Ruanjian Sanjie Decoction Induces Apoptosis in Triple-negative Breast Cancer Cells by Downregulating cIAP1/2 and XIAP

Background: Traditional Chinese medicine (TCM) comprises a unique theoretical system developed over thousands of years. The previous study reported that Ruanjian Sanjie (RJSJ) exerts anti-tumor effects by inducing cell apoptosis. However, the mechanism is not clear. Methods: In this study, we investigated the possible mechanism by the strategy of combining network pharmacology analysis with experiment (in vitro and in vivo). First, four kinds of breast cancer cell lines were used to conduct proliferation, apoptosis and cell cycle analysis. Secondly, to study pathophysiological processes of breast cancer at the molecular network level, we for the rst time constructed an “integrated apoptosis module network of breast cancer” by assembling the regulatory relationships of canonical apoptosis signaling pathways. Through the strategy of combining network analysis and experiments, we analyzed the main mechanism of RJSJ in breast cancer and screened out the core genes. We further studied the inhibitory effect of RJSJ combined with carboplatin (CBP) in vivo. Finally, the synergistic effect of RJSJ and CBP were analyzed and the potential active components in RJSJ were predicted. Results: This study demonstrated that RJSJ could signicantly inhibit breast cancer cell proliferation and induce apoptosis in a concentration-dependent manner. The primary mechanism of RJSJ in the treatment of breast cancer was pro-apoptotic. The core apoptosis genes regulated by RJSJ were cIAP1/2 and XIAP. We also found that RJSJ in combination with CBP tended to synergistically induce apoptosis, which might mainly be achieved through the regulation of multiple targets and pathways. Alexandrin (BX05, XKC02, SCG01), baicalin (BX22), guanosine (BX32), arjunglucoside I (XKC10) etc. were predicted as potential active components. Conclusions: These ndings provide the rationale for exploring the therapeutic effects of RJSJ against breast cancer and providing a bridge for the combined use of Chinese and Western medicine.

been illustrated to ameliorate symptoms, improve patient quality of life, and prolong survival [4][5][6][7][8][9] . In 2018, the WHO International Classi cation of Diseases listed traditional medicines, including TCM, in its classi cation system for the rst time 10 . TCM, an important component of complementary and alternative medicine, has gradually developed its own unique system of theories, diagnostics, and therapies over thousands of years. At present, TCM has become an increasingly attractive source of novel therapeutic agents with signi cant roles in preventing and treating cancer 11 . TCMs can alleviate the side effects of chemotherapy, enhance immunity, and kill cancer cells 12 . For example, Ru ai xiao is an experienced prescription developed by Professor Sun from more than 40 years of clinical practice, which plays a role in controlling the progress of breast cancer patients, improving the quality of life and prolonging survival to a certain extent.
Ruanjian Sanjie decoction (RJSJ) consists of four herbs, namely Ban xia (Pinellia ternata), Xia ku cao (Prunella vulgaris), Shan ci gu (Cremastra appendiculata), and Hai zao (Sargassum pallidum). It has been traditionally used for softening hard lumps and resolving hard tissue masses. Our previous study illustrated that RJSJ has de nite anti-tumor effects in vitro and in vivo. Meanwhile, RJSJ is safe and well tolerated in mice 13,14 . Although RJSJ has been demonstrated to have good curative effects against breast cancer, the fundamental molecular mechanisms have not been systematically explored. The bioactive compounds, potential targets, and related pathways of RJSJ remain unknown. The advancement of analytical tools including systems biology 15 , network biology 16 , and network pharmacology 17,18 potentially offer attractive strategies for elucidating the intricate and holistic mechanisms of Chinese herbal formulas.
In this study, we investigated the pro-apoptotic effects of RJSJ in a series of different breast cancer cell lines, including TNBC cell lines. The results suggested that the anti-tumor effect of RJSJ is mainly related to the induction of apoptosis. TCM-based network pharmacology analysis and experimental veri cation were employed to identify and verify the core targets of RJSJ in inducing apoptosis in breast cancer. In addition, when chemotherapy was combined with RJSJ, synergistic pro-apoptotic effects were observed, which might occur through multiple common or different targets in the "integrated apoptosis module network of breast cancer." Further, the potential active components in RJSJ that may play an important role in regulating apoptosis were predicted.

