2.1. Preparation of YCHD

The YCHD is composed of three herbs: capillaris, gardenia, and rhubarb [22]. The three active ingredients were supplied by Nanjing Pharmaceutical Materials Co., Ltd. (Nanjing, China). The preparation process involves the following steps: capillaris (18 g), gardenia (12 g), and rhubarb (6 g) were weighed and immersed in 360 mL of distilled water (1:10, w/v) for 12 h. Following this, the mixture was subjected to boiling twice for 1.5 h. The resulting extract was mixed and filtered through filter paper to remove particulates. The filtrate was concentrated to a volume of 36 mL, yielding a mixture containing 1 g of crude drug per milliliter [22]. For in vivo assay, the powder was dissolved in water and orally administered to rats at 12 g/kg/day (1.2 mL/100 g) [24].

2.2. Clinical Specimen Collection

Clinical serum samples were collected from healthy pregnant women (control) and patients with ICP at the Affiliated Nanjing Hospital of Nanjing University of Chinese Medicine. Written informed consent was obtained from all the participants. The study design was approved by the Ethics Committee of the Affiliated Nanjing Hospital of Nanjing University of Chinese Medicine (Approval No. 2021-LY-ky016). In total nine serum samples from women in late pregnancy were collected: three from normal pregnant women and six from ICP diagnosed women, of which three were mild and three severe cases. The diagnosis of ICP was based on obviously elevated TBAs, mild to moderate elevation of liver function during pregnancy, and cholestasis that returned quickly to normal levels after delivery. According to the “Clinical Diagnosis, Treatment, and Management Guidelines for Intrahepatic Cholestasis of Pregnancy (2024)” by the Obstetrics and Gynecology Branch of the Chinese Medical Association [25], the ICP division is as follows: mild: (1) fasting serum TBA levels in pregnant women range from 10 to 39 μmol/L or postprandial serum TBA levels from 19 to 39 μmol/L and (2) the clinical symptoms are mainly skin itching, with no apparent other symptoms, and severe: (1) pregnant woman’s serum TBA levels range from 40 to 99 μmol/L; (2) her serum bilirubin level is higher than the average value; (3) it is accompanied by other conditions, such as multiple pregnancies, preeclampsia, recurrent ICP, or perinatal death caused by ICP; and (4) early-onset ICP. All the clinical serum samples of pregnant women were rapidly frozen, stored in liquid nitrogen, and preserved at − 80°C for further analysis.

2.3. In Vitro Cell Culture and Establishment of ICP Model

HTR-8/SVneo cells were obtained from ATCC (number CRL-3271). The cells were cultured in RPMI-1640 medium (Procell) containing 10% fetal bovine serum (FBS) (Procell) and 1% penicillin/streptomycin (Procell). To establish an in vitro model of ICP, the cells were subsequently exposed to 100 μmol/L of taurocholic acid (TCA) for 24 h [26], with the experimental design segregated into two groups: HTR-8/SVneo group (control) and HTR-8/SVneo + ICP group. All the cells were propagated at 37°C in an incubator containing 5% carbon dioxide. The cell viability and proliferation rates were assessed via a cell counting kit-8 (CCK-8) (BS350b, Biosharp) according to the manufacturer’s instructions.

