Only one particle size of clay was used here, and diffusion rates were lower in this environment. In systems with low diffusion rates, exudate concentration is likely higher around the roots, which might lead to higher exudate re-uptake than in systems with larger particle sizes . Clays with different particle sizes might provoke a root morphology and exudation profile distinct from glass bead-grown plants and is worth further investigation. Further, substrate particle size might be a factor defining the amount of exudates present in soils.The largest difference in exudate profiles observed was between in situ clay-grown plants and other in situ conditions. Notably, the distinct exudation of clay-grown plants disappeared when exudates were collected in vitro, indicating that the differences observed resulted from the presence of clay, and not from an altered exudation of compounds by B. distachyon. About 20% of compounds were distinct between hydroponic and clay exudates, and most of these compounds were reduced in abundance in the presence of clay. Among these compounds were organic acids, amino acids, and nucleosides. When clay was incubated with a defined medium, 75% of compounds were reduced in abundance,container vertical farming among them negatively and positively charged compounds, as well as neutral compounds. The higher metabolite retention by clay in the defined medium experiment compared with the plant experiment might be due to several factors: The clay was incubated for two hours with the defined medium, but for three weeks with plants producing exudates.
Although exudates were also collected for two hours in the plant experiment, the clay was likely already saturated to some degree with exudates. The quantification of exudate amounts at different plant developmental stages in future studies would enable a better estimation of the total amount of compounds exuded and would correct for the difference in the two experimental setups. The reduction of metabolite abundance in the presence of clay is most likely due to its high ion exchange capacity, compared with quartz-based particles such as sand or glass beads . Previous studies investigating sorption of bacterial lysates to ferrihydrite found a depletion of more than half of the metabolites . Similarly, incubation of bacterial lysates with a soil consisting of 51% sand, 28% silt, and 21% clay resulted in low metabolite recovery rates . These findings are consistent with our data. Interestingly, two nitrogenous metabolites were higher in abundance in exudates of in situ clay-grown plants, . These compounds were not detected in clay negative controls, or in in vitro exudates of clay-grown plants, making it likely that the presence of plants leads to the release of these compounds from clay. Multiple examples exist in literature that describe a release of compounds from minerals by specific exudates. For example, plant-derived organic acids such as malate and citrate solubilize mineral-bound phosphate , and plant-derived oxalate releases organic compounds bound to minerals, making them available to microbial metabolism . Altered exudation depending on the growth substrate was also described for tomato, cucumber, and sweet pepper growing in stone wool, with higher exuded levels of organic acids and sugars compared with glass bead-grown plants .
The authors suggest that the presence of aluminum ions in stone wool might be responsible for the altered exudation observed. As the authors did not investigate in vitro-collected exudates of stone wool grown plants, it is unclear to which degree the observed effect was due to changes in plant metabolism or due to the presence of stone wool. In soils, metabolite sorption to minerals can lower decomposition rates . Also, the amount of clay in soil is correlated with retention of labeled carbon in soils . In clay-dominated soils, the size of clay particles shapes how much carbon can be retained: large clay aggregates were found to adsorb more carbon than smaller aggregates . Here, we only investigated one size of clay particles. Thus, it would be prudent to investigate the sorption behavior of clays with different particle sizes, and the ability of microbes to subsequently desorb these compounds. In natural systems, the presence of large amounts of clay with a specific particle size likely results in the sorption of plant-derived compounds to particles, changing the direct availability of these compounds to heterotroph organisms and, thus, altering soil processes.Microbes can release sorbed compounds from minerals, and they likely preferentially colonize minerals that are associated with compounds missing from the environment . The rhizobacterium Pseudomonas fluorescens utilized in this study was indeed able to desorb metabolites from clay, utilizing them as a carbon source for growth . In soils, root exudation creates zones with high metabolite concentrations.Although P. fluorescens was able to grow on particles conditioned with exudates, it did not grow on the effluent of the washed particles.
