Network Pharmacological Analysis and Animal Experimental Study on Osteoporosis Treatment with GuBen-ZengGu Granules

Aim We explored the molecular pathway and material basis of GuBen-ZengGu granules (GBZGG) in treating osteoporosis using network pharmacology and animal experiments. Methods The effective active components and potential targets of GBZGG were obtained from the TCMSP database and BATMAN-TCM database. Disease-related genes were obtained from GeneCard, NCBI, and DisGeNET. Next, a protein interaction network was established using the STRING database, and core genes were screened using the MCODE module. Cytoscape 3.8.0 was used to construct the network of component-disease-pathway-target, and KEGG pathway enrichment analyses were performed using the clusterProfiler R package to predict the mechanism of GBZGG in treating osteoporosis. An osteoporosis rat model was established by ovarian excision (OVX), and the partial results of network pharmacology were experimentally verified. Results Pharmacodynamic results showed that GBZGG increased bone mineral density (BMD) and significantly improved the indexes of femur microstructure in model rats. The network pharmacology results showed that quercetin, luteolin, stigmasterol, angelicin, kaempferol, bakuchiol, bakuchiol, 7-O-methylisomucronulatum, isorhamnetin, formononetin, and beta-sitosterol are the major components of GBZGG, with MAPK1, AKT1, JUN, HSP90AA1, RELA, MAPK14, ESR1, RXRA, FOS, MAPK8, NCOA1, MYC, and IL-6 as its core targets for treating osteoporosis. Biological effects could be exerted by regulating the signaling pathways of fluid shear stress and the signaling pathways of atherosclerosis, advanced glycation end products (AGE-RAGE) of diabetic complications, prostate cancer, interleukin (IL-17), tumor necrosis factor (TNF), hepatitis B, mitogen-activated protein kinase (MAPK), etc. The results of animal experiments showed that GBZGG could reduce the serum levels of IL-6 and TNF-α, increase the expression of bone morphogenetic protein-2 (BMP-2) and runt-related transcription factor 2 (RUNX2) protein, and inhibit the activity of extracellular-regulated protein kinases (ERK1/2) and phosphorylation ERK1/2 (p-ERK1/2) protein. Conclusion GBZGG reduces the expression of ERK1/2 and p-ERK1/2 proteins and mRNAs through the inhibitory effects on IL-6 and TNF-α and negatively regulates the MAPK/ERK signaling pathway. The osteoporosis model showed that it effectively improved the loss of bone mass and destruction of bone microstructure in rats and maintained a positive balance for bone metabolism.


