The unprecedented rise in atmospheric CO2 concentration and injudicious fertilization or heterogeneous distribution of Mg in the soil warrant further research to understand the synergistic and holistic mechanisms involved in the plant growth regulation. This study investigated the influence of elevated CO2 (800 μL L−1) on physiological and transcriptomic profiles in Arabidopsis cultured in hydroponic media treated with 1 μM (low), 1000 μM (normal) and 10000 μM (high) Mg2+. Following 7-d treatment, elevated CO2 increased the shoot growth and chlorophyll content under both low and normal Mg supply, whereas root growth was improved exclusively under normal Mg nutrition. Notably, the effect of elevated CO2 on mineral homeostasis in both shoots and roots was less than that of Mg supply. Irrespective of CO2 treatment, high Mg increased number of young leaf but decreased root growth and absorption of P, K, Ca, Fe and Mn whereas low Mg increased the concentration of P, K, Ca and Fe in leaves. Transcriptomics results showed that elevated CO2 decreased the expression of genes related to cell redox homeostasis, cadmium response, and lipid localization, but enhanced signal transduction, protein phosphorylation, NBS-LRR disease resistance proteins and subsequently programmed cell death in low-Mg shoots. By comparison, elevated CO2 enhanced the response of lipid localization (mainly LTP transfer protein/protease inhibitor), endomembrane system, heme binding and cell wall modification in high-Mg roots. Some of these transcriptomic results are substantially in accordance with our physiological and/or biochemical analysis. The present findings broaden our current understanding on the interactive effect of elevated CO2 and Mg levels in the Arabidopsis, which may help to design the novel metabolic engineering strategies to cope with Mg deficiency/excess in crops under elevated CO2.
Citation: Niu Y, Ahammed GJ, Tang C, Guo L, Yu J (2016) Physiological and Transcriptome Responses to Combinations of Elevated CO2 and Magnesium in Arabidopsis thaliana. PLoS ONE 11(2): e0149301. https://doi.org/10.1371/journal.pone.0149301
Editor: Jin-Gui Chen, Oak Ridge National Laboratory, UNITED STATES
Received: November 28, 2015; Accepted: January 29, 2016; Published: February 16, 2016
Copyright: © 2016 Niu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was supported by grants of the China Postdoctoral Science Foundation (517000-X91403, http://jj.chinapostdoctor.org.cn/V1/Program1/Default.aspx), the National Natural Science Foundation of China (31550110201, http://www.nsfc.gov.cn/), the Zhejiang Provincial priority projects in forestry sciences (KF201310 and KF201311, http://kjc2.zafu.edu.cn/). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Plants may simultaneously adapted to two or more stresses in their natural ecosystems. Numerous physiological studies have clearly indicated that plant responses to combinations of abiotic stresses represent new modes of stress responses [1–3]. Therefore, it is important to understand the in-depth mechanisms of multiple stress responses by comparing data on single stress with that of combined stresses in order to develop enhanced stress-tolerance strategies in plants. Such studies relevant to agronomically important traits  may answer basic questions on signaling “cross-talk” in systems biology .
In the past century, the concentration of atmospheric CO2 has increased from 318 to 392 ppm; and it is anticipated to exceed 1,000 ppm by the end of the 21st century (cdiac.ornl.gov/pns/current_ghg.html). Nevertheless, increasing concentrations of CO2 may have major impacts on plant growth and development . Elevated CO2 can be initially beneficial for plant growth [7–11]; however, plant responses to elevated CO2 often depend on the availability of soil nutrients [12–14]. Therefore, changes in the nutritional status may greatly influence the results in controlled conditions particularly under elevated CO2 which affect plant metabolism and growth.
Excluding H, C and O, magnesium is the fourth most abundant element in plants, after N, K and Ca, Mg is the 8th most abundant mineral element on earth . The level of Mg in the soil is basically maintained by natural genesis and/or fertilization practice and thus both deficiency and excess of Mg should be taken into consideration during developing management strategies. Abnormal Mg status in soil resulting from either Mg depletion or Mg excess is generally considered negative for the growth of the plants [16–23]. Importantly, aside from its crucial role in a vast number of enzymatic reactions including nucleotide metabolism and the turnover of nucleic acids in transcription, splicing or replication, Mg has an additional prominent role as the central atom in the chlorophyll molecules of photosynthesizing organisms [24, 25]. It has been reported that photosynthesis of crop plants depends on its Mg and CO2 status in several aspects and both of those resources may have a detrimental effect on plant photosynthesis, depending on the extremity in their levels [7, 26, 27], which eventually result in abnormal or restricted growth of plants . Thus, analysis of elevated CO2 and Mg stresses in Arabidopsis and other plants revealed that elevated CO2 or Mg stress applied separately affects plant biomass, photosynthesis, respiration, root growth and nutrient homeostasis. In spite of their indispensable roles in the central process of photosynthesis and other aspects of the plant, no previous study has investigated the combined effects of elevated CO2 and Mg status on plant performance.
In the present study, elevated CO2 and Mg availability were selected as the combined target perturbations to unveil possible physiological and molecular alterations in Arabidopsis by transcriptome sequencing. Our study brings new insights into interactive effects of elevated CO2 and Mg availability at molecular levels.
Materials and Methods
Plant materials and growth conditions
The seeds of Arabidopsis wild ecotype line (Col-0) were obtained from the Nottingham Arabidopsis Stock Centre (http://nasc.nott.ac.uk). All plants were grown in the controlled growth room under a 10 h light/14 h dark photoperiod at constant temperature of 22°C, 80% relative humidity and light intensity of 120 μmol photons m−2 s−1 as previously described by study of .
Experimental design and setup
The present study was designed with three levels of Mg2+ (supplied as MgSO4) and two concentrations of CO2 in various combinations. Initially, seeds of Arabidopsis thaliana genotype Col-0, were surface-sterilized, and germinated on nylon net with proper porosity floating on half-strength Hogland’s nutrient solution. The composition of the culture solution as previously described by study of . Altogether, seedlings were cultured hydroponically for 5 weeks (including germination period) before imposition of CO2 and Mg treatments. As CO2 treatments, plants were exposed to either ambient CO2 (350 ± 50 μL L−1) or elevated CO2 (800 ± 50 μL L−1) conditions. For each CO2 condition, plants were divided into three groups: one supplied with 1,000 μM Mg (Control Mg, C), one supplied with 1 μM Mg (Low Mg, L) and one supplied with 10,000 μM Mg (High Mg, H). Concentrations of Mg in the medium were adjusted by manipulating the concentration of MgSO4. The solution pH was adjusted to 6.0. The treatment solutions were renewed every 2 d. Ambient CO2 + control Mg (AC), ambient CO2 + low Mg (AL), ambient CO2 + high Mg (AH), elevated CO2 + control Mg (EC), elevated CO2 + low Mg (EL), elevated CO2 + high Mg (EH). Simultaneous treatments of CO2 and Mg lasted for 7 days before termination of the experiment.
