Pharmacological Mechanism of Pingxiao Formula against Colorectal Cancer

Colorectal cancer (CRC) is the most common cancer worldwide and develops due to a broad range of causative factors. Pingxiao (PX) formula and Xihuang (XH) formula are two commonly used drugs to treat CRC, especially as an alternative therapy for those patients who could not suffer surgery, chemotherapy, or immunotherapy, namely, elder or advanced CRC patients. However, the pertinent pharmacological mechanisms are still elusive. The investigation was designed to explain the pharmacological mechanisms of the PX formula. A murine model of CRC was established by injecting CT26.WT cells into the caecum of 4-week-old male Balb/c mice, following PX or XH treatment for 30 days. Network pharmacology analysis combined with weighted gene coexpression network analysis (WGCNA) predicted the pharmacological mechanisms and therapeutic value. High-throughput 16S rRNA sequencing determined the alterations in the gut microbiota communities. Western blotting, immunofluorescence, and flow cytometry examined the influence of PX on the tumor microenvironment (TME). Injection of CT26.WT-induced CRC in Balb/c mice was markedly attenuated by PX treatment. Compared with XH administration, PX exhibited a stronger antitumor effect, such as smaller tumor volume, lower interleukin 17 (IL-17), IL-6 and tumor necrosis factor-alpha (TNFα) serum levels, and higher interferon-gamma (IFN-γ) concentration. Network pharmacology analysis demonstrated that both PX and XH targets were enriched in cancers and inflammatory responses. RNA sequencing confirmed that PX treatment induced cancer cell apoptosis and inhibited inflammatory reactions within the tumor. Moreover, the PX formula considerably restored homeostasis of the gut microbiota, which was not observed in the XH group. PX targets, those associated with the survival probability of CRC patients, correlated with macrophage (Mφ) infiltration, which presented an independent risk factor for the CRC outcome. PX treatment promoted the transition of alternatively activated Mφs (M2 Mφs) to classically activated Mφs (M1 Mφs). Moreover, the peritoneal Mφs from the PX group inhibited the migration of CW26.WT cells, as evidenced by the wound healing experiment and transwell assay, which was consistent with the decreased expression of the vascular endothelial growth factor (VEGF). Furthermore, the coculturing system confirmed that PX-treated Mφs suppressed colorectal tumor-derived organoid proliferation. PX formula exhibits a potential antitumor effect against CRC by suppressing the colonization of pathological microorganisms, reshaping Mφ effector functions and hence inhibiting cancer cell proliferation.


Introduction
Colorectal cancer (CRC) is one of the most common cancers and the third most frequent cause of cancer-related deaths, and CRC-caused mortality has increased by 40% with the development of economics during the decades. Despite the access to specialists and efective therapy in recent years, it has been projected that there has been a rise in the number of young patients as well as subsequent deaths [1][2][3]. Te pathogenesis of CRC involves the environment, genetic susceptibility, nutrition, gut microbiota, and the natural history of infammatory bowel disease [4]. Surgical resection is the only curative choice for patients with CRC, followed by adjuvant chemotherapy or immunotherapy, especially for those who have clinically occult micrometastasis. However, for patients who are not suitable for surgery or long-period chemotherapy, namely, elderly or terminal cancer patients, adjuvant drug therapy is the backbone of anticancer treatment [5]. In China, traditional Chinese medicine (TCM) is widely administered after surgical resection to avoid tumor recurrence and metastasis due to its lower cost and lower toxicity than other adjuvant options. Especially for elderly patients with advanced CRC, TCM has been considered the only palliative therapy. More importantly, given that the pathogenesis of CRC, a TCM formula, mostly composed of several ingredients that are anticancer, immunity-promoting, and anti-infammatory, is promising to modulate the tumor microenvironment (TME), which determines tumor progression and outcomes. Multiple traditional Chinese anticancer drugs are utilized in the clinic to combat various cancer types including Pingxiao (PX) formula, Xihuang (XH) formula, and Xiangshaliujunzi decoction, stated Su et al. [6].
Te PX formula is an anticancer drug and encompasses alunite, Toxicodendron vernicifuum  Table 1). Te PX formula is an improved formula to protect against blood stasis, tubercles, and tumors by a Chinese physician, Jia Kun on the basis of "Xiaoshi-Fanshi-San" in Jin Gui Yao Lue (written by Zhang Zhong Jing, a Chinese pharmacologist of the Eastern Han dynasty) composed of alunite and nitre to treat liver and gall disorders. Its chemical compositions and metabolites in rats have been identifed by untargeted data-dependent tandem mass spectrometry [7], and its efcacy against liver cancer and CRC has been investigated in the clinic [8,9]. However, these studies are limited in subjects and unknown pharmacological mechanisms. Tere is an unmet need for more detailed experimental evidence of its efcacy in treating CRC. In this study, we aimed to delineate the molecular and cellular mechanisms of the PX formula against CRC by assessing the anticancer efect, predicting the therapeutic value in the clinic, and demonstrating its infuence on TME.

