The Tujia ethnic group is mysterious. According to China’s seventh national census, the Tujia ethnic group has a population of about 8.0 million, ranking eighth among ethnic minorities. They are known as the “毕兹卡 bil jix kar” and have lived for generations in the Wuling area, where the provinces of Hubei, Hunan, Guizhou, and Chongqing intersect. Due to its subtropical location and complex terrain, the area is suitable for plants with diverse climatic needs. Moreover, most of the plants here are in mountainous regions, and most of them have not been destroyed. Some medicinal botanical drugs that have become extinct elsewhere can survive here. Therefore, this place is known as the “Central China Natural Medicine Warehouse”, with abundant resources of various animals, plants, and mineral medicinal botanical drugs (Xiong and Shen, 2012). Unfortunately, although the Tujia ethnic group has its own language, it lacks a written form; most of its traditional medicinal practices are passed down orally, with little standardized organization and documentation. It is mainly passed down through family inheritance, apprenticeship, oral transmission, and other methods, and most of them are relatively conservative and unwilling to pass it on lightly to others (Huang et al., 2014).

The basic characteristics of Tujia humanistic medicine distinguish it from traditional Chinese medicine and other local ethnic medicines, such as Miao and Dong medicine. It has its own fundamental medical theories, unique medical methods, abundant drug resources, and special application methods (Tian and Pan, 1994). However, living in a region with diverse ethnic groups, Tujia culture, medicine, and economy are constantly exchanging, and some are even integrating to varying degrees. There are similarities and differences between the 36 symptoms of sudden illness in Tujia and Miao medicine. Both use the 36 meridian symptoms and the 72 symptoms to describe them. Still, the names of the diseases are almost identical, while their clinical manifestations differ (Tian and Pan, 1994).

From the era of indigenous primitive tribes to the Qin and Han dynasties, Tujia medicine relied entirely on “oral transmission and inheritance from generation to generation,” and this was also a period of its formation. From the Tang and Song dynasties to the late Ming and early Qing dynasties, feudal dynasties gradually implemented the “tusi system” in ethnic minority areas, including those inhabited by the Tujia ethnic group. During the Kangxi reign of the Qing Dynasty, with the introduction of traditional Chinese medicine, Tujia medicine entered a relatively mature stage guided by the theory of sensory dialectics. After the founding of the People’s Republic of China, the Party and the government attached great importance to the development of traditional medicine. They not only included the development of traditional medicine in the Constitution, but also took many measures to gradually excavate and organize the scattered Tujia medicine among the people (Yang D. S. et al., 2016).

The origin and development of Tujia medicine can be traced back to “natural philosophy. Tujia medicine believes that “all things are born for me, all things are for my use, and all things are naturally capable of nurturing and treating people. There is no incurable disease in the world, only an incurable life.” At the same time, it is influenced by dialectical thinking. Tujia medicine uses a simple materialist thinking to understand and explore the physiology and pathology of human life. He also absorbed the essence of traditional medicine. He established the “Trialistic Theory” with the Taiji Yin and Yang (The Book of Changes), the mathematical philosophy of Heluo, and the Yin Yang and Five Elements Theory (Neijing) (Zhao, 2005).