Growth parameters were determined in 8 plants per treatment. For the growth analysis, plants were photographed vertically after the 7-d treatments with a high resolution digital camera (Sony RX100, Japan). Then, plants were divided into shoots, roots, each of which was weighed on a precision balance. Both shoots and roots were recorded and quantified for rosette area and leaf area using the public domain image analysis program Image J version 1.43 (http://rsb.info.nih.gov/ij/). The scale was set for the picture within the program. Digital images were captured and processed using Image J. scanned at 300 dpi resolution for measurement of rosette diameter. Leaf number was determined by counting the number of true leaves (>1 mm long leaf blades) per plant.
Chlorophyll concentration in leaves
Leaf chlorophyll content was determined by using a portable chlorophyll meter (SPAD-502, Minolta, Japan) . Fully expanded leaves were randomly selected from three positions that corresponded to the old, mature, and young parts under different treatments. For each leaf position, 2 SPAD values were randomly collected avoiding main veins during measurement.
Root growth and morphological analysis
After 7-d treatment, root was scanned to analyze the root morphology (total root length and total number of root tips) using an automatic root scanner (STD1600, Seiko Epson Corp., Japan), and analyzed using the WinRHIZO image analysis software (Regent Instruments, WinRHIZO-EC, Canada). The root-to-shoot ratio was calculated from dry weights of roots and shoots. Additionally, root hairs in 3-cm apical root segments were imaged through a light microscopy with differential interference contrast optics. Micrographs were recorded using a CCD camera (Nikon Eclipse E600).
Analysis of elemental composition in plant tissue
After 7-d treatment, plants were harvested, washed thoroughly with deionized water, divided into shoots and roots, and dried in an oven at 75°C for 12 h. The samples were then weighed, digested in sulfuric acid/hydrogen peroxide, and analyzed for total P concentration using the vanadium-molybdenum-blue photometric method. For other elements, the dried root and shoot samples were wet-digested in the concentrated HNO3/H2O2 at 90, 120 and 140°C for 2 h, respectively, and then further digested at 180°C until the digest became clear as described by . Concentrations of K, Mg, Ca, Fe, S, Mn and Na in the digests were analyzed by ICP-MS (Inductively coupled plasma mass spectrometer, Agilent 7500a, USA). The concentration of nutrients was calculated on a dry-weight (DW) basis.
Tissue collection and RNA isolation
For each condition, two samples were collected, each of which was pooled from eight independent plants at the same growth stage. The sampling was done at the start of the light period of treatment (day-7). Total RNA was isolated from shoots and roots using the RNeasy mini kit (QIAGEN, Germantown, MD, USA) with an additional DNase I (QIAGEN) digestion step to remove any genomic DNA contamination. The concentration of the purified RNA was determined by a Qubit2.0 fluorometer (Invitrogen, Carlsbad, CA, USA). RNA integrity was assessed by the Agilent Technologies 2100 Bioanalyzer.
1 μg of total RNA from each sample (n = 2 per treatment) was collected for RNA-Seq library construction and sequencing. cDNA library was constructed using the TruSeq RNA Sample Prep Kit (Illumina, San Diego, CA) according to the manufacturer's instructions after mRNA purification and fragmentation. The samples were then clustered and sequenced on an Illumina HiSeq 2500. Deep sequencing was performed with two replicates for each treatment (twenty-four samples in total) for a 151 cycle pair end run.
RNA-Seq data analysis
RNA-Seq reads were assessed for quality control with FastQC (version 0.10.1; Babraham Bioinformatics, Cambridge, UK). Reads were mapped to a reference Arabidopsis genome (TAIR10, http://www.Arabidopsis.org) using TopHat with parameter (-I 30000) (version 2.08; ). The gene abundance values were measured as fragments per kilobase of exon per million fragments per kilobase of exon per million fragments (FPKM) mapped by Cufflinks 2.1.1 . Cuffdiff  was then used to determine differential expression (FDR ≤ 0.05). In present study, approximately 6 million pair-end reads from all 24 libraries were trimmed with Sickle and mapped to the Arabidopsis TAIR10 genome reference sequences representing 74% transcripts.
Gene function and pathway analysis
The list of differentially expressed genes generated from Cuffdiff was imported into agriGO tools for Gene Ontoglogy enrichment analysis (http://bioinfo.cau.edu.cn/agriGO/). Biological function and KEGG pathways were determined to be over-represented using the Fisher exact test with a false discovery rate (FDR) correction (FDR≤0.05).
Extraction of total RNA and quantitative real time PCR (qPCR)
Total RNA was extracted by RNAisoPlus (Takara, Otsu, Shiga, Japan) from about 50 mg of fresh root tissues. Four independent biological replicates were performed on independent root material from different plants. All RNA samples were checked for DNA contamination before cDNA synthesis. cDNA was synthesized, and possible residual genomic DNA contamination was verified as described in our previous study . The mRNA levels of all genes were detected by the Mix SYBR Green RT-PCR kit (Takara, Otsu, Shiga, Japan) with following pairs of gene-specific primers. UBQ10 was chosen as the housekeeping reference according to .
A control Mg concentration (1,000 μM) for Arabidopsis growth was selected according to the study of [33, 34]. Specially, 10,000 μM MgSO4 was selected as a maxmium-level Mg concentration in the present experiment based on our previous study by  which showed that Arabidopsis plants could sustain in this stress for seven days before severe detrimental effects becoming apparent. Thus, it was certain that within the timeframe of the present study, any observed physiological changes would reflect the response of the plants to the applied treatments without any interference from cellular death effects. Meanwhile, at the harvest, some growth parameters differed among the treatments. For the above reasons, three representative concentration of Mg were chosen to study the interaction with elevated CO2 in analysis of transcriptional and metabolic physiology of Arabidopsis (Fig 1).
Conceptual model (A) and photos of shoots (B) and images of roots (C) of five-week-old wild-type Arabidopsis grown for 7 d in low, control and high Mg nutrient solution under ambient or elevated CO2. Each arrow corresponds to a treatment; each node corresponds to a physiological state.
Morphological and physiological characterization
Elevated CO2 enhanced shoot growth under low Mg and normal Mg supply, whereas it enhanced root growth under exclusively normal Mg supply (Fig 2). Specifically, the fresh weight, rosette area, leaf area and chlorophyll content (SPAD value) in shoots were higher in EL and EC treatments whereas lower in AH and EH treatments than in the control treatment (Fig 2A and 2C–2E), suggesting that elevated CO2 could enhance the growth and chlorophyll content under both low Mg and normal Mg supply. By contrast, compared with the control, number of young leaf was increased in AH and EH treatments but not altered by AL or EL treatment (Fig 2B).
Means with a same letter within a measurement are not significantly different at P≤0.05.
The root fresh weight, total root length, number of root tips per plant, density of root hairs and root/shoot ratio were all greater under EC than under AC condition (Fig 2F–2J), which is in accordance with the results of other studies [9–11]. By comparison, root fresh weight, total root length, number of root tips per plant were increased in the AL and EL treatments but decreased in AH and EH where the decrease was greater in EH than in AH (Fig 2F–2H). It is worth mentioning that the root/shoot ratio was only increased in AL and EC (Fig 2I). Interestingly, the formation of root hairs was completely suppressed by high Mg (10 mM) but was partly restored by elevated CO2 (Fig 2J). Overall, these results indicated that elevated CO2 enhanced the growth of plant depending on Mg supply and the magnitude of the effect was substantially different in shoots and root tissue.
