To the best of our knowledge, the response of plants with decreased root exodermal suberin levels to water limitation has never been investigated. The importance of plant radial and cellular anatomy has also long been known as critical to our understanding of the role of plant roots in water uptake in the face of water deficit. Therefore, our findings provide direct evidence, via genetic perturbation, for the role of suberin in a specific cell type mediating tomato’s adaptive response to water deficit. Further, they impart a model by which exodermal suberin barriers contribute to whole-plant water relations in the absence of a suberized endodermis. While our findings are informative about the importance of suberin in the maintenance of transpiration and stomatal conductance under soil water deficit, our conclusions are limited to a particular stage of plant growth. Changes in response to water limitation in the field, particularly with genotypes with modified suberin that impart better maintenance of water potential, remains to be investigated. Suberin in plants roots has recently been proposed to be an avenue to combat climate change including via sequestration of atmospheric CO2 as well as conferring drought tolerance. This study provides evidence that root suberin is necessary for tomato’s response to water-deficit conditions. Increasing suberin levels within the root exodermis and/or the endodermis may indeed serve as such an avenue. The constitutive production of exodermal suberin in the drought-tolerant and wild relative of tomato, S. pennellii ,30 litre plant pots bulk certainly provides a clue that maintenance of suberin in non-stressed and stressed conditions may result in such a benefit.
However, trade-offs of such an increase must also be considered. Increased suberin levels have been associated with pathogen tolerance, but can also serve as a barrier to interactions with commensal microorganisms and constrain nutrient uptake, plant growth or seed dormancy. Regardless, this complex process serves as an elegant example of how plant evolution has resulted in a gene regulatory network with the same parts but distinct spatial rewiring and contributions of the different genes. Collectively, this rewiring results in the distinct but precise spatiotemporal biosynthesis and deposition of this specialized polymer to perform the equivalent function of endodermal suberin in a plant’s response to the environment.Seedlings of SlCO2p:TRAP and AtPEPp:TRAP cv. M82 were transplanted into 15 cm × 15 cm × 24 cm pots with Turface Athletic Profile Field & Fairway clay substrate pre-wetted with a nutrient water solution . Plants were grown in a completely randomized design for 31 days in a growth chamber at 22 °C, 70% relative humidity, 16 h/8 h light/dark cycle and 150–200 mmol m−2 s−1 light intensity. For ‘well-watered’ conditions, we maintained substrate moisture at 40–50% soil water content. For water-deficit treatment, we withheld water from the plants for 10 days before harvest, and for waterlogged conditions, we submerged the pot until the root–shoot junction. We harvested the roots as close to relative noon as feasible by immersing the pot into cool water, massaging the root ball free, rinsing three times sequentially with water, dissecting the root tissues and flash-freezing with liquid nitrogen. We harvested the lateral roots and 1 cm root tips of adventitious roots.
Sequencing libraries of adventitious roots were generated for each line in control and waterlogging conditions, and from lateral roots in control, waterlogging and water deficit conditions in four biological replicates per genotype/treatment, except for SlCO2p:TRAP lateral roots in control conditions . Total RNA was isolated from these roots as previously described, and non-strand specific random primer-primed RNA-seq library construction was performed as originally described. RNA-seq libraries were pooled and sequenced with the Illumina HiSeq4000 .Seedlings were transferred to 0.5 l cones containing Turface pre-wetted with a nutrient water solution . All pots were weight adjusted and a small set of pots were dried so that the percentage of water in the soil could be calculated. Plants were then grown in a completely randomized design for 3 weeks in a growth chamber at 22 °C, 70% relative humidity, 16 h/8 h light/dark cycle and ~150 µmol m−2 s−1 light intensity, and watered to soil saturation every other day. At the end of the first week, vermiculite was added to limit water evaporation from the soil. After 3 weeks, plants of each line were randomly assigned into two treatment groups and exposed to different treatments for 10 days. Water-limited plants were exposed to water deficit by adjusting pot weights daily with nutrient solution until a target soil water content of 40–50% was obtained. On the day of harvesting, between 09:00 to 12:00, stomatal conductance and transpiration were measured on the abaxial surface of the terminal leaflet of the third leaf or the youngest fully expanded leaf using a LICOR-6400XT portable photosynthesis system. Light intensity was kept at 1,000 µmol m−2 s−1, with a constant air flow rate of 400 µmol s−1 and a reference CO2 concentration of 400 µmol CO2 mol−1 air. The third primary leaflet was collected for measuring relative water content using a modified version of a previously established protocol. Fresh leaves were cut with a scalpel leaving a 1-cm-long petiole and the total fresh weight was measured. Leaves were then placed in individual zipper-locked plastic bags containing 1 ml of deionized water, making sure that only the leaf petiole is immersed in the solution.
