IL-6, and STAT3, with liquiritigenin, nobiletin, ginsenoside Rb1, and astragalin identified as quality markers. In vitro experiments demonstrated that ATF promoted HTR-8/SVneo cell viability, significantly reduced TNF-α and IL-1β levels, and upregulated IL-6.

Conclusions

These findings contribute to the identification and quantification of potential quality markers and elucidate the molecular mechanisms of ATF, thereby supporting its therapeutic potential in the treatment of APS-RPL.

Introduction

Women have a 5% chance of experiencing recurrent pregnancy loss (RPL). Antiphospholipid syndrome (APS), characterized by persistent antiphospholipid antibodies (aPLs), contributes to vascular thrombosis and/or pregnancy morbidity. APS is a crucial factor in the pathogenesis of RPL [1,2,3]. Inflammation, immune dysfunction, and thrombotic mechanisms characterize APS-related recurrent pregnancy loss (APS-RPL). Innate immune cells, trophoblasts, endothelium, and platelets play crucial roles in this process. Conventional treatment for APS-RPL involves the prophylactic administration of low-dose aspirin and low-molecular-weight heparin. However, this approach can result in adverse effects such as liver and kidney damage and bleeding. Furthermore, despite treatment, many RPL patients continue to experience pregnancy failure [4,5,6].

Anzi Tiaochong Fang (ATF), a Chinese Fangji, has gained popularity in treating APS-RPL. ATF comprises seven different herbs, namely CUSCUTAE SEMEN (CS, seeds of Cuscuta australis R.Br.), DIPSACI RADIX (DR, roots of Dipsacus asper Wall. ex Henry), SPATHOLOBI CAULIS (SC, rattans of Spatholobus suberectus Dunn), TYPHAE POLLEN CARBONISATUS (TP, pollen grains after charcoal of Typha angustifolia L.), GINSENG RADIX ET RHIZOMA (GR, roots and rhizomes of Panax ginseng C. A. Mey.), TAXILLI HERBA (TH, Stem branches with leaves of Taxillus chinensis(DC.)Danser), and CITRI RETICULATAE PERICARPIUM (CR, Trichosanthes rosthorinii Harms of Citrus reticulata Blanco). Clinically, ATF is commonly used in patients with APS-RPL [7]. However, since ATF is a chemical complex and not a monomer, it is challenging to identify and quantify quality markers and specific molecular mechanisms of APS-RPL processing. The absence of quality control measures for Fangji raises concerns about the effectiveness of ATF. Therefore, it is crucial to investigate the potential bioactive markers and mechanisms of ATF.

High performance liquid chromatography-mass spectrometry (HPLC-MS/MS) was employed to analyze the extracts of ATF, providing reliable and reproducible results. Herbal medicines are commonly tested using this approach because it enables greater accuracy in determining their components. To address the challenges posed by multitarget combination drugs, network pharmacology, which integrates systems biology and pharmacology, is a promising approach [8]. Previous research on traditional Chinese medicine (TCM) has primarily focused on individual targets and signaling pathways. However, identifying synergy among TCM compounds has been challenging. By analyzing target-drug and bioinformatics networks from the perspective of multiple targets, network pharmacology offers a promising approach to reveal the complex mechanisms of TCM action and facilitate the development of novel drugs.

Currently, network pharmacology is employed to study Fangji by identifying targets and predicting their functions. This investigation primarily involves analyzing the chemical compounds of each botanical drug in TCM using existing databases. However, Fangji may differ chemically from the original medicinal material due to factors such as extraction and purification processes. We recommend selecting identifiable and quantifiable active metabolites as appropriate quality markers in the final formulation to ensure accurate quality control information. This study determined the quality control characteristics of ATF in terms of safety and effectiveness.

In this study, a strategy that integrated chemical analysis, network pharmacology, molecular docking, and in vitro experiments was used to reveal the potential anti-APS-RPL effects of ATF. First, the main ATF chemical compounds were identified and quantified using HPLC-MS/MS, which laid the material foundation for further research. Next, network pharmacology was performed to predict the active compounds and potential targets of ATF against APS-RPL. To further evaluate the potential mechanism of action of APS-RPL, molecular docking technology was used to predict the affinity of the active compounds to potential targets. Finally, the underlying anti-APS-RPL mechanisms of ATF were further verified by cell counting kit-8 (CCK-8), fluorescein diacetate (FDA) staining, and enzyme-linked immunosorbent assay (ELISA) in vitro. This study is expected to validate that ATF has preventive and therapeutic effects on APS-RPL, further reveal its mechanism of action, and provide new insights into the discovery of therapeutic drugs for APS-RPL.

Figure 1 illustrates the design of the study.