The interactive effect of elevated CO2 and Mg on mineral homeostasis
It is well documented that Mg availability affects the ionome by impacting the uptake and distribution of other cations [19, 20, 35–37]. Meanwhile, analysis of elevated CO2 in Arabidopsis and other plants revealed that elevated CO2 remarkably affects plant biomass and nutrient homeostasis depending on level of nutrient supply [9, 14, 18]. Elevated CO2 decreased the concentration of P, Mg and Fe in leaves of normal Mg-supplied plants (Fig 3A). However, with supply of low Mg or high Mg, the plants had a similar level of these elements in shoot under both ambient and elevated CO2 treatments. Elevated CO2 increased Mn concentration in shoots of the low-Mg-supplied plants but did not alter it in normal Mg or high Mg-supplied shoots. In contrast, elevated CO2 decreased the level of Zn in shoots of low Mg and normal Mg-fed plants but did not change it in shoots of plants grown in high Mg treatment. In addition, high Mg decreased the concentrations of P, K, Ca, Fe and Mn while low Mg enhanced the concentrations of P, K, Ca and Fe in a given CO2 treatment (Fig 3A).
Values are the averages of at least five samples ± SD. Means followed by a same letter (italics) within a root segment are not significantly different at P < 0.05. DW, dry weight.
Similarly, elevated CO2 decreased P concentration in roots of the normal Mg-supplied plants but increased it in roots of the low Mg or high Mg-supplied plants. However, elevated CO2 decreased the concentration of Ca, Na and Cu roots of high Mg-fed plants but did not change that of low Mg-fed plants (Fig 3B). In addition, elevated CO2 increased the concentration of K and Mn in roots of low Mg-fed plants. Irrespective of CO2 concentration, a general response of roots to low Mg supply is the higher concentration of Ca, Na, Fe and Zn. Regardless of Mg supply, elevated CO2 did not affect the absorption of S, Fe and Zn in roots.
Transcriptome analyses under different CO2 and Mg conditions
The entire data set has been uploaded to the National Institutes of Health Gene Expression Omnibus database (GSE64501) and the raw sequence data have been deposited in the NCBI Gene Expression Omnibus (GEO). A summary of the complete transcriptomic analysis is presented in S1 Table. The Pearson correlation coefficients of sample pairs calculated with gene FPKM, were visualized using heat spectrum graphs where colors ranging from yellow to green correspond to correlation coefficients of 0.5 to 1.0, respectively (S2 Table). To focus on the differently expressed genes (DEGs) under CO2 and Mg treatments compared with the control condition, we only selected those with FDR ≤ 0.05. A total of 6345 genes showed ≥ two-fold change in expression upon CO2 and Mg treatments (|ΔS| = |log2SX-log2SCtrl| > 1, where X is the CO2 and Mg stress treatments) for at least one sample in shoots (Fig 4A) and roots (Fig 4B). 5644 and 1234 genes were identified to be differentially changed in shoots and roots, respectively. And the expression of the down- and up-regulated genes in any treatment groups is also presented in S3 Table.
Each list in the venn diagram denotes by a transparent shape and overlaps shape indicates elements shared between lists or more often the corresponding counts. Five lists input data of AL, AH, EL, EC and EH were highlighted for both shoots and root in green, blue, pink, yellow and orange accordingly. Number of the total differently regulated genes in shoots (C) and root (D) under each treatment as compared to the control (ACS and ACR, respectively). “+” indicates up-regulation, “-” indicated down-regulation. Some count values are displayed in the chart and some are outside the chart, using lines to line the count to its corresponding area. Common and specific components responded to one or multiple sets in shoots (E) and root (F). 2, 3, 4 and 5-set Venn diagram indicates the output set orders in the resultant extracted datasheet in panel.
Venn diagrams showed that there are 3,821 (AL/AC), 1,426 (AH/AC), 2,179 (EL/AC), 658 (EC/AC) and 2,222 (EH/AC) DEGs in shoots (Fig 4A). Meanwhile, as compared with the control, 1689, 311, 362, 103 and 436 DEGs was specifically respond to ‘AL’, ‘AH’, ‘EL’, ‘EC’ and ‘EH’ treatment, respectively (Fig 4A). There are 569 and 188 common differential expressed genes in shoots between ‘AL vs EL’ and ‘AH vs EH’, respectively (Fig 4A). Furthermore, the maximum and minimum DEGs were recorded in shoots of plant grown in AL and EC treatments, respectively (Fig 4C). It is suggested that low Mg had a greater while elevated CO2 had a smaller effect on transcript responses in shoots.
In roots, 241 (AL/AC), 313 (AH/AC), 240 (EL/AC), 336 (EC/AC) and 827 (EH/AC) genes altered their expression under combined treatments of CO2 and Mg (Fig 4B). As compared with the control, there were 87, 89, 56, 106 and 429 differentially expressed genes in roots under ‘AL’, ‘AH’, ‘EL’, ‘EC’ and ‘EH’, respectively (Fig 4B). In contrast with shoot, elevated CO2 had a few affect transcripts in the roots with low Mg (20% of variance) but transcripts in the roots with high Mg were more prominent under elevated CO2 (67% of variance) than under ambient CO2 (25% of variance). Moreover, there are only 6 but 98 common differential expressed genes in roots between ‘AL vs EL’ and ‘AH vs EH’, respectively (Fig 4B). The maximum and minimum DEGs were recorded in shoots of plant grown in the EH and EL treatments, respectively (Fig 4C and 4D).
There are nearly half of differently expressed genes in shoots were shared by more than two treatments (Fig 4E and 4F), revealing high overlap of transcript responses between elevated CO2 and Mg stresses treatments. We further classified all expression plots of DEGs into 20 (A-T) and 15 (A-O) clusters in shoots (Fig 5) and roots (Fig 6), respectively, according to genes expression patterns using hierarchical clustering with complete linkage. Several data adjustment procedures are available and often used prior to statistical analysis of a given data set . Gene expression is plotted on a log2 scale for each gene and details about individual genes can be found in Supporting Information S3 Table. It is noticed that clustering results of treatments showed a completely different pattern between shoots and roots. In shoots, the DEGs was clustered into three groups according to the concentrations of Mg (Fig 5) while in roots it was clustered into two groups according to CO2 treatments (Fig 6).
For each treatment, two admixture biological samples were performed on independent root material from ten different plants. Each column represents a single sample. Data adjustment procedures are employed and often used prior to statistical analysis of a given data set. The color scale ranges from saturated green for log ratios -3.0 to below saturated red for log ratios 3.0 and above. Each gene is represented by a single row of colored boxes; Cells with log ratio of 0 (genes unchanged) are colored black, increasingly positive log ratios with reds of increasing intensity, and increasing negative log ratios with greens of increasing intensity. Missing values usually appear gray. Gene expression is plotted on a log2 scale for each gene transcript and details about individual transcripts can be found in Supporting Information S3 Table.
The color scale ranges from saturated green for log ratios -3.0 to below saturated red for log ratios 3.0 and above. Each gene is represented by a single row of colored boxes; Cells with log ratio of 0 (genes unchanged) are colored black, increasingly positive log ratios with reds of increasing intensity, and increasing negative log ratios with greens of increasing intensity. Missing values usually appear gray. Gene expression is plotted on a log2 scale for each gene transcript and details about individual transcripts can be found in Supporting Information S3 Table.