Bags were incubated at 4 °C. After 8 h, leaves were taken out of the bags, placed between two paper towels to absorb excess water and then weighed to determine the turgid weight . Each sample was then placed into a paper bag and dried in a 60 °C dry oven for 3–4 days. Dried samples were weighed , and relative water content was calculated as: RWC = × 100/. A section of the fourth leaf, containing the terminal and primary leaflets, was used to measure stem water potential using a pump-up pressure chamber . The root systems were harvested by immersing the cone into water, massaging the root ball free, rinsing and removing excess water with paper towels. The middle section of the root system was sectioned using a scalpel. Around 300 mg of the dissected root tissue were added to Ankon filter bags . Bags were transferred into a glass beaker,wholesale plant containers an excess of chloroform:methanol was added and extracted for 2 h. Fresh chloroform:methanol was replaced and the extraction was repeated overnight under gentle agitation . Fresh chloroform:methanol was added and samples further extracted for 2 h. The extraction was repeated overnight twice with fresh chloroform:methanol . Finally, samples were extracted with methanol for 2 h. Methanol was removed and bags were dried in a vacuum desiccator for 72 h. Suberin monomer analysis was performed in these samples as described below.Co-expression network modules were generated with the WGCNA . Libraries were quantile normalized and a soft threshold of 8 was used to create a scale-free network. A signed network was created choosing a soft thresholding power of 8, minModuleSize of 30, module detection sensitivity deepSplit of 2 and mergeCutHeight of 0.3. Genes with a consensus eigengene connectivity to their module eigengene of lower than 0.2 were removed from the module . Modules were correlated with upregulated genes in DCRi lines described previously.Seven days after sowing, 50–100 primary roots per sample of length ~3 cm from the root tip were cut and placed in a 35-mm-diameter dish containing a 70 µm cell strainer and 4.5 ml enzyme solution , 20 mM KCl, 10 mM CaCl2, 0.1% bovine serum albumin and 0.000194% mercaptoethanol. Cellulase Onozuka R10, Cellulase Onozuka RS and Macerozyme R10 were obtained from Yakoult Pharmaceutical. Pectolyase was obtained from Sigma-Aldrich . After digestion at 25 °C for 2 h at 85 r.p.m. on an orbital shaker with occasional stirring, the cell solution was filtered twice through 40 µm cell strainers and centrifuged for 5 min at 500g in a swinging bucket centrifuge with the acceleration set to minimal. Subsequently, the pellet was resuspended with 1 ml washing solution , 20 mM KCl, 10 mM CaCl2, 0.1% bovine serum albumin and 0.000194% mercaptoethanol and centrifuged for 3 min at 500g. The pellet was resuspended with 1 ml of washing solution and transferred to a 1.7 ml microcentrifuge tube. Samples were centrifuged for 3 min at 500 × g and resuspended to a final concentration of ~1,000 cells per µl. The protoplast suspension was then loaded onto microfluidic chips with v3 chemistry to capture 10,000 cells per sample.
