Partial defoliation of Brachypodium distachyon plants grown in petri dishes under low light increases P and other nutrient levels concomitantly with transcriptional changes in the roots

Plant growth conditions

The grass species B. distachyon was used in this study. Seeds were surface sterilized by 75% ethanol for 5 min, then by 10% NaClO for 30 min, and were washed three times by sterilized distilled water. The seeds were then subjected to imbibition for another 12 h before sown in sterilized transparent plastic boxes (5 × 10 × 8 cm), which contain 200 mL 1/2-strength Murashige and Skoog medium (MS medium) with 1% sucrose and 0.65% agar, pH adjusted to 5.7. A total of 20 seeds were sown in each box. Seedlings in the boxes were grown in a plant growth chamber at 23 °C, 55% humidity, and 75 μmol·m²·s⁻¹ light intensity under a 16-h light/8-h dark photoperiod.

Leaf-cutting treatment and sample harvesting

For leaf-cutting assay, 2-week-old seedlings grown in each box were divided into two halves, one for treated and the other for control. Pruned seedlings were cut by scissors and 1/3 stub was left for each plant (Fig. S12). The remaining half in the same box was used as control plants. After treatment, plants were allowed to continue growth as indicated in the Figures and Tables before being harvested for further analyses.

ICP-MS analysis of ion contents

A total of 2-week-old B. distachyon plants were pruned as described above. Plant shoots and roots were separately collected, and then washed with deionized double-distilled water (DD H₂O) for four times and dried at 70 °C in an oven for 3 days. After cooling, 5–10 mg dried plant samples were digested with five mL nitric acids at 140 °C in a digestion block for 1.5 h. The digests were diluted 10-folds with sterile DD H₂O and then subjected to element measurements including Na, Mg, P, K, Ca, Mn, and Fe by using ICP-MS (NexION 300D; PerkinElmer, Waltham, MA, USA). Ion concentrations were calculated based on sample dry weights. Blank samples were run before and after the plant tissue samples to ensure the cleanness of the detection system. Indium as an standard reference materials was used and the recovery rate was between 99% and 105%. Values are mean ± SD, n ≥ 3 biological replicates. Asterisks indicate statistical significance, student t-test, p < 0.05.

Gene expression profiling by RNAseq

A total of 2 days after treatment (DAT), Trizol reagent (15596; Ambion, Suwanee, GA, USA) was used to isolate total RNA from roots of intact and pruned plants. RNA was treated with DNase (Turbo) to remove any genomic DNA contamination. Each sample has three biological replicates. Library construction and deep sequencing were performed by the Core Facility of Genomics in Shanghai Center for Plant Stress Biology, China. The sequence data have been deposited in NCBI’s Gene Expression Omnibus (GEO) and are accessible through GEO Series accession number GSE101500.

The libraries were sequenced on Illumina Genome Analyzer II and 150 bp paired-end reads (R1 and R2) were produced. We used FastQC (Andrews, 2010) to perform quality control checks and found the R2 reads in two treatment samples (T-2 and T-3) were of low quality (Fig. S4) while both R1 and R2 reads from other samples were of excellent quality, therefore we only used R1 reads for T-2 and T-3 samples and used paired-end reads for other samples to do subsequent analysis. We used SolexaQA and cutadapt to remove low quality regions and adapter sequences from the raw reads and the resulted clean reads with length larger than 25 nt and a Phred quality score larger than 17 were aligned to the genome of B. distachyon from phytozome v11 (International Brachypodium Initiative) using Tophat2.01 with default parameters (Chen, Lun & Smyth, 2014). Two mismatches were allowed for each read and both ends with reasonable distance were considered a good mapping result when using paired-end reads as input. The rate of unique mapping all reached to 80% in all samples (Fig. S10). The uniquely mapped reads were selected for gene counting. The correlation coefficients between replicates of each sample are more than 99% (Fig. S11) and the replicates are clustered in the PCA graph (Fig. S4). Differential expression analysis was performed using edgeR 3.8.6 (Kim et al., 2013). Differentially Expressed genes (DEGs) are defined as fold change ≥ 1.5 and FDR ≤ 0.05. DEGs were conducted with gene ontology (GO) enrichment analysis (Alexa, Rahnenfuhrer & Lengauer, 2006). GO terms with p-value ≤ 0.001 were selected as significant enriched. Orthologs identification was performed by using the web server viz. g:profiler (Reimand et al., 2016). TF target gene identification was performed by using the TF enrichment function of the web server-Plant Transcriptional Regulatory Map (Jin et al., 2017).