Validation of gene expression profiles using qPCR
To evaluate the accuracy of expression profiles obtained from RNA-Seq, we first measured a selected set of DEGs genes by real-time qPCR, using the same samples originally used for RNA-Seq (S4 Table). We analyzed the transcript levels of nuclear-encoded genes related to photosynthesis, phosphorus-containing anhydrides, cell wall, stress response, root development and auxin pathway by real-time qPCR. This analysis revealed a close correlation between the expression changes (fold difference) measured by each method (Pearson’s correlation coefficient r = 0.90).
Remarkable responses of genes in shoots
Functional enrichment analysis using agriGO for all DEGs is able to reveal biological functions based upon DEGs . The full functional annotation analysis is provided in supplementary S5 Table. We then investigated whether transcripts of the particular response modes could be associated with biological functions via their corresponding, significant GO terms. Firstly, it is found that the top ranked biological functions in the shoots included the most important and significant differences compared with the whole genome representation were in the response to ‘stimulus or stress’, the ‘cell wall type and organization’, the ‘antioxidant and oxidoreductase activity’, the ‘cellular metabolic and multi-organism process’, ‘plastid chloroplast, chloroplast part, chloroplast thylakoid’, ‘membrane part’ and ‘response to hormone pathway’ (S5 Table). Moreover, it is revealed that the strong common existence of genes involved in various treatments of CO2 and Mg with approximately 35% and 25% of the total functional annotations categorized as being cell part and stress/ stimulus-associated, respectively. A significant number of these DEGs were affected in more than four of the treatments, and the overlapping sets of genes are over-represented for C, K, and T clusters (Fig 5). Among the differently regulated gene functions with cell part, chloroplast and thylakoid, endomembrane system and cell wall were the maximum ranked cellular functions (Fig 7).
Regulation of transcripts in this category, with sub-categories identified, for six treatments as follows: ALS, ACS, AHS, ELS, ACS and AHS. The figure shows distinct clusters of expression patterns within the group of a treatment across the six treatments. The full cluster set is shown on the left; specific gene names are expanded to the right to allow closer inspection of the differential expression patterns. Higher levels are represented with color red of increasing intensity, and lower levels are represented with blue of increasing intensity. The increased or reduced expression is relative to the mean expression. All transcript levels shown were statistically significant (FDR ≤ 0.05) increase or decrease at any one treatment.
Under normal Mg supply, the significant differences in gene abundance in shoots between ambient and elevated CO2 had mainly focused on A and H clusters, respectively (Fig 5). In addition, genes in L cluster specifically respond to interaction of elevated CO2 and low Mg. This cluster disease resistance protein mainly included a disease-resistance proteins belonging to the TIR-NBS-LRR (Toll/Interleukin1 receptor–nucleotide binding site–leucine-rich repeat) domain signatures and genes encoding proteases and virulence-responsive proteins. On the other hand, genes in Q cluster related to cadmium-ion response (GO: 0046686), cell redox homeostasis (GO: 0045454) and lipid localization was specifically up-regulated by low Mg (Fig 5 and S6 Table). Most of the features of known genes (such as ATTRX5, RPT2A, GDH2, PAG1, ADL1E, NDPK1) in response to cadmium ion are thought to be cytosolic or associated with the plasma membrane. By comparison, elevated CO2 + high Mg had a weak impact on the response of transcripts in shoots. The only interaction between them was enriched in K cluster, which included chloroplast envelope (AOC1, AOC2, COR15B, LOX2, AT3G22620 etc) and monooxygenase activity genes. GO analysis showed that all those six monooxygenase activity genes were also associated with endomembrane system (Fig 5 and S6 Table). Other cluster genes GO term, such as response to hormones and water deficit, were shared between transcripts under EL and EH treatments.
Functional enrichment in roots
The functional categories most significantly enriched in the roots compared with the whole genome representation were in the response to ‘stimulus or stress’, the ‘cell wall type and organization’, the ‘antioxidant and oxidoreductase activity’, the ‘activity of structural molecule, electron carrier, transporter, catalytic’, the ‘endomembrane system’ and the ‘external encapsulating structure structural’ (S5 Table). This is in accordance with the findings of other studies [18–20]. Among these, a common set of genes responding to ‘stress or stimulus’, ‘cell wall type’ and ‘oxidoreductase activity’ were identified in both of low and high Mg supply irrespective of CO2 concentration, and were also shared with the response to shoot. Numerous up-regulated genes in cluster M, N, and O functional enrichment including: cytokinin-mediated signaling pathway, transferase activity, cell wall macromolecule biosynthetic process, water channel activity, rhythmic process, flavonoid metabolic process were found to be specifically located in the combined treatment of elevated CO2 and high Mg supply (Fig 6 and S6 Table).
The EH treatment induced specific genes in roots mainly focused in cluster N (Fig 6). These cluster genes functioned with lipid localization, cell wall and endomembrane system, and specifically increased heme binding (S6 Table). Notably, GO terms showed that all six genes AT5G46890, AT5G46900, AT3G22570, AT4G12510, AG4G12520 and AT4G22490 related to lipid localization were found encoding protease inhibitor, a lipid transfer protein (LTP) family (Fig 6A and 6B). In contrast, low Mg supply alone elevated a number of up-regulated genes in cluster D related to ‘trichoblast differentiation, epidermis development, root morphogenesis, structural constituent of cell wall, ion homeostasis and transport’. In addition, under both low Mg and high Mg conditions, some DEGs were elevated in those involved in immune system functions, in particular, functions related to cell death, and toxin catabolic and metabolic process and defense response (Fig 8 and S6 Table).
The order of RNA-seq samples along the x-axis is as follows: ALR, ACR, AHR, ELR, ACR and AHR. The figure shows distinct clusters of expression patterns within the group of a treatment across the six treatments. The full cluster set is shown on the left; specific gene names are expanded to the right to allow closer inspection of the differential expression patterns. Red color indicates high transcript level; Blue color indicates low transcript level. The 208 genes are highlighted in green (for root-hair genes).
The objective of this work is to understand responses in plants along with interactive effect of elevated CO2 and Mg levels by comparing data on single stress with that of combined stresses. Some remarkable transcriptomic responses to individual vs. combinatorial treatments of elevated CO2 and Mg levels were highlighted here. Contrasting changes were found between roots and shoots with the shoot transcriptome being more severely affected by low Mg while the root transcriptome more affected by high Mg. Elevated CO2 had a greater effect on transcript response in low Mg-fed shoots as well as in high Mg-fed roots.