2.4. In Vivo ICP Model and Therapeutic Intervention

Then, 15 pregnant SD rats were purchased from Animal Biotech Industries, housed in a temperature-controlled room with a 12-h light/dark cycle and 60%–65% humidity, and allowed to feed freely. All protocols of animal experiments were conducted in accordance with the Animal Care and Use Committee of the Nanjing University of Chinese Medicine (Approval No. ACU230301). The rats were randomly divided into five groups, with three rats in each group: NC (normal, intraperitoneal injection 2.5 mg/kg physiological saline, days for 5 days); ICP (intraperitoneal injection of estradiol benzoate (EB) 2.5 mg/kg/day for 5 days) [24]; ICP + NC (intraperitoneal injection of EB 2.5 mg/kg/day for 5 days and tail vein injection of 0.1 mg/kg physiological saline for 5 days); ICP + miR-370-3p antagomir (intraperitoneal injection of EB 2.5 mg/kg/day for 5 days + tail vein injection of the miR-370-3p antagomir 0.1 mg/kg for 5 days); and ICP + YCHD (intraperitoneal injection of EB 2.5 mg/kg for 5 days + YCHD orally administered at 12 g/kg/day (1.2 mL/100 g) for 5 days). Serum samples of rats were collected from the tail vein blood before and after injection to measure the levels of ALT, AST, alkaline phosphatase (ALP), and TBA via common biochemical kits (JianCheng Bioengineering Institute). After completing all procedures, pregnant rats were anesthetized via intraperitoneal injection of 3% (w/v) sodium pentobarbital (30 mg/mL in sterile saline; P3761, Sigma-Aldrich) at 100 mg/kg. Deep anesthesia was confirmed by loss of pedal reflex (no response to toe pinch for > 60 s). Euthanasia was then performed via cervical dislocation under sustained anesthesia, and death was confirmed by the absence of corneal reflex and cessation of heartbeat for > 5 min. This protocol complied with the AVMA Guidelines for Euthanasia of Animals (2020 edition). Immediately after confirming rats death, a midline laparotomy was performed to expose the abdominal cavity. A piece of liver tissue was immediately removed and fixed in 4% paraformaldehyde (PFA) (AS1018, Aspen Biotechnology) for the preparation of paraffin-embedded tissue sections, followed by standard HE staining. The cut placental tissue was thoroughly rinsed with precooled phosphate-buffered saline (PBS) (GNM20012, GENOM), and the sterile dry gauze was used to absorb water. After treatment with RNase-free water, it was quickly placed in liquid nitrogen and preserved at −80°C for further analysis.

2.5. Exosome Isolation and Purification

Exosomes in the culture medium (HyClone) or human and rat sera were separated according to the operating instructions in the Total Exosome Isolation Kit (UR52136, Umibio). Initially, an appropriate volume of ExoQuick solution (Umibio) was added to the samples, which were then incubated overnight at 4°C. Subsequently, the samples were centrifuged at 10,000 × g for 60 min at 4°C to sediment the exosome-containing fractions. The supernatant was discarded, and the sediment was resuspended in PBS at a ratio of V_PBS : V_initial = 1 : 5. The resuspended solution was transferred to a fresh centrifuge tube and centrifuged again at 12,000 × g for 2 min at 4°C. Finally, the supernatant containing the exosome particles was collected and subsequently loaded into the upper chamber of the exosome purification filter (EPF) column. The column was centrifuged at 3000 × g for 10 min at 4°C to purify exosomes. The purified exosomes were aliquoted into 50-μL aliquots and stored at −80°C for future use.

2.6. Electron Microscopy and Nanoparticle Tracking Analysis (NTA)

The extracted exosomes were dissolved in 2% PFA (P6148, SIGMA) and added 5 μL onto a Formvar carbon-loaded copper mesh (01753-F, PELCO). Subsequently, 100 μL of PBS was dispensed onto the parafilm, and the copper mesh was placed on the PBS droplet using tweezers for rinsing. The mesh was then placed in 1% glutaraldehyde solution (16051, Ted-pella) for 5 min and washed eight times in ddH₂O₂ (E130-01A, Novoprotein) for 1 min. The mesh was then exposed to phosphotungstic acid (PTA) (XW120679911, Sinopharm) at a pH of 7 for 5 min and methylcellulose (M0512, SIGMA) on ice for 10 min. Upon completion of these steps, the copper mesh was left to air-dry for 5–10 min before being stored in a box and used for electron microscopy imaging at 80 kV (1230, JEOL).

After dilution with PBS, the extracted exosome samples were introduced into the sample chamber. The NTA (ZetaView PMX 110, Particle Metrix) instrument was calibrated with 110-nm polystyrene microspheres (400168-110, Microtrac) prior to analysis. Subsequently, the diluted exosome samples were subjected to detection using the NTA.

2.7. Bioinformatics Analysis

The miRNA components were profiled via high-throughput RNA sequencing analysis (Illumina) in normal pregnant women (control), mild ICP, and severe ICP groups. The samples were sent to Xiamen Life Interconnect Technology Co., Ltd. (Illumina NovaSeq 6000, PE150 sequencing mode). Differentially expressed miRNAs were identified among the three groups. Subsequently, target genes associated with the identified miRNA were screened using the bioinformatics database starBase.

2.8. Dual-Luciferase Reporter Assay

When the density of HTR-8/SVneo cells cultured in the medium reaches 80%–90%, plasmid transfection was performed. Specifically, the wild-type (WT) or mutant-type (MUT) fragments of TM9SF4 and KIT were constructed into the pGL6-miR gene plasmid (D2106, ELK Biotechnology), with TM9SF4/KIT 3 ^′UTR (MUT) serving as a control. These constructs, along with miR-370-3p mimics, were transfected into HTR-8/SVneo cells using Lipofectamine 2000 Transfection Reagent (Invitrogen), according to the manufacturer’s instructions. Subsequently, the fluorescence intensity of the dual-luciferase reporter assay was measured to validate the targeting relationship between miR-370-3p, KIT, and TM9SF4. The sequences used for plasmid construction are listed in Table S1.