Network Analysis.
All the plant names have been checked with https://www.theplantlist.org. Component-target networks were computed by the active components according to the following criteria: 0.18 drug-likeness and 30% oral bioavailability. Te associated genes of CRC were collected from fve databases: GeneCards, Online Mendelian Inheritance in Man (OMIM), DrugBank, PharmGkb, and Statistics of Terapeutic Target Database (TTD). Te targets of active components were obtained from PubChem and the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform [10]. Te ingredient-target interaction network was produced by Cytoscape software. Te data of the protein-protein interaction (PPI) network were obtained from the Search Tool for Retrieval of Interacting Genes/Proteins database with a confdence score ≥0.7 and calculated by CytoNCA [11], sorting the data based on the median values of betweenness centrality (BC), closeness centrality (CC), degree centrality (DC), local average connectivity (LAC), eigenvector centrality (EC), and network centrality (NC).

Clinical Patient Data Acquisition and Consensus
Clustering Analysis. Te mRNA expression profle and clinical information of 426 patients with colon adenocarcinoma (COAD) were downloaded from Te Cancer Genome Atlas (TCGA) database, and weighted gene coexpression network analysis (WGCNA) was performed. Te tumor samples were clustered into two groups based on the mRNA levels of the PX targets by the "limma" and "ConsensusClusterPlus" packages. Te Kaplan-Meier (KM) plotter was plotted to utilize "survival" and "survminer" [12] packages. In the least absolute shrinkage and selection operator (LASSO) regression was computed by package "glmnet" to predict which PX-targets were associated with the survival probability with the penalty regularization parameter (λ). Based on the optimal lambda value as well as the corresponding coefcients, the risk signature genes were obtained and samples were clustered into low-risk and high-risk groups. Te receiving operating characteristic (ROC) curve with the area under the curve (AUC) representing the degree or measure of separability was constructed by "pROC" and "ggplot2" packages to determine the discrimination ability. Te closer to 1 means the validity. Single-sample gene set enrichment analysis (ssGSEA) was performed to identify the infuenced immune cells by PX treatment. In the PX formula, the weight ratio of CL, SS, nitre, AL, CR, alunite, FT, and TV is 5 : 3:3 : 3:3 : 2.5 : 1. Mixed CL, nitre, CR, alunite, FT, and TV powder were extracted by 70% ethanol, and the leftover and AL were boiled and concentrated. AL was added to the dried mixture. A total of 405 g of powder was obtained about 230 g of PX drug. To validate the components, a connected system of the LC-30 (Shimadzu)hybrid quadruple time-of-fight mass spectrometer (TOF-MS) with an electrospray ionization source (ESI) was used. . Data were processed in information-dependent acquisition (IDA) with a high sensitivity mode. XH Wan was bought from the Beijing Tong Ren Tang drug company. 10 5 CT26.WT cells were injected into the caecum of male Balb/c mice (Pengyue Animal Center of Shandong province, China) (male, about 22 g, 2 months old), and on the same day, the mice were treated with PX or XH by oral for the next 30 days. Te dose for a human being is 300 mg/kg per day based on the Chinese Pharmacopoeia 2020, and the correspondingly murine is 3.6 g/kg calculated by the Meeh Rubner formula: skin surface area = mass coefcient * weight * 2/3 * 10 −4 . Te positive control of the study is cyclophosphamide (CTX) (MedChemExpress, China) which facilitates antitumor immunity and suppresses CRC progression [13]. 90 mg (equal to 3.6 g/kg) or 45 mg (equal to 1.8 g/kg) were diluted in PBS and orally given to mice each day. Every 15 mice were grouped, and six groups were set as follows: 0.9% saline group, low concentration of the PXtreated group (1.8 g/kg), high concentration of the PXtreated group (3.6 g/kg), low concentration of the XHtreated group (1.8 g/kg), high concentration of the XHtreated group (3.6 g/kg), and the CTX (200 mg/kg) group. On day 31, isofurane (RWD Life Science, Shenzhen city, China) was used for anesthetization. Tumors were collected and measured.