Fig. 1

The workflow of this study

Materials and methods

Chemicals and reagents

We obtained Astragalin(9), Avicularin(8), Sweroside(2), Asperosaponin VI(16), Typhaneoside(5), Ginsenoside Rg1(13), Ginsenoside Rf(15), Ginsenoside Rb1(17), Ginsenoside Rb2(20), Ginsenoside Rc(19), Ginsenoside Rd(21), Hesperidin(10), Nobiletin(23), Narirutin(7), Medicarpin(22), Cianidanol(1), (-)-Epicatechin(3), Ononin(12), Liquiritigenin(14), Beta-sitosterol(11) from Sichuan Victory Biological Technology Co., Ltd. (Chengdu, China). In addition, we obtained Daidzin(4), Rutin(6), and Formononetin(18) from the National Institutes for Food and Drug Control (Beijing, China). For chromatography purposes, we used acetonitrile and formic acid, which we purchased from Merck (Germany). The milli-Q water purification system used in this study was purchased from Aquapro (USA).

Preparation of the sample solution

We performed two extractions of seven medicinal materials using 0.5 L of boiling water according to the prescribed ratio. We then filtered and freeze-dried the extracts and stored them in a refrigerator at − 80℃ to obtain the sample.

We accurately weighed 0.4 g of ATF powder and ultrasonically added it to 10 mL of 50% methanol. The mixture was ultrasonicated at 25℃ for 30 min. To compensate for the weight loss after extraction, we added 50% methanol. For analysis, we centrifuged at 4500 rpm for 10 min and then filtered the supernatant through a 0.22 μm membrane. We stored all solutions at 4℃ until they were used.

HPLC-MS/MS conditions

We used a Shimadzu LC-MS8045 HPLC system with a tandem dipole mass spectrometer (Shimadzu, Japan) to perform liquid chromatography separation and mass spectrometry. The instrument was equipped with an ESI source and a Waters XSelect HSS T3 column (2.1 × 150 mm, 3.5 μm) operating at a flow rate of 0.2 mL/min at a temperature of 30℃. Water (A) and acetonitrile (B) were used as mobile phases and gradient elution was used. Specifically, the program was as follows: 0–10 min, 12–15%(B); 10–52 min, 15–33%(B); 52–72 min, 33–36%(B); 72–80 min,36–44%(B). The injection volume was 1 µL. Mass spectra were recorded in the multiple reaction monitoring mode in both positive and negative ion modes. Ideal mass spectrometry conditions included a detector voltage of 1.88 kV, a nebulizer gas flow rate of 3.0 L/min, a heating gas flow rate of 10.0 L/min, and a drying gas flow rate of 10.0 L/min.

Prediction of potential target genes for the identification of compounds

We used HERB [9] (http://herb.ac.cn/), similarity ensemble approach [10](https://sea.bkslab.org/), PharmMapper [11] (http://www.lilab-ecust.cn/pharmmapper/), and Swiss Target Prediction [12] (http://www.swisstargetprediction.ch/) to predict candidate target genes for 23 chemical compounds identified using HPLC-MS/MS.

Prediction of potential targets for APS-RPL

The Gene Expression Omnibus [13] database (https://www.ncbi.nlm.nih.gov/geo/) offers access to functional genomic data, including gene expression data, chips, and microarrays. Humans were selected for analysis in this study. We used datasets GSE22490 and GSE121950, which contain samples obtained from RPL and induced abortion (IA) cases. To filter differentially expressed genes (DEGs) between RPL and IA, the R language limma package was employed [14]. Statistically significant DEGs with a P value < 0.05 were considered significant in this analysis.

We searched for RPL-related targets using keywords such as “recurrent pregnancy loss,” “recurrent spontaneous abortion,” “recurrent abortion,” and “recurrent miscarriage” on Genecards (https://www.genecards.org/) to identify RPL disease targets. In addition, we searched for APS-related targets using the keyword “antiphospholipid syndrome.” Furthermore, we employed DisGeNET [15] (v7.0) ( https://www.disgenet.org/) to search for RPL-related targets using the keywords “recurrent pregnancy loss” and “recurrent spontaneous abortion.”

Overlap target gene screening of ATF for APS-RPL

To collect more comprehensive targets related to RPL, we merged the targets obtained from RPL using DEGs, Genecards, and DisGeNET to identify potential target genes for RPL. Subsequently, we intersected these targets with APS-related targets. The resulting disease targets of APS-RPL were compared with the potential targets from the identified compounds of ATF and imported into jvenn [16] (http://www.bioinformatics.com.cn/static/others/jvenn/example.html). The overlapping section in the Venn diagram represents the overlapping targets of ATF and APS-RPL.