2.9. Western Blot Analysis

The exosomes were cleaved using RIPA buffer (AS1004, Aspen Biotechnology) containing a protease inhibitor. An SDS-PAGE kit (AS1012, Aspen Biotechnology) was used for protein separation from purified exosomes extracted from human and rat sera and culture media. After electrophoresis, the proteins on the SDS-PAGE gel were transferred onto a polyvinylidene fluoride (PVDF) membrane (IPVH00010, Millipore), blocked in 5% skimmed milk powder (AS1033, Aspen Biotechnology), and incubated with primary antibodies against CD9 (ab236630, Proteintech), CD63 (25682-1-AP, Proteintech), CD81 (27855-1-AP, Proteintech), TM9SF4 (ab253000, Abcam), KIT (18696-1-AP, Proteintech), or β-actin (TDY051, Abcam). After three TBST (TBS with Tween-20) (AS1100, Aspen Biotechnology) washes, the membranes were incubated with secondary antibodies for 30 min. As directed by the manufacturer, protein signals were detected via the ECL method and analyzed via the AlphaEaseFC software processing system.

2.10. RNA Extraction and miRNA qRT-PCR Analysis

TRIpure Total RNA Extraction Reagent (EP013, ELK Biotechnology) was used to extract total RNA from the human and rat sera, HTR-8/SVneo cells, and rat placental tissues. EntiLink™ 1^st Strand cDNA Synthesis Kit (EQ003, ELK Biotechnology) was used for the synthesis of the first strand cDNA. qRT-PCR was conducted according to the instructions of EnTurbo SYBR Green PCR SuperMix (EQ001, ELK Biotechnology) with the QuantStudio 6 Flex System PCR System (Life Technologies). The fold change was expressed via the 2^−ΔΔCt method. The internal reference for mRNAs was β-actin (ELK Biotechnology). The internal parameters of the miRNAs were U6 (ELK Biotechnology). Stem-loop reverse transcription (RT) primers of miRNAs and primers for PCR are listed in Table S1.

2.11. miRNA Mimic or Inhibitor Transfection

Transient transfection of miR-370-3p was achieved by transfection with miRNA oligonucleotides (miR-370-3p mimics, inhibitor, and corresponding negative control were synthesized by RiboBio, China). According to the manufacturer’s instructions, miR-370-3p mimics, inhibitors, and corresponding negative control were transfected into HTR-8/SVneo cells using the Lipofectamine 2000 Transfection Reagent (Invitrogen). In vitro models were divided into HTR-8/SVneo, HTR-8/SVneo+ICP, HTR-8/SVneo+ICP + inhibitor NC, HTR-8/SVneo+ICP + miR-370-3p inhibitor, HTR-8/SVneo+ICP + mimic NC, and HTR-8/SVneo+ICP + miR-370-3p mimics. The miR-370-3p mimics, inhibitor, and corresponding negative control sequences are listed in Table S1.

2.12. RNA-Pull Down Assay

Biotinylated miR-370-3p was constructed to generate biotin-miR-370-3p (miR-370-3p probes), and biotinylated miR-NC (NC probes) was used as a control. The cells were divided into HTR-8/SVneo+ICP + NC and HTR-8/SVneo+ICP + miR-370-3p probes, based on the ICP cell models. The cell extracts were incubated with biotinylated RNAs, followed by the addition of agarose beads (Invitrogen) for 1  h. We assess the correlation between miR-370-3p and both TM9SF4 and KIT using an RNA-Binding Protein Immunoprecipitation Kit (Magna RIP). The probe sequences are listed in Table S1.

2.13. Statistical Analysis

SPSS software (Version 23.0) was used for the statistical analysis. The results are presented as the means ± SDs from three independent experiments. Differences among groups were evaluated via unpaired Student’s t test or one-way ANOVA. A p value of less than 0.05 indicated a statistically significant difference.