Introduction
Osteoporosis (OP) is a systemic metabolic bone disease characterized by disruption of bone microarchitecture and impaired bone strength, resulting in increased susceptibility to fractures [1]. With the increasing incidences of OP annually and changes in the global population structure, fragility fractures in the spine, extremities, and other components at a later stage, it has started seriously afecting the quality of life of patients and signifcantly increased the burden on health care. Te current treatment for OP aims at reducing discomfort, delaying progression, reducing the risk of fracture, and improving the quality of life of patients.
Traditional Chinese medicine (TCM) follows the principles of "syndrome diferentiation and treatment variation" to regulate the body as a whole and exerts the regulatory efects of multiple molecules, targets, and pathways in preventing and treating OP. Tis approach has been accepted and recognized by an increasing number of patients because of its advantages of simplicity, convenience, examination, and cheapness. TCM believes that "kidneys store the essence to generate the marrow," and the health of bones depends on the sufciency of kidney essence, whereas kidney defciency and essence reduction largely contribute to OP [2]. In addition, it is intricately related to the decline of the digestive system, especially the functions of the spleen and stomach. Terefore, tonifying the kidneys and flling the essence are the roots of TCM in preventing and treating this disease, and tonifying the spleen, supplementing Qi, harmonizing the blood, and dredging collaterals are the core concepts of OP treatment [3]. GuBen-ZengGu granules (GBZGG), which are composed of nine kinds of Chinese herbal medicines, namely, Radix Astragali (Huangqi, HQ), Epimedii Folium (Yinyanghuo, YYH), Codonopsis pilosula (Dangshen, DS), Angelicae Sinensis Radix (Danggui, DG), Cistanches Herba (Roucongrong, RCR), Rehmannia glutinosa (Shu Di Huang, SDH), Psoraleae Fructus (Buguzhi, BGZ), Cibotii Rhizoma (Gouji, GJ), Linderae Radix (Wu Yao, WY), nine drugs in the compound prescription have been evaluated with http://www.worldforaonline.org. GBZGG is a clinical compound preparation developed by Professor Min Song, a famous doctor of TCM in Gansu Province (GYZ20220417000). It has been used for more than 20 years. Clinical trials have demonstrated that it can increase bone mineral density (BMD), improve clinical symptoms, be safe and efective, have a few adverse reactions, and have an excellent therapeutic efect on OP [4]. Previous studies have reported that GBZGG-treated serum exerts bone marrow promoting, stromal cell proliferation, osteogenic diferentiation, and inhibiting adipogenic differentiation efects, and its mechanism is intricately related to the regulation of key proteins in the Notch signaling pathway, Wnt/β-catenin signaling pathway, and bone morphogenetic protein (BMP) signaling pathway [5,6]. GBZGG can bidirectionally regulate bone metabolism in a rat model of postmenopausal osteoporosis [7], exert estrogen-like efects to improve BMD in rats, and regulate intestinal bacterial disorders by activating the intestinal bacterial RNA translation and transcription and the generation and repair of the cell wall [8]. Although certain breakthroughs have been made at the cellular and molecular levels and proteomics, the mechanism of action of TCM compounds in treating OP has not been completely elucidated due to their diverse components and complex afnity targets. Te interaction between molecules and diferent proteins results in a complex pharmacology of drugs. In this context, network pharmacology has emerged as a new disciplinary concept proposed in the context of big data [9] by constructing a drug-gene-target-disease network, combining the biological information with pharmacological analysis, and enriching the understanding of drug structural information and pharmacological efects from an interrelated perspective [10] based on the overall concept of TCM. Terefore, we used network pharmacology to analyze the active components, target genes, biological functions, and signaling pathways of GBZGG in treating OP and verifed them through animal experiments. We believe this study will be highly signifcant in revealing the material basis and molecular mechanism underlying OP treatment and providing a scientifc basis for the next step of research to broaden the ideas of TCM in preventing and treating OP.

Experimental Animals.
Specifc pathogen-free (SPF) grade SD female rats, aged 8 weeks and weighing about 180-200 g, were provided by the Animal Experimental Center of Gansu University of TCM. Te experimental site was the SPF-grade animal laboratory of the Animal Experimental Center of the Gansu University of TCM. Te room temperature was regulated at 22 to 26°C, and the air humidity varied from 45 to 65%. Six animals were housed in each cage; the light duration was 7:00 AM to 7:00 PM; sterile animal feeds were regularly and quantitatively fed; free access was provided to sterilized distilled water; and the bedding was changed every 5 days. Te experimental procedures were conducted according to the ethical review requirements for animal experiments of the Gansu University of TCM and followed the 3R principle, document approval no. 2020-310.

Main Reagents.
Tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) kits were purchased from Jiangsu Masha Industrial Co., Ltd. (F3056-A and F3066-A, respectively). Te main results are runt-related transcription factor 2 (Runx2), BMP-2, extracellular-regulated protein kinases (ERK1/2), and phosphorylation-induced ERK1/2 (p-ERK1/ 2) protein antibodies (Bioss, bs-1134R, bs-1012R, bs-0022R, bs-3292R), ERK1 + p-ERK1 protein antibody, and ERK2 + p-ERK2 protein antibody (Abcam, ab19282, ab32081). corresponding to the GBHD group, GBMD group, and GBLD group, respectively. In the EV group, rats were intragastrically administered estradiol valerate tablets administered at a concentration of 0.09 mg·kg −1 . Te sham group and OVX group received an equal volume of normal saline. Diferent concentrations of the drug were prepared as 2 mL of the solution according to the rat gavage volume of 10 mL·kg −1 standard for gavage once daily for 12 weeks. To ensure stable intragastric drug concentration and reduce concentration error, intragastric rats were weighed and measured regularly every week, and the intragastric concentration was adjusted according to the changes in their body weights.