Gene ontology enrichment analysis and reactome pathway analysis [17]

To perform gene ontology (GO) and Reactome analysis, we employed KOBAS-I [18] (http://kobas.cbi.pku.edu.cn/) to import overlap targets. We selected “Homo sapiens (humans)” as the species and “gene symbol” as the input type for GO and Reactome pathway enrichment analysis. We analyzed the targets at a significance level of P < 0.05. We selected the top 20 ranking items, and the results were visualized and examined using the bioinformatics website (http://www.bioinformatics.com.cn/). The identified compounds and diseases shared targets were enriched using the top 20 Reactome pathway items. We employed the STRING [19] database (https://cn.string-db.org/) to construct a protein–protein interaction (PPI) network of common targets. The lowest threshold of interaction was defined as “highest confidence (> 0.4) ”.

Fangji- herb- compound- disease- target- pathway interaction network analysis

We employed Cytoscape 3.9.1 software to construct an interaction network for Fangji, herbs, compounds, diseases, common targets, and pathways [20]. Each element is represented by a node in the network diagram, and the relationships between them are represented by edges. The Cytoscape software network analysis plugin was employed to analyze and filter the topological properties of the network. Quality markers were scored based on their degree in the network, which is a straightforward approach to characterize node centrality in network analysis.

Molecular docking

Although a probable relationship between metabolites and targets was predicted, the exact role of these interactions remains unclear. To further validate the accuracy of the predictions, computer-assisted molecular docking technology was employed. Molecular docking simulations were conducted to validate the binding between the target and its corresponding compound. We obtained four active compounds of the highest degree and identified five core targets of the highest degree based on the information provided in Sect. “Fangji- Herb- Compound- Disease- Target- Pathway Interaction Network Analysis” of the Methods. Data for constructing receptors targeting macromolecular proteins were obtained from the RCSB Protein Data Bank [21] (PDB, https://www.rcsb.org/). We employed PyMol 2.5 to remove protein hydrates and receptor ligands. For hydrogenation, charge calculation, charge distribution, and atom type specification, we imported the acceptors into AutoDockTools 1.5.7. The resulting files were saved in the “pdbqt” format. PubChem (https://pubchem.ncbi.nlm.nih.gov) was employed for the small-molecule compounds. We minimized the energy required by downloading the 3D structure in Spatial Data Format from PubChem and importing it into ChemBio3D Ultra 14.0. To accomplish this, the minimum RMS Gradient was set to 0.001. Subsequently, the small-molecule compounds were saved in the mol2 format. We imported the optimized small-molecule compounds into AutodockTools 1.5.7 to perform hydrogenation, charge calculation, charge distribution, and rotatable key setting. The compounds were then saved in the ‘pdbqt’ format. AutoDock Vina 1.2.5 was employed to simulate the docking of the receptor with its corresponding compounds. The docking results were analyzed using PyMol 2.5 to determine the mode of interaction.

Experimental verification in vitro

Cell culture

HTR-8/SVneo cells were provided by the Chinese Academy of Sciences Cell Bank (China), and the culture method followed our previous experiment [22].

Cell viability assay

HTR-8/SVneo cells (3 × 10^3 cells/100ul) were seeded in 96-well plates and incubated for 24 h. After the culture medium was discarded, each well was washed with phosphate-buffered saline (PBS, Beyotime, No.C0221A, Shanghai, CHN) and incubated with a culture medium containing various concentrations of ATF for 24 h. Cell viability was determined using CCK-8 (Beyotime, No.C0037, Shanghai, CHN) according to the instructions of the assay kit. The optical density (OD) of each well at 450 nm was measured using a microplate reader (BioTek, ELx800, USA).

FDA staining

Nanjing Detai Bioengineering Co., Ltd. (Nanjing, China) provided recombinant anti-β₂GP I antibodies (Ab) and rabbit IgG (RIgG). HTR-8/SVneo cells were divided into the following groups: control, RIgG, Ab, Ab + 0.8 mg/ml ATF (ATFL), Ab + 1.6 mg/ml ATF (ATFM), and Ab + 3.2 mg/ml (ATFH). Before exposure to Ab (10 µg/mL), the cultures were pretreated for 24 h. RIgG cells were then treated with RIgG (10 µg/mL). Control cells were incubated with equal amounts of medium. Six groups of cells were seeded in 96-well plates and incubated for 24 h. After the culture medium was discarded, each well was added with a solution (10 µg/mL) of FDA (Sigma, USA) and PBS 100µL so that the final concentration of FDA was 5 µg/mL, and the cells were incubated for 10 min. During this process, FDA entered the living cell membrane, and all solutions in all wells were discarded after 10 min. Then 200µL PBS was added to each well, and the PBS was discarded. Last, 200µL PBS was added again, and fluorescence was excited at 488 nm, which was photographed using a fluorescent microscope (TE300, Nikon, Japan).

Enzyme-linked immunosorbent assay (ELISA)

The levels of relative proteins were detected by ELISA. IL-1β, IL-6 and TNF-α were measured using an ELISA kit (Nanjing Jinlin Biotechnology Development Co. LTD). All operations were performed according to the manufacturer’s protocol.

