Desoximetasone

[图片]Towards the targeted activation of silent biosynthetic gene clusters by chemical elicitors（诱导剂）AbstractSystematic genome mining has revealed that microbes（微生物） encode numerous uncharacterized secondary metabolite biosynthetic gene clusters (BGCs). The efficient and selective activation of these silent or cryptic（隐性的） BGCs is crucial for the high-throughput（高通量） discovery of novel natural products. Recent influential studies have demonstrated that using small chemical elicitors is a practical and cost-effective（经济高效） method to unlock the secondary metabolic potential of microbes. However, the current approach mainly relies on high-throughput, non-targeted screening methods to discover chemical elicitors capable of activating these silent BGCs. Therefore, this study comprehensively reviews reported cases of small molecules that activate silent BGCs, covering the chemical structures of elicitors, resulting natural products, and target BGCs, thereby constructing an integrated knowledge graph. We also summarize the underlying activation mechanisms. Leveraging（利用） relationships captured（捕获） in this graph, we outline directions for targeted activation of silent pathways using small molecules, thereby facilitating more efficient natural product discovery.1. IntroductionThe discovery of novel natural products holds significant importance across various scientific fields due to their remarkable structural diversity, which leads to a wide range of biological activities and applications, including pharmaceuticals, food, cosmetics（化妆品）, and nutraceuticals（营养保健品）. With the advancement of next-generation sequencing technologies and the development of computational genome mining, the discovery of natural products has entered a genome-driven era (Navarro-Muñoz et al., 2020; Ziemert et al., 2016). In this era, genomic data is leveraged to uncover the secondary metabolic potential of microorganisms, guiding the discovery of novel natural products (Harvey et al., 2015). Numerous genome mining studies have revealed that the majority of the biosynthetic gene clusters (BGCs) encoded by microorganisms remain untapped（未开发的）. For example, Gavriilidou et al. estimated that only about 3% of the natural products encoded in bacterial genomes were experimentally characterized (Gavriilidou et al., 2022). Our recent study indicated that 96.8% of the secondary metabolic potential in marine prokaryotes（原核生物） remains unexplored (Wei et al., 2023).原核生物prokaryotes；真核生物eukaryotes The primary reason for this gap between the natural products that have been characterized and those identified through genome mining is that many BGCs are silent or expressed at low levels under laboratory conditions, thus limiting the production and characterization of these potential natural products (Xu et al., 2019).Many strategies have been used to activate silent BGCs, leading to the discovery of novel secondary metabolites from diverse microorganisms. These strategies are broadly divided into two categories: stimulation（刺激） and genetic activation（遗传激活） (Baral et al., 2018; El-Hawary et al., 2023; Mao et al., 2018; Zong et al., 2022). Stimulation involves using external factors to activate silent BGCs without altering the organism's genetic composition. This category of strategies can be classified into three types: chemical elicitation（诱导）, biological elicitation, and environmental perturbation（扰动）. When microorganisms are grown in pure culture in vitro, many activating signals are absent, leading to the downregulation of specialized metabolite BGCs. To address this, co-culturing with competing species (biological elicitation) or adding chemical elicitors (chemical elicitation) can induce changes in BGC expression (Rutledge and Challis, 2015). Additionally, altering environmental conditions (environmental perturbation), such as nutrient levels, pH, or oxygen availability, can stimulate BGC activity (Romano et al., 2018). Genetic activation, on the other hand, directly manipulates genetic elements to induce BGC expression. Examples include manipulating global or pathway-specific regulators (Lu et al., 2017; Mao et al., 2017; Otur and Kurt-Kızıldoğan, 2024), reporter-guided mutant selection (Mao et al., 2018), refactoring（重构） (He et al., 2024), and heterologous expression (Kang and Kim, 2021). However, these approaches frequently mandate（要求） intricate manual design and often have low success rates, particularly in heterologous expression because compounds may be biosynthesized through the collective contribution of multiple gene clusters (van Bergeijk et al., 2020). By contrast, chemical elicitation offers simplicity, ease of operation, and good reproducibility. In recent years, high-throughput（高通量） elicitor screening (HiTES) has been successfully employed to activate silent BGCs. HiTES can use various readouts, including transcriptional reporter assays (Xu et al., 2017), as well as genetics-free techniques such as bioactivity assays (bioactivity-HiTES) (Moon et al., 2019b), mass spectrometry detection (MS-coupled HiTES) (Covington and Seyedsayamdost, 2021), and fluorescence resonance energy transfer-based assays (FRET-HiTES) (Han et al., 2023), enabling efficient discovery of elicitors in a multi-well plate format. Over the past decade, chemical elicitation has gained widespread attention and application.In this review, we define a focused scope on studies in which chemical elicitors were explicitly（明确） shown to activate silent or lowly expressed BGCs, leading to the production of cryptic natural products. We note that, in some cases, the addition of small molecules results in the discovery of new structural analogues rather than genuine activation of a silent BGC. In such scenarios, the supplemented compounds often function as metabolic precursors, thereby altering substrate availability within an already active pathway. For example, feeding Streptomyces chartreusis NRRL 3882 with 20 different proteinogenic amino acids revealed that calcimycin production and the distribution of its analogs are strongly dependent on amino acid supply, with glutamine yielding the highest calcimycin levels (Arend and Bandow, 2021). Additional examples of precursor-driven modulation of product biosynthesis have been reviewed elsewhere (Yan et al., 2024). These effects, while valuable for expanding chemical diversity, do not constitute activation of a transcriptionally silent BGC and are therefore not discussed further in this review.While chemical elicitor activation is conceptually simple, the identification of effective elicitors remains a major challenge. Current strategies, such as HiTES, allow for efficient screening of large compound libraries but are largely empirical and lack a strong theoretical foundation. To address this limitation and enable more targeted elicitor selection, we first construct a comprehensive knowledge graph of known elicitors based on data curated（整理） from the literature, providing a structured knowledge base. In parallel, we review and generalize experimental paradigms for elucidating the regulatory mechanisms governing BGC expression, with an emphasis（强调） on omics（组学）-based regulatory networks. Building on these two foundations, we propose a set of guidelines to support the rational selection of small-molecule elicitors. Through this review, we aim to promote the targeted activation of silent BGCs and, ultimately, facilitate the discovery of novel natural products.2. Overview of chemical elicitors in activating silent BGCsDozens of small molecules have been reported to activate silent or poorly expressed gene clusters in microorganisms, leading to the production of new natural products. Here, we construct a comprehensive knowledge graph comprising 32 BGCs, which can be activated or upregulated by 54 small-molecule elicitors, leading to the synthesis or increased production of 87 representative secondary metabolites (Fig. 1 & Table S1). This knowledge graph was manually curated from published literature and designed as a conceptual integration framework. Specifically, elicitor–BGC–metabolite relationships were systematically extracted from peer-reviewed studies reporting experimentally validated activation or significant upregulation of silent or lowly expressed BGCs. Small-molecule elicitors were classified primarily based on their reported mechanisms of BGC activation. Accordingly, elicitors were grouped into three broad classes, including stress-associated elicitors, signaling-associated elicitors, and others that no clear mechanisms are discussed. Among these, stress-associated elicitors, representing the most extensively reported class, were further subdivided according to chemical nature and ecological context, including antibiotics, non-antibiotic clinical drugs, host-related molecules, and chemically modified molecules. This mechanism-oriented classification framework is intended to provide a conceptual guide for sourcing effective elicitors.2.1. Stress-associated elicitorsStress-associated elicitors refer to small molecules that activate silent or weakly expressed BGCs primarily by inducing physiological stress responses. In this context, elicitation arises from global or semi-global perturbations of cellular（细胞的） homeostasis（稳态）, including disturbances to redox（氧化还原反应） balance, DNA replication, transcription, translation or metabolic flux.