Part 2: Integrated Network Pharmacology And Zebrafish Model To Investigate Dual-effects Components Of Cistanche Tubulosa For Treating Both Osteoporosis And Alzheimer's Disease
Mar 30, 2022
Contact: joanna.jia@wecistanche.com / WhatsApp: 008618081934791
CLICK HERE FOR PART 1
3. Results
3.1 Candidate Compounds and Potential Targets
In the research of CT chemical constituents, a total of 75 ingredients were obtained from TCMSP. Then PCA was conducted to visualize the chemical distribution of CT. As shown in Figure 1, the ingredients of CT were multifarious in chemical space, and 28 of them satisfied Lipinski’s rule of five (Lipinski, 2003). Interestingly, there were many overlapping parts between the ingredients of CT and approved small molecule drugs for OP/AD. It illustrated that many compounds in CT had the potential druggability on OP/AD. To further evaluate their druggability, 43 compounds were screened by DL. After targets prediction, 26 candidate compounds were selected for the subsequent analysis. And they can be docked with a total of 847 target proteins.
cistamche has an anti-Alzheimer's disease effect.
The gene entries related to OP or AD were collected from the DisGeNET and GeneCards databases. As a result, 3052 and 8042 gene entries were respectively collected. We adopted the two scores from the DisGeNET and GeneCards databases as evaluation scores, screening the top fifth of targets. The one-fifth ratio was decided by pre-experiments. Then 211 OP-ED common targets were preserved to match with the targets of 26 candidate compounds. Finally, a total of 81 protein targets (Supplementary Table S2) connecting with 22 candidate compounds (Supplementary Table S3) were selected for molecular mechanisms of action analysis, forming a protein-protein interaction (PPI) network shown in Figure 2.
The 22 compounds were divided into 11 categories: 5 phenylethanoid glycosides (PhGs) (Decaffeoylacteoside, Cistanoside E, etc.), 3 phenylacryl oligosaccharides (Cistanoside H, Cistanoside F, etc.), 3 iridoids and iridoid glycosides (Leonuride, Geniposidic acid, etc.), 3 lignans and lignan glycosides (Yangambin, (+)-Pinoresinol-O-β-D-glucopyranoside, etc.), 2 flavonoids (quercetin, genistein), 1 alkaloid, 1 terpene, 1 sterol, 1 fatty alcohol, 1 fatty acid, and other. According to previous reports, these compounds are the main components or active functional ingredients of CT (Fu et al., 2018).
In Figure 2, a total of 81 common targets were found to have correlations with OP and AD. The degree of PPI was adopted as a characteristic parameter to define the significance of potential targets. The top 5 putative target proteins associated with OP and AD were albumin (ALB), insulin (INS), interleukin 6 (IL6), TNF-alpha (TNF), and epidermal growth factor (EGF).
Figure 1. Chemical distribution based on principal component analysis (PCA).
3.2 Integrated and Classified Network Analysis
81 putative target proteins in Table S2 were selected to initiate GO and KEGG pathway enrichment analysis. After filtering by p-value (GO cut-off of ≤ 0.05, KEGG pathway cut-off of ≤ 0.01), 15 GO terms and 66 KEGG pathway terms were returned, as shown in Figures 3 and 4. A total of 16 GO terms are included: 5 for molecular function, 5 for biological processes, and 6 for the cellular components. Since this study was aimed to discover the potential common pathogenesis of OP and AD, we removed the pathway terms directly related to other diseases and classified the rest into different functional categories. It suggested that the 22 active ingredients of CT might regulate a total of 66 pathways which mainly correlated with signal transduction, endocrine system, immune system, cell growth, and death to play a confrontational role against OP and AD.