Statistical analysis

GraphPad Prism 10.1.2 software (https://www.graphpad.com/) was used for statistical analysis. All experiments were performed individually, and the results are presented as mean ± standard error of the mean (SEM). Group comparisons were assessed using one-way analysis of variance (ANOVA), and P<0.05 indicated that the difference was statistically significant.

Results

Identification of active chemical metabolites of ATF

To fully control the quality of the ATF water extract, we developed an HPLC-MS/MS method to simultaneously detect 23 chemical compounds. Figure 2; Table 1 show the monitoring ions used for detection and analysis. We identified 23 active chemical markers and selected them for further network analysis.

Fig. 2

The high-performance liquid chromatography-mass spectrometry analysis SIM chromatogram of 23 analytes. (1) Cianidanol; (2) Sweroside; (3) (-)-Epicatechin; (4) Daidzin; (5) Typhaneoside; (6) Rutin; (7) Narirutin; (8) Avicularin; (9) Astragalin; (10) Hesperidin; (11) Beta-sitosterol; (12) Ononin; (13) Ginsenoside Rg1; (14) Liquiritigenin; (15) Ginsenoside Rf; (16) Asperosaponin VI; (17) Ginsenoside Rb1; (18) Formononetin; (19) Ginsenoside Rc; (20) Ginsenoside Rb2; (21) Ginsenoside Rd; (22) Medicarpin; and (23) Nobiletin

Table 1 Information for quantitative analysis from the monitoring ion

Method validation

Linear range

Through regression analysis, we established a linear relationship between the calibration curves using the peak areas and concentrations of each standard metabolite. The ratio between the concentrations of each standard substance and the internal standard substance is represented on the horizontal axis, whereas the ratio between the peak areas of each standard substance and its internal standard substance is represented on the vertical axis. Excellent linearity was observed across a broad concentration range for each calibration curve with a correlation coefficient of 0.9914. The regression equation, correlation coefficient, limit of detection, and limit of quantitation for each metabolite are shown in Table 2.

Table 2 Regression equations, linear range, limit of quantitation (LOQ), and limit of detection (LOD) of 20 detected metabolites

Precision, repeatability, and stability

We assessed the precision, repeatability, and stability of the proposed method. We measured the concentrations of 20 compounds in a mixed standard solution twice a day to assess precision, and the experiment was repeated three consecutive times over 3 days to evaluate variability. The method demonstrated satisfactory precision with a relative standard deviation (RSD) of < 5.00% (Table 3). In addition, we prepared and replicated six solutions in one day, resulting in RSD values ranging from 2.96 to 6.63%. For the analysis of sample solution stability, the RSD values ranged from 2.77 to 12.28% for 0, 2, 4, 8, and 12 h at room temperature. Overall, the proposed method demonstrated excellent precision, repeatability, and stability.

Table 3 Precision, repeatability, stability, and recovery of 20 detected metabolites

Recovery test

We extracted and analyzed the samples using the preparative method described in Methods Sect. “Preparation of the Sample Solution” for the recovery tests. The amounts of each analyte were determined six times using the corresponding standard curve. The recovery rate was computed using the following formula: recovery rate = (total amount detected − initial detection amount) / added amount × 100%. The recovery rate was 97.98–107.93%, and the RSD value was 2.02–8.84%. Table 3 illustrates the high accuracy of the method employed in this study.

Quantification of the compounds

According to Methods Sect. Preparation of the Sample Solution”, we used three ATF lots with unique lot numbers to prepare the test solutions. Sample injection and measurement were conducted under the conditions described in Sect. HPLC-MS/MS Conditions”. We analyzed 20 ATF metabolites and computed their concentrations using our method. The concentrations of these 20 metabolites varied significantly among the different samples, as shown in Table 4. Notably, asperosaponin VI exhibited the highest ATF content, measuring 8627.0909 ± 165.7464 µg/ml. Gao discovered that the upregulation of progesterone receptors and activation of Notch signaling can facilitate depersonalization in RPL [23].

Table 4 Concentrations of the 20 metabolites in ATF

Target Screening

From the Venn diagram (Fig. 3), we obtained 942 potential targets from 23 ATF discriminative elements. In addition, we identified 132 disease targets related to APS-RPL. We also identified 42 overlapping targets between ATF and APS-RPL using jvenn.

