Introduction
The World Health Organization estimates that about 80% population in
the developing countries apply herbal medicines for their primary health
care (Li et al., 2017). Additionally, herbal medicines also play an
important role in the treatment of some major diseases, such as COVID-19
(Yang et al., 2020), cardiovascular diseases (Hao et al., 2017), and
cancers (Wang et al., 2020), etc. Recently, plenty of experimental
studies have been carried out on the mechanism of action of herbal
medicines, and the herb targets have been identified. With the massive
accumulation of the previous researches (Chen et al., 2019; Liu et al.,
2018b; Liu et al., 2018c), the analysis of the active ingredients that
can regulate these targets will be the research focus in the future. The
active ingredients analysis of herbal medicines is crucial for their
quality control and efficacy/safety evaluation. However, the present
methods and strategies are still challenging, owing to the complex
disease pathophysiology and herbal composition. Here, we propose a new
strategy for effectively and rapidly analyzing the active ingredients of
herbal medicines based on the public bioinformatics platforms. This
strategy may extend and develop the achievements of the previous herb
targets researches, and also provide the references for the later
development and clinical application of herbal medicines, which serves
as a link between the preceding and the following.
PubChem BioAssay and STRING are the public bioinformatics platforms,
both of which are open-accessed and user-friendly. PubChem BioAssay
database contains the bioactive targets information of small molecules,
which is generated through the experiments and literatures (Wang et al.,
2017). STRING database aims to achieve a comprehensive and
objective protein-protein interaction networks and allows users to
visualize the interaction networks (Szklarczyk et al., 2019). Both
public resources can be applied in this strategy.
At present, there are mainly two ways to obtain the information of
herbal compounds. One method is directly based on the
compound analysis techniques (Li
et al., 2020; Liu et al., 2018a; Wang et al., 2019); the other method is
based on the retrieval from the database and literature (Ding et al.,
2020; Li et al., 2017; Zhang et al., 2017). Herbal medicine contains
multiple compounds with different polarities. Using the different
solvents for the preparation of herbal extracts will affect their
composition of compounds. Therefore, the herbal compounds in each
extract may not be exactly the same as those in the database and
literature. Compared with the retrieval from the database and
literature, the results of the compound analysis techniques (eg.
LC-MS/MS analysis) may better reflect the true composition of herbal
extracts (Liu et al., 2018a; Wang et al., 2019).
Previous research has shown that Herba Lysimachiae (HL), the dried
entire plant of Lysimachia paridiformis Franch. var.stenophylla Franch. (Primulaceae), can regulate the synovial
platelet aggregation through the biolabels (integrin alpha 2b (Itga2b)
and integrin beta 3 (Itgb3)) (Li et al., 2020). This is beneficial to
alleviate synovial injury in osteoarthritis (OA) (Li et al., 2020). The
prevalence rate of OA is 2-6% around the world and exceeds 40% in
people over 70 (Hsu and Siwiec, 2019). Biolabels, also the herb targets,
reflect the holistic effects of herbal medicines on the living organisms
(Li et al., 2020; Li et al., 2019). Biolabels can be used not only to
position the therapeutic effects of herbal medicines, but also to guide
the analysis of their active ingredients (Li et al., 2020; Li et al.,
2017; Zhang et al., 2017). Currently, the relevant active ingredients
are still unknown. Based on previous research (Li et al., 2020), we may
further use this strategy to carry out the subsequent analysis of the
active ingredients. Firstly, LC-MS/MS technique is used to analyze the
herbal compounds; subsequently, the compound targets are searched in
PubChem BioAssay database; then, STRING database is used to analyze the
association network between the biolabels and compound targets. Finally,
based on the association network, the links between herbal compounds and
biolabels are established, from which the herbal compounds with the
potential to regulate the biolabels are screened (Figure 1). These
compounds may be the active ingredients of herbal medicines in treating
diseases. In the current study, the treatment of OA by HL was used as an
example to practice this strategy. OA model was used to confirm the
analysis results.