To reveal the differences in the action of different types of compounds, the PPI network constructed in Figure 2 was analyzed according to their categories in Table S3. Figure 5 illustrated that 5 PhGs, 2 flavonoids, 1 terpene, and 1 sterol, particularly flavonoids, might be more important than other ingredients of CT against OP and AD. After that, these 4 series of active compounds were chosen to uncover their compressed pathways. As can be seen from Figure 6, PhGs, flavonoids, terpene, and sterol worked together or alone on 27 pathways, containing 11 modules: cell growth and death, endocrine system, immune system, signal transduction, development, and others. Results suggested that these compounds might play a synergistic role in the positive effects of CT on OP and AD. It’s worth noting that the protective effects of PhGs could be more remarkable than flavonoids in neurodegenerative disease since PhGs had direct interactions with the Alzheimer's disease pathway whereas flavonoids were related to signaling molecules and interaction. Only sterol and terpene were related to lipid metabolism and the excretory system. In addition, PhGs, terpene, and flavonoids showed a potential relation with endocrine and metabolic disease.
According to the PPI illustration, 2 flavonoids (GE, QU), 1 terpene (AA), and 1 sterol (BSS) were adopted as valuable compounds for further research on molecular mechanisms. Interestingly, we found that they are all interrelated with the MAPK signaling pathway and TGF-beta signaling pathway. And the p-value of the two pathways also was much higher. Putative targets of 4 valuable compounds and the distribution of their affected proteins in the two pathways were demonstrated in Figure 7.
3.3 Valuable Compounds Efficacy Validation
In this network analysis established in 3.2, we captured 4 valuable compounds for pharmacological activity validation. By the topologically structural analysis method in this study, we found that genistein (GE, C9, DL=0.21), quercetin (QU, C11, DL=0.28), abietic acid (AA, C19, DL=0.28), and β-sitosterol (BSS, C20, DL=0.21) showed feasible interplay to against OP and AD. In this part, zebrafish larvae in vivo models induced by Pre or AlCl3 were established respectively to evaluate the protective effects of 4 compounds against OP and AD.
3.3.1. Effect of 4 compounds on OP
To detect whether 4 compounds have protective effects against OP, Pre was used to establish the OP zebrafish model. The development of osteoblasts was first examined by alizarin red staining to investigate the direct effects of 4 compounds on bone formation. Compared with the control zebrafish, Pre apparently induced a change in bone morphology and bone density of zebrafish skulls in the model group (Figure 8). Declines of staining area and IOD in the Pre group and the opposite trend in the Ed group also visually demonstrated the success of the OP model. What’s more, the staining area and 2 IOD of zebrafish skulls in BSS/GE/QU/AA groups were significantly increased compared with the model group (Figure 9). The typical markers of bone formation and resorption such as ALP and TRAP were assayed to further evaluate the anti-osteoporosis effect of the 4 compounds. As shown in Figure 10, Pre significantly decreased ALP activity and increased TRAP activity in contrast with the control group. Conversely, Ed and 4 compounds significantly showed the opposite trend to the Pre group in the ALP and TRAP activities of zebrafish. These results combined together confirmed that BSS, QU, GE, and AA might play positive roles in bone formation in Pre-induced OP zebrafish.
3.3.2. Effect of 4 compounds on AD
To evaluate the protective effects against AD of 4 compounds, the well-recognized AlCl3-induced AD zebrafish model has been used. For this model of AD dyskinesia, the behavior analyzer was used to track the movement of zebrafish. As shown in Figure 11, the AS of the AlCl3 group was significantly lower than those of the control group, while for DPZ and all 4 compounds the ASs was close to the control. DRR and RE were listed in Table S4. The results showed that different concentrations of GE, BSS, QU, and AA increased DRR by 2.37-37.64%, 60.18-103.92%, 70.52-164.61%, 2.37-42.64%, respectively with partial significant (p-value < 0.001-0.01). The ∆Ss in a control group, AlCl3 group, and DPZ group further supported that the construction of the AD model was successful. RE for GE, BSS, QU, and AA were 13.12-50.45%, 11.41-63.09%, 1.63-66.89%, 13.12-51.09%, respectively. These results with significant differences demonstrated that GE, BSS, QU, and AA can improve dyskinesia of zebrafish to some extent. Determination of nerve conduction markers (AChE and ChAT) activities in the groups were shown in Figure 12. AlCl3 significantly increased AChE activity and decreased ChAT activity compared with the control group. But DPZ and 4 compounds showed the ability to partially recover the changes caused by AlCl3 in AChE and ChAT activities. It’s consistent with the results of rehabilitation effect dyskinesia evaluation. In conclusion, GE, BSS, QU, and AA might have the potential to become prominent anti-AD agents.