Fig. 3

Venn diagram of overlap target genes for ATF and APS-RPL. The blue circle represents the targets of APL, the red circle represents the targets of APS, the green circle represents the targets of 23 quantitative metabolites in ATF. The overlapping regions indicate the shared targets, demonstrating the potential mechanism of action of ATF in treating APS-RPL

Key pathways screening of ATF for treating APS-RPL

Analysis of the top 20 GO of the overlap targets, as illustrated in Fig. 4A, revealed that the biological processes were primarily cytokine-mediated signaling pathway, positive regulation of angiogenesis, response to hypoxia, inflammatory response, platelet degranulation, and 20 other biological processes. Figure 4B illustrates the top 20 findings of the Reactome analysis of the overlap targets. Among them, 42 genes were mainly associated with Immune System, Cytokine Signaling in the Immune system, Signaling by Interleukins, Platelet activation, signaling and aggregation, Signaling by VEGF, and other multiple signaling pathways. This result demonstrates that ATF may be effective in treating APS-RPL through multiple pathways that interact in a complex manner. Notably, 39 common targets were enriched from the top 20 Reactome pathway items.

Fig. 4

Biofunction analysis of targets. (A) The top 20 of gene ontology (GO) biological process analysis. Bubble size indicates the number of targets enriched, and bubble color represents the P-value, indicating statistical significance. (B) The top 20 Reactome pathway enrichment analyses. Bubble size and color follow the same representation as in (A), illustrating the involvement of these pathways in APS-RPL treatment

Fangji- herb- compound- disease- target- pathway interaction network

Figure 5 illustrates the Fangji- herb- compound- disease- target- pathway interaction network. The network comprises 91 nodes (1 Fangji, 7 herbs, 23 active compounds, 1 disease, 39 common targets, and Top 20 Reactome pathways) and 799 edges. This network summarizes the versatility and perplexity of Fangji’s functions and elucidates the mechanism of ATF with multi-metabolite, multi-target, and multi-pathway treatment for APS-RPL. Notably, the top four active compounds, namely liquiritigenin, nobiletin, ginsenoside Rb1, and astragalin, were highly potent and quality markers of ATF in treating APS-RPL. These quality markers in ATF may not only affect a single target gene but also the entire biological network system. Liquiritigenin, nobiletin, ginsenoside Rb1, and astragalin exhibited degrees of 13, 13, 12, and 10, respectively. The core targets, which were the top 5 of 39 common targets, were vascular endothelial growth factor A (VEGFA), tumor necrosis factor (TNF), interleukin-6 (IL-6), albumin (ALB), and signal transducer and activator of transcription 3 (STAT3). Their degrees were 59, 50, 47, 43, and 42, respectively. Table 5 shows information on core targets.

Fig. 5

Fangji- herb- compound- disease- target -pathway interaction network. The green ellipse represents Fangji, the blue hexagon represents herb, the yellow parallelogram represents compound, the orange diamond represents disease, the pink triangle represents target, and the red V represents pathway. Edge size and color indicate edge betweenness, reflecting the interaction strength

Table 5 The information of core targets

Molecular docking visualization

Table 6 shows the binding affinities of the four quality markers to the five core targets, which were discovered to be < − 5 kcal/mol. Figure 6 (A-E) show the simulations of the five highest molecular docking affinities, namely ALB- Liquiritigenin, TNF-Astragalin, STAT3-Ginsenoside Rb1, IL-6-Liquiritigenin, and VEGFA-Ginsenoside Rb1. The lowest binding energy was − 10.0 kcal/mol, from liquiritigenin to ALB, indicating an excellent binding effect between them. ALB interacted with liquiritigenin primarily through hydrogen bonds and hydrophobic interactions. Notably, the hydrogen bond with ALA-158 was 3.8 Å in length, and the hydrophobic interaction was with GLY-189, LEU-154, ILE-142, LEU-139, TYR-138, LEU-135, and TYR-161.

Table 6 Molecular docking affinity of quality markers and core targets in ATF

Fig. 6

Molecular docking simulations of the five highest affinity interactions. (A) ALB- Liquiritigenin. (B) TNF-Astragalin. (C) STAT3-Ginsenoside Rb1. (D) IL6-Liquiritigenin. (E) VEGFA-Ginsenoside Rb1

Experimental verification in vitro

Cell viability of HTR-8/SVneo cells treated with ATF

Cells were seeded in 96-well plates and treated with the specified concentrations (0 (control), 0.8, 1.6, 3.2, 6.4 mg/ml), and CCK-8 was used to detect the effect of ATF on the viability of HTR-8/SVneo cells in vitro. The results showed that ATF significantly increased the viability of HTR-8/SVneo cells in a dose-dependent manner (P<0.05), whereas there was no significant change in viability between the doses of 3.2 and 6.4 mg/ml (Fig. 7A). Therefore, doses of 0.8, 1.6, and 3.2 mg/ml were selected to treat HTR-8/SVneo cells to further study the potential action mechanism of ATF in APS-RPL treatment.