Elevated CO2 affects plant responses to Mg supply
The increasing interest has been on studying the effects of CO2 level on plant responses either phenotypically [40, 41] or transcriptomically [42, 43] and genome-scale studies of the model species Arabidopsis under Mg deficiency [19–20] and excess [21, 22]. Here, by analysis of individual vs. combinatorial effect of elevated CO2 and Mg levels on phenotypical or transcriptomic responses in Arabidopsis thaliana grown in hydroponic cultures (Fig 1), this study showed that elevated CO2 affected Arabidopsis growth and the response of transcripts to Mg supply and that the magnitude of the effect differed between shoots and roots. The resulted indicated that elevated CO2 significantly mitigated the negative impact of low Mg, as apparent in biomass reduction, photosynthesis inhibition, chlorophyll fluorescence decline (Fig 2). This seems suggested that elevated CO2-mitigation of low Mg stress associated with maintenance of positive carbon balance and carbohydrate accumulation in Arabidopsis. Besides, elevated CO2 promoted the root/shoot ratio in normal-Mg-fed plants but inhibited it in both low Mg and high Mg-fed plants (Fig 2I). CO2 exacerbated the negative effect of high Mg on root biomass and number of lateral roots but alleviated the effect on root hair density (Fig 2F and 2G). It has been reported that Mg deficiency and elevated CO2 changed sugar partitioning between shoot and roots, leading to an increase in the root: shoot biomass ratio [36, 44, 45]. Elevated CO2 enhanced both shoot and root growth of normal Mg-supplied plants (Fig 2F–2J). This is in accordance with the results of many studies [9, 11, 46]. Overall, these results indicate that elevated CO2 enhanced the growth of plant depending on Mg supply and the magnitude of the effect was similar in shoots and roots. Interestingly, leaf number was increased only in AH and EH treatments but not altered by AL or EL treatment (Fig 2B), indicating that high Mg could induce differentiation and division of leaves regardless of elevated CO2. Interestingly, phenotypes of narrow and thick leaves were observed in high Mg-fed leaves under ambient and elevated CO2. It is reported that WUSCHEL-RELATED HOMEOBOX1 (WOX1) acts for cell proliferation in the blade outgrowth and margin development downstream of adaxial/abaxial polarity establishment in Arabidopsis leaves . In present study, expression of WOX1 in leaves was greatly decreased by AH and EH treatment (S1 Table) indicating that the WOX1 are required for promoting cell proliferation in outgrowth of leaf blade under high Mg, maybe due to activated cell division.
However, the responses of transcripts coincided with measurable root parameters such as root weight and root hair number in AL and EL treatments, while the transcripts in roots treated by AH and EH were found not completely concomitant with their final morphological root phenotype (Fig 4 and S6 Table). These suggest that the molecular and physiological responses to low Mg and high Mg conditions differ not only in the number of genes or extents of expression changes, but also in the sets of genes induced.
Ionomic adjustment in response to elevated CO2 and Mg level
The current study showed that the EL increased the concentration of Mn and decreased the concentration of Zn in shoots (Fig 3A) while the EH decreased and the AH did not affect the concentration of Mn in roots (Fig 3B). These results were probably due to the competition of Mn2+ with Mg2+ for membrane transport and substitute Mg2+ for activating a number of enzymes (like ribulose-1,5-bisphosphate carboxylase/oxygenase)  because Mn2+ and Mg2+ have similar chemical properties. This study also indicates that high Mg supply had a profound impact on the ionic balance, noticeably by decreasing the concentration of macro-nutrients including P, K and Ca in both shoots and roots irrespective of CO2 treatment. Notably, the enhanced shoot and root growth and decreased concentrations of P and Mg in shoots and roots, Fe in shoots and Cu in roots of normal-Mg-supplied plants by elevated CO2 are in accordance with the findings of other studies [9, 11, 14]. The observed decrease in nutrient concentrations is most likely due to a dilution effect caused by enhanced biomass production at elevated CO2 .
Many transporters involved in metal ion homeostasis have been identified in the Arabidopsis genome [49–50]. It is worth mentioning that Mg deficiency did not induce the expression of genes encoding permeases potentially mediating Mg transport, such as the MITOCHONDRIAL RNA SPLICING2/MAGNESIUM TRANSPORTER (MRS2/MGT/CorA) family [51–53] and MAGNESIUM/PROTON EXCHANGER 1(MHX) [54–55]. One of the categories containing the highest number of Mg-regulated ion channel genes was the ‘anion channel protein family, KCO5 protein family, shaker family, glutamate receptor family’. Analysis revealed the existence of genes involved in various treatments of CO2 and Mg with approximately 80% of the ion binding functional annotations categorized as being P, S, Ca and Zn-associated.
Elevated CO2 mitigated against low Mg-induced photosynthesis-inhibition
As reported in this and previous studies [26, 28], both Mg deficiency and oversupply have detrimental effects on plant photosynthesis, and consequently resulting in restricted growth of plants. Transcripts of genes involving photosynthetic systems I (e.g. PSAN, PSAL, PSAD-1, PSAE-1, PSAG, PSAA and PSAB) and II (PSBO 1, PSBC, PSBD and PSBE) and photosystem II-Q (PSII-Q) and PSII-T subunits were significantly reduced by AL treatment (Fig 7). Moreover, the increased chlorophyll concentration by EL and EC but decreased by AL, AH and EH (Fig 2A and 2C–2E) indicate that low Mg and high Mg restrains the photosynthesis correlated with chlorophyll content. The up-regulation of photosynthesis-related genes in EL relative to AL indicates that elevated CO2 alleviated the effects of low Mg. This suggested that plant responses to combinations of elevated CO2 and low Mg stress induces a new response of photosynthetic. Mg plays a fundamental role in phloem export of photosynthates, so that Mg deficiency restricts the partitioning of dry matter between roots and shoots, which result in an accumulation of sugars, starch and amino acids in leaves, chlorophyll break-down, an over-reduction in the photosynthetic electron transport chain and the generation of excessive reactive oxygen species (ROS) impairing photosynthetic CO2 fixation [56–59]. Evidence for the involvement of the light-harvesting chlorophyll a/b-protein complex in thylakoid stacking and for effects of Mg2+ was reported by . Low Mg decreased but elevated CO2 restored the expression of genes correlated with chlorophyll a/b-protein and starch biosynthesis. These genes include SHOOT APICAL MERISTEM ARREST 1 (SHA1), ATP SYNTHASE DELTA-SUBUNIT GENE (ATPD), THIAMINC (THIC), SIGMA FACTOR 4 (SIG4), PHOTOTROPIN 2 (PHOT2), NDR1/HIN1-LIKE 12 (NHL12), HIGH CYCLIC ELECTRON FLOW 1 (HCEF1), PROTON GRADIENT REGULATION 3 (PGR3), GLYCINE DECARBOXYLASE P-PROTEIN 1 (AtGLDP1) and THIOREDOXIN F-TYPE 1 (TRXF1). Specifically, EL significantly increased the concentrations of homoserine, methionine and β-alanine.
Defense response upon a combined stresses of elevated CO2 and Mg
Interaction of elevated CO2 and low Mg specifically triggers disease-resistance proteins mainly including those belonging to the TIR-NBS-LRR (Toll/Interleukin1 receptor–nucleotide binding site–leucine-rich repeat) domain signatures and genes encoding proteases and virulence-responsive proteins. In the plant immune system, it has been shown that NBS-LRR disease-resistance proteins can monitor the homeostasis of type III effector targets [61–63]. On the other hand, elevated CO2 intensified the repression of the expression of genes induced by low Mg in Q cluster, which is related to the Cd response (GO:0046686), cell redox homeostasis (GO: 0045454) and lipid localization. Most features of known genes (such as ATTRX5, RPT2A, GDH2, PAG1, ADL1E, NDPK1) responding to cadmium ion are thought to be cytosolic or associated with the plasma membrane.
