A large number of reports have described the process of extracting and separating AM from A. pilosa. Low toxicity solvents, such as absolute methanol and 60–70% ethanol solutions, are usually used to extract AM. Conventional separation techniques, such as liquid–liquid extraction, silica gel column chromatography, recrystallization, macroporous resin, and preparative high-performance liquid chromatography, have been used to separate AM. AM yields vary from 3–4 mg/kg to 300–400 mg/kg ^{[1][8][11][22]}[1,8,11,22]. These classic separation methods are complicated, inefficient, and time consuming. However, in recent years, high-speed counter-current chromatography (HSCCC) coupled with ultraviolet detection or evaporative light-scattering detection has been considered an efficient protocol for separating AM from A. pilosa [23]. The AM yield obtained using this method is higher than that using conventional methods, producing an AM content of 770 mg/kg. Therefore, HSCCC is a powerful technique for separating AM from A. pilosa. In summary, there are substantial differences in AM yields extracted from A. pilosa owing to differences in the origin of plants, the extraction and separation methods used, extraction parts, and other potential factors. Notably, A. pilosa has a much higher content of AM than S. formosana, which explains why A. pilosa is the major source of AM.

2.2. Obtaining AM by Chemical Synthesis

The low accumulation of AM in plants, its cumbersome extraction and separation processes, and the overexploitation of natural resources are generally considered the main driving factors for its high production cost. These factors are also the main causes of supply shortages of AM. These obstacles have made chemical synthesis an appealing alternative method for obtaining AM. Although AM has a variety of pharmacological activities, there have been relatively few advances in its chemical synthesis. A few attempts have been made to synthesize AM, with variable success. In 1976, Yamato et al. synthesized the racemate of AM in five steps for the first time, and confirmed its structure using nuclear magnetic resonance imaging [24]. They started with compound 1 and obtained compound 2 by protecting two phenolic hydroxyl groups with a benzyl group. An overall yield of 2.6% AM was obtained through a series of reactions, including Stobbe condensation, ester hydrolysis, benzyl decarboxylation, bromine addition, and reduction [24]. The chemical synthesis route is shown in Figure 2A. Unfortunately, there were no subsequent reports of the chemical synthesis process of AM for many years. In 2018, a Chinese invention patent was published that contained a novel chemical synthesis method for AM [25]. The authors improved on the synthesis route of Yamato et al., and constructed the 3,4-dihydroisocoumarin core structure using microwave-assisted intramolecular esterification for the first time. The synthetic process began with 4-chlororesorcinol as the material, and the goal product was obtained in seven steps. The chemical synthesis route is shown in Figure 2B. This synthetic route increased the overall yield of AM to 20.7% ^[25][26][25,26], which is nearly eight times higher than that of Yamato et al. This exciting result suggests that microwave-assisted synthesis is a promising approach for the chemical construction of AM.

3.1. Physicochemical Properties

Physicochemical properties, such as solubility in water and organic solvents, the acid dissociation constant, the oil/water partition coefficient, and chemical stability, are key factors that affect the pharmacokinetics, biopharmaceutics, and quality of drugs. The physicochemical properties of AM are summarized in detail in Table 2 ^[27][28][29][27,28,29]. AM is a white powder with a molecular weight of 314.3 g/moL [28]. AM is soluble in chloroform, dichloromethane, ethyl acetate, dimethyl sulfoxide, acetone, and other solutions [30], but is only minimally soluble in petroleum ether. This solubility suggests that AM is lipophilic. The partition coefficient and distribution coefficient are important parameters that describe the lipophilicity of a compound, which can be used to help predict the in vivo permeability. The reported partition coefficient and distribution coefficient values of AM are 3.649 and 2.949, respectively [27], indicating that AM has low solubility and moderate permeability. Such compounds usually have good intestinal tract permeability because there is a good balance between dissolution and passive diffusion penetration. According to the biopharmaceutics classification system, drug substances are categorized into four classes on the basis of their solubility parameter and permeability to biomembranes for evaluating the desired results of a formulation on oral bioavailability [31]. A low water solubility and poor oral bioavailability limit the biological effects of many natural products in vivo. The existing data suggest that AM belongs to biopharmaceutics classification system Class II and is likely to exhibit dissolution rate-limited absorption. However, this possibility requires further confirmation by determining the solubility of AM in water. Regarding chemical stability, contact should be avoided between AM and strong oxidizing agents, reducing agents, strong acids, and alkalis. In the soluble form of AM, it should be sealed and stored below −20 °C to maintain its stability for several months. In the solid form of AM, it is stable at temperatures of 2–8 °C when kept in a dry place [30].