In early Tujia medicine, diseases were divided into Seventy-Two symptoms, Seventy-Two Winds, Seventy-Two Sha, Thirty-Six Tuberculosis, Thirty-Six Sores, Thirty-Six shock, and Thirty-Six injury diseases. Tujia medicine was also divided into Seventy-Two Shen, Seventy-Two Qi, Seventy-Two Lian, Thirty-Six Feng, Thirty-Six Huanyang, Thirty-Six Wugong, and Thirty-Six Xue. The Tujia people’s use of these numbers to name and classify diseases and medicines is not merely a simple mathematical description. It also has unique characteristics. Among them, “Seventy-Two” is derived from the Eight Trigrams of the Tai Chi Scripture of the “Shennong Yi” and the sixty-four hexagrams derived from it. Tujia medicine holds that it can account for the occurrence and development of all things in nature, including human life and diseases, as well as their auspiciousness and inauspiciousness. Originating from Qimen Dunjia, during the late Yin and early Zhou dynasties, Jiang Taigong and Lv Shang observed and calculated that there were three solar terms per section, one yuan per quarter, and twenty-four solar terms per year, seventy-two quarters. Tujia medicine believed in Jiang Taigong’s seventy-two divine calculation quarters. According to legend, the leader of the Tongtian Sect deployed the Thirty-Six Gang Formation, which was broken by Jiang Ziya, leading the Twenty-Eight Star Constellations and the Eight Marshals (Fang et al., 2007). Therefore, the Tujia Medical Thirty-Six were classified into some medicines and diseases. The Tujia medical values of thirty-six and seventy-two are regarded as the beliefs and value orientations of Tujia medical ethics, benevolence, and righteousness, and benevolent techniques.

In the Tujia language, “Lian” does not refer solely to lotus flowers, but is a general term for a specific type of plant that shares similar morphology or functions and is also related to its growth environment. It is usually an botanical drug aceous plant with leaves resembling lotus leaves, and it prefers moist conditions. Some have cleansing and detoxifying effects. These “Lian” drugs are primarily used to clear heat and detoxify, promote blood circulation, relieve pain, dispel dampness and reduce swelling, dispel wind and cold, among other uses. They are the core group of drugs used by the Tujia ethnic group to treat common conditions such as wind-dampness, snakebite, and inflammation. Knowledge of Tujia medicine is mostly passed down orally, heart to heart, and the “72 Lian,” as a medicinal experience, has been passed down from generation to generation, reflecting the profound understanding of the Tujia people towards local plant resources. These drug names are often combined with legends and folk customs (“Shenlian treats snake wounds”), becoming a part of national culture. Although the “Lian” of the Tujia ethnic group is mainly unrelated to lotus flowers, the term “Lian” may still borrow the symbolic meaning of “Lian” in Han culture, metaphorically referring to the “detoxification and purification” function of medicine. In some Tujia areas, influenced by Han culture, lotus flowers are also used in folk rituals. The naming of Tujia medicine combines form, efficacy, ecology, and culture, reflecting detailed observation of the natural world and embodying the simple pharmaceutical philosophy of “naming based on form and classifying based on efficacy.” This classification system crystallizes the wisdom of the Tujia ethnic group in adapting to the ecological environment of the Wuling mountainous area and in accumulating medical experience, with distinct ethnic and regional characteristics (Yuan, 2007).

The Wuling mountainous region, which spans across Hunan, Sichuan, Guizhou, and Hubei provinces, is rich in medicinal resources (Fan et al., 2018). However, Tujia medical knowledge is mostly passed down orally, and errors can occur during its transmission and use. A common issue is the misnaming of medicinal plants, with many species being called by the same name or a single species having multiple names. For example, four different species of Tujia medicinal plants are known as “Guanyinlian,” and three species are called “Huoxuelian.” Additionally, plants from the same family and genus but different species are often mistaken for one another. For instance, “Dengtailian,” which originates from various plants within the Araceae family and genus, and “Tiexianlian,” derived from different plants in the Ranunculaceae family and Clematis genus, are listed separately in different Tujia Medical Classics (Zhao, 2005). Moreover, there are inconsistencies in the collected contents.

In Tujia ethnomedicine literature, specific drug names exhibit homophonic variation in written form despite identical pronunciation, primarily due to dialectal transliteration, historical changes, and regional differences. Examples include: “岩桥莲” (Yanqiaolian) versus “岩荞莲” (Yanqiaolian), “乱脚莲” (Luanjiaolian) versus “乱角莲” (Luanjiaolian), “碧血莲” (Bixuelian) versus “鼻血莲” (Bixuelian). This review compiles 96 “Lian” category medicinal materials from 46 plant families, with 101 botanical origins (Fang et al., 2007). For each entry, we record: standardized names and aliases, botanical origins, property and flavor profiles (e.g., “cold, bitter”), and therapeutic uses (Supplementary Table 1).