3.1. Enrichment of miR-370-3p in ICP-Secreted Exosomes Is Positively Correlated With Disease Severity

In our initial experiment, we successfully isolated exosomes from serum samples of individuals in the control, mild, and severe ICP groups. We subsequently analyzed the exosomes via electron microscopy and NTA. Our observations revealed a distinct presence of highly condensed particles within the exosomes, displaying diameters spanning from 30 to 100 nm (Figure 1a,b). These findings confirm that patients with ICP secrete exosomes into their serum. Furthermore, we assessed the abundance of secretory protein markers, including CD9, CD63, and CD81, using western blot analysis. Interestingly, our analysis revealed no notable variations in the levels of these markers among the different ICP severity groups (Figure 1c).

Moreover, we employed a bioinformatic tool to search for miRNAs that are pivotal to serum secretion. Our analysis identified multiple miRNAs within serum-derived exosomes from patients with ICP (Figure 1d). Subsequently, qRT-PCR analysis revealed three clinical comparison groups of differentially expressed miRNAs: miR-370-3p, miR-323b-3p, miR-1469, and miR-10b-5p. Notably, miR-370-3p emerged as the most significantly differentially expressed miRNA among these, prompting its selection as the primary target miRNA for subsequent in-depth investigation (Figure 1e).

3.2. TM9SF4 and KIT Directly Interact With miR-370-3p

Using the bioinformatics tool starBase, we predicted the potential binding sites of miR-370-3p and identified TM9SF4 and KIT as candidate targets for further investigation. To validate these predictions, we used a dual-luciferase reporter assay. Our results demonstrated that the miR-370-3p mimics significantly reduced luciferase activity in constructs containing the wild-type 3 ^′UTR of both TM9SF4 and KIT (3 ^′UTR TM9SF4-WT and 3 ^′UTR KIT-WT) compared to the mimics NC and NC groups (Figure 2). Conversely, no notable change in luciferase activity was observed in constructs harboring the mutated 3 ^′UTR of TM9SF4 and KIT (3 ^′UTR TM9SF4-MUT and 3 ^′UTR KIT-MUT), confirming that miR-370-3p directly targets TM9SF4 and KIT.

3.3. miR-370-3p Overexpression and Low TM9SF4/KIT Expression in ICP-Secreting HTR-8/SVneo Cells

We isolated exosomes from both normal and ICP-secreting HTR-8/SVneo cells (Figures 3a, 3b, and 3c) and subsequently analyzed the expression of miR-370-3p in these cells using qRT-PCR. Our results showed that ICP-secreting HTR-8/SVneo cells exhibited significantly higher levels of miR-370-3p compared to normal HTR-8/SVneo cells (Figure 3d,e). Furthermore, qRT-PCR and western blot assays revealed that ICP-secreting cells had reduced the expressions of TM9SF4 and KIT compared to normal cells (Figures 3f, 3g, and 3h). To validate the direct interaction between miR-370-3p and its targets, TM9SF4 and KIT, we performed an RNA pull-down assay. This assay confirmed that cells transfected with miR-370-3p probes significantly upregulated TM9SF4 and KIT expressions compared to cells transfected with NC probes (Figure 3i). These findings suggest that miR-370-3p plays a role in ICP by regulating TM9SF4 and KIT expressions.

3.4. miR-370-3p Mediated Downregulation of TM9SF4 and KIT in ICP Cell Models

Quantitative analysis of cell viability and proliferation dynamics revealed marked suppression of cellular growth in ICP-modeled HTR-8/SVneo cells, whereas cells treated with miR-370-3p inhibitors exhibited a lesser degree of reductions (Figure 4a). Notably, transfection with miR-370-3p mimics induced the most significant attenuation of proliferative capacity compared to other experimental groups (Figure 4b).

We successfully isolated and characterized exosomes from various sets of HTR-8/SVneo cells (Figures 5a, 5b, and 5c). We then evaluated the expression levels of miR-370-3p, TM9SF4, and KIT in six groups of these cells using qRT-PCR and western blotting. miR-370-3p was overexpressed in ICP-secreting HTR-8/SVneo cells, with maximum expression observed with miR-370-3p mimics. In contrast, HTR-8/SVneo cells transfected with an miR-370-3p inhibitor exhibited a decrease in miR-370-3p levels compared to those in the inhibitor NC group (Figure 5d,e). Further validation revealed that the inhibition of miR-370-3p led to a significant increase in both TM9SF4 and KIT expressions in ICP-secreting HTR-8/SVneo cells compared to the inhibitor NC group (Figures 5f, 5g, and 5h). These findings confirmed that miR-370-3p exerts a negative regulatory effect on the expressions of TM9SF4 and KIT in ICP-secreting HTR-8/SVneo cells.