Compared with the ACS, a larger number of genes functionally responding to programmed cell death and lipid metabolism (e.g. FADB ATFAH2, WAX2, FATTY ACID DESATURASE 2, FATTY ACID DESATURASE 7, CYTIDINEDIPHOSPHATE DIACYLGLYCEROL SYNTHASE 5 and QUIRKY) in shoots were suppressed by AL, but were not affected by EL treatment (S3 and S5 Tables).
Interestingly, elevated CO2 plus high Mg had a weak impact on the response of transcripts in shoots. Their only interaction is intensified in K cluster, which includes chloroplast envelope (AOC1, AOC2, COR15B, LOX2, AT3G22620 etc) and monooxygenase activity genes. GO analysis showed that all those six monooxygenase activity genes are also associated with the endomembrane system. Not surprisingly, other genes GO term, such as response to hormones and water deficit, are shared between transcripts under AH and EH treatments.
Mg deficiency affects the root transcriptome irrespective of CO2 treatment
The present study and previous transcriptomic analysis of [19–20] suggest that low Mg affects the root development less than the shoot development in Arabidopsis. This was clearly distinct from reports on N, P and K, which have a severe impact on the root transcriptome and eventually on root development [44, 64–67]. Phenotype data showed that AL and EL increased root growth and proliferation (Fig 2). At the transcriptional level, low Mg increased but high Mg decreased cell wall synthesis and root hair cell differentiation (e.g. ATXTH17, ATXTH 26, LEUCINE-RICH REPEAT/EXTENSIN 1, EXORDIUM, PRP3, POLYGALACTURONASE INHIBITING PROTEIN 1 and GLYCINE-RICH PROTEIN 5) (Fig 6). Compared with the controls, there are 87 genes were specifically changed in low Mg2+-treated roots, and most of these genes involved in plant-type cell wall organization and oxidoreductase activity were greatly up-regulated in roots of the plant grown in AL treatment (S6 Table). This indicates that low Mg promoted root-hair growth, probably through the characterized ROS signals and the expression of plant-type cell wall genes. The result is consistent with our previous finding . Similarly, under supply of low Mg plus elevated CO2, 56 genes were specifically changed in low Mg2+ plus CO2 treated roots. The expression of low-Mg-induced DEGs in response to stress and stimuli, as well as the activity of cell wall and oxidoreductase, was still up-regulated by high CO2 concentration, but there was a large portion of gene expression reinstated as that in control, suggesting that increased CO2 level can alleviate the low Mg-induced stimuli-responsiveness and high expression of cell wall genes (Fig 8). Moreover, the up-regulation of cation-binding genes at a low Mg2+ plus elevated CO2 implies that increased CO2 concentration would facilitate the absorption of cations by low Mg-fed roots. On the contrary, elevated CO2 did not affect the expression of genes correlated with localization and transport in low Mg-supplied roots and suppressed the activity of ion transmembrane-transporters, the cell metabolism (metabolic processes of fatty acids), and metabolisms of benzene and its derivatives as well as a series of metabolic processes of organic acids, suggesting that the combination of low Mg2+ and elevated CO2 induced a unique response in the root system of Arabidopsis thaliana, particularly having an inhibitory effect on metabolic processes in root cells.
We proposed a model to show how the combined effect of elevated CO2 and Mg levels regulates the transcriptome profile in Arabidopsis (Fig 9). This model is mainly based on our analysis of gene functions and pathways. In shoots, both low Mg and high Mg induced but elevated CO2 restrained multiple-stress response and cell death, and subsequently immune system process. In addition, CO2 and Mg stresses commonly influenced the primary and secondary metabolism, displaying an inhibitory effect of low Mg but a promoting effect of both high Mg and elevated CO2. Moreover, low Mg specially inhibited photosynthesis including plasma-membrane and chloroplast as well as some related biosynthetic process, whereas high Mg altered leaf shape. On the other hand, low Mg enhanced but high Mg decreased cell wall synthesis and root-hair cell differentiation (e.g. ATXTH17, ATXTH 26, LEUCINE-RICH REPEAT/EXTENSIN 1, EXORDIUM, PRP3, POLYGALACTURONASE INHIBITING PROTEIN 1 and GLYCINE-RICH PROTEIN 5) (Fig 9). Under both low Mg and high Mg conditions, some DEGs were elevated in those involved in immune system functions, in particular, functions related to cell death, and toxin catabolic and metabolic process and defense response. Importantly, elevated CO2 mitigates the impact of low or high Mg on cell wall and root development, which is exactly the interaction nodes of combined treatment of CO2 and Mg. However, the current model still remains incomplete because the integrated output of hormonal cross-talks in response to the combined treatment of CO2 and Mg is complex due to the synergistic, additive or antagonistic effects on signaling pathways. These findings provide new insights in the interaction between CO2 and Mg nutrition on molecular physiology of the plants, which will help to design the novel metabolic engineering strategies to crops (Brassica as relatives to the model species) with Mg deficiency/excess in crop plants under elevated CO2.
Putative regulated processes are enclosed in boxes. For illustration, some representative genes are given in italics. Solid arrows indicate links established in the enhanced regulation and dashed circle denote moderation of elevated CO2 on cell wall in the root. T bars, Inhibition. Different colors indicate different regulatory signals. Abbreviations: LMg, low Mg, HMg, High Mg.
S1 Table. List of transcriptomic data under individual vs. combinatorial treatments of elevated CO2 and Mg levels.
S2 Table. The Pearson correlation coefficients of sample pairs.
S3 Table. List of differentially regulated genes expression in venn diagram.
S4 Table. List of primers used for the reverse transcription quantitative polymerase chain reaction (qPCR) assay and transcriptomic data of reconfirmed genes.
S5 Table. Gene ontology analysis of differentially expressed genes (DEGs).
We thank anonymous reviewers for their constructive comments on the manuscript. We thank Hangzhou Guhe Information and Technology Co., Ltd for the help with RNA-seq sequencing and bioinformatics analysis. This work was financially supported by Grants of the China Postdoctoral Science Foundation (517000-X91403, http://jj.chinapostdoctor.org.cn/V1/Program1/Default.aspx), the National Natural Science Foundation of China (31550110201, http://www.nsfc.gov.cn/) and the Zhejiang Provincial priority projects in forestry sciences (KF201310 and KF201311).
Conceived and designed the experiments: YFN JQY. Performed the experiments: YFN. Analyzed the data: YFN GJA. Contributed reagents/materials/analysis tools: YFN. Wrote the paper: YFN CXT LBG JQY.