Due to their age, some plant names and Latin names have been revised and updated. Among these plants, only Clematis chinensis, from the Ranunculaceae family, has been included in the list of “Tiexianlian” drugs; other varieties, such as Clematis peterae, Clematis lancifolia var. Ternata, C. peterae var. Trichocarpa, Clematis quinquefoliolata, Clematis kweichowensis, and “Tiexianlian” from the Acanthaceae family, have not been included in Supplementary Table 1 and still require further verification and confirmation.

The development of quality standards for Tujia ethnic medicine is at a critical stage, transitioning from local customs to national norms, and its system is multi-level but not yet fully unified. Due to the traditional Tujia medicine preaching method, which relies mainly on oral transmission, apprenticeship, and folk transmission, the phenomenon of “same name foreign object” and “same object but different names” is common. Owing to regional differences and oral transmission history, the same soil name may correspond to different plant families and genera, and the same medicinal botanical drug may have different Tujia-language names in other regions. Traditional quality evaluation primarily relies on sensory experience (seeing, touching, smelling, and tasting) and pharmacists’ practical experience, and lacks objective, quantitative indicators. At present, the primary basis is the inclusion of local standards. For example, the (Quality Standards for Traditional Chinese Medicine in Hubei Province) includes a large number of Tujia medicinal materials from Enshi and other places in Hubei Province, which is a pioneer and essential metabolite of Tujia medicine standardization; The (Hunan Province Traditional Chinese Medicine Standards) include the medicinal materials of the Tujia ethnic group in the Xiangxi region. The local standards of Chongqing city and Guizhou Province also include Tujia medicine in their respective jurisdictions. These local standards usually include: product name (Chinese name, Tujia language name, Latin scientific name), source (clarifying the family, genus, species, and medicinal parts), trait description (traditional experience identification), identification (microscopic identification, thin-layer chromatography identification), inspection (safety indicators such as moisture, ash content, impurities, heavy metals, and pesticide residues), extract, content determination, taste and functional indications based on Tujia medicine theory such as the “Three Elements” theory. There are also specialized books, such as “Tujia Medicinal Chronicles,” that systematically organize the origins and application experience of medicinal materials, serving as essential foundations for developing standards (Yuan, 2007). Although there is currently no fixed quality standard for Tujia medicine, given the current development trend, efforts are being made to translate Tujia medicine experience into modern scientific standards, using scientific language to safeguard ancient wisdom and enable it to safely and effectively benefit more people.