Bone Mineral Density Detection.
After the completion of drug intervention in each group of rats, anesthesia was induced by intraperitoneally injecting 1% pentobarbital sodium 3 mL·kg −1 . Te rats were placed in the prone position on the operating

Protein Interaction Network and Cluster Analysis.
Te common targets of GBZGG and OP were entered into the STRING database (https://www.string-db.org/), and protein-protein interaction (PPI) network was constructed.
Te biological species was set as "Homo sapiens." Te disconnected nodes were eliminated using a confdence >0.9 as a screening condition to obtain the PPI network. Te PPI network was imported into Cytoscape 3.8.0, and the MCODE module was used to analyze gene clusters as well as screen core targets.

Enrichment Analysis of Common Target Genes.
Te KEGG enrichment analyses of target genes of GBZGG in treating OP were performed using the clusterProfler R software package (p value cut-of <0.05, q value cut-of <0.05) to predict the related signaling pathways involved in the regulation of core genes in the treatment of OP by GBZGG. Finally, the "component-diseasesignalingpathway-target" network diagram was constructed using Cytoscape 3.8.0.

Animal Grouping, Modeling, and Administration.
It is the same as described in Section "3.1.1."

Serum IL-6 and TNF-α Content Evaluation by ELISA.
After the rats were anesthetized, blood sampling from the heart was performed using the "three-line localization" method. Te collected blood was allowed to stand at room temperature for 2 h. It was centrifuged at 3000 rpm at room temperature for 15 min. Te separated upper serum was aspirated using a micropipette. Te levels of IL-6 and TNF-α in the serum were determined using the ELISA kit, and the microplate was read at 450 nm.

Detection of BMP-2, Runx2, and ERK1/2 mRNAs by RT-PCR.
Te femoral samples were frozen at -80°C and cut into pieces. Next, 100 mg of samples from each group were placed into a mortar for grinding. Tey were pulverized into powder form and transferred into a centrifuge tube. Afterward, 1 mL of RNAiso Plus was added to dissociate the ribosomes. Next, 0.2 mL of chloroform was added and allowed to stand at room temperature for 3 min. It was centrifuged at 12,000 rpm for 10 min at 4°C, and the upper aqueous phase was drawn into a new centrifuge tube, to which 0.5 mL of isopropanol was added and which was made to stand at room temperature for 10 min. It was again centrifuged at 12,000 rpm for 10 min at 4°C and 1 mL of 75% ethanol was added, washed thoroughly by vortexing, and centrifuged at 7500 rpm for 5 min at 4°C. Te RNA was dissolved by adding 50 μL of enzyme-free water, and the RNA loading concentration was calculated. It was reverse transcribed into cDNA according to the instructions provided in the reverse transcription kit. Te results were calculated by the 2 -△△CT method, and the data were analyzed by statistical software to analyze the mRNA content of each group of samples. Te primer sequences used are shown in Table 1. 3.3.4. BMP-2, ERK1/2, p-ERK1/2, and Runx2 Protein Expression by Western Blotting. Frozen femoral samples were cut into pieces and placed in 2 mL centrifuge tubes and ground with a high-speed and low-temperature tissue grinder. A total of 100 mg of sample was transferred to a 2 mL centrifuge tube, and 1 mL of RIPA lysate containing PMSF was added. It was completely lysed on ice and centrifuged at 12,000 rpm for 15 min at 4°C; the supernatant was aspirated, and a BCA assay was used for protein quantifcation. Te protein was mixed with the sample bufer in a ratio of 4 : 1 and denatured with boiling water for 5 min. Electrophoretic separation was performed at a protein loading volume of 30 μL, transferred to polyvinylidene fuoride (PVDF) membranes, and blocked with 5% non-fat dry milk for 90 min. Tey were placed in protein primary antibody (BMP-2, Runx2, EKR1/2, p-ERK1/2) dilutions (1 : 1000) and incubated overnight at 4°C on a shaker, followed by incubation with II antibody for 1 h. Te luminescent solution was mixed in equal proportion, shaken well, and added to the PVDF membrane. Te membrane was developed and observed after exposure, and the bands were quantitatively analyzed using the ImageJ software.

