Traditional chinese medicine-based network pharmacology could lead to new multicompound drug discovery

Abstract

Current strategies for drug discovery have reached a bottleneck where the paradigm is generally "one gene, one drug, one disease." However, using holistic and systemic views, network pharmacology may be the next paradigm in drug discovery. Based on network pharmacology, a combinational drug with two or more compounds could offer beneficial synergistic effects for complex diseases. Interestingly, traditional chinese medicine (TCM) has been practicing holistic views for over 3,000 years, and its distinguished feature is using herbal formulas to treat diseases based on the unique pattern classification. Though TCM herbal formulas are acknowledged as a great source for drug discovery, no drug discovery strategies compatible with the multidimensional complexities of TCM herbal formulas have been developed. In this paper, we highlighted some novel paradigms in TCM-based network pharmacology and new drug discovery. A multiple compound drug can be discovered by merging herbal formula-based pharmacological networks with TCM pattern-based disease molecular networks. Herbal formulas would be a source for multiple compound drug candidates, and the TCM pattern in the disease would be an indication for a new drug.

Figure 1

A diagram of drug discovery based on TCM-based network pharmacology. Each circle or diamond represents one gene or protein. Functionally interconnected genes or proteins are grouped by bioinformatics and shown in different colors. The biological networks in a disease include the general biological network of the disease (middle part) and TCM pattern network of the disease (Pattern (A) and (B) as an example). The shared disease biological networks and TCM pattern networks are shown in different colors. Drug A and drug B (in lower part) is targeting to the general biological network of the disease and molecular network of TCM pattern (A) and (B) in the disease respectively, and thus the drug A could be good for the disease with TCM pattern (A) and drug B could be good for the disease with pattern (B).

Figure 2

A conceptual model for multipl compound drug discovery using TCM-based network pharmacology. In the left of the paradigm, the molecular network of the disease-TCM pattern (lower left) can be constructed by analyzing the omics data from patients classified with the TCM pattern or related information from public databases. The typical and major TCM patterns (indicated as A, B, and C in the middle left) can be determined based on an expert consensus or literature analysis. In the right of the diagram, the most commonly used TCM herbal combinations for the treatment of a disease with specific TCM patterns can be found using text mining based on publications from SinoMed database (indicated as A, B, and C, middle right). All of the targeted proteins for the active compounds in the TCM herbal formula can be obtained in PubChem, and these targeted proteins can be used to build up the pharmacological networks for potential multiple-compound drug candidates from the herbal formulas (lower right). By matching the pharmacological networks of herbal compound combinations from the herbal formula with the disease-pattern molecular network, those well-matched compound combinations might be found for new drug candidates (capsule 1, 2, and 3 in lower part).

Figure 3

The common herbal formula for the treatment of RA with TCM cold and heat pattern obtained by text mining. All publications about clinical trials in SinoMed were collected, and common combinations of herbs (herbal formula) for the treatment of RA with TCM cold and heat pattern were found.

Figure 4

The summary on the network based on the protein targets of Fuzi, Xixin, and Guizhi built up with IPA software. (a) The summary of network analysis results; (b) the merged networks of the protein targets; (c) the canonical pathways related with the protein targets; (d) the hot map of biofunctions related with the protein targets.

Figure 5

The merged networks of identified differentially expressed genes in RA-cold pattern and protein targets of Fuzi, Xixin, and Guizhi. Red color shows the upregulated genes in RA-cold pattern compared with health; green: the downregulated genes in RA-cold pattern compared with health; Blue color shows the protein targets of Fuzi, Xixin, and Guizhi; yellow color shows the common molecular both in networks of RA-cold pattern and networks of protein targets of Fuzi, Xixin, and Guizhi. (CP: the common canonical pathways related with differentially expressed genes in RA-cold pattern and protein targets of Fuzi, Xixin, and Guizhi both; orange lines: the molecular involved in the common canonical pathways.)

Figure 6

A conceptual paradigm for the repurposing of old drugs based on TCM network pharmacology. The pharmacological networks of an old drug can be built up with bioinformatics approaches (left). By merging the pharmacological networks of the old drug with the disease-TCM pattern molecular network (indicated with (A), (B), and (C) in the right), a better indication based on TCM pattern classification might be found.