Chemical screening campaigns require a miniaturized assay format for which 48-well or 96-well microplates are well suited. Correspondingly, the model plant Arabidopsis thaliana is a good choice as experimental system, because the small size seedlings can be directly grown in microplates supplemented with liquid medium. Any biological or biochemical readout that can be analyzed at the seedling stage is suitable for modulation by chemical interference, but a robust and reliable bioassay is essential because larger screening campaigns will usually be performed with single assays to generate first hits. A quantitative bioassay is preferable because it allows distinguishing between effective and less effective compounds. Below is a compilation of screens for bioactive chemicals that we recently performed. This is not only meant to demonstrate the range of experimental possibilities but rather to inspire your fantasy to plan and design your own chemical screens.

Modulation of PAMP-triggered immune responses in Arabidopsis

Using a transgenic Arabidopsis line harboring a elicitor-responsive reporter gene (ATL2p::GUS) we screened for chemical compounds modulating plant defense responses. First, we established an experimental setup using submerged cultures of Arabidopsis seedlings in 96-well microplates that permits chemical intervention of rapid elicitor-mediated immune responses. Screening of a chemical library comprising 120 small molecules with established biological activities revealed four compounds reducing cellulysin- or flg22-activated gene expression of the pathogen-associated molecular patterns (PAMP)-responsive ATL2p::GUS reporter gene. One compound, oxytriazine, was found to induce ATL2 gene expression in the absence of PAMP. By monitoring additional flg22-triggered immediate early plant responses, we found that two compounds, triclosan and fluazinam, interfere with the accumulation of reactive oxygen species (ROS) and internalization of the activated plasma membrane resident FLS2 immune receptor. Using triclosan structure types and enzyme activity inhibition assays, Arabidopsis enoyl-acyl carrier protein reductase (MOD1), a subunit of the fatty-acid synthase type II (FAS II) complex, was identified as a likely cellular target of triclosan. Inhibition of all tested elicitor-triggered early immune responses by triclosan indicates a potential role for signaling lipids in flg22-triggered immunity. Chemical profiling of eca mutants, each showing deregulated ATL2 gene expression, with the identified compounds revealed mutant-specific response patterns and allowed us to deduce tentative action sites of ECA genes relative to the compound targets.

Reference: Serrano et al. (2007) J Biol Chem 282: 6803-6811.

Suppression of RPM1 resistant gene-dependent hypersensitive cell death in Arabidopsis

The Arabidopsis thaliana resistance gene RPM1 encodes an intracellular immune sensor that conditions disease resistance to Pseudomonas syringae expressing the type III effector protein AvrRpm1. Conditional expression of this type III effector in a transgenic line carrying avrRpm1 under the control of a steroid-inducible promoter results in RPM1-dependent cell death that resembles the cell death response of the incompatible RPM1-avrRpm1 plant-bacterium interaction. We established a chemical screen for small molecules that suppress steroid-inducible and RPM1-avrRpm1-dependent cell death in Arabidopsis seedlings. Screening of a library comprising 6,800 compounds of natural origin identified two trichothecene-type mycotoxins, 4,15-diacetoxyscirpenol (DAS) and neosolaniol (NEO), which are synthesized by Fusarium and other fungal species. However, protein blot analysis revealed that DAS and NEO inhibit AvrRpm1 synthesis rather than suppress RPM1-mediated responses. This inhibition of translational activity likely explains the survival of the seedlings under screening conditions. Likewise, flg22-induced defense responses are also impaired at the translational, but not the transcriptional, level by DAS or NEO. Unexpectedly, both compounds not only prevented AvrRpm1 synthesis, but rather caused an apparent hyper-accumulation of RPM1 and HSP70. The hyper-accumulation phenotype is likely unrelated to the ribotoxic function of DAS and NEO and could be due to an inhibitory activity on the proteolytic machinery of Arabidopsis or elicitor-like activities of trichothecenes.

Reference: Serrano et al. (2010) Planta 231: 1013-1023.

Derepression of PAMP-inhibited anthocyanin accumulation in Arabidopsis

Recognition of microbe-associated molecular patterns (MAMPs) leads to the generation of MAMP-triggered immunity (MTI), which restricts the invasion and propagation of potentially pathogenic microbes. In Arabidopsis thaliana FLS2 and EF-TU function as receptors for the bacterial MAMP epitopes flg22 (of flagellin) and elf18 (of elongation factor Tu), respectively. Perception of flg22 leads to activation of plant defense responses (e.g. PR gene expression, ROS spiking, callose deposition) and additionally causes the repression of flavonoid accumulation upon simultaneous application of sucrose or UV light stress. However, the functional significance of this MTI-associated signaling output remains unknown. To further dissect the FLS2 signaling pathway, we screened a chemical library of 6,800 natural compounds and identified nine small molecules that de-repress flavonoid accumulation in the presence of flg22. Remarkably, one of the identified compounds uncoupled flavonoid repression and PR gene activation from ROS accumulation, MAP kinase activation, and callose deposition, corroborating a close link between flavonoid accumulation and PR gene expression. Collectively, the data suggest that repression of flavonoid accumulation has the function to relieve the inherent inhibitory action of flavonoids on plant defense expression. However, the molecular mode of action of the identified compounds still needs to be unraveled.

Reference: Serrano et al. (2012) Plant Physiol 158: 408-422.

Screening for bioactive small molecules by in vivo monitoring of luciferase-based reporter gene expression

We developed a chemical screening procedure that relies on inducible firefly luciferase (LUC) reporter constructs in Arabidopsis thaliana, specifically, a transgenic line harboring LUC under the control of the jasmonate-inducible promoter of the LIPOXYGENASE 2 (LOX2) gene. The advantage of reporter-based screens is that they afford quantitative data that allow discrimination between compounds with high and low bioactivity. The LUC reporter, in contrast to the β-glucuronidase reporter, further allows in vivo monitoring of activity without compromising plant viability, thus facilitating subsequent genetic screens. Furthermore, the choice of an inducible reporter gene system permits bidirectional screening for either activators of gene expression or inhibitors that impair induced gene expression. Sifting through a chemical library of about 1,700 small molecules of natural and semi-synthetic origin, we identified a single compound, 1-propyl-2-carboxy-3,8-dihydroxy-9,10-anthracenedione (766), that seemed to activate expression of the reporter gene LOX2p::LUC. Conversely, screening of the same library for inhibitors uncovered three small molecules that strongly impaired methyl jasmonate (MeJA)-induced expression of the LOX2p::LUC reporter gene. The identified compounds, cycloheximide, and the two trichothecene mycotoxins diacetoxyscirpenol and neosolaniol, have in common that they inhibit protein synthesis as previously recognized. Any of these compounds may serve as suitable positive control in reporter-based screens for inhibitors. The MeJA-stimulated LOX2p::LUC expression varies considerably across all samples of this primary screen. This is largely attributed to variable seedling size and orientation within individual microplate wells, thus leading to variable luminescence detection, which effectively renders the assay only semi-quantitative. However, stringent selection criteria for primary hits and their confirmation with increased samples numbers will reduce the false discovery rate.

Reference: Meesters et al. (2012) Plant Chem Biol (in press).