Fig. 7

Experimental verification in Vitro. (A) Cell viability of HTR-8/SVneo cells treated with ATF. **P < 0.01 vs. Control, ****P < 0.0001 vs. Control. (B) Fluorescein staining of β₂GPI-treated HTR-8/SVneo cells. (C-E) ELISA quantification of cytokines IL-1β, IL-6, and TNF-α in treated cells. Data were analyzed using GraphPad Prism 10.1.2, with n = 3 per group

FDA staining

As shown in Fig. 7B, Ab significantly disrupted the cell membrane of HTR-8/SVneo cells, causing the loss of fluorescein from cells, whereas ATF facilitated the cell membrane of Ab-treated HTR-8/SVneo cells, causing the gain of fluorescein from cells.

Validation of IL-1β, IL-6 and TNF-α

ELISA quantified the levels of IL-1β, IL-6 and TNF-α, and the results are shown in Fig. 7C, D and E. The results for IL-1β indicated that while there was an increase in the Ab group compared to the Control group, this difference was not statistically significant. On the other hand, when comparing the Ab group to the ATFL group (P = 0.0078), the ATFM group (P = 0.0016), and the ATFH group (P < 0.0001), the differences were found to be statistically significant. The IL-6 results indicated a statistically significant difference between the Ab group and the Control group (P < 0.0001). Furthermore, comparisons between the Ab group and the ATFL group (P = 0.0064), ATFM group (P = 0.0339), and ATFH group (P < 0.0001) also showed statistically significant differences. The TNF-α results indicated that there was no statistically significant difference between the Ab group (P = 0.3582) and the Control group. Similarly, there was no statistically significant difference between the Ab group and the ATFL group (P = 0.2095). However, a statistically significant difference was observed between the ATFM group (P = 0.0440) and the ATFH group (P = 0.0110) when compared to the Ab group. The in vitro experiments demonstrated that ATF promoted HTR-8/SVneo cell viability, which is critical for placental development. Furthermore, ATF treatment significantly reduced TNF-α and IL-1β levels while upregulating IL-6 production. These cytokine modulations are essential in addressing the inflammatory and immune dysregulation associated with APS-RPL.

Discussion

The unique mechanism of TCM Fangji involves multiple compounds and targets that work through various pathways. Assessing the therapeutic properties of Fangji accurately is impossible by examining only one of its compounds. Thus, ensuring Fangji’s quality decoctions and products presents a significant challenge, necessitating an appropriate approach. At our hospital, ATF is a TCM Fangji frequently combined with low-molecular-weight heparin calcium administered to patients with APS-RPL, which can improve the success rate of abortion and clinical efficacy, increase the negative conversion rate of APS, and improve the indexes of coagulation-fibrinolysis [7]; however, its mechanism is unclear. Furthermore, no reliable quality control approach is in place to ensure its efficacy. HPLC combined with MS has emerged as a standard quality control technique for Chinese herbal medicines. This approach enables qualitative and quantitative determination of major pharmaceutical compounds [24, 25]. However, despite technological advancements, identifying the most significant quality markers among numerous chemical markers remains challenging. Network pharmacology is an effective approach for developing multitarget Fangji [26]. This study employed HPLC-MS/MS and network pharmacology to quantify and identify potential bioactive compounds, clarifying the mechanism by which ATF treats APS-RPL.

To identify 23 compounds in ATF simultaneously, we employed HPLC-MS/MS analysis. Among them, SC contained nine compounds (flavonoids and steroids), GR contained six compounds (saponins), CR contained three compounds (flavonoids), DR contained two compounds (iridoids and saponins), and CS, TH, and TP contained one compound (flavonoids). We screened all Fangji compounds as active metabolites through network pharmacological target screening. Subsequently, the resulting Fangji- Herb- Compound- Disease- Target- Pathway network identified liquiritigenin, nobiletin, ginsenoside Rb1, and astragalin as quality markers of ATF in treating APS-RPL. Many pharmacological studies have demonstrated the potent role of flavonoids and saponins in various physiological and pharmacological processes, including antioxidant, anti-inflammatory, and immunomodulatory properties. Notably, several studies have shown that flavonoids extracted from CS, TH, and DR can effectively prevent miscarriage [27,28,29]. For example, Xia et al. reported the therapeutic effect of asperosaponin VI, a flavonoid from DR, in alleviating RPL. These aforementioned compounds can modulate the expression of key targets such as STAT3, PTGS2, JUN, SRC, and CASP3 in decidual cells [29]. Sun et al. demonstrated that the protective effects of astragalin, a flavonoid from CS, on mitochondrial quality control might be attributed to antioxidant effects [30]. Vineet Babu demonstrated that liquiritigenin, a flavonoid found in SCs, inhibits the production of pro-inflammatory cytokines, such as TNF-α and IL-6, in LPS-activated primary peritoneal macrophages [31]. In addition, Chiara Corrado et al. revealed that nobiletin, a flavonoid found in CR, could effectively counteract the negative impacts of TNF-α on angiogenesis and invasiveness mainly through the VEGF and NF-κB pathways [32]. Furthermore, Su et al. discovered that ginsenoside Rb1, a saponin present in GR, may be beneficial in reducing oxidative damage and inflammation in individuals with diabetes [33]. Although these compounds employed as quality markers of ATF have been investigated in various diseases through antioxidant, anti-inflammatory, and immunomodulatory activities, there is a lack of relevant research focused on APS-RPL. Thus, we aim to conduct studies specifically relevant to this condition.

