Chronic liver disease can irreversibly lead to cirrhosis, the most common and clinically relevant complication of cirrhosis (Gallo et al., 2020). Dysregulation of multiple intersecting biological pathways is intimately linked to the development of CAS. Current scientific strategies to discover therapeutic approaches for CAS usually focus on a single target: dehydration, so diuretic therapy is the standard traditional treatment method (Li et al., 2023). However, long-term use of diuretics can lead to endocrine disorders, diarrhea, and abdominal pain. Traditional Chinese Medicine (TCM) is multi-targeted and has been used clinically for thousands of years (Gao and Yao, 2019). Hence, exploring bioactive metabolites extracted from classical formulations of TCM with good therapeutic effects is a potentially effective way to discover novel therapeutics for CAS.

SZD, a very famous water-removing formula, is from “Treatise on Febrile Diseases” and has also been found in “Synopsis of the Golden Chamber” (Zhou et al., 2012). It comprises 4 botanical drugs: euphorbia (Euphorbia pekinensis Rupr.), flos genkwa (Daphne genkwa Siebold & Zucc.), gansui (Euphorbia kansui S.L.Liou ex S.B.Ho.), and jujube (Ziziphus jujuba Mill.). It is commonly used clinically to treat CAS and has achieved significant efficacy with few side effects (Zhao et al., 2018). Euphorbia has been widely used to treat urinary retention, malignant ascites, and edema because of its diuretic and decongestive properties. Euphorbia has been included in many formulas, such as the Ji Dai formula and Modified Shi Zao Decoction, for treating malignant ascites (Luo, 2013; He, 2015; He et al., 2017). The extracts of euphorbia have been found to have a variety of activities, such as antiviral and anti-inflammatory effects. Flos genkwa, a traditional laxative and water-removing medicine, is mainly used to treat the syndrome of fluid retained in the chest, hypochondrium, and abdominal fluid (Jiang et al., 2015). Gansuiis a dry root derived from Euphorbia kansui T., which is toxic and commonly used to treat pleural effusion, ascites, and edema (Zhao et al., 2019a). Euphorbia, Euphorbia kansui, and flos genkwa in SZD are toxic to the intestinal tract. Adding jujube can enhance immunity and show network analysis of the anti-inflammatory effects by regulating the gut microbiota and metabolites (Gao et al., 2023).

“Holistic regulation,” which views the organism as a whole, is what distinguishes Chinese herbal medicine. CHM practitioners focus more on the patients with diseases than on the ailments themselves, using the human body’s homeostasis as their guidance. It is challenging for a single medication to meet the needs of patients with variable and complex conditions due to the diversity of pathogenesis and the complexity of patients’ conditions. The TCM formula is derived from the wealth of rich experience in long-term clinical practice by fully utilizing the benefits of TCM differentiation and treatment. A large number of studies have confirmed the multi-target and multi-level characteristics of the mechanism of action of traditional Chinese medicine compounding, which means that modern scientific analysis should be used to study the active ingredients of compounding and the mechanism of compounding action. The characteristics of the mechanism of action of traditional Chinese medicine formulas are very similar to the completeness, systematicity, and comprehensiveness of network pharmacology, so network analysis is very suitable for the study of the pharmacological mechanism of traditional Chinese medicine formulas (Wang et al., 2022; Shang et al., 2023). There are multiple metabolites in TCM formulations. Network analysis can be used to screen potential metabolites and targets, while the potential PRM absorbed into the bloodstream at an effective concentration is a prerequisite for the therapeutic effects of TCM formulation. Therefore, pharmacokinetics revealed the exposure levels of the absorbed RPM from SZD (Zhao et al., 2021). Metabolic markers and pathways that may predict drug responses can be identified by metabolomics (Pang et al., 2019). Subsequently, integrating network analysis, pharmacokinetics, and metabolomics is a practical method to predict pharmacodynamic PRM and explore the effects of PRM.

To our knowledge, pharmacokinetic studies of CAS have never been reported before. In this paper, the metabolites and targets of SZD were screened by network analysis, and pharmacokinetic studies of the PRM of SZD were performed to explore the exposure levels in the blood. Comprehensive metabolomic studies of CAS in serum and urine were used to identify potential biomarkers of CAS. Finally, we explored the possible mechanisms of SZD for treating CAS by integrating pharmacokinetics and metabolomics. These findings contribute to the development of new therapeutic targets and treatments for CAS.