Microbiota and Microbiome
Analysis. Te feces and mucus within the colon were collected and mixed and processed by 16S rRNA sequencing. Raw data were fltered [14] and produced paired-end reads. Corrected paired-end reads generated circular consensus sequencing (CCS) reads. After removing chimeras, OTU (operational taxonomic unit) analysis was performed with a similarity >97%. Krona, as a powerful metagenomic visualization tool, demonstrates species annotation. Each fan means the corresponding annotation's proportion [15]. PICRUSt (Phylogenetic Investigation of Communities by Reconstruction of Unobserved States) analysis was conducted to predict the alterations in KEGG pathways. Te raw data are available with the accession number in SRA (PRJNA820603, PRJNA796207).
Te feces and mucus sample was treated with methanol and standard internal substances. After the ultrasound and frozen treatment, the samples were centrifugated. Te supernatant was subjected to ultrahigh-performance liquid chromatography (UHPLC) coupled with TOF-MS (Garcia and Barbas, 2011). Acquisition software (Analyst TF 1.7, AB Sciex) was utilized to assess the full scan survey MS data, which were then converted by ProteoWizard and generated the retention time (RT), mass-to-charge ratio (m/z) values, and peak intensity (Salido et al., 2009). According to the Inhouse MS2 database, substances were identifed and determined by orthogonal projections to latent structuresdiscriminant analysis (OPLS-DA). Te cut-of is fold change (FC) > 1, P value <0.05, and variable importance in the projection (VIP) > 1 [16].
2.5. RNA Sequencing. FastQ fles were processed for primary quality control. Read pairs that contain more than 3 or the proportion of bases with a quality value below 5 were removed. All the downstream analyses were completed on the clean data obtained. Diferentially expressed genes (DEGs) were identifed using the DEGSeq R package (PRJNA820958).

Macrophage (Mφ) Culturing and Phagocytosis
Experiment. For peritoneal Mφ preparation, mice were anesthetized and sterilized PBS was injected into the abdominal cavity. After the massage, the liquids were isolated and centrifuged. Te formula was diluted in a 1640 complete medium.
For the bone marrow-derived macrophage (BMDM), 4 week-old male mice were sacrifced and bone marrow from femurs was incubated with blood cell lysis bufer (Beyotime, China) for 5 minutes. Filtered cells were cultured with the 1640 medium in the presence of the granulocytemacrophagecolony-stimulating factor (STEMCELL Technologies, Canada) at a concentration of 55 ng/ml. Mφs were collected on day 7. 10 5 Mφs were cultured with prepared microparticle solution (Termo Fisher, USA) for 2 hours at 37°C and then mounted at the fuorescein isothiocyanate (FITC) wavelength (488 nm).

Colorectal Tumor Organoid Culture and Coculturing.
Te solid tumor was cut into small pieces and washed with cold PBS until the PBS solution became transparent and clear. After 35 minutes of digestion by digestive solution [17], the fltered medium was centrifugated and cultured in a growth medium. After 9 days, colorectal tumor-derived organoids (cTDOs) were cocultured with PX-treated Evidence-Based Complementary and Alternative Medicine peritoneal Mφs. After 24 hours, cTDOs were incubated at 37°C with propidium iodide (PI)/Hoechst33342 (480 nm/ 630 nm) for 20 minutes. For Ki67 staining, the Ki67 eFluor ™ 660 eBioscience ™ (Termo Fisher, USA) (1 : 1000) antibody was incubated for 24 hours and examined by using a fuorescent microscope (Zeiss, Japan).

Western
Blotting. Te total protein (40 μg) was subjected to SDS-PAGE (sodium dodecyl sulfate-polyacrylamide gel electrophoresis) and transferred. After blockage, diluted antibodies (1 : 1000) (BAX, Ki67, Arg1, Tgfβ1) were incubated with the membrane for 24 hours. After washing with PBS, secondary antibodies conjugated with horseradish peroxidase (HRP) were utilized and an enhanced chemiluminescence (ECL) substrate was added to visualize the protein bands. All chemicals were bought from Invitrogen, USA.