The “Lian” drugs, known for their anti-inflammatory and analgesic effects, along with their corresponding bioactive fractions or metabolites, are shown in Supplementary Table 2. Terpenes are among the most prominent classes of anti-inflammatory metabolites, including neo-clerodane diterpenoids, hederagenin, gypenoside L, XLIX, LXXV, and cucurbitacin-type triterpenoids. Flavonoids are the most widely present anti-inflammatory metabolites and form the material basis for the activity of most plant extracts, including cynaroside, hyperin, scutellarein, luteolin, and quercetin. Anti-inflammatory efficacy is closely related to the number and positions of hydroxyl groups (-OH), especially the adjacent dihydroxy group on the B ring and the 4-carbonyl group in the C ring. Quercetin, one of the most well-known flavonoids, is almost ubiquitous and a potent antioxidant and anti-inflammatory agent. Saponins have surface activity, and many have anti-inflammatory and immunomodulatory effects. Among them, Dioscin and Deltonin are derived from Dioscorea plants and are steroidal saponins with anti-inflammatory and anti-tumor activities. Aescin/escin is a classic anti-inflammatory drug metabolite; lignin metabolites are derived from phenylpropanoid units and often exhibit anti-inflammatory, antioxidant, and hepatoprotective activities. Among them, Podophyllotoxin, a famous anti-tumor lignan, also has anti-inflammatory activity. The vast majority of active metabolites have been reported to inhibit the two central inflammatory signaling pathways, NF-κB and MAPK. This underpins the “broad-spectrum” anti-inflammatory effect (reducing the expression of TNF-α, IL-6, IL-1β, iNOS, and COX-2). Most studies follow the complete logical chain of “network pharmacology prediction, in vitro cell validation (RAW 264.7/BV-2), and in vivo animal models (LPS, DSS, CIA, MCAO, etc.)”, with a transparent chain of evidence. However, this also highlights the shortcomings of its single, mechanized in vitro model. The RAW 264.7 mouse macrophage line has been overused, and its response to LPS may not fully represent the complex behavior of human primary macrophages or specific tissue macrophages. Many studies rely solely on NO inhibition as the primary indicator, neglecting the assessment of other critical inflammatory mediators and cellular functions. Acute models dominate, while chronic and low-grade inflammation models (aging models and high-fat diet-induced metabolic inflammation) are poorly studied, with the latter being closer to the pathological state of most human chronic diseases. A large amount of literature remains at the stage of isolating new metabolites and testing their activity at a single high concentration (10–40 µM). There is a Lack of systematic structural optimization and structure-activity relationship research on metabolites. Although target hypotheses have been proposed through molecular docking and inhibitor experiments, research directly demonstrating the eutectic structure, binding constants, or direct intracellular interactions between metabolites and hypothesized target proteins is rare. Network pharmacology predictions often provide dozens of potential targets, lacking experimental focus. This study believes that more human primary immune cells, organoids, and organ chip models should be used to validate experimental results and better simulate the human microenvironment. Their development and application should also be closer to complex animal models of human diseases. Directly identify and validate biological targets using chemical biology tools, such as active-molecule probes and photoaffinity labels, without relying on correlation-based speculation, and establish a standardized process for screening and evaluating the anti-inflammatory activity of natural products, including a unified positive control, cell model, and detection indicators. It is mandatory to complete pharmacokinetic and safety evaluations of GLP-compliant systems in the preclinical stage, laying a solid foundation for clinical trials.

The “Lian” drugs, known for their antibacterial properties, are also associated with active fractions or metabolites (Table 1). Earlier research mainly focused on flavonoids, saponins, and alkaloids, as well as solvent fractions separated by polarity (ethyl acetate and n-butanol). These studies correctly indicated that antibacterial activity was enriched in specific polarity ranges, and as research progressed, characteristic metabolite categories were isolated and identified. Flavonoids (quercetin, quercitrin, afzelin, scutellarein, and flavonoid glycosides) exhibit antibacterial activity, often via membrane interference and enzyme inhibition. The glycosylation type and hydroxylation pattern significantly affect their activity and selectivity (making them more effective against Gram-positive bacteria). Terpenes and their derivatives, including triterpenoid saponins, volatile oils, hyperforin, and anemin, have hydrophobic glycosides and hydrophilic sugar chains that together determine their surfactant properties, leading to membrane perforation and content leakage, with outstanding effects on Gram-positive bacteria and fungi. Phenolic acids and polyphenolic metabolites (phenolic acids, phenolic metabolites, polyphenols, ellagic acid, ellagitannins) show activity related to protein precipitation, metal ion chelation, and enzyme inhibition. Alkaloids, such as lappaconitine and benzilisoquinoline alkaloids, have diverse antibacterial mechanisms, often involving interference with nucleic acid or protein synthesis. The study not only reports MIC values but also uses techniques such as molecular docking, enzyme inhibition, and gene knockout/knockdown to link specific chemical structures to key bacterial targets directly. The agar diffusion method (inhibition zone) and the broth dilution method (MIC/MBC) are widely accepted mainstream methods for antibacterial models. However, over 90% of the research remains in static in vitro models. The lack of dynamic PK/PD models, validation of in vivo infection models, and consideration of the host immune environment results in model singularity and clinical disconnection. The definition of antibacterial activity is too broad, and many studies confuse antibacterial and antifungal activities. However, the mechanisms and targets of the two differ significantly, requiring more precise classification of mechanisms. The future research paradigm must shift from a single in vitro activity report to an integrated study centered on depth of mechanistic understanding, in vivo validation, drug efficacy evaluation, and clinical needs.

