Active ingredients of BSHX formula
After screening the TCMID database and TCMSP database with the ADME thresholds of OB≥30% and DL≤0.18, a total of 218 active ingredients were identified in BSHX formula, including 87 ingredients (40.0%) in GC, 36 (16.5%) in GQZ, 26 (11.9%) in DZ, 18 (8.2%) in TR, 17 (7.8%) in HH, 14 (6.4%) in SZY, 12 (5.5%) in SY, 5 (2.3%) in FZ, 2 (0.9%) in SDH, 0 (0%) in RG. As shown in Fig.1, a Herb-Ingredient network were further constructed using Cytoscape software. According to the descending order of edge number in this Herb-Ingredient network, we listed and analyzed the top four ingredients that were beta-sitosterol (MOL000358, DL = 0.75, OB = 36.91, found in DZ, GQZ, HH, TR, SZY), stigmasterol (MOL000449, DL = 0.76, OB = 43.83, found in GQZ, SY, SDH, HH, SZY), quercetin (MOL000098, DL = 0.28, OB = 46.43, found in DZ, GC, GQZ, HH) and sitosterol (MOL000359, DL = 0.75, OB = 36.91, found in FZ, GC, SZY, SDH). Thus, these four ingredients are the potential material foundation of BSHX formula against PMOP.
Target prediction and PPI network analysis
Target fishing on the 218 active ingredients was conducted in Uniprot databases, and we obtained 274 related targets of BSHX formula among which there were 220 in GC, 206 in DZ, 201 in HH, 190 in GQZ, 67 in SY, 46 in TR, 51 in SZY, 10 in, FZ 29 in SDH. Then, 292 related targets of PMOP were obtained from GeneCards and DisGeNet databases. After establishment of Venn diagram, we identified 64 overlapping genes between BSHX formula and PMOP (Fig. 2A). These 64 overlapping genes were considered as the potential therapeutic targets.
A Protein-Protein Interaction (PPI) network was built on these 64 overlapping genes using the String database, which contained 64 nodes and 1110 edges (Fig. 2B). Then, we used three main parameters, “degree (DC)”, “betweenness (BC)” and “closeness (CC)”, as the screening thresholds to select the central target genes. After the first screening round of DC≥12, BC≥0.002 and CC≥0.566, 41 nodes and 634 edges were obtained. Through the second screening round of DC≥24, BC≥0.008 and CC≥0.700, only 21 nodes and 202 edges were identified (Fig. 2C). These hub targets played a more important role in the therapeutic effects of BSHX formula, and their information were listed in Table 2.
GO enrichment analysis and KEGG enrichment analysis
Go enrichment analysis was performed on these 64 overlapping genes using DAVID database. Based on the filter of FDR<0.01, a total of 55 GO items were obtained, including 38 BP terms, 4 CC terms and 13 MF terms (Fig. 3A). As BP played a dominant role, we build a bubble diagram for the top 20 of them according to the descending order of log P-value (Fig. 3B). There were 6 BP terms concentrated into the category of angiogenesis, including positive regulation of angiogenesis (GO:0045766), angiogenesis (GO:0001525), cellular response to hypoxia (GO:0071456), positive regulation of endothelial cell proliferation (GO:0001938), response to hypoxia (GO:0001666) and positive regulation of vascular endothelial growth factor production (GO:0010575). Angiogenesis provides essential oxygen, nutrients as well as various bone cells for bone formation (22). Thus, angiogenesis is a key biological process through which BSHX formula exacts anti-PMOP effects.
To further determine the relevant pathways, KEGG enrichment analysis was conducted on the 64 overlapping genes. Based on the threshold of Number≥6, we screened a total of 99 pathways (20 of these listed in Fig. 3C), among which VEGF signaling pathway (hsa04370) directly regulates angiogenesis. A Target-Pathway network showed that 7 overlapping genes, including 4 hub genes (VEGFA, MAPK1, MAPK3 and PTGS2) were enriched in VEGF signaling pathway (Fig. 3D). The signal transduction of VEGF signaling were obtained from KEGG database and presented in Fig. 4.
BSHX formula preserves bone mass in OVX mice
To carry out an animal experimental validation, C57BL/6 J mice were subjected to an OVX surgery and continuously treated with BSHX formula for 8 weeks. The 3D images of μCT showed severe bone loss in the OVX mice compared to the sham ones, and BSHX formula effectively alleviated bone loss (Fig. 5A). We also found that bone microstructure parameters were significantly improved after treatment of BSHX formula, including the increase of BMD, BV/TV, Tb.Th and Tb.N and the decrease of Tb.Sp (Fig. 5B–5F). These results indicated that BSHX formula can preserve bone mass in the OVX mice.
BSHX formula promotes blood vessel formation in OVX mice
To verify the changes of angiogenesis, ABH staining was performed on the paraffin sections of each group, by which blood vessel can be dyed with red color. The representative images and quantitative analysis showed that thin trabeculae and massive lipid droplets in the OVX mice could be attenuated after treated with BSHX formula for 8 weeks (Fig. 6A-6C), confirming the anti-PMOP effects of BSHX formula. More importantly, we found the OVX mice presented sparse blood vessels in the bone marrow. While the number of blood vessel was significantly increased in the BSHX formula treated mice (Fig. 6A, 6D), indicating that BSHX formula can promote blood vessel formation in the OVX mice.
OVX-induced down-regulation of VEGF signaling are improved by BSHX formula
To determine the involvement of VEGF signaling pathway, we evaluated the expressions VEGF, COX2 and eNOS in each group by IHC staining. CD31, downstream target of VEGF signaling, is specifically expressed in vascular endothelial cells. The representative images and quantitative analysis showed that the levels of VEGF, COX2, eNOS and CD31 were significantly decreased in the OVX mice compared to the sham ones (Fig. 7A-7D). While the OVX-induced down-regulations of VEGF, COX2, eNOS and CD31 were both restored after treatment of BSHX formula for 8 weeks (Fig. 7A-7D), indicating that BSHX formula promotes angiogenesis possibly through activation of VEGF signaling.