Immunohistochemistry for ERK1/2 and p-ERK1/2
Protein Expression. Te prepared parafn sections were deparafnized and hydrated with xylene and ethanol. Te citric acid antigen was heat-repaired, washed with phosphate-bufered saline (PBS), and subsequently incubated with 3% H 2 O 2 for 15 min at room temperature. Te sections were blocked with sera homologous to the secondary antibody and incubated at 37°C for 30 min. Primary antibodies (ERK1/2 and p-ERK1/2 antibodies) were added dropwise and incubated overnight in a refrigerator at 4°C. Sections were removed for recovery, washed, and dropped with secondary antibody, incubated at 37°C for 30 min, developed by adding 3, 3-diaminobenzidine (DAB), and stained for imaging. Te Image Pro Plus software was used for data analysis.
3.3.6. Statistical Analysis. Te experimental data were statistically analyzed using SPSS 23.0 and GraphPad Prism 7 software. Measurement data are expressed as mean-± standard deviation (mean ± SD). An independent sample t-test was used for comparison between the two groups. One-way analysis of variance (ANOVA) was used for comparison between multiple groups. Statistical diferences were accepted when the p value was <0.05.

BMD Test Results.
Compared with the sham group, the BMD of the OVX group was signifcantly lower (P < 0.01). Compared with the OVX group, the BMD of the EV, GBHD, GBMD, and GBLD groups was signifcantly increased (P < 0.01). Te BMD of the GBHD group was signifcantly higher than that of the GBLD group (P < 0.01), as shown in Figure 1.

Histopathological Observation of Bones.
Te structure of the femoral bone was normal in the sham group, and the thickness and continuity of the bone under the epiphyseal plate were good. Te structure was normal, and the number was higher, as shown in Figure 2(a). In contrast, the bone structure became thinner and sparsely arranged in the OVX group; the continuity was interrupted, the structure was incomplete, numerous blind ends appeared, the thickness of the small beam wall was unequal, and osteoblast and osteoclast proliferation was observed around the bone, as shown in Figure 2(b). Compared with the OVX group, the EV group had increased bone thickness, an elevated number of osteocytes, a gradually intact structure, and enhanced connection between the two ends of the bone, as shown in Figure 2(c). In the GBZGG groups, the bone structure gradually became intact with an increase in drug dose; the thickness increased, the number of osteocytes increased, the connection between the bone and the surrounding tissues increased, and the overall structure gradually became intact, as shown in Figures 2(d)-2(f ).

Micro-CT Bone Morphological Findings.
Compared with the sham group, the OVX group showed signifcantly reduced the cancellous bone in the femoral cavity, sparse separation of the bone structure, and fracture of the connection, as shown in Figure 3. After three-dimensional reconstruction, the overall structure of the sham group degenerated, and the volume decreased. Following drug intervention, the number of bones increased signifcantly, the loose structure was improved, and the overall structure of cancellous bone was more complete. In addition, the efect was more evident in GBZGG groups with an increase in the dose, and the efect of GBHD was equivalent to that of EV, as shown in Figure 4.