3.4 Putative Targets Validation
To verify putative targets which were predicted in this work, targets related to MAPK and TGF-beta signaling pathways were chosen for qRT-PCR analysis. Their relative mRNA expression levels in the OP model were shown in Figure 13. Compared with the Pre group, GE increased levels of mapk14a, fgfr1b, tgfb1, bmp2, and decreased levels of tp53, TNF-α, jun, tgfbr1, and c-fos. QU increased levels of mapk14a, fgfr1b, tgfb1, bmp2, and decreased levels of tp53, jun, and il1b. BSS only showed reducing effects on tp53, TNF-α, and mapk3. As shown in Figure 14, in the AD model, GE, QU, BSS, and AA reduced the mRNA expression level of tp53. GE and QU both reduced levels of TNF-α, bmp2, tgfbr1, sp1, tgfb2, and ifnγ, and increased levels of jun, whereas GE and AA increased levels of mapk3. Only GE appeared to reverse the rise of mapk14a, fgfr1b, fgfr1a, and igf1ra. Besides, QU decreased the high levels of il1b and EGF. The relative mRNA expression levels of genes including tp53, TNF-α, mapk14a, mapk3, fgfr1b, bmp2, jun, il1b, and tgfbr1 significantly changed by contrast with the control group in each model. These results suggested that they may serve as a link between the two diseases. In addition, these results also suggested the possibility of synergistic effects of 4 compounds.
4. Discussion
OP and AD are the most common clinical diseases associated with aging. Their complex pathophysiological mechanisms have limited the effective treatment of two diseases for years. The risk factors of OP and AD demonstrated that the susceptible group of both diseases are partly similar, such as elderly women (Abraham et al., 2013; Patel, 2017), diabetics (Vieira et al., 2018; Xia et al., 2012), long-term smokers (Franic and Verdenik, 2018; Zhong et al., 2015), and inactive individuals (Koedijk et al., 2017; Sofi et al., 2011). Furthermore, current studies have identified that some key proteins in AD also had functions in bone metabolism as well (Li et al., 2016; Pan et al., 2018). There are reasons to believe that OP and AD are not completely independent, even have some potential connections. This suggested a potential homo-therapy for heteropathy, which means treating OP and AD with the same therapy.
On the grounds of the therapeutic objectives and principles of TCM, maladjustment in a disease can be classified into several “patterns”. So multiple diseases such as OP and AD might share one “pattern” and be treated by the same TCMs (Jiang, 2005). As a kidney tonic herb, CT is widely used in OP and AD treatment. Modern studies have provided some compact evidence that CT has positive pharmacologic actions on the skeleton and nervous system. However, it still needs more detailed studies to reveal the mechanisms of CT.
cistanche and tongkat ali have positive pharmacologic actions on the skeleton and nervous system.
In the present study, for a clear understanding of the effects of CT on OP and AD, a network pharmacology approach was set up to predict possible active ingredients and analyze related pathways. It’s different from previous network pharmacology studies on CT (Liu et al., 2017). The “one drug - two diseases” molecular network was conducted for the first time, instead of only focusing on one disease. On the one hand, integrated network analysis discovered a total of 22 active compounds of CT with dual effects for treating both OP and AD. On the other hand, the patterns of the two diseases provided support for revealing their links. Classified analysis showed that these active compounds may have synergistic effects on many biological function modules and PhGs, flavonoids, terpene, and sterol showed better efficacies. To verify the feasibility and suitability of this conducted network, zebrafish were used to construct OP and AD models in vivo and evaluated the efficacy of four compounds. And putative targets validation was applied to reveal their potential synergies. These combined studies in silico and in vivo provided a new starting point for revealing the protective effect of CT on two diseases. It may provide new clues to the correlation between OP and AD pathogenesis, as well as help for the future development of therapeutic strategies for two diseases. Nonetheless, the method still needs improvement and perfection to entirely explore synergetic mechanisms of CT for treating OP and AD. In this study, only certain compounds were selected for validation. It’s necessary to further verify the other 18 active constituents with underlying targets and pathways by experiments. And the commonalities between OP and AD still need to be further explored and discovered, especially from the perspective of treatment.