Transwell Assay and Wound Healing Experiment.
CT26.WT cells were added to a transwell insert (Corning, 8 μm diameter, USA) in the presence of PX serum for 24 hours. After the fxation by 4% paraformaldehyde (PFA), the cut insert membrane was subjected to DAPI (4′,6-diamidino-2-phenylindole) staining and examined by using the microscope. 10 5 CT26.WT cells were seeded in a 24-well plate at a density of 10 5 at 37°C in a 5% CO 2 incubator. When it reached 90% confuency, a scratch was made. After 24 hours, the migrated distance was measured. All chemicals were bought from Invitrogen, USA.

Statistical Analysis.
Data are means ± SEM. An unpaired Student's t-test was used to compare the diference between the two groups. A Cox proportional hazard model was established to indicate PX targets associated with the survival probability in COAD patients. A hazard ratio (HR) with 95% confdence intervals and a log-rank P value were calculated via univariate survival analysis, and a P value of <0.05 was considered statistically signifcant. Table 1) and shared 230 putative targets with a total of 10352-CRC genes (Figures 1(a) and 1(b)) (Supplementary Table 2). A HIT network representing the interactions between the active components of the PX formula and targets was established: pink rectangle nodes were the shared genes, and nodes inside represented active components (Figure 1(c)). As illustrated in the pie chart of target proportion, AL occupied 43% of the formula (Supplementary Figure 2A). KEGG analysis predicted the enriched pathways involved in the formula (Supplementary Figure 2B) and each herb (Supplementary Figure 2C) combating CRC.

HIT Network of PX and XH. Te PX ingredients had 171 active compounds (Supplementary
Te HIT network of XH was also established. Musk, Commiphora myrrha (Nees) Engl., Bovis calculus, and Boswellia sacra Flueck., shared 179 genes with CRC, and Myrrha was responsible for 78% of the targets (Supplementary Figure 6). KEGG analysis predicted the enriched pathways (Supplementary Figure 7).

PX-Targeted Genes Correlate with the Survival Probability in CRC Patients.
Utilizing consensus clustering, a total of 426 patients with COAD were divided into two clusters, which showed a correlation with the survival probability ( Figure 2(a)). Te DEGs between the two clusters were enriched in infammatory reactions, pathogen-recognition pathways including Toll-like receptor and NOD-like receptor signaling pathways, and infectious diseases (Figure 2 Table 3). Te schematic of the experimental design is shown (Figure 3(a)). As illustrated, the PX formula suppressed the tumor volume at both concentrations and exhibited a stronger tumorsuppressing capacity than cyclophosphamide (CTX) and XH (Figures 3(b) and 3(c)). Moreover, we observed that the high dose of PX mitigated the weight loss of CRC mice (Figure 3(d)). Te high dose of PX was selected for further experiments. Moreover, the high dose of PX or CTX signifcantly hampered the concentrations of IL-6, IL-17, and TNFα, whereas it increased IFNc secretion, whereas in this case, XH failed (Figure 3(d)). Altogether, PX is more potential than the XH formula.

PX Formula Promotes Gut Immunity and Induces
Colorectal Cancer Cell Apoptosis. Solid tumor tissue was collected from CRC mice with or without PX treatment. Saline and PX groups clustered distinctly from each other (Figure 4(a)). Among the diferentially expressed genes   (DEGs) (Figure 4(b)), there were 24 up-regulated genes and 384 down-regulated genes after PX administration, which were enriched in multiple cancers, cell adhesion molecules, the Toll and Imd signaling pathways, and cytochrome P450. Specifcally, the apoptosis-related gene BAX was signifcantly increased after PX treatment, while Naip1, Pparg, PYCARD, Tnfaip8, and Gadd45a were decreased (Figure 4(d)).