[图片]Fig. 1. A knowledge graph integrating reported chemical elicitors (triangles, n = 54, labeled as E1–E54), their induced BGCs (circles, n = 32), and metabolic products (diamonds, n = 87, labeled as C1–C87), corresponding to entries in Table S1. Elicitors are color-coded by category (stress-associated, signaling-associated, and others). Dashed arrows（虚线） indicate induction relationships, whereas solid arrows denote biosynthetic relationships.2.1.1. AntibioticsAmong stress-associated elicitors, antibiotics applied at sub-inhibitory concentrations represent one of the most extensively studied and experimentally validated classes. A series of HiTES-based studies by Seyedsayamdost group found that the majority of effective elicitors identified were antibiotics, highlighting their important role in activating silent biosynthetic pathways. For instance, the β-lactam（β-内酰胺类） piperacillin (E1) and dihydrofolate reductase inhibitor trimethoprim (E2) were found to activate the mal gene cluster in the model pathogen Burkholderia thailandensis E264. These compounds induced a 10–145-fold overproduction of malleilactone A (C1) and facilitated the discovery of a new analog, malleilactone B (C2). In addition to the mal gene cluster, trimethoprim, as a global activator, also activated at least four biosynthetic pathways. For example, it activated the tha gene cluster of B. thailandensis E264, leading to the production of thailandamide A (C3). Additionally, it induced the hmq gene cluster, leading to a sevenfold increase in the yield of the quinolone antibiotic and quorum-sensing (QS) modulator 4-hydroxy-3-methyl-2-(2-nonenyl)-quinoline (HMNQ, C4). Trimethoprim, as well as the Streptomycete-derived antitumor antibiotic drug mitomycin (E3) and the antimicrobial molecules flumequine (E4), were found to activate the bhc gene cluster in B. thailandensis E264, leading to the production of burkholdac A (C5), a histone deacetylase inhibitor (Fig. 2) (Seyedsayamdost, 2014).[图片]Fig. 2. A representative knowledge graph of stress-associated antibiotic elicitors with their induced BGCs and products. Piperacillin (E1), trimethoprim (E2), mitomycin (E3) and flumequine (E4) partially and collectively activate five BGCs in B. thailandensis E264 and lead to the production or overproduction of 14 secondary metabolites (C1–C14).A follow-up study found that trimethoprim triggered the production of over 100 compounds in B. thailandensis E264 that are not observed under standard growth conditions. Among these, 40 compounds were successfully identified using MS/MS and NMR, including a group of new molecules named acybolins (C6-C14), which originate from the bta gene cluster (Fig. 2) (Okada et al., 2016). Acybolins are bivalent metabolites composed of a bactobolin antibiotic scaffold covalently linked to an acyl chain derived from QS acyl-homoserine lactones (AHLs), a connection that requires a ligase not encoded within the bta cluster. Bactobolin itself is produced by the bta cluster and has been shown to be strictly regulated by AHL-mediated QS (Seyedsayamdost et al., 2010). By contrast, the combined presence of AHLs and trimethoprim elicits the formation of acybolins. This represents a distinct case in which antibiotic-induced stress reshapes the metabolic output of a known BGC, enabling the emergence of cryptic metabolites.Beyond B. thailandensis E264, subinhibitory exposure to antibiotics can also activate silent or lowly expressed BGCs in diverse microbial strains. Notably, several effective elicitors identified to date are antibiotics derived from Streptomyces. For instance, subinhibitory streptomycin (E5), an aminoglycoside antibiotic, induced the tam gene cluster in Microbispora sp. BCCAGE54, promoting a 33-fold enhancement in tetarimycin B (C15) production alongside several putative new tetarimycin analogs (Covington et al., 2018). Although tetarimycin B was detectable at low levels in unstimulated cultures, the corresponding gene cluster is therefore considered to be lowly expressed here. Similarly, actinomycin D (E6), a chemotherapeutic（化疗的） agent, was found through HiTES combined with cytotoxicity assays to activate the tun gene cluster in Streptomyces clavuligerus ATCC 27064, thereby enhancing the production of streptovirudin A2 (C16) and tunicamycins IIA (C17), IIB (C18), G (C19), and V (C20). Actinomycin D also activated the cla gene cluster in the same strain, leading to the production of clavorubin A (C21) (Han et al., 2022). In addition, subinhibitory lincomycin (E7) potentiated the act gene cluster in Streptomyces coelicolor A3(2) and Streptomyces lividans 1326, resulting in pronounced overproduction of the blue-pigmented antibiotic actinorhodin (C22), which is produced at low levels under standard conditions. Additionally, S. lividans 1326 produced several potential novel congeners of calcium-dependent antibiotics (CDAs) only under lincomycin exposure (Imai et al., 2015).2.1.2. Non-antibiotic clinical drugsIn addition to antibiotics, a broad range of non-antibiotic small molecules can also impose physiological stress on microbial cells. Notably, bioactive compounds that are already used clinically are particularly enriched in effective stress-associated elicitors, likely owing to their well-defined molecular targets and pronounced cellular effects. For example, pyronaridine (E8), an antimalarial drug, primarily targets nucleic acids by intercalating into DNA, thereby inducing DNA damage and triggering the SOS and/or oxidative stress. Such stress responses were shown to activate the tun gene cluster in S. clavuligerus ATCC 27064, as described previously. Similarly, etoposide (E9), a chemotherapeutic agent with broad-spectrum antiparasitic activity, has been shown to inhibit bacterial, and particularly actinobacterial, DNA gyrase, thereby inducing the SOS response. Through this stress pathway, etoposide acts as a potent elicitor in Streptomyces albus J1074, strongly activating the sur gene cluster and resulting in the discovery of seven surugamides and seven albuquinones (C23–C36) (Fig. 3 A) (Xu et al., 2017)[图片]Fig. 3. A representative knowledge graph of stress-associated non-antibiotic clinical drugs (A) and host-related molecules (B) with their induced BGCs and products. Etoposide (E9) and p-coumaric acid (E11) activate the sur and rbt gene clusters, leading to the production of surugamides (C23–C36) and roseobactin (C38), respectively.Spermidine (E10) was found to activate the gld gene cluster in Burkholderia gladioli ATCC 10248, resulting in enhanced production of the siderophore gladiobactin (C37). Spermidine is a key component of the intracellular polyamine pool, and its concentration balance with other polyamines, such as putrescine, is vital for microbial physiological state and metabolic regulation. At elevated concentrations, spermidine inhibited the growth of B. gladioli, indicative of a stress response that may underpin（支撑） enhanced gladiobactin biosynthesis. Notably, siderophore（铁载体） production to sequester（螯合） toxic metals is often a hallmark（标志） of oxidative stress response, and high spermidine level has been reported to trigger oxidative stress in E. coli (Yoshimura et al., 2020).2.1.3. Host-related moleculesFrom an ecological perspective, microorganisms are constantly exposed to a wide range of host-derived metabolites that shape their