Changes in Femoral Bone Morphometric Parameters.
Compared with the sham group, trabecular number (Tb.N) in the OVX group was signifcantly lower (P < 0.01). Compared with the OVX group, Tb.N number was significantly increased in the EV, GBHD, GBMD, and GBLD groups (P < 0.01). Compared with the sham group, the OVX group had a higher prevalence of trabecular thickness (Tb.T), which was signifcantly decreased (P < 0.01). Compared with the OVX group, the Tb.T of the EV, GBHD, GBMD, and GBLD groups was signifcantly improved (P < 0.05). Compared with the sham group, the trabecular separation (Tb.Sp) in the OVX group was signifcantly increased (P < 0.01). Compared with the OVX group, the Tb.Sp of the EV, GBHD, GBMD, and GBLD groups was signifcantly reduced (P < 0.01), as shown in Table 2.
Compared with the sham group, the OVX group connectivity density (Conn.D) was signifcantly reduced (P < 0.01). Compared with the OVX group, the Conn.D of EV, GBHD, GBMD, and GBLD groups was signifcantly increased (P < 0.01). Compared with the sham group, the Structure Model Index (SMI) was signifcantly higher in the OVX group (P < 0.01), whereas it was signifcantly decreased in the EV, GBHD, and GBMD groups (P < 0.01). Compared with the sham group, bone volume to tissue volume (BV/ TV) was signifcantly decreased in the OVX group (P < 0.01), and compared with the OVX group, the bone volume fraction was signifcantly improved in the EV, GBHD, GBMD, and GBLD groups (P < 0.01), as shown in Table 3.

Prediction of Common Target Genes of GBZGG-OP.
After retrieval, Radix Astragali was predicted to contain 20 compound components and 194 targets, Angelicae Sinensis Radix had 2 compound components and 51 targets, Codonopsis pilosula had 21 compound components and 106 targets, Rehmannia glutinosa had 2 compound components and 30 targets, Epimedii Folium had 23 compound components and 208 targets, Cistanches Herba had 7 compound  Figure 1: Changes of BMD in rats. ## P < 0.01 vs. the sham group; * * P < 0.01 vs. the OVX group; ΔΔ P < 0.01vs. the GBHD group. Mean ± SD, n � 3.  Te common targets were subsequently used as predicted targets of drugs acting on diseases for the following pathway enrichment analysis, as shown in Figure 5.

Serum IL-6 and TNF-α Expression.
Compared with the sham group, the rats in the OVX group had signifcantly elevated serum levels of IL-6 and TNF-α (P < 0.01). Compared with the OVX group, IL-6 and TNF-α levels in the EV, GBHD, GBMD, and GBLD groups decreased signifcantly (P < 0.05), as shown in Figure 11.

BMP-2, RUNX2, and ERK1/2 mRNA Expression.
Compared with the sham group, the OVX group showed signifcantly reduced expression of BMP-2 mRNA (P < 0.01). Compared with the OVX group, the expression of BMP-2 mRNA was signifcantly increased in the EV, GBHD, GBMD, and GBLD groups (P < 0.01), as shown in Figure 12(a). Compared with the sham group, Runx2 mRNA expression was signifcantly decreased in the OVX group (P < 0.01). Compared with the OVX group, the expression of Runx2 mRNA was signifcantly increased in the EV, GBHD, GBMD, and GBLD groups (P < 0.05), as shown in Figure 12(b). Compared Figure 6: Te GBZGG-ingredient-target-OP interaction network. Note: in the network, purple represents the active ingredient, yellow represents the target of the drug acting on the disease, pink represents the drug composed of GuBen-ZengGu granules, and the green rectangle represents osteoporosis.   Figure 8: Bar graph of the number of adjacency nodes in the protein interaction network. Note: topological analysis was performed using the NetworkAnalyzer tool. Degree ranking showed that genes with scores greater than the average score were selected as key targets. A total of 87 key targets were selected. Te top 20 targets were imaged using R 4.0.3, and the abscissa was the degree value of each target. with the sham group, the expression of ERK1/2 mRNA was signifcantly increased in the OVX group (P < 0.01). Compared with the OVX group, the expression of ERK1/ 2 mRNA was signifcantly decreased in the EV, GBHD, GBMD, and GBLD groups (P < 0.01), as shown in Figure 12(c).   Compared with the sham group, BMP-2 expression was signifcantly lower in the OVX group (P < 0.01). Compared with the OVX group, BMP-2 was signifcantly higher in the EV, GBHD, and GBMD groups (P < 0.01), as shown in Figure 13(b). Compared with the sham group, the expression of Runx2 was signifcantly lower in the OVX group (P < 0.01). Compared with the OVX group, the expression of Runx2 was signifcantly higher in the EV, GBHD, and GBMD groups (P < 0.05), as shown in Figure 13(c). Compared with the sham group, the expression of ERK1/2 was signifcantly increased in the OVX group (P < 0.01). Compared with the OVX group, the content of ERK1/2 protein was signifcantly decreased in the EV, GBHD, GBMD, and GBLD groups (P < 0.01), as shown in Figure 13(d). Compared with the sham group, the expression of p-ERK1/2 expression was signifcantly increased in the OVX group (P < 0.01). Compared with the OVX group, the expression of p-ERK1/2 was signifcantly decreased in the EV, GBHD, and GBMD groups (P < 0.01), as shown in Figure 13(e). Compared with the sham group, the expression of p-ERK1/2 was signifcantly increased in the OVX group (P < 0.01). Compared with the OVX group, the expression of p-ERK1/2 was signifcantly decreased in the EV, GBHD, and GBHD groups (P < 0.05), as shown in Figure 13(f ).