Fecal Microbiota and Metabolite Analysis.
Given the predicted antibacterial efects of PX and XH by network pharmacology analysis, we explored alterations in the gut microbiota. All sample libraries covered >99%, suggesting a sufcient library size to represent most microbes (Supplementary Table 4). Shannon curves with fat ends and rarefaction analysis showed that the OTUs' numbers of both groups adequately presented most species and reached saturation ( Figures 5(b) and 5(c)). PX treatment signifcantly enhanced Ace and Chao indices and also increased Shannon as well as Simpson, albeit without statistical signifcance ( Figure 5(a)). Te rank abundance curve ranking the OTU abundance of each sample by size represents both the species richness and species evenness in samples. Te wider the curve is, the more abundant the composition of species is; the fatter the shape is, the more even the composition of species is ( Figure 5(d)). Weighted unifracdistance with PERMANOVA showed that PX treatment signifcantly altered the β-diversity of commensal microorganisms ( Figure 5(e)). Linear discriminant analysis coupled with efect size measures (LEFSe) showed that PX treatment suppressed the abundance of Clostridium, Escherichia, and Enterococcus, while it increased PrevotellaceaeUCG 001 and Alloprevotella ( Figure 5(f)). Tax4Fun [18] analysis predicted that PX decreased pathogenic bacterial infection and the metabolism of fructose, mannose, cysteine, and methionine ( Figure 5(g)).
Compared with PX, XH failed to infuence the alpha diversity of the microorganisms in CRC murine models (Supplementary Figure 4A) (Supplementary Table 5). PCoA and a heatmap based on Bray Curtis suggested that XH treatment did not change the beta diversity of microbiota ( Supplementary Figures 4B and 4C). Krona demonstrated that the abundance of Clostridia increased from 21% to 26% (Supplementary Figure 4D).

PX Formula Stimulates the Classical Activation of Mφs.
Given the pivotal role of TME in carcinogenesis and the predicted involvement of Mφs by the WGCNA, peritoneal Mφs were collected from CRC murine models. PX treatment suppressed the proportion of P2-3 peritoneal Mφs from 38.2% to 8.51% (Figure 6(a)). Te protein abundance of Arg1 and Tgfβ1, as M2 Mφ markers, was signifcantly decreased in peritoneal Mφs after PX treatment (Figure 6(b)), indicating the inhibition of M2 Mφ transition.

PX Serum Suppresses CW26.WT Cell Migration and
Invasion. To examine the direct antineoplastic role of the PX formula, the serum from CRC murine models after PX treatment was added to CW26.WT cells. An MTT assay showed that PX serum treatment did not afect the viability of CT26.WT cells (Supplementary Figure 5). PX serum signifcantly suppressed the migration capacity of cancer cells after 24-hour incubation, as evidenced by wound healing assays and transwell experiments (Figures 7(a) and 7(b)), and markedly inhibited the expression of VEGF (Figure 7(c)).  Evidence-Based Complementary and Alternative Medicine

PX Formula Reshapes TME and Induces Cancer Cell
Apoptosis. Colorectal tumor-derived organoids (cTDOs) were cultured as a model to examine the proliferation capacity of cancer stem cells, which recapitulates the neoplastic progression within TME. An ex-vivo CRC cellular model was established by coculturing cTDOs with peritoneal Mφs from PX-treated CRC murine models. As illustrated, PXtreated peritoneal Mφs induced cTDO apoptosis, as evidenced by PI staining and elevated Bax protein abundance (Figures 8(a) and 8(c)). Moreover, the expression of Ki67 in cTDOs was signifcantly hindered by coculturing with PXtreated Mφs (Figure 8(b)).