The study confirmed that SZD significantly reduced liver injury and inflammatory cell infiltration in CAS rats, improved liver function, reduced ALT, AST, GLB, and TBIL levels, increased ALB levels, and also reduced NO and ET-1 levels in peripheral blood and portal vein pressure. However, systematic studies on the pharmacokinetics and metabolomics of SZD-treated CAS still need to be completed. In this study, 12 metabolites were screened through network analysis, and all have been found to have anti-CAS pharmacological effects according to relevant literature. Moreover, the pharmacokinetic parameters of the 12 potential PRM of SZD were revealed for the first time after intragastric administration. All these findings shed light on the pathogenesis of CAS and the mechanism of action of the RPM of SZD in the metabolic system of CAS.

The results of the pharmacodynamics showed that SZD could significantly attenuate liver injury and improve liver function in CAS rats. Pharmacokinetic studies indicated PA, CHA, CA, rutin, isovitexin, galuteolin, UL, luteolin, naringenin, apigenin, daphnoretin, and genkwanin have good blood exposure. Studies have shown that these 12 potential PRM of SZD have therapeutic effects on liver diseases such as hepatitis, hepatic cholestasis, cirrhosis, etc. While most chronic liver diseases have the pathological process of hepatitis to hepatic fibrosis to cirrhosis to hepatic carcinoma and cirrhotic ascites is one of the standard clinical manifestations of cirrhosis in its decompensated stage, these 12 potential PRM may hinder the development of cirrhotic ascites by acting on the compensatory stage of cirrhosis. So, we investigated these 12 metabolites. In addition, the metabolites with disordered expression levels were analyzed, which were mainly in the tryptophan metabolism, alpha-linolenic acid metabolism, glycerophospholipid metabolism, pantothenate and CoA biosynthesis, arginine biosynthesis and arginine, and proline metabolism.

High levels of NO production, as an important toxicity executor, play an important role in hepatotoxicity (Harstad and Klaassen, 2002; Hirunpanich et al., 2008). Arginine is a NO originator, and it can generate NO under the action of the NOS enzyme (Morris et al., 2017). To verify the effect of SZD on NO synthesis, we collected rat portal vein blood and detected the NO content. The NO level in the CAS group was significantly higher than in group A, and the NO content could be improved after SZD treatment (Supplementary Figure S4). Isovitexin can inhibit the expression of iNOS in cells and reduce NO production (Lin et al., 2005; Quintieri et al., 2008; Lucas et al., 2010). In the metabolomic study, SZD treatment led to a significant decrease in the arginine metabolite N-nervonoyl arginine. Arginine is a substrate for NO synthesis, and its downregulation can reduce NO production. Also, the changes in the levels of N-nervonoyl arginine were positively correlated with isovitexin, suggesting the mechanism of SZD in treating CAS may be involved in the L-arginine/NO pathway.

Different experimental models have shown glycine’s hepatoprotective properties (Heidari et al., 2016). Glycine decreased ROS and mitochondrial swelling (Pal et al., 2011). In addition, glycine could also restore liver mitochondrial ATP and histopathological changes. In liver cells, glycine can be conjugated with chenodeoxycholic acid to form bile salts, a primary bile acid generated in the liver from cholesterol, and elevated bile acids can cause cholestasis. After cholestasis, plasma contents of ALT, AST, TBIL, and taurocholic acid were increased, and this situation could develop into CAS over time (Zhu et al., 2018), which is the same as our pharmacodynamic study, where the levels of ALT, AST, and TBIL were higher in CAS rats than in healthy rats, and the SZD ameliorated the abnormal levels of ALT, AST, and TBIL. Studies have shown that CHA treatment reduces ALT, AST, and TBIL levels, improves liver pathological changes, and reduces cholestasis through metabolism and efflux of bile acid (Zhu et al., 2018). Apigenin can also regulate the metabolism of bile acid and improve cholestasis by reducing oxidative damage and inflammation (Zheng et al., 2021). Luteolin and rutin also have a protective effect on cholestatic liver injury (Wu et al., 2022). In addition, studies have shown that glycine can be metabolized to phenylacetylglycine, oleoyl glycine, and hippuric acid (De Simone et al., 2021). The metabolomic study in this paper found that the levels of the metabolites oleoyl glycine and hippuric acid in CAS rats were lower than those of the normal group, and phenylacetyl-glycine was higher than those of the normal rats. Phenylacetyl-glycine, oleoyl-glycine, and hippuric acid levers in CAS rats could be improved after treatment with SZD. Furthermore, after treatment with SZD, chenodeoxycholic acid and taurodehydrocholic acid in rat serum were downregulated to prevent liver damage. Also, changes in serum levels of chenodeoxycholic acid and taurodehydrocholic acid after SZD treatment were correlated with rutin, CHA, luteolin, and apigenin (Figure 10). This suggests that SZD may interfere with glycine metabolism.