Preparation of RJSJ
All herbs in the RJSJ formula were obtained from Tianjin Zhong Xin Pharmaceutical Group Corporation Ltd. (Tianjin, China). RJSJ was prepared as a lyophilized dry powder of hot water extracts as described previously 13 . The composition was as follows: Pinellia ternate (300 g), Prunella vulgaris (300 g), Cremastra appendiculata (200 g), and Sargassum pallidum (200 g). Herbs were soaked for 0.5 h and extracted twice with 10-and 8-fold volumes of water at 100 °C for 30 min each. The herbal broths were combined and ltered. Then, the ltrate was centrifuged twice at 12,000 rpm for 30 min. The supernatants were mixed with an equal volume of ethanol, maintained at 4 °C overnight, and centrifuged at 12,000 rpm for 30 min. The resultant supernatants were lyophilized, weighed, and stored at − 40 °C until use. The lyophilized dry powder was dissolved in RPMI-1640 medium (Invitrogen; Thermo Fisher Scienti c, Inc., Waltham, MA, USA) at the desired concentration and ltered through 0.22-µm sterilization lters before use 13,19 .

Cell lines and mice
The human breast cancer cell lines MDA-MB-231, MCF-7, Cal-51, and SKBR-3 were obtained from the American Type Culture Collection (Manassas, VA, USA) through the Cell Resource Center of the Tianjin Cancer Hospital (Tianjin, China). The cells were cultured in RPMI-1640 medium supplemented with 10% fetal calf serum and 2 mM l-glutamine (Gibco; Thermo Fisher Scienti c Inc. 3-(4, 5-Dimethylthiazol-2-yl)-2, 5 diphenyl tetrazolium bromide (MTT) assay Cal-51 and SKBR-3 cells in the logarithmic growth phase were seeded in 96-well plates at a density of 3 × 10 3 /well. Adherent cells were treated with RJSJ at different concentrations for 48 or 72 h in a humidi ed CO 2 incubator at 37 °C. Then, 20 µl of MTT solution were added to each well, and the cells were incubated at 37 °C for 4 h. The medium was aspirated carefully. The formazan crystals were solubilized with 150 µl of dimethyl sulfoxide, and the absorbance was measured at 570 nm to calculate the inhibition rate using the following formula: inhibition rate (% of control) = ([1 − absorbance of test sample]/absorbance of control) × 100%. IC 50 was determined via linear regression analysis using the inhibition rate against the corresponding concentration. Cellular morphology assay MDA-MB-231, MCF-7, Cal-51, and SKBR-3 cells in the logarithmic growth phase were seeded in 12-well plates at a density of 5 × 10 4 -10 × 10 4 /well. Based on the results of the MTT assay and our previous studies 13 , adherent cells were treated with different concentrations of RJSJ for 48 h. Changes in cellular morphology were monitored using a ZEISS Axio Observer inverted microscope (ZEISS, Germany).

Cell apoptosis analysis
Cal-51 and SKBR-3 cells in the logarithmic growth phase were seeded in 12-well plates at a density of 10 × 10 4 /well. After treatment with different concentrations of RJSJ, the cells were collected and washed with PBS. Collected cells resuspended in 500 µl of binding buffer. Then, 5 µl of annexin V-FITC and 5 µl of PI (BD Biosciences, Franklin Lakes, NJ, USA) were added to the buffer, followed by incubation at room temperature for 15 min in the dark. Cell apoptosis was analyzed via ow cytometry (BD FACSCanto II; BD Biosciences) within 1 h.