Discussion
In the present study, we provide evidence of the anticancer efcacy of the PX formula against CRC and demonstrate the underlying mechanisms. As the third most common cancer worldwide, CRC develops in the setting of hereditary cancer syndromes or from infammatory bowel disease (IBD) [19], and its complex etiology involves a broad spectrum of causal factors, including environment, genetic susceptibility, and diet [4]. Whilst the mortality rate of CRC patients has been decreased due to efective cancer screening measures and multiple anticancer therapies, those patients who are not   Evidence-Based Complementary and Alternative Medicine suitable for surgery or chemotherapy, or with cancer cachexia, are overlooked, and fnding a potential drug that is low-costing and efective yet of low toxicity is urgent. Accumulated investigations have demonstrated a strong regulation of cancer progression and metastasis by the reciprocal interactions between cell-intrinsic and cellextrinsic mechanisms within the TME [20,21]. In this context, targeting TME portends an ideal choice either as a monotherapy or an adjuvant in combination with other agents as well as chemotherapy. TCM has multiple components fulflling synergistic functions with minimized side efects based on Shang Han Lun, which fts with the concept of targeting TME.
PX and XH are widely used in China to treat cancers [8]. In the present study, we inoculated cancer cells into the mice and administered these prophylaxis remedies simultaneously and compared their antitumor efects on day 30. Our in-vivo experiments demonstrated that the PX formula exhibited a stronger antitumor efect than XH treatment, as evidenced by the suppressed tumor size and inhibited levels of IL-17, IL-6, and TNFα, which are the best characterized independent risk factors for CRC [22][23][24] and increased IFNc expression that inhibits the expansion of disseminated colorectal cancer cells [25]. KEGG analysis showed that PX-target genes were enriched in infammatory reactions, microbe infections, drug resistance, and carcinogenesis, including CRC and bladder cancer, which was consistent with the RNAseq results of murine CRC models after PX treatment. In short, the PX formula presents a promising drug that is able to orchestrate the complex interactions within the TME, for instance, the communications between cancer cells and immune cell populations, microbes, and antigenpresenting cells.
Given the high complexity of CRC pathogenesis, we delineated the pharmacological logic of the PX formula from seven aspects: (1) Reestablishing the immune status within the tumor: in developed cancers, immune surveillance is mostly missing and the immune-suppressive TME is established. TME is critical for carcinogenesis and cancer maintenance by supporting tumor growth, invasion, metastasis, and hiding from T-cell detection and traditional anticancer therapy. However, not all the immune cells within the TME represent an attempt to impair tumor immune surveillance and enable tumor cells to proliferate rapidly. Te combination of AF, CR, AL, and TV tends to induce immunity and eradicate neoplastic cells by shaping the plasticity potentials of Mφs and T17 cells, both of which are protumorigenic in CRC [26][27][28], as evidenced by the enriched IL-17 and TNF signaling pathways. Mφs are immune cells with high plasticity and can transit to diferent phenotypes in response to diverse stimuli. Te classically activated (M1) Mφs are proinfammatory and induce antitumor immunity, while the alternatively activated (M2) Mφs contribute to wound healing and facilitate cancer cells escaping from the host immune reactions [21,29]. Tumor-infltrating Mφs (TAMs) mostly take on an M2-like phenotype and contribute to immune escape, allowing the transformation of neoplastic cells. Herewith, changing M2 Mφs toward M1 Mφs is one of the prospective aims of cancer immunotherapy [30]. Based on TCGA database, we found a close association of PX targets with the survival probability in patients with CRC, as well as with TAM infltrating, which corroborates the therapeutic value of PX in the clinic. Additionally, our in vitro experiments corroborated the induction of proinfammatory polarization of Mφs, which in turn instigated the apoptosis of cancer stem cells and inhibited neoplastic cell migration. Collectively, the PX formula is able to reverse the tumor-promoting phenotype of TAMs and thus reshape the communication between TAMs and other cell populations or microbes, which has also been considered a causal factor for neoplastic processes and infuences the responsiveness to cancer therapeutics [31].
(2) Induction of metastatic cancer cells apoptosis: in our in vitro experiments, we observed that the peritoneal Mφs from the PX mice suppressed the migration capacity of cancer cells shown in the transwell experiments and wound healing assays and inhibited the expression of VEGF in cancer cells, indicating decreased metastatic dissemination [32]. Moreover, the peritoneal Mφs from the PX mice induced apoptosis as evidenced by the increased PI level and BAX protein abundance and reduced proliferationrelated Ki67 expression [33] of cTDOs, which are cancer stem cells responsible for early neoplastic progression and expansion of disseminated cancer cells [34]. Altogether, the PX therapy could inhibit metastatic cancer cell migration and proliferation, and promote apoptosis, which might be accomplished by AF, CR, AL, and TV [35,36]. (3) Altering the pathological microbiota communities: perturbations of the colonic microorganism communities are conducive to the carcinogenesis and metastasis of CRC, which exacerbate the colonization of pathogenic taxa at the expense of