Rhizosphere acidification detection

Rhizosphere pH change in the medium was determined by a pH indicator visualization assay (Yi & Guerinot, 1996). The medium contains 1/2-strength MS medium with 1% sucrose and 0.7% agar, 0.2 mM CaSO₄ and the pH indicator bromocresol purple (0.006%). The pH of the medium was adjusted to 6.5 with NaOH. B. distachyon plants were initially grown in medium without pH indicator and were treated as described above. At 1, 2, and 3 days after the cutting treatments, plants were transferred to the medium with pH indicator for 24 h before taking pictures. Quantification of rhizosphere acidification was performed according to Yao & Byrne (2001).

Quantitative real time PCR analysis

Total RNA (50–100 mg) was isolated from root and shoot of B. distachyon using RNAzol^®RT (MRC, Cincinnati, OH, USA) following manufacturer’s instruction. RNA sample were quantified using NanoDrop 2000 (Thermo Scientific Inc., Waltham, MA, USA) and stored at −80 °C till further use. Total RNA (1.0 μg) was converted to ss-cDNA using Transcript one-step gDNA removal and cDNA synthesis Super mix (Transgen, Beijing, China) as per manufacturer’s protocol. qRT-PCR reactions were performed in BIO-RAD C1000 Touch™ Thermal cycler system using iTaq™ universal SYBR^® green Supermix qPCR Kit (BIO-RAD, Hercules, CA, USA). Total cDNA was diluted to ∼25 ng/μL and a total 100 ng was used in a 20 μL reaction mixture. For each reaction, three technical replicates were used along with no template control to check for contaminants. The following thermal cycling programme was used for all qRT-PCR reactions: 3 min at 95 °C, 3 s at 95 °C and 45 s at 60 °C for 40 cycles, which includes data acquisition. Finally, a dissociation curve analysis was performed from 65 to 95 °C in increments of 0.5 °C, each lasting for 5 s, to confirm the presence of a specific product. Concentration of ACT2 (At3g18780) was used to normalize the gene expression in different samples. Log fold change in expression values were calculated using the 2^−ΔΔCT method (Livak & Schmittgen, 2001).

Cutting aerial portions alters nutrient uptake patterns in B. distachyon

To examine the impact of partial defoliation on nutrient uptake in B. distachyon, we performed ICP-MS to measure the levels of a group of macronutrients which including K, P, Ca Mg, and micronutrients including Na, Fe, Mn elements in both roots and shoots. Compared to the intact plants, roots of B. distachyon with defoliation showed 38% and 19% increases in Fe levels at 4 days and 7 DAT, respectively; whereas Fe levels in shoots were similar between the pruned and the intact plants (Fig. 1A). The elevation in Fe contents in pruned plants thus seems consistent with an increased demand of Fe for restoration of photosynthesis organ after defoliation. Similar to Fe, accumulation of Calcium (Ca) was increased in roots by cutting the aerial portion. At 4 and 7 DAT, the levels of Ca in pruned roots were 13% and 22% greater than those in intact roots (Fig. 1B). Interestingly, defoliation caused a reduction in shoot Ca levels initially at 4 DAT, while at 7 DAT, Ca levels were higher than those in intact plants.

Potassium (K) and phosphorus (P) are two major macronutrients necessary for normal plant growth and development. As expected, these two nutrient elements were detected at levels clearly higher than the other elements examined, including Ca, Fe, Mg, Mn, and Na (Fig. 1). Cutting the aerial portion also elevated P levels by 25% and 43% in the shoots and the roots, respectively, in B. distachyon at 4 DAT (Fig. 1D). The pattern of increased P accumulation in pruned plants was maintained at least by the time of 7 DAT, although the difference between the pruned and intact plants became less significant compared to that at 4 DAT.