This suggests that the organism is able to release mineral-bound metabolites as an additional source for growth—a trait that supports competitiveness and survival in the rhizosphere. Root-associated bacteria have distinct exudate substrate preferences from bulk soil bacteria , which might also define the kind of compounds bacteria are able to release from minerals . Our results are further evidence that minerals play an important role in plant–microbe interactions by sorbing root exudates, which can later be solubilized by microbes for growth. We conclude that alteration in particle size affects root morphology in B. distachyon. Root exudation was constant per root fresh weight, and the exudate metabolite profiles were similar across root morphologies. Mass spectrometry imaging detected ion abundances across various regions of the root system, suggesting involvement of different tissues in exudation. Exudates were strongly sorbed by clay, significantly reducing the availability of free metabolites. Some of the clay-bound metabolites however could be utilized by a rhizobacterium for growth. Soil clay content thus is likely an important factor to consider when investigating root exudates or plant–microbe interactions in natural environments.Substrates with various particle sizes and surface chemistries were chosen as experimental systems to assess changes in root morphology and exudation. The particle sizes used correspond to large soil particles , intermediary particles , and small particles . Glass beads constitute an inert experimental system, for which the diameter of the spheres is defined, and the particles have a defined mineral composition . The sand and clay substrates constitute more natural environments than glass beads . The sand and clay mineral composition was either defined by the manufacturer or determined here. To assess the chemical properties of the substrates, the sorption of metabolites to the substrates was assessed by incubating them with a defined medium . The defined medium consisted of amino acids, organic acids, sugars, nucleobases, nucleosides, and others, see Table S2. The various substrates were sterilized, and the defined medium was prepared as a sterile, 20 µM equimolar solution. The substrates were fully submerged in the defined medium in sterile conditions and incubated at 24°C for 8 hr. The sterility of the system was confirmed by plating an aliquot on LB plates, followed by a 3-day incubation. The defined medium was removed by pipetting. The recovered volume was recorded; for substrates with smaller particle sizes, the entire volume could not be reclaimed. Samples were filtered through a 0.45-µm filter and frozen at −80°C. See “Liquid chromatography–mass spectrometry sample preparation” for sample processing.The frozen samples were lyophilized , resuspended in 3 ml LC/MS grade methanol , vortexed three times for 10 s, sonicated for 20 min in a water bath at 24°C, and incubated at 4°C for 16 hr for salt precipitation. Samples were then centrifuged for 5 min at 5,000 g and 4°C, and supernatants were transferred to new micro-centrifuge tubes and evaporated at 24°C under vacuum until dry. The dried extracts were resuspended in 500 µl LC/MS grade methanol,hydroponic vertical garden and the above procedure centrifugation, drying and resuspension procedure were repeated. Finally, samples were resuspended in 100% LC/MS grade methanol with 15 µM internal standards , with the solvent volume being proportional to the root biomass .In the various experimental systems used here, exudation rates could be limited by diffusion. To determine the diffusion rates of various substrates, sterilized substrates were added to pipettes with a 50 ml volume. The pipettes were sealed at the bottom with parafilm, placed vertically, 50 ml of substrate was added, and approximately 25 ml of sterilized 0.5× MS was added to fully immerse submerse the substrate . The experimental setup was sterile, but the experiment was conducted in non-sterile conditions.
Congo red 4B was solubilized in water at a concentration of 20 mg/ ml, and 250 µl of the dye was added simultaneously to pipettes containing the various substrates. The front of the dye was followed recorded over time up to 4.5 hr. Initially, the movement of the dye front was supported by mass flow .Brachypodium distachyon Bd21-3 seeds were dehusked and sterilized in 70% v/v ethanol for 30 s, and in 6% v/v NaOCl with 0.1% v/v Triton X-100 for 5 min, followed by five wash steps in water. Seedlings were germinated on 0.5× Murashige & Skoog plates in a 150 µmol/ m2 s −1 16-hr light/8-hr dark regime at 24°C for three days. Weck jars were rinsed five times with MilliQ water, sprayed with 70% v/v ethanol, treated with UV light for 1 hr in a laminar flow hood, and dried over night. The jars were filled with 150 ml of the respective substrate, and 50 ml of 0.5× MS basal salts liquid medium. Three seedlings were transferred into each jar, with the roots buried in the substrate. As a control, jars without substrate were prepared: PTFE mesh was cut to fit the size of the jar, and autoclaved. Three openings were cut into the mesh to hold the seedlings. The mesh was transferred to jars with 50 ml 0.5× MS medium. For each condition, an experimental negative control was prepared containing substrate, but no seedlings. The experimental control jars were treated the same as the jars containing plants. To enable gas exchange, two strips of micropore tape were placed across the jar opening, and the lid was set on top and wrapped with micropore tape to ensure sterility. Plants were grown in a 16-hr light/8-hr dark regime at 24°C with 150 µmol/m2 s −1 illumination, and the growth medium was replaced weekly: The old medium was removed by pipetting, and new 0.5× MS was added. Sterility of the jars was tested in week 3 by plating 50 µl of medium on Luria-Bertani plates, following by three days incubation at 24°C. Plants were grown for 21 days before exudate collection and root morphology determination.Mass spectrometry imaging was used to investigate spatial patterning of root exudation across the root system. Brachypodium distachyon seeds were sterilized and germinated on 0.5 MS plates as described above. A stainless steel MALDI plate was cleaned with 100% v/v ethanol, and a 7 × 7 cm square of aluminum foil was affixed to the plate with double-sided scotch tape. The foil was overlayed with 4 ml 0.1% ultrapure agarose to create a thin layer of agarose. Four-day-old seedlings were transferred to the agarose layer, and gentle pressure was applied with a spatula to embed the roots in the agarose. The stainless steel plate was transferred into a petri dish plate to keep humidity constant and incubated for 6 hr in a growth chamber with 150 µmol/m2 s −1 illumination and 24°C. MALDI matrix was prepared as follows: 10 mg/ml a-cyano-4-hydroxycinnamic acid and 10 mg/ml Super-DHB were dissolved in 75% v/v methanol, 24.9% LCMS-grade water, and 0.1% formic acid. The plate with the seedlings was removed from the growth chamber, leaves were cut to ensure flatness of the sample, and the sample was sprayed with MALDI matrix, which simultaneously desiccated the tissue. The plate was incubated for 24 hr in a vacuum desiccator until completely dry. Mass Spectrometry Imaging was performed using a 5,800 MALDI TOF/TOF in positive reflector MS mode with an Nd:YAG laser acquiring spectra over a range of 50−2000 Da and accumulating 20 shots/spot. The 4,800 Imaging Tool software was used to raster across the sample and record spectra in x-y step-sizes of 75 × 75 μm. Data viewing and image reconstruction were performed using OpenMSI .Ion chromatograms corresponding to metabolites represented within our in-house standard library were extracted from LC/MS data with Metabolite Atlas.