Cells were barcoded with a Chromium Controller . Messenger RNA was reverse transcribed and Illumina libraries were constructed for sequencing with reagents from a 3’ Gene Expression v3 kit according to manufacturer instructions. Sequencing was performed with a NovaSeq 6000 .A trajectory analysis was run for the ground tissue cells after selecting and re-clustering the cell types annotated as exodermis and meristematic zone . Gene expression matrices, dimensionality reduction and clustering were imported into the dynverse wrapper from Seurat and a starting cell was decided within the meristematic zone cluster. Trajectory inference was run using the minimum spanning tree algorithm. The MST method and UMAP coordinates from Seurat were used as input for mclust. Predictive genes or genes that were differentially expressed along the trajectory, specific branches and milestones were identified and visualized with a heat map using dynfeature within the R package dynverse.For sections, roots were divided in 1-cm segments, embedded in 4% agarose and sliced in 120-µm sections using a vibratome. Sections were then incubated in FY088 for 1 h at room temperature in darkness, rinsed three times with water and counterstained with aniline blue for 1 h in darkness. Confocal laser scanning microscopy was performed on a Zeiss Observer Z1 confocal microscope with the ×20 objective and GFP filter . For whole roots, suberin was observed in 7-day-old S. lycopersicum wild-type or mutant seedlings. Whole roots were incubated in methanol for 3 days, changing the methanol daily. Once cleared, roots were incubated in fluorol yellow 088 for 1 h at room temperature in the dark, rinsed three times with methanol and counter stained with aniline blue for 1 h at room temerature in the dark. Roots were mounted and observed with the EVOS cell imaging system using the GFP filter . Root sections were also stained with basic fuchsin . 1 cm segments from the root tip were embedded in 3% agarose and sectioned at 150–200 µM using a vibratome . The sections were stained in Clearsee with basic fuchsin for 30 min and then washed two times and imaged with a Zeiss LSM700 confocal microscope with the ×20 objective; basic fuchsin: 550–561 nm excitation and 570–650 nm detection. Hairy roots of SlASFT transcriptional fusions were imaged with the same confocal and objective, but with excitation at 488 nm and emission at 493–550 nm for GFP, and excitation at 555 nm and emission at 560–800 nm for red fluorescent protein autofluorescence.An average of 80 mg fresh weight root tissue per biological replicate was washed and immediately placed in a 2:1 solution of chloroform:methanol. Subsequently, root samples were extracted in a Soxhlet extractor for 8 h, first with CHCl3, afterwards with methanol to remove all soluble lipids. The delipidated tissues were dried in a desiccator over silica gel and weighed. Suberin monomers were released using boron trifluoride in methanol at 70 °C overnight. Dotriacontane was added to each sample as an internal standard, saturated NaHCO3 was used to stop the transesterification reaction, and monomers were extracted with CHCl3. The CHCl3 fraction was washed with water and residual water removed using Na2SO4. The CHCl3 fraction was then concentrated down to ~50 µl and derivatized with N ,N-bis-trimethylsilyltrifluoroacetamide and pyridine at 70 °C for 40 min. Compounds were separated using gas chromatography and detected using a flame ionization detector as previously described. Compound identification was accomplished using an identical gas chromatography system paired with a mass spectroscopy selective detector . Compounds were identified by their characteristic fragmentation spectra pattern with reference to an internal library of common suberin monomers and the NIST database.Tomato roots were fixed in 2.5% glutaraldehyde solution in phosphate buffer for 1 h at room temperature and subsequently fixed in a fresh mixture of osmium tetroxide with 1.5% potassium ferrocyanide in PB buffer for 1 h. The samples were then washed twice in distilled water and dehydrated in acetone solution in a concentration gradient . This was followed by infiltration in LR White resin in a concentration gradient and finally polymerized for 48 h at 60 °C in an oven in atmospheric nitrogen. Ultrathin sections were cut transversely at 2, 5 and 8 mm from the root tip, the middle of the root and 1 mm below the hypocotyl–root junction using a Leica Ultracut UC7 , picked up on a copper slot grid 2 × 1 mm and coated with a polystyrene film .