Immunohistochemical Results.
Compared with the sham group, the expression of ERK1/2 protein was significantly increased in the OVX group (P < 0.01). Compared with the OVX group, the content of ERK1/2 protein was signifcantly reduced in the EV, GBHD, and GBMD groups (P < 0.05), as shown in Figures 14(a)-14(g). Compared with the sham group, the levels of p-ERK1/2 protein were signifcantly increased in the OVX group (P < 0.01). Compared with the OVX group, p-ERK1/2 protein staining was signifcantly reduced in the EV, GBHD, and GBMD groups (P< 0.05), as shown in Figures 15(a)-15(g).

Discussion
Low bone mass and microstructural degeneration of bone tissue constitute the major lesion characteristics of OP, resulting in increased bone fragility and fracture risk [11]. Te mechanism is intricately related to factors such as body hormone levels, genetic inheritance, lifestyle habits, underlying diseases, drug use, and exercise immobilization. Tis disease belongs to the category of "bone impotence" and "bone withering" in TCM. OP is manifested by kidney essence loss and loss of nutrition from bones. Te primary symptoms include kidney defciency and blood stasis, spleen, and stomach weakness. Te treatment of OP is based on tonifying the kidneys, invigorating the spleen, and flling the essence [12]. GBZGG is composed of nine herbs: Radix Astragali, Angelicae Sinensis Radix, Codonopsis pilosula, Epimedii Folium, Rehmannia glutinosa, Cistanches Herba, Psoraleae Fructus, Cibotii Rhizoma, and Linderae Radix. We used network pharmacology to analyze the active components, target genes, and signaling pathways of GBZGG in treating OP. Te results demonstrated that quercetin, luteolin, stigmasterol, angelicin, kaempferol, bakuchiol, 7-O-methylisomucronulatol, isorhamnetin, formononetin, and beta-sitosterol constitute the major components of GBZGG in treating OP. Quercetin, luteolin, beta-sitosterol, and kaempferol have been shown to exert important efects on bone metabolism, whereas the mechanism of action of other ingredients warrants further studies. Quercetin can maintain the balance of bone homeostasis through diferent signaling pathways, and Wong et al. [13] summarized the achievements of the application of quercetin in the feld of bone metabolism research and found that quercetin can inhibit RANKL-mediated osteoclastogenesis, osteoblast apoptosis, oxidative stress, and the infammatory response while promoting osteogenesis, angiogenesis, adipocyte apoptosis, and osteoclast apoptosis. Luteolin belongs to plant favonoids, and a study demonstrated that luteolin inhibits the diferentiation function of osteoclasts, increases bone mineral density, and reduces bone loss in OVX-induced mice, showing certain potential in preventing postmenopausal osteoporosis [14]. Beta-sitosterol is the most abundant phytosterol, and studies by Wang et al. [15] have stated that betasitosterol efectively prevents glucocorticoid-induced OP by protecting osteoblasts and inhibiting osteoclastogenesis. Kaempferol is a favonoid with osteogenic activity, and studies have demonstrated that kaempferol reduces glucocorticoidinduced bone loss, enhances bone regeneration capacity at the  fracture site, and exerts a positive efect on maintaining bone health [16]. Te PPI network and cluster analysis revealed IL6, MAPK1, MAPK8, and MAPK14 as the core targets of GBZGG. Te KEGG analysis of the consensus genes of TCM compound diseases demonstrated them to be involved in signaling pathways, including the MAPK signaling pathway. Te MAPK signaling pathway family is intricately related to the balance of bone homeostasis and exerts dual regulatory efects on the process of bone formation and bone resorption via the MAPK/ERK signaling pathway, p38 signaling pathway, and JNK signaling pathway, as shown in Figure 16. Moreover, it is tightly related to the proliferation, diferentiation, and apoptosis of