The “Lian” drugs, which exhibit antioxidant properties, contain various bioactive fractions or metabolites (Table 2). These include flavonoids (quercetin glycosides, luteolin-7-O-β- D-glucoside, scutellarin), phenolic acids, and polyphenolic metabolites (gallic acid derivatives, tannins, neochlorogenic acid). Except for a small number of single metabolites, most are oversimplified by “total flavonoids/total phenols, with only the content of “total flavonoids” or “total phenols” measured and correlated with antioxidant activity (DPPH scavenging). This approach lacks specificity, making it impossible to determine the truly effective monomers or to rule out synergistic or antagonistic effects among metabolites. The antioxidant mechanisms of the polysaccharide metabolites (acid-assisted polysaccharides, Yam polysaccharides) are usually associated with enhancing endogenous antioxidant enzymes (SOD, CAT) and immune regulation, rather than directly clearing free radicals. Biochemical assays (DPPH, ABTS, FRAP, ORAC, and other in vitro chemical assays) are necessary but insufficient for preliminary screening. Very few studies progress to cellular oxidative stress models (H₂O₂-induced injury), and even fewer enter animal models to verify their in vivo antioxidant efficacy (by detecting SOD, GSH, MDA in liver and brain tissues). The in vivo models used include the D-gal aging model, MCAO model, DSS model, and OVX model, which can link antioxidant activity to specific pathophysiological outcomes (cognitive function, infarct size, inflammation, and bone density), thereby significantly strengthening the evidence. However, there are still species differences in animal models, and the vast majority of studies lack data on the pharmacokinetics and bioavailability of metabolites in vivo, which is a key bottleneck in translating activity into practical applications.

The “Lian” drugs, along with their bioactive fractions or metabolites that exhibit anti-tumor effects, are shown in Supplementary Table 3. Flavonoids (quercetin, luteolin, scutellarin, kaempferol-3-O-glucoside) are commonly used in CCK-8, flow cytometry, WB, and CDX models. In vivo effects are often weaker than in vitro effects due to low bioavailability and rapid metabolism. Although the mechanisms are well studied, low bioavailability makes it difficult to achieve therapeutic concentrations. Terpenoids and their derivatives (triterpenoid saponins, steroid saponins, gypenosides, oleanolic acid, curcurbitacin) have poor oral absorption, unclear in vivo metabolism, high toxicity, and a narrow therapeutic index. Cucurbitacin, a classic inhibitor of the JAK2/STAT3 signaling pathway, induces cancer cell apoptosis. Alkaloids such as lappaconitine, cephalantine, and sinomontanine are commonly used for cytotoxicity testing of human cancer cell lines (MTT/CCK-8) and CDX models. Their toxicity to the heart and nervous system may limit their therapeutic window, and they lack targeted validation, have low bioavailability, and require formulation optimization or structural modification. Most studies on lignans, such as podophyllotoxin, sauchinone, and machilin, have been limited to the extract level, and the mechanisms of individual metabolites have not been thoroughly studied. Podophyllotoxin is a classic example of a natural product successfully transformed into modern drugs. Still, its high toxicity also warns us that potent anti-cancer activity often comes with serious side effects. Currently, most research is still at the stage of phenotype observation, lacking in-depth exploration of mechanisms and assessment of transformation potential. Future research should focus on target identification, structural optimization, pharmacokinetic improvement, and systematic analysis of mechanism networks for monomeric metabolites, and on validating these findings using more clinically relevant models.

The antiviral effects of “Lian” drugs derived from botanical sources, along with their specific active fractions or pure metabolites, are noteworthy (Table 3). Flavonoids generally exert µM-level activity by inhibiting viral entry (HA/NA/ACE2) and host protease (protease L), which reflects their non-specific effects. They may interfere with multiple viral life processes through slight membrane stabilization, antioxidant effects, or non-specific protein binding. However, in the human body, the concentrations of their original metabolites never reach the adequate micromolar levels observed in vitro. In addition, extracts account for a large proportion, and their scientific value is limited to suggesting that the source is worth further research. The vast majority of studies use laboratory-adapted viral strains in single, highly susceptible cell lines (Vero or MDCK), which are far from the complex multicellular environment, immune system stress, and the behavior of clinical isolates in the human body. The current research on plant antiviral metabolites is severely limited, and future breakthroughs will focus on chemical transformations from crude extracts to specific monomers and on biological transformations from simple cell models to physiologically relevant systems as precursors.