physiological states and metabolic behaviors. These compounds, produced by plants or animals in shared ecological niches, can act as chemical cues or stress signals that influence microbial secondary metabolism. Among such host-related molecules, p-coumaric acid (E11), a ubiquitous（普遍存在的） plant-derived phenolic compound, has emerged as a representative elicitor capable of activating cryptic metabolites. Algal（藻类的） p-coumaric acid was shown to activate the rbt gene cluster in its symbiotic bacterium Phaeobacter inhibens 2.10, leading to the discovery of an unreported siderophore, roseobactin (C38) (Fig. 3B). p-coumaric acid is a senescence（衰老）-associated molecule produced by the microalgal Emiliania huxleyi, and its accumulation marks a shift in microalgal-bacterial symbioses（共生细菌） from mutualistic（互利共生的） to parasitic phases. Meanwhile, p-coumaric acid is toxic to the bacteria at high concentrations but induces oxidative stress response by formation of reactive oxygen species (ROS) and secondary metabolism at sub-lethal doses（亚致死剂量）. These findings indicate that algal-derived phenylpropanoids, including p-coumaric acid as well as other compounds such as sinapic acid and ferulic acid, can induce cryptic secondary metabolism in symbiotic bacteria, underscoring（凸显） the potential of ecological interactions for elicitor discovery (Wang et al., 2022a).2.1.4. Chemically modified moleculesChemically modified elicitors represent a class of molecules that are generated through chemical modification of known bioactive compounds, often in the form of structural analogues. Rather than originating from natural or clinical contexts, these molecules are typically derived from pre-existing scaffolds and tailored to modulate microbial physiology. Importantly, this class of elicitors can be explored through rational design strategies, such as systematic structural optimization, to identify analogues with enhanced elicitor potency or selectivity.A representative example is the synthetic elicitor ARC2 (E12), an analogue of triclosan with a related mode of action. ARC2 activates the production of multiple specialized metabolites in Streptomyces. For instance, it activates the gcs gene cluster in Streptomyces coelicolor M145, resulting in a 3-fold increase in germicidins A–C (C39–C41) while decreasing the yields of prodiginines and the daptomycin-like CDA (Fig. 4). ARC2 also activates the des and dps gene clusters in Streptomyces pristinaespiralis ATCC 25486 and Streptomyces peucetius 27,952, enhancing the production of desferrioxamine B (C42) and E (C43) as well as doxorubicin (C44) and baumycin (C45) (Craney et al., 2012). Although ARC2 exhibits dose-dependent antibacterial toxicity at high concentrations, it primarily inhibits the enoyl reductase FabI, thereby blocking fatty acid biosynthesis. Inhibition of FabI induces a cellular stress response that depends on the global regulators AfsR and AfsS, linking ARC2-mediated stress to the activation of secondary metabolism (Calvelo et al., 2021a).[图片]Fig. 4. A representative knowledge graph of stress-associated chemically modified molecules with their induced BGCs and products. ARC2 (E12) activates three BGCs and lead to the production of seven secondary metabolites (C39–C45).Building on this scaffold, Cl-ARC (E13), a chemically modified derivative of ARC2, was shown to selectively enhance yields of secondary metabolites but has relatively little effect on the rest of the metabolomic profile. Because its molecular target, FabI, is highly conserved across Actinobacteria, Cl-ARC functions as a broadly effective elicitor in most Streptomyces species. Specifically, Cl-ARC activates the oxo, smd, and non gene clusters, increasing the yields of oxohygrolidin (C46) from S. ghanaensis ATCC 14672, 9-methylstreptimidone (C47) from S. hygroscopicus ATCC 53653, and nonactin (C48), monactin (C49), dinactin (C50), and trinactin (C51) from strain WAC0256 (Pimentel-Elardo et al., 2015).2.2. Signaling-associated elicitorsSignaling-associated elicitors represent a distinct class of elicitors that induce silent or weakly expressed BGCs by mimicking(模拟) or transmitting ecologically relevant signals rather than imposing physiological stress. These elicitors are often derived from host organisms or neighboring species and are sensed through dedicated regulatory systems, enabling microorganisms to coordinate secondary metabolism with specific ecological contexts.Zhang et al. utilized HiTES combined with matrix-assisted laser desorption/ionization mass spectrometry (MALDI-HiTES) and revealed that plant-derived amygdalin (E14), sarsasapogenin (E15), and homobutein (E16) in flask cultures of Streptomyces ghanaensis ATCC 14672 activated the cip gene cluster, resulting in a 4–10-fold increase in production of a novel non-ribosomal peptide, cinnapeptin (C52) (Zhang and Seyedsayamdost, 2020). In a related study, Li et al. demonstrated that two adjacent LuxR-type regulators in the gla cluster from Streptomyces globisporus C-1027, which is homologous to the cip cluster and also directs the biosynthesis of cinnamoyl moiety-containing compounds, function as cluster-situated positive regulators (Li et al., 2022). Notably, homologous LuxR-type regulators are also present in cip cluster. Based on these observations, we speculated that, as a plant-associated bacterium, S. ghanaensis may sense these plant-derived metabolites via these two LuxR-type regulators, thereby relieving repression of the silent BGC and initiating cinnapeptin biosynthesis as an ecologically grounded signaling response. However, this hypothesis requires further experimental validation.Vitamin B3 niacin (E17) was identified as a highly specific elicitor of the tqq gene cluster in Streptococcus suis ATCC 43765, whose activation led to the discovery of threoglucin A (C53), characterized by an aliphatic ether bond forming a substituted 1,3-oxazinane heterocycle within the peptide backbone (Covington and Seyedsayamdost, 2022). Previous evidence has indicated that Streptococcus species are capable of responding to extracellular niacin. In particular, the presence of NiaR has been identified in S. suis. NiaR is a niacin-responsive transcriptional repressor that regulates genes involved in nicotinamide adenine dinucleotide (NAD) metabolism (Rodionov et al., 2008). NiaR-mediated regulation links extracellular niacin availability to the coordinated expression of multiple genes involved in cellular metabolism (Afzal et al., 2017). From an ecological perspective, S. suis is a taxing agricultural pathogen and zoonotic agent, whereas niacin is an endogenous human metabolite. Therefore, niacin is likely an environmental signal that S. suis readily encounters and may use to modulate metabolic processes. Additionally, in the genomic context, the tqq gene cluster is located adjacent to an shp/rgg QS operon, suggesting that QS–related regulatory mechanisms may also be involved (Covington and Seyedsayamdost, 2022).Using Agar-HiTES, Lee et al. identified several steroids as the most effective elicitors in the plant fungal pathogen Sclerotinia sclerotiorum Ss-1. Among these, the glucocorticoid prednisone (E18) specifically activated the scl gene cluster, leading to the production of a previously uncharacterized metabolite, sclerocyclane (C54). Further studies revealed that the plant-derived steroids β-sitosterol and brassinolide, which are structurally similar to prednisone, also function as effective elicitors. Concurrently, the induced metabolite sclerocyclane exhibits antibiotic activity against Burkholderia plantarii, an ecological competitor of S. sclerotiorum. These findings suggest that secondary metabolites may be induced in regions of active plant growth where steroids, acting as host- or environment-associated signaling molecules, are released, thereby shaping ecological interactions (Lee and Seyedsayamdost, 2022).Similarly, Han et al. applied FRET-HiTES to Streptomyces clavuligerus ATCC 27064 and discovered that several steroids, including testosterone (E19), estrone (E20), cortisone (E21), danazol (E22), and 11α-hydroxyprogesterone (E23), activated the clv gene cluster, leading to the production of cryptic enediyne natural products, termed clavulynes A − J (C55–C64) (Fig. 5). This study suggests that steroids can play important roles in bacteria, functioning as signaling agents or carbon sources. Elicitation in S. clavuligerus may involve regulation of signal transduction pathways and/or an additional nutrient supply (Han et al., 2023).