Our ICP-MS analyses of B. distachyon ionome also detected defoliation-induced alteration in the levels of sodium (Na), Magnesium (Mg), and Manganese (Mn) (Table 1). As shown in Fig. 1E, accumulation of Na displayed higher levels in roots compared to those in shoots. Cutting B. distachyon leaves did not affect Na accumulation in roots at either 4 or 7 DAT, however, the treatment resulted in higher levels of Na in shoots at both 4 and 7 DAT (Fig. 1E). At 4 days after cutting, shoot Mg and Mn levels seemed to be reduced in the pruned plants, albeit no statistical difference was observed when comparing the pruned with the intact samples; in contrast, the cutting treatment resulted in altered levels of both Mg and Mn with statistical significance at 7 DAT (Figs. 1F and 1G).

Partial defoliation strongly impacts B. distachyon root transcriptome

To better understand how leaf-cutting affects the below-ground portion of B. distachyon at transcriptional level, we profiled root gene expression patterns on a whole-genome scale by performing RNAseq analyses. The transcriptome of intact roots and their pruned counterparts were compared at 2 days after the cutting treatment. A total of 1,268 DEGs were identified with ≥ 2-fold change, demonstrating a strong impact on the roots by cutting the aerial portion. Among all the DEGs, 731 and 537 DEGs are upregulated and downregulated (Fig. S1A). Although the genome of B. distachyon has been reported (International Brachypodium Initiative, 2010), detailed information of gene function is still largely limited. In the pruned B. distachyon, 55% of the DEGs showed GO annotations (Fig. S1B). Subsequently, all the genes were subjected to GO ontology assignment against the B. distachyon genome database (available at GO enrichment database of PlantRegMap), in order to classify them by various biological pathways. Three independent ontological classes were used for classifying the DEGs including biological processes, molecular function, and cellular components (Ashburner et al., 2000). DEGs that had GO annotation can be further classified into 110 significant GO terms (p < 0.01) with 72 biological processes, 33 molecular function, and five cellular components (Table S1). The top 20 biological processes GO terms and the top 20 molecular function GO terms are shown in Fig. S1, together with the GO terms of five cellular components.

Consistent with the elevated iron accumulation in pruned B. distachyon (Fig. 1A), the GO term “cellular response to iron ion starvation” was identified (Table S1). The enriched GO terms for ion homeostasis and binding also include “potassium ion binding,” “magnesium ion binding,” and “alkali metal ion binding,” which can be correlated with the altered accumulation of these two metal ions and Na⁺ as observed by ICP-MS (Fig. 1).

Partial defoliation induces transcriptional alterations on plant ion homeostasis, hormones, primary and secondary metabolites, and wound stress

All the GO terms obtained were further subjected to removal of redundancy by using REVIGO (Supek et al., 2011). The REVIGO analysis for 110 GO terms clearly grouped them into six major clusters (Fig. 2), which reveals that partial defoliation leads to differential expression of genes pertaining to stress and TF (Fig. 2A), hormone mediated ion homeostasis (Fig. 2B), secondary metabolism and transportation of its metabolites (Figs. 2C and 2D), cell organization (Fig. 2E) and primary metabolism (Fig. 2F) in B. distacyon. Moreover, these DEGs may further be classified into different gene functional groups based on the shared biological function (Table S2), with the top three most shared function being transcription regulation (60 DEGs), stress response (48 DEGs), and signal transduction (43 DEGs). Together these GO patterns display a complex regulatory network at transcriptional level underlying B. distachyon response to the partial defoliation treatment.

Among the DEGs, a group of 48 genes categorized as related to “response to hormone,” in addition to seven DEGs as related to “hormone transport” (Table S1; Fig. S2A). Among these DEGs, most of the gibberellin related DEGs were up-regulated (Table S3). This pattern may suggest that gibberellin-mediated plant growth is a key process during B. distacyon’s responses to partial defoliation, since gibberellin is known for its essential role in plant growth such as stem elongation, leaf expansion and trichome development (Davière & Achard, 2013). The involvement of phytohormones in B. distacyon’s responses to partial defoliation is also indicated by a subset of DEGs related to phytohormone-mediated wounding responses; these DEGs are associated with auxin (Bradi1g06670, Bradi4g21220), abscisic acid (Bradi4g24120, Bradi5g18830), ET (Table S3), JA (12 DEGs), and JA-mediated signaling (Bradi1g72610, Bradi3g23180, Bradi3g23190, Bradi5g08650).