According to TCM, CS, TH, and DR are believed to invigorate the kidney, a concept that has been applied in the treatment of recurrent pregnancy loss (RPL) [34]. Shoutai pills, which include these three botanical drugs, have been traditionally employed since the Qing Dynasty for treating threatened abortion, and recent studies have reported positive outcomes [35]. Our network pharmacology results have identified astragalin as a significant active compound, primarily associated with the monarch botanical drug. We suggest considering TP and SC as adjunct botanical drugs in ATF, given their potential effects on circulation [36], which could be important in conditions characterized by vascular thrombosis [37], such as APS. The adjuvant botanical drug also serves as a source of one of the quality markers, liquiritigenin. GR and CR are proposed as conductant botanical drugs in the context of ATF. These botanical drugs offer quality markers such as ginsenoside Rb1 and nobiletin. In addition, these botanical drugs are widely recognized for their anti-inflammatory effects and immunomodulatory activities [38,39,40,41]. This study explores potential mechanisms that may be consistent with the traditional TCM principle of ‘monarch-minister-adjuvant-conductant,’ as it has been applied in various TCM formulations. However, further studies are needed to confirm this hypothesis with scientific evidence. As shown in Table 1, the primary bioactive metabolites of ATF in the treatment of APS-RPL are flavonoids and saponins. This information is valuable for developing modern preparations of ATF because it demonstrates that lower dosages can achieve strong efficacy.

The analysis of the bio-functionality of GO demonstrated that the genes targeted by ATF in treating APS-RPL are closely associated with cytokine-mediated signaling pathway, positive regulation of angiogenesis, response to hypoxia, inflammatory response, and platelet degranulation. ATF may regulate various intricate biological processes involved in managing APS-RPL. Reactome pathway enrichment analysis demonstrated that 42 targets overlap, mainly associated with Immune System, Cytokine Signaling in the Immune system, Signaling by Interleukins, Platelet activation, signaling and aggregation, Signaling by VEGF, and other multiple signaling pathways. APS-RPL is an immune system disorder that occurs during pregnancy and is characterized by the presence of aPL. These aPL activate various cells in the body, including trophoblasts, platelets, endothelium, and innate immune cells, leading to endothelium dysfunction, inflammation, and thrombosis. The release of pro-inflammatory cytokines and activation of platelets exacerbate these processes. In addition, aPL can directly target trophoblasts, resulting in impaired implantation and placental insufficiency [5]. Signaling by the VEGF pathway is responsible for signaling angiogenesis, which is the formation of new blood vessels. VEGF serves as a vital growth factor in promoting angiogenesis and plays a crucial role in regulating blood vessel development during early embryonic growth [42]. During normal pregnancy, VEGFA, the first member of the VEGF family, is produced in a hypoxic environment. Its function may include the accumulation of macrophages in the decidua during early pregnancy [43]. Inhibition of STAT3 activation hinders endometrial decidualization and impairs blood vessel formation in the placenta and endometrium, leading to inadequate blood supply to the placenta and subsequent embryo development cessation and abortion. APS-RPL development is associated with inflammatory factors, VEGF, STAT3, and immune imbalance. Based on the pathway enrichment findings, it is suggested that ATF may exert therapeutic effects by targeting these pathways, potentially offering benefits in managing APS-RPL. Another important aspect to consider is the role of lipid rafts, which are small (10–200 nm) heterogeneous membrane domains enriched in glycosphingolipids and cholesterol. These lipid rafts play a crucial role in cellular signaling by serving as platforms for the assembly of signaling molecules. Given the known involvement of lipid rafts in various signaling transduction pathways, it is plausible that ATF could modulate APS-RPL pathogenesis by targeting lipid raft-associated pathways. Recent studies have demonstrated that disruption of lipid rafts by cholesterol-binding agents can reduce the activity of tissue factor (TF), a major initiator of the coagulation cascade, and restore nitric oxide (NO) production, thereby suggesting that lipid raft modulation could play a key role in the pathogenesis of APS [44, 45]. Understanding how ATF interacts with these lipid raft-associated pathways could open new avenues for therapeutic interventions in APS-RPL. This potential mechanism warrants further exploration, and we have highlighted it as a promising direction for future research. Another signaling pathway of interest is the Notch signaling pathway, particularly Notch4, which has been shown to play a significant role in trophoblast functions such as invasion, proliferation, and migration. These processes are crucial for proper placental development [46]. There is evidence suggesting that compromised Notch signaling, particularly in endothelial cells, can lead to placental defects, inadequate lymphangiogenesis, and miscarriages [47, 48]. We have discussed the possibility of ATF influencing Notch4 signaling to enhance vascular and placental function. Relevant studies on Notch4 signaling in lymphatic endothelial cells and animal models suggest that ATF could have a regulatory effect, potentially improving pregnancy outcomes by supporting adequate uterine perfusion and placental health. This aspect of ATF action presents a promising avenue for future exploration in APS-RPL therapy. Further research is needed to explore the precise mechanisms through which ATF modulates these biological processes, and experimental validation is required to confirm these findings. Therefore the pathway enrichment findings support the pathogenesis of APS-RPL, demonstrating that ATF may offer therapeutic benefits by acting on these pathways.