cells [17]. Te MAPK/ERK signaling pathway is one of the major signaling pathways of MAPK-an important pathway for major growth factors and cytokines to regulate cell proliferation and diferentiation. It plays a crucial role in osteogenic diferentiation as well as bone formation [18]. Te MAPK signaling pathway can be activated by infammatory factors, further increasing the expression of c-Fos and activating activated T cell nuclear factor 1 (NFATc1) to promote the diferentiation and maturation of osteoclasts and accelerate bone resorption [19]. During OP, the activity of osteoclasts increases, accelerating the apoptosis of osteoblasts. Apoptotic cells stimulate the surrounding osteocytes and macrophages and secrete several factors such as RANKL, IL-6, and TNF-α. Infammatory factors will promote the expression of ERK1/2 and its phosphorylation levels to accelerate the activation of osteoclasts and increase the RANKL/OPG ratio by activating ERK1/2 to promote the proliferation of osteoclasts [20]. Combined with the analysis results of network pharmacology, we speculated that GBZGG could afect bone metabolism via the MAPK/ERK signaling pathway via exerting the efect on the expression of infammatory factors, as verifed by animal experiments. After establishing the OVX OP rat model, the serum levels of IL-6 and TNF-α were measured by ELISA. Te levels of IL-6 and TNF-α were signifcantly increased in the OVX group, indicating altered levels of estrogen and activation of infammatory factors following OVX modeling in rats. Considering that BMD is an essential indicator for evaluating bone health and the degree of osteoporosis, we detected the levels of BMD. Te BMD of rats was signifcantly reduced after modeling to the extent of causing bone loss. Analysis of HE sections revealed that the structure of  Evidence-Based Complementary and Alternative Medicine femoral tissue in the OVX group became thinner and sparsely arranged, the continuity was interrupted, the structure was incomplete, numerous bone blind ends appeared, the thickness of the trabecular wall was unequal, and osteoblast and osteoclast proliferation was observed around the bone, indicating that the bone immune microenvironment was overexpressed and the bone resorption activity was increased after castration in rats. Similarly, Weitzmann [21] reported that immune activation and bone loss initiated simultaneously during menopause, and decreased estrogen levels activated T cell hyperactivation, producing excessive IL-6 and TNF-α. Te activation of bone resorption-related signaling pathways increased bone loss. Micro-CT, a scanning imaging technology, can be used to observe the detailed changes in bone structure as it more intuitively and accurately observes the microscopic changes in the bone following bone loss. It is an ideal method to observe the micromorphological changes in the bone [22].
To further observe the changes in bone microarchitecture, we applied micro-CT to scan the rat femur. Te 2D scanning sections of diferent positions of femoral samples showed major bone loss, resulting in the disappearance of the bone cavity in the OVX group. Te overall structure and quality changes in the cancellous bone in the femur were further evaluated after 3D pattern mapping and comparison of the content of bone in each group. Considering the overall structure of the skeleton, the cancellous bone will decay from a morphologically dense plate-like structure to a morphologically sparse rod-like structure during the development of the OP, thereby increasing the SMI. Simultaneously, since the rod-shaped structure of cancellous bone has a large gap, the Conn.D between bones decreases, and subsequently, the BV/TV used to describe the specifc gravity of cancellous bone volume in this space also decreases, refecting the degree of defect of the overall structure of cancellous bone at this site. After drug