Gold nanoparticles (AuNPs) with diverse physicochemical properties are reported to affect biological systems differently, but the relationship between the physicochemical properties of AuNPs and their biological effects is not clearly understood. Here, we aimed to elucidate the molecular origins of AuNP-induced cytotoxicity and their mechanisms, focusing on the surface charge and structural properties of modified AuNPs. We prepared a library of well-tailored AuNPs exhibiting various functional groups and surface charges. Through this work, we revealed that the direction or the magnitude of surface charge is not an exclusive factor that determines the cytotoxicity of AuNPs. We, instead, suggested that toxic AuNPs share a common structural characteristics of a hydrophobic moiety neighbouring the positive charge, which can induce lytic interaction with plasma membrane. Mechanistic study showed that the toxic AuNPs interfered with the formation of cytoskeletal structure to slow cell migration, inhibited DNA replication and caused DNA damage via oxidative stress to hinder cell proliferation. Gene expression analysis showed that the toxic AuNPs down-regulated genes associated with cell cycle processes. We discovered structural characteristics that define the cytotoxic AuNPs and suggested the mechanisms of their cytotoxicity. These findings will help us to understand and to predict the biological effects of modified AuNPs based on their physicochemical properties.

Live-cell-based biosensors have emerged as a useful tool for biotechnology and chemical biology. Genetically encoded sensor cells often use bimolecular fluorescence complementation or fluorescence resonance energy transfer to build a reporter unit that suffers from nonspecific signal activation at high concentrations. Here, we designed genetically encoded sensor cells that can report the presence of biologically active molecules via fluorescence-translocation based on split intein-mediated conditional protein trans-splicing (PTS) and conditional protein trans-cleavage (PTC) reactions. In this work, the target molecules or the external stimuli activated intein-mediated reactions, which resulted in activation of the fluorophore-conjugated signal peptide. This approach fully valued the bond-making and bond-breaking features of intein-mediated reactions in sensor construction and thus eliminated the interference of false-positive signals resulting from the mere binding of fragmented reporters. We could also avoid the necessity of designing split reporters to refold into active structures upon reconstitution. These live-cell-based sensors were able to detect biologically active signaling molecules, such as Ca and cortisol, as well as relevant biological stimuli, such as histamine-induced Ca stimuli and the glucocorticoid receptor agonist, dexamethasone. These live-cell-based sensing systems hold large potential for applications such as drug screening and toxicology studies, which require functional information about targets.

Cortisol, a stress hormone, plays key roles in mediating stress and anti-inflammatory responses. As abnormal cortisol levels can induce various adverse effects, screening cortisol and cortisol analogues is important for monitoring stress levels and for identifying drug candidates. A novel cell-based sensing system was adopted for rapid screening of cortisol and its functional analogues under complex cellular regulation. We used glucocorticoid receptor (GR) fused to a split intein which reconstituted with the counterpart to trigger conditional protein splicing (CPS) in the presence of targets. CPS generates functional signal peptides which promptly translocate the fluorescent cargo. The sensor cells exhibited exceptional performance in discriminating between the functional and structural analogues of cortisol with improved sensitivity. Essential oil extracts with stress relief activity were screened using the sensor cells to identify GR effectors. The sensor cells responded to peppermint oil, and L-limonene and L-menthol were identified as potential GR effectors from the major components of peppermint oil. Further analysis indicated L-limonene as a selective GR agonist (SEGRA) which is a potential anti-inflammatory agent as it attenuates proinflammatory responses without causing notable adverse effects of GR agonists.

Although in vitro sensors provide facile low-cost ways to screen for biologically active targets, their results may not accurately represent the molecular interactions in biological systems. Cell-based sensors have emerged as promising platforms to screen targets in biologically relevant environments. However, there are few examples where cell-based sensors have been practically applied for drug screening. Here, we used engineered cortisol-detecting sensor cells to screen for natural mimetics of cortisol. The sensor cells were designed to report the presence of a target through signal peptide activation and subsequent fluorescence signal translocation. The developed sensor cells were able to detect known biological targets from human-derived analytes as well as natural product extracts, such as deer antlers and ginseng. The multi-use capability and versatility to screen in different cellular environments were also demonstrated. The sensor cells were used to identify novel GR effectors from medicinal plant extracts. Our results suggest that decursin from dongquai had the GR effector function as a selective GR agonist (SEGRA), making it a potent drug candidate with anti-inflammatory activity. We demonstrated the superiority of cell-based sensing technology over in vitro screening, proving its potential for practical drug screening applications that leads to the function-based discovery of target molecules.

No abstract