[图片]Fig. 5. A representative knowledge graph of signaling-associated elicitors with their induced BGCs and products. Five steroids (E19–E23) activate the clv gene cluster and lead to the production of clavulynes (C55–C64).2.3. OthersBeyond the elicitors discussed above, many additional bioactive compounds have been reported to activate cryptic BGCs, although their mechanisms of action remain largely unexplored. A growing number of studies have demonstrated that diverse bioactive small molecules, many of which are clinically used drugs, can function as effective elicitors for cryptic BGCs when interrogated using HiTES-based strategies.Using HiTES, Seyedsayamdost first showed that in Burkholderia thailandensis E264, the bhc gene cluster could be activated by the DNA topoisomerase I inhibitor β-lapachone (E24), and the nuclear factor-kappa B (NF-κB) inhibitor BAY 11–7082 (E25), leading to efficient synthesis of burkholdac A (C5) (Seyedsayamdost, 2014).Subsequently, Xu et al. employed a genetics-dependent reporter-based HiTES to demonstrate that the broad-spectrum antiparasitic ivermectin (E26) strongly activated the sur gene cluster in S. albus J1074, leading to the production of surugamides and albuquinones (C23–C36) (Xu et al., 2017). They later developed a genetics-free strategy, HiTES in conjunction with imaging mass spectrometry (IMS-HiTES), to identify additional elicitors. Using this approach, kenpaullone (E27), a GSK3β inhibitor used for the treatment of chronic immune (idiopathic) thrombocytopenia, was identified as the most effective elicitor of the can gene cluster in Streptomyces canus NRRL B3980, inducing two novel cryptic lassopeptides, canucins A (C65) and B (C66). In contrast, dihydroergocristine (E28), a γ-secretase inhibitor clinically used to delay progressive mental decline in conditions such as Alzheimer's disease, activated the ker gene cluster in Amycolatopsis keratiniphila NRRL B24117, resulting in the production of seven glycopeptides, including four keratinimicins (C67–C70) and three keratinicyclins (C71–C73) (Xu et al., 2019).Using bioactivity-coupled HiTES, Moon et al. identified the clinical diuretic furosemide (E29) and the cholesterol-lowering agent fenofibrate (E30) as the most effective elicitors for Saccharopolyspora cebuensis SPE 10–1. Both compounds activated the ceb gene cluster, leading to the production of a novel lanthipeptide, cebulantin (C74), which exhibited selective antibacterial activity against Gram-negative bacteria (Moon et al., 2019a). Using the same approach, Moon et al. identified the β-blocker atenolol (E31) as an effective global elicitor for Streptomyces hiroshimensis A18. Atenolol activated the hrs and tfl/tox gene clusters, thereby inducing the production of a novel naphthoquinone epoxide hiroshidine (C75), a new toxoflavin-type analog taylorflavin A, (C76), and several known metabolites, including taylorflavin B (C77). Notably, both taylorflavins exhibited selective growth-inhibitory activity against Gram-negative bacteria, particularly Escherichia coli and Acinetobacter baumannii (Moon et al., 2019b).To streamline（简化） the analysis of HiTES-derived metabolomics data, Covington et al. developed MetEx, a metabolomics explorer for LC–MS-based HiTES workflows. Application of MetEx to Burkholderia gladioli ATCC 10248 enabled identification of multiple bioactive elicitors. Specifically, the β-lactam antibiotic oxacillin (E32), the calcium-channel blocker verapamil (E33), the antihypertension（抗高血压） agent methyldopa (E34), the imidazoline decongestant naphazoline (E35), and the anti-inflammatory drug nabumetone (E36) were found to activate the gld gene cluster, leading to the production of gladiobactin (C37). The oxacillin (E32) and the chemotherapeutic agent azacitidine (E37) activated the gbn gene cluster, inducing gladiolin (C78) biosynthesis. Meanwhile, anticancer agents bleomycin (E38) and carboplatin (E39) activated the ico gene cluster, strongly enhancing icosalide (C79) production. Finally, the synthetic corticosteroid desoximetasone (E40) and the HIV protease inhibitor ritonavir (E41) activated the bgd gene cluster, serving as potent elicitors for burriogladin A (C80) (Covington and Seyedsayamdost, 2021).Lee et al. extended the HiTES to solid media by developing Agar-HiTES, enabling systematic elicitor screening in fungi and bacteria. Using this approach, they identified the antidepressant isocarboxazid (E42), and the antibiotics paromomycin (E43) and sulfanilamide (E44) as the most effective elicitors of the plant fungal pathogen（病原体） Sclerotinia sclerotiorum Ss-1, where these compounds activated the scl gene cluster and induced the production of a novel metabolite sclerocyclane (C54). In Rhizoctonia solani AG-3, the calcium-channel blockers nisoldipine (E45) and nifedipine (E46), together with the antifungal clotrimazole (E47), potently activated the sol gene cluster, with nisoldipine yielding 12-fold overproduction of two new siderophores, solanibactins A and B (C81–C82) (Lee and Seyedsayamdost, 2022). Applying Agar-HiTES to Burkholderia species, Lee et al. further showed that in B. plantarii ATCC 43733, plant-derived tropane alkaloids, including ipratropium bromide (E48) and atropine (E50), as well as the antimigraine drug zolmitriptan (E49), robustly induced the bet cluster, leading to 12–15-fold enhanced production of burkethyl A (C83) and B (C84) at 90 μM ipratropium bromide. In B. gladioli ATCC 10248, chemically diverse elicitors, including the synthetic steroid fluocinolone (E51), the antibiotics imipenem (E52) and rifapentine (E53), and the HIV protease inhibitor saquinavir (E54), markedly altered secondary metabolism and triggered the production of modified gladiobactins [hydroxylated B/C (C85-C86) and glycosylated D (C87)] (Lee et al., 2025).3. Mechanisms of elicitor-mediated BGC activationA comprehensive understanding of the regulatory mechanisms underlying BGCs is essential for the targeted activation of silent clusters using chemical elicitors. Because BGC expression is typically orchestrated（协同调控） by complex, multilayered regulatory networks, network-based models provide a systematic framework for deciphering（解析） these processes. Such models can illuminate the interactions among diverse regulators and thereby offer theoretical guidance for designing effective elicitor-driven strategies to activate silent BGCs. In this section, we outline the principles of regulatory networks and discuss their utility in elucidating elicitor-mediated BGC activation.3.1. Regulatory network models of BGC expressionBGC expression is governed by intricate regulatory networks, which encompass interactions among DNA, RNA, proteins, and small molecules. The collective set of these interactions is often referred to as the interactome (Fig. 6A) (Luck et al., 2016; Sanchez et al., 1999). Among them, the typical gene regulatory network (GRN) models, transcription factor–target gene (TF–TG) interactions, which are often visualized as directed graphs where nodes represent genes or proteins and edges represent regulatory relationships (Fig. 6B) (Babu et al., 2004; Seshasayee et al., 2006). However, TF-based models alone cannot fully capture the complexity of BGC regulation, since gene expression is also influenced by epigenetic modifications (EMs)（表观遗传修饰）, post-transcriptional regulation (PTR) [such as RNA-RNA interactions (RRIs)] (Silverman and Melamed, 2025), post-translational modifications (PTMs) [such as protein-protein interactions (PPIs)] (Wang et al., 2022b), and other regulatory pathways. Extended GRNs therefore integrate additional regulatory components, including independently acting molecules such as cofactors, sigma factors, small RNAs (sRNAs) (Sesto et al., 2013; Waters and Storz, 2009), as well as the cis-regulatory elements (CREs) and riboswitches embedded within TGs or mRNA. In addition, extended GRNs encompass regulatory components associated with modification-based mechanisms, which commonly involve EMs, PTR, and PTMs (Fig. 6B). Together, these regulatory components provide a more complete view of BGC control (Augustijn et al., 2024; Badia-i-Mompel et al., 2023).