Cell cycle analysis
MDA-MB-231, MCF-7, Cal-51, and SKBR-3 cells were seeded in 12-well plates and treated with different concentrations of RJSJ for 48 h. After treatment, the cells were collected and washed with PBS. Then, the cells were xed with ice-cold 70% ethanol solution at 4 °C for 24 h. RNase A solution (100 µl) was added, and cells were incubated for 30 min at 37 °C. Finally, 400 µl of PI were added, followed by incubation for 15 min at room temperature. The DNA content was detected via ow cytometry.
Animal studies MDA-MB-231 cells and BALB/c-nude mice were used to construct the tumor-bearing animal model. A total of 1 × 10 7 MDA-MB-231 cells were suspended in 0.1 ml of PBS containing 50% Matrigel (BD Biosciences) and injected into the mammary fat pads of 4-6-week-old female nude mice. Tumor size was measured every other day in two dimensions using a caliper. When the diameter of the induration reached 3-4 mm, 24 mice were randomly divided into four groups according to the mode of administration: control (NS 0.2 ml/10 g via gavage), carboplatin (CBP, 0.1 ml/10 g, 60 mg/kg via intraperitoneal injection once), RJSJ (0.2 ml/10 g, 500 mg/kg via gavage every day), and combination groups (500 mg/kg RJSJ via gavage and 60 mg/kg CBP via intraperitoneal injection). Mice were monitored for 21 days following treatment, and the size of subcutaneous tumors was measured every 4 days using calipers to verify the effects of drugs. The mice were sacri ced 21 days after drug administration, and the tumors were collected for protein extraction. Tumor volumes were calculated using the following formula: (V) = (long diameter × short diameter 2 )/2. The inhibition rate was calculated as follows: inhibition rate (%) = ([1 -V experimental group ]/V control group ) × 100%.
Western blot analysis MDA-MB-231 and Cal-51 cells were collected after manipulation as described in section Cellular morphology assay, and tumor tissues were obtained as described in section Animal studies. These samples were lysed on ice for 30 min to isolate proteins. The protein concentration was determined using a BCA protein assay kit (Beijing Solabel Technology Co., Ltd.). Protein lysates (20 µg) were loaded on 12% SDS-PAGE gels and transferred to PVDF membranes (Merck Millipore, USA). Membranes were blocked in 5% nonfat milk for 1 h, washed with TBST buffer, and incubated with primary antibodies at 4 °C overnight. Next, the membranes were washed and incubated with a secondary antibody at room temperature for 1 h, followed by washing and transfer to ECL solution (Merck Millipore). Protein bands were visualized and detected using GE ImageQuant LAS 500 (GE, Japan). The results were measured using ImageJ software. Antibodies against cellular inhibitor of apoptosis protein 1 (cIAP1, YM3009), cellular inhibitor of apoptosis protein 2 (cIAP2, YM1343, (Immunoway), and X-linked inhibitor of apoptosis protein (XIAP, WL03561, Wanlibio Inc.) were used for immunodetection. RT² Pro ler PCR Array gene expression analysis MCF-7 cells were used to conduct the PCR array (PAHS-012A) after treatment with 500 µg/ml RJSJ or PBS. Mature RNA was isolated using an RNA extraction kit according to the manufacturer's instructions. RNA quality was determined using a spectrophotometer and reverse-transcribed using a cDNA conversion kit. The cDNA was subjected to the real-time RT² Pro ler PCR Array (QIAGEN, Cat. no. PAHS-012Z) in combination with RT² SYBR® Green qPCR Mastermix (Cat. no. 330529). C T values were exported to an Excel le to create a table. This table was then uploaded to http://www.qiagen.com/geneglobe. Samples were assigned to the control or test group. C T values were normalized via automatic selection from a full panel of reference genes. The data analysis web portal calculated the fold change using the ∆∆C T method, in which ∆C T was calculated between the gene of interest and the average of reference genes followed by ∆∆C T calculations (∆C T [test group) − ∆C T [control group]). Fold change was then calculated using the 2 ∆∆CT formula.

Network pharmacology analysis
(1) Molecular database construction. We collected the component information of the four herbs in the RJSJ prescription using the ETCM database 20 . The MetaCore (https://portal.genego.com/), MalaCards (www.malacards.org) 21 , and DisGeNET databases 22 were used to collect the causal genes of breast cancer. To obtain the apoptosis-related gene set de ned using Gene Ontology (GO) terms, we retrieved data from the UniProt database 23 using the following search strategy: goa = apoptosis AND organism = "Homo sapiens (human) [9606]." In addition, the AmiGO 2 database 24 was searched using the term "apoptosis," narrowed using "Genes and gene products associated with GO terms," and ltered using "Homo sapiens" as the organism. By integrating the aforementioned two parts, we obtained the apoptosis gene set. (2) Target shing and analysis. The potential target data of components in RJSJ were obtained from the ETCM database 20 . In addition, to improve the reliability and accuracy of the subsequent data analysis, we also collected the known activity data of the ingredients from the ChEMBL 25 and PubChem databases 26 and then integrated these data. Finally, the compound-target interaction network was constructed using Cytoscape v3.7.1 software 27 . For CBP, the target information was collected from the MetaCore (https://portal.genego.com/), ChEMBL 25 , PubChem 26 , and CTD databases 28 . (3) Pathway enrichment analysis. The MetaCore database was used to conduct the disease and pathway enrichment analysis to identify pathways closely related to breast cancer. (4) Mechanism analysis. Based on the apoptosis pathways closely related to breast cancer, the "integrated apoptosis module network of breast cancer" was constructed. The potential targets of RJSJ or CBP and the causal genes of breast cancer were mapped to analyze the pro-apoptotic mechanism of RJSJ. (5) Hub nodes identi cation. The Cytoscape plugin cytoHubba 29 , which provides 11 topological analysis methods for ranking nodes by their network features, was used to identify important nodes from the "integrated apoptosis module network of breast cancer." By comprehensively considering the information of several topological parameters, the important nodes were selected manually by biology experts. (6) Active compounds identi cation. To screen out the potential active compounds with favorable physicochemical and pharmacokinetic properties, we obtained the solubility and absorption properties of molecules from the ETCM database 20 . ADMET solubility level > 1 and ADMET absorption level > 1 were used to screen the potential active compounds.