To verify the mechanism of action of ATF for treating APS-RPL, CCK-8, FDA staining, and ELISA were used to conduct corresponding in vitro experiments. CCK-8 was first used to determine the effect of ATF on the viability of HTR-8/SVneo cells. ATF at concentrations from 0.8 to 6.4 mg/ml had a significant effect on cells compared with the control in a dose-dependent manner. In addition, FDA staining showed that ATF could recover the injury of HTR-8/SVneo cells treated by Ab. Last, ELISA was utilized to measure the levels of IL-1β, IL-6, and TNF-α in the supernatant of HTR-8/SVneo cells. The findings indicated that Ab significantly decreased IL-6 expression, while the impact on IL-1β and TNF-α was less pronounced. ATF notably enhanced IL-6 expression, particularly in the ATFH group; it also notably decreased IL-1β and TNF-α levels. IL-6 plays a crucial role in immune tolerance and protecting the fetus against maternal immune rejection [49], a role that our research findings aptly demonstrate. The use of Ab induced damage to HTR-8/SVneo cells that closely mimicked the pathological process of APS-RPL. Conversely, by inhibiting the production of TNF-α and IL-1β and promoting the production of IL-6, ATF may treat APS-RPL. The findings from our in vitro assays demonstrate that ATF’s modulation of cytokines such as TNF-α, IL-1β, and IL-6 can directly impact the inflammatory environment seen in APS-RPL. By reducing key pro-inflammatory cytokines and enhancing anti-inflammatory signals, ATF may help restore immune balance, thereby improving pregnancy outcomes in APS-RPL patients. These findings provide a mechanistic link between ATF’s actions in vitro and its therapeutic potential in managing APS-RPL, bridging the gap between experimental observations and clinical relevance.

This study has some limitations. A limitation of this study is the absence of protein-level validation of the core targets identified. While our results indicate significant changes at the transcriptional and functional levels, validation at the protein level for key targets, such as VEGFA, TNF, IL-6, and STAT3, would provide stronger evidence of the mechanisms of ATF. To further support the proposed mechanisms of action of ATF, future studies should include protein-level assays, such as Western blot or immunohistochemistry, to confirm these findings. This would help in elucidating the full molecular impact of ATF on APS-RPL pathophysiology. Additionally, to assess the in vivo translational potential of these findings, future research should include animal models of APS-RPL to evaluate ATF’s efficacy in regulating inflammatory responses, promoting angiogenesis, and improving pregnancy outcomes. Such in vivo validation would help bridge the gap between preclinical findings and clinical applications, providing a more comprehensive understanding of ATF’s therapeutic potential in APS-RPL.

Conclusion

In conclusion, this study used HPLC-MS analysis, network analysis, molecular docking techniques, and in vitro experiments to successfully identify and quantify quality markers and mechanisms of ATF in treating APS-RPL. This research has established a framework for controlling the quality of ATF and ensuring the safety and effectiveness of APS-RPL in clinical settings. Potential core targets for ATF treatment of APS-RPL include VEGFA, TNF, IL-6, ALB, and STAT3. This study indicated that ATF can regulate various aspects of the Immune System, Cytokine Signaling in the Immune system, Signaling by Interleukins, Platelet activation, signaling and aggregation, Signaling by VEGF, and other multiple signaling pathways. The therapeutic benefits of ATF in APS-RPL may be due to its ability to reduce TNF-α and IL-1β and upregulate IL-6. ATF follow-up can be based on these results to determine clinical indicators for assessing the effectiveness of APS-RPL management. In addition, this study provides new insights into the working mechanisms of ATF and provides a scientific basis for the clinical treatment of APS-RPL. However, further in vivo and in vitro experiments are necessary to definitively identify the major regulatory targets and pathways of ATF. We will elucidate the mechanism of ATF and lay an experimental foundation for the further development of new drugs.