Interestingly, CYP3A4, CYP1A2, and TNF may be the key targets of SZD in treating CAS through network analysis research, and DHA is closely related to these three targets. Many studies have shown that luteolin (Boeing et al., 2020), PA (Ya et al., 2021), CHA (Shi et al., 2021), CA, UL (Yin et al., 2018), galuteolin (Wang et al., 2021b), naringenin (Motallebi et al., 2022), and apigenin (Li et al., 2021a; Li et al., 2021b) have good antioxidant and anti-inflammatory effects, where CA can be used in combination with DHA to alleviate oxidative stress damage. In this study, the AUC_(0-∞) of luteolin, PA, CHA, CA, galuteolin, naringenin, and apigenin were 1,119.48 ± 193.52, 3,641.33 ± 1725.92, 1724.35 ± 570.50, 1833.80 ± 942.07, 1,492.72 ± 501.32, 4,406.82 ± 1878, and 2,492.03 ± 738.98 ng h/mL, respectively (Table 1). The AUC values were all larger. A larger AUC indicates greater drug absorption and increased bioavailability. Metabolomic results showed that ω-3 fatty acid metabolism (ω-3 PUFA) may be a potential metabolic pathway for SZD-treated CAS because of α-linolenic acid as the main precursor of ω-3 PUFA metabolites. Moreover, ω-3 PUFA can be metabolized to DHA by Cytochrome P450 (CYP) enzymes (Shoieb et al., 2020). In a CAS model, DHA inhibits fiber formation and prevents the development of cirrhosis by interfering with NF-κB and TGFβ signaling pathways. From the references, we know that CAS models are often induced by CCL₄, which upregulates the inflammatory cytokine TNF, while DHA ameliorates TNF (Boyer-Diaz et al., 2020). In conclusion, there was significant DHA depletion in the livers of patients with advanced cirrhosis. Studies have shown that the development of cirrhosis can be mitigated by enhancing antioxidant capacity. Linoleic acid (LA), or dihomolinoleic acid, is the most consumed PUFA in the human diet and can enhance liver cells’ antioxidant capacity and metabolism (Chen et al., 2019; Cai et al., 2022). LA-HPOD can be metabolized into LA in the body, and LA can be metabolized into 9,10-EPOME by CYP enzymes. Enguita et al. evaluated the association between LA and its derived lipid metabolite 9,10-EpOME and biomarkers of liver injury and systemic inflammation and showed that 9,10-EpOME was reduced in patients. In addition, metabolites associated with anti-inflammation in CAS rats, including DHA, alpha-linolenic acid, LA-HPOP, LA, and 9,10-EPOME, were found to be significantly altered by metabolomics studies, and administration of SZD significantly improved the changes of these metabolites. Interestingly, positive or negative correlations were found between alpha-linolenic acid, LA, DHA, LA-HPOD, and 9,10-EPOME and eight absorbed SZD pharmacokinetics markers, including luteolin, PA, CHA, CA, UL, galuteolin, naringenin, and apigenin (Figure 10). In summary, we believe that the occurrence of CAS may be related to inflammation caused by disorders of linolenic and linoleic acid metabolism. These results indicate that SZD may increase the metabolism of linolenic acid to produce DHA by promoting CYP enzymes, thus restoring the normal level of DHA in CAS rats and achieving therapeutic effects.