In addition to showing altered phytohormone gene expression, plants pruned by partial defoliation also displayed transcriptional regulation on DEGs related to homeostasis of secondary metabolites such as terpenoids and isoprenoids (Table S1; Fig. S2B). Plant primary metabolism may also be affected by partial defoliation, because the over-represented DEGs include those related to carbohydrate metabolic process, fructose, sucrose, di and oligo-saccharide metabolism, glycolytic process, pyruvate metabolic process, and glucose import and transport (Table S1; Fig. S2C). Particularly, sucrose synthase activity is the GO term that shows lowest p-value among all GO terms (Table S1). It is also possible that these membrane- or vacuole-related GO terms reflect in planta ion transportation and compartmentalization. Together with the ion accumulation patterns, the overrepresentation patterns of DEGs related to membrane and vacuole, to different transporters, and to ion homeostasis and ion uptake (Table S1; Fig. S2D), collectively suggest that plant responses to partial defoliation include not only elevated uptake of certain nutrients and altered primary and secondary metabolism, but also possibly cellular re-distribution of secondary metabolites and nutrient ions.

According to the InterPro classification, the enriched protein families belong to transporters (IPR020846, IPR013525, and IPR000109) and membrane specific proteins (IPR006016, IPR008004, and IPR005516). These patterns further support the hypothesis of plant transcriptional regulation for elevated uptake of nutrition from the rhizosphere. The protein families belonging to sugar binding (PF14416) and synthesis (IPR000368) reveal changes in energy homeostasis in mowed grasses, while protein families of different TF (like myb, bHLH bBox type, etc), and stress-related proteins (Fig. S3; Table 2) provide further mechanistic clues for plant responses to partial defoliation.

KEGG analysis revealed alteration in wound-responsive metabolism

The integration of genomic, biochemical and systemic function information can be found at KEGG (Kyoto Encyclopedia of Genes and Genomes), which is a biological system and resource management database (Kanehisa et al., 2012). We retrieved UNIPROT protein IDs for the DEGs identified in defoliated B. distachyon by using the PANTHER web server (Mi, Muruganujan & Thomas, 2013). Subsequently these protein IDs were screened for KEGG pathway annotation; the corresponding significance levels were calculated by using hypergeometric test/Fisher’s exact test, while FDR correction was achieved by using Benjamini & Hochberg (1995) method provided by the KOBAS web server (Mao et al., 2005). A total of 16 significant KEGG pathway categories were identified, with the criterion that at least four DEGs were represented (Fig. S4). Genes corresponding to biosynthesis of secondary metabolites (KEGG ID: 1110), glycolysis (KEGG ID: 10), and metabolic pathways (KEGG ID: 1100) were overrepresented in the transcriptome dataset (Table 2; Table S4). These results are consistent with wounding-induced transcriptional activation of several plant signaling cascades including oxidative phosphorylation, plant hormone signal transduction, protein processing in ER, and shikimic acid metabolism, presumably to activate the response to wounding stress and to prepare the plants for secondary metabolite synthesis and for mobilization of energy (Table S4).

Overrepresentation of a group of transcription factors in DEGs

Gene regulation by TFs is common and important for cellular response to biotic and abiotic stress conditions. To better understand the molecular mechanism underlying B. distachyon response to defoliation, we searched for TFs among the DEGs and identified their gene families according to the data base Plant TFDB (Guo et al., 2008). Out of the 1,263 DEGs (≥2-fold change), 95 are TF belonging to 23 TF family, with MYB, bHLH, C2H2, WRKY, MYB-related, ERF, NAC, and trihelix families containing more than five DEGs in the transcriptome (Fig. S5). Further, the 1,263 DEGs (≥2-fold change) were subjected to transcription factor enrichment analysis in Plant TFDB, which finds TFs with significant over-represented targets in the input genes. The results revealed 36 families of 236 TFs, with MYB, NAC, ERF, bZIP, and WRKY as the top five most abundant TF families (Fig. S6), targeting 1,213 DEGs of mowed grass transcriptome. From this analysis it is inferred that TFs of MYB, bHLH, C2H2, WRKY, ERF, and NAC domain are major players of partial defoliation- induced plant response. The expression of TF genes such as MYB, WRKY, and AP2-domain family has been reported in response to wounding (Cheong et al., 2002), while partial defoliation of grasses leads to differential expression of TFs regulating hormonal homeostasis (such MYB, bHLH, ERF, WOX, and HD-ZIP), of ion-binding TFs (C2H2, DBB, C3H, FAR1, Co-like, and ZF-HD), as well as of tissue specific (Myb related and LBD) and other transcription regulating TFs (Table S5). Therefore, compared to wounding, partial defoliation displayed broader impacts on plant transcriptional regulation of TFs and probably their downstream targets.