Table 4 presents the “Lian” drugs known for their insecticidal effects, along with their active fractions or metabolites. According to the records, 10 distinct “Lian” drugs show insecticidal effects. The primary insecticidal metabolite responsible for these effects is alkaloids. This is highly consistent with the ecological strategy of plant defense against insect feeding. Alkaloids (yajiaguan delcosine, belsoline, lepene, demethylenedelcorine, 18-O-methyligactonine) are the most important class of food repellents, gastric toxins, and nerve agents in the plant kingdom. They are central to research on natural insecticidal activity. Their mechanism is to block the transmission of nerve impulses, leading to paralysis and death. However, their most significant limitation is poor selectivity and an extremely high risk to non-target organisms, which severely limits their prospects for direct application as insecticides. Terpenoid metabolites (E)-β-farnesene, α-pinene, and oleic acid are typically highly volatile and primarily affect behavioral regulation and neurotoxicity. Their fumigation and contact-killing effects are usually non-selective and may harm pollinating insects such as bees and natural enemies. The study of their mechanisms of action usually remains at the behavioral level, lacking precise molecular target evidence. Model singularity, with the vast majority of studies using only 1-2 laboratory model insects, makes it difficult to predict their actual effects on complex pest populations in the field and lacks safety evaluations for non-target organisms such as natural enemies and pollinating insects. The confusion over dosage concepts has led many studies to provide only concentration gradients during screening, without calculating key toxicological parameters or comparing them with those of commercial insecticides, thereby preventing objective evaluation of their activity levels. Future research will focus on systematically studying the structure-activity relationships of the discovered highly active lead metabolites and on optimizing their efficacy, selectivity, and stability through chemical modification. Research on sustained-release technologies, such as microencapsulation and nanocarriers, for volatile and biodegradable metabolites to extend their shelf life, reduce usage frequency, and dosage.

Table 5 lists the “Lian“-derived antidiabetic drugs and their active fractions or pure metabolites. The data indicate that 17 different “Lian” drugs exhibit antidiabetic effects, primarily due to flavonoids and saponins (escins/escigens, dioscin, gypsides, dammarane-type triterpenoids), which have been widely shown to significantly improve insulin sensitization, glucose and lipid metabolism, and more. The frequent occurrence of flavonoids (quercetin, luteolin, kaempferol, apigenin) and water/alcohol extracts highlights the multitarget regulatory properties of flavonoids in metabolic diseases, reflecting direct research on traditional decoctions or tinctural forms. Natural products are often used as modulators with slow onset and weak efficacy, making it difficult to compete with potent chemical hypoglycemic drugs (such as insulin and sulfonylureas) for acute hypoglycemic effects. Moreover, the depth of the mechanism is insufficient, and many studies focus only on phenotype observation (lowering blood sugar and improving glucose tolerance), lacking in-depth exploration of exact molecular targets, signaling networks, and long-term medication safety. Therefore, natural products are more suitable for intervention at the early stage of diabetes or as adjunctive treatment for type-2 diabetes, combined with first-line drugs, thereby reducing the need for chemical drugs and their side effects. Taking specific saponins as lead metabolites and developing new insulin sensitizers with higher safety is one of the most promising directions at present.