[图片]Fig. 6. Conceptual frameworks for regulatory and induction mechanisms governing BGC expression. (A) Key cellular components and their interactions. (B) A typical TF-centered GRN and an extended GRN incorporating multi-level regulatory mechanisms. (C) Major classes of elicitor-induced regulatory mechanisms underlying BGC activation.GRNs typically exhibit a modular structure, consisting of recurring regulatory motifs (e.g., feed-forward loops [FFLs], single-input modules [SIMs], multi-input modules [MIMs]) as well as higher-order functional modules (e.g., stress response, metabolism). Global regulators (GRs) govern broad cellular functions, whereas cluster-situated regulators (CSRs) fine-tune（精细调控） the expression of individual BGCs. Additional layers of regulation, including various post-transcriptional and post-translational mechanisms, further contribute to the dynamic and multilayered nature of GRNs (Fig. 6B). Recent advances have facilitated the integration of GRN concepts into BGC studies. For instance, the antiSMASH7 now includes a transcription factor binding site (TFBS) prediction module, enabling the identification of key regulatory elements (Blin et al., 2023). Likewise, a recent review by Wezel and Medema emphasized the potential of GRN-based approaches to accelerate natural product discovery in Actinobacteria (Augustijn et al., 2024). Collectively, applying GRN models is expected to provide a rational framework for systematically analyzing BGC regulation and for guiding the selection of effective elicitors.3.2. Deciphering elicitor-mediated BGC activation through regulatory networksAs outlined in Section 3.1, activation of BGC expression can be systematically interpreted（阐释） within the framework of regulatory networks. Based on existing studies, we summarize four major classes of elicitor-associated regulatory mechanisms, including oxidative stress–driven regulation, SOS response–mediated regulation, ribosomal perturbation–driven regulation, and metabolic stress-driven regulation (Fig. 6C). Although the molecular targets of these elicitors vary widely, their effects are ultimately perceived and processed through specific regulatory pathways, which converge on gene regulatory networks to activate the corresponding BGCs. It should be noted that the depth of mechanistic understanding differs substantially among these categories, and additional regulatory mechanisms may exist beyond those currently recognized. Further investigation will be required to achieve a more comprehensive understanding of elicitor-mediated BGC activation.3.2.1. Oxidative stress-driven regulationIn the induction cases discussed in the previous section, a significant activation is associated with bacterial stress states, such as oxidative stress. Here, we first clarify the mechanisms related to stress, and then illustrate the mechanism by which BGC is induced using specific examples. Stress states induced by exogenous small molecules are mainly manifested as oxidative stress or DNA damage–triggered SOS responses, and these two processes are partially interconnected. Under normal conditions, bacterial cells maintain redox（氧化还原） homeostasis（稳态）, in which metabolically generated ROS are balanced by antioxidant systems to sustain a stable intracellular redox state. However, interference by exogenous molecules can disrupt this balance, driving cells into a stress state that may result in outcomes such as apoptosis-like cell death（类细胞凋亡）, autophagy（细胞自噬）, lipid peroxidation（脂质过氧化）, and DNA damage (Hong et al., 2024). Meanwhile, multiple peroxide-sensor TFs, including OxyR, PerR, and SoxRS, are directly activated by ROS, thereby initiating complex downstream transcriptional programs that rewire cellular metabolic networks (Imlay, 2015). Accumulating evidence indicates that these stress conditions activate stress-responsive transcription factors and lead to extensive changes in gene expression, highlighting the systems-level impact of stress-responsive regulatory networks (Guo et al., 2024; Seo et al., 2015).It has been proposed that many bactericidal antibiotics promote bacterial cell death not only through their specific drug–target interactions but also by inducing intracellular ROS, a hypothesis originally put forward by Kohanski et al. and widely discussed thereafter (Kohanski et al., 2007). Although this antibiotic–ROS–cell death model has remained controversial, accumulating evidence suggests that ROS can contribute to antibiotic-mediated cytotoxicity in a strain- and context-dependent manner. Based on this logic, low or sublethal doses of antibiotics may induce detectable ROS levels and engage stress-responsive pathways, such as redox-sensing TFs or the SOS response, thereby reprogramming global transcriptional and metabolic networks and potentially activating silent or weakly expressed BGCs. This notion is consistent with the pattern reported by the Seyedsayamdost group (Davies et al., 2006; Okada and Seyedsayamdost, 2017), and aligns with the concept of hormesis (Calabrese and Baldwin, 2003). For instance, Seyedsayamdost et al. demonstrated that the antifolate antibiotic trimethoprim and the β-lactam piperacillin upregulated five target genes in the malleicyprol (mal) gene cluster of Burkholderia thailandensis E264 (Seyedsayamdost, 2014). Li et al. further showed that piperacillin activates the mal gene cluster by inducing oxidative stress involving redox-sensing GRs. At low doses, piperacillin generates ROS, which trigger redox-sensitive TFs OxyR and SoxR. These TFs subsequently activate MalR, the CSR of the mal cluster (Fig. 7A) (Li et al., 2021). Meanwhile, as a LuxR-type transcriptional regulator, MalR activates the mal cluster in an AHL-independent manner by binding to a lux box-like element in the mal promoter, differently from the canonical AHL-based QS mechanism (Truong et al., 2015). Similarly, trimethoprim, a global regulatory elicitor, activates multiple metabolic pathways in B. thailandensis E264 through global regulatory mechanisms. It upregulates malR and thereby enhances mal expression (Okada et al., 2016). In parallel, it significantly boosts thailandamide production (8- to 36-fold), accompanied by strong induction of biosynthetic genes thaH (polyketide synthase, +26-fold) and thaR (oxidoreductase, +17-fold) (Seyedsayamdost, 2014). While the precise regulatory intermediates remain unknown, these findings illustrate how elicitors exploit global stress-response modules within GRNs to activate BGCs.[图片]Fig. 7. Representative regulatory mechanisms of chemical elicitors in activating silent BGCs, including (A) oxidative stress–driven regulation, (B) SOS response–driven regulation, (C) ribosomal perturbation–driven regulation, and (D) metabolic stress–driven regulation.Natural metabolites can function as elicitors by activating stress-related GRN modules. For example, the algae-derived compound p-coumaric acid induces the production of the siderophore roseobactin in Phaeobacter inhibens, increasing the expression of roseobactin biosynthetic genes by 3.3- to 6.7-fold. Concurrently, it upregulates multiple antioxidant and stress-response enzymes by 4.3- to 13.7-fold and enhances the expression of genes associated with cellular repair systems by 3.3- to 14.9-fold. These effects support a model in which p-coumaric acid triggers ROS formation, activating oxidative stress-response modules within the GRN. The induced stress response facilitates repair of damaged biopolymers and promotes secondary metabolite biosynthesis, while cellular energy may be conserved by downregulating non-essential processes, such as motility and flagellar synthesis, thereby redirecting resources toward roseobactin production (Fig. 7A) (Wang et al., 2022a).3.2.2. SOS response-driven regulationIn addition to ROS, some exogenous small molecules act as direct DNA-damaging agents, a phenomenon commonly encountered in natural environments, and can induce DNA lesions, thereby triggering the SOS response (Maslowska et al., 2019). The SOS response is regarded as one of the most fundamental and evolutionarily conserved global transcriptional regulatory systems in prokaryotes, specifically dedicated to coping with severe DNA damage. At the core of this system lies the coordinated action of two key regulators, LexA and RecA, which function as a transcriptional repressor and a damage sensor, respectively (Lima-Noronha et