Aim 1.1 – Physiological common garden experiment
We established three plant growth systems for our common garden experiments: an irrigated field plot in the Negev desert, pots, including a specialized soil mixture, and sand tubes, for exudate collection. We found that desert plants indeed differ from their Mediterranean relatives in response to drought conditions.
Aim 1.2 – Common garden experiment
We’ve set several iterations of common garden experiments. The first iteration was performed in a controlled field plot in the Negev region of Israel. We planted six species: S. erysimoides, S. officinale, M. chia, M. Africana, D. harra and D. erucoides. Endosphere, rhizosphere and soil clustered separately. In the endosphere, the effect of the drought treatment was statistically significant.
Aim 2.1 –culture collection
We plated ground roots on six different media: ½ LB, 1/10 LB, ½ TSB, 1/10 TSB, R2A and YEM, obtaining 464 isolates and applied a pre-screening approach, using a Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometer (MALDI-TOF). We clustered spectra by similarity using Cosine clustering to identify redundant isolates. Finally, 130 representative strains were chosen, and these were Sanger sequenced.
Aim 1.3 – Metagenomic data analysis
We developed a read-based annotation pipeline, in order to generate a feature table that we could use. We generated functional feature tables, and compared the metagenomes. Mean annual precipitation at the sampling site significantly affected microbiome composition at the site, for both soil and rhizosphere samples, supporting our hypothesis.
Aim 2.2: Community deconstruction for trait discovery
We are applying a new screening strategy for ecologically relevant traits within the microbiome, that we term ‘top-down community deconstruction. We constructed a closed model system consisting of Arabidopsis thaliana Col-0 plants grown in gamma-irradiated coconut coir within magenta boxes, covered with a transparent breathable membrane. We replaced the destroyed microbiota with a 150-member synthetic community (SynCom). We infected leaves of SynCom-inoculated plants with a lux-tagged Pseudomonas syringae strain (Pto), and measured the resulting population size of Pto per milligram of leaf tissue, six days post-infection.
Aims 3.1-3.2 Genomic adaptation to aridity
We grow Isatis lusitanica and Isatis microcarpa in 30 ml syringe barrels, with filter paper in the bottom, and irrigate with measured quantities of water. At the end of the experiment, we flush the sand, to collect root exudates. Samples were analyzed in an untargeted fashion, with both gas chromatography/mass spectrometry (GC/MS), using a reference library of organic acids, amino acids and sugars, and liquid chromatography–mass spectrometry (LC-MS), with a large reference library of natural products. Remarkably, the GC/MS analysis suggests a surprisingly strong response to drought conditions, in both plant species, reflected in a much higher level of exudation compared with the plants grown under full irrigation. Drought triggered the release of the sugars trehalose, sucrose and glucose, as well as several amino acids and other organic acids. LC-MS analysis also revealed a pronounced drought effect in exudation profiles. Interestingly, this effect is stronger in the case of I. macrocarpa, the desert species.