Table 6 presents a detailed overview of “Lian”-derived drugs with hepatoprotective effects, including their active fractions or pure metabolites. The data show that 16 “Lian” drugs possess hepatoprotective effects. Mechanistic research on saponin metabolites (aescin, dioscin, gypenosides, echinocystic acid, eclalbasaponin II) has progressed from simple antioxidant effects to energy metabolism, organelle homeostasis, and cell fate regulation, aligning more closely with the modern understanding of liver disease pathophysiology. Hepatoprotective activity has been confirmed in chemical (CCl₄, acetaminophen), alcoholic, and immune liver injury models. However, oral bioavailability is low, and some saponins may exhibit hemolytic or hepatorenal toxicity at high doses, resulting in a “double-edged sword” effect. Its effectiveness and safety window in the human body need to be strictly defined; The core mechanism of flavonoids is antioxidant stress. These metabolites have significant free-radical scavenging effects and can directly neutralize reactive oxygen species (ROS). The Nrf2 pathway has become a “classic” target for liver protection by these metabolites, and the evidence chain is relatively complete. However, most studies have focused on total flavonoids or individual known metabolites. There is a lack of systematic comparison of how glycosylation, acylation, and other modifications affect bioavailability, target affinity, and overall efficacy. In addition, their antioxidant effects may be weakened in complex in vivo environments, and high doses may produce pro-oxidative effects. The model is too simple, and a large number of studies rely on acute chemical liver injury models induced by CCl₄ or acetaminophen, which differ from the pathogenesis of major chronic liver diseases in humans (fatty liver, viral hepatitis, and alcoholic liver). Advanced in vitro models should be vigorously promoted and standardized, and liver-like organs and microfluidic liver chips should be included in the drug screening and mechanism research system to evaluate the comprehensive effects of metabolites in multidimensional pathological processes such as metabolism, inflammation, and fibrosis.

Table 7 provides a detailed overview of “Lian”-derived drugs that show efficacy in offering cerebral protection, along with their active fractions or metabolites. The data presented in the table indicate that six “Lian” drugs exhibit cerebral protective activity. However, no clear pattern has been identified among the metabolites responsible for these effects. Only certain alkaloids (stachydrine, indole, oxindole) can directly act on the central nervous system, penetrate the blood-brain barrier, and affect neurotransmitters, receptors, and ion channels. However, alkaloids exert a strong regulatory effect on the nervous system, making their treatment window usually very narrow. For example, alkaloids that act on the dopamine system may induce psychiatric symptoms or motor disorders. The neurotoxic risk of excessive sedation or dependence caused by actions on the GABA system must be rigorously evaluated. More extracts are used as active metabolites, and their mechanisms of action are mainly described in terms of phenotypes such as “antioxidant stress” and “anti-inflammatory”, which reflect the comprehensive effects of multiple metabolites. However, it remains unclear which metabolites achieve this, as the signaling pathways are poorly explored. Moreover, most studies in the literature use young, healthy male animals, ignoring the impact of key biological variables such as age and gender on stroke outcomes, which reduces the potential for clinical translation. In the future, activity-tracking methods should be adopted to isolate and identify the true key functional metabolites from crude extracts. Metabolomics and network pharmacology methods should be used to explore interactions within the metabolite-target group pathway network, establish quality control standards based on key marker metabolites, and ensure the reproducibility of experimental materials.

The “Lian” drugs, known for their neuroprotective effects and active sites or metabolites, are detailed in Table 8. Among them, saponin metabolites (aescin/escin, deltonin, steroid saponins, eclalbasaponin II, gypenosides, ombuoside, gypenoside XVII) exhibit strong anti-inflammatory, antioxidant, and anti-apoptotic activities and can protect neurons through multiple pathways. Saponins are polar macromolecules with an extremely weak ability to penetrate. The neuroprotective effects observed in animal models are likely due to their strong peripheral anti-inflammatory and antioxidant effects, which improve the systemic environment, indirectly reduce the stress load on the central nervous system, or enter pathological models of BBB damage (cerebral ischemia and trauma). For an intact blood-brain barrier (BBB), their ability to achieve effective concentrations in the brain is questionable, severely limiting their potential as direct therapeutic drugs for central nervous system diseases. Many studies have focused only on detecting changes in the expression of a few classic pathway proteins (p-Akt, Bcl-2/Bax), and have not revealed which metabolites in the extract, which upstream receptors or sensors activate this signal, or which disease models use a single approach and do not match clinical heterogeneity. All current studies are mostly acute or subacute intervention experiments, lacking evaluation of the toxicity, tolerability, and long-term neuroprotective effects of metabolites under long-term administration, which is a gap that must be filled before clinical translation.