al., 2022). Under normal growth conditions, LexA binds as a dimer to specific DNA sequences known as SOS boxes located in the promoter regions of genes within the SOS regulon. This binding effectively prevents the recruitment of RNA polymerase, thereby maintaining SOS genes at low basal expression levels. However, when cells are exposed to stressors that lead to the accumulation of single-stranded DNA (ssDNA), RecA is rapidly recruited to ssDNA and forms an activated ATP-dependent nucleoprotein filament, referred to as RecA*. Beyond its role in recombinational repair, RecA* acts as a co-protease for LexA, inducing LexA autoproteolysis. The cleavage of LexA results in loss of its promoter-binding affinity, thereby derepressing the transcription of numerous genes involved in DNA damage tolerance and repair, as well as global transcription reprogramming (Podlesek and Žgur Bertok, 2020).In such contexts, elicitors that induce DNA damage and activate the SOS response have the potential to trigger the expression of silent BGCs. A representative example is the activation of the sur gene cluster by the clinical DNA-damaging agent etoposide. Etoposide has been shown to inhibit actinobacterial DNA gyrase. At low concentrations, etoposide strongly induces the expression of recA and lexA in the SOS response pathway (Fig. 7B). Subsequent RT-PCR and gene knockout analyses revealed that etoposide downregulates SurR, a pathway-specific transcriptional repressor, thereby derepressing the cluster and enhancing the production of cryptic metabolites. Collectively, this case illustrates that elicitor-mediated activation of the SOS response can establish a global regulatory context in which relief of pathway-specific transcriptional repression enables BGC activation, linking DNA damage responses with secondary metabolism (Xu et al., 2017). In addition, mitomycin and bleomycin, as mentioned in the preceding examples, have also been reported to induce the SOS response in E. coli (Wei et al., 2001; Xu et al., 2012). However, their mechanisms of action as elicitors have not yet been experimentally examined.3.2.3. Ribosomal perturbations–driven regulationSome elicitors, particularly translation-inhibiting antibiotics, can rewire cellular metabolic networks by perturbing ribosomal homeostasis. Ochi and co-workers discovered that mutations in ribosomal proteins can profoundly reshape cellular metabolic phenotypes in Streptomyces lividans (Shima et al., 1996), an observation that later gave rise to the concept of ribosome engineering. In this strategy, mutations in ribosomal components or RNA polymerase (RNAP) are introduced to reprogram transcriptional and translational states (Zhu et al., 2019). These findings established the ribosome and RNAP as regulatory targets for inducing cryptic metabolites. Perturbation of ribosome or RNAP often disrupts translational dynamics, leading to ribosomal pausing or stalling. Such disturbances are rapidly sensed by stress-responsive regulatory networks, such as WhiB7/WlbC TFs, which coordinate ribosome rescue, translational reprogramming, and broader metabolic adaptation, ultimately leading to the activation of silent BGCs (Hurst-Hess et al., 2023; Lee et al., 2020).A representative example of this ribosome-centered elicitation mechanism is provided by lincomycin. The Hosaka group identified lincomycin at subinhibitory concentrations as an effective elicitor for actinorhodin biosynthesis in S. coelicolor A3(2) and further elucidated its regulatory mechanisms (Imai et al., 2015). Lincomycin, a ribosome-targeting antibiotic, temporarily inhibits ribosome function by binding to the peptidyl transferase loop of 23S rRNA in 70S ribosomes. This perturbation elicits a rapid response mediated by the WblC regulatory network (Mukai et al., 2023). As a master regulator, the TF WblC coordinates lincomycin resistance and ribosome reconstruction by upregulating genes that encode ABCF proteins (facilitating lincomycin efflux), HflX proteins (ribosome-splitting factors that resolve stalled ribosomes), and 23S rRNA methyltransferase (structurally protecting ribosomes against inhibition) (Ishizuka et al., 2018). In the context of WblC regulation, reduced 20S proteasome activity stabilizes ribosome-associated proteins and enhances translational efficiency. This multilayered regulation promotes the downstream expression of the CSR actII-ORF4, ultimately increasing actinorhodin production (Fig. 7C). Moreover, streptomycin, described in the above cases as an elicitor, can be classified as a ribosome-targeting antibiotic, and its elicitor mechanism remains unreported (Lin et al., 2018).3.2.4. Metabolic stress-driven regulationPrimary metabolism comprises the core biochemical pathways that sustain cellular growth, energy production, and structural integrity. Because these pathways operate at high flux and are deeply interconnected, their activities are highly sensitive to perturbation. Certain small molecules can directly target and inhibit key enzymes within primary metabolic pathways, disrupting metabolic flux balance and homeostasis, giving rise to the metabolic stress. Cells respond to this stress by engaging global TFs, leading to the reprogram of gene expression and metabolic flux.The ARC2 compound series provides an example of a metabolic stress-driven regulation, in which elicitors redirect metabolic flux by inhibiting core primary metabolic pathways. Due to its structural similarity to triclosan, a known FabI inhibitor, validation experiments and gene complementation studies confirmed that ARC2 functions as an effective inhibitor of the FabI enoyl reductase in fatty acid biosynthesis (Craney et al., 2012). Subsequent studies demonstrated that the ARC2-induced metabolic response depends on the global TFs AfsR and AfsS (Calvelo et al., 2021b). Notably, there is currently no evidence that AfsR or AfsS act as direct sensors of metabolic stress; instead, they are generally regarded as global regulators of antibiotic production in Streptomyces (Yan and Xia, 2024). As a regulatory outcome, inhibition of fatty acid biosynthesis diverts shared precursors such as acetyl-CoA and malonyl-CoA toward polyketide and siderophore production, including actinorhodin, germicidin, doxorubicin, baumycin, and the acetyl-CoA-dependent desferrioxamine, while reducing the yields of fatty acid-containing metabolites such as prodiginines and CDAs (Fig. 7D). Additionally, as mentioned above, elicitor trimethoprim inhibits folate metabolism, thereby blocking one‑carbon metabolism and nucleotide biosynthesis, and can therefore be regarded as a form of metabolic stress (Quinlivan et al., 2000).4 生物合成基因簇的靶向激活：诱导剂筛选指南结合上文整合的知识图谱与作用机制研究，本文提出一套筛选策略，用于靶向激活沉默或低表达的生物合成基因簇（BGCs）。该系列策略并非绝对化的既定规则，而是依托生物信息学与化学信息学理论、以知识为导向的经验性方法。由于不同微生物的调控通路架构差异极大，并非任意菌株都能实现生物合成基因簇的精准靶向激活。因此，下文将依据策略的可靠性由高至低依次阐述：从基于保守共有调控机制的经典手段，到依托同源序列推演的探索性方案。本研究构建的策略框架旨在为诱导剂筛选提供实践参考，助力发掘更多可通过诱导剂激活的生物合成基因簇调控案例。4.1 基于调控机制的诱导剂筛选依托作用机制或表型效应，可高效发掘功能性诱导剂。例如，李等人研究发现，泰国伯克霍尔德菌 E264 菌株中，活性氧（ROS）水平与mal 生物合成基因簇的表达量存在显著相关性。基于该结论，团队采用多种经典活性氧诱导剂（白花丹醌、百草枯、甲萘醌），成功激活该基因簇（Li 等，2021）。结合前文梳理的分子机制，诱导剂与微生物的互作规律可为诱导剂筛选提供重要理论依据：不同类型诱导剂通过特异性通路重塑微生物调控网络，引发代谢重编程，进而调控次级代谢。具体而言，能够诱导胞内活性氧显著累积的小分子化合物，是筛选诱导剂的合理候选。以 β- 内酰胺类抗生素为例，氨苄西林、美罗培南、头孢曲松等均可优先作为备选诱导剂。传统研究表明，该类药物通过结合转肽酶、阻碍细菌细胞壁肽聚糖交联发挥抑菌作用；后续研究进一步证实，细胞壁结构紊乱会加剧细胞呼吸作用，最终导致胞内活性氧水平异常升高（Dwyer 等，2014；Van Acker 等，2016）。除小分子物质外，非小分子处理方式同样可诱导活性氧生成（Memar 等，2018），本文不作赘述。同理，喹诺酮类等DNA 损伤型抗生素（萘啶酸、环丙沙星、诺氟沙星、莫西沙星），通过抑制 DNA 旋转酶与拓扑异构酶 Ⅳ 阻断 DNA 复制，造成基因组损伤，进而激活细菌 SOS 应激响应通路（Bush 等，2020；Dörr 等，2010；Newmark 等，2005）。因此，喹诺酮类药物可作为靶向 SOS 通路的候选诱导剂。以核糖体干扰为作用靶点、靶向翻译过程的抗生素，是一类极具应用价值的诱导剂。该类物质直接破坏核糖体功能，已被广泛证实可有效激活放线菌（链霉菌、分枝杆菌等）沉默基因簇，代表药物包括氨基糖苷类（链霉素、卡那霉素、庆大霉素）、四环素、红霉素、氯霉素等（Krause 等，2016；Lin 等，2018；Morris 等，2005）。除翻译抑制类抗生素外，其他可触发核糖体修复通路的小分子也具备诱导潜力：如烷基化试剂可损伤信使 RNA（Thomas 等，2020）；嘌呤霉素通过阻断多肽链延伸，诱发转录提前终止（Aviner，2020），二者均有望成为新型诱导剂（Keiler，2015）。从代谢应激角度出发，扰乱核心初级代谢通路的小分子化合物，同样可纳入诱导剂筛选范围。亚油酸会破坏微生物中心代谢与氧化还原稳态，引发强烈代谢胁迫（Senizza 等，2020）；浅蓝菌素与硫乳霉素可抑制脂肪酸合成，扰乱脂质代谢流向，限制微生物生长（Price 等，2001）；碘乙酸与碘乙酰胺通过抑制 3 - 磷酸甘油醛脱氢酶，阻断糖酵解通路、损伤能量代谢，诱发糖代谢应激（Schmidt，2009）；膦胺霉素靶向 MEP（甲基赤藓醇磷酸）途径，抑制 1 - 脱氧 - D - 木酮糖 - 5 - 磷酸还原异构酶，阻碍类异戊二烯前体合成并诱导代谢压力（Zhang 等，2011）。以上案例表明，依托分子机制系统性挖掘现有文献，能够高效筛选出尚未被开发的新型诱导剂。4.2 基于结构与活性关联的诱导剂筛选化合物结构相似度越高，作用机制通常越保守，这一规律已得到广泛验证。例如，结构高度相似的第三代、第四代头孢菌素，均能有效激活 mal 生物合成基因簇（Li 等，2021）。受此启发，可基于已报道诱导剂的化学结构，通过塔尼莫托系数量化分子结构相似度，从化合物库中快速筛选潜在同源诱导剂，这类结构类似物往往具备激活同源生物合成基因簇的潜力。