3.5. YCHD: Modulating miR-370-3p via the TM9SF4/KIT Pathway to Protect Against ICP

An ICP animal model was established through in vivo administration of 2.5 mg/kg EB daily for 5 consecutive days [24], which led to significant elevations in the rat serum levels of ALT, AST, ALP, and TBA. To investigate potential therapeutic interventions, we employed miR-370-3p antagomir, a chemically modified RNA oligonucleotide designed to specifically inhibit the function of miR-370-3p. Our results demonstrate that treatment with both YCHD and miR-370-3p antagomir resulted in a significant improvement of these elevated biochemical markers, suggesting their potential therapeutic effects against ICP-induced liver injury (Figures 6a, 6b, 6c, and 6d). This finding underscores the importance of miR-370-3p in the pathogenesis of ICP-related liver injury.

Before injection: This denotes the baseline conditions of five groups of rats prior to any treatment intervention. After injection: The NC group received intraperitoneal injections of 2.5 mg/kg physiological saline for 5 days, while the ICP groups underwent intraperitoneal injections of 2.5 mg/kg EB daily for 5 days to establish ICP models. This serves as the reference point for comparing subsequent treatment effects. After treatment: This denotes the final outcome after completion of the entire treatment protocol in five groups of rats.

The observation of liver slices from pregnant rats via light microscopy revealed no significant morphological changes in the livers of the NC group, and the hepatic lobule structure was clear. The arrangement of liver cells in the ICP group was slightly disordered, with some liver cells exhibiting vacuolar and granular degeneration and a small amount of necrosis (Figure 7a,b,b). To substantiate the correlation between miR-370-3p and YCHD in ICP, the liver slice results further revealed that the ICP + miR-370-3p antagomir and ICP + YCHD groups presented fewer liver cell lesions than did the ICP + NC group, indicating the therapeutic potential of YCHD in ICP management (Figure 7c, 7d, and 7e).

Similarly, exosomes were collected from rat serum and identified via a previously described method (Figures 8a, 8b, and 8c). We measured the levels of miR-370-3p, TM9SF4, and KIT in the serum and placental tissues of ICP rats. Our data revealed that miR-370-3p expression increased in the ICP model group, whereas KIT and TM9SF4 mRNA levels decreased. The YCHD and miR-370-3p antagomir groups reversed these trends, with decreased miR-370-3p expression and restored TM9SF4 and KIT levels (Figure 8d, 8e, 8f, 8g, and 8h). Thus, miR-370-3p exerts a pivotal regulatory influence on TM9SF4 and KIT, as demonstrated by genetic and proteomic analyses. YCHD regulates miR-370-3p during ICP progression by targeting the TM9SF4/KIT pathway, thereby conferring protection against ICP via the miR-370-3p/TM9SF4/KIT axis.

4.1. Limitations

While this study provides valuable insights into the therapeutic potential of YCHD in ICP, several limitations should be acknowledged. First, as a multicomponent herbal formulation, YCHD’s pharmacological effects may extend beyond the identified miR-370-3p/TM9SF4/KIT-mTOR axis due to potential synergistic or antagonistic interactions among its bioactive compounds (e.g., quercetin, chlorogenic acid, and geniposidic acid). Future network pharmacology and systems biology approaches will be essential to comprehensively map YCHD’s polypharmacology and identify potential off-target effects. Second, the experimental findings were derived from a rodent ICP model, which may not fully recapitulate the heterogeneity of human ICP pathophysiology. Future clinical studies involving larger patient cohorts are required to validate the translational relevance of our results, particularly regarding dose optimization and interindividual variability in treatment response. Third, while the selected YCHD dose (12 g/kg) was based on prior pharmacodynamic evidence, potential long-term safety concerns and tissue-specific effects remain to be explored. Addressing these limitations will refine our understanding of YCHD’s role in ICP management and guide its clinical application.

4.2. Future Perspectives and Challenges

These findings lay a foundation for translating YCHD’s therapeutic potential into clinical practice for ICP management. Future research should prioritize the following steps: First, dose optimization studies in nonhuman primates or human organoids are warranted to bridge the gap between rodent models and clinical applications, particularly given the complexity of YCHD’s multicomponent nature. Second, multicenter clinical trials should evaluate YCHD’s efficacy and safety in ICP patients, with stratification based on disease severity and biomarker profiles. Third, standardization of YCHD preparation is critical to ensure batch-to-batch consistency in bioactive compound ratios, which could be achieved through advanced quality control protocols integrating metabolomics and HPLC fingerprinting.