Through network analysis research, it was hypothesized that hepatitis B might be a key signaling pathway for SZD to treat CAS. HBV may cause cirrhosis or even liver cancer. Besides, some experiments have demonstrated that the RPM of SZD has obvious anti-HBV effects. Through pharmacokinetic studies, the T_1/2 and T_max of luteolin were the shortest, at approximately 4.5 ± 2.61 h and 2.06 ± 0.62 h, respectively, which indicates that luteolin was rapidly absorbed into the bloodstream and reached the site of action. Luteolin inhibits the production of TNF-α, IL-6, and IFN-γ in the blood and has potent anti-HBV effects (Wu et al., 2005; Wu et al., 2022). In addition, studies have shown that PRM PA, rutin, naringenin, isovitexin, CA, and CHA also exhibit anti-HBV effects (Dai et al., 2017; Jose-Abrego et al., 2022). HBV infection tends to lead to metabolic disorders, in which tryptophan metabolism can produce indoxyl sulfate and kynurenic acid (Kyn). Kyn can inhibit T-cell function and lead to T-cell apoptosis. This reaction puts the body in an immunosuppressed state and severely hinders liver function, so that tryptophan levels are consistently low in CAS patients (Jiang et al., 2020). Our experiment found that, compared with normal rats, CAS rats had higher levels of kynurenic acid and indoxyl sulfate and lower levels of tryptophan. Interestingly, SZD could regulate the levels of tryptophan, kynurenic acid, and indoxyl in CAS rats to return them to normal. Also, the contents of tryptophan, kynurenic acid, and indoxyl sulfate in serum and urine after SZD treatment were positively correlated with those of luteolin, isovitexin, and CHA, while indoxyl sulfate was negatively correlated with CA, PA, naringenin, and rutin (Figure 10). Based on the above results, it can be hypothesized that the PRM of SZD may act on the HBV signaling pathway by binding to the TNF target and interfering with the metabolism of tryptophan to exert anti-HBV effects and stop the progression of HBV to CAS.

Through network analysis, we hypothesized that the PI3K-Akt signaling pathway may play an essential role in SZD to treat CAS (Figure 1; Figure 2). In addition, many studies have shown that CHA, galuteolin (Ho et al., 2021), rutin, and genkwanin (Kim et al., 2023) can exert antioxidant effects by activating this pathway. CHA also prevents CCL₄-induced liver fibrosis by suppressing oxidative stress in the liver and hepatic stellate cells (Shi et al., 2021). Our pharmacokinetic results revealed a significant bimodal peak after CHA absorption into the blood, possibly caused by reabsorption from the enterohepatic circulation (Xu et al., 2020). The enterohepatic circulation of CHA can prolong the half-life and maintenance time of the drug, which is generally beneficial for exerting its therapeutic effect. Through metabolomics research, we found that the content of pantothenic acid in CAS rats was lower than that in the normal group. However, SZD can improve the metabolism of pantothenic acid. Studies have shown that pantothenic acid can prevent CCL₄-induced hepatotoxicity and oxidative stress (Eidi et al., 2012). In addition, pantothenic acid reduces oxidative stress levels mainly through the ROS-mediated PI3K-Akt signaling pathway, which may be the reason why pantothenic acid is resistant to CAS (Ya et al., 2021). Also, pantothenic acid has a positive correlation with the four active PRM of SZD, including rutin, galuteolin, and genkwanin, while CHA is negatively correlated with pantothenic acid (Figure 10). These demonstrate that SZD exerts its therapeutic effects by modulating pantothenic acid biosynthesis.

In this article, we refer to the studies in recent years that only combined network analysis and metabolomics. Our article identifies the potential RPM of SZD anti-CAS through network analysis and analyzes their migration in the bloodstream via pharmacokinetics. Additionally, an integrated network of metabolomics and network analysis was constructed using Cytoscape, ultimately hypothesizing that the key targets might be CYP3A4 and CYP1A2. This study preliminarily explored the possible mechanisms of SZD anti-CAS. In our future research, we will consider utilizing Western blot or PCR methods to validate the signaling pathways that were identified by network analysis. This would enhance the reliability of our results.

In this study, we first established a new integrated strategy to investigate the key targets and mechanisms of SZD in treating CAS. The results revealed that SZD can reduce liver tissue inflammation, inhibit collagen fiber hyperplasia, and improve liver function. In addition, we speculate that the effective RPM of SZD may have significant regulatory effects on the Hepatitis B signaling pathway, PI3K-Akt signaling pathway, tryptophan metabolism, arginine biosynthesis, alpha-linolenic acid metabolism, pantothenic acid biosynthesis, arginine, and proline metabolism through key targets. The data and theories provided in this study help to study the mechanism in depth and lay the foundation for clinical application.