probiotics and hence produce cancer-promoting metabolites [37]. PX treatment successfully rescued the decrease in microbiota richness in murine models of CRC and signifcantly suppressed the colonization of the genus Clostridium, Escherichia, and Enterococcus, which are considered human pathogens and associated with colorectal carcinogenesis by inducing colonic infammation and stimulating neoplastic signaling pathways [38][39][40]. Moreover, PX markedly increased the levels of probiotic gut microbiota [41] such as the  genus Alloprevotella and Prevotellaceae UCG 001. Based on KEGG analysis and RNA sequencing, PX exhibits a strong microbicidal efect by using AF, CR, AL, and TV, as well as white alunite, which is a natural product with a potential antivirus [42] and antibacterial capacity [43]. Compared with the 16S rRNA sequencing results of the PX formula, no signifcant alteration in microbial communities has been observed in the XH-treated CRC group, which might explain the poor efcacy of the XH formula.
(4) Reversing drug resistance: cancer-related mortality is mainly attributed to the development of drug resistance. Platinum-based drugs and epidermal growth factor receptor (EGFR) inhibitors have been used widely against human tumors [44,45], and endocrine therapy has been considered the major treatment of estrogen receptor (ER)-positive breast cancer [46]. Based on the system pharmacology analysis, the PX formula infuenced EGFR tyrosine kinase inhibitors, platinum drugs, and endocrine resistance, by the combination of CL, CR, AL, and TV. (5) Reprogramming of amino acid metabolism and nervous and humoral regulation: to sufce the biosynthetic demands of unlimited proliferated cancer cells from nutrient-poor TME, an abundant supply of amino acids is pivotal [47]. As a major component of the PX formula, SS was dedicated to the reprogramming of deregulated amino acid metabolism, including tyrosine, tryptophan, histidine, glycine, serine, and threonine, and was associated with serotonergic synapse activities, neuroactive ligand-receptor interaction, and neurotransmitter metabolic process, which orchestrates gut homeostasis and relieves mental stress [48][49][50]. Moreover, AF and CR modulate the thyroid hormone signaling pathway, which is crucial for aging and age-related diseases, including cancer [51]. In addition, the enriched estrogen signaling pathway also indicates that PX would also exert an efective anticancer capacity against breast cancer.
(6) Malignancy forms a thrombophilic state by increasing venous stasis, which portends a poor prognosis [52]. FTpromotes blood circulation and relieves blood stasis [53] and is also a source of dietary lignans with antiestrogenic, antiangiogenic, proapoptotic, and antioxidant mechanisms [54]. Despite the concerns regarding the risk of infection by zoonotic agents [55], the strong antibacterial combination of AF, CR, AL, and TV, as well as alunite, would decrease the possibility of FT-caused infection.
(7) Last but not least, the multiple components might raise a concern about the absorption and bioavailability of the PX formula. Lines of evidence have demonstrated that mirabilite could alter the pharmacokinetic behaviors of herb components by orchestrating the absorption of each compound [56]. Consistently, RNA sequencing corroborated that the PX formula modulated the cytochrome P450 system, the key enzyme responsible for drug metabolism [57].
Collectively, these herbs in the PX formula work synergistically to instigate the host anticancer immunity and thus stimulate cancer cell apoptosis, suppress infections, and reshape microbiota communities, establishing a tumor-eradicating TME, and restore the homeostasis of humoral. Although the 179 XH targets are also enriched in various cancers, infammation, apoptotic signaling, nervous and humoral regulation, and estrogen signaling pathways, the shortcomings of the XH formula might be attributed to the neglect of drug metabolism and resistance, humoral circulation, energy consumption, and microbiota composition. In short, the PX formula exerts a more successful therapeutic efect than XH and might be administered either as an adjuvant in combination with other therapeutic strategies or as a monotherapy for those patients who could not sufer surgery as well as chemotherapy. Especially, the combined utilization of the PX formula with other treatment modalities in traditional Chinese medicine such as acupuncture [58,59], which shows potent efcacy in alleviating fatigue and cancer pain, might considerably improve the quality of life of patients with CRC. Assuming that our murine CRC models established by orthotopic cecal injection of CT26.WT cells lack genetic diversity and are of modest value for predicting clinical relevance [60][61][62], WGCNA based on TCGA database was conducted and corroborated the therapeutic value.

Conclusion
We posit that the antineoplastic mechanism of PX involves the induction of anticancer immunity, the establishment of healthy microbiota composition, and the improvement of drug metabolism.

CC:
Colorectal cancer IOs: Intestinal organoids Mφs: Macrophages PX: Pingxiao XH: Xihuang TCM: Traditional Chinese medicine TME: Tumor microenvironment BC: Betweenness centrality DC: Degree centrality EC: Eigenvector centrality LAC: Local average connectivity NC: Network centrality GO: Gene ontology KEGG: Kyoto encyclopedia of genes and genomes AUC: Area under the curve VEGF: Vascular Endothelial growth factor.

Data Availability
Te data used to support the fndings of this study are available upon request.

Ethical Approval
All the authors approved the publication.

Conflicts of Interest
Te authors declare that they have no conficts of interest.