本研究以知识图谱中收录的 54 种已验证诱导剂为基础，设定塔尼莫托系数阈值为 0.6，在包含 11000 余种临床药物的 DrugBank 数据库中，筛选得到 38 种已知诱导剂的 118 个结构类似物（图 8A、附表 S2）。这些类似物涵盖多类分子家族：脂质及类脂质分子 37 种、苯丙烷与聚酮类化合物 14 种、苯环类物质 12 种、有机酸及其衍生物 10 种，其余类别化合物零散分布。深入解析诱导剂构效关系后，可进一步优化化合物聚类标准，大幅提升新型诱导剂的筛选精准度。例如，甲氧苄啶（E2）可激活 bhc、mal、bta、hmq、tha 等多个基因簇，其结构类似物（溴莫普林、四氧普林、2,4 - 二氨基 - 5-(3,4,5 - 三甲氧基苄基) 嘧啶鎓）大概率具备相同激活效果，但其对次级代谢产物合成的调控作用仍需实验验证。同理，哌拉西林（E1）的类似物（美洛西林、阿洛西林）可作为激活 mal 基因簇的备选诱导剂；丝裂霉素（E3）的 N - 甲基衍生物 —— 普卡霉素，是靶向 bhc 基因簇的高效潜在化学诱导剂（图 8B）。[图片]图 8 图例说明(A) 38 种已知诱导剂与从 DrugBank 数据库中筛选得到的 118 个结构类似物的化学相似性网络。该网络中，化合物之间的连线代表二者塔尼莫托系数＞0.6。可被诱导剂激活的 27 个生物合成基因簇（以三角形表示）也通过虚线箭头纳入网络分析；矩形色块的颜色代表药物的结构分类。(B) 从数据库中筛选获得、对应诱导剂 E1–E3 的 6 种结构类似物化学结构式，核心骨架颜色与其所属结构分类保持一致。4.3 基于生物合成基因簇同源性的诱导剂筛选微生物基因组中仍存在大量未被解析的生物合成基因簇。本研究中已报道的 32 个可激活生物合成基因簇，序列差异度极高；在 BIG-SCAPE 软件默认阈值 0.3 的分析条件下，这 32 个基因簇均被判定为独立单系基因簇，无近缘聚类关系。与此同时，同源生物合成基因簇通常可被同一种化学诱导剂激活。利用已验证的激活型基因簇作为探针，可在微生物基因簇数据库中批量挖掘大量同源基因簇，且这类同源基因簇大概率也能被同种诱导剂靶向激活。本研究利用 cblaster 软件（v1.3.19），在包含 130 余万个微生物生物合成基因簇的自建数据库中，比对筛选得到1274 个与上述激活型基因簇高度同源的基因簇（图 9A、附表 S3）。在 32 个激活型基因簇中，23 个存在同源基因簇，剩余 9 个未检索到同源序列。这些同源基因簇来源于 8 个菌属，主要集中于伯克霍尔德菌属、链霉菌属与链球菌属。借助已知激活基因簇对应的诱导剂，可靶向挖掘同源基因簇编码的新型次级代谢产物类似物，大幅提升天然产物的挖掘效率。此外，本研究以sur基因簇为模型开展初步验证：证实伊维菌素可显著诱导白浅灰链霉菌 LHW61002中同源sur基因簇的高表达（图 9B）。采用代谢组学软件 MZmine3 对二级液质联用（LC-MS²）数据进行特征峰提取与预处理后，将数据上传至全球天然产物分子网络平台（GNPS），开展基于特征峰的分子网络分析（FBMN）。结果显示：同一分子家族下的 2 个特征峰被注释为已知化合物骏河酰胺 A、骏河酰胺 D，另有 1 个特征峰暂无匹配注释，推测为未被报道的新型结构类似物（图 9C）。结合已有文献与质谱镜像比对结果，该未知化合物初步推断为酰基化骏河酰胺 I。该案例充分证实：同源生物合成基因簇能够响应同种诱导剂并被特异性激活，为靶向发掘新型微生物天然产物提供了高效可行的研究策略。[图片]图 9 图例说明(A) 序列相似性网络由自建细菌基因簇数据库中、与已激活生物合成基因簇（BGCs）高度同源的基因簇构建而成。每个节点代表一个生物合成基因簇；节点填充颜色对应基因簇的来源菌属；节点形状表示 BIG‑SCAPE 软件划分的基因簇类型。(B) 从实验室保藏菌株白浅灰链霉菌 Streptomyces albidoflavus LHW61002中鉴定得到的sur同源基因簇对比分析。(C) 伊维菌素可诱导白浅灰链霉菌 LHW61002 合成一系列骏河酰胺类化合物，包含一种潜在的新型结构类似物。4.4 基于调控基因的诱导剂筛选来源迥异、无序列同源性的细菌生物合成基因簇，也可能共享高度相似的调控机制。例如，一类受 LuxR 型群体感应（QS） 调控的生物合成基因簇，可合成多种活性代谢产物，包括临床常用抗生素（碳青霉烯、莫匹罗星）及巴托波林、伊那西菌素等抗菌物质。LuxR 型群体感应的调控机制核心为 LuxR 蛋白激活：该蛋白作为信号分子受体（配体多为酰基高丝氨酸内酯 AHL 或其他未知信号分子），激活后进而调控下游关联生物合成基因簇的表达。已有研究证实，外源添加费氏弧菌 AHL 信号分子（3 - 氧代己酰 - L - 高丝氨酸内酯），可使群体感应调控的生物发光表型提前启动、且发光强度显著提升（Septer & Stabb, 2012）。该结果表明：筛选靶向 LuxR 型调控蛋白的特异性诱导剂，有望批量激活所有受 LuxR 型群体感应调控的同类生物合成基因簇。格林伯格团队系统调研了变形菌门中luxR 同源基因关联的生物合成基因簇，发现 luxR 同源基因与次级代谢基因簇的调控关联，在演化历程中发生过多次独立起源（Brotherton 等，2018）。这类调控关联广泛存在于功能各异的非同源基因簇中，说明不同来源、序列无同源性的 BGC，其调控模式发生了趋同进化。该团队借助生物信息学手段，批量鉴定出大量受 LuxR 型群体感应管控的生物合成基因簇，可为后续利用诱导剂实现靶向激活提供丰富的研究靶点。[图片]图 10自建细菌基因簇数据库中、含有已知调控基因同源序列的生物合成基因簇（BGC）序列相似性网络。每个节点代表一个生物合成基因簇；节点填充颜色表示基因簇的来源菌属；节点边框颜色代表该基因簇所含调控基因的类型；节点形状为 BIG‑SCAPE 软件划分的基因簇类别。机制研究表明：化学诱导剂可通过改变生物合成通路中调控基因的表达水平或蛋白活性，大幅促进微生物次级代谢产物合成。这提示：此类诱导剂同样有潜力激活携带同源调控基因的其他生物合成基因簇。据此，依托保守调控基因，有望实现大范围沉默基因簇的靶向激活。本研究选取已证实可被特定化学诱导剂调控的关键基因（malR、thaR、actII‑ORF4、surR、recA、lexA），在自建细菌基因簇数据库中开展本地 BLAST 比对。结果共筛选得到608 个含有上述基因高度同源编码序列的生物合成基因簇（附表 S4）。利用 BiG‑SCAPE 1.1.9 版本、默认阈值 0.3 进行聚类分析：其中 334 个基因簇归为62 个基因簇家族（GCFs）（图 10），剩余 274 个为独立单特异基因簇。值得注意的是，其中 5 个基因簇家族共包含 144 个生物合成基因簇，均分别对应前文已证实可被小分子激活的核心基因簇（des、tha、sur、act、mal）。说明这些基因簇不仅与上述 5 类经典基因簇序列高度相似，还共享同源调控基因，被对应诱导剂激活的概率显著更高。该类基因簇来源广泛，覆盖 182 个不同菌属，其中链霉菌属占比最高。recA与lexA基因广泛分布于各类基因簇与微生物类群中，在 DNA 修复及 SOS 应激通路调控中发挥核心作用。后续可探究对应化学诱导剂对上述同源基因菌株次级代谢谱的影响，为基于保守调控机制、靶向挖掘新型天然产物提供理论依据。5 结论与展望综上，本文系统综述了化学诱导剂激活沉默生物合成基因簇的研究现状，及其在新型天然产物靶向挖掘中的应用。基因组挖掘研究已证实微生物体内存在大量未解析的沉默基因簇，合理利用化学诱导剂，是充分释放微生物次级代谢潜能的关键手段。本文整合构建了诱导剂 — 生物合成基因簇 — 代谢产物关联知识图谱，并依据作用模式将诱导剂分为三大类：应激相关诱导剂、信号通路相关诱导剂及其他化合物。目前，大量诱导剂激活基因簇的效应及其深层调控机制仍尚不明确，也是未来重点研究方向。本文系统总结了诱导剂激活沉默基因簇的分子调控机制。从调控网络层面来看：尽管诱导剂化学结构差异巨大，但最终会汇聚至少数保守的上游调控通路与应激响应途径，主要分为四大类：氧化应激调控、DNA 损伤介导的 SOS 应答、核糖体扰动调控、代谢应激调控。不同诱导剂的初始作用靶点各不相同，但各类细胞扰动信号会经由基因调控网络整合，最终解除基因簇抑制、实现沉默 BGC 激活。现阶段，诱导剂介导基因簇激活的分子机制解析仍不完善。随着多组学技术快速发展，结合时间序列转录组、蛋白质组、代谢组与表观基因组的联合分析，有望构建完整的基因调控网络（GRN）及拓展调控通路。同时，基于机器学习、深度学习的网络推演算法，为鉴定基因簇激活的通用上游调控因子与功能模块提供了强大工具。此外，单细胞转录组可弥补常规混合测序的不足，解析菌群中不同个体对诱导剂的响应异质性。上述技术进步将助力深入解析诱导剂 — 调控网络互作机制，为基因簇激活与天然产物挖掘提供理性设计策略。同时，依托知识图谱与机制解析结果，本文提出四类沉默基因簇靶向激活核心策略：（1）基于调控机制筛选诱导剂：依据保守应激通路与信号调控通路选择适配诱导剂；（2）基于结构‑活性关系筛选诱导剂：挖掘已知诱导剂的结构类似物，用于激活同类基因簇；（3）基于 BGC 同源性筛选诱导剂：利用成熟诱导剂，激活序列与功能特征相似的同源基因簇；（4）基于调控基因筛选诱导剂：靶向筛选与已激活基因簇共享同源调控基因的基因簇。实际研究中，对于已有基因组数据的菌株，可优先参考近缘物种研究成果指导诱导剂筛选；若同属或近缘菌株已有小分子扰动报道，可优先选用该类化合物及其结构类似物。结合基因组注释与 BGC 预测结果，评估菌株基因组中是否含有知识图谱收录的同源基因簇与调控基因，进一步缩小诱导剂筛选范围。需客观说明的是，该套策略存在固有局限性：仅适用于与现有知识图谱存在关联的菌株与诱导剂，无法为任意微生物菌株提供普适性筛选方案。但相较于盲目随机筛选，该知识驱动型筛选框架可大幅提升阳性诱导剂的筛选成功率，尽管无法保证百分百激活效果。整体而言，该体系有助于丰富诱导剂激活基因簇的研究案例，同时为机制研究与应用研究提供支撑。利用诱导剂实现特定基因簇的靶向激活仍面临诸多挑战，需综合考虑多项关键影响因素：诱导剂浓度是核心因素，尤其抗生素类诱导剂，通常需在亚抑菌浓度下发挥代谢激活作用；培养体系同样关键，培养基组分可显著调控沉默基因簇表达，碳源、氮源、磷源的含量变化会直接改变次级代谢产物的种类与产量。例如：高浓度葡萄糖、磷酸盐、铵盐通常抑制次级代谢，而低水平营养胁迫可显著促进次级代谢合成。大量研究证实，同一诱导剂在不同培养基中会产生完全不同的代谢表型，充分说明次级代谢高度依赖培养条件；固体 / 液体培养方式差异，也会导致相同诱导剂处理下代谢谱截然不同。温度、pH 等环境条件同样会影响基因簇表达，需纳入实验设计。因此，在筛选诱导剂之外，还应优化培养基组合与培养环境参数。可结合液质联用（LC‑MS）、生物活性筛选、报告基因检测等技术，表征不同条件下的新型代谢产物与激活基因簇。对于微生物而言，诱导剂可视为一类特殊的 “调控药物”：通过唤醒休眠沉默基因簇，解除次级代谢沉默状态，挖掘微生物隐秘的代谢潜力。如同药物理性设计推动新药研发，基于已知诱导剂的结构优化与机制改造，也有望开发出活性更强、靶向性更高的新型诱导剂。随着诱导剂激活案例不断积累、分子机制逐步阐明，人们对诱导剂与微生物调控网络的互作认知将更加完善，进而实现诱导剂的理性设计与机制深度解析，推动特异性代谢产物高效合成，赋能医药研发、生物技术及相关领域发展。本文配套补充数据见文末。

糖皮质激素是什么？ 糖皮质激素（glucocorticoids, GCs）是一类由肾上腺皮质以胆固醇为前体合成的甾体激素，主要在皮质束状带生成；在人类以内源性皮质醇（cortisol）为主（皮质酮通常更低；而大鼠/小鼠以皮质酮为主）。容易混淆的点：“皮质激素（corticosteroids）”是更大的总称，通常包括1. 糖皮质激素（GCs）：偏代谢调控 + 抗炎/免疫调节 + 应激适应2. 盐皮质激素（mineralocorticoids, 如醛固酮）：偏电解质/体液容量与血压（主要受 RAAS 与血钾调控，ACTH 可提供短期刺激，但长期主控仍是 RAAS/血钾。）糖皮质激素有哪些“典型特征”？ 把它想成“能穿膜的化学信号”，糖皮质激素（GCs, glucocorticoids）的典型特征可用四句话概括：1.脂溶性：游离态 GCs 可跨膜进入细胞，结合糖皮质激素受体（GR, glucocorticoid receptor），经典效应多通过调控基因表达实现（也可有部分快速非基因组效应）。2.作用广但“因地制宜”：多数组织具备响应能力，参与代谢稳态、应激适应与炎症/免疫调节，但效应强弱高度依赖组织/细胞背景与共调控网络。3.节律 + 负反馈：下丘脑–垂体–肾上腺轴（HPA axis, hypothalamic–pituitary–adrenal axis）驱动昼夜节律并在应激时升高，同时通过负反馈抑制上游信号以防失控。皮质醇也有超日节律/脉冲（ultradian pulses）。4.临床价值高但“用量决定代价”：抗炎/免疫抑制效应强，因此治疗价值大；但全身用药的不良反应与剂量/疗程（以及给药途径）强相关，常见风险包括代谢异常、骨质疏松/骨折、感染风险上升与 HPA 轴抑制等。它们在哪些领域起作用？分别“在做什么事”？A. 生理学：身体为什么需要它？1.应激与能量调度：糖皮质激素（GCs, glucocorticoids）在感染/炎症/饥饿/压力等情况下，帮助维持葡萄糖供应与能量动员，并对血管反应性与循环稳态具有“许可/维持”作用。2.免疫与炎症“刹车”：通过糖皮质激素受体（GR, glucocorticoid receptor）重排炎症基因网络（如抑制 NF-κB、AP-1 等通路），把免疫反应从“失控放大”拉回“可控范围”，并促进炎症分辨。3.日常稳态：GCs 也参与昼夜节律下的稳态维护，不只是“应激时才上场”。 B. 医学与药学：为什么它是“超级常用药”？ 合成糖皮质激素在多个学科高频使用（过敏/哮喘、风湿免疫、皮肤、移植、血液肿瘤方案支持、危重症等），核心原因是其可通过GR 介导的转录调控强力压制炎症程序；同时给药路径多（局部与全身均可），便于在不少场景实现“靶器官获益更大、全身副作用尽量更小”。糖皮质激素受体GR（glucocorticoid receptor）是什么？ GR 是糖皮质激素作用的主要受体：一种细胞内、配体依赖的转录调控因子，属于核受体（nuclear receptor）超家族；人类基因为 NR3C1（nuclear receptor subfamily 3 group C member 1）。GR 怎么工作？（从细胞质到细胞核的“4 步动作”）1.未结合激素时：GR 多位于细胞质，与伴侣蛋白复合物（Hsp90/Hsp70，heat shock proteins；以及 FKBP51/52，FK506 binding proteins 等）结合并处于可激活状态（也可一定程度穿梭于核/胞质）。2.激素进入细胞并结合 GR：诱导构象变化并触发伴侣蛋白重排。3.GR 转位入核：GR 转位入核后可直接结合 GRE（glucocorticoid response element），或以“系链（tethering）”方式调控 NF-κB、AP-1 等转录因子网络。4.改变基因表达网络：既能上调也能下调，效果强烈依赖细胞类型与染色质环境。结构上，GR 典型由 N 端转录激活区（NTD/AF-1）、DNA 结合域（DBD，含锌指）、铰链区与配体结合域（LBD/AF-2）构成。可以形象理解成：激素是钥匙，GR 是锁；但这把锁一旦被打开，会去“改写一大批基因的开关状态”。备注：伴侣蛋白（molecular chaperones）：帮助受体蛋白维持正确构象并完成装配/运输的“蛋白质助手”；FKBP51/52 属于共伴侣（co-chaperones）。GR 重要吗？ 重要，是很多疾病与药物效应的“总阀门”原因主要有三条：1.几乎所有治疗性糖皮质激素的效果，最终都要通过 GR（及其网络）落地。2.同样剂量有人很有效/有人副作用大，与 GR 相关机制（受体异构体、伴侣蛋白、代谢酶、共调节因子等）有关；但现实中我们缺少可靠的“用药前预测指标”。 3.“激素抵抗（glucocorticoid resistance）”是临床与研究的大问题（如部分哮喘、炎症性疾病、某些肿瘤/免疫状态），常涉及 GR 信号层层环节。如何对 GR“分类”？分类轴子类/选项一句话定义关键细节/常见误区1 按“受体本体”分（isoforms/异构体）GRα（glucocorticoid receptor alpha）经典主力形式：能结合糖皮质激素并介导大多数经典效应不止一种蛋白形态：同属 GRα，仍可因不同翻译起始位点产生多种 N 端长度不同的亚型（文献常用 GR-A/B/C/D 指代），活性谱/基因选择性可不同；因此“同为 GRα”也能出现组织差异GRβ（glucocorticoid receptor beta）由可变剪接产生；通常认为不结合常规糖皮质激素，并可在部分情境下对 GRα 起负向调节/主导负性（dominant-negative）作用误区：把 GRβ 只当“刹车”。