Statistical analysis
Statistical analysis was performed using SPSS version 17.0 (SPSS, Inc., Chicago, IL, USA). All data are presented as the mean ± SD. One-way analysis of variance was used to determine statistically signi cant differences between groups. P < 0.05 was considered to indicate a statistically signi cant result.

RJSJ concentration-dependently inhibited breast cancer cell proliferation
In previous studies, we obtained the IC 50 values of RJSJ in MDA-MB-231 and MCF-7 cells as 0.578 and 0.635 mg/ml, respectively 13 . We included Cal-51 and SKBR-3 cells in the analysis to further verify the growth-inhibitory effects of RJSJ. The results demonstrated that the IC 50 values of RJSJ in Cal-51 and SKBR-3 cells were 0.865 ± 0.071 and 1.654 ± 0.170 mg/ml, respectively, after 48 h of exposure and 0.855 ± 0.092 and 1.566 ± 0.188 mg/ml, respectively, after 72 h of exposure ( Fig. 1A-B). This study indicated that the viability of these two cell lines was markedly reduced by RJSJ in a concentration-dependent manner.
Using an inverted microscope, control MDA-MB-231, MCF-7, Cal-51, and SKBR-3 were demonstrated to adhere to the plate, and the adjacent cells fused into pieces. As the RJSJ concentration was increased, the density of cells declined, the cell space increased, and cells began to necrose, indicating that RJSJ inhibited cell growth remarkably (Fig. 1C).

RJSJ induced breast cancer cell apoptosis
The effect of RJSJ on apoptosis in MDA-MB-231 and MCF-7 cells was previously reported by our group 13 . In this study, we conducted apoptosis assays using Cal-51 and SKBR-3 cells. The results indicated that RJSJ also induced apoptosis in these cell lines in a concentration-dependent manner (Fig. 2).