- 1. Ahuja I, de Vos RC, Bones AM, Hall RD (2010) Plant molecular stress responses face climate change. Trends Plant Sci 15:664–674. pmid:20846898
- 2. Rasmussen S, Barah P, Suarez-Rodriguez MC, Bressendorff S, Friis P, Costantino P, et al. (2013) Transcriptome responses to combinations of stresses in Arabidopsis. Plant Physiol 161:1783–1794. pmid:23447525
- 3. del Carmen Rodríguez-Hernández M, Moreno DA, Carvajal M, del Carmen Martínez-Ballesta M (2014) Genotype influences sulfur metabolism in Broccoli (Brassica oleracea L.) under elevated CO2 and NaCl stress. Plant Cell Physiol 55:2047–2059. pmid:25246493
- 4. Oerke EC, Dehne HW, Schönbeck F, Weber A (1994) Crop production and crop protection: estimated losses in major food and cash crops. Elsevier Science, Amsterdam, The Netherlands.
- 5. Mundy J, Nielsen HB, Brodersen P (2006) Crosstalk. Trends Plant Sci 11:63–64. pmid:16406754
- 6. Taub D (2010) Effects of rising atmospheric concentrations of carbon dioxide on plants. Nature Education Knowledge 1:21.
- 7. Stitt M, von Schaewen A, Willmitzer L (1991) “Sink” regulation of photosynthetic metabolism in transgenic tobacco plants expressing yeast invertase in their cell wall involves a decrease of the Calvin-cycle enzymes and an increase of glycolytic enzymes. Planta 183:40–50. pmid:24193531
- 8. Stulen I, Den Hertog J (1993) Root growth and functioning under atmospheric CO2 enrichment. Vegetation 104:99–115.
- 9. Jin CW, Du ST, Chen WW, Li GX, Zhang YS, Zheng SJ (2009) Elevated carbon dioxide improves plant iron nutrition through enhancing the iron-deficiency-induced responses under iron-limited conditions in tomato. Plant Physiol 150: 272–280. pmid:19329565
- 10. Niu YF, Jin CW, Jin GL, Zhou QY, Lin XY, Tang CX, et al. (2011) Auxin modulates the enhanced development of root hairs in Arabidopsis thaliana (L.) Heynh. under elevated CO2. Plant Cell Environ 34:1304–1317. pmid:21477123
- 11. Niu YF, Chai RS, Dong HF, Wang H, Tang CX, Zhang YS (2013) Effect of elevated CO2 on phosphorus nutrition of phosphate-deficient Arabidopsis thaliana (L.) Heynh under different nitrogen forms. J Exp Bot 64:355–367. pmid:23183255
- 12. Berntson GM, Bazzaz FA (1997). Nitrogen cycling in microcosms of yellow birch exposed to elevated CO2: simultaneous positive and negative below-ground feedbacks. Glob Change Biol 3:247–258.
- 13. BassiriRad H, Gutschick VP. Lussenhop J (2001) Root system adjustments: regulation of plant nutrient uptake and growth responses to elevated CO2. Oecologia 126:305–320.
- 14. Kogawara S, Norisada M, Tange T, Yagi H, Kojima K (2006) Elevated atmospheric CO2 concentration alters the effect of phosphate supply on growth of Japanese red pine (Pinus densiflora) seedlings. Tree Physiol 26:25–33. pmid:16203711
- 15. Maguire ME, Cowan JA. (2002) Magnesium chemistry and biochemistry. BioMetals 15: 203–210. pmid:12206387
- 16. Cakmak I, Hengeler C. Marschner H (1994a) Changes in phloem export of sucrose in leaves in response to phosphorus, potassium and magnesium deficiency in bean plants. J Exp Bot 45:1251–1257.
- 17. Cakmak I, Hengeler C, Marschner H (1994b) Partitioning of shoots and root dry matter and carbohydrates in bean plants suffering from phosphorus, potassium and magnesium deficiency. J Exp Bot 45:1245–1250.
- 18. Tewari RK, Kumar P, Sharma PN (2006) Magnesium deficiency induced oxidative stress and antioxidant responses in mulberry plants. Sci Hort-Amaterdam 108:7–14.
- 19. Hermans C, Vuylsteke M, Coppens F, Craciun A, Inzé D, Verbruggen N (2010a) Early transcriptomic changes induced by magnesium deficiency in Arabidopsis thaliana reveal the alteration of circadian clock gene expression in roots and the triggering of abscisic acid–responsive genes. New Phytol 187:119–131.
- 20. Hermans C, Vuylsteke M, Coppens F, Cristescu SM, Harren FJM, Inzé D, et al. (2010b) Systems analysis of the responses to long–term magnesium deficiency and restoration in Arabidopsis thaliana. New Phytol 187:132–144.
- 21. Visscher AM, Paul AL, Kirst M, Guy CL, Schuerger AC, Ferl RJ (2010) Growth performance and root transcriptome remodeling of Arabidopsis in response to Mars-Like levels of magnesium sulfate. PloS one 5: e12348. pmid:20808807
- 22. Niu YF, Chai RS, Liu LJ, Jin GL, Liu M, Tang CX, et al. (2014a) Magnesium availability regulates the development of root hairs in Arabidopsis thaliana (L.) Heynh. Plant Cell Environ 37:2795–2813.
- 23. Niu YF, Jin GL, Zhang YS (2014b) Root development under control of magnesium availability. Plant Signal Behav 9:1–4.
- 24. Müller AH, Hansson M (2009) The barley magnesium chelatase 150-kD subunit is not an abscisic acid receptor. Plant Physiol 150:157–166. pmid:19176716
- 25. Neuhaus C, Geilfus CM, Zörb C, Mühling KH (2013) Transcript expression of Mg-chelatase and H+-ATPase isogenes in Vicia faba leaves as influenced by root and foliar magnesium supply. Plant Soil 368:41–50.
- 26. Shabala S, Hariadi Y (2005) Effects of magnesium availability on the activity of plasma membrane ion transporters and light-induced responses from broad bean leaf mesophyll. Planta 221:56–65. pmid:15645306
- 27. Reddy AR, Rasineni GK, Raghavendra AS (2010) The impact of global elevated CO2 concentration on photosynthesis and plant productivity. Curr Sci 99:46–57.
- 28. Shaul O (2002) Magnesium transport and function in plants: the tip of the iceberg. BioMetals 15:307–321.
- 29. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL (2013) TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol 14:R36. pmid:23618408
- 30. Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, van Baren MJ, et al. (2010) Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol 28:511–515. pmid:20436464
- 31. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. (2012) Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nature Protocols 7:562–578. pmid:22383036
- 32. Gan YB, Kumimoto R, Liu C, Ratcliffe O, Yu H, Broun P (2006) GLABROUS INFLORESCENCE STEMS modulates the regulation by gibberellins of epidermal differentiation and shoot maturation in Arabidopsis. Plant Cell 18:1383–1395. pmid:16679458
- 33. Møller IS, Gilliham M, Jha D, Mayo GM, Roy SJ, Coates JC, et al. (2009) Shoot Na+ exclusion and increased salinity tolerance engineered by cell type–specific alteration of Na+ transport in Arabidopsis. Plant Cell 21:2163–2178. pmid:19584143
- 34. Morel M, Crouzet J, Gravot A, Auroy P, Leonhardt N, Vavasseur A, et al. (2009) AtHMA3, a P1B-ATPase allowing Cd/Zn/Co/Pb vacuolar storage in Arabidopsis. Plant Physiol 149:894–904. pmid:19036834
- 35. Karley AJ, White PJ (2009) Moving cationic minerals to edible tissues: potassium, magnesium, calcium. Curr Opin Plant Biol 12:291–298. pmid:19481494
- 36. Hermans C, Johnson GN, Strasser RJ, Verbruggen N (2004) Physiological characterisation of magnesium deficiency in sugar beet: acclimation to low magnesium differentially affects photosystems I and II. Planta 220:344–355. pmid:15378366
- 37. Ding Y, Luo W, Xu G (2006) Characterization of magnesium nutrition and interaction of magnesium and potassium in rice. Ann Appl Biol 149:111–123.