Sauchinone in Baihelian (Saururus chinensis (Lour.) Baill.) also exhibits notable pharmacological properties (Shi et al., 2019). Xiang et al. (2011) also found that the extract and metabolite emodin-8-O-β-D-glucoside (E.G.,) from Daosilian (Bistorta amplexicaulis subsp. Sinensis (F. B. Forbes & Hemsl. ex Steward) Soják) directly stimulates the proliferation and differentiation of osteoblasts. The purified lectin from Dengtailian (Arisaema erubescens (Wall.) Schott) agglutinates red blood cells of rabbits (Liu et al., 2011), mice, and dogs but not from chickens and humans. The flavonoid metabolite Schaftoside in Dengtailian inhibits the increase in melanin production in B16F1 cells stimulated by α-melanocyte-stimulating hormone. It also downregulates the expression of tyrosinase (TYR) and tyrosinase-related protein 1 (TRP1) and inhibits melanin production by activating autophagy in melanocytes (Kim et al., 2018). In the drug-induced liver injury (DILI) mouse model (Jiang et al., 2021), Gouyabanzhilian (Sedum sarmentosum Bunge) reduces inflammation and regulates the Nrf2-ARE cascade. It not only supports normal liver cell growth and inhibits APAP-induced liver cell apoptosis but also suppresses the expression of Nrf2-ARE proteins in the liver tissue of the DILI mouse model. This suggests that regulating the Nrf2 signaling pathway helps the active metabolites in Gouyabanzhilian prevent drug-induced liver injury. Zhou J et al. (Zhou et al., 2020) found that Guanyinlian (Dioscorea zingiberensis C. H. Wright) can significantly reverse the destruction of the blood-testis barrier (BTB), testicular tissue damage, and abnormal sperm morphology in diabetic mice, and protect against diabetes-induced testicular damage. Du et al. (Du et al., 2017) isolated a diarylheptane from D. zingiberensis C. H. Wright and observed that its anti-pancreatitis activity against taurocholic acid sodium salt hydrate (NaT)-induced pancreatic acinar necrosis was superior to that of caffeine. Research shows that Jizhualian (Sceptridium ternatum (Thunb.) Lyon) extract has a significant therapeutic effect on radiation-induced pulmonary fibrosis (RIPF) in vitro and in vivo, suggesting that it may exert anti-RIPF effects by regulating the EGFR/p38-MAPK/NF-κB/CEACAM1 signaling pathway (Zhang Y. et al., 2025). It may also exert beneficial effects against pulmonary fibrosis by targeting the SETDB1/STAT3/p-STAT3 pathway, indicating its therapeutic potential (Zou et al., 2023). When administered at 1.4 g/kg for 12 weeks, Mohanlian (Eclipta prostrata L.) botanical drug extract (Feng et al., 2019) effectively increased bone mass and suppressed weight gain in ovariectomized rats, with no adverse effects observed. This outcome is attributed to dual inhibition of bone formation and resorption, with resorption being more prominent, possibly involving downregulation of the key osteoclast factor RANKL. Shen et al. (Shen et al., 2024) found that the total glycosides of Gynostemma pentaphyllum in Qiyelian (G. pentaphyllum (Thunb.) Makino) significantly increased the expression of key proteins in the LOX1-PI3K-AKT-eNOS pathway, thereby improving hyperlipidemia. Leonurine in Yimulian (Leonurus japonicus Houtt.) prevents Ang II-induced cardiac remodeling and dysfunction by inhibiting the MAPK/NF-κB pathway (Shen et al., 2023). It also alleviates acute ischemic kidney injury by activating Nrf2 to counteract oxidative stress and inhibiting TLR4/NF-κB-mediated inflammatory gene expression (Han et al., 2022).