Inviting complexity: herbal medicines & human health

Application of Network Medicine & Network Pharmacology / Polypharmacology

Lately I have been thinking about the complexity of plant chemistry and the complexity of the human body. I have been wondering if we can explore the complexities of these interactions using scientific methods. Historically, scientists try to understand complex systems by breaking them down into their smaller, simpler, and fundamental parts – a reductionist approach. Are we ready for a wholistic approach?

We have successfully applied the reductionist approach to studying medicinal plants and the result has been numerous well-known drugs. Examples of these drugs include morphine and codeine from the opium poppy, and cocaine from the cocoa leaves for pain relief; atropine from the leaves and berries of belladonna to increase heart rate and dilate the pupils in the eye; vinblastine from the leaves of the Madagascar periwinkle for cancer chemotherapy and artemisinin produced in the leaves and flowers of sweet wormwood plant for antimalarial effects.

But plants contain hundreds of chemical constituents, and often, more than one of these has health benefits. On a basic nutritional level, plants contain the substances necessary for life: carbohydrates, proteins and fats. Secondarily plants contain compounds that help them to interact with the environment around them. These “secondary” compounds protect plants from bacterial and viral infections; they deter foraging animals; and attract pollinators. They also have pharmacological actions in humans.

For example, St John’s wort (SJW) flowers are known for their anti-depressant properties. Initially the active principal was deemed to be hypericin. Later, another constituent, hyperforin, was deemed to be the active. Following that, flavonoids (quercetin, rutin, isoquercitrin and hyperoside) were implicated in the herb’s mood-elevating effects. Ultimately, the overall antidepressant effect appears to be an amalgamation of each of these constituents with varying mechanisms for boosting the neurotransmitters associated with mood. In May of 2025, Professor Rudi Bauer and his research team published a paper with evidence that a SJW extract acts via the microbiome-gut-brain axis in a bidirectional manner. Their data suggests that gut bacteria intensively metabolize the phytochemicals in SJW into bioactive compounds while simultaneously altering the composition of the organisms in the microbiome.

This is just one example of the potential complexity of herbal medicines. Can we explore the simultaneous actions of multiple phytochemicals? Not just one chemical and one action at a time?

Recently I became aware of a relatively new concept called Network Medicine, which seeks to provide an alternative to the ‘one disease–one target–one drug’ approach. While acknowledging the positive outcomes for mono-target diseases, Network Medicine seeks to move beyond studying individual genes or proteins. It seeks to focus on how these elements are interconnected and how these connections are disrupted in disease.

For example, high blood pressure diagnosed as primary hypertension, means that the blood pressure is elevated, but we do not know why. Patients are treated with blood vessel-dilating drugs, such as thiazide-type diuretics or calcium channel blockers, targeting mechanistically unrelated proteins until the symptom, elevated blood pressure, disappears. We may have prevented myocardial infarctions and strokes, but the molecular cause of hypertension remains unknown. The treatment addresses the symptom but is not curative as they require chronic treatment.

Network Medicine seeks to redefine disease in a move from symptom and organ to mechanism and cause. Maps, which look something like those for computer networks, have been generated to help conceptualize this concept of the inter-connectiveness of metabolic pathways associated with disease symptoms. Network Pharmacology is a new approach to drug design that seeks to integrate a drug action network with a biological network. It proposes to reframe drug development from single pharmacological targets to a comprehensive network analysis that allows for simultaneous modulation of multiple targets.

A publication by Korean scientists (Ye et al, 2025) reported on a computational analysis attempting to map the multiple actions of Traditional Chinese Medicine (TCM) remedies with multiple herb ingredients. They mined several databases for information on the ingredient plants, their constituents and collected pharmacological activities affecting multiple targets. They then used Association Rule Mining (a computerized learning technique) to look at herbal material combinations frequently found in prescriptions for a specific disease phenotype, namely asthma, diabetes, arthritis, stroke, or inflammation. They created networks for both the pharmacological targets of the plant constituents and the disease pathways. And then laid the maps on top of each other to look at the proximity of the herb bio-targets and phenotype-related genes. In one experiment, the association rule “(Morus alba, Ephedra sinica, Perilla frutescens, Pinellia ternate) to asthma” had a confidence of 0.67, indicating that ∼67% of prescriptions containing this combination are known to be effective for asthma.

This research hints at a new way of looking at herbal medicine. It just might be the scientific approach to investigating multi component therapies for complex diseases. Time will tell whether this approach will translate to the clinic- whether this remains an academic exercise or whether it can help us redefine how we treat diseases.

Let me know what you think. Let’s imaging together!

Resources:
https://www.ebmconsult.com/articles/herbal-supplement-st-john-wort-hypericum-perforatum-mechanism-depression
https://pubmed.ncbi.nlm.nih.gov/40327993/#:~:text=Among%20others%2C%20the%20Firmicutes/Bacteroidetes,microbiome%2Dgut%2Dbrain%20axis
Ye, H. Brief Bioinform. 2025 Jul 2;26(4):bbaf328