- 38. Sturn A (2000) Cluster analysis for large scale gene expression studies (Doctoral dissertation, Graz University of Technology).
- 39. Du Z, Zhou X, Ling Y, Zhang Z, Su Z (2010) agriGO: a GO analysis toolkit for the agricultural community. Nucleic Acids Research 38: W64–W70. pmid:20435677
- 40. Long SP, Ainsworth EA, Rogers A, Ort DR (2004) Rising atmospheric carbon dioxide: plants FACE the future. Ann Rev Plant Biol 55:591–628.
- 41. Lake JA, Wade RN (2009) Plant–pathogen interactions and elevated CO2: morphological changes in favour of pathogens. J Exp Bot 60:3123–3131. pmid:19470658
- 42. Osuna D, Usadel B, Morcuende R, Gibon Y, Bläsing OE, Höhne M, et al. (2007) Temporal responses of transcripts, enzyme activities and metabolites after adding sucrose to carbon-deprived Arabidopsis seedlings. Plant J 49:463–491. pmid:17217462
- 43. Usadel B, Blaesing OE, Gibon Y, Poree F, Hoehne M, Guenter M, et al. (2008) Multilevel genomic analysis of the response of transcripts, enzyme activities and metabolites in Arabidopsis rosettes to a progressive decrease of temperature in the non-freezing range. Plant Cell Environ 31:518–547. pmid:18088337
- 44. Hermans C, Hammond JP, White PJ, Verbruggen N (2006) How do plants respond to nutrient shortage by biomass allocation? Trends Plant Sci 11:610–617. pmid:17092760
- 45. Hermans C, Verbruggen N (2005) Physiological characterization of Mg deficiency in Arabidopsis thaliana. J Exp Bot 56:2153–2161. pmid:15983014
- 46. Wang H, Xiao WD, Niu YF, Jin CW, Chai RS, Tang CX, et al. (2013) Nitric oxide enhances development of lateral roots in tomato (Solanum lycopersicum L.) under elevated carbon dioxide. Planta 237:137–144. pmid:22990909
- 47. Nakata M, Matsumoto N, Tsugeki R, Rikirsch E, Laux T, Okada K (2012) Roles of the middle domain-specific WUSCHEL-RELATED HOMEOBOX genes in early development of leaves in Arabidopsis. Plant Cell 24:519–535. pmid:22374393
- 48. Azam A, Khan I, Mahmood A (2013) Yield, chemical composition and nutritional quality responses of carrot, radish and turnip to elevated atmospheric carbon dioxide. J Sci Food Agr 93:3237–3244.
- 49. Mäser P, Thomine S, Schroeder JI, Ward JM, Hirschi K, Sze H, et al. (2001) Phylogenetic relationships within cation transporter families of Arabidopsis. Plant Physiol 126:1646–1667. pmid:11500563
- 50. Krämer U, Talke I, Hanikenne M (2007) Transition metal transport. FEBS Lett 581:2263–2272. pmid:17462635
- 51. Gardner RC (2003) Genes for magnesium transport. Curr Opin Plant Biol 6:263–267 pmid:12753976
- 52. Li LG, Sokolov LN, Yang YH, Li DP, Ting J, Pandy G, et al. (2008) A mitochondrial magnesium transporter functions in Arabidopsis pollen development. Mol Plant 1:675–685. pmid:19825572
- 53. Gebert M, Meschenmoser K, Svidova S, Weghuler J, Schweyen R, Eifler K, et al. (2009) A root-expressed magnesium transporter of the MRS2/MGT gene family in Arabidopsis thaliana allows for growth in low-Mg2+ environments. Plant Cell 21:4018–4030. pmid:19966073
- 54. Shaul O, Hilgemann DW, de-Almeida-Engler J, Van Montagu M, Inzé D, Galili G (1999) Cloning and characterization of a novel Mg2+⁄H+ exchanger. The EMBO J 18:3973–3980. pmid:10406802
- 55. David-Assael O, Saul H, Saul V, Mizrachy-Dagri T, Berezin I, Brook E, et al. (2005) Expression of AtMHX, an Arabidopsis vacuolar metal transporter, is repressed by the 5′ untranslated region of its gene. J Exp Bot 56:1039–1047. pmid:15710632
- 56. Cakmak I, Kirkby EA (2008) Role of magnesium in carbon partitioning and alleviating photooxidative damage. Physiol Plant 133:692–704. pmid:18724409
- 57. Hermans C, Bourgis F, Faucher M, Strasser RJ, Delrot S, Verbruggen N (2005) Magnesium deficiency in sugar beets alters sugar partitioning and phloem loading in young mature leaves. Planta 220:541–549. pmid:15580527
- 58. Huber DM, Jones JB (2013) The role of magnesium in plant disease. Plant Soil 368:73–85.
- 59. Gerendás J, Führs H (2013) The significance of magnesium for crop quality. Plant Soil 368:101–128.
- 60. Carter DP, Staehelin LA (1980) Proteolysis of chloroplast thylakoid membranes. II. Evidence for the involvement of the light-harvesting chlorophyll ab-protein complex in thylakoid stacking and for effects of magnesium ions on photosystem II-light-harvesting complex aggregates in the absence of membrane stacking. Arch Biochem Biophys 200:374–386. pmid:7002041
- 61. Dangl JL, Jones JDG (2001) Plant pathogens and integrated defence responses to infection. Nature 411:826–833. pmid:11459065
- 62. Chisholm ST, Coaker G, Day B, Staskawicz BJ (2006) Host-microbe interactions: Shaping the evolution of the plant immune response. Cell 124:803–814. pmid:16497589
- 63. Jones JDG, Dangl JL (2006) The plant immune system. Nature 444:323–329. pmid:17108957
- 64. Armengaud P, Breitling R, Amtmann A (2004) The potassium-dependent transcriptome of Arabidopsis reveals a prominent role of jasmonic acid in nutrient signaling. Plant Physiol 136:2556–2576. pmid:15347784
- 65. Shin R, Schachtman DP (2004) Hydrogen peroxide mediates plant root cell response to nutrient deprivation. Proc Nat Acad Sci USA 101:8827–8832. pmid:15173595
- 66. Misson J, Raghothama KG, Jain A, Jouhet J, Block MA, Bligny R, et al. (2005) A genome-wide transcriptional analysis using Arabidopsis thaliana Affymetrix gene chips determined plant responses to phosphate deprivation. Proc Nat Acad Sci USA 102:11934–11939. pmid:16085708
- 67. Iyer-Pascuzzi A, Simpson J, Herrera-Estrella L, Benfey PN (2009) Functional genomics of root growth and development in Arabidopsis. Curr Opin Plant Biol 12:165–171. pmid:19117793