3.2. Predicted Absorption, Distribution, Metabolism, Excretion, and Toxicity Properties

Pharmacokinetic behaviors of drugs in vivo include absorption, distribution, metabolism, excretion, and toxicity (ADMET). The ADMET properties account for 50% of drug research and development (R&D) failures. Computer-aided design is an effective and alternative method of biological experimental evaluation, and helps to improve the R&D success rate. Computational approaches have increasingly been used to predict ADMET properties of compounds, especially in evaluating the ADMET properties of herbal medicines [32]. The predicted ADMET profiles of AM regarding its absorption, solubility, permeability across the blood-brain barrier (BBB), interactions with cytochrome P450 2D6, hepatotoxicity, and plasma protein binding (PPB) are shown in Table 3 [27]. ADMET absorption levels: 0, 1, 2, and 3 represent good, moderate, low, or very low absorption, respectively. ADMET BBB levels: 0, 1, 2, 3, 4, and 5 represent very high, high, medium, low, undefined, and molecules with one or more unknown AlogP98 types, respectively. ADMET solubility levels: 0, 1, 2, 3, 4, 5, and 6 represent extremely low, very low but possible, low, good, optimal, too soluble, and molecules with one or more unknown AlogP98 types, respectively. ADMET hepatotoxicity: 0 and 1 represent nontoxic and toxic effects, respectively. ADMET CYP2D6: 0 and 1 represent non-inhibitor and inhibitor, respectively. ADMET PPB levels: 0, 1, and 2 represent binding <90%, binding ≥90% and binding ≥95%, respectively. AM appears to show a good absorption capacity in vivo with a predicted absorption level of 0. However, AM is predicted to have a low aqueous solubility, with a solubility level of 2, which contradicts the prediction result for in vivo absorption and needs to be further confirmed by in vivo testing. Regarding the prediction of BBB penetration, AM exhibits a moderate BBB penetration capability, with a level of 2. This indicates that AM may enter the brain tissue through the BBB and could be used to treat brain diseases. Furthermore, the ADMET predictor shows that AM exhibits potential hepatotoxicity, with a level of 1. Preliminary explorations and in-depth investigations are required to determine the specific mechanism of hepatotoxicity and whether it is dose dependent. In addition, AM is predicted to be a non-inhibitor of the cytochrome P450 2D enzyme and may be metabolized and excreted successfully. Therefore, drug-drug interactions are less likely when AM and the cytochrome P450 2D6 substrates are used simultaneously. Moreover, the PPB level is predicted to be 2, indicating that the binding rate of AM with plasma protein is ≥95%. The high degree of PPB limits the partitioning of AM from the blood into the tissues, where it could be metabolized. This limited partitioning may result in a delayed onset of action and longer half-life period, thereby reducing the elimination of AM. AM has been predicted to have a good drug-likeness, with a drug-likeness weight of 0.842. Generally, AM demonstrates promising ADMET profiles. However, to fully confirm the ADMET properties of AM, real-world tests are required to validate these properties, and more animal and human studies are required. AM possesses a wide range of pharmacological activities, such as antitumor activity, antioxidation and hepatoprotection, antidiabetic activity, anti-inflammatory activity, myocardial protection, and α1A adrenergic receptor antagonist activity. The mechanisms of action of these effects are shown in Table 4.