where to buy cistanche
In our study, BSS, GE, AA, and QU were proved to have protective effects on OP and AD in zebrafish. More delightfully, these valuable constituents also have been experimentally validated in other literature and showed pharmacological activities consistent with that in this paper (Ayaz et al., 2017; Chauhan et al., 2018; King et al., 2015; Ramnath et al., 2018; Thummuri et al., 2018; Vargas-Restrepo et al., 2018; Wang et al., 2019; Yuan et al., 2018). It further proved the effectiveness and rationality of this network and indicated the dual effects of the remaining 18 compounds. As for putative targets validation, 9 effective targets consisting of TP53, JUN, TNF, IL1B, MAPK14, MAPK3, FGFR1, BMP2, and TGFBR1 were screened as common targets of OP and AD finally. TP53 is a key tumor suppressor, and its activation could induce neuronal apoptosis (Xiao et al., 2019). TP 53, JUN, and SP1 are active transcriptional factors in primary OP (Xie et al., 2015). TNF and IL1B are common inflammatory cytokines with functions of promoting osteoclastogenesis (Geissler et al., 2018) and bone loss (Sang et al., 2017). Furthermore, they both are therapeutic targets for AD through the inhibition of neuroinflammation to protect neurons (Liu et al., 2017). TGFB1, TGFB2, TGFBR1, and TGFBR1 have complex effects on neuronal inflammation (Lippa et al., 1998) and osteolysis (Quinn et al., 2001). Bone morphogenetic proteins (BMPs) are one of the factors involved in the glial differentiation of neural progenitor cells (Kwak et al., 2014). The upregulated expression of BMP2 has been demonstrated to serve an essential role in osteoblast differentiation (Li et al., 2017). In addition to these experimentally validated targets, the 81 common targets listed in Table S2 were also expected to be therapeutic targets, but have not been researched or have had preliminary studies in other research.
It is particularly noteworthy that these valuable targets all related to MAPK and TGF-beta signaling pathways based on the KEGG enrichment analysis. MAPK signaling pathway regulates cell proliferation, differentiation, survival or apoptosis, inflammation, and innate immunity. It has been reported that the compromised MAPK signaling pathway contributed to the pathology of neurodegeneration such as AD (Kim and Choi, 2015) as well as inhibiting osteoblasts directly (Xiao et al., 2019). As for TGF-β family members, they played different roles in the skeleton with direct effects on bone impairment (Sun et al., 2016), whereas the activation of neuronal TGF-beta signaling increases neurodegenerative disorders and AD-like disease(Tesseur et al., 2006). Thus, MAPK and TGF-beta signaling pathways may be expected to become shared mechanisms for revealing the pathogenesis of OP and AD or slowing the progressions. Besides the two pathways, the 22 active compounds regulated a total of 66 pathways, especially the pathways with the highest p-value, such as the Prolactin signaling pathway, FoxO signaling pathway, and Th17 cell differentiation will also be worthy of attention and research and as the key to uncovering the correlation between OP and AD. Finally, the discoveries summarized may imply the link between the two diseases in the immune system and the endocrine system.
5. Conclusion
In this work, we proposed a network pharmacology approach, combining PCA analysis, DL screening, multiple targets collection and prediction, PPI network construction， as well as GO and KEGG pathway analysis to probe the efficiency of CT for the treatment of OP and AD. Our results suggested that the 22 active ingredients of CT might mainly regulate signal transduction, the endocrine system, the immune system, cell growth, and death to play an important role in the treatment of OP and AD. To make a better understanding of the mechanisms of CT, this network was deeply excavated and analyzed depending on the type of compounds. In the end, we applied zebrafish OP and AD models separately to identify 4 valuable compounds and related targets, providing a feasible method to connect the genome with pharmacodynamics and find dual-effects compounds.