Effect of RJSJ on the cell cycle distribution in breast cancer cells
To explore whether the inhibition of cell proliferation by RJSJ was associated with cell cycle arrest, we examined the cell cycle distribution of MDA-MB-231, MCF-7, Cal-51, and SKBR-3 cells using ow cytometry to analyze the cellular DNA content. The results indicated that RJSJ did not induce cell cycle arrest (Fig. 3), suggesting that the anti-tumor effect of RJSJ is mainly related to the induction of apoptosis.
RJSJ induces apoptosis in breast cancer cells by downregulating cIAP1/2 and XIAP From the ETCM database, we constructed the component dataset for RJSJ. The number of compounds in Ban xia, Xia ku cao, Shan ci gu, and Hai zao are 40, 39, 16, and 27 respectively (Table S1). In total, 276 targets corresponding to components in RJSJ were obtained from the public databases ChEMBL and PubChem. Meanwhile, 495 targets were obtained from the ETCM database. There were 89 overlapping targets in these two target sets (Table S2). The interaction information of "component and target" is presented in Table S3.
Disease biomarker networks were used to conduct the enrichment analysis for RJSJ-related targets (Fig.  S1, Table S4). In agreement with the experimental results in the aforementioned cell lines, the enrichment analysis results also suggested that the regulatory mechanism of RJSJ in breast cancer mainly occurs through the induction of apoptosis.
To elucidate the potential pharmacological mechanisms, especially the apoptosis-related mechanisms of RJSJ in breast cancer, we collected breast cancer genes from the MetaCore, MalaCards 21 , and DisGeNET databases 22 and apoptosis genes from the UniProt 23 and AmiGO 2 databases 24 . In total, 3410 causal genes of breast cancer (Table S5) and 132 apoptosis-related genes were collected (Table S6). The "integrated apoptosis module network of breast cancer" was assembled by integrating canonical apoptosis signaling pathways related to breast cancer in MetaCore (Fig. 4A). Details about this process are presented in Fig. S3. Network topology parameters were calculated using the cytoHubba plugin 29 to evaluate the importance of nodes in the network. Notably, the experimental targets of RJSJ validated in our previous study, namely Bcl-2, survivin, caspase-3, caspase-7, and caspase-9, were highly ranked in this list (Table S7), suggesting that the analysis and calculation could provide potential targets involved in RJSJ treatment that deserve experimental validation.
As forecasted using the aforementioned "integrated apoptosis module network of breast cancer" and demonstrated in our previous study, RJSJ obviously induced apoptosis in MCF-7 cells. To validate the regulatory effects of RJSJ on the core apoptotic targets, MCF-7 cells were used to conduct the RT² Pro ler PCR Array (PAHS-012A) after treatment with 500 µg/ml RJSJ or PBS. RT² Pro ler PCR Arrays are highly reliable and sensitive gene expression pro ling tools for analyzing focused panels of genes in signal transduction, biological processes, or disease research pathways using real-time PCR. The analysis identi ed 27 differentially expressed apoptotic genes (Table S8).
In the RT² Pro ler PCR results of MCF-7 cells without evaluating XIAP (BIRC4), cIAP1 (BIRC2, fold change = − 5.99) and cIAP2 (BIRC3, fold change = − 4.18) were differentially regulated by RJSJ, which is consistent with the aforementioned result of the "integrated apoptosis module network of breast cancer" (Table S7, Fig. 4A) and which was partly validated by our previous study 13 . Additionally, the regulatory relationship in network also revealed the direct regulatory effects of cIAP1/2 and XIAP on caspases. As reported in the literature, cIAP1/2 and XIAP participate in anti-apoptotic mechanisms in various cancer cells including breast cancer cells [30][31][32] . cIAP1/2 and XIAP can bind and effectively inhibit caspase-3, caspase-7, and caspase-9 [33][34][35][36] . Consequently, by comprehensively considering the aforementioned results, cIAP1/2 and XIAP were selected for further experiments. The protein expression of cIAP1/2 and XIAP was detected in two TNBC cell lines. The results indicated that RJSJ inhibited cIAP1/2 and XIAP expression in MDA-MB-231 and Cal-51 cells in a concentration-dependent manner (Fig. 4B).

RJSJ in combination with CBP inhibited transplanted tumor growth in mice
In previous studies, we found that RJSJ inhibited the growth of transplanted tumors, prolonged the survival of Ehrlich ascites carcinoma-bearing mice, and enhanced the effects of 5-uorouracil and doxorubicin 13 . To con rm whether RJSJ has broad-spectrum synergistic, we further studied the inhibitory effect of RJSJ combined with CBP in vivo. The animal weight increased slowly in each group during the entire test period (Fig. 5), but no abnormalities regarding diet consumption, bowel movements, and behavior were noted, indicating that CBP and RJSJ were well tolerated (Fig. 5A). As expected, tumor growth was slower in the three treatment groups than in the control group (Fig. 5B). The tumor growth inhibition rates in the RJSJ, CBP, and combination groups were 26.45, 79.43, and 85.78%, respectively. RJSJ or CBP alone downregulated cIAP1/2 and XIAP, and combined treatment led to further inhibition of their expression (Fig. 5C-D). These results suggest that combining RJSJ with chemotherapy can enhance the tumor response by further inducing apoptosis.
Analysis of the potential synergistic mechanism of RJSJ and CBP in promoting apoptosis in breast cancer A total of 136 CBP targets (Table S9) were collected from the MetaCore, ChEMBL 25 , PubChem 26 , and CTD databases 28 , including 14 apoptosis-related genes. Disease biomarker networks were used to conduct the enrichment analysis for CBP targets in breast cancer (Fig. S2). The results suggested that the regulatory mechanism of CBP in breast cancer mainly involves the induction of apoptosis, which is in line with some published studies 37,38 . Based on the compounds in RJSJ, we obtained 682 targets, 33 of which were also targets of CBP, including ve apoptosis-related genes (Fig. 6A).
To identify potential targets exerting pro-apoptotic effects on breast cancer, we mapped the targets of RJSJ and CBP onto the "integrated apoptosis module network of breast cancer" (Fig. 4A). In total, we screened out 63 targets, among which RJSJ and CBP had 42 and 31 targets, respectively, including 10 common targets (Fig. 6B). The apoptosis module network obviously illustrated that RJSJ and CBP could exert synergetic effects by regulating common targets such as AKT1, CASP3, IL6, and TP53 or different targets in apoptosis pathways related to breast cancer (Fig. 4A).
Based on the 63 targets and their interactions with CBP or compounds in RJSJ, we constructed the compound-target network to further clarify the mechanism (Fig. 6C). Overall, 72 compounds in RJSJ with the potential to induce apoptosis in the "integrated apoptosis module network of breast cancer" were identi ed (Table S10). We screened out 25 potential active compounds (4 in Ban xia, 12 in Xia ku cao, 2 in Shan ci gu, and 7 in Hai zao) with favorable physicochemical and pharmacokinetics properties (ADMET solubility level > 1 and ADMET absorption level > 1).
Further, based on the compound-target network, 30 compounds from RJSJ with potential synergistic effects with CBP on apoptosis with 10 common targets (overlapping part of Fig. 6B) were further identi ed (Fig. 6D). Of these compounds, 13 had good solubility and absorption.