intervention, the levels of IL-6 and TNFα in each group were reduced, and the BMD index was signifcantly improved, with the most signifcant efect in GBHD and EV groups. In addition, micro-CT results demonstrated that Tb.N and Tb.T were signifcantly increased, SMI and Tb.Sp values were signifcantly decreased, and Conn.D and BV/TV values were signifcantly elevated in the femoral tissue of rats in the EV and GBHD groups. Tis suggests that GBZGG can act directly on the bones to enhance the quality of cancellous bone, repair defective bone, prevent the degeneration of bones, improve bone strength, and enhance the stability of the overall structure of cancellous bone in the femur.
To further clarify the positive regulation of GBZGG on bone metabolism, we conducted mechanistic studies. BMP-2, a member of the BMP family of growth factors, is primarily distributed in bone tissue and activates the production of bone formation-specifc transcription factors by receiving extracellular stimulatory signals from osteoblasts, thus promoting bone formation [23], whereas Runx2 is a specifc transcription factor for bone formation. BMP-2 can promote osteogenesis by activating ERK1/2 and increasing the expression of Runx2 by elevating its phosphorylation level [24]. RT-PCR and western blotting  revealed that BMP-2 and Runx2 proteins and mRNA expression were signifcantly decreased in the OVX group, indicating the reduced osteogenic potential of rats following OVX treatment. After drug treatment, the expression of BMP-2, Runx2, and mRNA in the GBHD and EV groups was signifcantly increased, confrming that GBZGG promoted bone formation in rats. We next found that the ERK1/2 mRNA was highly expressed in the OVX group, whereas the GBHD and EV groups showed signifcantly decreased expression, indicating that GBZGG inhibited the expression of the ERK1/2 gene. Combined with the detection results of serum IL-6 and TNF-α levels, the MAPK/ERK signaling pathway could be activated by inducing the expression of infammatory factors, further improving the activity of osteoclasts to promote bone resorption. To assess this, we detected the expression of ERK1/2 and p-ERK1/2 proteins. Te results showed that the expression of ERK1/2 and p-ERK1/2 proteins was signifcantly increased, whereas that of BMP-2 and Runx2 proteins was signifcantly decreased in the OVX group, indicating that ERK1/2 and p-ERK1/2 exacerbated bone resorption. Te expression of ERK1/2 and p-ERK1/2 proteins was signifcantly decreased, whereas that of BMP-2 and Runx2 proteins was signifcantly increased after drug treatment, indicating that GBZGG inhibited the activation of ERK1/2 and p-ERK1/2 and reduced the occurrence of bone resorption. To further verify the expression of the ERK signaling pathway in OP, we detected the expression of ERK1/2 and p-ERK1/2 proteins in rat bone tissues using immunohistochemical staining. Te results were consistent with those of western blotting. Tis proves that GBZGG promotes bone formation, not through the BMP-2-ERK1/ 2-Runx2 pathway for conduction but via the negative regulation of the MAPK/ERK signaling pathway. Tese results also explain from another perspective that the functional expression of the MAPK/ERK signaling pathway is afected by the overall environment of the body and is not the embodiment of a single function. In all, combined with the above results, GBZGG improved the BMD levels, delayed the degree of bone loss and structural degeneration of the femoral bone in OVX model rats, and promoted bone formation by upregulating the expression of BMP-2, Runx2 protein, and mRNA. Te negative regulation of the MAPK/ERK signaling pathway could be achieved by downregulating the expression of IL-6 and TNF-α, providing an experimental basis for the mechanism study of GBZGG in preventing and treating OP.

Data Availability
Te datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.