GRβ 在一些模型中可能还具有GRα 之外的独立功能（与细胞背景强相关）其他 GR 异构体除 α/β 外，NR3C1 还可产生其他剪接/起始位点组合（综述中常见如 GRγ、GR-P 等命名）“GR 不止 α/β”2 按“信号输出方式”分（signaling modes/信号模式）基因组作用（genomic actions，基因组途径）GR 入核后调控转录，是最系统、最核心的作用方式3 种典型“核内落地方式”：①直绑 GRE（glucocorticoid response element，糖皮质激素反应元件）；②tethering（系链）到炎症转录因子如 NF-κB（nuclear factor kappa B）、AP-1（activator protein-1）；③通过负向调控元件（如 nGRE, negative GRE）等实现下调非基因组作用（non-genomic actions，非基因组途径）更快（秒–分钟级）的效应，可通过膜/胞质层面的相互作用影响信号通路非基因组效应在多种情境都可能出现，机制可涉及膜起始信号、离子通道/激酶级联、以及经典受体参与的快速过程3 按“药理学调节方式”分（pharmacological modulation/药理调节）激动剂（agonists，激动剂；多数临床用药）激活 GR，重塑炎症/免疫与代谢相关网络（疗效强）关键现实点：作用面广 → 副作用也来自‘太广’；并且“全身 vs 局部给药”（吸入/外用/关节腔等）会显著改变风险–收益拮抗剂/阻断剂（antagonists，拮抗剂）阻断 GR 信号，用于需要降低皮质醇效应的特定情境（如库欣综合征相关策略）属于更专门的路径，临床适应证与监测要求更严格选择性 GR 调节剂（SEGRMs/SEGRAs：selective glucocorticoid receptor modulators/agonists）目标：提高治疗指数——尽量保留关键疗效，同时减少部分不良反应疗效/副作用的分离不等同于旧观点“转录激活=副作用、转录抑制=疗效”的简单二分；更像是不同的基因调控谱/共调节因子招募谱/偏向性（biased）常见科研或应用场景场景（科研/应用）典型“在做什么事”常用读出/方法（科研端常见）关键注意点生理学/应激生物学（HPA axis）在应激/炎症/饥饿时：协调能量与免疫反应；通过负反馈把反应拉回稳态皮质醇/ACTH 动态；GR 核转位；转录组变化“稳态维护 + 应激峰值”是同一套系统；时序很重要炎症免疫基础研究通过 GR 重排炎症基因网络：压制过强炎症，促进“分辨/收敛”ChIP-seq/ATAC-seq；GRE 报告基因；与 NF-κB/AP-1 的互作/“tethering”机制不是简单“关免疫”，而是重塑网络；细胞类型依赖很强呼吸：哮喘/慢阻肺（ICS 为核心治疗之一）抑制气道炎症、改善症状与急性发作风险；研究重点之一是激素不敏感/抵抗气道细胞 GR/HDAC2；炎症因子谱；临床表型分层“激素抵抗”涉及多环节（GR 异构体、染色质、氧化应激等）风湿免疫/自身免疫（RA、SLE、血管炎等）快速压制炎症“高峰”，桥接到免疫调节/生物制剂起效炎症评分、CRP；剂量递减试验；靶器官评价获益与风险高度依赖剂量/疗程/给药途径（局部 vs 全身）皮肤科/眼科/耳鼻喉（外用、滴眼、鼻喷）在局部组织把炎症压下来，同时尽量减少全身暴露局部症状量表；皮肤屏障/角膜指标“局部强效 + 低全身暴露”是很多场景常用逻辑（但仍需警惕长期使用）移植/过敏与免疫抑制方案抑制排斥/过敏级联，作为方案组件之一免疫细胞谱；排斥指标；感染监测免疫抑制与感染风险的平衡，是长期使用的核心矛盾肿瘤学与血液科（“辅助用药 + 某些肿瘤的 GR 生物学”）方案支持（止吐/抗过敏/减炎/减水肿等）；同时 GR 在某些肿瘤细胞命运与耐药中也被研究细胞凋亡/转录组；与化疗/免疫治疗联用研究同一通路在不同肿瘤可能“利弊不同”，需要分癌种/分情境解释内分泌：库欣综合征（高皮质醇）与药物对抗用 GR 拮抗/调节降低“过量皮质醇效应”带来的高血压/高血糖等临床终点（血压/糖代谢/症状）；安全性监测代表方向：选择性 GR 调节剂（如 relacorilant 的临床研究）神经精神/压力相关疾病研究研究 GR 过度激活与应激相关表型；探索更“精准”干预动物应激模型；脑区特异多组学；GR 相关基因调控新工具方向：GR 降解剂（PROTAC）等用于“关掉过强 GR”药物研发：提高治疗指数目标是“保留抗炎等获益、减少代谢/骨等副作用”共调节因子招募谱；组织选择性；长期安全性模型共识：不再等同于旧的“转录激活=副作用/转录抑制=疗效”二分法，机制更复杂“解耦副作用”的代表案例“解耦”或“选择性”类固醇/调节剂的临床落地临床获益 vs 骨/糖代谢不良反应例：vamorolone（Agamree）作为“选择性/解耦”类固醇的代表，已获 FDA 批准（DMD）缩写备注：1.HPA axis = hypothalamic–pituitary–adrenal axis（下丘脑–垂体–肾上腺轴）；2.GRE = glucocorticoid response element（糖皮质激素反应元件）；3.NF-κB = nuclear factor kappa B；4.AP-1 = activator protein-1；5.SEGRMs/SEGRAs = selective GR modulators/agonists（选择性 GR 调节剂/激动剂）；6.PROTAC = proteolysis-targeting chimera（靶向蛋白降解嵌合体）。最新研究进展层面方向性结论代表性要点/例子意义机制层面GR 不再被看作“简单进核开关”结构与生物物理新图景：解析 GR 构象、域间通讯与伴侣蛋白（Hsp90/Hsp70 等）如何协同调控 GR 成熟/活化，解释“同样配体为何转录谱不同”把“为什么同药不同效/不同副作用”从现象推进到机制：差异可能来自构象与复合体动态经典问题仍在演进：单体/二聚体/高阶状态多聚化/构象状态被刷新：2025 研究讨论 GR 多聚化构象与功能关系，提示“单体/二聚体/更高阶形式”的旧问题仍在更新影响如何解释“转录调控方式/剂量效应/细胞类型差异”的来源“异构体→个体差异”证据链更系统化GRα/GRβ 及更多异构体（剪接、翻译起始位点等）被更系统梳理，可解释组织特异性与个体差异/抵抗给“个体差异/激素抵抗”一个更可引用、可落地的机制框架药物层面从“激动剂/拮抗剂”走向“选择性调节 + 新型工具”选择性/解耦类固醇临床落地：vamorolone 获 FDA 批准用于 DMD，被视为提高获益/风险的重要里程碑证明“选择性/解耦”不只是概念，开始出现真实世界落地样例新型 SEGRM/SGRM 持续出现（但多在早期）GRM-01（2025）：动物模型提示“抗炎效应与糖/骨代谢不良反应分离”的探索方向（仍偏早期/一期）传达出方向在推进，但仍需谨慎表述证据等级（早期/动物）库欣综合征方向：临床证据持续积累relacorilant：含 Phase 3 与延伸研究，III 期（GRACE/GRADIENT）与长期延伸研究已给出改善高血压/糖代谢等临床终点的证据；监管审批仍以官方最终结论为准。代表“阻断/调节过强皮质醇效应”的现实临床路径仍在加固“不是拮抗而是降解”的新路线GR PROTAC 降解剂（Nature Communications 2023）：作为研究工具/潜在治疗概念，代表新工具链扩展工具箱：从“调节活性”到“直接移除受体”疾病层面激素抵抗与精准分层更受重视严重哮喘激素不敏感/抵抗仍是核心难题；综述强调需个体化分层并总结多层级机制把“为什么有人无效”从单一靶点问题升级为“分层+多环节网络”问题遗传变异与表型关联持续出现NR3C1 变异与哮喘严重度/控制等关联研究持续出现（效应依赖队列与表型定义）提醒：可用来解释差异，但需要避免过度外推主要瓶颈瓶颈本质原因（为什么难）对研发/临床/科研意味着什么疗效与副作用难以彻底解耦旧二分法（转录激活=副作用/转录抑制=疗效）被削弱；更像在不同细胞/染色质/共调节因子背景下，GR 选择性重排网络设计“对所有人都完美”的 SEGRM 很难；同一分子在不同组织可能出现不同收益/风险个体差异 & 用药前预测指标不足受 GR 异构体比例、伴侣蛋白/共调节因子、局部代谢、表观遗传等多因素影响，但缺少稳定、可推广的预测体系现实上难以在用药前精准预测谁更有效/谁更易副作用；试错成本高激素抵抗（glucocorticoid resistance/insensitivity）长期痛点在严重哮喘/慢性气道炎症等，抵抗涉及氧化应激、染色质调控（如 HDAC2 链条）、GR 信号多环节失衡，单点修复往往不够需要“分层 + 组合干预/多靶点”思路；单一机制解释经常不够非基因组（non-genomic）快速效应统一框架不够稳快速效应（秒–分钟）被确认，但受体定位、膜相关机制、不同组织贡献比例复杂“如何利用/规避快速效应”尚难工程化；不同模型结论可能不一致糖皮质激素 & 受体相关代表产品选型表（GR/MR 工具化合物 × 系统/基准与可溶性形式 × 局部制剂/缓释）先定在做什么，再选表：1、要验证 GR/MR 通路、做拮抗/激动对照或机制拆分 → 看表1（受体工具）2、要做定量/质控、建立“系统用基准对照”，或需要水溶/注射形式便于加样/给药 → 看 表2（系统与基准）3、目标是局部治疗/局部模型（气道、鼻腔、眼表、皮肤）或关注“局部滞留/前药/缓释/制剂形态” → 看 表3（局部与制剂）表1｜受体工具化合物：GR/MR 调节、拮抗与探针（机制验证/通路拆分）分类Cas编号阿拉丁货号名称规格或纯度产品应用或特色糖皮质激素受体相关｜GR 调节/拮抗剂1496510-51-0R669597RelacorilantMoligand™, ≥98%选择性 GR 调节/拮抗方向小分子；用于 GR 通路阻断、耐药/代谢效应研究与对照。糖皮质激素受体相关｜GR 拮抗剂2222344-98-9O648769ORIC-101≥99%GR 拮抗剂方向小分子；常用于 GR 阻断、联用机制与“抗糖皮质激素”药理研究对照。糖皮质激素受体相关｜选择性 GR 调节剂1018679-79-2C171679CORT-108297≥97%选择性 GR 调节剂方向工具化合物；用于区分不同 GR 下游通路（转录激活/抑制等）机制研究。糖皮质激素受体相关｜高亲和 GR 激动剂（研究探针）1110-40-3C667938Cortivazol_超高亲和力 GR 激动剂探针；用于受体结合/占有率、信号放大与机制验证实验。糖皮质激素受体相关｜GR 调节/拮抗剂1496508-34-9D649407Dazucorilant_GR 调节/拮抗方向小分子；用于“抗糖皮质激素效应”、应激/代谢相关药理与机制对照。糖皮质激素受体相关｜GR（兼 PR）拮抗剂84371-65-3M126999米非司酮Moligand™, ≥98%常用 GR 阻断对照（亦具 PR 拮抗活性）；用于应激轴/代谢与“抗糖皮质激素效应”机制研究。糖皮质激素受体相关｜选择性 GR 激动剂（研究探针）74915-64-3R613329RU28362Moligand™研究级选择性 GR 激动剂工具化合物；用于受体信号通路、选择性与机制拆分实验。矿皮质激素受体相关｜GC 合成前体/代谢通路对照152-58-9C302974脱氧可的松Moligand™, ≥98%矿皮质激素样（偏 MR 激动）对照；用于盐皮质激素/电解质稳态模型与受体选择性研究。矿皮质激素受体相关｜MR 激动剂64-85-7D13397211-去氧皮质酮Moligand™, ≥97%(HPLC)MR 激动方向的经典类固醇对照；常用于区分 GR vs MR 相关效应与受体通路比较。矿皮质激素样｜MR 激动剂（兼 GC）514-36-3F420036醋酸氟氢可的松≥98%矿皮质激素样代表（MR 活性突出，兼具一定 GC 活性）；用于受体选择性与电解质稳态相关研究对照。表2｜系统/基准用糖皮质激素 + 标准品 + 水溶/注射可用形式分类Cas编号阿拉丁货号名称规格或纯度产品应用或特色分析标准品/标准品｜糖皮质激素52-21-1P129282醋酸泼尼松龙分析标准品常用于 LC/LC-MS 定量校准、方法学开发与质控；亦可作局部制剂/眼科用激素对照品。分析标准品/标准溶液｜糖皮质激素83-43-2M121766甲基泼尼松龙标准溶液Moligand™, 分析标准品, 100 ng/ul in Acetonitrile预配标准溶液（乙腈中），便于直接用于 LC/LC-MS 建曲线与回收率/基质效应评估。分析标准品/标准溶液｜糖皮质激素378-44-9B121764甲醇中倍他米松溶液标准物质Moligand™, 分析标准品, 1.00mg/ml预配标准溶液（甲醇中），适合快速用于定量、方法确认与批间一致性对照。分析标准品｜内源性糖皮质激素（皮质酮）50-22-6C119329皮质甾酮Moligand™, 分析标准品啮齿类常见主导 GC（HPA 轴/应激研究常用）；用于定量检测与内分泌模型对照。内源性糖皮质激素｜细胞培养级50-23-7H657409氢化可的松Moligand™, 无动物源, 低内毒素, 用于细胞培养, ≥98%内源性“皮质醇”对照；无动物源/低内毒素，更适合细胞培养补充 GC 信号或受体激活实验。内源性糖皮质激素｜代谢/酶学对照53-06-5C119445可的松Moligand™, ≥97%需经 11β-HSD 转化为活性皮质醇；常用于代谢/酶学、内分泌与受体研究对照。酯化前体｜糖皮质激素50-04-4C129280醋酸可的松≥98%(HPLC)可的松乙酸酯（前体/酯化形式）；用于代谢/受体比较与制剂对照。系统用｜糖皮质激素53-03-2P116562强的松Moligand™, ≥98%经典系统用 GC 前体；用于体内抗炎/免疫模型与不同类固醇横向比较。系统用/对照｜糖皮质激素50-24-8P276607泼尼松龙≥98%经典系统用 GC；常作体内药效学对照、与其盐/酯之间溶解性与效应差异比较。系统用｜糖皮质激素14484-47-0D129585地夫可特Moligand™, ≥98%(HPLC)系统用激素代表；用于体内抗炎/免疫抑制相关模型与不同 GC 的效力比较研究。API/GMP 原料｜糖皮质激素50-02-2D137736地塞米松≥98%适合制剂开发、工艺/质量研究相关实验用对照。水溶性盐/溶液｜糖皮质激素（磷酸盐）125-02-0P1499394泼尼松龙磷酸二钠Moligand™, 10 mM in Water水溶性磷酸盐溶液，便于水相体系（细胞/体外）直接加样；也常作注射/水溶性形式对照。水溶性盐/注射前体｜糖皮质激素（磷酸盐）2392-39-4D107196地塞米松磷酸钠Moligand™, ≥98%高水溶性磷酸钠盐；适合水相细胞实验与动物给药，常作 GR 激动/抗炎强效对照。水溶性盐/注射前体｜糖皮质激素（半琥珀酸盐）2375-03-3M1688956α-甲基泼尼松龙21-半琥珀酸钠盐≥98%水溶性半琥珀酸钠盐（注射常见形式）；便于水相配制与体内给药/快速起效对照。水溶性盐/注射前体｜糖皮质激素（琥珀酸盐）125-04-2H329245氢化可的松琥珀酸钠≥98%高水溶性琥珀酸钠盐；适合水相体系与动物快速给药，对照“快速可用的注射形式”。注射前体（琥珀酸酯）｜糖皮质激素2203-97-6H132728氢化可的松琥珀酸酯≥95%琥珀酸酯提高水溶性（常与钠盐配套思路）；用于注射/水相配制与体内给药对照。糖皮质激素母体｜研究/对照124-94-7T129207曲安西龙Moligand™, ≥99%曲安西龙母体（非缩酮）；用于 GC 抗炎对照、结构-活性比较或衍生物/制剂研究。表3｜局部制剂为主（吸入/鼻用/眼科/皮肤外用）+ 缓释注射/制剂前体分类Cas编号阿拉丁货号名称规格或纯度产品应用或特色局部注射缓释/外用｜中效糖皮质激素76-25-5T101294曲安缩松Moligand™, ≥99%曲安奈德缩酮类代表；常用于外用或局部注射缓释相关研究，对照“较长作用/较强局部效应”。局部注射缓释｜糖皮质激素（长效）5611-51-8T346633己曲安奈德≥98%典型长效缓释（关节腔/局部注射）代表；用于慢释放动力学、关节炎/局部炎症模型对照。吸入/鼻用｜糖皮质激素5534-09-8B129370二丙酸倍氯米松Moligand™, ≥98%(HPLC)经典吸入/鼻用激素代表；二丙酸酯提高脂溶性与局部作用，适合呼吸道局部抗炎对照。吸入/鼻用｜前药型糖皮质激素126544-47-6C287438环索奈德Moligand™, ≥98%(HPLC)吸入/鼻用前药型（局部活化）代表；适合研究“局部作用/降低系统暴露”的机制与制剂策略。吸入/鼻用（亦可肠道局部）｜糖皮质激素51333-22-3B129941布地奈德Moligand™, ≥98%经典吸入/鼻用激素；也常作为局部抗炎对照用于气道/肠道相关模型与制剂研究。吸入/鼻用｜糖皮质激素80474-14-2F129894丙酸氟替卡松Moligand™, ≥97%经典吸入/鼻用激素对照；适合气道局部抗炎模型、制剂与稳定性研究。吸入/鼻用｜糖皮质激素（长效）397864-44-7F413479糠酸氟替卡松Moligand™, ≥98%高效长效吸入/鼻用激素代表；适合呼吸道局部作用、受体效力与滞留特性比较研究。外用/鼻用｜强效糖皮质激素83919-23-7M157957糠酸莫米他松≥98%(HPLC)强效局部激素代表（皮肤外用/鼻用常见）；适合高效局部抗炎对照与配方研究。眼科局部｜糖皮质激素（软甾体）82034-46-6L129196依碳酸氯替泼诺Moligand™, ≥98%(HPLC)眼科“软甾体”代表（易代谢失活的设计思路）；适合滴眼剂/眼表炎症模型与配方研究。眼科局部｜糖皮质激素426-13-1F138831氟米龙Moligand™, ≥98%(HPLC)常见眼科用激素对照；适合眼部局部抗炎、滴眼制剂筛选与稳定性研究。眼科局部｜强效糖皮质激素23674-86-4D129472二氟孕甾丁酯Moligand™, ≥98%眼科强效含氟激素代表；适合滴眼制剂/眼内或眼表炎症模型对照与方法学研究。眼科局部｜糖皮质激素49697-38-3R337073利美索龙Moligand™眼科用激素代表；适用于滴眼制剂筛选、眼表/前节炎症模型与对照实验。皮肤外用｜低～中效糖皮质激素638-94-8D129996地索奈德Moligand™, ≥99%低～中效外用激素对照；适合皮肤炎症模型、配方与渗透/滞留筛选。皮肤外用｜中效糖皮质激素（酯）57524-89-7H336850氢化可的松17-戊酸酯≥99%17-戊酸酯提高脂溶性与皮肤滞留；适合外用配方、透皮与局部药效对照。皮肤外用｜中效糖皮质激素（酯）13609-67-1H15736917-丁酸氢化可的松酯≥98%(HPLC)丁酸酯外用代表；用于外用制剂筛选、皮肤局部抗炎对照与渗透评价。皮肤外用｜前药型糖皮质激素74050-20-7H648363Hydrocortisone aceponate_外用前药型设计（局部活化/滞留）；常用于皮肤炎症模型、配方与透皮评价。皮肤外用｜中效糖皮质激素（酯）66734-13-2A354948丙酸氯米松_外用酯化激素代表；适合皮肤局部抗炎对照与制剂筛选。皮肤外用｜中效糖皮质激素25122-57-0C303211丁酸氯倍他松≥98%中效外用激素代表；适合皮肤局部抗炎对照、配方与刺激性/耐受性比较研究。皮肤外用｜前药型糖皮质激素73771-04-7P348401泼尼甲酸盐Moligand™, ≥99%外用前药型设计（局部活化/降低全身暴露倾向）；常用于皮肤制剂与药效学对照。皮肤外用｜中～强效糖皮质激素2152-44-5B123298倍他米松 17-戊酸酯≥98%戊酸酯提高脂溶性与局部滞留；用于外用药效、渗透与配方筛选对照。皮肤外用｜中～强效糖皮质激素（新戊酸酯）2002-29-1F486837新戊酸氟米松_新戊酸酯提高脂溶性与局部滞留；用于外用配方、渗透/滞留与药效对照研究。皮肤/黏膜局部｜糖皮质激素152-97-6F337919氟可龙≥96%局部用 GC 代表；适合皮肤/黏膜局部抗炎对照与相关制剂研究。皮肤外用｜强效糖皮质激素356-12-7F129893氟轻松Moligand™, ≥98%(HPLC)强效外用激素代表；用于皮肤炎症模型、透皮/滞留评价与配方优化。皮肤外用｜强效糖皮质激素（含氟）67-73-2F129244醋酸肤轻松Moligand™, ≥98%典型强效外用糖皮质激素对照；常用于皮肤炎症/免疫反应模型、外用制剂筛选与“局部滞留/透皮”比较研究。皮肤外用｜强效糖皮质激素382-67-2D336852去羟米松Moligand™, ≥98%强效外用激素代表；常用于皮肤抗炎对照、红斑/水肿等药效学模型与配方筛选。皮肤外用｜强效糖皮质激素33564-31-7D337813双醋酸双氟松Moligand™, ≥98%强效外用“二醋酸酯”代表；用于提高脂溶性/皮肤滞留的配方策略与药效对照。皮肤外用｜强效糖皮质激素59198-70-8D304084双氟可龙戊酸酯≥98%强效外用“戊酸酯”代表；用于局部抗炎对照、配方与皮肤滞留/渗透研究。皮肤外用｜强效糖皮质激素（双酯）5593-20-4B12329717,21-二丙酸倍他米松酯≥97%双酯提高脂溶性与皮肤滞留；用于强效外用对照、配方与渗透/滞留评价。皮肤外用｜超强效糖皮质激素25122-46-7C123300丙酸氯倍他索Moligand™, ≥98%超强效外用激素代表；用于强抑炎对照、受体高激动效应研究或高门槛皮肤模型。皮肤外用｜超强效糖皮质激素66852-54-8H129467卤贝他索丙酸酯≥98%超强效外用激素代表；用于强抑炎对照、难治皮肤模型与受体效力上限对照。酯化/脂溶性制剂前体｜糖皮质激素1177-87-3D773875乙酸地塞米松Moligand™, ≥96%脂溶性乙酸酯；适合有机溶剂配制、局部制剂/缓释或眼科对照研究。酯化/局部制剂｜糖皮质激素50-03-3H102192醋酸氢化可的松≥98%乙酸酯常用于外用/局部制剂对照；适合配方筛选、稳定性与局部药效研究。酯化/制剂前体｜糖皮质激素53-36-1M3361976α-甲基泼尼松龙21-乙酸盐≥98%甲基泼尼松龙乙酸酯类；常用于制剂/缓释策略、结构-活性比较与对照实验。阿拉丁：