Discussion
In this study, experimental and network pharmacology strategies were combined to study the mechanism of action of RJSJ in breast cancer. Both the experimental data and results of network pharmacology analysis suggested that the induction of apoptosis may be the primary mechanism of the therapeutic effects of RJSJ on breast cancer, which is consistent with the experimental results of a previous study 13 .
To elucidate the mechanism by which RJSJ induces apoptosis, we constructed an "integrated apoptosis module network of breast cancer" for the rst time by assembling the regulatory relationships of canonical apoptosis signaling pathways that are closely related to breast cancer, which systematically concretized the phenotype of disease physiological processes at the molecular network level. The MetaCore database contains high-quality and high-reliability data obtained via manual curation from studies. The regulatory relationship between the nodes has clear mechanisms (such as binding, transcriptional regulation, and phosphorylation) and effects (activation and inhibition). Deepening our understanding of the role of apoptosis in the development of breast cancer has great signi cance. Meanwhile, the integrated network could illuminate the key regulatory genes and pathways and provide a theoretical basis for the development of low-toxicity, high-e cacy drugs, such as combination drugs and multi-target drugs [39][40][41] . This is an e cient strategy for studying the pathophysiological processes of diseases. Subsequently, the potential core apoptotic genes playing important roles in the pro-apoptotic effects of RJSJ on breast cancer cells were screened using integrating network analysis, regulatory relationships, and experimental data in MCF-7 cells. The potential targets of RJSJ, such as Bcl-2, survivin, caspase-3, caspase-7, caspase-9, cIAP1/2, XIAP, AKT (PKB), p53, AKT1, and Bax, were the top ranked genes. Intriguingly, genes such as Bcl-2, survivin, caspase-3, caspase-7, and caspase-9 might be the potential targets of RJSJ, as indicated by our previous study 13 . cIAP1/2 and XIAP can bind and effectively inhibit caspase-3, caspase-7, and caspase-9, which participate in the anti-apoptotic mechanisms in various cancer cells including breast cancer cells [30][31][32][33][34][35][36] . The in uence of RJSJ on these proteins was veri ed both in vitro and in vivo, supporting the e ciency of this strategy. Other hub nodes, such as AKT (PKB), p53, AKT1, and Bax, may be potential targets of RJSJ, but additional studies are required for validation. Generally, the high-quality pathophysiological process network could deepen our understanding of disease physiological processes and provide valuable clues for the discovery of important drug targets.
TCM holds the unique advantages of multi-target and multi-pathway regulation in the treatment of complex diseases with high e cacy, low toxicity, and few side effects [4][5][6][7][8][9] . Previous research illustrated that RJSJ is safe and well tolerated, and it does not induce body weight loss, immune dysregulation, or myelosuppression in mice 13 . In recent years, the combination of Chinese and Western medicines has attracted increasing attention. Reasonable combinations of Chinese and Western medicines can lead to enhanced curative effects, reduced side effects, and lower drug doses [42][43][44][45] . However, improper combinations of Chinese and Western medicines will lead to reduced curative effects and greater side effects. CBP is a broad-spectrum anti-tumor drug that is widely used clinically. However, its side effects include myelosuppression, gastrointestinal reactions, and allergic reactions 46 . Using CBP as an example, we preliminarily studied the effect of RJSJ combined with Western medicine to provide evidence for clinical usage. The results revealed that the combination had no side effects, and the effects of the regimen on tumor volume were not signi cantly different from those of CBP alone. However, RJSJ and CBP in combination synergistically increased the expression of apoptotic proteins, which was consistent with the aforementioned experimental results at cellular level.
To elucidate the synergistic mechanism of RJSJ and CBP, we analyzed their targets at the molecular and network levels. The results illustrated that the synergistic effect of CBP and RJSJ was re ected in the regulation of the "apoptosis module network of breast cancer" and other pathways, such as those related with in ammatory responses (Table S4). The molecular network involved in cancer is extremely complex, and alternative pathways usually exist 47,48 . After blocking one speci c signaling pathway, tumors can activate signal transduction through other pathways, which makes it di cult for single-target drugs to achieve good therapeutic effects 47 . In this study, by mapping the targets of RJSJ and CBP onto the "apoptosis module network of breast cancer," we clearly revealed that they have both common and unique targets. The combination of RJSJ and CBP could increase coverage for targets distributed across multiple apoptotic pathways. This may be the main principle of the synergistic effects of RJSJ and CBP on apoptosis in breast cancer cells.
To further elucidate the material basis of the induction of apoptosis by RJSJ in breast cancer cells, we predicted and screened the active components. We identi ed active components with potential proapoptotic effect in all four plants, which re ects their synergistic effects. This is consistent with the principles of formulating TCM prescriptions. According to the compatibility of "monarch, minister, adjuvant and guide," the entire prescription for a speci c indication can enhance e cacy and reduce toxicity 49 . It has been reported that the four herbs in RJSJ inhibited tumor cell growth and induced apoptosis. For example, when human medullary thyroid carcinoma TT cells treated with Prunella vulgaris, mitochondrial membrane potential was signi cantly decreased, CYCS expression in mitochondria was increased, procaspase-3 expression was reduced, and activated caspase-3 expression was increased 50 .
Although some important results were obtained and preliminary understanding of the pro-apoptotic mechanism for RJSJ in breast cancer was achieved, additional experiments are required to further investigate the biological functions and mechanism of this decoction.