In order to find out partial defoliation-triggered B. distachyon DEGs that are conserved among different species, the orthologs for these DEGs were searched in Arabidopsis thaliana. Subsequently, the hormone-responsive genes identified in GO analysis study were subjected to clustering (Fig. 3). Most of the genes showed one to one clustering with its counterpart in A. thaliana, such as TFs (like MYB, WRKY, bHLH, and homeobox domain), kinases, synthesase, and other enzymes. As can be seen in Fig. 3 and Fig. S7A, the JA related proteins (wound stress responsive protein) from two species clustered separately, revealing the divergence of grass proteins from their counterpart in Arabidopsis. This pattern may indicate a difference in stress adaptability between B. distachyon and Arabidopsis because JA is known to mediate plant stress response and because grasses are more often subjected to wounding than Arabidopsis. Similarly, some extent of divergence can be observed among genes from two species involved in ion homeostasis (Fig. S7B).

Cutting the aerial portion mildly decreases B. distachyon root rhizosphere acidification

Soil pH at the vicinity of roots, that is, rhizosphere, is an important determinant of the mobility of many nutrient elements. Roots releases proton (H⁺), in exchange of cations, through root plasma membrane H⁺-ATPase. Our RNAseq analysis identified three root ATPase-related DEGs (Table S3). To examine the effect of partial defoliation on root rhizosphere acidification, both pruned and intact plants were transferred to new growth medium with the pH indicator bromocresol purple, which exhibits a yellowish color when pH is below 5.2. As shown in Fig. 4, B. distachyon roots were surrounded by yellowish color independently of the partial defoliation treatment, although the pruned plants displayed a weaker intensity of the yellowish color in the rhizosphere compared to the intact plants. Such a mild reduction in rhizosphere acidity was similarly observed with B. distachyon plants at 1, 2, and 3 DAT (Fig. 4; Fig. S8), indicating the partial defoliation treatment quickly and continuously affected the capacity of roots in releasing H⁺ into the rhizosphere. It would be interesting to determine whether such a fine tuning of rhizosphere acidity contributes to the altered nutrient uptake in B. distachyon after partial defoliation.

Partial defoliation regulates B. distachyon gene expression similarly in shoots as in root

In order to compare transcription patterns between B. distachyon roots and shoots, we extended gene expression analysis by doing quantitative real-time PCR using both roots and shoots. Plant samples were collected at 24 and 48 h after partial defoliation treatment to disclose early transcriptional regulation. A group of eight genes were examined including Bradi5g02760 (probable inorganic phosphate transporter 1–4), Bradi3g56330 (sodium-coupled neutral amino acid transporter 1), Bradi2g42550 (metal ion binding protein), Bradi5g12020 (sugar transport protein 5-like), BD3g13370 (pyruvate decarboxylase 2), BD5g08650 (aba receptor 8), BD1g34250 (two-component response regulator orr3 isoform x4), and BD3g38140 (acc oxidase). As expected, all the examined genes displayed similar patterns as observed in RNAseq (Fig. 5; Table S6). Besides, partial defoliation-triggered gene expression patterns in roots were similarly observed in shoots in general (Fig. 5; Table S6), indicating that root RNAseq results are representative of the whole B. distachyon plants, in addition to providing organ-specific accuracy of transcriptional regulation patterns. Consistent with elevated phosphorus levels in roots and shoots (Fig. 1), partial defoliation increased gene expression of Bradi5g02760, which is a putative transporter of inorganic phosphate, at both 24 and 48 h after the treatment and in both roots and shoots (Fig. 5). Therefore these observations further support the conclusion that, during recovery from partial defoliation, B. distachyon plants elevate root uptake of nutrient elements such as Fe and P, at least partially, through transcriptional regulation.