Conclusions
In conclusion, this study demonstrated that RJSJ could signi cantly inhibit breast cancer cell proliferation and induce apoptosis in a concentration-dependent manner. The primary mechanism of RJSJ in the treatment of breast cancer was the induction of apoptosis. Based on the "apoptosis module network of breast cancer," we screened and veri ed that RJSJ could induce apoptosis in breast cancer cells by regulating the core nodes cIAP1/2 and XIAP in vitro and in vivo. We also found that RJSJ in combination with CBP tended to synergistically induce apoptosis, which might mainly be achieved through the regulation of multiple targets and pathways in breast cancer apoptosis networks. Through the molecular network analysis of disease physiological processes, we can systematically understand disease physiological processes at the molecular level, which could lay the foundation for the development of effective, low-toxicity drugs. In addition, through network pharmacology technology, the complex mechanism and material basis of the RJSJ prescription were explained in a relatively clear manner, and its synergistic effect with CBP was preliminarily explained, providing directions for future research and experimental support for monotherapy or combination therapy with RJSJ in the clinical treatment of tumors. Therefore, the anti-tumor activities of RJSJ make it a promising TCM for malignancy. The results both signi cantly improve our understanding of breast cancer and clarify the mechanism of action of RJSJ, which should accelerate the application of TCM in modern medicine and promote the modernization of TCM.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analysed during this study are included in this published article [and its supplementary information les].

Competing interests
The author reports no con icts of interest in this work.    Table S7. cytoHubba score calculated using 11 methods; Table S8 PCR results of gene expression following Ruanjian Sanjie treatment in MCF-7 cells; Table S9. Target set of carboplatin; Table S10. Potential components in Ruanjian Sanjie exerting pro-apoptotic effect on breast cancer cells.